(The Systematics Association Special Volume Series) M.F. Claridge, A.H. Dawah, M.R. Wilson - Species - The Units of Biodiversity-Springer (1997)

Secies
M.R Claridge, HA Dawah

and M,R, Wilson
CHAPMAN & HALL Systerriatics

J
ASSOCIATION
Species
JOIN US ON THE INTERNET VIA WWW, GOPHER, FTP OR EMAIL

WWW: http://www.thomson.com
GOPHER: gopher.thomson.com
'*"•' AAservice
corwiroQf
nfI(f)p
<
ICTW®
FTP: ftp.thomson.com
EMAIL: findit@kiosk.thomson.com
The Systematics Association Special Volume Series
Series Editor
David M. John
Department of Botany, The Natural History Museum, Cromwell Road,

London SW7 5BD, UK.
The Systematics Association provides a forum for discussing systematic

problems and integrating new information from cytogenics, ecology and
other specific fields into taxonomic concepts and activities. It has achieved
great success since the Association was founded in 1937 by promoting
major meetings covering all areas of biology and palaeontology, support-
ing systematic research and training courses through the award of grants,
production of a membership newsletter and publication of review vol-
umes by its publisher Chapman & Hall. Its membership is open to both
amateurs and professional scientists in all branches of biology who are
entitled to purchase its volumes at a discounted price.
The first of the Systematics Association's publications, The New Systematics,
edited by its then president Sir Julian Huxley, was a classic work. Over 50
volumes have now been published in the Association's 'Special Volume'
series often in rapidly expanding areas of science where a modern syn-
thesis is required. Its modus opemndi is to encourage leading exponents to
organise a symposium with a view to publishing a multi-authored volume
in its series based upon the meeting. The Association also publishes vol-
umes that are not linked to meetings in its 'Volume' series.
Anyone wishing to know more about the Systematics Association and its
volume series are invited to contact the series editor.
Forthcoming titles in the series:

Relationships of the Major Arthropod Groups
Edited by R. A. Fortey and R. H. Thomas
Other Systematics Association publications are listed after the index for
this volume.
The Systematics Association Special Volume Series 54
^pecies
The units of biodiversity
Edited by
M. F. Claridge
School of Pure and Applied Biology,
University of Wales Cardiff, UK
H. A. Dawah
School of Pure and Applied Biology,
University of Wales Cardiff, UK
and
M. R. Wilson
Department of Zoology,
National Museums and Galleries of Wales,
Cardiff, UK
CHAPMAN & HALL

London • Weinheim • New York • Tokyo • Melbourne • Madras
Published by Chapman & Hall,
2-6 Boundary Row, London SE1 8HN, UK
Chapman & Hall, 2-6 Boundary Row, London SE1 8HN, UK

Chapman & Hall GmbH, Pappelallee 3, 69469 Weinheim, Germany
Chapman & Hall USA, 115 Fifth Avenue, New York, NY 10003, USA
Chapman & Hall Japan, ITP-Japan, Kyowa Building, 3F, 2-2-1
Hirakawacho, Chiyoda-ku, Tokyo 102, Japan
Chapman & Hall Australia, 102 Dodds Street, South Melbourne, Victoria
3205, Australia
Chapman & Hall India, R. Seshadri, 32 Second Main Road, CIT East,
Madras 600 035, India
First edition 1997

© 1997 The Systematics Association
Typeset by Saxon Graphics Ltd, Derby
Printed by TJ International, Padstow, Cornwall
ISBN 0 412 63120 2
Apart from any fair dealing for the purposes of research or private study, or
criticism or review, as permitted under the UK Copyright Designs and
Patents Act, 1988, this publication may not be reproduced, stored, or
transmitted, in any form or by any means, without the prior permission in
writing of the publishers, or in the case of reprographic reproduction only in
accordance with the terms of the licences issued by the Copyright Licensing
Agency in the UK, or in accordance with the terms of licences issued by the
appropriate Reproduction Rights Organization outside the UK. Enquiries
concerning reproduction outside the terms stated here should be sent to the
publishers at the London address printed on this page.
The publisher makes no representation, express or implied, with
regard to the accuracy of the information contained in this book and
cannot accept any legal responsibility or liability for any errors or
omissions that may be made.
A catalogue record for this book is available from the British Library
Library of Congress Catalog Number: 96-70872
^-^ Printed on acid-free text paper, manufactured in accordance with

ANS1/NISO Z39.48-1992 (Permanence of Paper).
sc\
Contents
List of contributors xi
Preface xv
1 Practical approaches to species concepts for living

organisms 1
M. F. Claridge, H. A. Dawah and M. R. Wilson
1.1 Introduction 2
1.2 Linnaeus and early species concepts 3
1.3 Biological species concepts 5
1.4 Phylogenetic species and related concepts 9
1.5 Species concepts and speciation: pattern and process 10
1.6 Species in practice - an evolutionary synthesis 12
1.7 References 13
2 Viral species 17
M. H. V. Van Regenmortel
2.1 Introduction 18
2.2 Semantics 19
2.3 Continuity versus discontinuity and the problem of
species demarcation 20
2.4 Species as polythetic classes 21
2.5 Species fuzziness 22
2.6 Species or quasispecies 23
2.7 References 23
3 Towards a practical species concept for cultivable bacteria 25

M. Goodfellow, G. P. Manfio and J. Chun
3.1 Introduction 26
3.2 Early species concepts 27
3.3 The new bacterial systematics 31
vi Contents
3.4 Minimal and nomenclatural standards for the
description of new species 48
3.5 Conclusions 49
3.6 References 50
4 Species in practice: exploring uncultured prokaryote

diversity in natural samples 61
T. Martin Embley and E. Stackebrandt
4.1 Introduction 61
4.2 Species among eukaryotes 62
4.3 Species among cultured prokaryotes 62
4.4 Practical recognition of cultured prokaryote species 63
4.5 Exploring uncultured prokaryote diversity using
molecular sequences 65
4.6 16S rRNA analyses to infer phylogenetic relationships
between prokaryotes 66
4.7 16S rDNA sequences to investigate the relationships of
uncultured prokaryotes in natural samples 68
4.8 Prokaryote 16S rDNA sequence diversity in natural
samples: the relationship between 16S rRNA sequence
similarity and prokaryote species recognized using
DNA : DNA pairing assays 69
4.9 Results of some studies using SSU rRNA sequences to
explore prokaryote diversity in nature 71
4.10 Concluding remarks 75
4.11 References 76
5 Species problems in eukaryotic algae: a modern perspective 83

D. M. John and C. A. Maggs
5.1 Introduction 84
5.2 The morphological species concept 85
5.3 The biological species concept 93
5.4 The phylogenetic species concept 96
5.5 Role of molecular data 97
5.6 Conclusions 100
5.7 References 102
6 The species concept in lichens 109
O. W. Purvis
6.1 Historical aspects 110
6.2 What are lichens? Ill
6.3 Current taxonomic concepts 112
6.4 Vegetative characters, asexual propagules and
'species pairs' 118
Contents vii
6.5 The importance of environmental factors in
influencing phenotypic expression 120
6.6 Secondary metabolites, lichen chemotaxonomy and
molecular studies 123
6.7 What is an individual lichen? Fused thalli and
mechanical hybrids 125
6.8 Conclusions 129
6.9 References 130
7 Fungal species in practice: identifying species units in fungi 135

C. M. Brasier
7.1 Introduction 136
7.2 The changing fungal species concept 138
7.3 Identifying operational species units 140
7.4 Operational species units in asexual fungi 155
7.5 Operational species units among sterile mycelia 157
7.6 Criteria and concepts for the future: the way ahead 157
7.7 References 164
8 Practical aspects of the species concept in plants 171

R. J. Gornall
8.2 The taxonomic species 172
8.3 Infraspecific variation 173
8.4 Uniparental inheritance 175
8.5 Hybridization 176
8.6 Cryptic or sibling species 177
8.7 Alternative concepts 179
8.8 Commentary 185
8.9 References 186
9 Cultivated plant diversity and taxonomy 191

/. G. Hawkes
9.2 Polyploidy 192
9.3 Species concepts in cultivated plants 193
9.4 Conclusions 196
9.5 References 198
10 Species of marine invertebrates: a comparison of the

biological and phylogenetic species concepts 199
N. Knowlton and L. A. Weigt
10.2 Corals and shrimps as case studies 201
viii Contents
10.3 Other marine invertebrates 210
10.4 Broader implications 213
10.5 Conclusions 215
10.6 References 216
11 Nematode species: concepts and identification strategies

exemplified by the Longidoridae, Steinernematidae and
Heterorhabditidae 221
D. /. Hunt
11.2 Phytoparasitic nematodes of the family Longidoridae
(with particular emphasis on the genus Xiphinema
Cobb, 1913) 225
11.3 Entomopathogenic nematodes of the families
Steinernematidae and Heterorhabditidae 232
11.4 References 241
12 Species in insect herbivores and parasitoids - sibling

species, host races and biotypes 247
M. F. Claridge, H. A. Dawah and M. R. Wilson
12.2 Status of host-associated populations - host races
and biotypes 250
12.3 Biological species, specific mate recognition and
sibling species 255
12.4 Herbivore/parasitoid food webs - grass-feeding
chalcid wasps 263
12.5 Discussion and conclusions 265
12.6 References 267
13 The species concept in blood-sucking vectors of human

diseases 273
R. Lane
13.2 Theoretical species concepts 274
13.3 Species complexes 276
13.4 Molecular data and species concepts 278
13.5 Nomenclature and species concepts 279
13.6 Practical species concepts 280
13.7 Identification - a reverse look at species 284
13.9 References 285
Contents ix
14 Recognition of parthenogenetic insect species 291
R. G. Foottit
14.2 Nature and extent of parthenogenetic reproduction 293
14.3 Past treatment of asexual and parthenogenetic species 294
14.4 Evolutionary considerations 297
14.5 Practical considerations 301
14.7 References 304
15 The species in terrestrial non-insect invertebrates

(earthworms, arachnids, myriapods, woodlice and snails) 309
A. Minelli and D. Foddai
15.2 Diagnostic characters versus specific mate recognition
systems (SMRS) 311
15.3 Pitfalls of morphological evidence: sexual
dimorphism, fuzzy developmental schedules 313
15.4 Morphological and molecular evidence:
chromosomes, allozymes, isozymes and mtDNA 314
15.5 Hybrids 316
15.6 Dispersal, distribution and species 316
15.7 The temporal dimension 318
15.8 Uniparentals 318
15.9 Epilogue 320
15.10 References 320
16 Species concepts in systematics and conservation biology

- an ornithological viewpoint 325
/. Cracraft
16.2 Species in theory and practice 326
16.3 The phylogenetic species concept: some myths and
misrepresentations 329
16.4 Two species concepts: a comparison 332
16.5 Taxonomic units and conservation biology 332
16.6 References 337
17 The species in mammals 341

G. B. Corbet
17.2 Historical review 342
x Contents
17.3 Sympatric cryptospecies 344
17.4 Parapatric series: subspecies or species? 346
17.5 Allopatry 348
17.6 Discussion 349
17.7 Conclusion 353
17.8 References 354
18 The ideal species concept - and why we can't get it 357

D. L. Hull
18.2 Present-day species concepts 360
18.3 Universality and monism 365
18.4 Applicability and theoretical significance 371
18.5 Summary and conclusion 374
18.6 References 377
19 A hierarchy of species concepts: the denouement in the

saga of the species problem 381
R. L. Mayden
19.2 Methodology 383
19.3 Important qualities to consider 384
19.4 Concepts, definitions, groups, categories and names:
the unfortunate conflation of terminology 384
19.5 The twin meanings of species 387
19.6 Species concepts versus empirical data 388
19.7 The species concepts 389
19.8 Discussion 412
19.9 Hierarchy of concepts: species in theory and
practice 418
19.10 References 422
Index 425
Contributors
C. M. Brasier
Forest Research Station, Alice Holt Lodge, Farnham, Surrey GU10 4LH,
UK.
J. Chun
Department of Microbiology, The Medical School, Framlington Place,
Newcastle upon Tyne NE2 4HH, UK.
M. F. Claridge
School of Pure and Applied Biology, University of Wales Cardiff, P.O. Box
915, Cardiff CF1 3TL, Wales, UK.
G. B. Corbet
Little Dumbarnie, Upper Largo, Leven, Fife KY8 6JQ, Scotland, UK
J. Cracraft
Department of Ornithology, American Museum of Natural History,
Central Park West at 79th Street, New York, New York 10024, USA.
H. A. Dawah
School of Pure and Applied Biology, University of Wales Cardiff, P.O. Box
915, Cardiff CF1 3TL, Wales, UK.
T. Martin Embley
Department of Zoology, The Natural History Museum, Cromwell Road,
London SW7 5BD, UK.
D. Foddai
Universita di Padova, Dipartimento di Biologia, Via Trieste 75, 135121
Padova, Italy.
R. G. Foottit
Eastern Cereal and Oilseed Research Centre, Research Branch,
Agriculture and Agri-Food Canada, K.W. Neatby Bldg., Central
Experimental Farm, Ottawa, Ontario, KIA OC6, Canada.
xii Contributors
M. Goodfellow
R. J. Gornall
School of Biological Sciences, Department of Botany, University of
Leicester, University Road, Leicester LEI 7RH, UK.
J. G. Hawkes
School of Continuing Studies, The University of Birmingham, Edgbaston,
Birmingham B15 2TT, UK.
D. L. Hull
Department of Philosophy, Northwestern University, Evanston, IL 60208,
USA.
D. J. Hunt
International Institute of Parasitology, 395A, Hatfield Road, St Albans AL4
OXU, UK.
D. M. John
London SW7 5BD, UK.
N. Knowlton
Smithsonian Tropical Research Institute, Apartado 2072, Balboa, Republic
of Panama.
R. Lane
Department of Entomology, The Natural History Museum, Cromwell
Road, London SW7 5BD, UK.
C. A. Maggs
School of Biology and Biochemistry, The Queen's University of Belfast,
Belfast BT9 7BL, Northern Ireland, UK.
G. Paulo Manfio
R. L. Mayden
Department of Biological Sciences, P.O. Box 0344, University of Alabama,
Tuscaloosa, AL 35487, USA.
A. Minelli
Universita di Padova, Dipartimento di Biologia, Via Trieste 75, 135121
Padova, Italy.
Contributors xiii
O. W. Purvis
London SW75BD, UK.
E. Stackebrandt
DSM-German Collection of Micro-organisms and Cell Cultures Gmbh,
Mascheroder Weg 1 b, 38124 Braunschweig, Germany.
M. H. V. Van Regenmortel
Institut de Biologic Moleculaire et Cellulaire, CNRS, 15 rue Rene
Descartes, 67084 Strasbourg, France.
L. A. Weigt
Smithsonian Tropical Research Institute, Apartado 2072, Balboa, Republic
of Panama; now at The Field Museum, Roosevelt Road at Lake Shore
Drive, Chicago, IL 60605, USA.
M. R. Wilson
Department of Zoology, National Museums and Galleries of Wales,
Cathays Park, Cardiff CF1 3NP, Wales, UK.
Preface
The idea of bringing together a wide diversity of specialists on different

groups of living organisms to discuss the practical nature of species had
been a longstanding ambition of one of us (M.F.C.). A recommendation
'that an interdisciplinary clarification of species and population biology
concepts be accorded high priority, and that a special meeting devoted to
this topic be convened at an early date' was unanimously agreed at the
2nd Workshop on the Identification and Characterization of Pest
Organisms, held at the International Mycological Institute in 1993 [D.L.
Hawksworth (ed.) (1994), The Identification and Characterization of Pest
Organisms, CAB International, p.475]. This recommendation provided the
catalyst and final stimulus for us to organize a Systematics Association
Symposium on 'The Units of Biodiversity: Species in Practice'. This was
eventually held at University Hall, a conference centre of the University of
Wales Cardiff, from 19-21 April, 1995. A total of 66 participants heard 20
different papers on a diversity of topics concerning species concepts and
their application in a very wide selection of living organisms. Most impor-
tantly there was the opportunity for extensive informal discussion. The
present volume represents the considered and revised proceedings of that
meeting. Unfortunately, two contributors were unable to provide us with
manuscripts and so are not included in this book. We are particularly
grateful to Richard Mayden who kindly offered to present his overview
on^ species concepts for publication, though he did not present it formally
at the meeting.
As organizers of the Cardiff meeting we are deeply indebted to the
enormous number of colleagues, students and friends who gave freely
of their time to make the meeting a success. In particular we thank
Gareth Holmes and Stella Shackel who played vital roles in welcoming
and seeing to the needs of participants, David and Diana Edwards
who designed and produced the programme cover, John Morgan who
ensured the functioning of the audiovisual systems, Dr M. Al-Yaseen
for assistance with converting word processing systems, and
Rosemary Jones who provided efficient secretarial assistance. We are
xvi Preface
also indebted to the chairmen of the formal sessions for keeping con-
trol, but not limiting too much the very lively discussions - David
Hawksworth (CAB International Mycological Institute), Joel Cracraft
(American Museum of Natural History) and Bob Foottit (Agriculture
and Agri-Food Canada). Of critical significance to the success of the
meeting and of the subsequent book were the generous financial con-
tributions from the Wellcome Trust, the UK Federation of Culture
Collections, and the Systematics Association. We are deeply indebted
to those organizations for their generosity and support.
We are very grateful to all of the contributors to our book for their
forebearance and patience. We hope that the resulting volume justifies
their efforts. The quality of the book has certainly been enhanced by
the work of our independent reviewers, to whom we offer our sincere
thanks. They were: Dr Kathryn Benson-Evans, Prof. Arthur J. Cain,
Prof. John C. Fry, Prof. David L. Hawksworth, Prof. Peter Morgan, Dr
Graham Oliver, Prof. Sir Ghillean T. Prance, Dr Jeremy A. Roberts and
Prof. Harford H. Williams.
M.F.Claridge,
H.A. Dawah,
M.R.Wilson
1
Practical approaches to species
concepts for living organisms
M. E Claridge, H. A. Dawah and M. R. Wilson
Contacting address: School of Pure and Applied Biology, University of Wales Cardiff, P.O. Box 915,
Cardiff CF1 3TL, Wales, UK.
ABSTRACT
From a practical viewpoint species are generally the units of biodiver-
sity. Traditionally since before Linnaeus species have been defined in
terms of clear morphological differentiation - the morphospecies. In""
practice most species are still described on a basis of dead preserved j
material and are therefore morphospecies.
The increasing recognition by naturalists, geneticists and evolu-
tionists over the past 200 years that species occur as reproductively
isolated natural entities in the field led to the various biological species
concepts. Reproductively isolated species are separate evolutionary ^
entities characterized by unique specific mate recognition systems. An
important consequence of the biological species is the recognition of
reproductively isolated sibling species that show no clear morpholog- j
ical differentiation but which are reproductively isolated^ In practice
biological species are diagnosed by markers that may be morphologi-
cal, cytological, behavioural, molecular, etc., but which indicate the
presence of high levels of reproductive isolation.
The biological species can only be applied to biparental sexually
reproducing organisms, or at least organisms that regularly exchange
genetic material. Thus, only some form of morphospecies is available
for asexual and obligately parthenogenetic forms (agamospecies). Also
application of biological species to populations isolated in space -
allopatry - is difficult and usually subjective.
These difficulties and the desire to apply cladistic techniques at the
species level have led to widespread rejection of the biospecies by sys-
tematists in favour of a broadly phylogenetic species. Here, species are
essentially equated with diagnosably distinct clades. Advantages are
that allopatric and asexual populations can be treated in the same way
Species: The Units of Biodiversity. Edited by M.F. Claridge, H.A. Dawah and M.R. Wilson.
Published in 1997 by Chapman & Hall. ISBN 0 412 63120 2
Practical approaches to species concepts for living organisms
as sympatric sexually reproducing ones. Disadvantages include the
difficulties of deciding objectively on what is a diagnosably distinct
clade and the possibility of ignoring sibling species. There is clearly
common ground between these two general concepts for describing
biological diversity and together they form a unitary taxonomic or
evolutionary species.
1.1 INTRODUCTION
'There is probably no other concept in biology that has remained so
consistently controversial as the species concept'. (Mayr, 1982)
'What are species? Perhaps no other issue in comparative or evolu-
tionary biology has provoked quite so much disparate opinion as
this simple question'. (Eldredge, 1995)
These views of two of the most influential of recent evolutionary biolo-
gists are borne out by the plethora of publications on species concepts in
recent years. After a lapse in the 1950s and 1960s when a consensus
seemed to have been achieved, the basic philosophy and biology of
species concepts has once again been opened up widely for discussion.
This renewed interest is exemplified by the publication of review volumes
(Otte and Endler, 1989; Lambert and Spencer, 1995; Wheeler and Meier,
1977) and reprinted collections of classic papers (Ereshefsky, 1992). Many
authors now apparently feel the need to come up with yet other and
apparently new personal species concepts. Mayden (1997: Chapter 19) has
identified 22 concepts to date that he recognizes as distinct, though some
of us may regard many of them as essentially synonymous. These con-
cepts include a variety of approaches, some purely theoretical and some
entirely empirical. Hull (1997: Chapter 18) has attempted to bring some of
these approaches together.
The prolonged wrangle among scientists and philosophers over the
nature of species has recently taken on added and wider significance.
The belated recognition of the importance of biological diversity to the
survival of mankind and the sustainable use of our natural resources
makes it a matter of very general and urgent concern. Species are nor-
mally the units of biodiversity and conservation (Wilson, 1992) so it is
important that we should know what we mean by them. One major con-
cern has been with estimating the total number of species of living
organisms that currently inhabit the earth (May, 1990). In addition,
many authors have attempted to determine the relative contributions of
different groups of living organisms to the totality of living biodiversity
in which usually some sort of morphological species concept is used
(Figure 1.1). Unless we have some agreed criteria for species such dis-
cussions are of only limited value. Above all we need to know whether
species units are comparable between different major groups of organ-
Linnaeus and early species concepts 3
isms. This volume is an attempt to find common ground in the practical
use of species in documenting biodiversity by bringing together special-
ists on as wide a range of organisms as possible.
1.2 LINNAEUS AND EARLY SPECIES CONCEPTS

The term species derives from classical Greek logic. As Cain (1958) has
emphasized, it would have been natural for all scholars of the 17th and
18th centuries to adopt the Aristotelian system of logic with its precise set
of terms, including Definition, Genus, Differentia, and Species, in attempt-
ing to classify living organisms. In this system the Genus is that part of the
Definition which refers to the general kind while Species refers to the par-
ticular, as qualified by the Differentia.
Linnaeus was not only the founder of the modern binomial system of
nomenclature but he was also one of the few 18th century systematists
who wrote down precisely what he was doing and what he thought he
should do (Cain, 1958). To Linnaeus species were the lowest particular
kinds of organisms in his classifications, though varieties were also some-
times noted. From a careful analysis of Linnaeus's many writings
Ramsbottom (1938) showed that his notion of species was characterized
by three different attributes. To Linnaeus, species were:
1. Distinct and monotypic.
2. Immutable and created as such.
3. Breeding true.
Later, however, he developed a complex theory of speciation by
hybridization (see Cain, 1993).
Protoza Fungi Bacteria Viruses Insects

(1.6%) (8.0%) (3.2%) (4.0%) (64.4%)
Algae
(1.6%)
Plants
(2.4%)
Vertebrates
(0.4%)
Other
invertebrates
(6.7%)
Other arthropods
(7.7%)
Figure 1.1 Estimates of proportions of species of major groups of organisms con-

tributing to the total of living biological diversity. (Adapted from data in
Hammond, 1992.)
4 Practical approaches to species concepts for living organisms
These features are interesting because in his practical work Linnaeus
was more and more confronted with large numbers of specimens from all
over the world which he had to describe. Thus, usually all that was avail-
able to him were characteristics of dead museum specimens so that in
practice he relied almost exclusively on the obvious morphological fea-
tures of his material. However, even in the mid-18th century the notion
that these characters were markers representing some sort of breeding
units was obviously present in however vague a form.
Until the times of Linnaeus most early taxonomists were familiar with
the organisms with which they worked as living entitities in the field.
During the later 18th and the 19th centuries the natural history and
museum traditions were to become more and more separate as increas-
ingly large amounts of material needing description arrived in museums
from all parts of the known world. Inevitably taxonomists were forced
back more and more to describing dead specimens, usually with little
knowledge of the habits and habitats of their organisms. Thus, the tra-
dition was reinforced that species, and indeed higher taxa, must be
based on morphological characters recognizable in preserved specimens.
This is the morphological species concept or morphospecies which clear-
ly arose as an empirical approach to a practical problem. It is not truly a
concept but a technique of description. Nevertheless, most species in the
most species-rich groups of living organisms are still today effectively
morphospecies, often known from little more than preserved specimens.
i The amount of difference required to allow the recognition of different
species is wholly determined by the subjective judgement of the indi-
vidual taxonomist. This is epitomized by the well-known statement of
Regan (1926) that 'A species is a community, or a number of related com-
munities, whose distinctive morphological characters are, in the opinion
of a competent systematist, sufficiently definite to entitle it, or them to a
specific name'.
Naturalists immediately following Linnaeus were well aware that
species had some biological reality in the field irrespective of degrees of
morphological distinctiveness. For example, Gilbert White (1789) recog-
nized the morphologically very similar species of breeding song birds in
Britain, the Willow Warbler (Phylloscopus trochilus) and Chiffchaff (P.
collybita), as different species on a basis of their songs, now recognized as
elements of their specific mate recognition systems. Unfortunately the two
traditions diverged during the late 18th and 19th centuries with natural-
ists continuing to emphasize breeding criteria and species as reproductive
communities, but with systematists tending to emphasize morphological
differences.
Darwin and Wallace both represented the natural history tradition from
which emerged the theory of natural selection and the overwhelming
evidence for descent with modification. Modern species then became the
Biological species concepts 5
end terms of lines of descent. The controversies surrounding evolution itself
meant that the nature of species was not widely regarded as an important
problem in the mid~1800s. Indeed, Darwin himself seemed to regard species
as rather arbitrary stages in the process of evolutionary divergence.
In the early 20th century Poulton (1908) made what was probably the
most important advance towards what has since become known as the
biological species, originally in his 1904 Presidential Address to the
Entomological Society of London. In this he emphasized the importance
of interbreeding in nature as the species criterion. He also very clearly dif-
ferentiated this from simple hybridization. This was the modern biologi-
cal species in all but name.
1.3 BIOLOGICAL SPECIES CONCEPTS

The most significant development in the establishment of a broadly bio-
logical species concept was the unification of genetics, systematics and
evolutionary biology in the 1930s and 1940s, as exemplified by the pub-
lication of Genetics and the Origin of Species by Theodosius Dobzhansky
_[1937) and Systematics and the Origin of Species by Ernst Mayr (1942). Of
the various definitions provided during this period probably the most
useful is - 'species are groups of actually or potentially interbreeding
natural populations which are reproductively isolated from other such
groups' (Mayr, 1942: 120). Reproductive isolation is seen to be main-
tained by what Dobzhansky originally termed species isolating mecha-
nisms which are any properties of species populations that reduce the
likelihood of interbreeding with other species and thus the breakdown
of co-adapted genetic systems. Of course it was recognized early on that
reproductive isolation may not always be complete, but it must always
be adequate to maintain the essential integrities of the interacting
species populations. (Thus, species were seen as beginning at the stage
when different lineages become separate evolutionary entities.
Speciation, the origin of new biological species, was, according to
Dobzhansky, the origin of reproductive isolation and thus of distinct iso-
lating mechanisms. The biological species has been developed and
refined over the past 50 years or so, particularly by Cain (1954) and Mayr
(1963, 1982), and has been accepted and widely used, mostly by field
biologists, geneticists and evolutionists.
A major set of criticisms of the biological species has been developed in
a series of papers by Hugh Paterson (summarized in 1985,1993). In these
he developed what he regards as a new concept which he termed the
recognition concept in contrast to the biological species sensu Mayr which
he prefers to call the isolation concept.
Paterson's main concern was with the conception of species being
reproductively isolated by isolating mechanisms. The clear implication of
this is that these mechanisms have been evolved under natural selection
in order to achieve and maintain such isolation. They would thus obvi-
ously be primary adaptations. In this, Paterson clearly identified an area
of imprecise thinking in the original formulation of the biological species
concept. While Dobzhansky was a major advocate of the evolution of iso-
lating mechanisms by reinforcement, Mayr has always regarded them as
the result of incidental divergence between evolving nascent species
(Mayr, 1988). To avoid this controversy Avise (1994) has suggested the
more neutral term reproductive isolating barriers.
Paterson views species as groups of organisms with common fertiliza-
tion systems - 'We can, therefore, regard as a species that most inclusive
population of individual biparental organisms which share a common
fertilization system'. He further recognizes an important adaptive subset
of the fertilization system, the specific mate recognition system (SMRS)
which is 'involved in signalling between mating partners or their cells'
(Paterson, 1985). In most groups of animals the SMRS is well exemplified
by exchanges of signals during courtship sequences, as described by
ethologists. A classic example is the courtship sequence of the three-
spined stickleback (Gasterosteus aculeatus), a small freshwater fish, first
described and analysed by Tinbergen (1951). In such courtship sequences
successive signals release successive responses via tuned receptors in the
opposite sex, usually initiated first by the male. Unless the appropriate
responses are made at each stage and the signals are recognized as appro-
priate, then the exchange will be terminated and fertilization will not be
achieved (Figure 1.2)AThus, to Paterson the defining properties of partic-
ular species are their unique specific mate recognition systems and the
process of speciation is the evolution of new and different SMRSs\ He has
on several occasions strongly differentiated what he terms his recogni-
tion concept from the earlier biological species of Mayr, Dobzhansky and
Cain, which he prefers to term the isolation concept (Paterson, 1985).
Vrba (1995) has recently cogently argued the same case, but others
(Claridges 1988, 1995a; Coyne et al, 1988; Mayr, 1988) have doubted the
validity of the extreme contrast between the two concepts. Indeed, Mayr"
had earlier referred to what Paterson and his followers now term specif-
ic mate recognition systems as species recognition, 'the exchange of
appropriate stimuli between male and female to ensure mating of con-
specific individuals' (Mayr, 1963: 95).
The differentiation between the isolation and recognition concepts has
generated much controversy and is discussed at length in a recent multi-
author volume (Lambert and Spencer, 1995) and by Mayr (1988) and
Coyne et al. (1988). We take the view that, in practice, the two concepts are
similar and will usually allow the recognition of the same entities as
species (Claridge, 1988,1995a). Thus, a broadened or composite biological
species concept recognizes that different species are characterized by
Biological species concepts
' X X X
///, Female Male
Female 'response Male
Male response
-> ' tuned '-t* 'released1' •** tuned •>> •etc. Fertilization
signal 'receiver,
released
, -f signal / receiver - signal
f f
Female Male
terminates terminates
sequence sequence
Figure 1.2 Diagrammatic representation of the exchange of signals and respon-

ses that constitute a specific mate recognition system (SMRS) (after Paterson,
1985).
distinct SMRSs which result in the reproductive isolation observed

between different sympatric species in the field. This species concept is
widely used in various groups of biparentally reproducing organisms.
In practice, species are rarely recognized by direct studies of the SMRS
though that must remain the ultimate arbiter for biological species (Claridge,
1988,1994; Claridge et al, 1997: Chapter 12). Usually, biological species are
recognized by distinctive markers not necessarily directly associated with
the SMRS. These most often in the past have been diagnostic morpholo-
gical differences between species, but in more recent years have increasing-
ly included characteristics of cytology, behaviour, biochemistry, etc. Most
excitingly, molecular markers involving characteristics of DNA, the heredi-
tary material itself, are more and more being used (Avise, 1994). However,"
all of the diversity of characters now available to taxonomists are used as
indicators of levels of reproductive isolation under the biological species
concept. Molecular techniques can often be used directly to measure levels
of gene flow between populations, and therefore of reproductive isolation.
One very important consequence of the broad concept of biological
species advocated here is that reproductive isolation may occur without
associated morphological or other obvious differentiation. Thus, real bio-
logical species may exist that show little or no obvious differentiation to
the human observer - so-called sibling or cryptic species. Such species are
well known and abundant in many groups of organisms (Claridge, 1988;
Claridge et al., 1997: Chapter 12; Knowlton, 1997: Chapter 10; Lane, 1997:
Chapter 13). These are just as real biological entities as are morpholo-
gically recognizable species, but they pose practical and theoretical diffi-
culties if non-biological species concepts are used.
A large body of opinion among systematists has always been unhappy
with the biological species (Sokal and Crovello, 1970) and has preferred
either an overtly morphospecies approach or some sort of phenetic sys-
tem. In particular, botanists have been widely critical because of the fre-
quent occurrence in the field of interspecific hybrids between plant
species (Gornall, 1997: Chapter 8). However Mayr (1992) showed that in at
least one North American local flora the biological species was easily
applicable to all but between 6% and 7% of the plants present.
There is certainly no dispute even among the most ardent advocates of
biological species that there are two important areas where application is
either very difficult or impossible:
1. Asexual or parthenogenetic forms. The biological species can only
apply to biparental sexually reproducing organisms in which a SMRS is
present leading to reproductive isolation. Neither asexual nor partheno-
genetic organisms can have SMRSs, since functional mating and fusion of
gametes does not take place. Thus, the biological species concept cannot
be used. Such organisms exist as clones which may differ in features of
morphology, biochemistry, behaviour, etc. (Foottit, 1997: Chapter 14).
Distinctive and diagnosable clones may be recognized as species but they
can not be biological species. Cain (1954) coined the term agamospecies for
such entities and De Bach (1969), in a useful contribution, termed them
uniparental species. Clearly the very many groups of microorganisms
which have no exchange of gametes or other genetic material in repro-
duction, can only be agamospecies.
2. Allopatric forms. A further difficulty with biological species con-
cepts is that reproductive isolation in the field can only be observed
between sympatric populations; in these alone opportunities exist for
testing the effectiveness of presumed SMRSs. The problem of determin-
ing the status of allopatric populations is a longstanding one. Varying
degrees of observed differentiation from virtually nothing to large differ-
ences at least comparable with those between sympatric species, may be
observed. The criterion of gene flow and degree of reproductive isolation
in the field cannot be tested. The polytypic nature of biological species
has long been recognized and a series of categories from superspecies to
subspecies has been advocated in attempts to document such essentially
continuous variation (Mayr, 1942; Cain, 1954). However, the allocation of
allopatric forms within this series has always been largely subjective. This
problem is undoubtedly a weakness in the applicability of biological
species concepts.
These difficulties, together with a desire to eliminate the priority of one
set of characteristics of organisms - the SMRS and reproductive isolation
- over others have led many modern systematists to reject completely bio-
logical species and to favour some variant of what may be termed a gen-
eral phylogenetic species concept.
Phylogenetic species and related concepts 9
1.4 PHYLOGENETIC SPECIES AND RELATED CONCEPTS
With the development and widespread acceptance of cladistic methods by

English-speaking systematists over the past 30 years or so, dissatisfaction
with the biological species has developed. Curiously Hennig (1966), the
founder of cladistic methodologies, regarded species as reproductive com-
munities and his species concept was close to the biological species of
Ernst Mayr (Nixon and Wheeler, 1990). Of course Hennig was also con-
cerned with extending species back in time as diagnosable clades for
which the biological species is not adequate. In this he was continuing
what Simpson (1951) had begun by developing a broader evolutionary
concept which has since been taken up by others, including Cain (1954),
Wiley (1978) and Mayden (1997: Chapter 19).
Opinion has been widely divided among cladists on the nature and sta-
tus of species. Some even take the view that species have no special signifi-
cance and are no different from higher taxa ('A species is only a taxon';
Nelson, 1989), but this certainly does not seem to be the general view.
Hennig (1966) saw species as the very important level in the systematic hier-
archy above which cladistic methods were applicable and below which they
were not. At and above the species level phylogenetic relationships were
important. Below the species level interbreeding relationships dominate
and these he differentiated from phylogenetic, as tokogenetic relationships.
As Nixon and Wheeler (1990) state, 'Species are uniquely different from
higher level taxa, in that species do not have resolvable internal phyloge-
netic structure among the individual organisms included'.
Various authors have given definitions of phylogenetic species.
Perhaps the most widely cited and influential is that of Cracraft (1983) in
which species were defined as the 'smallest diagnosable cluster of indi-
vidual organisms within which there is a parental pattern of ancestry and
descent'. Contrary to the views of some critics (e.g. Avise, 1994; Mallet,
1995), it is clear that this concept can and does apply to populations
(Cracraft, 1997: Chapter 16), in which it therefore resembles the biological
species. It is, though, applicable also to non-sexually reproducing organ-
isms. This is emphasized in a more recent refinement of the concept ifr"l
which Nixon and Wheeler (1990) define species as the 'smallest aggrega- \
tion of populations (sexual) or lineages (asexual) diagnosable by a unique j
combination of character states in comparable individuals'. Thus, the \
essence of the phylogentic species concept is concerned with the recogni- —'
tion of diagnosably distinct clades. The major question that has to be
asked in practice is - What precisely is diagnostic? How different do two
populations or lineages have to be to be regarded as diagnosably distinct?
What is distinct to one worker may not be so to another.
Any of the markers discussed above as indicative of biological species
may be used to characterize phylogenetic species, though in most groups of
larger organisms the characters used are generally mainly morphological
ones. In microorganisms they are more usually biochemical and molecular.
As presently applied (Cracraft, 1997: Chapter 17) the differences in
practice between a phylogenetic concept and a broadly biological species
do not seem to us to be very great. The advantages of the former are clear:
• Phylogenetic species can be applied to non-sexual and parthenogenetic

forms and will depend on the diagnosability of the different clones or
lineages. They will effectively be the same as agamospecies (Cain, 1954).
• Phylogenetic species may also be applied to allopatric forms.
Diagnosably distinct allopatric populations will therefore be regarded as

separate species, presumably however slight the difference. The result is
almost always that more allopatric populations are recognized as different
species than has traditionally been the case for the polytypic biological
species. For example Cracraft (1992), in reviewing the Birds-of-Paradise
(Aves, Paradisaeidae), recognized more than twice as many phylogenetic
species (90) as had previous applications of the biological species to the
same data (40+). However, this decision still remains a matter of personal
subjective judgement. In our view there is no fundamental difference of
principle between the two concepts on this matter, but it would seem best
to us, when in doubt, to err on the side of recognizing more rather than
less species.
A major disadvantage of phylogenetic species concepts that is not often
recognized is the improbability that they will reveal the existence in many
groups of organisms of complexes of sibling species. The emphasis on the
criterion of reproductive isolation and specific mate recognition in the bio-
logical species concept means that sibling species will be revealed there.
The philosophy of the phylogenetic species gives no reason or incentive
to search for further divisions of existing diagnosably distinct forms. In
view of the significance of sibling species in many groups this a major
weakness. The phylogenetic species should have little difficulty in accom-
modating sibling species once discovered, but only application of biologi-
cal species concepts will positively identify them.
1.5 SPECIES CONCEPTS AND SPECIATION: PATTERN AND

PROCESS
Most authors will presumably agree that in recognizing species we are
attempting to provide a framework for describing and understanding the
diversity of living organisms and their evolutionary relationships.
However, there have been longstanding controversies about the interac-
tion between particular species concepts and the theories of speciation
upheld by different workers. In principle, it is surely necessary to establish
Species concepts and speciation: patterna nd process 11
a system for describing the diversity of living organisms that is indepen-
dent of the various modes by which that diversity may have evolved.
Many evolutionary biologists do not agree with this. For example, Guy
Bush has for long been a proponent of theories of sympatric speciation, in
which, contrary to widely supported theories of allopatric speciation
(Mayr, 1942, 1963; Cain, 1954), no period of spatial isolation is required
before initial genetic divergence can take place (Bush, 1975, 1993, 1994).
These interesting ideas are controversial, but are clearly supported by some
studies, particularly on specialist feeders and parasites. Our understanding
of the pattern of diversity in the field should not vary according to which
theory of speciation we support. The recognition of biological species and
the inevitable difficult cases, where some level of gene flow occurs between
species, should be completely compatible with either allopatric or sym-
patric models of speciation (Claridge, 1995b, but see also Bush, 1995).
A similar argument in which species concepts and models of speciation
have become completely intertwined is that of the recognition concept
(here regarded as part of a wider biological species). Paterson (1985,1993)
is a passionate advocate of what we might term pure allopatric speciation
which, despite his differences of opinion with Mayr, follows closely the
original proposals of Mayr (1942,1963). In these it is envisaged that speci-
ation is completed in allopatry, often after the isolation of a relatively
small subgroup of populations from the original ancestral group. Species
differences thus evolve completely in allopatry. Paterson's rejection of the
concept of species isolating mechanism follows since, if such characteris-
tics evolve entirely in isolation, then obviously they cannot be adaptations
to ensure reproductive isolation.
Contrary to theories of speciation in complete allopatry, Wallace (1889),
Dobzhansky (1937) and others, have developed the theory of reinforcement
of species isolating mechanisms after diverging incipient species popula-
tions have once again become sympatric. Any differences so evolved could
truly be isolating mechanisms as a result of natural selection favouring
homogametic matings. The weight of evidence currently is generally
thought to be against the importance of reinforcement (Butlin, 1989), but
recent theoretical studies once again suggest that it may be sometimes more
significant than has been thought (Liou and Price, 1994; Butlin, 1995).
These important and interesting controversies should not affect the
species units that we recognize in nature and that form the basis for spe-
ciation studies. This matter has been well discussed at length by Chandler
and Gromko (1989). We believe that the biological species concept must be
formulated to allow any of the theories of speciation to be fairly tested
without prior assumptions concerning speciation. Thus, in principle, we
agree with many cladists on the particular point that it is necessary to
describe patterns in nature, so far as possible, independent of theories
concerning the origins of those patterns (Nixon and Wheeler, 1990).
However, we would not go so far as Wheeler and Nixon (1990) who state
that 'the responsibility for species concepts lies solely with systematists'. If
we take an evolutionary view of species then we cannot separate species
also from genetics and evolutionary biology.
"~ Though we emphasize the need to separate so far as possible the recog-
nition of pattern, in the form of species, from process in the form of modes
of speciation, we cannot - and indeed should not - divorce species con-
cepts from evolution. If we accept the generality of evolution and species
_as the result of evolutionary divergence then the species itself must be an
evolutionary concept.
1.6 SPECIES IN PRACTICE - AN EVOLUTIONARY SYNTHESIS

"'We are concerned here entirely with species as used for extant living
organisms and not with the difficult problems of using species in a time
dimension when fossil forms are studied. However, the different species
concepts we have briefly reviewed all attempt to describe the extant ter-
minal branches of different evolutionary lineages. Simpson (1951) was one
of the first to develop an explicitly e^volutionary species concept, by integ-
rating species as used for living organisms with species as segments of
evolutionary lineages so as to include fossils. Cain (1954) took up this integ-
ration. Thus in Cain's system the taxonomic species includes
palaeospecies which are named lineages over time, which might be akin
to the internodal species of Kornet (1993) and other similar recent con-
cepts (see Mayden, 1997: Chapter 19). The biological species are modern
sexually reproducing forms evolving separately from their closest relatives
and from which they are reproductively isolated. Other separate lineages
of organisms lacking biparental sexual reproduction are clones which may
be recognized as separate agamospecies on a basis of particular character-
istics. Thus, these forms are all part of the general taxonomic species con-
cept and what Simpson (1951) first developed as an evolutionary species
concept. This useful umbrella term was taken up more recently by Wiley
(1978) and Mayden (1997: Chapter 19), who resuscitates it in this volume
as a general species concept that we can all use. The development of phy-
logenetic species concepts in recent years (Cracraft, 1997: Chapter 16) has
established a framework which many taxonomists find useful. However,
we do not regard it as so different to the biological species and agamo-
species in the above system (after Cain, 1954).
The detailed considerations by specialists on a wide diversity of living
organisms in this volume give reason for some hope that species may be
used consistently as the basic units of biodiversity. We agree with
Cracraft that this is even more important now that high priority is right-
ly being given to the conservation of biodiversity. It is vital that some
general agreement on species for all groups of organisms should be
References 13
achieved. We believe that there are no insuperable difficulties to using
biological species, agamospecies and phylogenetic species in different
groups and particular situations as appropriate. They all represent part of
the taxonomic or evolutionary species. The particular problem of the sta-
tus of allopatric populations is common to all these approaches and
should have similar resolutions.
Acknowledgements
We thank all colleagues who have devoted many hours to discussing
ideas on species concepts with us. In particular we are grateful to all of the
contributors and participants at the 1995 meeting in Cardiff who clarified
much of our thinking.
We are particularly indebted to Arthur Cain for his critical comments on
a draft of this chapter, John Morgan for helping in the preparation of the
figures and Rosemary Jones for preparing the manuscript.
1.7 REFERENCES
Avise, J.C. (1994) Molecular Markers, Natural History and Evolution. Chapman & Hall,
New York.
Bush, G.L. (1975) Modes of animal speciation. Annual Review of Ecology and
Systematics, 6, 339-64.
Bush, G.L. (1993) A reaffirmation of Santa Rosalia, or why are there so many kinds
of small animals, in Evolutionary Patterns and Processes (eds D.R. Lees and D.
Edwards), Academic Press, London, pp. 229-49.
Bush, G.L. (1994) Sympatric speciation. Trends in Ecology and Evolution, 9, 285-8.
Bush, G.L. (1995) Species and speciation, reply from G. Bush. Trends in Ecology and
Evolution, 10, 38.
Butiin, R. (1989) Reinforcement of premating isolation, in Speciation and its
Consequences, (eds D. Otte and J.A. Endler), Sinauer Associates, Sunderland,
Massachusetts, pp. 158-79.
Butiin, R. (1995) Reinforcement: an idea evolving. Trends in Ecology and Evolution,
10,432-4.
Cain, A. J. (1954) Animal Species and their Evolution. Hutchinson, London.
Cain, A.J. (1958) Logic and memory in Linnaeus's system of taxonomy. Proceedings
of the Linnean Society of London, 169,144-63.
Cain, A.J. (1993) Linnaeus's Ordines naturales. Archives of Natural History, 20,405-15.
Chandler, C.R. and Gromko, M.H. (1989) On the relationship between species con-
cepts and speciation processes. Systematic Zoology, 38,116-25.
Claridge, M.F. (1988) Species concepts and speciation in parasites, in Prospects in
Systematics (ed. D.L. Hawksworth), Clarendon Press, Oxford, pp 92-111.
Claridge, M.F. (1995a) Species concepts and speciation in insect herbivores: plant-
hopper case studies. Bolletino di Zoologia, 62, 53-8.
Claridge, M.F. (1995b) Species and speciation. Trends in Ecology and Evolution, 10,38.
Claridge, M.F., Dawah, H.A. and Wilson, M.R. (eds) (1997) Species: the Units of
Biodiversity, Chapman & Hall, London.
Coyne, J.A., Orr, H.A. and Futuyma, D.J. (1988) Do we need a new species con-
cept? Systematic Zoology, 37,190-200.
Cracraft, J. (1983) Species concepts and speciation analysis, in Current Ornithology
(ed. R. Johnston), Plenum, New York, pp. 159-87.
Cracraft, J. (1992) The species of the birds-of-paradise (Paradisaeidae): applying
the phylogenetic species concept to complex patterns of diversification.
Cladistics, 8,1-43.
Cracraft, J. (1997) Species concepts in systematics and conservation - an ornitho-
logical viewpoint, in Species: the Units of Biodiversity (eds M.F. Claridge, H.A.
Dawah and M.R. Wilson), Chapman & Hall, London, pp. 325-39.
De Bach, P. (1969) Uniparental, sibling and semi-species in relation to taxonomy
and biological control. Israel Journal of Entomology, 4,11-27.
Dobzhansky, T. (1937) Genetics and the Origin of Species. Columbia University Press,
New York.
Eldredge, N. (1995) Species, selection, and Paterson's concept of the specific-mate
recognition system, in Speciation and the Recognition Concept (eds D.M. Lambert
and H.G. Spencer), The Johns Hopkins University Press, Baltimore, London,
pp. 464-77.
Ereshefsky, M. (ed.) (1992) The Units of Evolution: Essays on the Nature of Species. MIT
Press, New York.
Foottit, R. (1997) Recognition of parthenogenetic insect species, in Species: the Units
of Biodiversity (eds M.F. Claridge, H.A. Dawah and M.R. Wilson), Chapman &
Hall, London, pp. 291-307.
Gornall, R. J. (1997) Practical aspects of the species concept in plants, in Species: the
Units of Biodiversity (eds M.F. Claridge, H.A. Dawah and M.R. Wilson),
Chapman & Hall, London, pp. 171-90.
Hammond, P.M. (1992) Species inventory, in Global Biodiversity: Status of the earth's
Living Resources (ed. B. Groombridge), Chapman & Hall, London, pp. 17-39.
Hennig, W. (1966) Phylogenetic Systematics. University of Illinois Press, Urbana.
Hull, D.L. (1997) The ideal species concept - and why we can't get it, in Species: the
Knowlton, N. (1997) Species of marine invertebrates: current practices and options
for the future, in Species: the Units of Biodiversity (eds M.F. Claridge, H.A. Dawah
and M.R. Wilson), Chapman & Hall, London, pp. 199-219.
Kornet, D. (1993) Permanent splits as speciation events: a forward reconstruction
of the internodal species concept. Journal of Theoretical Biology, 164,407-35.
Lambert, D.M. and Spencer, H.G. (eds) (1995) Speciation and the Recognition Concept.
The Johns Hopkins University Press, Baltimore, London.
Lane, R. (1997) The species concept in bloodsucking vectors of human diseases, in
Species: the Units of Biodiversity (eds M.F. Claridge, H.A. Dawah and M.R.
Wilson), Chapman & Hall, London, pp. 273-89.
Liou, L.W. and Price, T.D. (1994) Speciation by reinforcement of premating isola-
tion. Evolution, 48,1451-9.
Mallet, J. (1995). A species definition for the modern synthesis. Trends in Ecology
and Evolution, 10, 294-9.
May, R. (1990) How many species?. Philosophical Transactions of the Royal Society B,
330, 293-304.
References 15
Mayden, R. L. (1997) A hierarchy of species concepts: the denouement in the
saga of the species problem, in Species: the Units of Biodiversity (eds M.F.
Claridge, H.A. Dawah and M.R. Wilson), Chapman & Hall, London, pp.
381-424.
Mayr, E. (1942) Systematics and the Origin of Species from the Viewpoint of a Zoologist.
Columbia University Press, New York.
Mayr, E. (1982) The Growth of Biological Thought. Harvard University Press,
Cambridge, Massachusetts.
Mayr, E. (1988) The why and how of species. Biology and Philosophy, 3,431-41.
Mayr, E. (1992) A local flora and the biological species concept. American Journal of
Botany, 79,222-38.
Mayr, E.(1963) Animal Species and Evolution. Harvard University Press, Cambridge,
Massachusetts.
Nelson, G. (1989) Cladistics and evolutionary models. Cladistics, 5, 275-89.
Nixon, K. and Wheeler, Q.D. (1990) An amplification of the phylogenetic species
concept. Cladistics, 6,211-23.
Otte, D. and Endler, J.A. (eds) (1989) Speciation and its Consequences. Sinauer
Associates, Sunderland, Massachusetts.
Paterson, H.E.H. (1985) The recognition concept of species, in Species and Speciation
(ed. E.S. Vrba), Transvaal Museum, Pretoria, pp. 21-9.
Paterson, H.E.H. (1993) Evolution and the Recognition concept of Species, Collected
Writings of H.E.H. Paterson (ed. S.F. McEvey), The Johns Hopkins University
Press, Baltimore, London.
Poulton, E.B. (1908) What is a species? in Essays on Evolution 1889-1907, Clarendon
Press, Oxford, pp. 46-94.
Ramsbottom, J. (1938) Linnaeus and the species concept. Proceedings of the Linnean
Society of London, 150,192-219.
Regan, C.T. (1926) Organic Evolution. Report of the British Association for the
Advancement of Science, 1925, 75-86.
Simpson, G.G. (1951) The species concept. Evolution, 5,285-98.
Sokal, R.R. and Crovello TJ. (1970) The biological species concept: a critical evalu-
ation. The American Naturalist, 104,127-53.
Tinbergen, N. (1951) The Study of Instinct, Oxford University Press, Oxford.
Vrba, E. (1995) Species as habitat-specific, complex systems, in Speciation and the
Recognition Concept (eds D.M. Lambert and H.G. Spencer), The Johns Hopkins
University Press, Baltimore, London, pp. 3—44.
Walker, F. (1832) Monographia Chalcidum. Entomological Magazine, 1,12-29.
Wallace, A.R. (1889) Darwinism: An exposition of the Theory of Natural Selection.
Macmillan, London.
Wheeler, Q.D. and Meier, R. (eds) (1997) Species Concepts and Phylogenetic Theory: A
Debate. Columbia University Press, New York (in press).
Wheeler, Q.D. and Nixon, K.C. (1990) Another way of looking at the species prob-
lem: a reply to de Queiroz and Donoghue. Cladistics, 6, 77-81.
White, G. (1789) The Natural History of Selborne. White, Cochran, London.
Wiley, E. (1978) The evolutionary species concept reconsidered. Systematic Zoology,
27,17-26.
Wilson, E.O. (1992) The Diversity of Life. Harvard University Press, Cambridge,
Massachusetts.
Viral species
M. H. V Van Regenmortel
Contacting address: Institut de Biologic Moleculaire et Cellulaire, CNRS, 15 rue Rene
Descartes, 67084 Strasbourg, France
ABSTRACT
Species is the universally accepted term for the lowest taxonomic
cluster of living organisms. It has been argued that species taxa
should be regarded as individuals and not as classes or categories
because species change during evolution while classes are
immutable and timeless. This viewpoint is based on the notion that
species correspond to so-called Aristotelian classes or universal class-
es that can be defined by one or more properties that are both nec-
essary and sufficient for class membership. However, because of the
inherent variability of the organisms constituting a species taxon, the
species category does not fit the classical notion of class but is more
like a fuzzy set with no clear-cut boundaries.
Although viruses are not living organisms, it is possible to use the
species concept in virology because viruses are biological entities, not
simply chemicals. Viruses have genomes, replicate, evolve and occu-
py particular ecological niches. In 1991, the International Committee
on Taxonomy of Viruses (ICTV) accepted the following definition of
virus species: A virus species is a polythetic class of viruses that con-
stitutes a replicating lineage and occupies a particular ecological niche.
The definition incorporates the notions of genome, biological replica-
tion and natural selection, since the term 'replicating lineage' indicates
an inherited genealogy extending over many generations and unified
by a common descent. The reference to 'ecological niche occupancy'
in the definition brings in the role played by environmental determi-
nants such as host, tissue and vector tropisms in maintaining species
identity. This definition does not provide a list of diagnostic properties
for recognizing members of particular virus species. The characters
most commonly used for recognizing members of individual species
are certain features of genome, the presence of antigenic cross-reac-
18 Viral species
tions and various biological properties such as host range and reac-
tions, tissue tropism, type of vector and transmission route.
It should be stressed that the definition of species as a polythetic
class rules out the possibility of finding a single character that could be
used as a differential diagnostic property. The members of a virus
species do not have a single defining property in common that is nec-
essary and sufficient for class membership (i.e. a property that is com-
mon and peculiar only to members of the species). Certain common
properties such as morphological features or genome composition are
shared also by members of other virus species and such properties
define higher categories such as genera and families and are not
species-defining properties. A single diagnostic property such as a
particular level of genome homology, the extent of antigenic similari-
ty or a certain host reaction will always fail as a criterion for member-
ship of a particular virus species.
The classification of viruses should not be confused with the classi-
fication of viral genome sequences. Viruses are biological entities and
the notion of ecological niche is a crucial component for demarcating
individual viral species.
2.1 INTRODUCTION
For many years the world community of virologists could not agree on the
status and nomenclature of the taxa to be used in virus classification
(Matthews, 1983, 1985). The virologists who study the viruses that infect
plants were particularly reluctant to apply the species concept in virology,
arguing that entities that reproduce by clonal means could not be accom-
modated within the classical definition of biological species (Harrison,
1985; Milne, 1985). Those plant virologists who were opposed to the use of
the species concept in virology took the view that the only legitimate def-
inition of species was that of biological species characterized by gene pools
and reproductive isolation, and applicable only to sexually reproducing
organisms. After several years of vigorous debate concerning the validity
of various alternative species concepts (Matthews, 1983; Bishop, 1985;
Kingsbury, 1988; Milne, 1988; Van Regenmortel, 1989, 1990), the
International Committee on Taxonomy of Viruses (ICTV) agreed in 1991
that the usual categories of species, genus and family should also be used
in virus classification (Pringle, 1991; Van Regenmortel et al., 1991).
Eventually, it was accepted that the species concept is applicable in virol-
ogy because viruses have genomes, replicate, evolve and occupy particu-
lar ecological niches. The following definition of virus species was
endorsed by the ICTV: 'A virus species is a polythetic class of viruses that
constitutes a replicating lineage and occupies a particular ecological niche'
(Van Regenmortel, 1990). The earlier reluctance of some virologists to
accept viral species was due in part to the well-known difficulties posed
by the classification of asexual organisms (Mayr, 1982; Holman, 1987). The
Semantics 19
difficulties experienced by virologists in arriving at an acceptable defini-
tion of virus species are thus an illustration of the more general issues that
arise in any classification when attempts are made to deal with organisms
that reproduce in a clonal or parthenogenic manner.
2.2 SEMANTICS
Part of the confusion surrounding the debates about species is of a seman-
tic nature. The term 'species' can be used to refer to a taxonomic category
in which case it corresponds to an abstract concept devoid of any spa-
tiotemporal location. This use of the word species either as a class of cate-
gories used in taxonomy or as a class of organisms is often confused with
another meaning of the word, namely that of a concrete collective entity
made up of real organisms localized in space and time, i.e. a taxon. When
the word species is used to refer to a practical entity of real organisms, it
may enter into what logicians call part-whole relations applicable only to
spatiotemporally localized entities (Hull, 1976). The part-whole relation
applies for instance to a particular dog which is part of the taxon dog com-
prising all animals with dog features. On the other hand, the abstract con-
cept of species cannot enter into part-whole relations but instead can take
part in relations known as class-inclusion or class-membership. Although
an organism may thus be considered as a member of a species (viewed as
a class), it is logically impossible for it to be part of an entity of different
logical type such as the abstract concept of class.
Confusion between the abstract and practical usages of the term species
is responsible for innumerable idle debates about the reality of species.
Many biologists readily accept that genera and families are artificial,
abstract constructions of the mind, but insist that species are real, i.e.
endowed with an objective reality and individuality. The inability to con-
ceive of species as a conceptual construction led Milne (1984) for instance
to assert 'Linnaeus did not create species, he found them'. The same reluc-
tance to view species as classes led to the proposal that particular species
should be regarded as individuals. According to this viewpoint, species
are constituted of organisms in the same way that an individual organism
is constituted of cells and organs (Ghiselin, 1974; Hull, 1976). The proposal
that species should be regarded as individuals, i.e. as practical entities and
not as abstract classes, stems from the belief that all classes are necessarily
Aristotelian classes, immutable and timeless. Since species taxa change
during evolution, they cannot correspond to universal Artistotelian class-
es and this is advanced as an argument against viewing species as a class.
However, as discussed below, the concepts of polythetic class and fuzzy
set make it possible to reconcile phylogenetic change with class member-
ship and this removes the rationale for considering species only as real
individuals. The suggestion that biological classification is concerned with
20 Viral species
individuals and not with classes was not extended to higher taxa although
there is, of course, no reason why genera could not be considered as indi-
viduals constituted of species. Presumably the protagonists of the species-
as-individual thesis recognized that if genera, families and kingdoms are
not allowed to be classes, any taxonomy becomes impossible.
As pointed out by Quine (1987), universal classes and properties are
related abstract entities. Ascribing a property to a thing, for instance
spherical shape or possession of an RNA genome, amounts to assigning
the thing to a universal class, i.e. the class of spheres and of entities con-
taining a RNA genome. In this sense, viruses or organisms can be mem-
bers of various universal classes corresponding to higher taxonomic cate-
gories such as genera or families. These taxonomic classes correspond to
Aristotelian classes defined by a single property or by a set of properties
necessary and sufficient for membership in the class. In contrast, species
are not universal classes and their members do not have a single defining
property in common (Beckner, 1959). A final note of caution about the
distinction between practical species taxa and abstract species classes
should be made. When species are viewed as taxa it is impossible to
define them. They can only be given proper names in an arbitrary man-
ner analogous to baptism (Kitts, 1984). Only when they are viewed as
abstract classes can species be defined. However, such a definition of the
concept is of little help for identifying the members of a particular
species. For example, the definition of biological species in terms of gene
pools and reproductive isolation is of little use for identifying members of
the species. The diagnostic properties of real objects should not be con-
fused with the theoretical, defining properties of abstract classes
(Ghiselin, 1984).
2.3 CONTINUITY VERSUS DISCONTINUITY AND THE PROBLEM

OF SPECIES DEMARCATION
Ever since the demise of the theory of spontaneous generation of living
organisms, it has been recognized that life is a continuum linking indi-
viduals to their parents, unicellular organisms to multicellular organisms
and animals to man. Although the study of biology is thus concerned
with a continuous historical development embodied in the uninter-
rupted chain of replicating DNA molecules, there is also clearly a need
to recognize discontinuities among different life forms. Species undergo
continuous variation that is regulated by discontinuous single
nucleotide changes. Transition from one species to another during evo-
lution occurs within the continuity of gene pools. In spite of this conti-
nuity, however, so-called evolutionary species are considered separate
at one particular point in time (Lovtrup, 1979). Difficulties in reconciling
Species as polythetic classes 21
continuous and discontinuous frames of references are not unique to
biology. There is, for example, vigorous debate concerning the possible
continuous or discontinuous nature of space and time. On a less abstract
note the Swiss sometimes find it difficult to say exactly where the
Jungfrau and neighbouring Monch mountains start and stop, but this
does not lead them to doubt the reality of these two mountains because
their limits are unclear. Although one cannot make absolutely clear dis-
tinctions where none exists, it is useful to identify and give names to
parts of a continuum, whether it be a geological rock formation, a colour
in the visible spectrum of electromagnetic waves or a particular influen-
za virus responsible for a worldwide pandemic.
The classical concept of biological species demands that we identify
clear-cut breeding discontinuities that are mostly absent. Similarly, the
concept of evolutionary species requires that we demarcate boundaries in
time that coincide with the appearance of new species, a task equally
impossible to do in practice. The solution lies in viewing species as poly-
thetic classes since this does away with the need to use any single discon-
tinuity as criterion for demarcation.
2.4 SPECIES AS POLYTHETIC CLASSES

The polythetic species concept was introduced by Beckner (1959) to
replace the classical notion of universal class. He gave the name polytypic
(later changed to polythetic) to classes that are defined by a combination
of characters, each of which may occur also outside the given class and
may be absent in any member of the class. The nature of polythetic class-
es can be illustrated by the following example (Sattler, 1986). Suppose a
species is defined by a set of five properties Fl, F2, F3, F4 and F5. If these
properties are distributed in the way shown in Table 2.1, the class will be
polythetic. This example represents a polythetic class because each indi-
vidual possesses a large number of the properties (i.e. four out of five),
each property is possessed by a large number of individuals and no prop-
erty is possessed by all individuals. Contrary to the situation with univer-
sal classes, no single property is either necessary or sufficient for member-
ship in a polythetic class. The concept of polythetic class is extremely use-
ful for dealing with biological entities endowed with intrinsic variability,
since it can accommodate individual members that lack one or other char-
acter considered typical of the class. In this kind of class, certain elements
may evolve and there is no difficulty in reconciling class membership with
phylogenetic change. This makes a polythetic species similar to a fuzzy set
(Beatty, 1982; Kosko, 1994) with boundaries that are modifiable and not
uniquely defined. The view that species are sets has been elaborated by
Kitcher (1984).
22 Viral species
Table 2.1 Distribution of five properties, Fl to F5, among five members of a poly-
thetic class
Individual Properties
\ Fl F2 F3 F4
2 Fl F2 F3 F5
3 Fl F2 F4 F5
4 Fl F3 F4 F5
5 F2 F3 F4 F5
2.5 SPECIES FUZZINESS

It is generally accepted that species sets consist of organisms that are phy-
logenetically connected, although in the case of structures as simple as
some viruses it cannot be excluded that the same pathogenic entity might
have arisen or evolved from a different parent more than once during bio-
logical evolution. The fuzziness inherent in species sets arises from the
rejection of so-called binary or bivalent logic, according to which every
statement or sentence is either true or false (i.e. A or not A). This type of
logical dichotomy is absent in fuzzy logic, which on the contrary assumes
that everything is a matter of degree, including truth and set membership.
Fuzzy logic and fuzzy sets were introduced by Lofti Zadeh (1965) and
have been popularized by Bart Kosko (1994). When fuzzy logic is applied
to the diagnostic characters that are used for identifying members of a
species, these properties are taken as multivalent and capable of assuming
several intermediate states. Life itself is a fuzzy concept, as shown for
instance by the ambiguities that result when its beginning in man is
defined at conception and its end is made to coincide with brain damage.
Viruses are an interesting example of the difficulty of defining life since
they represent a type of 'borrowed' life. Viruses are entirely dependent on
living cells for their replication and are devoid of any metabolic activity,
although they are endowed with life-like properties such as genetic conti-
nuity and adaptability. It is also a sobering thought that as the amount of
information increases together with greater precision of collected data, the
fuzziness actually increases rather than decreases (Kosko, 1994). It used to
be believed that when the entire genome sequences of many different
viruses would become available, a simple examination of these sequences
expressing the complete viral blueprint would enable one to assess if indi-
vidual viruses belonged to the same or to different species. This in fact has
not happened and it is now evident that the boundary between two viral
species in terms of percentage of sequence identity cannot be drawn in a
non-fuzzy manner. In the family Potyviridae, for example, it is accepted
that different potyvirus species exhibit coat protein sequence identities of
References 23
40-70% while strains of the same potyvirus have sequence identities of
90-99%. However, two potyviruses showing 85% coat protein sequence
identity could be considered either as two different species or as two
strains of the same species (Van Regenmortel, 1992). There is no single
sequence identity figure that can be stated as the absolute cut-off between
species and strains but there is a range of values (80-90%) within which
the alternative categories can be considered sensible options, depending
on the relative significance of other properties (Shukla et al, 1994: 207).
2.6 SPECIES OR QUASISPECIES

It should be stressed that it is no longer accepted that a virus species can be
defined by a single genome sequence. Because RNA viruses have genomes
that replicate in the absence of repair mechanisms, they evolve very rapid-
ly with a mutation frequency per nucleotide site in the viral genome of 10~3
to 10~5. Since RNA viral genomes usually contain about 104 nucleotides, a
clone of an RNA virus consisting of say 1012 particles will always consist of a
complex mixture of millions of different genomes all of which compete dur-
ing replication of the clone (Holland et al., 1992). In view of their genome
plasticity, RNA viruses are usually considered to be quasispecies popula-
tions. The term quasispecies was introduced by Eigen (for review, see Eigen,
1993) to describe the distribution of self-replicating RNAs believed to be the
first genes on earth. A quasispecies population consists of a master sequence
corresponding to the most fit genome sequence with respect to a given
environment together with innumerable competing virus mutants. It
should be noted that the use of the term quasispecies for a virus population
does not imply that there exists an entity called virus species with a single,
invariant genome sequence. The current status of virus classification,
including a discussion of the various taxonomic categories used in virology,
can be found in Shukla et al. (1994) and Murphy et al. (1995).
2.7 REFERENCES
Beatty, J. (1982) Classes and cladists. Systematic Zoology, 31, 25-34.
Beckner, M. (1959) The Biological Way of Thought. Columbia University Press, New
York.
Bishop, D.H.L. (1985) The genetic basis for describing viruses as species.
Intervirology, 24, 79-93.
Eigen, M. (1993) Viral quasispecies. Scientific American, 269,32-9.
Ghiselin, M.T. (1974) A radical solution to the species problem. Systematic Zoology,
23,536-44.
Ghiselin, M.T. (1984) Definition, character and other equivocal terms. Systematic
Zoology, 33,104-10.
Harrison, B.D. (1985) Usefulness and limitations of the species concept for plant
viruses. Intervirology, 24, 71-8.
24 Viral species
Holland, J.J., de la Torre, J.C. and Steinhauer, D.A. (1992) RNA virus populations
as quasispecies, in Genetic Diversity of RNA Viruses (ed. J.J. Holland), Springer-
Verlag, Vienna, pp. 1-20.
Holman, E.W. (1987) Recognizability of sexual and asexual species of rotifers.
Systematic Zoology, 36,381-6.
Hull D.L. (1976) Are species really individuals? Systematic Zoology, 25,174-91.
Kingsbury, D.W. (1988) Biological concepts in virus classification. Intervirology, 29,
242-53.
Kitcher, P. (1984) Species. Philosophy of Science, 51, 308-33.
Kitts, D.B. (1984) The names of species: a reply to Hull. Systematic Zoology, 33,
112-15.
Kosko, B. (1994) Fuzzy Thinking. HarperCollins, London.
Lovtrup, S. (1979) The evolutionary species: fact or fiction? Systematic Zoology, 28,
386-92.
Matthews, R.E.F. (1983) A Critical Appraisal of Viral Taxonomy. CRC Press, Boca
Raton, Florida.
Matthews, R.E.F. (1985) Viral taxonomy for the nonvirologist. Annual Review of
Microbiology, 39, 451-74.
Mayr, E. (1982) The Growth of Biological Thought. Diversity, Evolution, and Inheritance.
Harvard University Press, Cambridge, Massachusetts.
Milne, R.G. (1984) The species problem in plant virology. Microbiological Science, 1,
113-22.
Milne, R.G. (1985) Alternatives to the species concept for virus taxonomy.
Intervirology, 24, 94—8.
Milne, R.G. (1988) Species concept should not be universally applied to virus tax-
onomy - but what to do instead? Intervirology, 29, 254-9.
Murphy, F.A., Fauquet, C.M., Bishop, D.H.L., Ghabrial, S.A., Jarvis, A.W., Martelli,
G.P., Mayo, M.A. and Summers, M.D. (eds) (1995) Virus Taxonomy, Sixth Report
of the International Committee on Taxonomy of Viruses. Springer-Verlag, Vienna.
Pringle, C.R. (1991) The 20th meeting of the executive committee of ICTV. Virus
species, higher taxa, a universal database and other matters. Archives of
Virology, 119, 303-4.
Quine, W.V. (1987) Classes versus properties, in Quiddities, Belknap Press of
Harvard University Press, Cambridge, pp. 22-24.
Sattler, R. (1986) Biophilosophy. Analytic and Holistic Perspectives. Springer, Berlin.
Shukla, D.D., Ward, C.W. and Brunt, A.A. (eds) (1994) The Potyviridae, CAB
International, Cambridge, UK.
Van Regenmortel, M.H.V. (1989) Applying the species concept to plant viruses.
Archives of Virology, 104,1-17.
Van Regenmortel, M.H.V. (1990) Virus species, a much overlooked but essential
concept in virus classification. Intervirology, 31, 241-54.
Van Regenmortel, M.H.V. (1992) What is a virus? in Potyvirus Taxonomy (ed. O.W.
Barnett), Archives of Virology, Suppl. 5, Springer-Verlag, Vienna, pp. 47-53.
Van Regenmortel, M.H.V., Maniloff, J. and Calisher, C. (1991) The concept of virus
species. Archives of Virology, 120, 313-14.
Zadeh, L.A. (1965) Fuzzy sets. Information and Control, 8, 338-53.
Towards a practical species
concept for cultivable bacteria
M. Goodfellow, G. P. Manfio and J. Chun
Contacting address: Department of Microbiology, The Medical School, Framlington Place,
Newcastle upon Tyne NE2 4HH, UK
ABSTRACT
The basic unit in bacteria systematics has long been recognized as the
species. However, despite this, there is still no universally accepted
definition of species in bacteriology. The traditional view is that bac-
terial species can be distinguished by correlated phenotypic characters
and, as such, members of a given species have a combination of char-
acters peculiar to it. In practice, the number of such phenotypic
species in a genus is influenced by the aims of the taxonomist, the
extent to which the taxon has been studied, the criteria adopted to
define the species and the ease by which strains can be brought into
pure culture. The phenotypic species concept has been useful in prac-
tice but has severe limitations.
Bacterial species can now be defined in molecular terms. Indeed,
DNA : DNA relatedness is often seen as the gold standard for the cir-
cumscription of bacterial species. This method is attractive as it can be
applied to all prokaryotes, irrespective of their growth requirements.
Although the exact level below which organisms are considered to
belong to different species varies, extensive studies with the family
Enterobacteriaceae and related taxa have led to the recommendation
that genomic species should encompass strains with approximately
70% or more DNA : DNA relatedness with a difference of 5°C or less
in thermal stability. 16S ribosomal RNA is now routinely used to high-
light novel - that is, previously undescribed - species, but it is not
always possible to detect diverged species in this way.
It is now becoming increasingly accepted that the integrated use of
genotypic and phenotypic characteristics - that is, polyphasic taxono-
my - is necessary for the delineation of bacterial taxa, including species.
This polyphasic species concept will be considered with reference to
suitable examples of organisms of medical and industrial importance.
26 Towards a practical species concept for cultivable bacteria
3.1 INTRODUCTION
It is widely recognized that the species is the basic unit in biological clas-
sification. This emphasis on the species raises the intriguing problem of
the nature and comparability of species across the whole range of biolog-
ical diversity. The most widely accepted species concept is the biological
species which can be considered as an interbreeding or potentially inter-
breeding community of populations. A major problem with the biological
species concept is that it is not universally applicable, even in sexual
organisms! It is, for example, not possible to test members of all possible
pairs of species for their ability to interbreed. The small size of bacteria
(including Archaea), their mainly asexual reproductive behaviour and the
dearth of knowledge on the genetics of bacterial populations necessitate a
pragmatic approach to the bacterial species concept. More generally, the
term species implies distinctness between organisms, an approach which
encompasses all organisms irrespective of whether they are classified as
Archaea, Bacteria or Eukarya.
Little is known about the taxonomic structure of bacteria (Sneath, 1985).
It is generally accepted that the vast majority of strains fall into distinct
phenetic clusters separated by definite gaps. However, some (Cowan,
1955, 1962) consider that bacteria might instead form a continuous spec-
trum. This may prove to be the case with some groups though it seems
highly unlikely that there are no gaps at all as this would mean that all
possible combinations of properties would occur among bacteria.
Recent attempts to define bacterial species have tended to reflect the
methods used to classify individual strains. This is highlighted by the dra-
matic impact which modern taxonomic methods have had on the ways in
which bacteria are classified (Stackebrandt and Goodfellow, 1991;
Goodfellow and O'Donnell, 1993,1994). Technique-driven approaches to
the circumscription of bacterial species are reasonably sound in an opera-
tional sense, but they overlook the fact that species are the product of bio-
logical processes. It is the pattern of distinctive properties shown by bac-
teria not the process which gave rise to them which is currently seen to be
paramount in bacterial systematics.
The development of a universally accepted species concept for bacteria
is proving to be a formidable task. In practice, bacterial species are usual-
ly taken to be groups of strains which individually show high levels of bio-
chemical, genetical, morphological, nutritional and structural similarity.
This somewhat ill-defined operational species concept is widely applied in
bacteriology. It is, for example, used by diagnostic bacteriologists to tease
out the complex taxonomy of new and emerging pathogens (McNeil and
Brown, 1994), by industrial microbiologists searching for novel, commer-
cially significant microbial products (Bull et al, 1992), by molecular ecolo-
gists monitoring the impact of genetically manipulated bacteria released
Early species concepts 27
into the environment (Edwards, 1993), by soil microbiologists trying to
establish relationships between microbial diversity and sustainable land
management (Hawksworth, 1991), by molecular biologists engaged in
reconstructing bacterial evolution (Woese, 1987, 1992), and by bacterial
systematists intent on unravelling the extent of bacterial diversity in nat-
ural habitats (O'Donnell et al, 1994).
This contribution is designed to demonstrate how recent developments
in bacterial systematics are being used to provide an improved operational
species concept for cultivable bacteria. However, it must be remembered
that the number of bacterial species known and described represents only
a tiny fraction of the estimated species diversity (Bull et al., 1992; Embley
and Stackebrandt, 1997: Chapter 4).
3.2 EARLY SPECIES CONCEPTS

The species concept is still a difficult and controversial issue in bacterial
systematics (Cowan, 1955, 1959, 1962; Ravin 1960; Gordon, 1978;
Goodfellow and O'Donnell, 1993). Early definitions of bacterial species
were often based on monothetic groups described by subjectively select-
ed sets of phenotypic properties. This species concept had severe limita-
tions as strains which varied in key characters could not be accommodat-
ed in existing taxonomies. Moreover, such classifications often lacked uni-
formity as different criteria were frequently used by different investigators
for the same group of organisms. Another consequence of this approach
was that the rank of species became very unevenly defined over the
whole range of bacterial variation. The legacy of this approach is still
apparent. Typical examples are the current treatments of the families
Bacillaceae and Enterobacteriaceae with organisms in the first group
underdifferentiated (Rainey et al., 1993; White et al, 1993), whereas in the
latter taxon different generic descriptions are preserved for bacteria relat-
ed at the species level (Palleroni, 1993).
It is evident from the early literature that the number of species in a
genus was influenced by the aims of the taxonomist, the extent to which
the taxon had been studied, the criteria adopted to define the species and
the ease by which the strains could be brought into pure culture (Williams
et al, 1984). Some idea of the variations in the number of species within
bacterial genera can be gleaned from Table 3.1. In general, members of
ecologically and medically important genera have been underclassified
and those in industrially significant taxa overclassified. This latter point is
well illustrated by reference to the genus Streptomyces. Over 3000 species
have been named, mainly in the patent literature, on the basis of their
capacity to produce novel secondary metabolites (Trejo, 1970). However,
Pridham and Tresner (1974) only recognized 463 Streptomyces species
using a limited number of standardized phenotypic tests.
Table 3.1 Number of species in selected bacterial genera
Genus Bergey's Manual Approved lists
1974 1980 1995
A. Medically important bacteria

Actinomyces 5 6 20
Aeromonas 3 3 15
Campylobacter 3 4 15(9)*
Clostridium 61 75 123(12)
Helicobacter - - 12
Legionella - 1 34(6)
Mycobacterium 29 41 70
Mycoplasma 36 57 98(6)
Nocardia 23 20 15(11)
Staphylococcus 3 13 33
Streptococcus 21 27 42(16)
Vibrio 5 9 35(3)
B. Ecologically important bacteria
Azospirillum - 2 5
Azotobacter 6 6 6
Beijerinckia 4 4 4
Bradyrhizobium - - 3
Curtobacterium 6 6 6
Nitrobacter 1 1 1
Phyllobacterium - - 2
Rhizobium 6 6 11(3)
Rhodospirillium 5 5 6
C. Industrially important bacteria
Actinoplanes 6 9 21(1)
Amycolatopsis - - 9
Bacillus 48 31 82(14)
Lactobacillus 27 35 60(11)
Micromonospora 16 12 15
Rhodococcus - 10 11(11)
Saccharomonospora - 1 4
Streptomyces 463 337 463(3)
Streptosporangium 11 13 13
Thermoactinomyces 4 5 8
* The numbers in parentheses refer to the number of species transferred to other genera.
Cowan (1968) recognized the subjective nature of the traditional

species concept when he whimsically described a species as 'a group of
organisms defined more or less subjectively by the criteria chosen by the
Early species concepts 29
taxonomist to show to best advantage and as far as possible put into prac-
tice his individual concept of what a species is'. Other attempts to define
bacterial species were also descriptive. Stanier et al. (1986), for example,
stated that, 'a species consists of an assemblage of clonal populations that
share a high degree of phenotypic similarity coupled with an appreciable
dissimilarity from other assemblages of the same general kind'. The sub-
jective nature of the species concept led not just to the circumscription of
phenotypic species but also to the classification of taxonomists into
lumpers and splitters! Lumpers tend to emphasize the similarities
between strains, thereby recognizing relatively few taxa, whereas splitters
highlight differences in the belief that the clarity inherent in small groups
is paramount.
The limitations of the classical approach to circumscribing phenotypic
species are readily exemplified. Thus, the sugar tests used to separate
Bacillus circulans from other facultative bacilli are of no value in distin-
guishing between other members of the genus, such as Bacillus sphaericus
which have an oxidative metabolism and as such cannot metabolize sug-
ars (Priest, 1993). This particular approach to the delineation of Bacillus
species was intrinsically flawed as most sugar-utilizing strains formed
homogeneous species whereas their oxidative counterparts belonged to
genetically diverse groups.
Buchanan (1955), a firm believer in the nomenclatural type concept,
considered that 'A bacterial species may (then) be defined as the type cul-
ture together with such other cultures or strains of bacteria as are accept-
ed by bacteriologists as sufficiently closely related'. This definition
acknowledges the fact that the type strain of a species need not be the
most characteristic strain of that species. Gordon (1967,1978) took a very
different view; she considered that species descriptions should be based
on type strains, other authentic reference strains, fresh isolates, and old
stock cultures and their variants.
Three very practical kinds of species were recognized by Ravin (1963):
namely genospecies, which encompass mutually interfertile forms and
correspond most closely to the biological species concept; nomenspecies,
made up of individuals resembling the nomenclatural type strain; and tax-
ospecies, groups of strains which share a high proportion of common
properties. Genospecies should not be confused with genomic species,
that is, with organisms which share high DNA relatedness values.
The genetic concept of species based on fertility has contributed little
towards a more precise definition of bacterial species (Ravin, 1960; Jones,
1989). Genetic material can pass between bacteria in two ways: (i) verti-
cally, from mother to daughter cell during asexual reproduction by binary
fission; and (ii) horizontally between cells through the main genetic
exchange mechanisms, transduction, transformation and conjugation.
Populations in which vertical transmission predominates are clonal. Thus,
as bacteria divide and spread, new alleles derived from mutations are
restricted to the direct descendants of the cell in which they arose and can-
not reassert with alleles of other genes into novel genetic combinations. In
contrast, horizontal genetic exchange provides a mechanism for sexual
transmission of genetical material allowing the reassortment of mutant
alleles among strains and thereby promoting the genetic diversity of the
species (Maynard Smith et al., 1991; Maynard Smith, 1995). This process
works against the natural tendency of an asexual population to be clonal.
Consequently, there is a spectrum of population structures that bacterial
populations can exhibit ranging from highly clonal to panmictic (Maynard
Smith et al., 1993). Evidence of genetic isolation between members of closely
related taxa has been observed (Maynard Smith et al., 1991; Roberts et al.,
1994) and could be due to physical barriers to gene exchange, codon usage
patterns and host restriction/modification mechanisms.
In general, genetic exchange plays a limited role as a force of cohesion
among prokaryotes (Cohan, 1994). Horizontal gene transfer has mainly
been recorded for members of genera which have a special capacity for
the uptake and incorporation of chromosomal DNA (Maynard Smith et al,
1991; Maynard Smith 1995). Such naturally transformable bacteria are
found among both Gram-negative (e.g. Haemophilus and Neisserid) and
Gram-positive bacteria (e.g. Bacillus and Streptococcus). Bacteria differing in
DNA sequence by up to 25% can and do exchange chromosomal DNA.
The exchange is usually local, often involving only a few hundred base
pairs, and hence does not destroy clonal structures. It is possible that most
of the genes which are exchanged frequently confer some immediate
adaptive advantage, such as those encoding cell-surface proteins or
restriction-modification systems, in contrast to 'housekeeping' genes,
such as those encoding enzymes studied by multilocus enzyme elec-
trophoresis (Selander et al., 1994).
Some phages and plasmids mediate transfer between closely related
strains (Jones, 1989; Harwood, 1993). This might appear to provide a basis
on which to develop a universal species concept for bacteria but other
phages and plasmids have a broad host range (Holloway, 1993).
Promiscuous plasmids have the potential to transfer gene operons across
widely divergent taxa, even across phyla. The biological basis of such
extensive host ranges is not known, though some plasmids appear to have
developed complex and highly flexible replicative systems. However,
wide host-range plasmids do not necessarily display conjugal competence
in members of diverse species.
It is perhaps not surprising given the problems outlined above that
little interest has been shown recently in the bacterial species concept. In
practice, species level taxonomy has been based implicitly or explicitly on
the detection of phenotypic discontinuities. The species concept was
The new bacterial systematics 31
never formally abandoned but it would appear to have become a concept
that dare not speak its name!
3.3 THE NEW BACTERIAL SYSTEMATICS

The current renaissance in bacterial systematics can be traced to the intro-
duction and application of new concepts and methods, notably to the emer-
gence of numerical taxonomy (Sokal and Sneath, 1963), chemosystematics
(Goodfellow and Minnikin, 1985) and molecular systematics (Stackebrandt
and Goodfellow, 1991). Views on how bacterial species should be circum-
scribed have radically changed in light of these developments.
3.3.1 Numerical taxonomy

In the late 1950s, the call for a more objective taxonomy led to the devel-
opment of numerical taxonomy (Sokal and Sneath, 1963; Sneath and
Sokal, 1973). As the name implies, numerical taxonomy consists of apply-
ing various mathematical procedures to numerically encoded character
state data; organisms are assigned to groups on the basis of overall simi-
larity. In the 1960s and early 1970s, bacterial classifications, considered
almost entirely in phenetic terms, were generated from numerical analy-
ses of biochemical, nutritional and physiological data.
The primary objective of early numerical taxonomic studies was to assign
individual bacterial strains to homogeneous groups or clusters, which could
be equated with taxospecies, using large sets of phenotypic data. The resul-
tant quantitative data on numerically defined taxospecies were used to
design improved identification schemes. When introduced, the numerical
taxonomic procedure was in sharp contrast to the prevailing orthodoxy as
species were recognized using many equally weighted features, not by a
few subjectively chosen morphological, staining and behavioural proper-
ties. The theoretical basis of numerical taxonomy is well documented
(Sneath and Sokal, 1973; Goodfellow et at, 1985; Sackin and Jones, 1993); the
main steps involved in numerical classification and identification are out-
lined in Figure 3.1. Numerical taxonomic data are usually stored and man-
aged using computer systems given the widespread availability of special-
ized software (Canhos et al., 1993; Sackin and Jones, 1993) and the need to
study large numbers of strains and properties.
Numerical taxonomic procedures have been applied to most groups of
cultivable bacteria (Goodfellow and Dickinson, 1985; Sackin and Jones,
1993) both to revise existing classifications and to classify unknown strains
isolated from diverse habitats. It is evident from these studies that numer-
ical taxonomic procedures are effective in delineating taxospecies. The
method has been less successful in the construction of higher taxonomic
Selection of strains
and characters
I
Collection of data
I
Coding of characters
Rejection of non-differential
or unreproducible data
Final data matrix
Hierarchical cluster analysis
I Ordination methods
Definition of clusters
Frequency matrix
Selection of characters
Identification matrix
Theoretical evaluation
Routine identification
Figure 3.1 Major steps in numerical classification and identification.
ranks, but this is almost certainly due to the types of data used rather than
to fundamental flaws in numerical methods. Thus, representative strains
from diverse genera may have different metabolisms and growth require-
ments which can make studies across generic boundaries difficult.
Numerical taxonomic surveys have been used to circumscribe many tax-
ospecies, including those encompassed in taxonomically complex taxa
such as Bacillus (White et al, 1993; Nielson et al., 1995), Mycobacterium
(Wayne, 1985), Pseudomonas (Strenstom et al, 1990) and Streptomyces
(Kampfer et al, 1991).
Taxonomic clusters or taxospecies are 'operator unbiased' representa-
tions of natural relationships between strains though group composition
may be influenced by the choice of strains and tests, experimental proce-
dures, test error and statistics (Sackin and Jones, 1993). It is, therefore,
essential to evaluate the taxonomic integrity of taxospecies by examining
representative strains using independent taxonomic criteria derived from
the application of chemotaxonomic and molecular systematic methods.
There is evidence that Curie-point pyrolysis mass spectrometry provides
a quick and effective way of evaluating the taxonomic status of tax-
ospecies circumscribed in numerical phenetic surveys (Goodfellow et al.,
1994a).
Comprehensive databases - one of the end-products of numerical tax-
onomy - contain extensive information on the biochemical, nutritional,
physiological and tolerance properties of test strains. These data can read-
ily be arranged into tables which list the percentage of strains in each tax-
ospecies that are positive for each unit character (percentage positive
tables). These data can be used for several purposes, notably:
• To construct frequency matrices for the identification of unknown bac-
teria.
• To design media for the selective isolation of target organisms from nat-
ural habitats.
• To choose representative strains for additional taxonomic studies.
A frequency matrix is a reduced version of a percentage positive matrix
in which a combination of unit characters is selected for the identification
of unknown isolates to taxospecies. The selection of an optimal combina-
tion of features is achieved using intuitive mathematical and statistical
routines (Sneath, 1979a,b, 1980a,b,c). The practical and theoretical devel-
opments in computer-assisted numerical identification have been
described in detail elsewhere (Pankhurst, 1991; Priest and Williams, 1993).
The polythetic nature of probabilistic identification matrices gives
them several advantages over conventional identification keys and diag-
nostic tables as no single property is either sufficient or necessary for the
identification of a strain to a previously defined group. Frequency matri-
ces are also theoretically robust as they can be used to accommodate nat-
ural variation in test results presented by bacteria isolated from diverse
sources. Theoretically sound and practically useful frequency matrices
are available for the identification of industrially and medically impor-
tant taxospecies (Bryant, 1993; Canhos et al., 1993), including campy-
lobacters (On et al., 1996), slowly growing mycobacteria (Wayne et at.,
1984), acidophilic actinomycetes (Seong et al., 1995) and neutrophilic
streptomycetes (Kampfer and Kroppenstedt, 1991). Some probabilisitic
identification matrices can be accessed through the internet (Canhos et
al, 1993; http://www.bdt.org.br/cgl-bin/msdn/matrices). Many commer-
cial diagnostic kits and automated instruments used for the identifica-
tion of unknown pathogenic bacteria are based on numerical taxonomic
methods.
Commercially available automated bacterial identification systems are
now available for the routine identification of isolates in the laboratory
and for the construction of databases (Mauchline and Keevil, 1991).
Identification tests are assembled in microtitre plates or disposable kits
and results read and collected automatically by a plate reader connected
to a microcomputer or visually by the operator (Bochner, 1989). Rapid
automated identification systems are used for rapid and reliable identifi-
cation of clinical isolates, but few systems are available for the identifica-
tion of environmentally important strains (Klinger et al.r 1992).
The wealth of information held in numerical taxonomic databases can
also be used to determine the nutritional and tolerance limits of members
of individual taxospecies in order to devise media formulations selective
for one or more taxospecies. This taxonomic approach to selective isola-
tion has been used to isolate target and novel streptomycete species of
potential industrial importance from soils (Williams et al., 1984; Williams
and Vickers, 1988; Goodfellow et al., 1994a). These studies clearly show
that there is no such thing as a 'general' isolation medium for strepto-
mycetes; several selective media must be used to gain a more accurate pic-
ture of the qualitative nature of bacterial populations in environmental
samples. The taxonomic approach to selective isolation has a crucial part
to play in establishing the extent of bacterial variation in natural habitats
(Bull et al, 1992).
It is evident that the ability to delineate taxospecies has had a profound
influence in bacteriology, notably for the identification of pathogenic bac-
teria. Indeed, the strengths of the numerical taxonomic procedure far out-
weigh its limitations (Table 3.2). Any tendency to see numerical taxonomy
as a method with a long past and an uncertain future should be resisted.
Improved methods and automated data acquisition systems will facilitate
the generation of high-quality phenotypic databases for a variety of pur-
poses. It can, for example, be anticipated that with the developing interest
in bacterial species diversity, these databases, which often reflect the func-
tional diversity of a habitat, will be put to other more fundamental uses.
3.3.2 Molecular systematics

The motive force in contemporary bacterial systematics owes much to
developments in molecular biology. These advances underpin an impres-
sive array of methods which can be used to derive taxonornically useful
information from analyses of nucleic acids. The estimation of the mean
overall base composition of DNA, indirect comparisons of nucleotide
sequences by DNA: DNA pairing and sequence analyses of conserved
homologous genes, notably 16S rRNA, all provide grist for the circum-
scription of bacterial species.
(a) DNA base composition

In bacteria, DNA base composition values range from ~ 25 to 80 mol% GC
(Tamaoka, 1994). The DNA base composition of members of well circum-
Table 3.2 Advantages and disadvantages of numerical classification and identi-
fication
Advantages
Circumscription of taxospecies
Provision of simple phenotypic properties for identification
Formulation of taxon-specific selective media
Choice of representative strains for additional studies
Disadvantages
Collection of data time-consuming and laborious
Some diagnostic tests are difficult to read
Little value for classification above the genus level
Needed
Automated data acquisition systems
scribed taxospecies does not usually differ by more than 3 mol% GC,
whereas members of species within a genus should not usually differ from
one another by more than about 10 mol% GC. Firm guidelines have yet to
be set for the range of DNA base compositions that can be encompassed
at these taxonomic ranks but a range of > 15 mol% GG can be taken to
indicate heterogeneity within a genus. Bacillus (33-64 mol% GC),
Clostridium (22r-55mol% GC) and Flavobacterium (31-68 mol% GC), as
defined in the current edition of Bergey's Manual of Systematic Bacteriology
(Krieg and Holt, 1984; Sneath et al, 1986), are all examples of heteroge-
neous taxa.
DNA base composition data need to be interpreted with care as the
choice of analytical method and experimental conditions influence results
(Tamaoka, 1994). The sensitivity and reproducibility of the thermal melt-
ing point method provide sufficiently good data for taxonomic purposes
though discrepancies occur between different laboratories. The high-
pressure liquid chromatography (HPLC) method, which is more accurate,
should be adopted when DNA base composition data are used to deter-
mine hybridization conditions for DNA : DNA relatedness studies
(Kusunoki and Ezaki, 1992).
(b)DNA: DNA pairing

A unique property of DNA and RNA is the ability for reassociation or
hybridization . The complementary strands of DNA, once denatured, can,
under appropriate experimental conditions, reassociate to reform native
duplex structures. The specific pairings are between the base pairs, adenine
with thymine and guanine with cytosine, and the overall pairing of the
nucleic acid fragments is dependent upon similar linear arrangements of
these bases along the DNA. When comparing nucleic acids from different
organisms, the amount of molecular hybrid formed and its thermal stability
provide an average measurement of nucleotide sequence similarity.
Since the 1970s information from DNA : DNA pairing studies has been
used extensively to delineate bacterial species (Krieg, 1988; Stackebrandt
and Goebel, 1994). An important practical advantage of DNA : DNA pair-
ing was that the continua occasionally found between phenotypically
defined groups were usually resolved as organisms tended to be either
closely related or not. The rationale of using DNA: DNA pairing as the
gold standard for the delineation of bacterial species is based on the
results of numerous investigations where good agreement was found
between DNA relatedness values and corresponding results based on
numerical phenetic and chemotaxonomic data (Nielsen et al., 1995).
The term genomic species is applied to strains which show DNA:
DNA relatedness values greater than some specified figure and thermal
denaturation values less than some specific rating. Extensive studies with
members of the family Enterobacteriaceae and related taxa led to the rec-
ommendation that genomic species should encompass strains with
approximately 70% or more DNA : DNA relatedness with a difference of
5°C or less in thermal stability (ATm; Wayne et al, 1987). A definition
based on these values has been used to clarify species relationships in
many bacterial genera. Genetically closely related bacteria that diverge in
phenotype can be designated subspecies. Members of related subspecies
show approximately 70-85% DNA: DNA relatedness and ATm values
below 3%. Values from 30% to 70% reflect a moderate degree of relation-
ship, but values become increasingly unreliable once they fall below the
30% level as they can be attributed to experimental artefacts.
Caution is sometimes needed in reconciling classifications at the species
level based on phenotypic and DNA : DNA relatedness data. The ideal sit-
uation is where an equivalence is found between genomic species and
taxospecies. DNA: DNA hybridization data have usually confirmed the
taxonomic status of phenotypically defined species, thereby showing that
the phenotypic characters used to define them were predictive. In addi-
tion, many newly described species have been proposed using the genomic
group as a criterion. However, there are several examples of taxospecies
which show marked genetic diversity. For example, in Xanthomonas,
hybridization values range from 0 to 100% between pathovars
(Hildebrandt et al., 1990) known to be indistinguishable using biochemical
tests (Van den Mooter and Swings, 1990). Similar incongruities have been
reported with Bacillus (Priest, 1993; Nielsen et al., 1995), Flavobacterium
(Ursing and Bruun, 1991) and Pseudomonas species (Rossello et al, 1991). In
the first instance, representatives of such taxospecies need to be examined
to find new phenotypic features for distinguishing between the
constituent genomic species. This strategy has been applied successfully
to unravelling the classification of organisms classified as Bacillus circulans
(Nakamura, 1984) and Bacillus sphaericus (Alexander and Priest, 1990).
Thus, strains previously known as Bacillus circulans were assigned to five
species, B. amylolyticus, B. circulans sensu stricto, B. lautus, B. pabuli and B.
validus, when conclusions drawn from DNA relatedness experiments
were reinforced by phenotypic features.
The genomic species concept may be difficult to apply where there is
extensive overlap between DNA relatedness values. This seems to be the
case with members of the genus Xanthomonas (Hildebrandt et al., 1990)
and the family Enterobacteriaceae (Gavini et al, 1989). It is possible that
chromosomal rearrangements may result in changes in phenotype with-
out influencing corresponding DNA: DNA pairing values (Egel et al.,
1991). Similarly, chromosomal rearrangement might change relatedness
values without affecting phenotypic expression (Krawiec, 1985). Such
possibilities underline the need for prudence in setting arbitrary related-
ness values to define species.
DNA : DNA pairing studies are hampered by technical difficulties and
results impaired by experimental error (Sneath, 1983; Johnson, 1991a).
Studies employing complete matrices of DNA : DNA relatedness values
are the exception rather than the rule. The usual practice is to take a few
organisms as reference strains, and to compare all of the other strains
against this restricted set. The challenge is to recover a taxonomic struc-
ture that is in accordance with the position which would have been
reached from the analysis of a complete matrix. This is not straightforward
for it has been shown (Sneath, 1983; Hartford and Sneath, 1988) that infor-
mation on the underlying taxonomic structure is necessarily lost when a
small number of reference strains are studied. It is important that refer-
ence strains are widely spaced and representative of the constituent
species. Such problems together with the diversity of available experi-
mental procedures (Figure 3.2) emphasize the need for a critical appraisal
of all DNA : DNA pairing data.
It is not always appreciated that DNA : DNA relatedness values do not
reflect the actual degree of sequence similarity at the level of primary struc-
ture. Thermal stabilities of heteroduplexes have been shown to fall by 1 to
1.5°C for every 1% of unpaired bases (Britten and Kohne, 1968). This means
that in bacteria DNA heteroduplexes will not be formed, even under non-
stringent conditions, unless the strains show DNA relatedness values of
more than 50 to 70%. It has been estimated that organisms which have 70%
or more DNA similarity will also have at least 96% DNA sequence identity
(Stackebrandt and Goebel, 1994). These authors also noted that significant
differences in the phenotype of Escherichia coli strains could be anticipated
given a 2% difference in the nucleotide base sequence of the genome. It
seems unlikely that the primary structure of most of the genes of this
Genomic DNA
Target DNA Target DNA

immobilized not - immobilized
Filter Solution
hybridization hybridization
Direct binding Competition HA method Tm method
Competitor
I I Same as
i I
DNA DH
None target DNA None None
Removal
unbound probe
. I I Wash Wash Hydroxyapatite
column
Hybridization
I i UV 260 nm
spectrophotometry
Detection
Scintillation counting
Labelled Unlabelled
Reference DNA
Figure 3.2 Schematic representation of some of the different DNA reassociation

assays used to determine DNA sequence similarity. HA, hydroxyapatite; Tm, ther-
mal melting.
organism are affected by genomic rearrangements caused by horizontal

gene transfer or by the presence of mobile elements.
Despite the problems, the advantages of DNA : DNA pairing analyses
outweigh their limitations. The method is still seen as the final arbiter for
delineating species (Wayne et al., 1987; Stackebrandt and Goebel, 1994)
and is attractive as it can be applied to all prokaryotes irrespective of their
growth requirements. In particular, DNA : DNA pairing studies provide:
• A unified concept for the delineation of bacterial species.
• Confirmation of the taxonomic integrity of phenotypically defined
species.
• Detection of taxospecies that accommodate substantial amounts of
genetic diversity.
• Recognition that species assigned to the same genus (or even different
genera) belong to a single genomic species.
DNA: DNA pairing studies can also be used to detect and identify
unknown isolates. However, classical DNA pairing techniques are too
finicky and time-consuming to apply to large numbers of strains, since
DNA must be extracted and purified from each of the test strains before
the pairing experiment (Johnson, 1991b). However, recent studies show
that blotting and lysis of whole organisms on membrane filters followed
by pairing with labelled DNA from reference strains (chromosomal
probes) provides a rapid means of identifying isolates, notably those of
medical importance (Schleifer et al, 1993).
3.3.3 Nucleotide sequence analysis

It is well known that the sequences of the genes that code for ribosomal
RNA are relatively highly conserved and hence can be used to determine
taxonomic relationships between species which show little DNA related-
ness. It is becoming increasingly popular to propose new bacterial species
using data generated from 16S rRNA sequencing studies (Table 3.3). Two
basic assumptions underline this approach, namely that lateral gene trans-
fer has not occurred between 16S rRNA genes, and that the amount of
evolution or dissimilarity between 16S rRNA sequences of a given pair of
organisms is representative of the variation shown by the corresponding
genomes. The good congruence found between phylogenies based on 16S
rRNA sequences and those derived from alternative molecules, such as
23S rRNA (Ludwig et al, 1993a), ATPase subunits (Ludwig et al, 1993b),
elongation factors (Ludwig et al, 1993b) and RNA polymerases (Zillig et al,
1989), give substance to this latter point. It also seems likely that lateral
gene transfer between 16S rRNA genes is rare as this gene is responsible
for the maintenance of functional and tertiary structural consistency.
However, the possibility of horizontal gene transfer in 16S rDNA should
not be overlooked (Sneath, 1993).
Fox et al (1992) pointed out that 16S rRNA molecules from members of
closely related species may be so conserved that they cannot be used to
differentiate between strains at the species level. This important observa-
tion means that strains of related species with identical, or almost identi-
cal, 16S rRNA nucleotide sequences may belong to different genomic
species. This is the case with species of Aeromonas (Martinez-Murcia et al,
100
98
r.
JB
1
'«
<
Z
DC
W
CO
94
92
10 20 30 40 50 60 70 80 90 100
DNA relatedness (%)
Figure 3.3 Comparison of 16S rRNA and DNA relatedness values of mycolic acid-
containing actinomycetes. (Based on data taken from molecular systematic stud-
ies on mycolic acid-containing organisms; Chun, 1995.)
Table 3.3 New species described within existing genera or recommended as

synonyms in the International Journal of Systematic Bacteriology
Techniques employed
DNA:DNA Serological 16S rRNA Others
Year hybridization tests sequencing
1987* 60 10 0 30
1993* 75 8 14 3
1995 68 7 90 3
Data from Stackebrandt and Goebel (1994).
1992), Bacillus (Ash et al, 1991) and Legionella (Fry at al, 1991). It is clear
from these observations that the resolution of DNA hybridization is high-
er than that of 16S rRNA sequence analysis and that DNA : DNA pairing
remains the method of choice for measuring the degree of relatedness
between closely related organisms. Nevertheless, even in this context 16S
rRNA sequence data can be used to select appropriate reference strains for
the more exacting DNA: DNA pairing studies, thereby reducing the
number of reference strains which need to be examined. The terms rRNA
species complex and rRNA superspecies have been proposed for organ-
isms which have virtually identical 16S rRNA sequences but can be distin-
guished using DNA : DNA relatedness data (Fox et al., 1992).
The correlation between 16S rRNA sequence and DNA relatedness data
is not linear (Figure 3.3) though rRNA similarity values below 97% invari-
ably correspond to DNA relatedness values below 60%. Similar findings
were reported by Stackebrandt and Goebel (1994) who argued that
genomic species sensu Wayne et al. (1987) usually have more than 97%
sequence identity. This cut-off point is plausible given the results shown
in Figure 3.4.
16S rRNA sequencing analyses are easier and more cost-effective than
DNA hybridization studies due to developments in molecular biology,
notably the use of the polymerase chain reaction (PCR) and the avail-
ability of automatic DNA sequencers (Figure 3.5). It is well known that
the 3% or 45 nucleotide sequence differences that can be used to distin-
guish most species are not evenly scattered along the primary structure
of the 16S rRNA macromolecule but tend to be concentrated in hyper-
variable regions. There is evidence that the hypervariable regions can be
taxon-specific (Stackebrandt and Goebel, 1994). It is clear, therefore, that
only complete 16S rRNA sequences allow reliable comparisons of novel
organisms with available databases containing complete or almost com-
plete nucleotide sequences (Canhos et al., 1993). Information on rRNA
sequences can be accessed through the internet (http://www.bdt.org.br/
structure/molecular.html).
The taxonomic relationships of both potentially novel and poorly mis-
classified organisms can readily be determined by comparing their 16S
rRNA sequences with corresponding results held in databases. It is, for
example, evident from the example (Figure 3.6) that the unknown actino-
mycete isolated from activated sludge belongs to the genus Tsukamurella
and that an organism until recently known as Nocardia amarae forms a dis-
tinct species in the genus Gordona. It is also clear that a new taxonomic
niche is needed for actinomycetes classified as Nocardia pinensis (Chun et
al, 1996).
Evolutionary relationships between bacteria need to be interpreted
with care as estimates of phylogeny are based on relatively simple
assumptions when considered against the complexities of evolutionary
processes. All methods of phylogenetic inference are based on certain
assumptions that may be violated by the data to a greater or lesser extent
(Swoffold and Olsen, 1990; Hillis et al., 1993). Such questions were raised
by O'Donnell et al. (1993) who also pointed out that potential problems
in nucleotide sequence data include alignment artefacts, non-indepen-
dence of sites, inequalities in base substitution frequencies between
sequences, and lineage-dependent inequalities in rates of change.
100
200 400 600 800 10001200 1400 1600

Nucleotide position
Figure 3.4 Nucleotide sequence variation in different regions of the 16S rRNA
molecule. Data points correspond to the average nucleotide sequence variation
calculated from aligned 16S rRNA sequences. (From Chun, 1995.)
Genomic DMA 50to100ngDNA

I
PCR 30 cycles
Check the size of approximately 1.4kb

amplified fragments
Purification of
16SrDNA fragment
Cloning and Dir ect

plasmid sequencing of
preparation PCRpr oducts
Automated or
manual sequencing
Figure 3.5 Schematic representations of 16S rRNA sequencing methods.

————— Turicella otitidis

-Corynebacterium cystitidis
• Corynebacterium bovis
100 Corynebacterium glutamicum
— Corynebacterium variabilis
75 • Corynebacterium amycolatum
- Corynebacterium xerosis
Dietzia maris
74 Tsukamurella paurometabola
100 Tsukamurella paurometabola M334
Tsukamurella paurometabola N663
Activated sludge isolate N1171
r Gordona aichiensis
"— Gordona sputi
Gordona bronchialis
100 Gordona rubropertincta
82 Gordona terrae
Gordona amarae
75
Nocardia pinensis
31 Mycobacterium smegmatis
Mycobacterium chlorophenolicum
Mycobacterium fortuitum
100
Mycobacterium intermedium
Mycobacterium leprae
Mycobacterium tuberculosis
72 Mycobacterium simiae
Rhodococcus rhodnii
——— Rhodococcus ruber
90 — Rhodococcus rhodochrous
75
Rhodococcus coprophilus
——— Rhodococcus fascians
'Rhodococcus (Tsukamurella) wratislaviensis"
Rhodococcus opacus
24
Rhodococcus erythroplis
Rhodococcus globerulus
Rhodococcus marinonascens
63 Rhodococcus sp. DSM 43943
"Rhodococcus (Nocardia) corynebacteroides"
• Rhodococcus equi
— Norcardia asteroides
—— Norcardia brevicatena
—— Norcardia carnea
Norcardia nova
0.01 Norcardia otitidiscaviarum
— Norcardia seriolae
Norcardia vaccinii
——— Norcardia brasiliensis
— Norcardia transvalensis
Norcardia farcinica
Figure 3.6 An unrooted phylogenetic tree showing relationships between isolate

Nil71 and other mycolic acid-containing actinomycetes. The tree was construct-
ed by using the Jukes and Cantor distance (1969) and neighbour-joining methods
(Saitou and Nei, 1987). The numbers at the nodes indicate the level of bootstrap
support on 1000 resamplings. The scale bar indicates 0.01 substitutions per
nucleotide position.
Problems such as these are compounded by bacterial taxonomists who
regard the various data handling techniques as 'blackboxes' integrated
into computer software. The literature is replete with examples of inad-
equate practices, as witnessed by the following examples taken from a
single issue of the International Journal of Systematic Bacteriology:
• 'The ODEN program package was used to align the sequences, and
phylogenetic distances were calculated by using both the unweighted
pair group method and neighbor-joining method' (Ezaki et al., 1994:
130). In this study, the authors did not say how evolutionary distances
were calculated.
• 'Evolutionary distances were determined by using the neighbor-
joining method' (Briglia et al., 1994: 494). The method used in this study
was designed for constructing trees from distance matrices not for gen-
erating distances!
• 'The dendrogram was constructed by using the software program
PILEUP obtained from the Genetic Computer Group Inc'. (Robertson et
al., 1994: 836). The authors did not give any information on how the
data were analysed.
• 'An unrooted phylogenetic tree was produced by using the DNADIST
and FITCH programs in the PHYLIP package' (Dupuy et al., 1994: 461).
It is not clear which distance method was used as the DNADIST pro-
gram contains four different distance methods.
• A common omission was a failure to mention which distance models
were used (e.g. Cai and Collins, 1994; Collins et al., 1994; Jagoueix et al.,
1994).
There are, of course, many examples of good practice in the analysis of
molecular sequence data (Chun and Goodfellow, 1995; Chun et al., 1996;
Kim et al., 1996). Despite the blind applications of computer packages, there
is no doubt that the advantages of using 16S rRNA sequencing for helping
to delineate new species far outweigh the limitations. However, there can
be no bacterial species definition based solely on sequence similarity of
rRNAs or their genes as absolute values for delineating species cannot be set
because of different rates of sequence divergence. Nevertheless, good
agreement is generally found between relationships derived from 16S
rRNA sequence data and those based on DNA: DNA pairing data, for
example, in species of the genus Serratia (Dauga et al., 1990).
The advantages of using 16S rRNA sequencing for the recognition of
novel species far outweigh the deficiencies (Table 3.4). Further, the abil-
ity to obtain rRNA sequence data from difficult to culture and uncul-
tured bacteria will help in the exploration of the huge diversity of
prokaryotic species which await discovery (see Embley and
Stackebrandt, 1997; Chapter 4). Nucleic acid sequence data can also be
used to design probes for in situ hybridizations (Schleifer et al., 1993) and
thereby facilitate the development of appropriate selective isolation
strategies by showing whether environmental samples contain members
of target species.
3.3.4 Chemotaxonomy
Chemical data derived from the analysis of cell components can be used
to classify bacteria at different taxonomic ranks according to the pattern
of distribution of the different compounds within and between members
of different taxa. Chemotaxonomic analyses of chemical macromole-
cules, particularly amino acids and peptides (e.g. from peptidoglycan
and pseudomurein), lipids (lipopolysaccharides), polysaccharides and
related polymers (e.g. methanochondroitin, wall sugars), proteins (e.g.
bacteriochlorophyll, whole-organism protein patterns), enzymes (e.g.
hydrolases, lyases), and other complex polymeric compounds, such as
isoprenoid quinones and sterolsr all provide valuable data for the
chemotaxonomic cornucopia (Goodfellow and O'Donnell, 1994). The
base composition of DNA is also a chemical property sensu stricto but is
usually considered as a molecular feature. Chemical fingerprints of tax-
onomic value can be obtained using analytical chemical techniques,
notably Curie-point pyrolysis mass spectrometry (Goodfellow et al,
1994a). Other promising approaches which provide valuable data for
delineating species include analyses of cellular fatty acids (Stead et al.,
1992; Vauterin et al., 1996) and whole-organism proteins (Vauterin et al.,
1993; Verissimo et al., 1996), and the elucidation of enzyme profiles based
on chromogenic and fluorogenic substrates (Manafi et al., 1991).
Developments in molecular systematics should not be seen as a threat
to chemosystematics as the two approaches are complementary.
Table 3.4 The advantages and limitations of using 16S rRNA sequence data for
the circumscription of bacterial species
Advantages
Full sequence analysis has become rapid and inexpensive
Provision of high-quality databases
More objective definition of species
Species presented within a supra-generic framework
Nucleic acid probes for identification
Limitations
Only near-complete sequences allow reliable comparisons with other
near-complete sequences from databases
Resolution limited when closely related organisms are compared
Strains belonging to different species may have identical sequences
Taxonomic relationships are affected by the choice of statistical methods
Phylogenetic data provide a hierarchic framework of relationships
among bacterial species but do not give reliable information for the
delineation of taxa above the species level. In contrast, chemical markers
are unevenly distributed across taxa but rarely give information on the
hierarchic rank of taxa. It is very encouraging that good congruence
exists between the distribution of chemical markers and the relative
positions of species in phylogenetic trees (Goodfellow and O'Donnell,
1994; Chun et al., 1996). Chemical data are not only employed to evalu-
ate existing phylogenies but can also be used to adjudicate between con-
flicting phylogenetic trees. The phylogenetic positions of activated
sludge isolate N1171, Gordona amarae and Nocardia pinensis (Figure 3.6)
are supported by chemotaxonomic evidence (Blackall et al., 1989;
Goodfellow et al, 1994b; Chun, 1995).
3.3.5 Towards a new operational species concept

It is evident from the previous section that none of the individual
approaches used to circumscribe bacterial species is without problems.
Molecular systematic techniques, especially those based on 16S rRNA
sequencing and DNA hybridization, will continue to play an important
role in the revision of established classifications and in the delineation of
new species. Nucleic acid sequence data can be used to generate a frame-
work of the perceived evolutionary relationships between representatives
of all bacterial taxa, but this phylogeny needs to be evaluated and refined
using independent taxonomic data as phylogenetic relationships between
organisms can be distorted by differences in evolutionary rates, as well as
by technical and statistical problems. Further, simple phenotypic proper-
ties are still at a premium for the identification of unknown strains to
validly described species.
A more reliable and comprehensive approach to the delineation of
species is emerging based on the integrated use of genotypic and pheno-
typic data. This approach, known as polyphasic taxonomy, was intro-
duced by Colwell (1970) to signify successive or simultaneous taxonomic
studies on groups of organisms using an array of methods designed to
provide genotypic and phenotypic data. Polyphasic taxonomic studies
can by their nature be expected to yield well-defined species, a stable
nomenclature and improved species definitions.
Polyphasic taxonomy remains in its infancy for a variety of reasons. All-
embracing approaches to the circumscription of species only recently
became practical given the availability of rapid data acquisition systems and
improved data handling procedures. A lack of suitably trained taxonomists
and a reluctance of funding agencies to support taxonomic research seen to
be 'non-molecular' have compounded the situation, despite significant
developments in chemo- and numerical taxonomy (Goodfellow and
O'Donnell, 1993,1994). It is, therefore, far from surprising that only limited
phenetic data are available to evaluate even the major phylogenetic groups
of prokaryotes (Murray et al., 1990). Nevertheless, it is encouraging that
most descriptions of new cultivable prokaryotes in the 1995 issues of the
International Journal of Systematic Bacteriology are based on a judicious selec-
tion of genotypic and phenotypic criteria.
The subgeneric classification of several bacterial groups, notably aero-
bic, endospore-forming bacilli (White et al., 1993), bartonellae (Birtles et al.,
1995), flavobacteria (Bernardet et al., 1996), microtetrasporae
(Kroppenstedt et al., 1990), mycobacteria (Wayne et al., 1996), oceanospir-
illa (Pot et al, 1989) and xanthomonads (Vauterin et al, 1990, 1995), have
been clarified using the polyphasic taxonomic approach. Similarly, several
new species, for instance, Amycolatopsis methanolica (De Boer et al, 1990),
Burkholderia vietnamiensis (Gillis et al, 1995), Nocardia pseudobrasiliensis
(Ruimy et al, 1996), Paenibacillus validus (Heyndrickx et al, 1995), Rhizobium
mediterraneum (Nour et al, 1995), Rhodococcus percolatus (Briglia et al, 1996)
and Tsukamurella inchonensis (Yassin et al, 1995) have been described using
a combination of genotypic and phenotypic data. Such studies, while
admirable, have not addressed the question of exactly what type of pro-
cedures and analyses constitute a polyphasic taxonomy. At present,
polyphasic taxonomic studies tend to reflect the particular expertise of the
laboratory or of the investigators.
Ideally, new species should be described using data derived from an
analysis of a representative set of strains. Sneath (1977) visualized a
species as a cluster of a large number of strains in a character space. As an
approximation, this space can be regarded as roughly spherical, and can
thereby be described, using sampling theory, by its centre, while its radius
defines an envelope that encompasses most of the strains. Sneath consid-
ered that about 25 strains were needed to define accurately the centre and
radius of a cluster (taxospecies) but went on to suggest that never less than
10 strains be used for this purpose. In practice, most new species are still
described on the basis of a few strains, not infrequently on a single strain!
The proposition that nomenclature should reflect genomic relationships
(Wayne et al, 1987) and that all pre-conceived notions be re-examined
within this context (Murray et al, 1990) has been widely accepted in prac-
tice. There are no problems in naming new species given congruence
between the various data subsets, notably genotypic and phenotypic data.
However, it has already been pointed out that some taxospecies encom-
pass two or more genomic species. Wayne et al (1987) recommended that
distinct genomic species which cannot be differentiated from other genom-
ic species using known phenotypic properties should not be formally
named until they can be differentiated by suitable phenotypic markers.
This proposition, while laudable, has tended to leave taxospecies which
accommodate more than one genomic species in a state of limbo. Such tax-
ospecies can also be considered as nomenspecies.
It has been suggested that genomic species accommodated in nomen-
species should be referred to as genomovars (Rossello et al., 1991; Ursing
et al., 1995). This concept is useful as it indicates that genomic species are
an integral part of a nomenspecies and hence should not be overlooked in
subsequent taxonomic work. Once phenotypically delineated,
genomovars can be given formal names. Rossello et al. (1991) also suggest-
ed that genomovars encompassed in nomenspecies should be numbered;
the first genomovar being the genomic species that contains the type
strain of the nomenspecies. Each of the following genomovars should
have a designated reference strain, which should be deposited in a service
culture collection.
3.4 MINIMAL AND NOMENCLATURAL STANDARDS FOR THE

DESCRIPTION OF NEW SPECIES
Since 1 January 1980, priority of bacterial names has been based on the
Approved Lists of Bacterial Names (Skerman et al., 1980). Names not includ-
ed in such lists lost standing in bacterial nomenclature, though old names
are available for reuse individually, provided that the provisions for doing
so are met. Following the introduction of the new starting date for bacte-
rial nomenclature, valid publication of new names and new combinations
can only be achieved in the International Journal of Systematic Bacteriology,
although they may be effectively published elsewhere and then validated
by announcement in the validation lists published periodically in the jour-
nal. To date, 56 validation lists have been published. All validated names
can be accessed through an electronic database on the internet
(http://www.bdt.org.br/cgi-bin/bdtnet/bacterianame). Before publication
of the name of a new cultivable species, a culture of the type strain should
be deposited in at least one of the permanently established culture collec-
tions. It is also a requirement that the accession number designated to the
strain by the culture collection is quoted in the published description.
One of the responsibilities of the taxonomic subcommittees of the
International Committee on Systematic Bacteriology of the International
Union of Microbiological Societies is to recommend minimal standards for
the publication of new species (Sneath, 1992). This practice is intended to
prevent the literature becoming cluttered with inadequately described
species, as was the case before the introduction of the approved lists
(Skerman et al., 1980). Such standards need to include tests for the estab-
lishment of generic identity and for the delineation of species. In practice,
the properties recommended for the description of species in different
genera vary as they reflect the biological features of the organisms, but
Conclusions 49
they should follow the tenets of polyphasic taxonomy (Figure 3.7).
Guidelines for the minimal requirements necessary for the description of
species are available (Sneath, 1977; Triiper and Schleifer, 1992). Minimal
standards have been recommended recently for the description of new
species of campylobacters (Ursing et aL, 1994), methanogenic bacteria
(Boone and Whitman, 1988), mollicutes (ICSB Subcommittee on the
Taxonomy of Mollicutes, 1995), root and stem-nodulating bacteria
(Graham et al., 1991) and slowly growing mycobacteria (Levy-Frebault and
Portaels, 1992).
3.5 CONCLUSIONS
Bacterial systematics as a core discipline is practised by few, but the appli-
cations of the subject are important to most - if not all - bacteriologists. It
is the implementation of taxonomic concepts and practices which give rise
to identification and typing systems, procedures for quality control and
risk assessment, protocols for the analysis and characterization of bio-
diversity, hypotheses about the evolution of prokaryotes, and improved
procedures for the selective isolation and use of microorganisms in
Recognition of a new pattern of properties
r
Genotypic properties
\
Phenotypic properties
V V
Mol. % G+C Morphology and staining
DNA:DNA relatedness Chemical markers
Nucleic acid sequencing Enzyme tests

Nucleic acid fingerprinting Protein electrophoresis
Serology
Numerical taxonomy
Figure 3.7 Methods relevant to the generation of genotypic and phenotypic prop-
erties for setting minimal standards for the description of new species.
biotechnological processes. Consequently, the nature of the bacterial
species is not simply a matter for philosophical discourse, but is one of real
practical significance.
Recent advances in bacterial systematics have promoted a more unified
approach to the delineation of bacterial species. The concept of the
'polyphasic species' has distinct advantages over traditional more descrip-
tive species concepts, especially since it can be expected to yield well-
described species, a stable nomenclature and better identification systems.
This approach to the circumscription of species assumes a population that
is reproducing asexually is adapted to particular microhabitats, and is
maintained as a relatively stable entity by natural selection. The ever-
increasing availability of new methods for the generation of genotypic
and phenotypic data, associated with new software tools, will help expe-
dite polyphasic taxonomic studies. In practice, bacteriologists have a more
standard set of comparative taxonomic methods than botanists and zool-
ogists, but cross-checking of findings is still critical to sound work.
The polyphasic species concept is universally applicable and will
become ever easier to apply, though the details of the approach need to
be tailored to take into account the differing behavioural properties of
members of taxonomically diverse genera. The concept does not address
the question of the origin of bacterial species, that is, an understanding of
the evolutionary processes that generate taxonomic diversity. This is
clearly a major omission but is also a fertile area for collaboration between
microbial population geneticists and bacterial systematists. However, will
a theoretically sound species concept prove to be as practical as the pre-
sent operational species concept?
Acknowledgements
Thanks are due to Professor P.H.A. Sneath and Professor F.G. Priest for
critically reviewing the manuscript.
3.6 REFERENCES
Alexander, B. and Priest, F.G. (1990) Numerical classification and identification of
Bacillus sphaericus including some strains pathogenic to mosquito larvae. Journal
of General Microbiology, 136, 367-76.
Ash, C, Farrow, J.A.E., Wallbanks, S. and Collins, M.D. (1991) Phylogenetic
heterogeneity of the genus Bacillus revealed by comparative analysis of small
subunit-ribosomal RNA sequences. Letters in Applied Microbiology, 13,202-6.
Bernardet, J.-F., Segers P., Vancanneyt, M., Berthe, F., Kersters, K. and Vandamme,
P. (1996) Cutting a Gordonian knot: Emended classification and description of
the genus Flavobacterium, emended description of the family Flavobacteriaceae
and proposal of Flavobacterium hydatis nom. nov. (Basonym, Cytophaga aquatilis
Strohl and Tait, 1978). International Journal of Systematic Bacteriology, 46,128-48.
References 51
Dirties, R.J., Harrison, T.G., Saunders, N.A. and Molyneux, D.H. (1995) Proposal to
unify the genera Grahamella and Bartonella, with descriptions of Bartonella talpae
comb, nov., Bartonella peromysci comb. nov. and three new species, Bartonella
grahamii sp. nov., Bartonella taylorii sp. nov., and Bartonella doshiae sp. nov.
International Journal of Systematic Bacteriology, 45,1-8.
Blackall, L.L., Parlett, J.H., Hayward, A.C., Minnikin, D.E., Greenfield, P.P. and
Habers, A.E. (1989) Nocardia pinensis sp. nov., an actinomycete found in activat-
ed sludge foams in Australia. Journal of General Microbiology, 135,1547-58.
Bochner, B.R. (1989) Sleuthing out bacterial identities. Nature, 339,157-8.
Boone, D.R. and Whitman, W.B. (1988) Proposal of minimal standards for describ-
ing new taxa of methanogenic bacteria. International Journal of Systematic
Bacteriology, 38, 212-19.
Briglia, M., Eggen, R.I.L., van Elsas, DJ. and De Vos, W.M. (1994) Phylogenetic evi-
dence for transfer of pentachlorophenol-mineralising Rhodococcus chlorophenoli-
cus PCP-1 to the genus Mycobacterium. International Journal of Systematic
Briglia, M., Rainey, F.A., Stackebrandt, E., Schraa, G. and Salkinoja-Salonen, M.S.
(1996) Rhodococcus percolatus sp. nov., a bacterium degrading 2,4,6-trichlorophe-
nol. International Journal of Systematic Bacteriology, 46, 23-30.
Britten, R.J. and Kohne, D.E. (1968) Repeated sequences in DNA. Science, 161,
529-40.
Bryant, T.N. (1993) A review of probabilistic identification matrices. A compilation
of probabilistic bacterial identification matrices. Binary, 5,207-10.
Buchanan, R.E. (1955) Taxonomy. Annual Review of Microbiology, 9,1-20.
Bull, A.T., Goodfellow, M. and Slater, J.H. (1992) Biodiversity as a source of inno-
vation in biotechnology. Annual Review of Microbiology, 46, 219-57.
Cai, J. and Collins, M.D. (1994) Phylogenetic analysis of species of the meso-
diaminopimelic acid-containing genera Brevibacterium and Dermabacter.
International Journal of Systematic Bacteriology, 44, 583-5.
Canhos, V.P., Manfio, G.P. and Blaine, L.D. (1993) Software tools and databases for
bacterial systematics and their dissemination via global networks. Antoine van
Leeuwenhoek, 64,205-29.
Chun, J. (1995) Computer Assisted Classification and Identification of Actinomycetes.
Unpublished PhD Thesis, University of Newcastle upon Tyne, UK.
Chun, J. and Goodfellow, M. (1995) A phylogenetic analysis of the genus Nocardia
with 16S rRNA gene sequences. International Journal of Systematic Bacteriology,
45,240-5.
Chun, J., Kang, S.O., Hah, Y.C. and Goodfellow, M. (1996) Phylogeny of mycolic
acid-containing actinomycetes. Journal of Industrial Microbiology (in press).
Cohan, P.M. (1994) Genetic exchange and evolutionary divergence in prokaryotes.
Trends in Ecology and Evolution, 9,175-80.
Collins, M.D., Lawson, P.A., Willems, A., Cordoba, J.J., Fernandez-Garayzabal, J.,
Gercia, P., Cai, J., Hippe, H. and Farrow, J.A.E. (1994) The phylogeny of the
genus Clostridium: proposal of five new genera and eleven new species combi-
nations. International Journal of Systematic Bacteriology, 44, 812-26.
Colwell, R.R. (1970) Polyphasic taxonomy of bacteria, in Culture Collections of
Microorganisms (eds H. lizuka and T. Hasegawa), University of Tokyo Press,
Tokyo, pp. 421-36.
Cowan, S.T. (1955) The principles of microbial classification. Introduction: the phil-
osophy of classification. Journal of General Microbiology, 12, 314-19.
Cowan, S.T. (1959) Bacterial classification - problems and developments, in
Microbiology Yesterday and Today (ed. V. Bryson), Institute of Microbiology,
Rutgers, The State University, New Brunswick, pp. 54-79.
Cowan, S.T. (1962) The microbial species - a macromyth? in Microbial Classification
(eds G.S. Ainsworth and P.H.A. Sneath), Cambridge University Press,
Cambridge, pp. 433-55.
Cowan, S.T. (1968) A Dictionary of Microbial Taxonomic Usage. Oliver and Boyd,
Edinburgh.
Dauga, C, Grimont, F. and Grimont, P.A.D. (1990) Nucleotide sequences of 16S
rRNA from 10 Serratia species. Research in Microbiology, 141,1139-49.
De Boer, L., Dijkhuizen, L., Grobben, G., Goodfellow, M., Stackebrandt, E., Parlett,
J.H., Whitehead, D. and Witt, D. (1990) Amycolatopsis methanolica sp. nov., a fac-
ultatively methylotrophic actinomycete. International Journal of Systematic
Bacteriology, 40,194-204.
Dupuy, N., Willems, A., Pot, B., Dewettinck, D., Vandenbruaene, L, Maestrojuan,
G., Dreyfus, B., Kersters, K., Collins, M.D. and Gillis, M. (1994) Phenotypic and
genotypic characterization of bradyrhizobia nodulating the leguminous tree
Acacia albida. International Journal of Systematic Bacteriology, 44,461-73.
Edwards, C. (ed.) (1993) Monitoring Genetically Manipulated Microorganisms in the
Environment. John Wiley and Sons, Chichester.
Egel, D.S., Graham, J.H. and Stall, R.E. (1991) Genomic relatedness of Xanthomonas
campestris strains causing diseases of citrus. Applied and Environmental
Embley, T.M. and Stackebrandt, E. (1997) Species in practice: exploring uncultured
prokaryote diversity in natural samples, in Species: the Units of Biodiversity (eds
M.F. Claridge, H.A. Dawah and M.R. Wilson), Chapman & Hall, London, pp.
61-81.
Ezaki, T., Li, N., Hashimoto, Y., Miura, H. and Yamamoto, H. (1994) 16S ribosomal
DNA sequences of anaerobic cocci and proposal of Ruminococcus hansenii comb,
nov. and Ruminococcus productus comb. nov. International Journal of Systematic
Felsenstein, J. (1993) PHYLIP (phylogenetic inference package), version 3.5c
Department of Genetics, University of Washington, Seattle.
Fry, N.K., Saunders, N.A., Warwick, S. and Embley, T.M. (1991) The use of 16S
ribosomal RNA analyses to investigate the phylogeny of the family
Legionellaceae. Journal of General Microbiology, 137,1215-22.
Fox, G.E., Wisotzkey, J.D. and Jurtshuk, P., Jr (1992) How close is close: 16S rRNA
sequence identity may not be sufficient to guarantee species identity.
Gavini, F., Mergaert, J., Beji, A., Mielcarek, C., Izard, D., Kersters, K. and De Ley, J.
(1989) Transfer of Enterobacter agglomerans (Beijerinck 1888) Ewing and Fife 1972
to Pantoea gen. nov. as Pantoea agglomerans comb. nov. and description of
Pantoea dispersa. International Journal of Systematic Bacteriology, 39,337-45.
Gillis, M., Van, V.T., Bardin, R., Gvor, M., Hebbar, P., Willemes, A., Segers, P.,
Kersters, K., Heulin, T. and Fernandez, M.P. (1995) Polyphasic taxonomy in the
genus Burkholderia leading to an emended description of the genus and propo-
References 53
sition of Burkholderia vietnamiensis sp. nov. for nitrogen-fixing isolates from rice
in Vietnam. International Journal of Systematic Bacteriology, 45,274-89.
Goodfellow, M. and Dickinson, C.H. (1985) Delineation and description of micro-
bial populations using numerical methods, in Computer-Assisted Bacterial
Systematics (eds M. Goodfellow, D. Jones and F.G. Priest), Academic Press,
Orlando, pp. 165-225.
Goodfellow, M. and Minnikin, D.E. (eds) (1985) Chemical Methods in Bacterial
Systematics. Academic Press, London.
Goodfellow, M. and O'Donnell, A.G. (eds) (1993) Handbook of New Bacterial
Goodfellow, M. and O'Donnell, A.G. (1994) Chemical Methods in Prokaryotic
Systematics. John Wiley and Sons, Chichester.
Goodfellow, M., Jones, D. and Priest, F.G. (eds) (1985) Computer-Assisted Bacterial
Goodfellow, M., Chun, J., Atalan, E. and Sanglier, J.-J. (1994a) Curie-point pyro-
lysis mass spectrometry and its practical application to bacterial Systematics, in
Bacterial Diversity and Systematics (eds F.G. Priest, A. Ramos-Cormenzana and
BJ. Tindall), Plenum Press, New York, pp. 87-104.
Goodfellow, M., Chun, J., Stubbs, S. and Toboli, A.S. (1994b) Transfer of Nocardia
amarae Lechevalier and Lechevalier 1974 to the genus Gordona as Gordona
amarae comb. nov. Letters in Applied Microbiology, 19,401-5.
Gordon, R.E. (1967) The taxonomy of soil bacteria, in The Ecology of Soil Bacteria (eds
T.R.G. Gray and D. Parkinson), Liverpool University Press, Liverpool, pp.
293-321.
Gordon, R.E. (1978) A species definition. International Journal of Systematic
Graham, P.H., Sadowsky, M.J., Keyser, H.H., Barnet, Y.M., Bradley, R.S., Cooper,
J.E., De Ley, D.J., Jarvis, B.D.W., Roslycky, E.B., Strijdom, B.W. and Young,
J.P.W. (1991) Proposed minimal standards for the description of new genera
and species of root-and-stem-nodulating bacteria. International Journal of
Systematic Bacteriology, 41,582-7.
Hartford, T. and Sneath, P.H.A. (1988) Distortion of taxonomic structure from
DNA relationships due to different choice of reference strains. Systematic and
Applied Microbiology, 10,241-50.
Harwood, C.R. (1993) Plasmids, transposons and gene flux, in Handbook of New
Bacterial Systematics (eds M. Goodfellow and A.G. O'Donnell), Academic Press,
London, pp. 115-50.
Hawksworth, D.L. (ed.) (1991) The Biodiversity of Microorganisms and Invertebrates:
Its Role in Sustainable Agriculture. Redwood Press, Melksham.
Heyndrickx, M., Vandemeulebroeke, K., Scheldeman, P., Hoste, B., Kersters, K.,
De Vos, P., Logan, N.A., Aziz, A.M., Ali, M. and Berkeley, R.C.W. (1995)
Paenibacillus (formerly Bacillus) gordonae (Pichinoty et al. 1986) Ash et al. 1994 is a
later subjective synonym of Paenibacillus (formerly Bacillus) validus (Nakamura
1984) Ash et al. 1994: Amended description of P. validus. International Journal of
Hildebrandt, D.C., Palleroni, N.J. and Schroth, M.N. (1990) Deoxyribonucleic acid
relatedness of 24 xanthomonad strains representing 23 Xanthomonas campestris
pathovars and Xanthomonas fragariae. Journal of Applied Bacteriology, 68,263-9.
Hillis, D.M., Allard, W. and Miyamoto, M.M. (1993) Analysis of DNA sequence
data: phylogenetic inference. Methods in Enzymology, 224,456-87.
Holloway, B.W. (1993) Genetics for all bacteria. Annual Review of Microbiology, 47,
659-84.
ICSB Subcommittee on the Taxonomy of Mollicutes. (1995) Revised minimum
standards for description of new species of the class Mollicutes (Division
Tenericutes). International Journal of Systematic Bacteriology, 45, 605-12.
Jagoueix, S., Bove, J.M. and Garnier, M. (1994) The phloem-limited bacteria of
greening disease of citrus is a member of the a subdivision of the Proteobacteria.
Johnson, J.L. (1991a) DNA reassociation experiments, in Nucleic Acid Techniques in
Bacterial Systematics (eds M. Goodfellow and A.G. O'Donnell), John Wiley and
Sons, Chichester, pp. 21-44.
Johnson, J.L. (1991b) Isolation and purification of nucleic acids, in Nucleic Acid
Techniques in Bacterial Systematics (eds M. Goodfellow and A.G. O'Donnell),
John Wiley and Sons, Chichester, pp. 1-19.
Jones, D. (1989) Bacterial classification. IV: Genetic methods, in Bergey's Manual of
Systematic Bacteriology, Volume 4 (eds S.T. Williams, M.E. Sharpe and J.G. Holt),
Williams and Wilkins, Baltimore, pp. 2310-12.
Jukes, T.H. and Cantor, C.R. (1969) Evolution of protein molecules, in Mammalian
Protein Metabolism (ed. H. N. Munro), Academic Press, New York, pp. 21-132.
Kampfer, P. and Kroppenstedt, R.M. (1991) Probabilistic identification of strepto-
mycetes using miniaturised physiological tests. Journal of General Microbiology,
137,1893-902.
Kampfer, P., Kroppenstedt, R.M. and Dott, W. (1991) A numerical classification of
the genera Streptomyces and Streptoverticillium using miniaturized physiological
tests. Journal of General Microbiology, 137,1831-91.
Kirn, D., Chun, J., Sahin, N., Hah, Y.C. and Goodfellow, M. (1996) Analysis of ther-
mophilic clades within the genus Streptomyces by 16S rDNA sequence compar-
isons. International Journal of Systematic Bacteriology, 46, 581-7.
Klinger, J.M., Stowe, R.P. Obenhuber, D.C., Grover, T.O., Mishra, S.K. and Pierson,
D.L. (1992) Evaluation of the Biolog automated microbial identification system.
Applied and Environmental Microbiology, 58, 2089-92.
Krawiec, S. (1985) Concept of a bacterial species. International Journal of Systematic
Krieg, N.R. (1988) Bacterial classification: an overview. Canadian Journal of
Microbiology, 34, 536^0.
Krieg, N.R. and Holt, J.G. (eds) (1984) Bergey's Manual of Systematic Bacteriology,
Volume I, Williams and Wilkins, Baltimore.
Kroppenstedt, R.M., Stackebrandt, E. and Goodfellow, M. (1990) Taxonomic revi-
sion of the actinomycete genera Actinomadura and Microtetraspora. Systematic
and Applied Microbiology, 13,148-60.
Kusunoki, S. and Ezaki, T. (1992) Proposal of Mycobacterium peregrinum sp. nov.,
and elevation of Mycobacterium chelonae subsp. abscessus (Kubica et al.) to species
status: Mycobacterium abscessus comb. nov. International Journal of Systematic
Levy-Frebault, V.V. and Portaels, F. (1992) Proposed minimal standards for the
genus Mycobacterium and for description of new slowly growing Mycobacterium
species. International Journal of Systematic Bacteriology, 42, 315-23.
References 55
Ludwig, W., Kirchhof, G., Klugbauer, N., Weizenegger, M., Betzyl, D., Ehrmann,
M., Hertel, C, Jilg, S., Tatzel, R., Zitzelsberger, H., Liebl, S., Hochberger, M.,
Shah, J., Lane, D., Wallnofer, P.R. and Schliefer, K.H. (1993a) Complete 23S ribo-
somal RNA sequences of Gram-positive bacteria with a low G+C content.
Systematic and Applied Microbiology, 15,487-501.
Ludwig, W., Neumaier, J., Klugbauer, N., Brockmann, E., Roller, C., Jilg, S., Reetz,
K., Schachtner, I., Ludvigsen, A., Bachleitner, M., Fischer, U. and Schleifer, J.H.
(1993b) Phylogenetic relationships of Bacteria based on comparative sequence
analysis of elongation factor Tu and ATP-synthase (3-subunit genes. Antonie van
Leeuwenhoek, 64, 285-305.
McNeil, M.M. and Brown, J.M. (1994) The medically important actinomycetes: epi-
demiology and microbiology. Clinical Microbiology Reviews, 7, 357-417.
Manafi, M., Kneifel, W. and Bascomb, S. (1991) Fluorogenic and chromogenic sub-
strates used in bacterial diagnostics. Microbiological Reviews, 55, 335-48.
Martinez-Murcia, A.J., Benllock, S. and Collins, M.D. (1992) Phylogenetic interre-
lationships of members of the genera Aeromonas and Plesiomonas as deter-
mined by 16S ribosomal DNA sequencing: Lack of congruence with results of
DNA: DNA hybridisation. International Journal of Systematic Bacteriology, 42,
412-21.
Mauchline, W.S. and Keevil, C.W. (1991) Development of the BIOLOG substrate
utilization system for identification of Legionella spp. Applied and Environmental
Maynard Smith, J. (1995) Do bacteria have population genetics? in Population
Genetics of Bacteria (eds S. Baumberg, J.P.W. Young, E.M.H. Wellington and J.R.
Saunders), Cambridge University Press, Cambridge, pp. 1-12.
Maynard Smith, J., Dowson, C.C. and Spratt, B.G. (1991) Localised sex in bacteria.
Nature, 349, 29-31.
Maynard Smith, J., Smith, N.H., O'Rourke, M. and Spratt, B.G. (1993) How clonal
are bacteria? Proceedings of the National Academy of Sciences, USA, 90, 4384-8.
Murray, R.G.E., Brenner, D.J., Colwell, R.R., De Vos, P., Goodfellow, M., Grimont,
P.A.D., Pfennig, N., Stackebrandt, E. and Zavarzin, G.A. (1990) Report of the ad
hoc committee on approaches to taxonomy within the Proteobacteria.
Nakamura, L.K. (1984) Bacillus amylolyticus sp. nov., nom. rev., Bacillus lautus sp.
nov., nom. rev. Bacillus pabuli sp. nov., nom. rev., and Bacillus validus sp. nov.,
nom rev. International Journal of Systematic Bacteriology, 34, 224-6.
Nielsen, P., Fritze, D. and Priest, F.G. (1995) Phenetic diversity of alkaliphilic
Bacillus strains: Proposal for nine new species. Microbiology, 141,1745-61.
Nour, S.M., Cleyet-Marel, J-C, Normand, P. and Fernandez, M.P. (1995) Genomic
heterogeneity of strains nodulating chickpeas (Cicer arientum L.) and descrip-
tion of Rhizobium mediterraneum sp. nov. International Journal of Systematic
O'Donnell, A.G., Embley, T.M. and Goodfellow, M. (1993) Future of bacterial sys-
tematics, in Handbook of New Bacterial Systematics (eds M. Goodfellow and A.G.
O'Donnell), Academic Press, London, pp. 513-24.
O'Donnell, A.G., Goodfellow, M. and Hawksworth, D.L. (1994) Theoretical and
practical aspects of the quantification of biodiversity among microorganisms.
Philosophical Transactions of the Royal Society B, 345, 65-73.
On, S.L.W., Holmes, B. and Sackin, M. (1996) A probabilistic identification matrix
for campylobacters, helicobacters and allied taxa. Journal of Applied Bacteriology,
81,425-32.
Palleroni, N.J. (1993) Structure of the bacterial genome, in Handbook of New Bacterial
Systematics (eds M. Goodfellow and A.G. O'Donnell), Academic Press, London,
pp. 57-113.
Pankhurst, R.J. (1991) Practical Taxonomic Computing. Cambridge University Press,
London.
Pot, B., Gillis, M., Hoste, B., Van de Velde, A., Bekaert, F., Kersters, K. and De Ley,
J. (1989) Intra-and intergeneric relationships of the genus Oceanospirillum.
Pridham, T.G. and Tresner, H.D. (1974) Genus I. Streptomyces Waksman-and
Henrici, in Bergey's Manual of Determinative Bacteriology (eds R.E. Buchanan and
N.E. Gibbons), Williams and Wilkins Co., Baltimore, pp. 748-829.
Priest, F.G. (1993) Systematics and ecology of Bacillus, in Bacillus subtilis and other
Gram-Positive Bacteria: Biochemistry, Physiology and Molecular Genetics (eds A.L.
Sonnenheim, J.A. Hoch and R. Losick), American Society for Microbiology,
Washington, DC, pp. 3-16.
Priest, F.G. and Williams, S.T. (1993) Computer-assisted identification, in Handbook
of New Bacterial Systematics (eds M. Goodfellow and A.G. O'Donnell), Academic
Press, London, pp. 361-81.
Rainey, F.A., Janssen, P.H., Morgan, H.W. and Stackebrandt, E. (1993) A biphasic
approach to the determination of the phenotypic and genotypic diversity of
some anaerobic, cellulolytic, thermophilic, rod-shaped bacteria. Antonie van
Ravin, A.W. (1960) The origin of bacterial species: genetic recombination and fac-
tors limiting it between bacterial populations. Bacteriological Reviews, 24,201-20.
Ravin, A.W. (1963) Experimental approaches to the study of bacterial phylogeny.
American Naturalist, 97, 307-18.
Roberts, M.S., Nakamura, L.K. and Cohan, P.M. (1994) Bacillus mojavensis sp. nov.,
distinguishable from Bacillus subtilis by sexual isolation, divergence of DNA
sequence, and differences in fatty acid composition. International Journal of
Systematic Bacteriology, 44, 256-64
Robertson, J.A., Howard, L.A., Zinner, C.L. and Stemke, G.W. (1994) Comparison
of 16S rRNA genes within the T960 and Parvo biovars of ureaplasmas isolated
from humans. International Journal of Systematic Bacteriology, 44, 836-8.
Rossello, R., Garcia-Valdes, J., Lalucat, J. and Ursing, J. (1991) Genotypic and
phenotypic diversity of Pseudonomas stutzeri. Systematic and Applied Microbiology,
8,124-7.
Ruimy, R., Riegal, P., Carlotti, A., Boiron, P., Bernardin. G., Monteil, H., Wallace,
R.J., Jr and Christen, R. (1996) Nocardia brasiliensis sp. nov., a new species of
Nocardia which groups bacterial strains previously identified as Nocardia
brasiliensis and associated with invasive disease. International Journal of
Systematic Bacteriology, 46, 259-64.
Sackin, M.J. and Jones, D. (1993) Computer-assisted classification, in Handbook of
New Bacterial Systematics (eds M. Goodfellow and A.G. O'Donnell), Academic
References 57
Saitou, N. and Nei, M. (1987) The neighbor-joining method: a new method for
constructing phylogenetic trees. Molecular Biology and Evolution, 4,40&-25.
Schleifer, K.H., Ludwig, W. and Amann, R. (1993) Nucleic acid probes, in Handbook
of New Bacterial Systematics (eds M. Goodfellow and A.G. O'Donnell), Academic
Selander, R.K., Li, J., Boyd, E.F., Wang, F-S. and Nelson, K. (1994) DNA sequence
analysis of the genetic structure of populations of Salmonella enterica and
Escherichia coli, in Bacterial Diversity and Systematics (eds F.G. Priest, A. Ramos-
Cormenzana and B.J. Tindall), Plenum Press, New York, pp. 17-49.
Seong, C.N., Park, S.K., Goodfellow, M., Kirn, S-B. and Hah, Y-C. (1995)
Construction of probability identification matrix and selective medium for aci-
dophilic actinomycetes using numerical classification data. Journal of
Skerman, V.B.D., McGowan, V. and Sneath, P.H.A. (1980) Approved lists of bacte-
rial names. International Journal of Systematic Bacteriology, 30,225-420.
Sneath, P.H.A. (1977) The maintenance of large numbers of strains of microorgan-
isms and the implications for culture collections. FEMS Microbiology Letters, 1,
333-4.
Sneath, P.H.A. (1979a) BASIC program for identification of an unknown with
presence-absence data against an identification matrix of percent positive
characters. Computers and Geosciences, 5,195-213.
Sneath, P.H.A. (1979b) BASIC program for character separation indices from an
identification matrix of percent positive characters. Computers and Geosciences, 5,
349-57.
Sneath, P.H.A. (1980a) BASIC program for the most diagnostic properties of
groups from an identification matrix of percent positive characters. Computers
and Geosciences, 6, 21-6.
Sneath, P.H.A. (1980b) BASIC program for determining the best identification
scores possible from the most typical examples when compared with an identi-
fication matrix of percentage positive characters. Computers and Geosciences, 6,
27-34.
Sneath, P.H.A. (1980c) BASIC program for determining overlap between groups in
an identification matrix of percent positive characters. Computers and
Geosciences, 6, 267-78.
Sneath, P.H.A. (1983) Distortions of taxonomic structure from incomplete data on
a restricted set of reference strains. Journal of General Microbiology, 129,1045-73.
Sneath, P.H.A. (1985) Future of numerical taxonomy, in Computer-Assisted Bacterial
Systematics (eds M. Goodfellow, D. Jones and F.G. Priest), Academic Press,
Orlando, pp. 415-431.
Sneath, P.H.A. (1992) International Code of Nomenclature of Bacteria. American
Society for Microbiology, Washington, DC.
Sneath, P.H.A. (1993) Evidence from Aeromonas for genetic crossing-over in ribo-
somal sequences. International Journal of Systematic Bacteriology, 43,626-9.
Sneath, P.H.A. and Sokal, R.R. (1973) Numerical Taxonomy: The Principles and
Practice of Numerical Classification. Freeman, San Francisco.
Sneath, P.H.A., Mair, N.S., Sharpe, M.E. and Holt, J.G. (eds) (1986) Bergey's Manual
of Systematic Bacteriology, Volume 2. Williams and Wilkins, Baltimore.
Sokal, R.R. and Sneath, P.H.A. (1963) Principles of Numerical Taxonomy. Freeman,
San Francisco.
Stackebrandt, E. and Goebel, B.M. (1994) Taxonomic note: A place for DNA:DNA
reassociation and 16S rRNA sequence analysis in the present species definition
in bacteriology. International Journal of Systematic Bacteriology, 44, 846-9.
Stackebrandt, E. and Goodfellow, M. (eds) (1991) Nucleic Acid Techniques in Bacterial
Systematics. John Wiley and Sons, Chichester.
Stanier, R.Y., Ingraham, J.L., Wheelis, M.L. and Paintes, P.R. (1986) The Microbial
World, Fifth Edition. Prentice Hall, Englewood Cliffs, New Jersey.
Stead, D.E., Sellwood, J.E., Wilson, J. and Viney, I. (1992) Evaluation of a commer-
cial microbial identification system based on fatty acid profiles for rapid, accu-
rate identification of plant pathogenic bacteria. Journal of Applied Bacteriology, 72,
315-21.
Strenstom, I.-M., Zakaria, A., Ternstrom, A. and Molin, G. (1990) Numerical taxon-
omy of fluorescent Pseudomonas associated with tomato roots. Antonie van
Swofford, D.L. and Olsen, G.J. (1990) Phylogenetic reconstruction, in Molecular
Systematics (eds D. Hillis and C. Moritz), Sinauer Associates, Sunderland, USA,
pp. 411-501.
Tamaoka, J. (1994) Determination of DNA base composition, in Chemical Methods in
Prokaryotic Systematics (eds M. Goodfellow and A.G. O'Donnell), John Wiley
and Sons, Chichester, pp. 463-469.
Trejo, W.H. (1970) An evaluation of some concepts and criteria used in the specia-
tion of streptomycetes. Transactions of the New York Academy of Sciences, Series II,
32, 989-97.
Triiper, H.G. and Schleifer, K.H. (1992) Prokaryote characterisation and identifica-
tion, in The Prokaryotes (eds A. Balows, H.G. Triiper, M. Dworkin, W. Harder
and K.H. Schleifer), 2nd edn, Springer-Verlag, New York, pp. 126-48.
Ursing, J. and Bruun, B. (1991) Genotypic heterogeneity of Flavobacterium group lib
and Flavobacterium breve demonstrated by DNA:DNA hybridisation. Acta
Pathologica Micobiologica et Immunologica Scandinavica, 99, 780-6.
Ursing, J.B., Lior, H. and Owen, R.J. (1994) Proposal of minimal standards for
describing new species of the family Campylobacteriaceae. International Journal
of Systematic Bacteriology, 44, 842-5.
Ursing, J.B., Rossello-Mora, R.A., Garcia-Valdes, E. and Lalucat, J. (1995)
Taxonomic note: a pragmatic approach to the nomenclature of phenotypically
similar genomic groups. International Journal of Systematic Bacteriology, 45, 604.
Van den Mooter, M. and Swings, J. (1990) Numerical analysis of 295 phenotypic
features of 266 Xanthomonas strains and related strains and an improved taxon-
omy of the genus. International Journal of Systematic Bacteriology, 40,348-69.
Vauterin, L., Vantomme, R., Pot, B., Hoste, B., Swings, J. and Kersters, K. (1990)
Taxonomic analysis of Xanthomonas campestris pv. legoniae and X. campestris pv.
pelargonii by means of phytopathological, phenotypic, protein electrophoretic
and DNA hybridisation methods. Systematic and Applied Microbiology, 13,166-76.
Vauterin, L., Swings, J. and Kersters, K. (1993) Protein electrophoresis and classifi-
cation, in Handbook of New Bacterial Systematics (eds M. Goodfellow and A.G.
O'Donnell), Academic Press, London, pp. 251-80.
References 59
Vauterin L., Hoste, B., Kersters, K. and Swings, J. (1995) Reclassification of
Xanthomonas. International Journal of Systematic Bacteriology, 45,472-89.
Vauterin, L., Yang, P. and Swings, J. (1996) Utilisation of fatty acid methyl esters
for the differentiation of new Xanthomonas species. International Journal of
Systematic Bacteriology, 46, 298-304.
Verissimo, A., Morais, P.V., Dioga, A., Gomes, C. and da Costa, M.S. (1996)
Characterisation of Legionella species by numerical analysis of whole-cell pro-
tein electrophoresis. International Journal of Systematic Bacteriology, 46,41-9.
Wayne, L.G. (1985) Computer-assisted analysis of data from co-operative studies
on mycobacteria, in Computer-Assisted Bacterial Systematics (eds M. Goodfellow,
D. Jones and F.G. Priest), Academic Press, Orlando, pp. 91-105.
Wayne, L.G., Krichevsky, M.I., Portyrata, D. and Jackson, C.K. (1984) Diagnostic
probability matrix for the identification of slowly growing mycobacteria in clin-
ical laboratories. Journal of Clinical Microbiology, 20, 722-9.
Wayne, L.G., Brenner, D.J, Colwell, R.R., Grimont, P.A.D., Kandler, P., Krichevsky,
M.I., Moore, L.H., Moore, W.E.C., Murray, R.G.E., Stackebrandt, E., Starr, M.P.
and Triiper, H.G. (1987) Report of the ad hoc committee on reconciliation of
approaches to bacterial systematics. International Journal of Systematic
Wayne, L.G., Good, R.C., Bottger, B.C., Butler, R., Dorsche, M., Ezaki, T., Gross,
W., Jonas, V., Kilburn, J., Kirschner, P., Krichevsky, M.I., Ridell, M., Shinnick,
T.M., Springer, B., Stackebrandt, E., Tarnok, I., Tarnok, Z., Tasaka, H., Vincent,
V., Warren, N.G., Knott, C.A. and Johnson, R. (1996) Semantide- and chemo-
taxonomy-based analyses of some problematic phenotypic clusters of slowly-
growing mycobacteria, a cooperative study of the International Working
Group on Mycobacterial Taxonomy. International Journal of Systematic
White, D., Sharp, RJ. and Priest, F.G. (1993) A polyphasic taxonomic study of ther-
mophilic bacilli from a wide geographical area. Antonie van Leeuwenhoek, 64,
357-86.
Williams, S.T. and Vickers, J.C. (1988) Detection of actmomycetes in the natural
environment, in Biology of Actinomycetes '88 (eds Y. Okami, T. Beppu and H.
Ogawara), Japan Scientific Societies Press, Tokyo, pp. 265-270.
Williams, ST., Vickers, J.C. and Goodfellow, M. (1984) New microbes from old
habitats? in The Microbe 1984: Part II: Prokaryotes and Eukaryotes (eds D. Kelly and
N.G. Carr), Cambridge University Press, Cambridge, pp. 219-56.
Woese, C.R. (1987) Bacterial evolution. Microbiological Reviews, 51, 221-71.
Woese, C.R. (1992) Prokaryotic systematics: the evolution of a science, in The
Prokaryotes (eds A. Balows, H.G. Truper, M. Dworkin, W. Harder and K.H.
Schleifer), 2nd edn, Springer-Verlag, New York, pp. 1-8.
Yassin, A.F., Rainey, F.A., Brzezinka, H., Burghardt, J., Lee, H.J. and Schaal, K.P.
(1995) Tsukamurella inchonensis sp. nov. International Journal of Systematic
Zillig, W., Klenk, H.P., Palm, P., Piihler, G., Gropp, F., Garrett, R.A. and Leffers, H.
(1989) The phylogenetic relations of DNA-dependent RNA-polymerases of
archaebacteria, eukaryotes and eubacteria. Canadian Journal of Microbiology, 35,
73-80.
Species in practice: exploring
uncultured prokaryote diversity in
natural samples
T. Martin Embley and E. Stackebrandt
Contacting address: Department of Zoology, The Natural History Museum, Cromwell
Road, London SW7 5BD, UK
ABSTRACT
Traditionally prokaryote species have been recognized using phenet-
ic methods incorporating information from the phenotype and geno-
type. Molecular methods have now made it possible to use 16S rRNA
sequences to analyse the prokaryote community in natural habitats
without the need to culture. The sequences recovered can be directly
compared with sequences from cultured organisms through the medi-
um of tree diagrams, and relationships can be interpreted in terms of
inferred common ancestry. These studies have revealed patterns of
prokaryote diversity which were unavailable using traditional micro-
biological methods. Not only are most of the lineages uncovered new
to science, but some are of such clear phylogenetic or ecological
importance that their further study is imperative.
4.1 INTRODUCTION
Indeed taxa are all much the same, even if some taxa include others. I hes-
itate to suggest that if there are taxonomic units of evolution, the units are
taxa generally' Nelson (1989: 61)
The main subjects of the present paper are those prokaryotes for which
there are no recognized laboratory cultures and hence no phenotypic
information. This apparent contradiction occurs because in the past it was
necessary first to isolate microorganisms into pure culture, before taxo-
nomic analysis could proceed and a newly discovered prokaryote be rec-
ognized as representing a new centre of taxonomic variation. Isolation of
a pure culture for study requires that a microorganism will grow under a
62 Exploring uncultured prokaryote diversity in natural samples
set of more or less artificial conditions, which many, perhaps most, have
seemed remarkably reluctant to do.
It is now possible to explore natural prokaryote taxonomic diversity
using molecular sequences without the need to isolate and culture
microorganisms. This approach has been incredibly successful, but it does
raise issues about how this uncultured microbial diversity can be appreci-
ated in familiar terms, and in the context of this volume, whether the term
species can be meaningfully applied to prokaryotes for which we only
have sequences as evidence. Before we discuss these issues, it is necessary
to discuss how species are currently recognized among prokaryotes which
can be cultured in the laboratory. It is also useful briefly to consider the
treatment of eukaryotes under the biological species concept, since this
probably dominates how species are generally viewed and discussed.
4.2 SPECIES AMONG EUKARYOTES

The biological species concept (BSC) was formulated for 'groups of inter-
breeding natural populations which are reproductively isolated from
other such groups' (Mayr, 1970: 12). So in a strict sense it should be used
only for those individuals for which these criteria have been satisfied.
However, as discussed by Sokal and Crovello (1970) most species, even
among Metazoa, are practically recognized using taxonomic information,
and reproductive isolation is assumed but seldom tested.
The recognition that taxonomic analysis is the practical means whereby
most species are recognized, has frequently raised the question of
whether species are somehow different from other taxa (some of this lit-
erature has been reviewed by Nelson, 1989; Mallet, 1995). Nelson (1989)
has argued persuasively from the perspective of a taxonomist that this is
not the case, commenting: 'There seems to be no basic taxonomic unit and
no particular taxonomic unit of evolution'. From this perspective the
major question transmutes into how best to carry out taxonomic analysis,
and this applies equally to all types of data and to all potential taxa -
eukaryote or prokaryote. It is of course perfectly legitimate subsequently
to try and understand the reasons why a particular group might be rec-
ognized in a particular taxonomic analysis.
4.3 SPECIES AMONG CULTURED PROKARYOTES

Gene flow does occur between prokaryotes but it occurs by different
mechanisms and has different effects compared with sexual eukaryotes
(Maynard-Smith et al., 1993; Cohan, 1994). Put simply, prokaryotes do not
produce gametes and do not have meiosis; thus, any concept such as the
BSC which requires these characteristics cannot be applied to prokaryotes.
Gene transfer among individuals of different bacteria varies in frequency
Practical recognition of cultured prokaryote species ' 63
(Maynard-Smith et al., 1993), and the rate of recombination varies among
loci encoding proteins of different types (Selander et al., 1994). At the
moment it is difficult to appreciate how bacterial population genetics can
provide another 'biological species concept' for prokaryotes (Maynard-
Smith, 1995). Bacterial population genetic studies clearly can and do pro-
vide characters for fine-resolution taxonomic analysis, and for studying
patterns of gene flow between individuals (Selander et al., 1994).
It has been recognized that asexual organisms do not fit the definition of
species under the BSC. Indeed, Mayr (1987) has previously suggested
using the term paraspecies to describe asexual organisms, but this has not
been adopted by prokaryote systematists. In fact, there would have to be
very convincing objective reasons for now limiting use of the word species
for the description of one particular subset of cellular life. In our opinion
such convincing arguments have not been presented. Mayr has suggested
that 'trouble' (Mayr, 1987:165) will be the inevitable consequence of adopt-
ing a concept of species that applies to all living organisms and also
'Inevitably in such a broadened concept of species, all those criteria will
have to be excluded that are particularly characteristic for the majority (our
emphasis) of species, that is the characteristics of biological species'. The
process of exploration of the natural diversity of prokaryotes has barely
begun, so arguments about numbers of species are almost certainly pre-
mature. However, it is already apparent (see later) that two of the possibly
three fundamental domains of life, the Archaea and Bacteria, are exclu-
sively prokaryote (Woese et al., 1990). Moreover, the sexual habits of many
microbial eukaryotes, a not insubstantial fraction of eukaryote diversity
and history, are also not well documented (Sogin, 1991).
4.4 PRACTICAL RECOGNITION OF CULTURED PROKARYOTE

SPECIES
Since the late 1950s (Sneath, 1957a,b) bacterial taxonomists have mainly
used numerical taxonomy to recognize entities which have subsequently
formed the basis for new species. The method as generally practised
analyses a group of strains for many unweighted characters, codes the
responses of each strain, calculates a measure of similarity using one or
more coefficients, and organizes the results by way of a phenogram or
ordination plot (Sneath and Sokal, 1973).
The characters which have traditionally been used have comprised the
ability to degrade or metabolize certain chemicals, the ability to produce,
or survive exposure to antibiotics, the ability to grow on particular carbon
or nitrogen sources, particular staining reactions and external morpholog-
ical features (Williams et al., 1983). A serious problem with some of the
more traditional tests and their corresponding analysis, are the difficulties
in inferring homologous (sensu Owen, see Patterson, 1987) features, i.e.
those which it is logically legitimate to compare when attempting to infer
historical relationships. For example, different microorganisms might pro-
duce a zone of clearing (the feature scored) in a bacterial lawn due to the
production of different antimicrobials, or may produce the same indicator
reaction in a biochemical test but using different enzymes. The uncertain
quality of tests sometimes used has been blamed (Sackin and Jones, 1993)
for numerical taxonomy's past short-comings in indicating consistent
higher groupings among microorganisms, as opposed to its relative suc-
cess and convenience for organising large numbers of prokaryotes into
homogeneous groupings which can be equated to species. More recently
there have been moves towards using protein banding patterns, fatty acid
profiles and rapid enzyme tests, which are perhaps more precise in what
they measure, and which can be semi-automated to deal with large num-
bers of strains (Goodfellow and O'Donnell, 1993).
An inferred amount of genomic similarity measured using DNA : DNA
reassociation assays was subsequently incorporated with overall pheno-
typic resemblance into the taxonomic method (the so-called polyphasic
approach) for identifying prokaryote species, the stated desire being to
provide a common basis for establishing species among different bacterial
groups (Johnson, 1973). It has been suggested that strains of a species
should share at least 70% DNA : DNA reassociation under optimal condi-
tions with 5°C or less ATm: the difference between the thermal stability of
the homologous and heterologous duplex (Wayne et al., 1987). This appar-
ently arbitrary level has been justified by reference to empirical observa-
tion that some strains of bacteria which are highly related phenetically
share at least this amount of DNA: DNA reassociation (Brenner, 1973),
and the observation that heteroduplexes which share less than 70% reas-
sociation are sometimes less thermally stable, implying a greater degree of
mispairing than the corresponding homoduplexes (Johnson, 1973).
DNA : DNA pairing assays have advantages in that they theoretically
sample a very large number of characters, but different methods may give
higher or lower values for the same strains (Grimont et al., 1980), and all
methods are subject to significant experimental error, typically .in the
region of 5-6% (Hartford and Sneath, 1990). The level of DNA : DNA reas-
sociation between strains of established taxonomic species can be greater
than 70%. For example, the two species Neisseria gonorrhoea and Neisseria
meningitidis share over 74% DNA : DNA reassociation (Rossau et al., 1989),
but are kept as separate species because they can be distinguished pheno-
typically (Barrett and Sneath, 1994), and because they cause different
diseases. So far there have been about 4000 prokaryote taxonomic species
validly described using a more or less polyphasic approach.
A potential problem with the phenetic approach to recognize clusters
which are then called species, is that these species may not be mono-
phyletic groups; non-monophyletic in that they may contain individuals
Using molecular sequences 65
which resemble each other phenetically, but some of them are more close-
ly related to individuals outside of the species by the criterion of common
ancestry. Classification based upon monophyletic groups is considered
desirable because only then will a hierarchical taxonomy be isomorphic
with a phylogenetic tree (Harvey and Pagel, 1991). Phenetics makes no
distinction between synapomorphy (i.e. shared derived characters),
homoplasy (false similarity including convergence), or plesiomorphy
(ancestral similarity) when it measures resemblance. There is an apparent
assumption among some systematists that prokaryote species identified
using phenetic methods are necessarily phylogenetic, e.g. 'at present the
species is the only unit which can be defined in phylogenetic terms'
(Wayne et al., 1987). It has been pointed out previously that in this context
the use of the term phylogenetic is potentially confusing (Sneath, 1988).
Sneath and Sokal have argued that phenetic groups defined using
'good' (see previous comments, Sackin and Jones, 1993) characters and
rigorous numerical phenetic methods are usually monophyletic (Sneath
and Sokal, 1973: 46). However, it has been demonstrated that under some
conditions phylogenetic and phenetic analyses give different trees for the
same data sets (Farris, 1977; Platnick, 1989). Moreover, in some recently
published papers dealing with prokaryotes there are differences between
branching diagrams based upon analyses of different character sets or
using different coefficients of similarity (e.g. compare composition of
some clusters in Figures 1 and 2, Dupuy et al., 1994; Figures 1 and 3, Rainey
et al., 1993; Figures 1 and 2, Segers et al., 1994). Unfortunately, there are no
estimates of branching pattern reliability (e.g. from bootstrapping) for the
figures compared from these cited papers, so it is not possible to gauge the
strength of support for the particular differences observed. It is probably
unrealistic, because of the effort that would be necessary, to think that
monophyly will be practically investigated for strains of the 4000 or so cur-
rently recognized prokaryote species, apart from a few model species
which have particular medical or ecological importance. It is outside of
our expertise to appreciate how widely, if at all, non-monophyly of
species is a potential problem for metazoan taxonomy.
4.5 EXPLORING UNCULTURED PROKARYOTE DIVERSITY USING

MOLECULAR SEQUENCES
The current polyphasic approach to recognize prokaryote species requires
pure cultures. Yet it is possible to extrapolate, from comparing direct and
viable counts, and from molecular studies, that much of prokaryote diver-
sity in nature is probably not recognized in pure culture (Ward et al., 1992).
Thus, most prokaryotes cannot at present be considered directly in terms
of prokaryote species under the polyphasic system. The lack of pure cul-
tures stems from the obvious difficulties in isolating prokaryotes from
nature when nothing is known of their growth requirements. There are,
however, persuasive arguments based upon consideration of global bio-
diversity (Hawksworth, 1995), nutrient cycling and ecosystem function
(Brock et al., 1994), and understanding the evolution of life itself (Woese,
1994), which suggest that these uncultured prokaryotes are some of the
more interesting and important candidates for biological study.
Environmental DNA samples have been analysed using reassociation
kinetics to estimate community complexity and to estimate the number of
contributing genomes (Torsvik et al., 1990), but this method is too impre-
cise to identify individual genomes, or to place them within a hierarchical
taxonomic framework. In contrast, analyses of molecular sequences,
which are now commonly used to infer phylogenetic relationships
between established prokaryote species, can be applied to individual
uncultured prokaryotes. It is this approach which currently holds greatest
promise for the exploration of uncultured natural microbial diversity
within a taxonomic framework (Pace et al., 1985).
4.6 16S rRNA ANALYSES TO INFER PHYLOGENETIC

RELATIONSHIPS BETWEEN PROKARYOTES
In prokaryote systematics it is comparisons of 16S rRNA sequences (now
more likely to be 16S rRNA gene sequences or 16S rDNA) which have
been used most to infer phylogenetic relationships between species. The
reasons for choosing 16S rRNA, and its useful features for classification
and identification of prokaryotes, have been fully discussed in articles by
Carl Woese, who pioneered this field of endeavour (Woese, 1987; Woese
et al., 1990). They include its functional constancy and universal distribu-
tion in cellular life, its high information content, ease of alignment, and its
mosaic structure which allows both ancient and recent relationships to be
inferred. Before molecular sequence comparisons, prokaryote taxonomists
had largely given up on trying to infer phylogeny and prokaryote
systematics was an endeavour aimed at most efficiently putting taxa into
groups, whereby their identification for medical, ecological or commercial
purposes could be best facilitated (Goodfellow and O'Donnell, 1993).
By 1977 sufficient sequence data were available to suggest that cellular
life, at least that in culture for study, could be organized into three
primary divisions based upon SSU rRNA sequence comparisons, which
Woese and Fox (1977) called the Eubacteria, the Archaebacteria and the
Urkaryotes (eukaryotes). The discovery of a second great kingdom of
prokaryote life - the Archaebacteria - is arguably one of the great achieve-
ments of 20th century biology. Sequence comparisons also supported
inferences from the fossil record, that the classical higher eukaryote king-
doms Plantae and Animalia are but recent branches in an already highly
diversified tree. Most of the branches in the tree of life based upon SSU
16S rRNA analyses to infer phylogenetic relationships • 67
rRNA sequences, and all of the deepest ones, are represented by microor-
ganisms. To signify the primacy of the three divisions of cellular life over
the classical metazoan kingdoms, Woese et al. (1990) subsequently sug-
gested that the term domain be used. Thus, Eubacteria became domain
Bacteria, Archaebacteria became domain Archaea and eukaryotes became
domain Eucarya.
Concerns have been expressed about basing the taxonomic hierarchy
for prokaryotes so heavily upon a single molecule (Olsen and Woese,
1993), but it is difficult at present to identify any other molecule likely to
usurp it as the most generally useful molecular tool. Anyway, it is impor-
tant to consider any gene tree as a hypothesis of relationships which
should be challenged using other genes. So far there have been few explicit
tests of congruence, but in the cases where topologies derived from other
genes have been compared with the 16S rRNA tree, the agreement about
relationships has sometimes been good, when allowance is made for the
different resolving power of different molecules (Ludwig et al., 1993;
Springer et al., 1995). At the very deepest levels of relationship, i.e.
between domains, it has proved more difficult to reconcile different gene
trees and there is a lively debate about different tree topologies and how
to interpret them from an evolutionary perspective (Forterre et al., 1993;
Gogarten, 1995; Golding and Gupta, 1995).
A second concern is what is the best method for inferring phylo-
genetic relationships from sequence data, a debate which affects all mole-
cular studies, not just those dealing with prokaryotes. It is outside of the
scope of this review to go into details of what are considered the strengths
and weaknesses of the different methods, and how they perform for infer-
ring phylogeny (Felsenstein, 1988; Sneath, 1989; Hillis et al, 1993).
Computer simulations and experiments have revealed that all methods
fail when the assumptions upon which they are based are badly violated,
and almost all will detect strong signals in the data which may reflect phy-
logeny (Jin and Nei, 1991; Hillis et al, 1993; Huelsenbeck and Hillis, 1993).
Most investigations of microbial relationships using 16S rRNA have
continued to use relatively simple phenetic methods such as neighbour
joining, for making phylogenetic inferences. The apparent greater success
of these methods in inferring consistent higher groupings for prokaryotes
(see Sackin and Jones, 1993) is probably due in part to data (sequences)
which allow more reliable hypotheses of homology to be inferred for the
characters (aligned base positions) and sequences compared (but see
Patterson, 1987, 1988). There is also increasing comparison, and greater
caution in the interpretation, of trees generated using different methods
such as parsimony and maximum likelihood (Felsenstein, 1981).
Bootstrapping has also served to introduce a necessary degree of caution
in interpreting the strength of support for particular groupings
(Felsenstein, 1985). Some of these changes in habits are moving towards
what Sneath and Sokal (1973) perhaps meant when they argued for rigor-
ous methods in numerical taxonomy. The trend towards increasing
sophistication of analysis of molecular data is likely to continue and it can
be expected that newer approaches such as spectral analysis (Lento et al,
1995), and transformations akin to paralinear distances (Lake, 1994;
Lockhart et al,, 1994), will soon make their appearance in the prokaryote
literature.
4.7 16S rDNA SEQUENCES TO INVESTIGATE THE

RELATIONSHIPS OF UNCULTURED PROKARYOTES IN NATURAL
SAMPLES
The methods whereby 16S rDNA sequences may be recovered and
analysed from natural populations of prokaryotes, without culturing
them, have been extensively reviewed, since Norman Pace and colleagues
first suggested the general approach (Pace et al., 1985; Olsen et al., 1986;
Ward et al., 1992). The current most common method is based upon PCR
amplification of 16S rRNA genes or gene fragments using 16S rRNA gene-
specific primers, followed by segregation of individual gene copies by
cloning into Escherichia coli (Giovannoni et al., 1990). The process produces
a library of community 16S rRNA genes, the composition of which can be
estimated by sampling clones and comparing their sequences by restric-
tion endonuclease digestion, their reaction to specific probes, or by full or
partial sequencing (Ward et al., 1992). This information can be analysed to
infer clone abundance and representation in the library. Unique clones
can be completely sequenced and their relationships to sequences from
cultured taxa in a taxonomic hierarchy based upon rRNA sequences can
be discovered.
A concern about using PCR to analyse microbial communities is how
accurately does library composition reflect gene composition in situ, i.e.
how well does PCR sample the pool of natural prokaryote diversity? Are
there biases in the system which mean that some sequences will amplify
more readily, or will some sequences be under-represented or missed
entirely? Simple computer models (Wagner et al., 1994) suggest that indi-
vidual PCR reactions may produce skewed distributions (PCR drift) of
products starting from template mixtures of known composition. PCR
drift is due to stochastic processes which occur in early cycles of PCR and
it can be reduced but not eliminated, by pooling many PCR reactions.
Selection for certain templates in a mixture can also occur if they have
different GC base compositions (Reysenbach et al., 1992), and it can be
predicted by simulation that some templates in mixtures may have higher
replication probabilities for other reasons (discussed in Wagner et al.,
1994). The contribution of starting templates from different bacteria will
also vary, even if there are the same number of individuals of each species
Prokaryote 16S rDNA sequence diversity in natural samples 69
in the original population. Different bacteria will be more or less suscepti-
ble to lysis or differential sampling may occur due to shape or size.
Different PCR products may be cloned or be retained more or less effi-
ciently in the new host, and different bacteria have different numbers of
rRNA genes and different sizes of genome, e.g. cyanobacteria may have
between two and six rRNA operons and their genomes may vary between
2.2 and 3.6 x 109 Da (Nichols et al., 1982). It is therefore necessary to treat
the distribution of clone sequences in libraries with caution, and to realize
that they may not accurately reflect sequence abundance in situ (Farelly et
al, 1995).
Measures of the relative abundance of a sequence in a sample can be
estimated by using probes (to the sequence) to analyse total ribosomal
RNA extracts (Giovannoni et al., 1990; DeLong et al., 1994). This approach
also has some limitations, including the fact that different prokaryotes
may contain different numbers of ribosomes and thus variable amounts of
probe target (Ward et al., 1992). A more direct measurement of cell abun-
dance can be obtained by labelling probes with fluorescence, to probe and
identify intact microorganisms directly in environmental samples
(DeLong et al., 1989). It is this approach that currently offers the greatest
potential for accurate counts of organisms in natural samples without the
need to culture them. It can also be used to link a sequence to a microbial
morpho-type and to identify samples which may contain cells from which
a sequence of particular taxonomic interest originates, thus providing a
tool for use in isolation strategies (Huber et al., 1995).
4.8 PROKARYOTE 16S rDNA SEQUENCE DIVERSITY IN NATURAL

SAMPLES: THE RELATIONSHIP BETWEEN 16S rRNA SEQUENCE
SIMILARITY AND PROKARYOTE SPECIES RECOGNIZED USING
DNA : DNA PAIRING ASSAYS
Before describing some of the results of community analysis of uncultured
prokaryotes using 16S rRNA sequences it is essential, in the context of this
symposium, to discuss the relationship between 16S rRNA sequence sim-
ilarities and prokaryote species as recognized using the polyphasic
approach. First of all it is important to realize that there is no precise cor-
relation between 16S rRNA sequence similarity and prokaryote nomen-
clature (Fox et al., 1992; Martinez et al., 1992). The 16S gene samples only a
fraction of the genome, and by itself it tells nothing about the important
phenotypic component which in practice often provides a guide to draw-
ing the taxonomic line. However, strains of taxonomic species which share
about 70% DNA: DNA reassociation often demonstrate inter-strain 16S
rRNA sequence similarities of over 98%, and strains which show less than
98% 16S rRNA sequence similarity seldom demonstrate DNA : DNA pair-
ing values of greater than 60% under optimal conditions (Amann et al.,
1992; Stackebrandt and Goebel, 1994). Within this framework, it is a rea-
sonable working hypothesis to infer that any environmental sequence,
which shares 98% 16S rRNA sequence or less similarity to sequences from
established prokaryote species, potentially originates from a new species
of prokaryote.
The 98% value is just a rule of thumb which has been developed using
similarities measured between complete or very nearly complete 16S
rRNA sequences. Clone sequences from environmental samples are often
only fragments. This is because some of the PCR primers designed to be
specific for particular groups will amplify only a fragment of the gene, e.g.
for beta-proteobacteria ammonia oxidizers (McCaig et al., 1994) or because
only short fragments of individual clones are actually sequenced.
Ribosomal RNA sequences are mosaics of variable and conserved
sequence regions depending on structure and function (Woese et al.,
1983), and different partial fragments of sequence will sample this mosaic
in different ways. Restriction of the analysis to more conserved stretches
or to more variable regions will lead to higher or lower similarity values
between sequences, respectively.
In addition to the known mosaic structure of 16S rRNA it is apparent that
the position of hypervariable regions may vary in different groups of
bacteria. It is therefore difficult to estimate precisely the degree of similarity
of complete sequences from partial sequences (Stackebrandt and Rainey,
1995). Using selected sequence stretches from members of the important
genus Vibrio as an example, i.e. region 50 through 500 and region 550
through 1500, the similarity is about 3% higher and 2% lower, respectively,
when compared with the values obtained with complete sequences. In this
genus the hypervariable regions are located in the 3' half of the molecule,
while, for example, in actinomycetes these regions are found within the 5'
half of the 16S rDNA molecule. In this group of organisms the similarity
values of the 5' half of the molecule are on the average 1.5% lower than
determined for the total sequence. While most studies have analysed the 5'
500 nucleotides, several studies have derived their conclusions about taxon
composition from other parts of the molecule.
A more important concern about analysing only partial fragments of
16S rRNA sequences for the exploration of microbial diversity is the nega-
tive effects it has on the ability to infer relationships. One of the great
strengths of sequence data is that it accumulates, so that old and new
sequences can be directly compared, to produce an increasingly compre-
hensive taxonomic analysis of the microbial world. However, this can only
be done if the stretches of sequence from different studies cover the same
parts of the molecule. Trees based upon small fragments of 16S rRNA
sequences are often less stable than those based on more characters, and
resolving power is lost concerning the relationships of environmental
sequences to those from cultured microorganisms (Schmidt et al., 1991).
SSU rRNA sequences to explore prokaryote diversity 71
4.9 RESULTS OF SOME STUDIES USING SSU rRNA SEQUENCES
TO EXPLORE PROKARYOTE DIVERSITY IN NATURE
The analysis of uncultured prokaryote communities using 16S rRNA

sequences has now become commonplace and it has already been wide-
ly reviewed (Ward et al, 1992; Embley et al., 1994; Olsen, 1994), so here
we will concentrate on a few generalizations and recent examples which
illustrate the power and potential of the approach. Almost all of the
sequences recovered so far from natural samples have been new to sci-
ence. In part this undoubtedly reflects the still small number of cultured
prokaryotes which have been analysed for their 16S rRNA sequences,
but it also suggests that there is enormous prokaryote diversity in nature
which remains to be discovered. Whether the pace of discovery of com-
pletely new lineages will continue, is not easy to appreciate from the
small number of samples and habitats so far investigated. In samples
taken from the oceans it is becoming apparent that sequences which
form monophyletic groups with previously discovered sequences (e.g.
the SAR 11 and SAR 7 bacterial clusters, and archaeal group I and II clus-
ters) have been found in most samples (DeLong et al., 1994; Mclnerney
et al., 1995; Mullins et al., 1995).
Some of the environmental sequences may be artefacts of the PCR
process, which has been shown sometimes to produce chimeric molecules
from mixtures of 16S rRNA templates (Liesack et al., 1991). However, as far
as can be ascertained by current analytical approaches (Robison-Cox et al.,
1995) most of the new sequences appear to be genuine; for example Barns
et al. (1994) reported that four of 98 clones analysed were chimeras, and
Choi et al. (1995) found seven chimeras among 81 clones analysed.
Because of their depth of branching in the rRNA tree based upon
sequences from cultured organisms, some of the environmental
sequences can be predicted to originate from organisms which may rep-
resent entirely new major groups of prokaryotes. For example, one of the
most interesting recent discoveries is that of 16S rRNA sequences from
Archaea in marine samples. Before describing the results of these studies
it is worth saying something about the diversity of the Archaea which are
already in laboratory culture. Woese and co-workers (Woese et al., 1990)
have identified two Archaea kingdoms, the Crenarchaeota and
Euryarchaeota, within the SSU rRNA tree for cultured Archaea. The two
groups are defined with reference to the root relative to Bacterial out-
group sequences, which split cultured Archaea into these two putative
monophyletic groups. The Archaea which make up the Euryarchaeota
and Crenarchaeota are physiologically heterogeneous and have tradition-
ally been collected from environments such as acidic hot springs, anaero-
bic sediments or saturated brines. Because they were seldom isolated from
(for example) the open oceans, freshwaters, or agricultural soils, there was
a perception that Archaea, with the exception of methanogenic Archaea,
might not be important in global ecology (Olsen, 1994).
In 1992 this perception was shown to be fundamentally flawed. Two
independent studies recovered two groups (termed group I and group II)
of novel Archaea 16S rRNA sequences from small samples of deep and
shallow waters taken from the Atlantic and Pacific Oceans (DeLong, 1992;
Fuhrman et al., 1992). Analysis of these sequences has revealed that group
I marine Archaea sequences are so divergent (Figure 4.1) from those of cul-
tured Archaea that some have suggested that they might be designated a
new Archaeal kingdom (Olsen and Woese, 1993). Sequences which form a
strongly supported monophyletic group with group I marine Archaea
sequences have subsequently been recovered from gut samples from a
Holothurian collected from 4870 m in the North Atlantic Abyss
(Mclnerney et al., 1995), and in samples from frigid marine surface waters
in Antarctica (DeLong et al., 1994). A recent survey of sequence diversity
within a single paddy field in Japan has reported Archaeal sequences
which formed a sister group to the marine Archaea group I, suggesting
that members of this group may be present in non-marine systems (Ueda
et al., 1995). Unfortunately, the sequence fragments for the paddy field
clones are rather short (c. 300 bases) and this has made assessment of their
detailed position relative to other clones uncertain. Sequences which form
a deeply branched monophyletic group with group II marine Archaea
(Figure 4.1) have recently been detected in salt marsh sediments on the
south-east coast of England, which are sometimes covered during high
tides (M. Munson, D. Nedwell and T. M. Embley, unpublished data).
These data, from small samples taken over vast geographic distances,
suggest that group I and II marine Archaea are extraordinarily important
Figure 4.1 Tree based upon SSU rRNA sequences from representative cultured
Archaea including sequences from clone libraries prepared from different environ-
mental samples. The tree is based upon analysis of 791 sequence positions using the
Jukes and Cantor (Jukes and Cantor, 1969) correction and neighbour joining (Saitou
and Nei, 1987). Essentially the same topology was recovered in maximum parsi-
mony analysis. Interstingly, analysis using a method (Lockhart et al., 1994) designed
to deal with unequal base compositions between sequences, as occurs here, consis-
tently placed environmental clones env. pJP27 and env. pJP78 on the Eucarya
branch. Key: env. pJP, clones from mud volcano area of Jim's Black Pool, hot spring,
Yellowstone National Park (Barns et al., 1994); env. carna, Carna Bay bacterioplank-
ton, west coast of Ireland, unpublished M. Malarkey and R. Powell, University of
Galway; env. pPM7, deep water sample Atlantic, unpublished R. Mclnerney and R.
Powell; env. ANT, Antarctic bacterioplankton (DeLang et al., 1994); env. WHAR,
Woods Hole DNA clone (DeLang 1992); env. SBAR, Santa Barbara Channel bacterio-
plankton (DeLang, 1992); env. C and env. P, coastal salt marsh clones, south-east
coast of England, unpublished M. Munson, D. Nedwell and T.M. Embley.
Published aligned sequences were obtained from the Ribosomal Data Base Project
(Maidakrffl/., 1994).
SSL7 rRNA sequences to explore prokaryote diversity 73
in ocean processes, a suggestion confirmed by recent data which suggest
that at certain times of the year marine Archaea comprise about 34% of the
prokaryote biomass in coastal Antarctic surface waters (DeLong et al.,
1994). The published data also suggest that pelagic marine Archaea are
broadly represented by these two groups. This apparent limited diversity,
in the sense of the number of new monophyletic groups so far discovered,
would potentially greatly facilitate analysis of community dynamics using
nucleic acid probes (Mullins et al., 1995). None of the marine Archaea has
yet been cultured so their role(s) in the ocean systems, and the reasons
why they are apparently so extraordinarily successful, are completely
unknown.
Pyrodictium occultum
Sulfolobus shibatae
Pyrobaculum islandicum
Crenarchaeota
Thermofilum pendens
env. pJP33
env. pJP89
env. carnall
env.ANT12
- env. SBAR1 Marine group
env. SBAR5
- env. WHARQ
env. PM7
env. pJP41
Methanobacterium bryantii
Methanothermus fervidus
——— Haloferax valcanii
- Methanoculleus olentangyi
- Thermoplasma acidophilum
r- env. C25 "
-PL env. P8 Salt marsh DMA clones Euryarchaeota
J I—— env. C84
'—— env. P1
env. ANTS
C env.WHARN
— env. SBARIA Marine group II
_i— env. OARB
'— env. SBAR16
- Methanococcus jannaschii
- Methanopyrus kandleri
env. pJP27
env. pJP78
Giardia lamblia
Giardia muris
Hexamita inflata Eucarya
Tritrichomonas foetus
Aquifex pyrophilus
Geotoga petraea
Thermus ruber Bacteria
—— Deinococcus radiodurans
10%
A second major discovery of Archaea relationships came from Barns,
Pace and co-workers (Barns et al., 1994) who used primers designed selec-
tively to retrieve a variety of different Archaea SSU rRNA sequences from
5 ml of sediment taken from a hot spring termed 'Jim's Black Pool' in
Yellowstone National Park. Several of the recovered sequences showed
no close phylogenetic affinity to cultivated species of Archaea (pJP 33, 41,
89, 27, and pJP 78 in Figure 4.1). In different analyses (Barns et al., 1994)
these five sequences branched closer to the root of the Crenarchaeota than
did sequences from cultured Crenarchaeota. In some analyses JP 27 and JP
78 were the sister group to all other Archaea (as in Figure 4.1 and as in the
RDP maximum likelihood tree, Maidak et al., 1994).
Other examples of widely distributed marine monophyletic groups of
16S rRNA sequences (Maidak et al., 1994), are clearly bacterial in character
(Mullins et al., 1995). They include a large group of alpha-proteobacterial-
like sequences (proteobacteria are a major phylum of cultured bacteria)
termed the SAR 11 cluster and a group of cyanobacterial-like sequences
called the SAR 7 cluster (Giovannoni et al., 1990), both of which have been
recovered from the Atlantic and Pacific Oceans and are probably very
important in these systems (Giovannoni et al., 1990; Schmidt et al., 1991;
Fuhrman et al., 1993).
An apparently monophyletic group of widely distributed soil clone
sequences has also recently been discovered (E. Stackebrandt, unpub-
lished). Related sequences have been isolated from a forest soil in
Queensland, Australia, a peat bog in North Germany, and mud from a hot
spring in New Zealand. Recent sequence entries in the ribosomal data
base project (server@rdp.lefe.uiuc.edu or http://rdp.life.uiuc.edu/
RDP/data/ssu.html) suggest that short fragments of similar sequences
have now been recovered from paddy field and soybean field samples. In
some analyses representatives of these new sequences form a clade at the
base of the Actinomycete phylum (a group noted for its abilities to pro-
duce antibiotics), along with an iron-oxidizing culture TH3 isolated from
a copper bioleaching pond, and a filamentous strain 'Microthrix parvicella'
isolated from activated sludge. Whether or not suitable isolation strategies
for uncultured members of this group can be inferred from the limited
physiological information on TH3 and 'Microthrix' remains to be tested.
As well as these highly divergent and geographically widespread mono-
phyletic groups of sequences, most studies have also recovered sequences
which bear more resemblance to those from different cultured prokaryote
species. A single example will serve to illustrate the potential for discovery
in even well-studied habitats. Choi et al. (1994) focused on spirochaete
bacteria of the medically important genus Treponema occupying a gingival
crevice in the mouth of a patient with severe destructive periodontitis.
After extracting DNA from gingival material and amplifying rRNA genes,
81 clones related to Treponema were identified Further analysis revealed
Concluding remarks 75
that these new sequences fell into 23 clusters defined at 98% or less 16S
rRNA sequence similarity, calculated over the 5' 500 bases of 16S rRNA
sequence. Only two of these groups contained representatives of cultured
oral Treponema species, suggesting that the other 21 clusters represent
novel centres of variation, and potentially new species. Goebel and col-
leagues (unpublished) have subsequently managed to isolate and grow
spirochaetes with the same sequences as two of the closely related (about
96.5% similar) sequence groups discovered by Choi et al. The isolated
strains indeed show significant phenotypic differences from each other,
and under current taxonomic practice would be classified as new species.
4.10 CONCLUDING REMARKS

At the beginning of this chapter we posed the question of how uncultured
prokaryote diversity can be considered in familiar terms, since current prac-
tice in prokaryote systematics requires pure cultures for species description.
The most useful approach to the analysis of prokaryote natural diversity is
through tree diagrams of relationships based upon taxonomic analysis of
sequence characters. Within this context environmental sequences can be
directly compared with sequences from cultured taxa, prokaryote or
eukaryote, and hypotheses of common ancestry and estimates of sequence
divergence can be inferred. In this context, the relationships inferred for
uncultured taxa and the information they convey about prokaryote distrib-
utions and evolutionary diversity, are no less real because they deal with
organisms which have yet to be brought into recognized laboratory culture
or classified as prokaryote species. The biological implications of the results
so far are enormous, not only in terms of the number of new lineages which
have been revealed and the degree of phylogenetic novelty displayed by
some of them, but also in the promise of an enormous prokaryote pheno-
typic diversity still to be discovered. Although taxonomic analyses can pro-
ceed using sequences alone, only the isolation of living cultures will allow
organism biology to be more fully explored. Some of the new prokaryotes
uncovered by molecular analysis of environmental samples, are of such
clear biological, phylogenetic or ecological significance that they merit the
most determined efforts to do precisely this.
The excitement and potential for a microbiology which succeeds in
linking historical analysis through sequence comparisons, with the study
of organismal biology have been captured by the naturalist Edward O.
Wilson (1994: 364):
' If I could do it all again, and relive my vision for the twenty first
century, I would be a microbial ecologist. Ten billion bacteria live in
a gram of ordinary soil, a mere pinch held between thumb and fore-
finger. They represent thousands of species, almost none of which
are known to science. Into that world I would go with the aid of
modern microscope and molecular analysis. I would cut my way
through clonal forests sprawled across grains of sand, travel in an
imagined submarine through drops of water proportionally the size
of lakes, and track predators and prey in order to discover new life
ways and alien food webs. All this and I need venture no more than
ten paces outside my laboratory building. The jaguars, ants, and
orchids would still occupy distant forests in all their splendour, but
now they would be joined by an even stranger and vastly more com-
plex living world virtually without end/
4.11 REFERENCES
Amann, R.I., Lin, C, Key, R., Montgomery, L. and Stahl, D.A. (1992) Diversity
among Fibrobacter isolates: towards a phylogenetic classification. Systematic and
Applied Microbiology, 15, 23-31.
Barns, S.M., Fundyga, R.E., Jeffries, M.W. and Pace, N.R. (1994) Remarkable
archaeal diversity detected in a Yellowstone National Park hot spring environ-
ment. Proceedings of the National Academy of Sciences, USA, 91,1609-13.
Barrett, S.J. and Sneath, P.H.A. (1994) A numerical phenotypic taxonomic study of
the genus Neisseria. Microbiology, 140,2867-91.
Brenner, D.J. (1973) Deoxyribonucleic acid reassociation in the taxonomy of enteric
bacteria. International Journal of Systematic Bacteriology, 23,298-307.
Brock, T.D., Madigan, M.T., Martinko, J.M. and Parker, J. (1994) Biology of
Microorganisms. Prentice Hall International Inc., New York.
Choi, B.K., Paster, B.J., Dewhirst, F.E. and Gobal, U.B. (1994) Diversity of cultivable
and uncultivable oral spirochaeates from a patient with severe destructive peri-
odontitis. Infection and Immunity, 62,1889-95.
Cohan, P.M. (1994) Genetic exchange and evolutionary divergence in prokaryotes.
Trends in Ecology and Evolution, 9,175-80.
DeLong, E.F. (1992) Archaea in coastal marine environments. Proceedings of the
National Academy of Sciences, USA, 89, 5685-9.
DeLong, E.F., Wickham, G.S., and Pace, N.R. (1989) Phylogenetic stains: ribosomal
RNA-based probes for the identification of single cells. Science, 243,1360-3.
DeLong, E.F., Wu, K.Y., Prezelin, B.B. and Jovine, R.V.M. (1994) High abundance
of Archaea in Antarctic marine picoplankton. Nature, 371, 695-7.
Dupuy, N., Willems, A., Pot, B., Dewettinck, D., Vandenbruaene, I., Maestrojuan,
G., Dreyfus, B., Kersters, K., Collins, M.D. and Gillis, M. (1994) Phenotypic and
genotypic characterisation of bradyrhizobia nodulating the leguminous tree
Acacia albida. International Journal of Systematic Bacteriology, 44,461-73.
Embley, T.M., Hirt, R.P. and Williams, D.M. (1994) Biodiversity at the molecular
level: the domains, kingdoms and phyla of life. Philosophical Transactions of the
Royal Society London B, 345,21-33.
Farelly, V., Rainey, F.A. and Stackebrandt, E. (1995) Effect of genome size and rrn
copy number on PCR amplification of 16S rRNA genes from a mixture of
bacterial species. Applied and Environmental Microbiology, 61, 2798-801.
Farris, J.S. (1977) On the phenetic approach to vertebrate classification, in Major
Patterns in Vertebrate Evolution (eds M. K. Hecht, P. Goody and B. M. Hecht),
Plenum, New York, pp. 823-50.
References 77
Felsenstein, J. (1981) Evolutionary trees from DNA sequences: a maximum likeli-
hood approach. Journal of Molecular Evolution, 17, 368-76.
Felsenstein, J. (1985) Confidence limits on phylogenies: an approach using the
bootstrap. Evolution, 39, 783-91.
Felsenstein, J. (1988) Phylogenies from molecular sequences: inference and relia-
bility. Annual Review of Genetics, 22, 521-65.
Forterre, P., Benachenhou-Lahfa, N., Confalonieri, F., Duguet, M., Elie, C. and
Labedan, B. (1993) The nature of the last universal ancestor and the root of the
tree of life, still open questions. BioSystems, 28,15-32.
Fox, G.E., Wisotzkey, J.D. and Jurtshuk, P., Jr (1992) How close is close: 16S rRNA
sequence identity may not be sufficient to guarantee species identity.
Fuhrman, J.A., McCallum, K. and Davis, A.A. (1992) Novel major archaebacterial
group from marine plankton. Nature, 356,148-9.
Fuhrman, J.A., McCallum, K. and Davis, A.A. (1993) Phylogenetic diversity of sub-
surface marine microbial communities from the Atlantic and Pacific Oceans.
Applied and Environmental Microbiology, 59,1294-302.
Giovannoni, S.J., Britschgi, T.B., Moyer, C.L. and Field, K. G. (1990) Genetic diver-
sity in Sargasso Sea bacterioplankton. Nature, 345, 60-3.
Gogarten, J.P. (1995) The early evolution of cellular life. Trends in Ecology and
Evolution, 10,147-51.
Golding, G.B. and Gupta, R.S. (1995) Protein-based phylogenies support a chimeric
origin for the eukaryotic genome. Molecular Biology and Evolution, 12,1-6.
Goodfellow, M. and O'Donnell, A.G. (1993) Handbook of New Bacterial Systematics.
Academic Press, London.
Grimont, P.A.D., Popoff, M.Y., Grimont, F., Coynault, C. and Lemelin, M. (1980)
Reproducibility and correlation study of three deoxyribonucleic hybridisation
procedures. Current Microbiology, 4, 325-30.
Hartford, T. and Sneath, P.H.A. (1990) Experimental error in DNA-DNA pairing:
a survey of the literature. Journal of Applied Bacteriology, 68, 527-42.
Harvey, P.H. and Pagel, M.D. (1991) The Comparative Method in Evolutionary Biology.
Oxford University Press, Oxford.
Hawksworth, D.L. (1995) Biodiversity, Measurement and Estimation. Chapman &
Hall, London.
Hillis, D.M., Allard, M.W. and Miyamoto, MM. (1993) Analysis of DNA sequence
data: phylogenetic inference. Methods in Enzymology, 224,456-87.
Huber, R., Burggraf, S., Mayer, T., Barns, S.M., Rossnagel, P. and Stetter, K.O.
(1995) Isolation of a hyperthermophilic archaeum predicted by in situ RNA
analysis. Nature, 376, 57-8,
Johnson, J.L. (1973) Use of nucleic acid homologies in the taxonomy of anaerobic
bacteria. International Journal of Systematic Bacteriology, 23, 308-15.
Jukes, T.H. and Cantor, C.R. (1969) Evolution of protein molecules, in Mammalian
Protein Metabolism (ed. H.N. Munro), Academic Press, New York, pp. 21-132.
Lake, J.A. (1994) Reconstructing evolutionary trees from DNA and protein
sequences: Paralinear distances. Proceedings of the National Academy of Sciences,
USA, 91,1455-9.
Lento, G.M., Hickson, R.E., Chambers, G.K. and Penny, D. (1995) Use of spectral
analysis to test hypotheses on the origins of pinnipeds. Molecular Biology and
Evolution, 12, 28-52.
Liesack, W., Weyland, H. and Stackebrandt, E. (1991) Potential risks of gene ampli-
fication by PCR as determined by 16S ribosomal DNA analysis of a mixed-
culture of strict barophilic bacteria. Microbial Ecology, 21,191-8.
Lockhart, P.J., Steel, M.A., Hendy, M.D. and Penny, D. (1994) Recovering evolu-
tionary trees under a more realistic model of sequence evolution. Molecular
Biology and Evolution, 11, 605-12.
Ludwig, W., Neumaier, J., Klugbauer, N., Brockmann, E., Roller, C, Jilg, S., Reetz,
K., Schachtner, I., Ludvigsen, A., Bachleitner, M., Fischer, U. and Schleifer, K.H.
(1993) Phylogenetic relationships of Bacteria based on comparative sequence
analysis of elongation factor Tu and ATP-synthase beta-subunit genes. Antonie
van Leeuwenhoek, 64,285-305.
Maidak, B.L., Larsen, N., McCaughey, J., Overbeek, R., Olsen, G.J., Fogel, K.,
Blandy, J. and Woese, C.R. (1994) The ribosomal database project. Nucleic Acids
Research, 22, 3485-7.
Mallet, J. (1995) A species definition for the modern synthesis. Trends in Ecology and
Martinez, A.J., Benlloch, S. and Collins, M.D. (1992) Phylogenetic interrelation-
ships of members of the genera Aeromonas and Plesiomonas as determined by
16S ribosomal DNA sequencing: lack of congruence with results of DNA-DNA
hybridisations. International Journal of Systematic Bacteriology, 42,412-21.
Maynard-Smith, J. (1995) Do bacteria have population genetics? in Population
Genetics of Bacteria (eds S. Baumberg, J.P.W. Young, E.M.H. Wellington and J.R.
Saunders), Cambridge University Press, Cambridge, pp. 1-12.
Maynard-Smith, J., Smith, N.H., O'Rourke, M. and Spratt, B.C. (1993) How clonal
are bacteria? Proceedings of the National Academy of Sciences, USA, 90,4384-8.
Mayr, E. (1970) Populations, Species and Evolution. Belknap Press of Harvard,
Mayr, E. (1987) The ontological status of species: scientific progress and philo-
sophical terminology. Biology and Philosophy, 2,145-66.
McCaig, A.E., Embley, T.M. and Prosser, J.I. (1994) Molecular analysis of enrichment
cultures of marine ammonia oxidisers. FEMS Microbiology Letters, 120, 363-8.
Mclnerney, J.O., Wilkinson, M., Patching, J.W., Embley, T.M. and Powell, R. (1995)
Recovery and phylogenetic analysis of novel archaeal rRNA sequences from a
deep-sea deposit feeder. Applied and Environmental Microbiology, 61,1646-8.
Mullins, T.D., Britschgi, T.B., Krest, R.L. and Giovannoni, S.J. (1995) Genetic com-
parisons reveal the same unknown bacterial lineages in Atlantic and Pacific
bacterioplankton communities. Limnology and Oceanography, 40,148-58.
Nelson, G. (1989) Species and taxa: systematics and evolution, in Speciation and its
Consequences (eds D. Otte and J.A. Endler), Sinauer Associates, Sunderland,
Nichols, J.M., Foulds, I.J., Crouch, D.H. and Carr, N.G. (1982) The diversity of
cyanobacterial genomes with respect to ribosomal RNA cistrons. Journal of
General Microbiology, 128,2739-46.
Olsen, G.J. (1994) Archaea, Archaea, everywhere. Nature, 371, 657-8.
Olsen, G.J., Lane, D.J., Giovannoni, S.J., Pace, N.R. and Stahl, D.A. (1986) Microbial
ecology and evolution: a ribosomal RNA approach. Annual Review of
References 79
Olsen, G.J. and Woese, C.R. (1993) Ribosomal RNA: a key to phylogeny. The FASEB
Journal, 7,113-23.
Pace, N.R., Stahl, D.A., Lane, D.J. and Olsen, G.J. (1985) Analysing natural micro-
bial populations by rRNA sequences. American Society for Microbiology News, 51,
4-12.
Patterson, C. (1987) Introduction, in Molecules and Morphology in Evolution: Conflict
or Compromise (ed. C. Patterson), Cambridge University Press, Cambridge, pp.
1-22.
Patterson, C. (1988) Homology in classical and molecular biology. Molecular Biology
and Evolution, 5, 603-25.
Platnick, N.I. (1989) Cladistics and phylogenetic analysis today, in The Hierarchy of
Life (eds B. Fernhom, K. Bremer and X. Jornvall), Excerpta Medica, Amsterdam,
pp. 17-24.
Rainey, F.A., Janssen, P.H., Morgan, H.W. and Stackebrandt, E. (1993) A biphasic
approach to the determination of the phenotypic and genotypic diversity of
some anaerobic, celluloytic, thermophilic, rod-shaped bacteria. Antonie van
Reysenbach, A.-L., Giver, L.J., Wickham, G.S. and Space, N.R. (1992) Differential
amplification of rRNA genes by polymerase chain reaction. Applied and
Environmental Microbiology, 58, 3417-18.
Robison-Cox, J.F., Bateson, M.M. and Ward, D.M. (1995) Evaluation of nearest
neighbor methods for detection of chimeric small-subunit rRNA sequences.
Applied and Environmental Microbiology, 61,1240-5.
Rossau, R., Vandenbussche, G., Thielemans, S., Segers, P., Grosch, H., Gothe, E.,
Mannheim, W. and De Ley, J. (1989) Ribosomal ribonucleic acid cistron similar-
ities and deoxyribonucleic acid homologies of Neisseria, Kingella, Eikenella,
Simonsiella, Alysiella and centres for disease control groups EF-4 and M-5 in the
emended family Neissseriaceae. International Journal of Systematic Bacteriology, 39,
185-98.
Sackin, M.J. and Jones, D. (1993) Computer assisted classification, in Handbook of
New Bacterial Systematics (eds M. Goodfellow and A.G. O'Donnell), Academic
Saitou, N. and Nei, M. (1987) The neighbor-joining method: new method for
reconstructing phylogenetic trees. Molecular Biology and Evolution, 4,406-25.
Schmidt, T.M., DeLong, E.F. and Pace, N.R. (1991) Analysis of a marine pico-
plankton community by 16S rRNA gene cloning and sequencing. Journal of
Segers, P., Vancanneyt, M., Pot, B., Torck, U., Hoste, B., Dewerrinck, D., Falsen, E.,
Kersters, K'. and de Vos, P. (1994) Classification of Pseudomonas diminuta Leifson
and Hugh 1954 and Pseudomonas vesicularis Busing, Doll and Freytag 1953 in
Brevundimonas gen. nov. as Brevundimonas diminuta comb. nov. and
Brevundimonas vesicularis comb, nov., respectively. International Journal of
Selander, R.K., Li, J., Boyd, F. and Wang, F. S. (1994) DNA sequence analysis of the
genetic structure of populations of Salmonella enterica and Escherichia coli, in
Bacterial Diversity and Systematics (eds F.G. Priest, A. Ramos-Cormenzana and
B.J. Tindall), Plenum Press, New York, pp. 17-49.
Sneath, P.H.A. (1957a) The application of computers to taxonomy. The Journal of
General Microbiology, 17,201-26.
Sneath, P.H.A. (1957b) Some thoughts on bacterial classification. Journal of General
Microbiology, 17,184-200.
Sneath, P.H.A. (1988) The phenetic and cladistic approaches, in Prospects in
Systematics (ed. D.L. Hawksworth), Clarendon Press, Oxford, pp. 252-73.
Sneath, P.H.A. (1989) Analysis and interpretation of sequence data for bacterial
systematics: the view of a numerical taxonomist. Systematic and Applied
Microbiology, 12,15-31.
Sneath, P.H.A. and Sokal, R.R. (1973) Numerical Taxonomy. W.H. Freeman and
Company, San Francisco.
Sokal, R.R. and Crovello, T.J. (1970) The biological species concept: a critical evalu-
ation. American Naturalist, 104,127-53.
Sogin, M.L. (1991) Early evolution and the origin of eukaryotes. Current Opinion in
Genetics and Development, 1,457-63.
Springer, E., Sachs, M.S., Woese, C.R. and Boone, D.R. (1995) Partial gene
sequences for the A subunit of methyl-coenzyme M reductase (mcrl) as a phy-
logenetic tool for the family Methanosarcinaceae. International Journal of Systematic
Stackebrandt, E. and Goebel, B.M. (1994) Taxonomic note: a place for DNA : DNA
reassociation and 16S rRNA sequence analysis in the present species definition
in bacteriology. International Journal of Systematic Bacteriology, 44, 846-9.
Stackebrandt, E. and Rainey, F.A. (1995) Partial and complete 16S rDNA
sequences, their use in generation of 16S rDNA phylogenetic trees and their
implications in molecular ecological studies, in Molecular Microbial Ecology
Manual (eds A.D.L. Akkermans, J.D. van Elsas and F.J. de Bruin), Kluwer
Academic Publishers, Dordrecht, The Netherlands, pp. 1-17.
Torsvik, V., Salte, K., Sorheim, R. and Goksoyr, J. (1990) Comparison of phenotyp-
ic diversity and DNA heterogeneity in a population of soil bacteria. Applied and
Environmental Microbiology, 56, 776-81.
Ueda, T., Suga, Y. and Matsuguchi, T. (1995) Molecular phylogenetic analysis of a
soil microbial community in a soybean field. European Journal of Soil Science, 46,
415-21.
Wagner, A., Blackstone, N., Cartwright, P., Dick, M., Misof, B., Snow, P., Wagner,
G.P., Bartels, J., Murtha, M. and Pendleton, J. (1994) Surveys of gene families
using polymerase chain reaction: PCR selection and PCR drift. Systematic
Biology, 43, 250-61.
Ward, D.M., Bateson, M.M., Weller, R. and Ruff-Roberts, A.L. (1992) Ribosomal
RNA analysis of microorganisms as they occur in nature. Advances in Microbial
Ecology, 12, 219-86.
Wayne, L.G., Brenner, D.J., Colwell, R.R., Grimont, P.A. D., Kandler, O.,
Krichevsky, M.I., Moore, L.H., Moore, W.E.C., Murray, R.G.E., Stackebrandt, E.,
Starr, M.P. and Truper, H.G. (1987) Report of the ad hoc committee on reconcil-
iation of approaches to bacterial systematics. International Journal of Systematic
Williams, S.T., Goodfellow, M., Alderson, G., Wellington, E.M.H., Sneath, P.H.A.
and Sackin, M.J. (1983) Numerical classification of Streptomyces and related gen-
era. Journal of General Microbiology, 129,1743-813.
References 81
Wilson, E.O. (1994) Naturalist. Island Press, Washington.
Woese, C.R. (1987) Bacterial evolution. Microbiological Reviews, 51, 221-71.
Woese, C.R. (1994) There must be a prokaryote somewhere: microbiology's search
for itself. Microbiological Reviews, 58,1-9.
Woese, C.R. and Fox, G.E. (1977) Phylogenetic structure of the prokaryotic
domain: the primary kingdoms. Proceedings of the National Academy of Sciences,
USA, 74, 5088-90.
Woese, C.R., Gutell, R., Gupta, R. and Noller, H.F. (1983) Detailed analysis of the
higher-order structure of 16S-like ribosomal ribonucleic acids. Microbiological
Reviews, 47, 621-69.
Woese, C.R., Kandler, O. and Wheelis, M.L. (1990) Towards a natural system of
organisms: Proposal for the domains Archaea, Bacteria, and Eucarya.
Proceedings of the National Academy of Sciences, USA, 87,4576-9.
Species problems in eukaryotic
algae: a modern perspective
D. M. John and C. A. Maggs
Contacting address: Department of Botany, The Natural History Museum, Cromwell
ABSTRACT
The algae are a very diverse, unnatural assemblage of seven distinct
phylogenetic lineages of oxygen-producing photosynthetic organ-
isms. The number of species presently recognized, about 36 000, is
estimated to be about 10% of the true worldwide total. Species-level
taxonomy is based explicitly or implicitly on the detection of morpho-
logical discontinuities in sets of field-collected or cultured algae. New
data on phenotypic variation, breeding compatibility, and molecular
analyses are clarifying species concepts. Culture studies have demon-
strated that the species concept traditionally applied to many mor-
phologically simple algae is too narrow. Polyploidy, for example, can
cause spontaneous changes in the morphology of some clonal cultures
of green algal groups. The biological species concept has not been
widely examined since relatively few algae are available as clonal
cultures and sexual reproduction is either unknown or rare and
unpredictable in the majority of algal classes. It has been tested most
frequently in rhodophytes, chlorophytes and diatoms: congruence
has been shown to exist in some genera between morphological data
and sexual compatibility. Cryptic variation has been demonstrated by
the discovery of mating complexes or sibling species within tradition-
al morphospecies. In diatoms, for example, investigations of sexual
compatibility indicate that many 'morphospecies' are masking signifi-
cant variation. Molecular data are assisting in calibrating or testing the
limits of morphospecies and may provide the touchstone for the inter-
pretation of other data. In the future the traditional morphological
species concept will increasingly operate alongside less formalized
concepts involving data from other disciplines. These data will enable
larger suites of concordant characters to be used for calibrating species
concepts and defining species boundaries. If cladistic methods are
84 Species problems in eukaryotic algae: a modern perspective
applied more frequently in future to study species-level relationships
in algae then the 'phylogenetic species concept' will come into wider
use.
5.1 INTRODUCTION
It is not an easy task to review species concepts and species delimitation
in an unnatural assemblage of oxygen-producing photosynthetic organ-
isms possessing enormous morphological, cytological, molecular and
reproductive diversity. Ultrastructural, biochemical and molecular studies
have demonstrated the 'algae' to be divisible into seven evolutionary lin-
eages, the chlorophytes, chromophytes, rhodophytes, dinophytes,
euglenophytes, cryptophytes and glaucophytes. There is an increasing
tendency to adopt a 'protistan' view and to place these lineages of so-
called primitive eukaryotes into the Protista, a Kingdom that can no
longer be justified (Corliss, 1994). For convenience, all those photosyn-
thetic eukaryotic protists traditionally referred to as algae are treated in
this review. Excluded are the 'blue-green algae', a prokaryotic group now
more correctly referred to as the Cyanobacteria or 'blue-green bacteria'
(see Castenholz, 1992, for a review of species concepts in this group).
Global concern over biodiversity issues, the impact of molecular data
on taxonomic decision making, and the increasing adoption of cladistic
methods of data analysis are just a few of the reasons why it is timely to
debate species concepts in algae. Increasingly, agendas for future envi-
ronmental research depend upon comparisons of estimates of species
diversity. It is tacitly assumed that the units compared are equivalent -
an assumption that is clearly untenable when dealing with a diverse and
unnatural assemblage like the algae. Despite this non-equivalence, such
comparisons continue to be made along with estimates, by taxonomic
specialists in particular groups, of the numbers of species still to be
described. About 36 000 species have been described and educated
'guesstimates' predict the true global total to be in excess of 200 000
(Andersen, 1992; John, 1994). Recently it has been suggested that the
estimate for the diatoms, a chromophyte group, should been increased
by a factor of 10 to 200 000 species (Mann and Droop, 1996), so that the
number of species currently recognized would be only about 11% of the
real total (see Table 5.1).
A vast literature exists on the nature of species and species concepts
(see Grant, 1981, for reviews) with no consensus emerging on what con-
stitutes a species, nor agreement on the philosophical approach to be
taken towards species as classes or individuals. Towards the end of the
19th century, discontinuities or unevenness in the pattern of morpholog-
ical variation were the universally accepted criteria for distinguishing
species (the 'morphological species concept'). Other species concepts
The morphological species concept 85
Table 5.1 Diversity of algae: estimated known number of species and predicted
world species totals within the major series or lineages of algae
Taxonomic Known Estimated Proportion

grouping species species known (%)
15 345 79 000 19
SUBKINGDOM
BILIPHYTA 4263 12800 33
(two algal classes)
KINGDOM
CHROMISTA
Chromophyta
sensulato 13776 217200 6
(10 algal classes)
KINGDOM
PROTOZOA 2853 9270 31
(three algal classes)
TOTAL 36237 318270 11
developed subsequently include the 'biological species concept' (Mayr,

1992), which has found wide acceptance as the most objective way of
defining species limits in sexually reproducing animals. The seven species
concepts reviewed by Steussy in 1990 (morphological, biological, genetic,
palaeontological, evolutionary, phylogenetic, biosystematic) are all applic-
able to algae, although the morphological species concept is still the one
most frequently applied. It is not our intention to review the long-running
debate on the relative merits of adopting one concept as opposed to
another, but to reflect on how algal species are dealt with in practice by
reference to selected examples.
5.2 THE MORPHOLOGICAL SPECIES CONCEPT
5.2.1 Historical overview

The morphological species concept has dominated algal systematics: the
vast majority of species are recognized by discontinuities in morphological
characters. It applies implicitly or explicitly even to those genera or
groups in which attempts have been made to use other species concepts.
The morphological species or 'morphospecies' concept - sometimes
referred to as the Linnaean, classical or typological concept - considers a
taxon described at one point in time. It is assumed that morphological
characters are stable until a quite different form is exhibited under
changed physiological conditions or at another growth stage. Of the 14
algal genera recognized by Linnaeus in 1753, only four (Conferva, Ulva,
Fucus, Chara) are attributable to the 'algae' as defined in a modern sense;
these macroalgae were recognized by readily observed morphological
criteria. William Hudson was the first English botanist to use the
Linnaean system of classification and nomenclature. The species con-
cepts he used when describing a large number of new algae (Hudson,
1762) can still be deduced by reference to his herbarium, literature
sources and his own publications (Irvine and Dixon, 1982). In the follow-
ing century, the period from 1800-1875 has been referred to as the
'Golden Age of plant taxonomy' (Prescott, 1951) because major advances
were made in plant systematics, including the description of the majori-
ty of the genera recognized today. These advances were in large part due
to considerable improvement in optical instruments (e.g. invention of the
achromatic lens in 1827) that facilitated critical observations of taxonomi-
cally important features of both microscopic and macroscopic algae.
Some of the taxonomic problems encountered today when working with
microalgae date back to the 19th century. Many are related to the failure
of early taxonomists to provide complete species diagnoses, take account
of variability in key diagnostic features or preserve material for future
examination. The incompleteness of the diagnoses was often due to an
examination of limited amounts of field-collected material with the result
that some sexual, life-history and ontogenetic stages were not observed.
Nevertheless, key characters for identification of macroalgae were recog-
nized by some talented taxonomists working in the early and mid-19th
century. For example, the description of Fucus devoniensis [now
Ahnfeltiopsis devoniensis (Greville) Silva et De Cew, see below] by Greville
(1821) was based on acute morphological and phenological observations
by Mrs Amelia Griffiths; the importance of one of the features she noted
was only understood a century-and-a-half later, after a comprehensive
study of this group (Maggs et al., 1992).
A new wave of interest in algal systematics in the 1950s followed the
introduction of the transmission electron microscope (TEM), improve-
ments in interference light microscopy, and the availability of a vast array
of equipment developed for the cell biologist. The introduction in the mid-
1960s of the scanning electron microscope (SEM) was of particular signifi-
cance to phycologists studying algae with interesting surface features.
These powerful and versatile research tools led to the discovery of new
suites of morphological characters and a re-interpretation of those
described earlier. The synthesis and evaluation of ultrastructural data
strengthened the morphological species concept and gave rise to a dra-
matic increase in the number of recognized species. For example, all but
one of almost 50 species now attributed to the chrysophyte genus
Paraphysomonas were described after the advent of the electron micro-
scope, when the fine details of scale morphology became evident (Preisig
and Hibberd, 1982). Several other groups (e.g. diatoms, stoneworts) bene-
fited greatly from SEM examination: many new 'morphospecies' were
described, while some existing species were reduced to synonymy.
The ability to grow clones of algae (axenically or non-axenically)
under rigorously controlled laboratory conditions has permitted great
advances in our understanding of algal species by providing new infor-
mation on morphology and cytology, reproduction, life-history and
ontogeny. Laboratory culture of algae had its beginning towards the end
of the last century but it was not until the 1920s and 1930s that 'type cul-
tures' were established. During the 1940s, culture-based investigations
of algae became fully established through the work of Pringsheim, and
they were continued in the following decade by Bold and his co-work-
ers who took full advantage of new culture techniques and media then
available (see reviews by Pringsheim, 1967; Bold, 1974). Today, species
are often recognized after detailed culture investigations; eventually the
synthesis of the results of culture-based studies with other types of data
may lead to the development of sounder species concepts and taxonom-
ic systems. Species descriptions should ideally be based on a study of
several clones since a single clone will not possess the range of variation
found in natural populations (Lewin, 1975). The establishment of formal
'Culture Collections' has been of tremendous significance, particularly as
they can hold cultures of the 'type' when new species are described.
These are of paramount importance for maintaining populations of
microalgae in the same way as herbaria house critical collections of pre-
served macroalgal material.
In relatively simple algae there is a limited range of morphological
characters available for species diagnosis, so other features increase in
importance. These include characters associated with nucleic acids,
metabolism, descriptive biochemistry, physiology, ontology, phenology,
breeding relations, life-history and ecology. As it has only been possible to
study some of these aspects relatively recently, microalgal taxonomy is
going through a revolutionary period. Attention is often directed towards
these investigations in order to resolve problems remaining after the dili-
gent application of alpha taxonomy (naming and describing of organ-
isms). However, to date only a small number of algal taxa have been inves-
tigated using a full range of sophisticated molecular, genetic, cytological,
breeding and culture techniques. Selected examples that demonstrate the
contribution of molecular data and crossability experiments to our under-
standing of algal species will be considered later.
5.2.2 The type method in algae

The classical taxonomic method involving comparison of material with
type specimens ('the type method', see Silva, 1952) is of fundamental
importance when studying macroalgae and some microscopic groups like
diatoms (Williams, 1993). Diatoms are one of the few classes of microalgae
that retain many taxonomically important characters on preservation,
because their systematics are based largely on the highly resistant siliceous
frustule. For seaweeds in particular, preparation of monographs depends
on the macroscopic and microscopic examination of dried herbarium
mounts, liquid-preserved material and, when available, living material.
Morphologically defined species are considered more reliable and well-
founded if the variability of taxonomically important characters has been
thoroughly investigated. Assessing the significance of morphological vari-
ation is easier in seaweeds and other macroalgae than in microalgae, pro-
viding that there are large numbers of herbarium specimens available for
examination and that key characters remain unaffected by long-term
preservation. However, sometimes staggeringly large numbers of speci-
mens must be examined to assess the full range of variation in potentially
important taxonomic characters. For example, Littler and Littler (1992)
studied all available herbarium material including types, approximately
1500 specimens, when monographing the green alga Avrainvillea in the
tropical western Atlantic. They concluded from this analysis and experi-
mental evidence (see below) that although traditionally important in
species recognition, some of the more obvious gross morphological traits
(e.g. number of stipes per holdfast, blade thickness and texture) exhibit
substantial variation in response to substratum features and other envi-
ronmental parameters. In amending diagnoses, emphasis was placed
therefore on more reliable anatomical characters (e.g. form and shape of
medullary and cortical filaments) that were well preserved in dried
herbarium material.
Recent quantitative morphological studies (Necchi and Zucchi, 1995;
Vis et al, 1995) have greatly clarified the systematics of freshwater repre-
sentatives of the red algal genera Batrachospermum and Audouinella.
Necchi and Zucchi evaluated qualitative and quantitative data on all
potentially useful morphological characters for 45 Brazilian populations
of Audouinella and all relevant type specimens. Although the size ranges
of many morphological features overlapped between species, multivari-
ate analysis unequivocally grouped the populations into five clusters,
four of which included a type specimen (one type was unavailable). Five
species were therefore clearly discriminated by this elegant application of
the type method. In contrast, the large majority of rnicroalgal species lose
many of their most valuable taxonomic characters on preservation and
field-collected material frequently lacks features required for precise
determination (e.g. reproductive structures, certain life-history stages).
As a consequence, culturing of clones under carefully manipulated labo-
ratory conditions is often indispensable for describing and identifying
microalgae. It is common to see references in the literature to 'type cul-
tures'. These are only living elements of the type since, according to the
International Code of Botanical Nomenclature, Article 9.5, 'Type specimens of
names of taxa must be preserved permanently and cannot be living
plants or cultures' although it recommends (9A.1) that 'Wherever practi-
cable a living culture should be prepared from the holotype material of
the name of a newly described taxon of fungi or algae and deposited in a
reputable culture collection'. The maintenance of a cultured element of
the type has proven invaluable for comparisons with the original descrip-
tion and for the examination of characters subsequently recognized to be
of taxonomic significance. There are many problems associated with
using clonal cultures to characterize species, however, particularly if mor-
phological features are thought to be species-specific on the basis of their
state in single clonal isolates of each species.
5.2.3 Significance of experimental data

The contribution of experimental studies in the field and laboratory to our
understanding of morphologically defined algae cannot be overestimated.
Such studies have demonstrated the extraordinary degree of plasticity
exhibited by many algae and the extent to which morphological charac-
ters are environmentally and genetically controlled. Frequently there has
been a failure to recognize just how phenotypically plastic are algae, lead-
ing to doubt attaching to the validity of many earlier-described species.
Even at the present time insufficient account is taken of the taxonomic, as
opposed to the physiological or ecological, significance of transplants and
other manipulative field experiments. An exception to this is the study by
Littler and Littler (1992) of populations of all western Atlantic species of
Avrainvillea in which they were able to demonstrate, for example, that A.
levis Howe develops a bulbous holdfast in sand or soft sediment and a
conical holdfast when growing on hard substrata. In several other species
of Avrainvillea quantitative characters considered of taxonomic impor-
tance (e.g. stipe length) were discovered to be unreliable since influenced
by light and depth conditions. There are many other genera in which mor-
phological characters show continuous variation, with each character a
graded series of expressions that overlap between species. Although con-
tinuity of ranges of variation in key characters between putative species is
suggestive of conspecificity, this has not always proven to be the case
when thoroughly investigated. For example, species discrimination in the
notoriously difficult green algal genera Cladophora and Enteromorpha is
often based on overlapping quantitative characters (e.g. cellular dimen-
sions, number of pyrenoids per cell); tables are often more useful than
keys for comparing character combinations and identifying species.
Several investigations of crossability within these two genera (Eliding,
1963; Hoek, 1963,1964,1982) have demonstrated a close correspondence
between the results obtained using biological and morphological species
criteria.
As already stated, the contribution of culture studies to the evaluation
of microalgal morphospecies cannot be overestimated. These have
enabled the validity of algal morphospecies originally described from
field-collected material to be tested. Several studies have demonstrated
that species are mere 'ecomorphs' or 'ecophenes' characteristic of particu-
lar environmental conditions. One genus that has received considerable
attention is the very polymorphic and widely distributed green alga
Stigeoclonium. As a result of culture studies, the most reliable and conserv-
ative characters for distinguishing its species are those associated with the
system of prostrate filaments and germination of the zoospores. Of the 28
species recognized by Islam (1963), only three were recognized by Simons
et al. (1986). In his review of algal species diversity, Andersen (1992) men-
tions another well-researched example, Scenedesmus, a green algal genus
containing 330 species and 1300 varieties according to its most recent revi-
sion (Hegewald and Silva, 1988). Culture studies by Trainor and co-work-
ers (Trainor, 1991; Trainor and Egan, 1991) have demonstrated that the
principal characters used for species discrimination in Scenedesmus (e.g.
presence and form of spines, colony size) are temperature-dependent.
Many species of Scenedesmus are now considered to be 'ecomorphs' that,
in response to environmental cues, undergo a seasonal or aseasonal suc-
cession termed cyclomorphosis. As a result of these studies, the number of
valid species is estimated by Trainor (personal communication, in
Andersen, 1992) to lie somewhere between 12 and 30.
Similarly, culture studies have demonstrated considerable phenotypic
variation in desmids, a green algal group whose species are typically dis-
tinguished on the basis of morphological characteristics observed in field-
collected material. In view of the variation exhibited by desmid species,
the recent floristic treatment of British desmids by Brook (1997) illustrates
populations of cells in order to demonstrate the full form range of each
species as encountered in nature. Breeding experiments indicate the pos-
sible existence of syngens within desmid morphospecies (Blackburn and
Tyler, 1987), as reported in other green algal groups and diatoms (see sec-
tion 5.3). Desmidologists will continue to define desmids using the tradi-
tional typological method since the majority only reproduce asexually.
Many culture-based studies on diatoms also reveal that taxonomically
important features of frustule morphology are influenced by a wide range
of environmental factors that include temperature, salinity, eutrophica-
tion and silica supply (see Cox, 1993, for references).
It is important not to fall into the trap of considering all environmen-
tally associated morphological variation to be the result of phenotypic
plasticity. Recent studies have borne out Russell and Fielding's (1981)
prediction that many variable characters once attributed to phenotypic
plasticity will be shown to be at least partly heritable. This has proven the
case in Chondrus crispus (Stackhouse) Guiry, a very variable red alga to
which a large number of infraspecific epithets have been applied. In this
alga Guiry (1992) discovered that morphological variation is associated
with particular shore habitats. Of the 26 isolates examined by him, all were
fully interfertile in culture and yet each maintained its own distinct mor-
phology. It is clear, therefore, that morphological variability in this single
biological species is under both environmental and genetical control.
Another example is the brown alga Pilayella littoralis (L.) Kjellman in
which the significance of morphological variation, particularly differences
in axis diameter between Baltic and Atlantic populations, has been hotly
debated during the past two decades. It has now been demonstrated by
Russell (1994) that although, in culture, axis diameter shows a relationship
to salinity and other environmental conditions, the limits are genetically
determined. Baltic populations have significantly narrower axes, both in
the field and in culture, and therefore might be regarded as a geographi-
cal subspecies. The species is apomictic, however, so the biological species
concept is not applicable and all distinctive variants are analogous to the
microspecies of certain higher plants.
An interesting example of taxonomic insights gained from study of a
single clone was reported by McCourt and Hoshaw (1990). A clone of the
green alga Spirogyra underwent spontaneous change after 45 months of
culturing to produce by autopolyploidy four distinct sub-clones. These
sub-clones were morphotypes distinguished by differences in filament
breadth and three corresponded to recognized species. McCourt and
Hoshaw (1990) consider each morphotype to be a 'species complex' and
argue against describing further species based on minor morphological
variations including filament width. Similar complexes are believed to be
widespread in the genus and on this basis Hoshaw estimated for
Andersen (Andersen, 1992) the number of Spirogyra species to be nearer 50
rather than the 386 reported by Kadlubowska in 1984.
A further contribution of culture studies to microalgal taxonomy is to
enable species to be defined using suites of characters only evident when
clonal cultures are grown under carefully defined laboratory conditions.
Many species of aquatic algae and the majority of soil and symbiotic algae
have been described only after isolation and study in laboratory culture.
For example, the taxonomy of freshwater representatives of
Pseudendodonium, a filamentous green algal genus, is entirely culture-
based. Characters of the prostrate system of filaments as evident in culture
have been used to compare and group clonal cultures; groups correspond
to species and subgroups to varieties (John and Johnson, 1989). As in the
majority of green algae, the species of Pseudendodonium described in
culture are 'morphospecies' and the biological species concept is not
applicable since sexual reproduction is unknown or, at least, a rare and
unpredictable event. The morphologically simple green alga Trebouxia, the
most common phycobiont of lichens, is often impossible to identify to
species level when in situ. Only in culture does Trebouxia exhibit all the
morphological characteristics (e.g. form of the chloroplast, cell size and
shape during different growth phases) necessary for recognizing its 25
species (Gartner, 1985). Phylogenetic analysis of the nuclear-encoded
small-subunit ribosomal RNA (18S rRNA) in four genera of green algae,
including three Trebouxia species (Friedl and Zeltner, 1994), resolved a
'lichen algae group'. One of the Trebouxia species was more closely related
to members of other genera than to its congeners suggesting the 'genus'
to be paraphyletic.
5.2.4 Interpretation and data analysis

The majority of algal taxonomists are still 'Linnaean in their thinking'
(Guiry, 1992) since the recognition of species is often based largely on the
observer's a priori (intuitive) opinion concerning phylogeny. This taxo-
nomic 'method' relies on the experience and competence of the specialist
in deciding the importance of each character (whether morphological, bio-
chemical or otherwise) and determining whether differences between
populations are sufficient to warrant species-level recognition. This 'evo-
lutionary approach' is highly subjective in terms of the selection and
weighting of characters. Taxonomic judgements are particularly difficult
to make when putative 'species' are not separated by absolute morpho-
logical discontinuities but rather by combinations of overlapping quanti-
tative characters (as in Cladophora).
Only in the past three decades has there been significant progress in
the development of rigorous and objective methods for understanding
and interpreting morphological data, including distinguishing between
different kinds of similarity (Hennig, 1966). Of the three analytical
approaches (phenetics, evolutionary systematics, cladistics), cladistics is
finding widest acceptance. This method, developed from a body of taxo-
nomic theory known as phylogenetic systematics, is organized around
three principles: (i) monophyletic taxa are natural; (ii) organisms are relat-
ed through descent; and (iii) evolutionary modifications uniquely shared
by organisms are evidence of their unique phylogenetic history (Hennig,
1966). Use of the method leads to the formation of groups on the basis of
The biological species concept 93
nested sets of derived characters. Unlike other approaches, it is scientific
in the sense that the results can be repeated and hypotheses tested using
the Popperian falsification model. There is no doubt that cladistics is a
powerful tool for testing phylogenetic hypotheses, but it is not without
conceptual and practical problems (Tehler, 1994).
Several reviews of cladistics (Kociolek et al, 1989; Theriot, 1989, 1992;
Williams, 1992) were intended to draw the attention of phycologists to its
potential for examining phylogenetic relationships within algae. Cladistics
has been widely adopted for addressing taxonomic problems in algae at a
level above that of the 'species'. There are a few exceptions amongst
macro- and microalgal groups where it has been used at the species level.
Several diatom studies (Williams, 1985; Theriot et al., 1987) have used
cladistics in order to analyse morphological character sets when revising
lower taxonomic groups. In a study by Littler and Littler (1992) of the
green alga Avrainvillea (see above), phylogenetic relationships were
demonstrated by carrying out a cladistic analysis based largely upon
anatomical characters. The analysis distinguished only species groups and
highlighted the problems of working with morphologically plastic algae.
5.3 THE BIOLOGICAL SPECIES CONCEPT
5.3.1 To what extent can this be applied to algae?

Mann (1995) stated that 'The biological species concept can perhaps be
viewed as a practical guide to the interpretation of the variation pattern
and to the equivalence of the species category between different plant and
animal groups'. However, this guidance can be used only for biparental,
sexually reproducing organisms. The application of the biological species
concept to the algae must therefore be very limited, as eight classes are not
known to reproduce sexually; sexual reproduction in many of the remain-
der is a rare and unpredictable event. Some algal lineages show a tenden-
cy to have abandoned sexual processes in the course of evolution. Even in
sexually reproducing algae, the biological species concept has been
explored only for comparatively few groups. The results of algal breeding
studies cannot always be reconciled with other data, and as Guiry (1992)
has pointed out 'phycologists, while professing to approve o f . . . a biolog-
ical species concept, quickly abandon it when difficulties arise'. The bio-
logical species concept is criticised by cladists on the grounds that sexual
compatibility is a primitive trait (a 'universal symplesiomorphy') and con-
sequently cannot be used for defining monophyletic taxa (Manhart and
McCourt, 1992). Notwithstanding such criticisms, attempts to employ a
biological species concept in the algae have provided valuable new
insights into algal taxonomy. Here, we present examples of recent
research on the biological species concept in three algal classes, the
Rhodophyta (red algae), the Bacillariophyta (diatoms) and the
Chlorophyta (green algae).
5.3.2 Rhodophyta
The red algae are ideal for breeding experiments since many readily grow
and reproduce sexually in culture. Many hybridization experiments have
been performed following the demonstration by Edwards (1970) that two
morphologically distinct species of Polysiphonia were not interfertile. In a
recent review by Guiry (1992), it is concluded that intra- and interpopu-
lation crossability patterns in red algae are equally as complex as those
found in other organisms. He recognized four categories of congruence
or otherwise of the morphological and biological species concepts (Table
5.2). The first and last of these categories are quite straightforward, show-
ing a good correspondence between the two species concepts.
Considerable research effort has been devoted to some examples in the
second and third categories, and some progress has been made recently
in resolving the apparent incompatibility of the two species concepts. For
example, Koch (1986) reported that Danish isolates of Polysiphonia fibril-
losa (Dillwyn) Sprengel and Polysiphonia violacea sensu Harvey were inter-
sterile although morphologically very similar. Maggs and Hommersand
(1993) showed that Koch's 'P. fibrillosa' was a misidentification of P. har-
veyi Bailey, a species readily distinguishable by several morphological
characters (e.g. position of plastids within the periaxial cells). Of the cases
placed by Guiry (1992) in his third category (see Table 5.2), one can now
be clarified. A more extensive breeding study of the P. harveyi complex
has shown (C.A. Maggs, unpublished data) that two of the Japanese
forms currently recognized as separate species are actually elements of a
nearly globally distributed interbreeding complex, or single biological
species. The morphological differences between field populations, which
can be maintained in culture to some extent, presumably indicate infra-
specific genetic variability.
5.3.3 Bacillariophyta
Conventionally diatoms have been recognized and classified according to
details of the acid-cleaned siliceous wall or frustule revealed on examina-
tion of field-collected samples under the light or electron microscope. One
of the principal problems in diatom taxonomy has been to decide which
discontinuities in the morphological characteristics correspond to bound-
aries at different levels of classification (genera, species, infraspecific taxa).
The major taxonomic revisions of diatoms by Krammer and Lange-
Bertalot (1986,1988,1991a,b) have adopted a wide morphological concept
and abandoned many infraspecific taxa in the belief that minor variations
The biological species concept 95
Table 5.2 Congruence and non-congruence of the morphological and biological
species concepts in the red algae (From Guiry, 1992.)
1. Morphologically indistinguishable plants that are capable of interbreeding

with each other throughout their geographical ranges: complete congruence
between the two species concepts.
2. Morphologically indistinguishable plants that are incapable of crossing with
each other. In this case the variation is cryptic and such sibling species are not
afforded species status in algae. Cryptic variation is detectable in various ways
including the analysis of allozymes and molecular data.
3. Morphologically dissimilar plants that will cross, some having various levels of
genetic and ecological distinctness. These are recognized to be very problem-
atical entities with fertility sometimes depending on the nature of the cross.
4. Morphologically dissimilar plants that will not cross. Many of these were poor-
ly known morphologically and, although separated initially by breeding exper-
iments, further morphological studies showed clear differences.
in cell size, shape, stria density, pattern and, possibly, ultrastructural fea-
tures, are of little or no taxonomic significance. Mann (1989b) considers
that a narrower rather than a broader morphological concept is needed,
stating there to be 'a general failure to look critically at the variation pat-
tern and a belief that variation is often continuous within, and sometimes
also between, traditionally recognized species. Arguing logically from this
unsubstantiated, and I believe incorrect premise, various authors have
suggested that most infraspecific taxa are worthless and that many species
should be combined'. Mann and co-workers have explored the mating
patterns in mixed semi-natural populations of raphid diatoms and discov-
ered incompatibility barriers between infraspecific taxa recognized by
small morphological differences. For example, Mann and Droop (1996)
discovered intrinsic barriers to hybridization existing between six mor-
phologically distinct, sympatric populations of Sellaphora pupula (Kiitzing)
Mereschkowsky and recommend species status for these incompatible
'morphotypes'. Other closely related species of Sellaphora have proved to
be reproductively isolated (Mann, 1989a, 1995), indicating congruence
between the morphological and crossability data.
5.3.4 Chlorophyta
The 'mating type' phenomenon is widespread in algae including many
unicellular and colonial green algae (e.g. Chlamydomonas moewusii Gerloff,
Eudorina species, Gonium pectorale Mueller, Cosmarium botrytis Meneghini
ex Ralfs, Micrasterias thomasiana Archer, Closterium ehrenbergii Meneghini
ex Ralfs). It involves multiple, genetically controlled mating types existing
within a morphologically defined species. These mating units might be
regarded as sibling or incipient biological species whose cryptic variation
is often only detectable by means of physiological, biochemical or molec-
ular analysis. One alga that has been extensively investigated is the 16-
celled volvocine green alga Pandorina morum Bory. Coleman (1977)
discovered that within this morphological species exist 20 distinct mating
complexes, or syngens. These subordinate sibling or biological species are
the functional or operational species units. In the case of Pandorina morum
the morphological species is a much more comprehensive entity than the
biological species. Mating groups (biological species) are thought to be of
polyploid origin in the desmid Closterium ehrenbergii (Ichimura and Kasai,
1990). As soon as a ploidy change occurs, the new morphotypes are nor-
mally incapable of interbreeding so may be regarded as incipient new
species.
Compatibility (interfertility) is used to reveal relationships between
clones or 'species', but caution has to be exercised when interpreting the
findings. For example, the identification of strains provisionally identi-
fied as Chlamydonomas reinhardtii Dangeard is routinely tested by carrying
out crossability experiments with an authenticated laboratory clone.
Spanier et al. (1992) found that despite crossing experiments demonstrat-
ing partial fertility between some clones, these differed in a wide range of
non-morphological traits including heavy metal tolerance, protein com-
position, mitochondrial DNA length and nuclear, chloroplast and mito-
chondrial DNA restriction fragment length polymorphisms (RFLPs). The
majority of Chlamydomonas species cannot be tested by breeding experi-
ments because reproduction is known in less than 20% of described
species (Ettl and Schlosser, 1992).
5.4 THE PHYLOGENETIC SPECIES CONCEPT

This concept is favoured by the cladistic school of systematics with its
practitioners continuing to debate it although there seems general accep-
tance that 'species' are terminal taxa on their cladistic trees (Theriot, 1992).
Two types of phylogenetic species concept are recognized by cladists: one
identifies species as groups of organisms possessing at least one diagnos-
tic character (terminal taxa on cladistic tree), and the other focuses on
species as monophyletic and sharing one or more derived characteristics
(for a detailed discussion, see Nelson, 1989; Baum, 1992; Manhart and
McCourt, 1992; and references therein). In the context of species the term
monophyletic has at least two meanings: all the descendants of a common
ancestor together with that ancestor, or organisms more closely related to
one another than to any other. The argument among cladists revolves
around the question as to whether species are like any other taxa, or are
they in some way special. Cladistic studies of intrageneric relationships in
algae using morphological data tend to address relationships between
Role of molecular data 97
species groups rather than focusing on individual species (cf. Williams,
1985; Mrozinska, 1991,1993; Littler and Littler, 1992). A notable exception
to this is the detailed phylogenetic analysis of several genera of green sea-
weeds by a research group at Groningen in the Netherlands (e.g. Kooistra
et al, 1993; Bakker et al, 1995). A problem not unique to algae, but partic-
ularly acute in the group, is that the data on potentially important charac-
ters are insufficient to warrant rigorous statistical treatment. Organisms
with a higher level of organization than algae generally have a greater
number of more robust characters for defining species limits.
5.5 ROLE OF MOLECULAR DATA

Various types of molecular data provide a wealth of information with
which to address problems at all taxonomic levels in the major algal lin-
eages. Detailed analysis of genetic variation by protein sequencing,
immunological comparisons and enzyme (isozyme) electrophoresis have
been used to examine phylogenetic relationships and address species-
level problems. Enzyme electrophoresis has been used widely to provide
species markers, but interpretation remains problematical due to a lack of
information on the stability of enzyme loci and the effects of physiological
state and environment on their expression. For a brief review of algal
species problems addressed by protein-based comparisons, see Manhart
and McCourt (1992). These molecular approaches have been largely
superseded by recent advances in nucleic acid technology.
To date, the majority of studies involving the analysis of nucleic acids
have concentrated upon resolving higher-order relationships rather than
those at lower taxonomic levels. In addressing species problems, molecu-
lar characters are being used to test established species concepts (princi-
pally morphological), at least until 'the development of a "molecular
species concept" employing disembodied sequences as the sine qua non of
species definition' (Manhart and McCourt, 1992). Studies addressing
species-level problems have involved the analysis of nuclear, chloroplast
and mitochondrial DNA at different levels of sequence resolution using
RFLPs, randomly amplified polymorphic DNA (RAPD), DNA sequencing
and single copy DNA-DNA hybridization. This last method is useful for
testing evolutionary divergence (e.g. Cladophom; Bot et al., 1989a,b), but it
has been largely replaced by more powerful molecular techniques. Of the
four methods, RAPD is potentially the most valuable for population
analysis (Patwary et al., 1993) although it is beset with many technical dif-
ficulties that affect the reliability of the results (Dutcher and Kapraun,
1994). The methods of choice for algal researchers are RFLPs and sequenc-
ing, particularly of rRNA and rDNA genes (nuclear: 18S, 26S; chloroplast:
16S, 23S). The moderately fast-evolving spacer regions (ribosomal internal
transcribed spacers; spacer between large and small Rubisco subunits) are
becoming important sources of informative sequence variation in algae
now that more is known of evolution rates in different coding and non-
coding regions (Bakker et al., 1995).
The chloroplast DNA of land plants is highly conserved, whereas in the
more ancient algal lineages its astonishing diversity enables it to be used
for examining all levels of relationships from populational to those
between orders and classes (Freshwater et al., 1994). Chloroplast DNA
inheritance, like that of mitochondrial DNA, is usually uniparental and so
any boundaries do not necessarily correlate with those obtained by
nuclear DNA analysis. Mitochondrial DNA previously received little
attention in algae (Coleman and Goff, 1991) but as the genome has now
been better characterized (Boyen et al., 1994) it may be more widely used
in future.
Nuclear and plastid sequence data are now applied with increasing
frequency to resolving taxonomic problems in difficult groups where tradi-
tional morphological comparisons appear to have failed. However, nucleic
acid data are most informative in taxonomic decision-making when com-
bined with data on other types of character. For example, an unusually
complete analysis of the red algal genus Gymnogongrus (now split into
Gymnogongrus and Ahnfeltiopsis) involved comparative molecular, morpho-
logical, and culture studies (Maggs et al., 1992). In this study plastid DNA
RFLP analysis and nucleotide sequences for the Rubisco spacer region
revealed the existence of what appeared to be two distinct taxa in the
Atlantic, both going under the name Gymnogongrus devoniensis. These were
distinguished not only by RFLP analysis, probing and nucleotide sequence
data, but on vegetative, reproductive and life-history characters. The
approach enabled resolution of complex taxonomic problems in the genus,
including the relationship between populations with different life-histories,
and answered questions concerning its distribution in the Atlantic.
Congruence between molecular, morphological and other data types is
often regarded as evidence in support of the original morphospecies con-
cept. Conversely, the absence of congruence between datasets indicates
that the original concept was either too narrow or too wide. As might be
expected, there tends to be less consensus (although there may be no
greater conflict) between molecular and morphological data in relatively
simple algae where uncertainty frequently surrounds species and generic
boundaries due to the existence of few discrete and easily definable char-
acters. For example, an attempt was made by Kooistra et al. (1993) using
cladistics to discover whether morphological data supported the findings
of the molecular analysis of four genera of green algae belonging to the
same order. As the limits of morphologically well-defined species showed
little correlation with the results of cladistic analysis, the genera were
revealed to be paraphyletic. All the genera were seen to be artificial: none
was monophyletic and two were paraphyletic.
Role of molecular data 99
There are many examples where molecular data have been used to
define the limits of morphologically distinguished species. For example,
morphospecies of the green alga Microthamnion were tested by rigorous
experimentation in culture and revealed to be too narrow, so that several
previously recognized species were considered to be mere growth forms
or 'ecomorphs' (John and Johnson, 1987). RFLP analysis by John et al.
(1993) of 18S rRNA for 20 isolates from different habitats and geographical
regions showed that there was only one pattern; this supported the
culture-based view that Microthamnion is monospecific. Similarly Lange et
al. (1994) compared sequences of the same subunit in three colony-form-
ing species of the haptophyte Phaeocystis and discovered congruence
between the molecular, morphological, and physiological datasets. In the
brown seaweed, Costaria costata (Turner) Saunders, morphologically dis-
tinct wave-sheltered and wave-exposed populations were compared
using nuclear DNA RFLPs probed with an array of anonymous randomly
cloned fragments (Druehl and Saunders, 1992). Only one RFLP separated
the two populations; the results supported previous hypotheses that mor-
phological variation in this species reflects a plastic phenotypic response
to environmental variables.
Genomic relationships in Chlamydomonas have been examined by
Buchheim et al. (1990), who discovered close agreement between mor-
phological data and 18S rRNA partial sequences in 14 species. As in most
other algal groups, delimitation of the 500-plus recognized Chlamydomonas
species is possible only by the experienced specialist. In the past the pres-
ence of very specific sporangial wall autolysins was used by Schlosser
(1976) to distinguish species. The majority of these 'species' are so similar
in terms of cellular morphology, reproduction, development and physio-
logical properties that they are now considered to be clones of the same
species (Ettl and Schlosser, 1992).
If breeding and morphometric studies in diatoms continue to lead to
the separation of distinctive allopatric or sympatric morphotypes (cf.
Mann, 1989b; Mann and Droop, 1995), then the current species concept is
too wide or 'coarse-grained'. This view was supported by Medlin et al.
(1991) when describing a new species, Skeletonema pseudocostatum, on the
basis of small differences in morphology and nucleotide substitutions in
16S rRNA sequences among four clones originally attributed to
Skeletonema costatum, two of which were reassigned to the new species.
Medlin (1995) considers polymorphisms in rRNA sequence data among
morphologically indistinguishable clones to indicate the presence of a
'species complex'. The current view of the diatom species concept is sum-
marized by Cox (1993) as follows: 'While it may be possible to distinguish
taxa on certain non-morphological criteria,.... if the taxa cannot be readily
and routinely distinguished, formal description of a new species is large-
ly irrelevant.... If there is no significant detectable difference in ecology,
physiology, or reproductive behaviour of morphologically alike individu-
als or populations, they should be considered the same taxon'.
The red alga Gracilaria is an example of a genus in which systematic
revision required extensive molecular studies. The discreteness of
Gracilaria species defined on the basis of morphological criteria has been
tested by comparing morphological and molecular data (Goff and
Coleman, 1988; Bird and Rice, 1990). In several morphological species
there was close agreement between morphology and DNA data. For
example, Goff and Coleman (1988) demonstrated different RFLP patterns
for chloroplast DNA in Gracilaria pacifica and G. robusta. Rice and Bird
(1990) investigated various populations previously identified as G. verru-
cosa (Hudson) Papenfuss, a terete species with few taxonomically useful
morphological characters. They were able to demonstrate that a core
group of interfertile isolates had similar RFLP patterns, but outside this
group, various levels of difference in RFLP patterns were evident. The 'G.
verrucosa-like algae' included representatives of several other species of
Gracilaria and a separate genus, Gracilariopsis. The longstanding taxonom-
ic and nomenclatural confusion in this group has now been largely
resolved by use of molecular data in combination with traditional mor-
phological studies (Steentoft et al., 1995). Similarly, relationships within
the morphologically very variable red algal genus Gelidium have been clar-
ified using rbcL sequences (Freshwater and Rueness, 1994). Various
species complexes were recognized and this enabled a problematic species
to be correctly identified as G. attenuatum (Turner) Bornet.
5.6 CONCLUSIONS
The vast majority of algae are distinguished by morphological discontinu-
ities - the 'morphological species concept' dominates algal systematics.
Problems associated with the continued acceptance of traditionally recog-
nized 'morphospecies' or 'morphotypes' relate to disagreements over the
weighting of characters, the discovery of cryptic molecular variation and
to extreme levels of phenotypic plasticity. Breeding experiments have
enabled clusters of reproductively isolated cryptic or sibling species (syn-
gens) to be identified within traditional morphological species. Groups of
sibling species are equivalent to what have been termed 'species complex-
es' when morphologically indistinguishable clones show nucleotide sub-
stitutions in a reasonably highly conserved region of the genome. Sibling
species have generally not been afforded any taxonomic status: only when
characters are discovered that make it possible readily and routinely to
distinguish them are they formally diagnosed. The results of these com-
patibility experiments involving biparental, sexually reproducing algae
should perhaps be viewed as a practical guide to assist in interpreting
morphological variation and making taxonomic decisions. However, algae
Conclusions 101
are largely asexual so the biological species concept is often not applicable
and genomic relationships are untestable through a breeding programme.
Sexuality is regarded as a primitive characteristic by cladists who point out
that genomically very different clones or 'species' are often capable of
interbreeding. There is a considerable body of evidence from algae and
other organisms to indicate the occurrence of considerable evolutionary
divergence without the development of strong isolating mechanisms.
Compatibility data therefore need to be viewed with caution as they are
not necessarily always a good indication of the degree of taxonomic relat-
edness.
Systematists are increasingly applying molecular approaches to exam-
ine relationships in algae at different taxonomic levels. One advantage of
using molecular data is that what is measured represents differences in
the genome rather than the phenotype. Nucleic acid sequences are prov-
ing especially useful in testing the limits of morphological species and
detecting the existence of 'species complexes' within indistinguishable
clones. Sequence data are most powerful when combined with other
datasets to test the morphological species hypotheses. Inevitably congru-
ent datasets give more reliable and well-corroborated hypotheses for the
recognition of monophyletic taxa such as species. It is generally becoming
recognized that cladistics is a valuable tool for rigorously analysing these
datasets. It is important to recognize that it still fails to address the classi-
cal problem of how to prevent personal bias and a priori opinion influenc-
ing selection of phylogenetically informative characters and equally parsi-
monious cladistic trees. Ideally, different morphological and molecular
datasets should be treated in as similar a fashion as possible in order to
facilitate comparison (Williams, 1993). The 'phylogenetic species concept'
is favoured by practising cladists, although to date it has been applied
principally to diatoms where many putative natural groupings have been
revealed using cladistics (cf. Williams, 1985).
Determining discontinuities in morphological variation will undoubted-
ly remain the principal practical approach to species-level taxonomy in the
algae. In the future the traditional morphological species concept is likely
increasingly to operate alongside less formalized concepts involving data
from other disciplines (e.g. molecular systematics). These data will enable
larger suites of concordant characters to be used for calibrating species con-
cepts. If cladistics is accepted as the analytical method of choice for study-
ing species-level relationships then the 'phylogenetic species concept' will
come into more frequent use. Finally, we agree with the statement by
Wilmotte and Golubic (1991: 4) that 'for practical reasons, all these results
[molecular data] will finally have to feed back to the taxonomy based on
the morphology and simple testing methods, so that an improved system-
atic practice based on morphology will then be able to deliver fast and reli-
able determinations, as pressing ecological questions demand'.
Acknowledgements
We should like to thank Dr David Williams for bringing us up-to-date on
current debate on the 'phylogenetic species concept', Drs David Mann
and Stephen Droop for commenting upon the diatom section, and Dr
Judith John for reading the manuscript and offering valuable suggestions
for its improvement.
5.7 REFERENCES
Andersen, R.A. (1992) Diversity of eukaryotic algae. Biodiversity and Conservation, 1,
267-92.
Bakker, F.T., Olsen, J.L. and Stam, W.T. (1995) Evolution of nuclear rDNA ITS
sequences in the Cladopora albida/sericea clade (Chlorophyta). Journal of Molecular
Baum, D. (1992) Phylogenetic species concept. Trends in Ecology and Evolution, 2,1-3.
Bird, CJ. and Rice, E.L. (1990) Recent approaches to the taxonomy of the
Gracilariaceae (Gracilariales, Rhodophyta) and the Gracilaria verrucosa problem.
Hydrobiologia, 204/205,111-18.
Blackburn, S.I. and Tyler, P.R. (1987) On the nature of eclectic species - a tiered
approach to genetic compatibility in the desmid Micrasterias thomasiana. British
Phycological Journal, 22,277-98.
Eliding, C. (1963) A critical survey of European taxa in Ulvales. Part I. Capsosiphon,
Percursaria, Blidingia, Enteromorpha. Opera Botanica Lund, 8(3), 1-160.
Bold, H.C. (1974) Phycology, 1947-1972. Annals of the Missouri Botanic Garden, 61,
14-44.
Bot, P.V.M., Holton, R.W., Stam, W.T. and Hoek, C. van den (1989a) Molecular
divergences between North Atlantic and Indo-West Pacific Cladophora albida
(Cladophorales, Chlorophyta) isolates as indicated by DNA-DNA hybridiza-
tion. Marine Biology (Berlin), 102, 307-33.
Bot, P.V.M., Stam, W.T., Boele-Bos, S.A. Hoek, C. van den and Delden, W. van
(1989b) Biogeographic and phylogenetic studies in three North Atlantic species
of Cladophora (Cladophorales, Chlorophyta) using DNA-DNA hybridization.
Phycologia, 28,159-68.
Boyen, C., Leblanc, C., Kloareg, B. and Loiseaux-de Goer, S. (1994) Physical and
gene map organization of the mitochondrial genome of Chondrus crispus
(Rhodophyta, Gigartinales). Plant Molecular Biology, 26, 651-97.
Brook, AJ. (1997) A Revision of the West's British Desmidiaceae. Volume 1. The Ray
Society, London.
Buchheim, M.A., Turmel, M., Zimmer, E.A. and Chapman, R.L. (1990) Phylogeny
of Chlamydomonas (Chlorophyta) based on cladistic analysis of nuclear 18S
rRNA sequence data. Journal of Phycology, 26, 689-99.
Castenholz, R.W. (1992) Species usage, concept, and evolution in the cyanobacte-
ria (blue-green algae). Journal of Phycology, 28, 737-45.
Coleman, A.W. (1977) Sexual and genetic isolation in the cosmopolitan algal
species Pandorina morum. American Journal of Botany, 64, 361-8.
Coleman, A.W. and Goff, L. (1991) DNA analysis of eukaryotic algal species. Journal
of Phycology, 27, 463-73.
References 103
Corliss, J.O. (1994) An interim utilitarian ('user-friendly') hierarchical classification
and characterization of the protists. Ada Protozoologica, 33,1-51.
Cox, E.J. (1993) Diatom systematics - a review of past and present practice and a
personal vision for future development. Beiheft Nova Hedwigia, 106,1-20.
Druehl, L.D and Saunders, G.W. (1992) Molecular explorations in kelp evolution.
Progress in Phycological Research, 8,47-78.
Dutcher, J.A. and Kapraun, D.F. (1994) Random amplified polymorphic DNA
(RAPD) identification of genetic variation in three species of Porphyra
(Bangiales, Rhodophyta). Journal of Applied Phycology, 6, 267-73.
Edwards, P. (1970) Attempted hybridization in the red algal genus Polysiphonia.
Nature, 226,467-8.
Ettl, H. and Schlosser, U.G. (1992) Towards a revision of the systematics of the
genus Chlamydomonas (Chlorophyta). 1. Chlamydomonas applanata Pringsheim.
Botanica Acta, 105, 323-30.
Freshwater, D.W. and Rueness, J. (1994) Phylogenetic relationships of some
European Gelidium (Gelidiales, Rhodophyta) species, based on rbcL nucleotide
sequence analysis. Phycologia, 33,187-94.
Freshwater, D.W., Fredericq, S., Butler, B.S., Hommersand, M.H. and Chase, M.
(1994) A gene phylogeny of the red algae (Rhodophyta) based on plastid rbcL.
Proceedings of the National Academy of Science, USA, 91, 7281-5.
Friedl, T. and Zeltner, C. (1994) Assessing the relationships of some coccoid green
lichen algae and the Microthamniales (Chlorophyta) with 18S ribosomal RNA
gene sequence comparisons. Journal of Phycology, 30, 500-6.
Gartner, G. von (1985) Die Gattung Trebouxia Puymaly (Chlorellales,
Chlorophyceae). Archivfur Hydrobiologia, suppl. 71, 4,495-548.
Goff, L.J. and Coleman, A.W. (1988) The use of plastid DNA restriction endonu-
clease patterns in delineating red algal species and populations. Journal of
Phycology, 24, 357-68.
Grant, V. (1981) Plant Speciation, 2nd edn. Columbia University Press, New York.
Greville, R. K. (1821) Description of a new species of Fucus, found in Devonshire.
Memoirs of the Wernerian Natural History Society, 3, 396-9.
Guiry, M. (1992) Species concepts in marine red algae. Progress in Phycological
Research, 8, 251-78.
Hegewald, E. and Silva, P. (1988) Annotated catalogue of Scenedesmus and nomen-
claturally related genera, including original descriptions and figures. Bibliotheca
Phycologica, 80,1-587.
Hennig, W. (1966) Phylogenetic Systematics. University of Illinois Press, Urbana.
Hoek, C. van den (1963). Revision of the European Species of Cladophora. Brill, Leiden.
Hoek, C. van den (1964) Criteria and procedures in present-day taxonomy, in
Algae and Man (ed. D.F. Jackson), pp. 31-58.
Hoek, C. van den (1982) A taxonomic revision of the American species of
Cladophora (Chlorophyceae) in the North Atlantic Ocean and their geographic
distribution. Verhandelingen. Koninklijke akademie van wetenschappen Tweede Sectie,
78,1-236.
Hudson, W. (1762) Flora Anglica. London.
Ichimura, T. and Kasai, F. (1990) Mating systems and speciation in haplontic uni-
cellular algae, desmids, in Biological Approaches and Evolutionary Trends in Plants
(ed. H. Hara), Academic Press, London, pp. 195-201.
Irvine, L. M. and Dixon, P. S. (1982) The typification of Hudson' s algae: a taxo-
nomic and nomenclatural reappraisal. Bulletin of the British Museum of Natural
History (Botany), 10, 91-105.
Islam, A.K.M.N. (1963) A revision of the genus of the genus Stigeoclonium. Beiheft
Nova Hedwigia, 10,1-164.
John, D.M. (1994) Biodiversity and conservation: an algal perspective. The
Phycologist, 38,15-18.
John, D.M. and Johnson, L.R. (1987) Observations on the developmental mor-
phology, growth rate, and reproduction of Microthamnion kuetzingianum Nageli
(Pleurastraceae, Pleurastrales) in culture and a taxonomic assessment of the
genus. Nova Hedwigia, 44, 25-53.
John, D.M. and Johnson, L.R. (1989) A cultural assessment of the freshwater
species of Pseudendodonium Wille (Ulotrichales, Ulvophyceae, Chlorophyta).
Algological Studies, 54, 79-112.
John, D.M., Bhoday, R., RusseU, S.J., Johnson, L.R., and Gacesa, P. (1993) A molec-
ular and morphological analysis of Microthamnion (Chlorophyta,
Microthamniales). Archivfiir Protistenkunde, 143, 33-9.
Kadlubowska, J.Z. (1984) Conjugatophyceae I Zygnemales = Chlorophyta VIII, in
Susswasserflora von Mitteleuropa. G. Fischer, Stuttgart and New York.
Koch, C. (1986) Attempted hybridization between Polysiphonia fibrillosa and P. vio-
lacea (Bangiophyceae) from Denmark; with culture studies primarily on P. fib-
rillosa. Nordic Journal of Botany, 6,123-8.
Kociolek, J.P., Theriot, B.C. and Williams, D.M. (1989) Inferring diatom phylogeny:
a cladistic perspective. Diatom Research, 4,289-300.
Kooistra, W.H.C.F., Olsen, J.L., Stam, W.T. and Hoek, C. van den (1993) Problems
relating to species sampling in phylogenetic studies: an example of non-
monophyly in Cladophoropsis and Struvea (Siphonocladales, Chlorophyta).
Phycologia, 32, 419-28.
Krammer, K. and Lange-Bertalot, H. (1986) Bacillariophyceae 1. Teil: Naviculacae,
in Susswasserflora von Mitteleuropa (eds H. Ettl, J. Gerloff, H. Heynig and D.
Mollenhauer), vol. 2(1), G. Fischer, Stuttgart, New York, pp. 1-876.
Krammer, K. and Lange-Bertalot, H. (1988) Bacillariophyceae 2. Teil: Bacillariaceae,
Epithemiaceae, Surirellaceae, in Susswasserflora von Mitteleuropa (eds H. Ettl, J.
Gerloff, H. Heynig and D. Mollenhauer), vol. 2(2), G. Fischer, Stuttgart, New
York, pp. 1-596.
Krammer, K. and Lange-Bertalot, H. (1991a) Bacillariophyceae 3. Teil: Centrales,
Fragilariaceae, Eunotiaceae, in Susswasserflora von Mitteleuropa (eds H. Ettl, J.
Gerloff, H. Heynig and D. Mollenhauer), vol. 2(3), G. Fischer, Stuttgart, New
York, pp. 1-576.
Krammer, K. and Lange-Bertalot, H. (1991b) Bacillariophyceae 1. Teil:
Achnanthaceae. Kritische Erganzungen zu Navicula (Lineolatae) und
Gomphonema, in Susswasserflora von Mitteleuropa (eds H. Ettl, J. Gerloff, H. Heynig
and D. Mollenhauer), vol. 2(4), G. Fischer, Stuttgart, New York, pp. 1-437.
Lange, M., Medlin, L.K. and Baumann, M.E.M (1994) Genetic differentiation
among three colony-forming species of Phaeocystis: further evidence for the
phylogeny of the Prymnesiophyta. Phycologia, 33,199-212.
Lewin, R.A. (1975) A Chlamydomonas with black zygospores. Phycologia, 14, 71-4.
Linnaeus, C. (1753). Species Plantarum. Vols. 1, 2. Laurenti Salvii, Stockholm.
References 105
Littler, D.S. and Littler, M.M. (1992) Systematics of Avrainvillea (Bryopsidales,
Chlorophyta) in the tropical western Atlantic. Phycologia, 31, 375-418.
Maggs, C.A. and Hommersand, M.H. (1993) Seaweeds of the British Isles. Volume 1
Rhodophyta, Part 3A Cemmiales. Natural History Museum/HMSO, London.
Maggs, C.A., Douglas, S.E., Fenety, J. and Bird, C.J. (1992) A molecular and mor-
phological analysis of the Gymnogongrus devoniensis (Rhodophyta) complex in
the North Atlantic. Journal ofPhycology, 28,214-32.
Manhart, J.R. and McCourt, R.M. (1992) Molecular data and species concepts in the
algae. Journal ofPhycology, 28, 730-7.
Mann, D.G. (1989a) The diatom genus Sellaphora: separation from Navicula. British
Phycological Journal, 24,1-20.
Mann, D.G. (1989b) The species concept in diatoms: evidence for morphologically
distinct, sympatric gamodemes in four epipelic diatoms. Plant Systematics and
Evolution, 164, 215-37.
Mann, D.G. (1995) Gamodemes and genodemes in diatoms: the biological species
concept in action. The Phycologist, 40, 24-5.
Mann, D.G. and Droop, S.J.M. (1996) Biodiversity, biogeography and conservation
of diatoms. Hydrobiologia, 331, in press.
Botany, 79, 222-38.
McCourt, R.M. and Hoshaw, R.W. (1990) Noncorrespondence of breeding groups,
morphology, and monophyletic groups in Spirogyra (Zygnemataceae:
Chlorophyta) and the application of species concepts. Systematic Biology, 15,
69-78.
Medlin, L.K. (1995) Can molecular techniques enhance our understanding of a
species concept? The Phycologist, 40,25.
Medlin, L.K., Elwood, H.J., Stickel, S. and Sogin, M.L. (1991) Morphological and
genetic variation within the diatom Skeletonema costatum (Bacillariophyta): evi-
dence for a new species, Skeletonema pseudocostatum. Journal of Phycology, 27,
514-24.
Mrozinska, T. (1991) A preliminary investigation of the taxonomical classification
of the genus Oedogonium Link (Oedogoniales) based on the phylogenetic rela-
tionship. Archivfiir Protistekunde, 139, 85-101.
Mrozinska, T. (1993) A preliminary investigation of the taxonomical classification
of the genus Bulbochaete Agardh (Oedogoniales, Chlorophyta) based on the
phylogenetic relationship. Archivfiir Protistenkunde, 143,113-23.
Necchi, O., Jr and Zucchi, M.R. (1995) Systematics and distribution of freshwater
Audouinella (Acrochaetiaceae, Rhodophyta) in Brazil. European Journal of
Phycology, 30, 209-18.
Nelson, G. (1989) Species and taxa: systematics and evolution, in Speciation and
Patwary, M.U., MacKay, R.M. and van der Meer, J.P. (1993) Revealing genetic
markers in Gelidium vagum (Rhodophyta) through the random amplified poly-
morphic DNA (RAPD) technique. Journal ofPhycology, 29,216-22.
Preisig, H.R. and Hibberd, D.J. (1982) Ultrastructure and taxonomy of
Paraphysomonas (Chrysophyceae) and related genera 2. Nordic Journal of Botany,
2,601-38.
Prescott, G.W. (1951) History of phycology, in Manual ofPhycology: An Introduction
to the Algae and their Biology (ed. G.M. Smith), Chronica Botanica, Massachusetts,
pp. 1-11.
Pringsheim, E.G. (1967) Phycology in the field and in the laboratory. Journal of
Phycology, 3, 93-5.
Rice, E.L. and Bird, CJ. (1990) Relationships among geographically distant popu-
lations of Gracilaria verrucosa (Gracilariales, Rhodophyta) and related species.
Phycologia, 29, 501-10.
Russell, G. (1994) A Baltic variant of Pilayella littoralis (Algae, Fucophyceae). Annales
Botanici Fennici, 31,127-38.
Russell, G. and Fielding, A.J. (1981) Individuals, populations and communities, in
The Biology of Seaweeds (eds C.S. Lobban and M.J. Wynne), Blackwell Scientific,
Oxford, pp. 393-420.
Schlosser, U.G. (1976) Entwicklungsstadien -und sippenspezifische Zellwand-
Autolysine bei der Freisetzung von Fortpflanzungszellen in der Gattung
Chlamydomonas. Bericht der Deutschen Botanischen Gesellschaft, 89,1-56.
Silva, P. C. (1952) A review of nomenclatural conservation in the algae from the
point of view of the type method. University of California Publications in Botany,
25,241-324.
Simons, J., Beem, A.P. van and Vries, P.J.R. de (1986) Morphology of the prostrate
thallus of Stigeoclonium (Chlorophyceae, Chaetophorales) and its taxonomic
implications. Phycologia, 25, 210-20.
Spanier, J.G., Graham, J.E. and Jarvik, J.W. (1992) Isolation and characterization of
three Chlamydomonas strains interfertile with Chlamydomonas reinhardtii. Journal
ofPhycology, 28, 822-8.
Steentoft, M., Irvine, L.M. and Farnham, W.F. (1995) Two terete species of Gracilaria
and Gracilariopsis (Gracilariales, Rhodophyta) in Britain. Phycologia, 34,113-27.
Steussy, T.F. (1990) Plant Taxonomy The Systematic Evaluation of Comparative Data.
Tehler, A. (1994) Cladistic analysis in ascomycete systematics: theory and practice,
in Ascomycete Systematics: Problems and Perspectives in the Nineties (ed. D.L.
Hawksworth), Plenum Press, New York, pp. 185-97.
Theriot, E. (1989) Phylogenetic systematics for phycology. Journal of Phycology, 25,
407-11.
Theriot, E. (1992) Phylogenetic systematics, the theory and practice of taxonomy,
in Modern Approaches to the Analysis of Algal Morphology and Systematics (eds
D.M. John and D. Garbary), vol. 8, Progress in Phycological Research, pp.
180-207.
Theriot, E., Stoermer, E.F and Hakansson, H. (1987) Taxonomic interpretation of
the rimoportula of freshwater genera in the centric family Thalassiosiraceae.
Diatom Research, 2, 251-65.
Trainor, F.R. (1991) The format for a Scenedesmus monograph. Algological Studies,
61,47-53.
Trainor, F.R. and Egan, P.P. (1991) Discovering the various ecomorphs of
Scenedesmus. Archivfur Protistenkunde, 139,125-32.
Vis, M.L., Sheath, R.G. and Entwistle, T. J. (1995) Morphometric analysis of
Batrachospermum section Batrachospermum (Batrachospermales, Rhodophyta)
type specimens. European Journal of Phycology, 30, 35-56.
References 107
Williams, D.M. (1985) Morphology, taxonomy and inter-relationships of the ribbed
araphid diatoms from the genera Diatoma and Meridian (Diatomaceae:
Bacillariophyta). Bibliotheca Diatomologia, 8,1-228.
Williams, D.M. (1992) Analytical techniques for systematic molecular data. Progress
in Phycological Research, 8, 209-49.
Williams, D.M. (1993) Diatom nomenclature and the future of taxonomic databas-
ing studies. Beiheft Nova Hedwigia, 106, 21-31.
Wilmotte, A. and Golubic, S. (1991) Morphological and genetic criteria in the tax-
onomy of Cyanophyta/Cyanobacteria. Algological Studies, 64,1-24.
The species concept in lichens
0. W. Purvis
Contacting address: Department of Botany, The Natural History Museum, Cromwell
ABSTRACT
Lichens are a combination of at least one fungus and an alga or
cyanobacterium (blue-green alga). The formation of the lichen body
(thallus), typically morphologically, physiologically and biochemically
distinct from either partner in the free living state, is one of the greatest
enigmas in biology. The classification of lichens at a species level is dis-
cussed on a historical basis and evaluated in relation to the varying
importance attached to external morphology and internal anatomy of
fungal tissues, the significance of the alga in morphogenesis, and the
importance of sexual reproductive structures, asexual vegetative
propagules and reproductive strategies. Environmental factors are
shown to have a profound influence on phenotypic expression and
careful field evaluation must therefore be undertaken when describing
new species. The wide range of chemical characters including sec-
ondary metabolites are considered as an important tool for defining
species. The difficulties in maintaining the composite organism under
laboratory conditions have resulted in little experimental attention
being applied to the nature of species in lichen fungi, artificial crosses
being impossible for technical reasons, although recent attempts have
been made to study speciation using chemical markers. Molecular tech-
niques have as yet hardly been applied at the specific level in lichen sys-
tematics and species specific markers have yet to be developed. The
concept of a lichen as an individual is explored with reference to fusions
between individuals within a lichen population, species within a genus
or species in different genera, which results in the incorporation of
different fungal and algal partners within the same thallus. Although
documentations of thallus fusions are rare, these pose important conse-
quences for experimental work and also in classification. Personal expe-
rience and intuition continue to play a major role in defining species.
The morphological species concept, based on phenetic characters,
including secondary chemistry, remains therefore of major importance.
This must lead the work of molecular systematists.
110 The species concept in lichens
6.1 HISTORICAL ASPECTS
Since lichens were first recognized as members of the genus Lichen by
Tournefort in 1694, many schemes of classification have been outlined
(Smith, 1921; Zahlbruckner, 1926; Lorch, 1988). Linnaeus unfortunately
did not understand the dual nature of lichens and regarded cryptogams
generally as 'rustici pauperrimi' - the poor little peasants of nature. He
described only 109 species of lichens, mostly placed in his genus Lichen but
some were mistakenly treated within the 'Algae' (J0rgensen et al., 1994).
That this great pioneer of phanerogamic systematics did not understand
these organisms no doubt greatly contributed to the study of lichenology
and of cryptogams as a whole being greatly held back. It was not until
Acharius first made lichens a subject of special study by his scientific sys-
tem of classification in which he introduced a new terminology for
lichenology in his Methodus Lichenum (1803), when he raised them to the
rank of the classes of the other great classes of plants, describing 906
species (Smith, 1921). He and other authors to the mid-19th century based
nearly all taxonomy on externally visible characters, such as general mor-
phology, colour, hairiness, the external shape of apothecia, and the pres-
ence of a range of structures specific to lichens including several asexual
reproductive structures (Almborn, 1965). An increased value was placed
on anatomical characters by particularly Massalongo and Korber towards
the mid-19th century. However, Nylander, although describing a very
large number of species unwittingly put the clock back by establishing a
classification based almost entirely on thalline form overlooking the
importance of characters of the fruiting bodies or ascomata (see section
6.3.1 (a)), contrary to the situation in mycology where the thallus is usual-
ly not visible. Following the proposal of the composite nature of the lichen
thallus by Schwendener in 1868, there was a prolonged and often heated
debate over the dual nature of lichens which continued for about 50 years
(Lorch, 1988). Today it is generally accepted that a lichen consists of at
least one fungal (myco-) and one algal (phyto-) biont and that licheniza-
tion is an extremely successful life strategy for fungi that is taken up by
almost 50% of all ascomycete and 20% of all known fungi (Hawksworth,
1988b). There have been various fashions in assigning a taxon to a specif-
ic rank. During the 19th and early part of the 20th century, many genera
were based on relatively few characters and many infraspecific taxa were
recognized. This trend has been dramatically reversed during the past 30
years, with crustose genera such as Lecidea Ach. originally united on a sin-
gle spore character, having been divided into numerous segregates
including the resurrection of several originally described, e.g. Psilolechia
Massal. (1850). For instance the current checklist of lichens of Great Britain
and Ireland accepts 42 species of Lecidea, a further 80 species since the pre-
vious 1980 checklist being assigned to 30 other genera (Purvis et al., 1994).
Conversely, the numbers of infraspecific taxa recognized has dramatically
What are lichens? Ill
decreased following realization of the importance of environmental fac-
tors on phenotypic expression (see section 6.5). Estimates of known lich-
enized taxa have varied from a conservative 13 500 (Hawksworth et al,
1983) to 17 000 (Hale, 1974). Britain is one of the most intensively surveyed
countries in the world for lichens and yet since publication of the recent
Lichen Flora of Great Britain and Ireland (Purvis et al, 1992) including 1700
taxa, over 50 additional species have been added within 2 years. Galloway
(1992) draws attention to the large number of recent monographic revi-
sions where new taxa are discovered in even quite well-known
macrolichen genera and suggests the world total may be closer to 20 000.
Although for vascular plants the tropics are regarded as major sites of bio-
diversity, much less is known about tropical lichens whose biodiversity
tends to be richest in canopy vegetation, which is still very poorly sampled
in many tropical areas. Furthermore, as crustose lichens have been so
inadequately studied even in many temperate areas with their wide vari-
ations of habitat, geology and climate, a conservative estimate of the total
number of species would seem to be 30 000 and this would not encompass
chemical variation considered by some to indicate sibling speciation (see
section 6.6).
6.2 WHAT ARE LICHENS?

In simple terms a lichen is an association between a fungus and one or
more photosynthetic partners (photobiont) which may be an alga or a
cyanobacterium (blue-green alga). A precise definition of a lichen,
although much debated, has not yet been satisfactorily resolved
(Hawksworth, 1988a, 1994; Ahmadjian, 1993a) due in part to the great
diversity of fungal-algal associations (Table 1 in Hawksworth, 1988a)
exhibiting a range of biological strategies from apparently mutualism to
antagonism where up to five bionts or possibly more may be involved. A
key feature of the symbiosis is the formation of the lichen body or thallus
within which photobionts are housed to best advantage for photosynthe-
sis, although the two organisms behave from a physiological point of view
rather as a single organism subjected to natural processes of selection. This
successful mode of nutrition for the fungus, taken up by widely different
groups of fungi, has enabled these organisms to colonize habitats
unfavourable to most plants, contributing a significant component of bio-
diversity in many of the world's ecosystems. The association, although tra-
ditionally considered to be of mutual benefit to both partners, appears to
represent more of a controlled parasitism on behalf of the fungus at the
level of the individual cells (Smith and Douglas, 1987; Kappen, 1994). The
challenge of maintaining the composite organism under laboratory condi-
tions has resulted in little experimental attention being applied to the
nature of species in lichenized fungi as compared with free-living fungi
which are generally easier to culture, artificial crosses being impossible for
technical reasons. However, recent significant advances in tissue culture
techniques (Yamomoto et al., 1993; Yoshimura et al.r 1994) and resynthesis
studies (Yoshimura et al, 1993, 1994; Stocker-Worgotter and Turk, 1994;
Crittenden et al., 1995) provide many exciting opportunities to test the
specificity of particular bionts, their combined influence on morphogene-
sis and role in speciation as currently understood.
Given the great complexities of lichen thalli, on what basis are species
defined? Almborn (1965) stated that it is essential that lichen taxonomy be
founded on firm principles not deviating from those generally accepted in
other plant groups. Firm principles are clearly important, but to what
extent should we be examining other plant groups? These aspects are now
considered in relation to key developments in modern systematic
research.
6.3 CURRENT TAXONOMIC CONCEPTS

6.3.1 Fungal partner (mycobiont)
The mycobiont is an obligate symbiont unable to occur free-living other
than as a reproductive propagule; the reasons for the dependence of the
fungus on the alga are not known. The mycobiont is typically unique to
each lichen and the Code of Botanical Nomenclature rules that 'for nomen-
clatural purposes names given to lichens shall be considered as applying
to their fungal component'; [Article 13.1 (d), Greuter, 1994]. Thus, it follows
that those characters derived from the fungus are regarded as most impor-
tant taxonomically. Herein lies a problem - what are the characters that
are derived from the fungus? In general a lichen thallus receives its shape
from a mycobiont (Jahns, 1988) but it is still unclear what processes regu-
late the formation of the specific thallus which allows us to classify the
lichen taxonomically. In pure culture many mycobionts remain relatively
featureless, mostly lacking the delimitation of specific layers and struc-
tures, so to a certain extent the key to this must be the influence of the
photobiont on thallus morphogenesis (see section 6.3.2). Most lichen fungi
are ascomycetes; only a few form associations with basidiomycete or
deuteromycete fungi. Non-lichenized fungi, however, typically lack a
perennial thallus and fruiting bodies (ascomata). During the past 30 years
lichen taxonomists have increasingly become aware of the importance of
mycological characters, notably anatomical structures of the thallus and
reproductive structures such as ascomata (see below) that provide a more
natural means of delimiting genera mainly based on spore characters
adopted by Zahlbruckner (1926a). In view of the polyphyletic origin of
lichen fungi (Gargas et al, 1995a) it is hardly surprising that different char-
acters are diagnostic at a species level according to the group of fungi.
Current taxonomic concepts 113
(a) Reproductive fungal characters
The type and basic morphology of ascomata, the sexual fruiting body in
the ascomycetes ('sac fungi') containing sac-like asci in which spores are
formed, are useful characters at a specific and also at higher taxonomic
levels. Generic and also higher level taxonomic concepts in fungi forming
crustose lichens are nowadays based on microscopic features of the asco-
mata and asci, including their ontogeny - characters which have only
exceptionally been critically examined in the foliose and fruticose lichens
- which partly explains the reasons for recent segregates of several
macrolichen genera including foliose and fruticose representatives of
Teloschistaceae by Karnefelt et al. (1992), although more critical studies are
needed in many cases, especially within the genus Parmelia Ach. sensu lato
(Hawksworth, 1994).
A major problem facing lichenologists is our scant knowledge of many
details of lichen reproduction, which has not been directly observed in
nature or the laboratory and it is assumed that what happens in free-
living ascomycetes has a parallel in lichenized ascomycetes, although this
need not necessarily be true. However, indirect evidence suggests that
sexual reproduction does occur, as shown by the extreme variability of
some crustose lichen species which form interlocking mosaics (Poelt, 1994)
and from laboratory investigations including isozyme and molecular stud-
ies (see section 6.6). During the early stages of ascomatal development in
some species specialized cells (trichogynes) are produced which are
assumed to be able to receive conidia (spermatia) produced by special
reproductive structures producing conidia (pycnidia) of the same lichen
and probably also from other lichens of the same species (Culberson and
Culberson, 1994). Formed through the fusion of dikaryotic cells, the ascus
is the characteristic cell of ascomycetes which brings about karyogamy
(the fusion of two sex nuclei after cell fusion), undergo genetic recombi-
nation through meiosis, and during ascosporogenesis divide mitotically to
form ascospores, which are subsequently liberated (Bellemere, 1994). Data
are sparse concerning the dikaryotization process (where cells having two
genetically distinct haploid nuclei fuse), and although spermatia have
been observed in contact with trichogynes in Cladonia furcata (Huds.)
Schrader (Honegger, 1984), the sexual process of dikaryotization has not,
however convincingly been seen in nature or the laboratory (Ahmadjian
and Jacobs, 1987). There is no evidence for the presence of the parasexual
cycle occurring in lichens where plasmogamy, karyogamy and hap-
loidization result in genetic recombination during mitosis, although the
presence of thallus fusions (see section 6.7) offer intriguing possibilities
that genetic exchange between different mycobionts may occur, surely
one of the most challenging questions which molecular biologists need
address as an urgent priority.
The relative constancy of ascus types linked to ontogenetic differences
in the ascomata are considered to be the result of a long evolution, sug-
gesting that they can be usefully used to define taxa at a higher level
(Bellemere, 1994). The special types of ascus apex structures, and particu-
larly their iodine reactions, has recently been used to rearrange the classi-
fication of lichens (Hafellner, 1984). There is no doubt that studies of large
unwieldy heterogeneous genera in the past 10 years such as Lecidea and
Lecanora Ach. have benefited from the use of this additional character.
However, there has not been universal acceptance for its use and
Bellemere (1994) concludes that ascus characters are of secondary impor-
tance in ascomycete systematics. He also points out technical problems in
interpreting observations made only by light microscopy on the 1+ blue
patterns which may reflect the position of certain polysaccharides. There
is clearly here a need for more standardized techniques as well as detailed
ontogenetic studies using TEM.
Ascospores, the reproductive cell in ascomycete fungi, are important
species characters varying greatly in lichens in shape, size, structure, num-
ber per ascus, septation and may be colourless, greenish or brown.
Ascospore septation is nowadays considered diagnostic at a specific level
thus species in recent genera such as Vezdaea Tscherm.-Woess and Poelt
may contain simple or varyingly septate ascospores with 8- to 16-spored
asci (Purvis et at, 1992).
Conidiomata, specialized multi-hyphal structures bearing conidia
occur as mostly flask-shaped pycnidia in lichens. Different types of pyc-
nidia relating to hyphal arrangement are recognized which are useful
characters at a generic and also specific levels (Vobis, 1980). For instance,
Ramalina siliquosa (Huds.) A.L. Sm. and R. cuspidata (Ach.) Nyl. have dif-
ferently coloured pycnidia and thus may be conveniently distinguished
in the field on this basis (Purvis et al, 1992). Conidia vary in shape, size
and septation and are particularly useful characters at the specific level as
in Opegrapha. Some species produce more than one type of conidium as
in M. denigrata (Fr.) Hedl. producing three types (Coppins, 1983) and the
function of the different types has yet to be conclusively established (see
discussion above). Although conidia may be formed in a range of widely
different structures), they have not been recorded in all lichenized fungi.
The ability of lichens to produce pycnidia varies widely, even within the
same genus. For example, Lecanactis abietina (Ach.) Korber produces many
pycnidia and rarely fruits, while others such as L. subabietina Coppins and
P. James have abundant pycnidia but fruiting bodies are unknown.
Specialized asexual conidia termed thalloconidia occur in 17 species with-
in the genus Umbilicaria Hoffm. as well as in other, including crustose
lichen genera (Poelt and Obermayer, 1990; Hestmark, 1991; Poelt, 1991).
These have been interpreted as hypothalline, hyphomycetous
anamorphs occurring attached to the teleomorph. Differences in struc-
ture of these are useful at a specific level.
Current taxonomic concepts 115
6.3.2 Lichen algae (photobionts) and photomorphs
Most algae occurring as photobionts belong to the class Chlorophyceae, to
a lesser extent the Cyanobacteria (Cyanophyceae or blue-green algae) and
rarely the Xanthophyceae (yellow-green algae) with most occurring with-
in three genera: Nostoc, Trentepohlia and Trebouxia (Tschermak-Woess,
1988). The number of species currently identified as photobionts has been
estimated to include about 100 green algae in relatively few (about 23)
genera, and an uncertain, but much lower number of cyanobacterial
species in some 15 genera (Tschermak-Woess, 1988; Hawksworth and
Honegger, 1994). However, given the woefully little attention paid to the
taxonomy of these organisms many statements and/or identifications,
particularly in the older literature are questionable. As reported in this
volume (John and Maggs, 1997: Chapter 5) there are many problems sur-
rounding species concepts in green algae belonging to the order
Chlorococcales, e.g. Trebouxia, with an additional problem in lichens being
the algae are modified through symbiosis resulting, for instance, in a sup-
pression of sexual reproduction and the formation of filaments. The pop-
ular perception that the photobiont has little taxonomic value in lichens
has no doubt further contributed to the lack of study of these organisms
(Ahmadjian, 1993b). Ahmadjian argued that the type of photobiont select-
ed by a mycobiont might legitimately be considered a taxonomic charac-
ter along with thallus or secondary compound characteristics. However,
several studies have shown that mycobionts are not very selective in their
choice of photobionts. For example, some crustose lichens may contain
different photobionts as Chaenotheca chlorella (Ach.) Mull. Arg. lichenized
with a trebouxioid alga in Costa Rica rather than Stichococcus with which
it is associated in more northerly areas of the northern Hemisphere (Tibell,
1982). On the other hand Tschermak-Woess (1995) examined some 150
thalli of Phlyctis argena (Sprengel) Flotow from Europe and North America
concluding that there was strong selectivity for the photobiont
Dictyochloropsis splendida Geitler since all cultures were identical in both
morphology and development. Many lichen species contain different
genera of algae within their thallus. Thus species with green algal photo-
bionts may also contain accessory cyanobacteria normally localized in dis-
tinct structures termed cephalodia which are considered valuable species
characters in several genera including Peltigera (Vitikainen, 1994) and
Stereocaulon (Lamb, 1977). Other photobiont differences have also been
used to distinguish between species and certain genera. For example,
Staurothele Norman is unusual in having algae within the spore-bearing
layer of the fruiting bodies (the hymenium). These hymeneal algae differ
morphologically from the algae within their thalli and also between
species, ranging from cuboid to cylindrical and are considered diagnostic
species characters. The pattern of arrangement of algae within fungal
tissues is also an important species character in many lichen genera,
including Acarospora (Clauzade and Roux, 1981) and Thelotrema (Salisbury,
1971). In other cases closely related species containing different photo-
bionts have been assigned to different genera, as for instance the separa-
tion of lonaspis Th. Fr. with a Trentepohlia photobiont from Hymenelia
Krempelh. with a Trebouxia photobiont; these are currently being evaluat-
ed (P.M. J0rgensen, personal communication).
However, while the fungus has traditionally been regarded as playing
the major role in thallus formation since the early decades of this century,
it has been recognized that the photobiont nevertheless also plays a sig-
nificant role. The first to describe composite lichen thalli, with two mor-
phologically different structures, attached together, were the pioneering
studies of Dughi during the 1930s and 1940s on Dendriscocaulon Nyl.
(James and Henssen, 1976). In one, the mycobiont was associated with a
green alga and had a leaf-like (foliose) morphology; in the other with a
cyanobacterium it had a shrubby (fruticose) morphology. The different
morphotypes were also found as independent thalli. The differences in
morphology, ecology, distribution and occasionally chemistry of the two
independent morphotypes can be so striking that taxonomists at times
have even ascribed the unattached entities to different genera. In such
cases, the true relationship between the two morphs has only been eluci-
dated by the discovery of composite thalli as, for example, in Sticta felix
(Rausch.) Nyl. and Dendriscocaulon. This phenomenon has since been
observed in many species, particularly occurring within the Peltigerales
and Pannariaceae and are usually termed photosymbiodemes or photo-
biont morphotypes (Hawksworth, 1994) or, more recently, photomorphs
(Laundon, 1995). Although histological observations of hyphal continuity
suggests the same fungus to be present (James and Henssen, 1976), several
authors have suggested they might not be identical, given the frequency
with which taxonomically unrelated lichens associate with one another
(Hawksworth, 1988). Artificial resynthesis experiments offer one opportu-
nity to establish whether the same fungus occurs in both green- and
cyanobacterial photomorphs; however, this is not always possible in the
case of the genera Lobaria (Schreber) Hoffm., Peltigem, Pseudocyphellaria
Vainio and Sticta (Schreber) Ach. since these mycobionts are difficult to
grow in laboratory culture (Ahmadjian, 1993a). Armaleo and Clerc (1991)
therefore applied molecular methods to define the relatedness between
paired fungal components in the photomorphs. Using Southern
hybridization and the polymerase chain reaction they demonstrated a
genetic near-identity of the mycobionts in two photomorphs of
Pseudocyphellaria and Sticta (Figure 6.1). While they concluded that paired
mycobionts within two photomorphs should be assigned to the same
species, they pointed out that detailed quantitative definitions of 'related-
ness versus molecular variation' in lichen mycobionts will need the inclu-
sion of additional taxa and data. Further direct evidence that the green
Current taxonomic concepts \\*j
thallus and cephalodiate structures are formed by the same mycobiont has
recently been obtained by artificial resynthesis of the photosybiodeme
Pdtlgera leucophlebm (Nyl.) Gyelnik from the one fungus and two alga"
S
partners (Stocker-Worgotter and Turk, 1994).
According to the International Code of Nomenclature, which restricts the
name of the lichen to the mycobiont, in those lichens with paired mvco-
bionts priority is given to the first published name. There is therefore no
indication of which photomorph is implied when the name alone is used
The use of a single name can therefore result in a loss of ecological and
Kgure 6.1 Variations in morphology within the lichen species, Sticta canariensis,
containing the same fungus, but different photobionts. (a) leaf-like growth form
contaming cyanobacteria; (b) shrub-like growth form containing cyanobacteria;
(c) leaf-like growth form containing green alga; (d) composite thalli with lobules
contaming green algal morph joined to lichen containing cyanobacteria. Scale
~™ I*"' AgrS (a~C)/ AlaSadi?os'Terceira' A^res; Purvis, James and Smith
); (d) from Galloway, SW Scotland, James (BM).
geographical information about the morphs unless qualified by a very
cumbersome communication system. Thus, for example, in the Azores the
blue-green photomorph of Sticta canariensis ('S. dufourii') is frequent while
the green algal photomorph ('S. canariensis') is extremely rare and poorly
developed on the island of Pico, though locally abundant on the islands of
Terceira and Flores (O.W. Purvis and P.W. James, unpublished data, see
Figure 6.1). There is therefore a good case to be made for maintaining such
morphs as distinct taxa, which are not only morphologically dissimilar but
may also be ecologically, geographically and physiologically distinct
(J0rgensen, 1991) perhaps best accomplished at a subspecific level
(Laundon, 1995).
Many lichen species contain different algae during different stages of
their life-cycles. Psoroma durietzii P. James and Henss. is unusual in devel-
oping sorediate cephalodia (James and Henssen, 1975) capable of forming
small, independent, structured thalli containing only a blue-green photo-
biont (Nostoc) which can secondarily capture a coccoidal green alga. In
Solorina crocea (L.) Ach. fungal propagules lacking algae form initially an
association with cyanobacteria and latterly with coccoidal green algae
resulting in a layered thallus (Jahns, 1987). Lichenicolous lichens may also
associate with different photobionts at different stages of their life-
histories. For example, in Diploschistes muscorum (Scop) R. Sant. where the
thallus starts as a parasite on squamulose thalli on different species of
Cladonia and is first associated with the photobiont of the Cladonia,
(Trebouxia irregularis), but later exchanges it for another species of the same
green algal genus, T. showmanii (Friedl, 1987).
6.4 VEGETATIVE CHARACTERS, ASEXUAL PROPAGULES AND

'SPECIES PAIRS'
Most non-lichenized fungi consist of a diffuse mycelium below ground
and seasonal ascomata with usually hardly any vegetative characteristics
and hence lacking in important taxonomic characters. In the lichenized
ascomycetes vegetative structures are mostly evident, highly complex and
it is hardly surprising therefore that great significance has been applied to
these, particularly at the species level. Although gross morphology, pig-
mentation and anatomy have traditionally been used at specific and even
generic levels, recently even fruticose and crustose lichens have been rec-
ognized as being closely related as, for example, the crustose lichens,
Toninia leucophaeopsis (Nyl.) Th. Fr. and Bilimbia tornense Magnusson which
share a similar chemistry and apothecial anatomy with the shrub-like
Stereocaulon and are now placed in that genus (Purvis and James, 1985). In
many macrolichens the great diversity of gross morphological features as
compared with crustose lichens, has often led to a lack of critical attention
paid to ascomatal anatomy and ontogeny, e.g. in the genus Parmelia and
Vegetative characters, asexual propagules and 'species pairs' 119
its segregates. There exists wide variation in the taxonomic importance
given to different characters between genera. In some genera ascospores
provide useful characters, whereas in Usnea Hill these are apparently
remarkably uniform and not useful at a specific level, although detailed
studies involving statistical analyses of length : breadth ratios are lacking.
However, the wide range of asexual reproductive structures including
soredia, isidia and schizidia are routinely used to differentiate between
taxa at a species level in many lichen genera, including Usnea (Clerc, 1987).
Since production and morphology of soralia correlates with populations
of Usnea having distinct chemical, geographic and other morphological
properties, they are considered to be genetically determined. However,
caution is required in interpreting such characters as these characters may
also be influenced by environmental factors, thus emphasizing the impor-
tance of carrying out field investigations when describing new taxa.
Species concepts have traditionally been partly based on sexual strategies
and characters associated with reproductive structures. In those lichens
where no fruiting bodies are known to exist, ascospores and ascomatal
characters cannot be used and morphological and chemical characters are
therefore crucially important. Only relatively few lichens are regularly
able to simultaneously propagate by ascospores and in addition by some
type of sexual diaspore. One example is the well-known Baeomyces rufus
(Huds.) Rebent., a successful colonizer of eroded soil which develops
apothecia, schizidia and soredia at the same time (Jahns, 1987). In many
cases, however there exists a continuum between fertile species and sore-
diate counterparts as in Catillaria pulverea (Borrer) Lettau, which may
occur as rarely fertile, continuously granular, sorediate thalli (Catillaria pul-
verea) and thin, frequently non-sorediate, richly fertile thalli (C. albocincta
Degel.), but which as these intergrade are now considered a single species
(Purvis and James, 1993). Where lichens have sexual and sorediate phases
these entitities have also occasionally been accorded specific status. For
example, in the genus Peltigera Willd. where P. spuria (Ach.) DC lacking
soredia, bearing small erect fertile lobes was regarded as distinct from
sorediate morphs ('P. erumpens (Taylor) Elenkin/) and morphs bearing
apothecia and soralia on the same thallus as ('P. hazslinszkyi Gyeln.') but
which are now collectively regarded as the single species Peltigera didacty-
la (With.) Laundon (Vitikainen, 1994).
Many cases are now known of 'Artenpaare' or species pairs, one of
which is fertile with ascomata (a 'primary species') and the other non-
fertile, bearing isidia or soredia (a 'secondary species'); these pairs do not
intergrade and appear to be genetically isolated, unable to exchange genes
(Tehler, 1982). The character sorediate versus non-sorediate is often used
as a species criterion, provided that the taxa share different geographic
distributions. Vegetative, asexual reproduction by means of soredia, isidia
or other vegetative diaspores has the clear advantage of both fungal and
algal partners being distributed simultaneously without a need for resyn-
thesis in an environment where there may be few suitable algal hosts. As
a rule the sorediate element is more widely distributed, as shown in pol-
luted environments where there is typically a loss of sexuality, and many
islands have disproportionately fewer fertile species than on the adjacent
mainland. It has been suggested that the sorediate species might be
formed in response to unfavourable conditions; thus, species at the edges
of their ranges might be expected to be sorediate. Asexual reproduction
apparently also has its disadvantages, however, as the organism also pre-
sumably loses its genetic flexibility and can no longer adapt to new or
changing environments through gene recombination. The situation may
well be far more complex than we currently believe with not only sexual
and sorediate counterparts, but possibly also various asexual counterparts
with alternative asexual propagules. In the genus Menegazzia, M. eperforata
P. James and D.J. Galloway is thought to be the isidiate counterpart of the
sorediate M. nothofagi (Zahlbr.) P. James and D.J. Galloway which is relat-
ed to the primary sexual species M. prototypica P. James (James and
Galloway 1992; see Figure 6.2 for further explanation). Arguments to rou-
tinely relegate secondary 'species' to form rank are not now generally
accepted (Hawksworth, 1994; Poelt, 1994), but the existence of such pairs
or groups of organisms provides a valuable opportunity to study specia-
tion by the application of, for example, molecular techniques and isozyme
analysis to determine just how distinct such organisms are and also
whether they have evolved on single or multiple occasions.
Jahns and Ott (1994) draw attention to the apparently low number of
specialized cells and determined cells and tissues in lichens, which has
important consequences for their taxonomy. Lichens might therefore be
viewed as a complicated building constructed from few types of stones.
Jahns and Ott have shown through various ontogenetic studies a multi-
functionality of tissues and considerable plasticity, i.e. old pycnidia can
produce apothecia, root-like rhizines may acquire algae and produce lay-
ered thalli and out-growing cortical cells in Peltigera aphthosa (L.) Willd.
might start the development of cephalodia. Therefore, it follows naturally
that it is not sufficient to describe the adult organs but that the ontogenetic
aspect should also be considered.
6.5 THE IMPORTANCE OF ENVIRONMENTAL FACTORS IN

INFLUENCING PHENOTYPIC EXPRESSION
Lichens are intimately associated with their environment, many species
having subtly different ecological requirements, e.g. to the extent of pre-
ferring particular tree or rock types as substrate. It is those species with a
wide ecological amplitude which usually exhibit the greatest range of
morphological variation. An understanding of the sources of phenotypic
variation in lichens, therefore, is critical to the evaluation of criteria for tax-
Environmental factors influencing phenotypic expression 121
Figure 6.2 Three species of Menegazzia found in temperate rainforest

distinguished by having different asexual, vegetative propagules. The possible
'primary' fertile species is much rarer than the 'secondary' asexual counterparts,
though the evolutionary relationships of these and many other groups of species,
characterized essentially by alternative methods of reproduction, are in need of
critical evaluation, (a) M. nothofagi (sorediate: having powdery soredia, consisting
of fungal hyphae intermixed with algae), Tasmania, G. Kantvilas 603/80, BM); (b) M.
eperfomta (isidiate: having cylindrical isidia, peg-like outgrowths from the upper
fungal layer of the thallus, the cortex), Tasmania, J. 6494, CBG); (c) Menegazzia pro-
totypica (lacking soredia and isidia and rarely fertile and bearing disc-shaped fruits
as in Figure 6.1(c), Projection Bluff, Tasmania, G. Kantvilas, HO). Scale bar = 1 mm.
onomic delimitation. Such variation may reflect either genetic differences

between individuals due to mutational and recombinational processes, or
result from changes in developmental patterns in response to heteroge-
neous habitats. The unfortunate practice of taxonomists describing lichens
on the basis of fragmentary material without field study has resulted in
the description of many superfluous taxa (Weber, 1968; Hawksworth,
1973). This is especially the case in crustose lichens occurring on smooth,
hard rocks which are difficult to sample adequately. For instance, several
species within the crustose lichen genus Aspicilia, which is notoriously
lacking in diagnostic characters, are distinctively pigmented when fresh,
but these colours are lost in storage in the herbarium. Similarly, extreme
morphological variation observed in Acarospora has resulted in the
description of many superfluous names; numerous species in both
Acarospora and Aspicilia therefore require careful field evaluation.
Lichenized ascomycetes are characterized by the great longevity of
their thalli and ascomata. Given their exposure to harsh environmental
conditions, it is perhaps hardly surprising that they typically show great
variation in morphology. Indeed, two taxa displaying the extreme ranges
in morphology of a species may appear more different than distinct
species (Poelt, 1994). In the past a large number of taxa were introduced at
species and sub-specific level to accommodate such morphotype varia-
tion. For instance, Zahlbruckner (1927) recognizes 59 varieties and forms
of Cladonia squamosa Hoffm. Many lichens are restricted to distinct habitats
and it is reasonable to assume that ecological factors have played an
important role in speciation. Such is the importance of ecological factors in
phenotypic expression that these have been emphasized in definitions of
species where these correlate with other features. Thus, Hawksworth
(1973) claimed that species concepts in lichens are 'currently based on
sharp discontinuities in one or several morphological and anatomical (or)
chemical characters, particularly where there is evidence that genotypic
differences are involved (e.g. the two entities growing side by side in a
uniform ecological situation and retaining their identities) or there are dif-
ferences in either ecological requirements or geographical distribution, or
both/ Some lichens have a relatively wide ecological amplitude and their
thalli become modified by the environment in many ways, resulting in
changes in growth form, anatomy, pruinosity in those lichens growing on
calcareous substrates caused by superficial deposits of calcium oxalate,
and alterations in pigment due to differences in light intensity. The varia-
tion in morphology observed within taxa in different ecological situations
is vast and the problem facing lichenologists is deciding when such varia-
tion is important or trivial, especially difficult when examining herbarium
material with inadequate field data. A detailed discussion of ecological fac-
tors is beyond the scope of this review but has been outlined by
Hawksworth (1973).
Among the most extreme environments in which modifications are
numerous are metal-rich ones characterized by specialized lichen com-
munities with different rock types possessing distinctive assemblages
(Wirth, 1972; Purvis and James, 1985; Purvis, 1993). On copper-rich rocks
certain lichens accumulate copper as a chelated Cu-lichen acid complex
Secondary metabolites, lichen chemotaxonomy, molecular studies 123
and turn greenish (Purvis et at, 1987). As colour is normally regarded as an
important character at a specific level it is perhaps hardly surprising that
this phenomenon has been the cause of taxonomic confusion on several
occasions (Purvis, 1984; Purvis et al., 1990). Thus the Cu-rich morph of
Acarospora smaragdula (Wahlenb.) Massal. has been described on at least
three separate occasions previously as A. undata G. Clauzade, C. Roux and
V. Wirth, A. isortoquensis Alstrup and A. alberti Tavares (Purvis et al, 1985).
A further consequence of the complexing of metals by lichen acids to form
stable ligands is that the free acid may be difficult to detect using thin-
layer chromatography, thus adding further weight to the unsuspecting
researcher that his taxon might be distinct. Additionally, many mine envi-
ronments are characteristically acidic and under these conditions
cyanobacterial containing lichens are absent perhaps owing to the sensi-
tivity of the key enzyme, nitrogenase involved in nitrogen metabolism, to
low pH (Farmer et al, 1992). For this reason, Purvis et al. (1992) considered
the mine 'species' Placopsis lambii Hertel and V. Wirth (lacking cephalodia
containing cyanobacteria) doubtfully distinct from P. gelida (L.) Lindsay
(with cephalodia). While several ecotypes inevitably occur in such spe-
cialized environments, there are other species which are more or less
restricted, such as Psilolechia leprosa Coppins and Purvis, which also occurs
beneath copper lightning conductors on church walls throughout Britain
(Coppins and Purvis, 1987), as well as on sheltered rock faces in cloud for-
est in the Azores (O.W. Purvis and P.W. James, unpublished data). Such
environments might be regarded as analogous to islands and thus provide
models for the study of evolution of isolated populations. Several other
lichens occur in anthropogenic habitats whose precise wild origin is
unknown, e.g. Lecanora vinetorum Poelt and Huneck growing on vine sup-
ports sprayed with Bordeaux mixture in Austria (Poelt and Huneck, 1968),
Buellia pulverea Coppins and P. James subjected to fluoride pollution aris-
ing from nearby aluminium smelters, and the ubiquitous pollution lichen
Lecanora conizaeoides Nyl. ex Crombie first recorded in herbarium collec-
tions in the early 1860s.
6.6 SECONDARY METABOLITES, LICHEN CHEMOTAXONOMY

AND MOLECULAR STUDIES
When colour was accepted as a generic or specific character, chemical dis-
crimination was being inadvertently applied. For example, the grey genus
Physcia Nyl. (containing the colourless substance atranorin in the cortex) is
distinguished from the superficially similar yellow-orange genus
Xanthoria (Fr) Th. Fr. (containing an orange pigment parietin in the cortex)
based on colour and hence chemical characteristics. Similarly at a species
level, the yellow Parmeliopsis ambigua (Wulfen) Nyl. (containing yellow
usnic acid) is separated from the grey P. hyperopta (Ach.) Arnold (containing
colourless atranorin). Most secondary metabolites are, however, colourless
and require analytical procedures for their detection. There has been a
long history of using chemical tests for taxonomic purposes dating back to
the simple, though unrefined, spot test reactions performed by Nylander
in the 1860s followed by a wide range of more sophisticated analytical
methods (Santesson, 1973). Indeed, the discipline of 'chemosystematics'
was pioneered by lichenologists. Over 400 substances in 6000 lichen
species have been characterized, many unique to lichen fungi and new
compounds are being continually discovered (Culberson and Elix, 1989).
The occurrence of these heterogeneous secondary metabolites has pro-
vided the lichenologist with useful additional characters for classifying
lichens at the supraspecific, specific and varietal levels and is essential
when describing sterile crustose lichens which are unknown fertile.
Recent research has also the demonstrated the ability of certain lichenized
fungi to produce these characteristic lichen acids in mycobiont culture
(Hamada, 1989).
The taxonomic value of chemical criteria in lichen taxonomy has, how-
ever, been a matter of considerable controversy for over 130 years
(Hawksworth, 1976), although chemical investigations now form an inte-
gral part of all serious taxonomic investigations of lichen-forming fungi.
There are three common types of chemical pattern studied based upon
replacement type compounds, chemosyndromic variation and accessory
type compounds. Varying taxonomic significance is applied according to
the relatedness of the compounds, although this depends on the group of
lichens and the specific monographer. Clearly chemical variation when
associated with morphological differences, merit formal taxonomic recog-
nition, but where chemical variation occurs within morphologically simi-
lar plants, then the question as at what rank these should be treated
remains a cause of considerable debate. The taxonomic interpretation of
chemical variation is discussed in detail elsewhere (Hawksworth, 1976;
Brodo, 1978, 1986; Culberson and Elix, 1989). Of key importance is to
ascertain the genetic basis for the variation in chemistry which will then
allow a more informed judgement of how and at what level to differenti-
ate between different taxa.
There are many well-defined lichens having a uniform secondary
chemistry throughout their range and in these chemistry is as much a key
character as the morphology characterizing the species. This is not always
the case with other species exhibiting minor chemical variation not asso-
ciated with obvious morphological, distributional or ecological differ-
ences. Most lichenologists interpret morphologically similar chemotypes
as components of single polymorphic species mainly because they believe
the genetic basis of chemical differences to be trivial. The Culbersons
(Culberson, 1986; Culberson and Culberson, 1994), making a not clearly
proven parallel principally with the animal kingdom, have recognized a
What is an individual lichen? Fused thalli and mechanical hybridsl25
number of such morphs in the Ramalina siliquosa and Cladonia chlorophaea
(Florke ex Sommerf.) Sprengel complexes as sibling species where
extremely similar and visually indistinguishable cryptic species exist, the
individuals of which, when co-existing, do not interbreed with each other
but maintain the integrity of their respective gene pools (Mayr, 1992). In
many groups of organisms, this question would have been readily
resolved by the appropriate experimental crosses, not possible in the
lichen fungi. Culberson and Culberson (1994) used lichen secondary prod-
ucts produced in either mycobiont culture (in Ramalina siliquosa agg.) or
resynthesis experiments (in Cladonia chlorophaea agg.) as genetic markers
for the indirect analysis of progeny from maternal individuals in nature.
These findings, although extremely interesting, require wider testing
using complementary analytical procedures including molecular studies
to conclusively determine the reasons for the chemical variation observed
before the full implications of these results can be assessed.
Molecular techniques have as yet hardly been applied to the systemat-
ics of the lichen-forming fungi, especially at the species level, although
several studies have been successfully applied at either higher taxonomic
levels or at a population level (DePriest, 1994). Analysis of small subunit
ribosomal DNA sequences suggests at least five independent origins of
the lichen habit in several groups of fungi, thus confirming evidence from
morphological studies (Gargas et al, 1995a). At a species level, Armaleo
and Clerc (1991) used restriction-fragment polymorphism of anonymous
portions of the genomic DNA and ribosomal DNA (rDNA), to show that
lichen chimeras are fungi of a single species forming distinct photomorphs
with photobionts from different kingdoms (see section 6.3.2). We now
know that lichens show extreme variability for rDNA, there being a large
amount of length and restriction site variation, even within a species (De
Priest, 1993; Gargas et al, 1995b). At the Nato Conference on Ascomycete
Systematics held in Paris during 1994, a need for developing species spe-
cific markers was identified (see Berbee and Taylor, 1994: 222-3). Studies
using enzymes and proteins - the first apparent products of genes - have
been used with some success to demonstrate intrapopulational enzyme
polymorphisms as in Umbilicaria species in central Ellesmere Island
(Fahselt, 1989) and at a species level, two lichen species Ramalina cuspidata
and R. siliquosa have also been separated on the basis of banding patterns
(Mattsson and Karnefelt, 1986).
6.7 WHAT IS AN INDIVIDUAL LICHEN? FUSED THALLI AND

MECHANICAL HYBRIDS
There is now good evidence that fusions of lichen thalli may occur within
populations of the same species, between different species of the same
genus or between different genera (Figures 6.3-6.5), although how
Figure 6.3 Life-cycle of Xanthoria parietina (simplified after Ott, 1987a) where the
yellow colour due to the substance parietin serves as a marker, (a) Germinating
fungal ascospores; (b) developing network of fungal hyphae; (c) free-living lichen
alga - Pseudotrebouxia normally involved in lichenization; (d) mature Xanthoria
thallus formed through the interaction between the fungal hyphal network and
lichen alga; (e) foreign coccoid green algae not involved in lichenization; (f) undif-
ferentiated, areolated 'lichen' crust containing fungal hyphae and foreign coccoid
green algae, enabling the mycobiont to persist until it meets the right algal part-
ner for lichenization to occur; (g) Physcia sp. producing powdery soredia consist-
ing of fungal hyphae intermixed with lichen algae; (h) Physcia thallus and soredia
infected by Xanthoria fungal spores or the undifferentiated crust in (f) resulting in
Physcia disseminating Xanthoria through powdery soredia which Xanthoria itself
does not produce; (i) Physcia sp. producing spore-bearing fruiting bodies -
apothecia; (j) Physcia thallus and apothecia infected by Xanthoria resulting in
Physcia thallus producing Xanthoria spores. Yellow colour of Xanthoria dotted.
frequently these occur is unknown. It is hardly surprising that this phenom-

enon has caused some confusion in the past. For instance, Phaeophyscia
orbicularis is typically grey in colour, but individuals growing adjacent to
the bright yellow Xanthoria parietina (L.) Th. Fr. containing areas with an
atypical yellow pigmentation due to parietin, were described as f. virella
A.L. Sm. We now know this to be one of a few documented mechanical
hybrids between lichens belonging to different genera and demonstrates
What is an individual lichen? Fused thalli and mechanical hybrids 127
a b c
Figure 6A Thallus fusion in Cladonia (adapted from Jahns, 1987). (a) The 'reindeer
lichen', Cladonia rangiferina with smooth surface; (b) C. squamosa with abundant
squamules; (c) C. rangiferina-C. squamosa chimera with scattered squamules.
Figures (a) and (b) reproduced with kind permission from Societe Botanique du
Centre Quest, Le Clos de la Lande, Saint-Suplice-de-Royan, 17200 Royan, France;
(c) reproduced with kind permission from J. Cramer, Gebriider Borntraeger,
D-1000 Berlin, D-7000 Stuttgart, Germany.
how easy it is to describe taxa based on phenetic characteristics without

proper consideration of other factors. Ott (1987a,b) demonstrated that the
mycobiont of Xanthoria parietina (L.) Th. Fr. is capable of living together
and obtaining nutrition not only from Trebouxia but also from other coc-
coid green algae unable to undergo lichenization in the true sense (see
Figure 6.3). However, with foreign algae the fungus does not undergo dif-
ferentiation into a normal thallus but exists as an areolated crust and is
even capable of extracting suitable algae from foreign soredia of Physcia
(Schreber) Michaux, Parmelia and Evernia Ach. or else parasitizing mature
thalli and taking over the hymenium as in the case of Physcia tenella (Scop)
DC. that has yellow apothecia containing Xanthoria spores (Ott, 1987a,b).
These have been considered analogous to graft-chimeras in higher plants
(Hawksworth, 1988a). Such combined phenotypes do not apparently
represent true or 'sexual' hybrids, as there seems to be no exchange of
genetic material although this needs rigorous testing. There are several
examples of fusions between species within the same genus, as in Cladonia
subgenus Cladina, which are rarely fertile. Many of these so-called rein-
deer lichens grow in cushions with the individual plants touching one
another and the branches may fuse (Jahns, 1987). Therefore, most hyphae
will contain hyphae from different parental plants and it is probable that
the phenotype is influenced by the genetic material of all hyphae.
Figure 6.5 Fusion of Rhizocarpon geographicum thalli (simplified after Letrouit-

Galinou and Asta, 1994). (a) Confluence of areoles on black young fungal hypothal-
lus; (b) fusion of young thalli; (c) fully mature thallus with a continuous margin and
whose central yellow part is entirely covered with areoles and apothecia.
Different species rarely fuse, though fusions do occur as, for instance,
between C. rangiferina (L.) Weber ex F.H. Wigg. and C. mitis Sandst. and
perhaps, more surprisingly, between C. rangiferina (L.) Weber ex F.H.
Wigg. and C. squamosa Hoffm. (subgen. Cladonia) when the former may
develop small phyllocladia typical for C. squamosa (Figure 6.4(c)). Many
crustose lichens often form mosaics. Letrouit-Galinou and Asta (1994)
demonstrated that in Rhizocarpon geographicum (L.) DC. several thalli may
fuse at an early stage of development, resulting in the formation of a sin-
gle thallus with a well-delimited margin. It is therefore likely that individ-
ual thalli will not be genetically uniform and that populations of myco-
bionts and photobionts may exist within a single thallus, though this
remains to be tested. Yet other lichens, such as in species of Toninia Massal,
commence development on other lichens, usually species containing
cyanobacteria (Timdal, 1991). Although some thalli in mosaic-forming
lichens may fuse together, there are many other species in which this does
not happen even between thalli of the same species, as for instance in the
genus Pyrenula Massal. where thalli are typically marked by clearly delim-
ited prothalline boundaries. Poelt (1994) draws attention to the variable
mosaics formed by several crustose lichen genera typically occurring on
siliceous rocks (e.g. Fuscidea V. Wirth and Vezda, Bellemerea Hafellner and
Roux, Lecidea) and in Graphis scripta (L.) Ach. on trees. Such variation has
in the past been accorded subspecific rank, e.g. Zahlbruckner (1923) enu-
merates no less than 72 varieties and forms of Graphis scripta.
The extent to which mixed thalli occur in nature remains unclear.
Indeed, the extraordinary specificity of many fungi growing on lichens
('lichenicolous fungi') to particular genera or species of lichen would sug-
gest there are great incompatabilities and that it is therefore unreasonable
to suggest a free-for-all. Molecular biologists need to be aware, however,
of the possibilities for thallus fusions and to formulate hypotheses to test
their frequency of occurrence and influence on lichen biology .
Conclusions 129
6.8 CONCLUSIONS
Lichen systematists involved in monographic revisions rarely discuss their
basis for defining species. Considerably greater effort has recently been
focused on defining the far less stable higher levels. Thus, Hafellner (1989)
in discussing the 'principles of classification' deals exclusively with genera
and higher taxa and does not define a species. Systematics has two prin-
cipal objectives, namely to communicate the identity of an organism by
means of latinized names, and to indicate the probable evolutionary rela-
tionships of organisms. The basic assumption of lichenologists that every
lichen species consists of a distinct and unique fungus and its specific alga
is flawed. We now know that individual thalli can be composed of popu-
lations of different mycobionts, as well as different species of algae. The
preliminary evidence of sexual outcrossing provided by the Culberson
and Culberson (1994) and De Priest (1994) and consequent formations of
sibling species, if found to be widespread, will result in a significant
increase in taxa, but would we be performing a service if we recognized
these at a specific level and re-defined our concepts? The reviewer adopts
a pragmatic view in such matters and believes this would cause undue
complexity for the user.
Systematics, if considered a mirror of evolution is faced with one major
problem - we were not there to watch it happening (Jahns, 1988). We
attempt to reconstruct the course of evolution through comparative eval-
uation of characters from characteristics of present day organisms but the
fossil record of fungi is non-existent. This means a subjective component
is inevitable. This is also true for cladistics and systematics based on mol-
ecular techniques. The cladistic method has forced a more logical way of
thinking and molecular data additional characters. Molecular systema-
tists, certainly in ascomycete lichenized and non-lichenized taxa have
tended to work with supraordinal taxa to avoid problems of identification
at a specific level. We now need to focus attention on particular problem
groups and apply a range of techniques to test our methodologies. There
are woefully few researchers and results are often accepted without
repeated testing in other laboratories.
The phylogenetic species concept has yet to play an important role in
defining species, mycologists have been slow to take up cladistics, and
only recently have phylogenetic studies been used to study relationships
between taxa at higher levels (Tehler, 1994). The biological species concept
is inappropriate for lichens owing to technical problems in studying
breeding behaviour in culture. It is to be hoped that the many exciting
new analytical techniques at our disposal, including molecular and ultra-
structural studies, which when combined with morphological, ontogenetic
and environmental studies will further contribute to our basic under-
standing of these intriguing organisms. Only by such a multi-disciplinary
approach involving both comparative morphologists as well as other biol-
ogists, who have mutually much to learn from each other's disciplines, can
we hope to progress beyond the morphological species concept (Mayr,
1992) which presently remains the only practical way of naming species.
In conclusion, lichen species are based on clear discontinuities in one or
more unrelated fungal characters. A character which is of fundamental
importance in one group may be much less important in another. Indeed,
it has been said the 'art of taxonomy is in devising the most appropriate
scheme of character weighting' (Brodo, 1986). The lichen symbiosis is com-
plex and its existence poses one of the most fundamental questions in biol-
ogy, namely how two such distinct organisms can combine to form such
distinctive organisms so dissimilar from either component. The time has
come for the barriers to come down between molecular and morphologi-
cal systematists. Morphologists need to become more involved in molecu-
lar work and vice versa as they know the taxa and interesting questions to
be answered.
Acknowledgements
I am particularly grateful to Mr P.W. James and Professor P.M. J0rgensen
for stimulating discussions on all aspects of lichen biology on many occa-
sions. Dr M. Gibby, Professor D.L. Hawksworth, Dr D.M. John and Mr J.
Vogel are thanked for their valuable comments on an earlier draft of the
paper. Professor M. Blackwell is thanked for advice regarding molecular
studies.
6.9 REFERENCES
Ahmadjian, V. (1993a) The Lichen Symbiosis. John Wiley and Sons, Inc, New York.
Ahmadjian, V. (1993b) The lichen photobiont - what can it tell us about lichen sys-
tematics? Bryologist, 96,310-13.
Ahmadjian, V. and Jacobs, J.B. (1987) Studies in the development of synthetic
lichens. In Progress and Problems in Lichenology in the Eighties. Bibliotheca licheno-
logica, 25, 47-58.
Almborn, O. (1965) The species concept in lichen taxonomy. Sartryck ur Botaniska
Notiser, 1965 (4), 454-7.
Armaleo, D. and Clerc, P. (1991) Lichen chimeras, DNA analysis suggests that one
fungus forms two morphotypes. Experimental Mycology, 15,1-10.
Bellemere, A. (1994) Asci and ascospores in ascomycete systematics, in Ascomycete
Systematics: Problems and Perspectives in the Nineties (ed. D.L. Hawksworth),
Plenum Press, London, pp. 111-25.
Berbee, M.L. and Taylor, J.W. (1994) 18S Ribosomal DNA, sequence data and dat-
ing, classifying and ranking the fungi, in Ascomycete Systematics: Problems and
Perspectives in the Nineties (ed. D.L. Hawksworth), Plenum Press, London, pp.
213-23.
Brodo, I.M. (1978) Changing concepts regarding chemical diversity in lichens.
Lichenologist, 10,1-11.
Brodo, I.M. (1986) Interpreting chemical variation in lichens for systematic pur-
poses. Bryologist, 89,132-8.
References 131
Clauzade, G. and Roux, C. (1981) Les Acarospora de 1'Europe occidentale et de la
region Mediterraneenne. Bulletin du Museum d'Histoire Naturelle de Marseille, 41,
41-93.
Clerc, P. (1987) On the morphology of soralia in the genus Usnea, in Progress and
Problems in Lichenology in the Eighties, Bibliotheca lichenologica 25 (ed. E. Peveling),
Cramer, Berlin, pp. 99-102.
Coppins, B.J. (1983) A taxonomic study of the lichen genus Micarea in Europe.
Bulletin of the British Museum (Natural History). London, 11,17-214.
Coppins, B.J. and Purvis, O.W. (1987) A review of Psilolechia. Lichenologist, 19,
29-42.
Crittenden, P.D., David, J.C., Hawksworth, D.L. and Campbell, F.S. (1995)
Attempted isolation and success in the culturing of a broad spectrum of lichen-
forming and lichenicolous fungi. New Phytologist, 130,267-97.
Culberson, C.F. and Elix, J.A. (1989) Lichen substances, in Methods in Plant
Biochemistry (eds P.M. Dey and J.B. Harborne), Academic Press, London, pp.
509-35.
Culberson, W.L. (1986) Chemistry and sibling speciation in the lichen-forming
fungi: ecological and biological considerations. Bryologist, 89,121-31.
Culberson, W.L. and Culberson, C.F. (1994) Secondary metabolites as a tool in
ascomycete systematics: lichenized fungi, in Ascomycete Systematics: Problems
and Perspectives in the Nineties (ed. D.L. Hawksworth), Plenum Press, London,
pp. 155-63.
DePriest, P.T. (1993) Variation in the Cladonia chlorophaea complex I, morphological
and chemical variation in Southern Appalachian Populations. Bryologist, 96,
555-63.
DePriest, P.T. (1994) Variation in the Cladonia. chlorophaea complex II, ribosomal
DNA variation in a southern Appalachian Population. Bryologist, 97,117-26.
Fahselt, D. (1989) Enzyme polymorphism in sexual and asexual umbililcate lichens
from Sverdrup Pass, Ellesmere Island, Canada. Lichenologist, 21,279-85.
Farmer, A.M., Bates, J.W. and Bell, J.N. (1992) Ecophysiological effect of acid rain
on bryophytes and lichens, in Bryophytes and Lichens in Changing Environments
(eds J. Bell and A. Farmer), Oxford Science, Oxford, pp. 284-313.
Friedl, T. (1987) Thallus development and phycobionts of the parasitic lichen
Diploschistes muscorum (Cop.) Ant, in Progress and Problems in Lichenology in the
Eighties, Bibliotheca Lichenologica 25 (ed. E. Peveling), Cramer, Berlin, pp. 81-93.
Galloway, D.J. (1992) Biodiversity, a lichenological perspective. Biodiversity
Conservation, 1, 312-23.
Gargas, A., DePriest, P.T., Grube, M. and Tehler, A. (1995a) Multiple origins of
lichen symbioses in fungi suggested by SSU rDNA phylogeny. Science, 268,
1492-4.
Gargas, A., DePriest, P.T. and Taylor, J. (1995b) Positions of multiple insertions in
SSU rDNA of lichen-forming fungi. Molecular Biology and Evolution, 12,208-18.
Greuter, W. (1994) International Code of Botanical Nomenclature (Tokyo). Koeltz,
Konigstein.
Hafellner, J. (1984) Studien in Richtung einer natiirlichen Gliederung der
Sammelfamilien Lecanoraceae und Lecideaceae. Bibliotheca Lichenologica, 79,
241-371.
Hafellner, J. (1989) Principles of classification and main taxonomic groups, in
Handbook of Lichenology (ed. M. Galun), Vol. 1, CRC, Boca Raton, pp. 41-52.
Hale, M.E. (1974) The Biology of Lichens, 2nd edn, Edward Arnold, London.
Hamada, N. (1989) The effect of various culture conditions on depside production
by an isolated lichen mycobiont. Bryologist, 92, 310-13.
Hawksworth, D.L. (1973) Ecological factors and species delimitation in the lichens,
in Taxonomy and Ecology (ed. V.H. Heywood ), Vol. 5, Academic Press, London,
pp. 31-69.
Hawksworth, D.L. (1976) Lichen chemotaxonomy, in Lichenology: Progress and
Problems (eds D.H. Brown, D.L. Hawksworth and R.H. Bailey), Academic Press,
London, pp. 139-184.
Hawksworth, D.L. (1988a) The variety of fungal-algal symbioses, their evolution-
ary significance, and the nature of lichens. Botanical Journal of the Linnean Society,
96, 3-20.
Hawksworth, D.L. (1988b) The fungal partner, in Handbook of Lichenology (ed. M.
Galun), Vol. 1, CRC, Boca Raton, pp. 35-8.
Hawksworth, D.L. (1994) The recent evolution of lichenology: a science for our
times. Cryptogamic Botany, 4,117-29.
Hawksworth, D.L. and Honegger, R. (1994) The lichen thallus: a symbiotic pheno-
type of nutritionally specialized fungi and its response to gall producers, in
Plant Galls (ed. M.A.J. Williams), Clarendon Press, Oxford, pp. 77-98.
Hawksworth, D.L., Sutton, B.C. and Ainsworth, G.C. (1983) Ainsworth & Bisby's
Dictionary of the Fungi., 7th edn, Commonwealth Mycological Institute, Kew.
Hestmark, G. (1991) Teleomorph-anamorph relationship in Umbilicaria 1. Making
the connections. Lichenologist, 23, 343-59.
Honegger, R. (1984) Scanning electron microscopy of the contact site of conidia
and trichogynes in Cladonia furcata. Lichenologist, 16,11-19.
Jahns, H.M. (1987) New trends in developmental morphology of the thallus, in
Progress and Problems in Lichenology in the Eighties. Bibliotheca lichenologica (ed. E.
Peveling), Vol. 25, Cramer, Berlin, pp. 17-33.
Jahns, H.M. (1988) The establishment, individuality and growth of lichen thalli.
Botanical Journal of the Linnean Society, 96,21-9.
Jahns, H.M. and Ott, S. (1994) Thallic mycelial and cytological characters in
ascomycete systematics, in Ascomycete Systematics: Problems and Perspectives in
the Nineties (ed. D.L. Hawksworth), Plenum Press, London, pp. 57-62.
James, P.W. and Galloway, DJ. (1992) Menegazzia, Flora of Australia, 54,213-46.
James, P.W. and Hesssen, A. (1975) A new species of Psoroma with sorediate
cephalodia. Lichenologist, 7,143-7.
James, P.W. and Henssen, A. (1976) The morphological and taxonomic significance
of cephalodia, in Lichenology: Progress and Problems (eds D.H. Brown, D.L.
Hawksworth and R.H. Bailey), Academic Press, London, pp. 27-77.
John, D.M. and Maggs, C.A. (1997) Species problems in Eukaryotic algae: a mod-
ern perspective, in Species: the Units of Biodiversity (eds M.F. Claridge, H.A.
J0rgensen, P.M. (1991) Difficulties in lichen nomenclature. Mycotaxon, 40,497-501.
J0rgensen, P.M., James, P.W. and Jarvis, C.E. (1994) Linnaean lichen names and
their typification. Botanical Journal of the Linnean Society, 115, 261-405.
Kappen, L. (1994) The lichen, a mutualistic system - some mainly ecophysiological
aspects. Cryptogamic Botany, 4,193-202.
Karnefelt, A., Mattsson, J.-E. and Thell, A. (1992) Evolution and phylogeny of
cetrariod lichens. Plant Systematics and Evolution, 183,113-60.
References 133
Lamb, I.M. (1977) A conspectus of the lichen genus Stereocaulon (Schreb.) Hoffm.
Journal of the Hattori Botanical Laboratory, 43,191-355.
Laundon, J.R. (1995) On the classification of lichen photomorphs. Taxon, 44,387-9.
Letrouit-Galinou, M.A. and Asta, J. (1994) Thallus morphogenesis in some lichens.
Cryptogamic Botany, 4,274-82.
Lorch, J. (1988) The true nature of lichens - a historical survey, in Handbook of
Lichenology (ed. M. Galun), Vol. 1, CRC, Boca Raton, pp. 3-32.
Mattsson, J.E. and Karnefelt, J. (1986) Protein banding patterns in the Ramalina
siliquosa group. Lichenologist, 18, 231-40.
Botany, 79,222-38.
Ott, S. (1987a) Reproductive strategies in lichens, in Progress and Problems in
Lichenology in the Eighties (ed. E. Peveling), Bibliotheca lichenologica, J. Cramer,
Berlin, pp. 81-93.
Ott, S. (1987b) Sexual reproduction and developmental adaptations in Xanthoria
parietina. Nordic Journal of Botany, 7, 219-28.
Poelt, J. (1991) Homologies and analogies in the evolution of lichens, in Frontiers in
Mycology (ed. D.L. Hawksworth), CAB International, Wallingford, pp. 85-97.
Poelt, J. (1994) Different species types in lichenized ascomycetes, in Ascomycete
Systematics: Problems and Perspectives in the Nineties (ed. D.L. Hawksworth),
Plenum Press, London, pp. 273-8.
Poelt, J. and Huneck, S. (1968) Lecanora vinetorum nova spec., ihre
Vergesellschaftung, ihre Okolgie und ihre Chemie. Oserreichische Botanische
Zeitschrift, 115,411-22.
Poelt, J. and Obermayer, W. (1990) Uber Thallosporen bei einigen Krustenflechten.
Herzogia, 8,273-88.
Purvis, O.W. (1984) The occurrence of copper oxalate in lichens growing on cop-
per sulphide-bearing rocks in Scandinavia. Lichenologist, 16,197-204.
Purvis, O.W. (1993) The botanical interest of mine spoil heaps - the lichen story.
Journal of the Russell Society, 5, 45-8.
Purvis, O.W. and James, P.W. (1985) Lichens of the Coniston Copper Mines.
Lichenologist, 17, 221-37.
Purvis, O.W. and James, P.W. (1993) Studies on the lichens of the Azores. Part 1.
Caldeira do Faial. Arquipelago, 1,1-15.
Purvis, O.W., Gilbert, O.L. and James, P.W. (1985) The influence of copper miner-
alization onAcarospora smaragdula. Lichenologist, 17,111-14.
Purvis, O.W., Elix, J.A., Broomhead, J.A. and Jones, G.C. (1987) The occurrence of
copper-norstictic acid in lichens from cupriferous substrata. Lichenologist, 19,
193-203.
Purvis, O.W., Elix, J.A. and Gaul, K.L. (1990) The occurrence of copper-psoromic
acid in lichens from cupriferous substrata. Lichenologist, 22, 345-54.
Purvis, O.W., Coppins, B.J., Hawksworth, D.L., James, P.W. and Moore, D.M.
(1992) Lichen Flora of Great Britain and Ireland. Natural History Museum, London.
Purvis, O.W., Coppins, B.J. and James, P.W. (1994) Checklist of Lichens of Great
Britain and Ireland. British Lichen Society, London, pp. 1-75.
Salisbury, G. (1971) Thelotrema wightii (T. Tayl.) Nyl. Portugaliae Acta Biologica, series
B, 11, 35-7.
Santesson, J. (1973) Identification and isolation of lichen substances, in The Lichens
(eds V. Ahmadjian and M.E. Hale), Academic Press, New York, pp. 633-652.
Smith, A.L. (1921) Lichens. Cambridge University Press, Cambridge.
Smith, D.C. and Douglas, A.E. (1987) The Biology of Symbiosis. Edward Arnold,
London.
Stocker-Worgotter, E. and Turk, R. (1994 ) Artificial resynthesis of the photosym-
biodeme Peltigera leucophlebia under laboratory conditions. Cryptogamic Botany,
4,300-8.
Tehler, A. (1982) The species pair concept in lichenology. Taxon, 31, 708-14.
Tehler, A. (1994) The phylogeny of Roccellina (Rocellaceae), revisited. Ada Botanica
Fennica, 150,185-95.
Tibell, L. (1982) Caliciales of Costa Rica. Lichenologist, 14,219-54.
Timdal, E. (1991) A monograph of the genus Toninia (Lecideaceae, ascomycetes)
Opera Botanica, 110,1-137.
Tschermak-Woess, E. (1988) The algal partner, in Handbook of Lichenology (ed. M.
Galun), Vol. 1, CRC, Boca Raton, pp. 39-92.
Tschermak-Woess, E. (1995) Dictyochloropsis splendida (Chlorophyta), the correct
phycobiont of Phlyctis argena and the high degree of selectivity or specificity
involved. Lichenologist, 27,169-87.
Vitikainen, O. (1994) Taxonomic revision of Peltigera (lichenized ascomycotina) in
Europe. Ada 'Botanica fennica, 152,1-96.
Vobis, G. (1980) Bau und Entwicklung der Flechten-Pycnidien und ihren
Conidien. Bibliotheca lichenologica, 14,1-141.
Weber, W.A. (1968) Environmental modification and the taxonomy of the crustose
lichens. Sartryck ur Svensk Botanisk Tidskrift, 56,293-333.
Wirth, V. (1972) Die Silikatflechten-Gemeinschaften in Ausseralpinen
Zentraleuropa. Dissertationes Botanicae, Lehre, 17,1-305.
Yamomoto, Y., Miura, Y., Higuchi, M., Kinoshita, Y. and Yoshimura, I. (1993)
Using lichen tissue cultures in modern biology. Bryologist, 96, 384-93.
Yoshimura, I., Kurokawa, T., Yamamoto, Y. and Kinoshita, Y. (1993) Development
of lichen thalli in vitro. Bryologist, 96, 412-21.
Yoshimura, L, Kurokawa, T., Yamamoto, Y. and Kinoshita, Y. (1994) In vitro devel-
opment of the lichen thallus of some species of Peltigera. Cryptogamic Botany, 4,
314-19.
Zahlbruckner, A. (1922-24) Catalogus Lichenum Universalis, 2. Borntrager, Leipzig.
Zahlbruckner, A. (1926-27) Catalogus Lichenum Universalis, 4. Borntrager, Leipzig.
Zahlbruckner, A. (1926a) B. Spezieller Teil. Lichenes (Flechten), in Die Naturlichen
Pflanzenfamilien. Englelmann, Leipzig, pp. 61-270.
7
Fungal species in practice:
identifying species units in fungi
C. M. Brasier
Contacting address: Forest Research Station, Alice Holt Lodge, Farnham, Surrey GU10 4LH,
UK
ABSTRACT
Through the strong historical association of mycology with botany,
taxonomic species in fungi continued to be almost entirely morpho-
logically based up to the middle of this century, despite a paucity of
suitable morphological characters in some fungal groups. Even
though considerable progress was made in understanding the genet-
ics of fungal breeding systems in the early 1900s, evolutionary biology
made little initial impact on fungal systematic concepts. However, the
early genetical studies did result in the emergence of the modern fun-
gal genetics that, together with microbial genetics, contributed signif-
icantly to recombinant DNA theory.
More evolutionary based fungal species concepts began to
emerge in the 1950s, and an accelerated use of population, genetical
and molecular tools to assess variation and species diversity occurred
from the 1960s to the 1970s, particularly through the efforts of fungal
geneticists, pathologists and ecologists. One important result of these
studies is that clusters of biological and sibling species have been iden-
tified within many traditional morphological species. For example the
well-known basidiomycete pathogen Armillaria mellea, or 'honey fun-
gus', has been subdivided into 10 biological species in North America
alone. In such cases the original morphological species approximates
to a superspecies, the functional or operational species unit occurring
at the level of subordinate sibling or biological species (which may or
may not subsequently be shown to have useful morphological differ-
ences). The extent of fungal biodiversity may therefore have been seri-
ously underestimated in the past.
A great deal of complexity is also being revealed that is as yet little
understood. For example, many behaviourally or molecularly dis-
tinct, partially reproductively isolated subpopulations are being iden-
tified; some will undoubtedly prove to be independent operational
136 Fungal species in practice: identifying species units in fungi
species units. In addition, operational species are being identified at
very different levels of genetic hierarchy within a genus. Other
longer-standing complexities are being resolved. Thus, molecular
tools are enabling the many asexual fungal taxa to be assigned to sex-
ually defined genera, so throwing new light on the question of their
individual status.
Despite these developments, applying a unified species concept
within the fungi is likely to prove difficult. Indeed, while such ques-
tions are being explored, the fungal species concept must be consid-
ered to be in a transitional phase. Currently, traditional morphological
species concepts often operate alongside the emerging, largely non-
formalized (and sometimes supportive and sometimes contradictory)
ecological or molecularly-based species concepts. Ultimately, a new,
more flexible, set of hierarchical terms may be needed. The application
of genetical, ecological and molecular approaches to species recogni-
tion has therefore achieved a belated conjunction of fungal genetics
and fungal systematics, and perhaps a much needed reassessment of
the fungal species concept, although the outcome is still unclear.
7.1 INTRODUCTION
Current species concepts in fungi, as in most other groups of organisms,

are the product of historical, philosophical and technological influences,
plus the varied experiences and interests of individual observers. This
paper discusses recent progress in the characterization of species units
and the development of species concepts in fungi from the standpoint of
a practising plant pathologist and fungal population biologist.
An inherent problem, as the author sees it, is that in the past species dis-
crimination in fungi has been too anthropocentric. Too great an emphasis
has been put on classification and too little on defining the operational
species as it functions in nature. A major challenge for today's mycologist
is that of correctly identifying and accurately biologically characterizing
these fungal operational species units, or OSUs - the principal population
units, sharing a common gene pool and exhibiting a common set of phys-
iological, ecological and morphological attributes, by which fungi have
defined themselves making use of all appropriate methods. Since in
general OSUs are ecologically distinct entities the methods used to dis-
criminate them should, wherever practicable, extend beyond morpholog-
ical or molecular criteria to behavioural criteria such as breeding system,
habitat and physiological attributes. Indeed, at its most sophisticated, the
process of discriminating OSUs can itself be a developmental and evolu-
tionary one. The character and rank of population units becomes appar-
ent to the observer as information accumulates (Brasier and Rayner, 1987),
and more gradual progress in their identification may be punctuated by
leaps of understanding. The processes involved may require experience
Introduction 137
and even a degree of intuition on the part of the observer, together with
laboratory experimentation and statistical analysis of data.
The need for accuracy in the recognition and labelling of OSUs in fungi
extends well beyond the requirements of scientific communication or the
assessment of fungal biodiversity to a range of issues of environmental and
socio-economic significance (Table 7.1). These include the requirements of
international quarantine legislation and the legal specification of organ-
isms used in production of pharmaceuticals or in genetic engineering.
Uncritical identification of species units in fungal pathogens, for example,
may result in inadequate quarantine protection and so can contribute
unwittingly to the spread of highly destructive organisms. A surprising
number of important plant pathogens have recently been shown to have
been inadequately defined. In particular, distinctive taxa with very differ-
ent host ranges or geographical distributions are being shown to have been
labelled collectively under a single species. Equally, the designation of taxa
that are too narrowly defined can create loopholes in quarantine legislation
that might be exploited by the unscrupulous. There is, therefore, a hidden
environmental and social cost to the species concept debate.
Table 7.1 Accurate species recognition in fungi: applications
Scientific communication reduce ambiguity

Natural history, biodiversity assessment of rain forest resources, etc.
Diagnosis plant and animal pathogens
Quarantine legislation plant and animal pathogens
Patenting pharmaceutical, industrial, genetic
engineering, biological control organisms,
mycorrhizal symbionts
Table 7.2 Fungal groups (mycelial fungi)
Chytridomycota Potato wart disease fungus (Synchytriuni)

Zycomycota Pin moulds (e.g. Mucor), VAM mycorrhizae
Animal affinity Ascomycota Yeasts, Dutch elm disease pathogens, coral
spot fungus, discomycetes (Peziza)
Basidiomycota Mushrooms, toadstools, bracket fungi,
puffballs, truffles, rust pathogens,
ectomycorrhizae
Asexual fungi 'Deuteromycota' - Penidllium, Aspergillus,
grey mould Botrytis, Trichoderma
Algal affinity Oomycota Potato blight (Phytophthora infestans), Jarrah

root-rot fungus (P. cinnamomi), Pythium
(damping off fungi)
A point that may not be appreciated by non-mycologists is that fungi
encompass a highly diverse and polyphyletic assemblage of organisms
(Table 7.2). Some are of near-animal and others of near-plant affinity
(Cavalier-Smith, 1987, 1993; Barr, 1992; Wainwright et al, 1993; Corliss,
1994); they exhibit a wide range of morphological structure from
ephemeral unicellular yeasts to large relatively well-differentiated thalli
capable of living a thousand years or more (Brasier, 1992; Smith et al.,
1992); involve very different genetic systems from haploid, to dikaryotic
(two haploid nuclei per cell and functionally diploid), to multinucleate to
truly diploid; and have breeding systems that, within most fungal groups,
range from obligatorily outcrossing types to largely inbreeding types to
totally asexual mitotic forms exhibiting clonal lineages. This organismal
diversity and the associated variation in ecological strategies and breeding
systems increases the problems of finding a unified species concept appro-
priate to all fungal groups.
7.2 THE CHANGING FUNGAL SPECIES CONCEPT

Until the middle of the 20th century, fungal species, in practice, were
largely morphologically based (cf. Bisby, 1953; Hawksworth, 1974). This
derived in part from the necessary and inevitable historical importance
of morphological criteria in the development of taxonomic concepts, as
in most organisms, and in part from a long tradition of the teaching of
mycology as an aspect of botany. The latter is also reflected to this day in
the now somewhat anachronistic foundation of terms and naming pro-
cedures on the International Code of Botanical Nomenclature, and the
requirement under the Code for the deposition in a 'herbarium' or cul-
ture collection of a single, often fortuitously selected, type specimen (cf.
Leonian, 1934; Brasier, 1983, 1990b; Brasier and Rayner, 1987; Hansen,
1990). During the early 1900s considerable progress was made in under-
standing fungal mating systems, extending in some cases to the demon-
stration of inter-sterile breeding groups within traditional morphological
taxa (Shear and Dodge, 1927; Verrall, 1937; Mounce and Macrae, 1938;
reviewed in Burnett, 1983; Brasier, 1987). However, these important
early developments and other philosophical milestones of wider evolu-
tionary biology from Darwin and Dobzhansky to Mayr and Stebbins
made remarkably little impact on fungal systematic philosophy. Indeed,
with insufficient dialogue between them, the fields of fungal systematics
and fungal genetics drifted apart (fungal genetics emerging as an
advanced discipline that later contributed significantly, along with bac-
terial genetics, to recombinant DNA theory).
The separation of fungal systematics from fungal genetics and the con-
tinued historical pre-eminence of morphological characters in species
delimitation inevitably restricted mycological perceptions. While morpho-
The changing fungal species concept 139
logical characters are a valuable tool in the fungal systematist's armoury,
they can have profound limitations if used as sole criteria, especially when
applied to the myriad relatively simple structured microfungi and yeasts.
These limitations have been highlighted many times in the past (Leonian,
1934; Nelson, 1965; Brasier, 1983, 1990b; Kurtzman, 1985; Brasier and
Rayner, 1987) and are summarized in Table 7.3. For example, emphasis on
morphological characters takes little account of the processes of fungal
speciation, especially the possibility that morphologically useful differ-
ences between taxa may not have accumulated or that parallel or conver-
gent evolution of morphological structures may have occurred (Kemp,
1977; Brasier, 1987). A major practical limitation is that many fungi are iso-
lated into artificial culture as mycelia from a natural substrate before their
identification, a state of confinement in which their fruiting structures
may be difficult or impossible to obtain.
Following the contributions made by early pioneers such as Shear and
Dodge (1927) who took a genetical approach to species identification,
more evolutionarily related species concepts began to emerge from
around the 1950s (Nelson, 1965) and an accelerated application of statisti-
cal, genetical and molecular tools to investigate the ecological status of
taxa and to define their limits of variation occurred from the 1970s
onwards. Initially, such work was often carried out by plant pathologists,
fungal ecologists or fungal geneticists responding to a conceptual problem
with a particular organism or group of organisms which demanded an
understanding of their population structure, rather than by those with a
fundamentally systematic intent. They increasingly applied multivariate
analysis to continuous and discontinuous and to behavioural and mor-
phological characters, investigated karyotypes, conducted genetic analy-
sis of mating types and breeding units, and introduced molecular analysis
of population structure. Fungi were at last beginning to be considered
primarily as field populations responding to a set of environments rather
than as a specimen entombed alive in a Petri dish or dead in an herb-
arium box (Brasier, 1989).
This activity has led to an enormous mass of new information, a process
which from the literature appears still to be accelerating. It includes a
Table 7.3 Limitations of morphological characters in fungal systematics. (Based

on Brasier, 1991.)
Useful morphological differences between taxa may not have accumulated

Small number of characters available
Plasticity of structures, size range overlaps
Difficulty in finding structures in field (rare, seasonal)
Difficulty in obtaining structures in laboratory
Absence of sexual structures: asexual fungi
range of intriguing, frequently important but not altogether unexpected
(Brasier, 1983) developments across the entire range of fungal groups. For
example, multiple OSUs are being identified within many traditional mor-
phological species, paralleling recent revelation of multiple sibling species
within arthropod taxa (Knowlton, 1993; Claridge and Boddy, 1994). At the
same time a great deal of inter- and intra-specific complexity is being
revealed that at present remains uninterpreted or poorly understood. To
illustrate the diverse approaches used in identifying OSUs in fungi, three
case studies each selected from a different major fungal group will now be
considered. Emphasis has deliberately been placed on mycelial fungi
rather than on the mainly unicellular microbe-like yeasts. For a discussion
of similar issues within the yeasts, see especially Kurtzman (1985) and
Kurtzman and Robnett (1994).
7.3 IDENTIFYING OPERATIONAL SPECIES UNITS
7.3.1 Identification from evidence of total reproductive isolation:

Armillaria mellea
A striking example of recent developments is that of the basidiomycete
fungus Armillaria mellea, a common tree root pathogen known to foresters
and horticulturalists as the 'honey fungus' (Figure 7.1).
Until recently, A. mellea was viewed in the literature as a single, variable
morphospecies of worldwide distribution. In the 1970s, however, a
Finnish forest pathologist demonstrated an outcrossing mating system in
the fungus (Hintikka, 1973) and, using this system, another Finn identified
five totally inter-sterile breeding groups among population samples of the
fungus from Europe (Korhonen, 1978a), which came to be known as bio-
logical species groups. Similar interfertility tests (Figure 7.2) demonstrated
at least ten biological species groups among A. mellea isolates in North
America (Anderson and Ullrich, 1979). These results have been corrobo-
rated by a number of molecular studies which have also revealed the
extent of genetic divergence between the groups (Figure 7.3; reviewed in
Guillaumin et al, 1991).
At least three of the North American biological species groups are con-
specific with a European species group, and several American groups
have been named as new taxa (Figure 7.3) (Guillaumin et al., 1991).
Extensive field studies have also been stimulated by these observations.
These have revealed that the European species groups exhibit different
ecological characteristics, e.g. A. ostoyae is a primary killer of trees while A.
gallica is a saprotroph that can also colonize declining or stressed trees. It
seems likely that different ecological characteristics will be identified for
the North American species as field information accumulates.
These biological species groups are the newly recognized OSUs within
what until recently was previously viewed, in practice, as a single mor-
Identifying operational species units 141
Figure 7.1 Fruit bodies of Armillaria mellea. Scale bar = 2 cm.
phospecies, A. mellea. Within a continent, the boundaries of the units are

clearly defined by total sterility barriers. They therefore represent classical
biological species, sensu Mayr (1942). The European OSUs correlate with
some previously proposed morphological taxa, and also show some
potentially useful diagnostic differences in their colony characteristics
(reviewed by Watling et al, 1991). However, at present most of these OSUs
can only be accurately identified on the basis of an interfertility test in cul-
ture. Ultimately a series of molecular protocols to aid their identification
seems desirable (Harrington and Wingfield, 1995).
Similar studies are revealing additional Armillaria taxa worldwide (Cha
et al., 1994). Accurate identification of Armillaria taxa is imperative for
many of the reasons outlined in Table 7.1, in particular because of their
different pathological attributes and ecological characteristics. Indeed
much of the enormous literature on the biology and pathology of A. mel-
lea before 1980 is now of limited value because of uncertainty over which
species was being investigated.
Figure 7.2 Appearance of interfertile and intersterile pairings between Armillaria

OSUs on culture, (a) Four sets of intersterile pairings, in which the haploid
colonies involved belong to different OSUs. A zone of antagonism occurs between
the paired colonies, (b) Four sets of interfertile pairings in which the haploid
colonies involved belong to the same OSU. Secondary, fruiting competent diploid
mycelia develop as a result of fusion between the colonies. These secondary
colonies also producing black boot-lace-like rhizomorphs (asexual foraging struc-
tures). (Courtesy of S.C. Gregory, Forestry Commission, UK.)
Biological
species groi p Other
A.ostoyae
(highly
pathogenic)
]————————————————————————————————
v'
L —————i—————— A.gallica
1———————————| vil (weak pathogen)
1——1 11 A.gemina
r_ '—————————————————————————————— III A.calvescens
1———————————————| vi A.mellea
|—————————1 IX
11 V
A
20 40 60 80 100
Percent similarity
Figure 7.3 Dendrogram of molecular similarity between isolates of Armillaria

(revealed by EcoRl digest of mitochondrial DNAs) showing the relationships
between eight of the ten North American biological species groups (revealed by
sexual mating tests) and the major molecular groupings. (Redrawn from Smith
and Anderson, 1989.)
Discovery of multiple OSUs within traditional taxonomic species via
the demonstration of intersterile breeding units is a process that has
occurred many times in recent decades, particularly with the culturable
ascomycetes and basidiomycetes. Ascomycete examples include Nectria
haematococca, Gibberella fujikoroi, Neurospora spp., Ascobolus immersus,
Phomopsis oblonga and Ophiostoma piceae (Matuo and Snyder, 1973; Perkins
et al, 1976; Kuhlman, 1982; Webber and Gibbs, 1982; Meinhardt et al., 1984;
Perkins and Raju, 1986; Brayford, 1990; Brasier, 1993; Brasier and Kirk,
1993). Basidiomycete examples include Coniophom puteana, Coprinus spp.,
Paxillus involutus, Rhizoctonia solani, Pleurotus ostreatus, Potyporus abietinus,
Collybia dryophila and Gloeocsystidium tenue (Boidin, 1951; Macrae, 1967;
Parmeter et al, 1969; Kemp, 1983, 1985; Fries, 1985; Ainsworth, 1987;
Vilgalys and Johnson, 1987).
7.3.2 Identification from cultural and behavioural properties and

evidence of a degree of reproductive isolation: the Dutch elm disease
pathogens
Where breeding barriers cannot be readily investigated for practical rea-
sons, or where the barriers involved are only partial, the process of iden-
tifying OSUs can be a more protracted one. This was so in the recent case
of the Dutch elm disease fungi.
Dutch elm disease, a vascular wilt disease of elms, is one of the most
catastrophic plant diseases known to man. Until the late 1970s it was
believed to be caused by a single morphospecies, the ascomycete,
Ophiostoma ulmi (Figure 7.4). In the 1970s, population studies carried out
in Britain in the context of an unexpected second epidemic of the disease
revealed two distinct colony types (Figure 7.5). One proved to be a weak
and the other a highly aggressive pathogen of elms. They were initially
referred to as the non-aggressive and aggressive strains, and then as sub-
groups of the fungus. Before long, field samples across the northern hemi-
sphere showed the aggressive and non-aggressive subgroups to be widely
geographically distributed, and also recently sympatric as a result of their
independent epidemic spread into much of North America and Eurasia
from unknown origins (reviewed in Brasier, 1990a, 1991).
Sexual crosses between the aggressive and non-aggressive subgroups,
conducted shortly after the aggressive subgroup was discovered in
Britain, signalled powerfully that they were separate species. It was
shown that a strong but not total pre-zygotic reproductive barrier was
operated by the aggressive subgroup against the non-aggressive. Thus,
compared with intra-group pairings, perithecial maturation in aggressive
x non-aggressive crosses was disrupted and the numbers of viable
perithecia (Figure 7.4(d)) were greatly reduced. In addition, the resultant
ascospores were of low viability, and the Fjs exhibited a remarkable range
of non-parental colony types (Figure 7.6) and showed strong negative
Figure 7.4 Morphological structures of Ophiostoma ulmi. (a) Asexual mycelial spores;
(b) yeast-like phase; (c) asexual synnemata and synnemiospores; (d) perithecium
(sexual stage) and ascospores. Scale bar = 5mm and refers to spores only.
Synnemata and perithecia are approximately 0.5 mm high. (From Brasier, 1981.)
interactions for heritability of continuous characters such as growth-rate

and pathogenic ability (Figure 7.7) (Brasier and Gibbs, 1976; Brasier, 1977).
Simply put, the genomes of the two subgroups, when recombined, gave
rise to Fjs that were unlikely to survive in nature (Kile and Brasier, 1990).
In numerous additional studies, involving both in vitro tests and field
experiments, the aggressive and non-aggressive subgroups were shown to
differ in many important physiological, ecological and morphological attrib-
utes. These included characters such as temperature optima for growth
(widely different at c. 22° and c. 28°C respectively), pathotoxin production
and mean perithecial neck lengths, though there was too much overlap in
the latter for this to be diagnostic for many individual fungal isolates. By the
late 1970s the two subgroups were considered divergent, probably at the
Figure 7.5 Representative colonies of the non-aggressive (left) and aggressive

(right) subgroups of Ophiostoma ulmi. Note that non-aggressive subgroup colonies
are slow-growing and waxy in appearance, the aggressive colonies fibrous-striate
petaloid. Subsequently, it was realized that the two subgroups were separate
OSUs, the non-aggressive subgroup being retained as O. ulmi s.s. and the aggres-
sive being designated O. novo-ulmi sp. nov.
level of subspecies. By the early 1980s, they were considered, at least in the
author's mind, separate OSUs. Molecular studies in the late 1980s confirmed
that they were widely genetically divergent entities (Figure 7.8), and in 1991
they were formally designated as separate species, the non-aggressive being
retained as O. ulmi sensu stricto and the aggressive recognized as the new
species O. novo-ulmi (Brasier, 1991).
Turning to the practical relevance of these developments (Table 7.1), O.
ulmi s.s, was the pathogen responsible for the first pandemic of Dutch elm
disease between the 1920s and 1940s, while O. novo-ulmi is that responsi-
ble for the enormously destructive current pandemic. Had the existence of
0. novo-ulmi been recognized before 1970, its importation into Britain
might have been prevented or perhaps delayed by enactment of appro-
priate quarantine legislation, or its arrival met by a more immediate sani-
tation response. However, as with the OSUs recently identified within A.
mellea, failure to distinguish O. novo-ulmi from O. ulmi s.s. earlier must be
seen in the context of the mycological philosophy of the 1950s and 1960s,
when mainstream mycology was relatively isolated from the disciplines of
population biology and fungal genetics. Also, the two taxa could not read-
ily have been separated on the basis of traditional morphological criteria,
since the morphological differences are limited and their application
requires an experimental approach.
In fact, recognition of the aggressive and non-aggressive 'strains' of O.
ulmi s.l as separate OSUs was a relatively slow process (some 18 years
from start to finish), involving a gradual accumulation of information.
Figure 7.6 Colony characteristics of the progeny of a cross between a non-

aggressive subgroup isolate (M35, colony a) and an aggressive subgroup isolate
(027, colony b) of Ophiostoma ulmi showing the highly unusual non-parental
colony types exhibited by the FjS (colonies c-1). (From Brasier and Gibbs, 1976.)
M 35 0 27
48
I
15
10 U.procera
5-
o
(a) 6
15- M 35 0 27
{ I
10- X
IT U.laevis
5- - —,
FI
0 50 100
Pathogenicity (% defoliation)
(b)
Figure 7.7 Inheritance of pathogenicity (as % defoliation) among 62 progeny of a

cross between an aggressive (027) and a non-aggressive (M35) subgroup isolate of
Ophiostoma ulmi, showing final disease levels: (a) on the moderately susceptible
English elm, Ulmus procem; (b) on the highly susceptible European white elm U.
laevis. x, mean of Fj progeny. Note the strong negative interaction for inheritance
of pathogenicity among the progeny relative to the parents. (From Brasier and
Gibbs, 1976.)
Genetic similarity
1.00 0.75 0.50 0.25
Non-aggressive subgroup
= O.ulmi
EAN form
Aggressive subgroup
= O.novo-ulmi
NAN form
Figure 7.8 Dendrogram of similarity between isolates of the aggressive and non-
aggressive subgroups of Ophiostoma ulmi, determined by RAPD markers. Also
shown is the molecular separation of the two distinct biotypes within the aggres-
sive subgroup, termed the EAN and NAN races (Brasier, 1990a, 1991). (From Pipe
et al, 1995.)
This initially amounted to evidence of strong but not total reproductive

isolation and of behavioural differences, supported later by evidence of
morphological and molecular differences. Similar processes were involved
in the identification of OSUs in Neurospora (Shear and Dodge, 1927;
Perkins et al., 1976). Recently, when yet another Dutch elm disease fungus
was discovered in the Himalayas (Brasier, 1994), its characterization as a
third OSU (O. himal-ulmi sp. nov.) was achieved in only 8 months using
many of the techniques already developed for distinguishing O. ulmi s.s.
and O. novo-ulmi (Brasier and Mehrotra, 1995).
7.3.3 Identification from cultural and behavioural properties alone: the

case of Phytophthora megasperma
Identification of OSUs within morphological taxa does not always involve a
direct demonstration of genetic isolation, as can be seen in the case of
Phytophthora megasperma (Figure 7.9), a member of the diploid algal-like
Oomycota and a relative of the potato blight fungus, P. infestans (Table 7.2).
Until the late 1970s, this taxonomic species was viewed as an economically
important pathogen on a remarkably wide range of crops from alfalfa and
soybean to forest trees worldwide. Some authors considered it to contain
two varieties with different oogonial diameters (Fig 7.9), and it was known
to contain physiological races specialized in attacking particular crops. Since
P. megasperma was self-fertile (homothallic), simple sexual mating tests in
culture could not be used for defining OSUs. More importantly, it was not
even suspected that multiple OSUs existed within this taxon.
However, in the 1980s, two North American pathologists investigated
samples of P. megasperma isolates from different hosts for a range of mor-
phological and cultural characters and, using multivariate analysis, distin-
guished distinct host-associated clusters (Hamm and Hansen, 1982;
Hansen and Hamm, 1983) (Figure 7.10). Subsequent research demonstrat-
Figure 7.9 Sexual and asexual reproductive structures of Phytophthora megasperma:

(a) sexual structures, oogonia (9) and antheridia (#); (b) asexual sporangia and
motile flagellate zoospores. Scale bar (a) = 50mm; (b) = 30mm. (From the
original description of Drechsler, 1931.)
78 78
70- -70
® 62 62
E
CO
I 54
54-
46- 46
37 43 49 55 61
Mean oogonium diameter (urn)
Figure 7.10 Morphometric analysis of host-related subgroups within Phytophthora

megasperma. SB, soybean group; AL1, alfalfa group 1; AL2, alfalfa group 2; Dl,
Douglas fir group 1; D2, Douglas fir group 2; CL, clover group. (From Hansen and
Hamm, 1983.)
ed additional host-related characters including colony patterns, chromo-

some number differences and soluble protein banding patterns. This led
to the proposal that P. megasperma comprised at least six behaviourally
different species groups which could be distinguished by their protein
patterns (Hansen et al., 1986). A number of molecular polymorphism stud-
ies have since confirmed and extended these observations. For example,
analysis of mitochondrial DNA and nuclear DNA RFLPs and RAPDs by
Forster and Coffey (1993) and Whisson et al. (1993) revealed at least seven
major groups in P. megasperma, six corresponding to those identified by
Hansen et al. (1986) (Figure 7.11), and a seventh representing isolates from
asparagus.
It became increasingly evident from the correlations between their
physiological and morphological characteristics and host affinities and
later from their wide genetic divergence that the original clusters identi-
fied by Hansen and Hamm (1983) were separate OSUs within the old mor-
phospecies 'P. megasperma''. Several of these groups such as the ALF and
SOY 'protein groups' of Hansen et al. (1986) have now been formally
Genetic similarity is a
O3
0 1
I CD O
ro
| Protein
Other
group
A2
A8
B5
r.
B15 BHR
C3 megasperma
C1
B13
JJ6* *apple
D5 AC
D2
P. drechsleri
13 I SOY P. sojae
E1
-c E3
E2
asparagus
P. cryptogea
JJ4* *alfalfa
H2 ALF P. medicaginis
G1 CLO P. trifolii
F1 DF
Figure 7.11 Molecular subunits identified within Phytophthom megasperma isolates

on the basis of a nuclear DNA RFLP analysis, showing the close correspondence
with the host-associated protein banding groups of Hansen et al. (1986). P. drechsleri
and P. cryptogea were used as outgroups. BHR, broad host range protein group; AC,
apple/cherry group; SOY, soybean group; ALF, alfalfa group; CLO, clover group;
DF, Douglas fir group. * JJ6, JJ4, representative examples of unique molecular types
currently of uncertain status as OSUs. (Redrawn from Forster and Coffey, 1993.)
named as new species, the BHR protein group being redefined as P.

megasperma sensu stricto (Figure 7.11). The positions of two other tradition-
al morphological taxa, P. drechsleri and P. cryptogea, in the middle of the
same dendrogram when they had been intended as outgroups (Figure
7.11) indicates the extent of the earlier misclassification of 'P. megasperma'
s.l Indeed, this morphological taxon appears to be highly polyphyletic, a
situation which is in part a product of the limited morphological criteria
available to Phytophthora systematists in the past (Brasier, 1990b), and in
part a product of convergent evolution.
The initial demonstration of these OSUs was achieved solely by a care-
ful analysis of morphometric, behavioural and cultural data, and was later
reinforced by molecular analysis. The latter is also providing the first real
measures of genetic distance between the OSUs, which ranges from small
to very large (Figure 7.11). Evidence for the extent of reproductive isola-
tion between these OSUs is lacking but these questions are under investi-
gation in a number of laboratories (Forster et al, 1994; Whisson et al., 1994).
Owing to their enormous economic importance, many other tradition-
al Phytophthora taxa are under similar scrutiny. In some cases, wide genet-
ic diversity indicating the existence of multiple OSUs has been found
(Forster and Coffey, 1991; Oudemans and Coffey, 1991; Brasier and
Hansen, 1992). Note here, for example, the wide genetic divergence
between the two P. cryptogea outgroup isolates shown in Figure 7.11. In
other cases such as that of P. nicotianae, the behavioural and genetic diver-
sity being revealed is very limited, the species being shown to be a single
OSU (Oudemans and Coffey, 1990; Hall, 1993).
7.3.4 The wider perspective: underlying complexities

The above three cases are the tip of a large iceberg. They cannot adequate-
ly represent the myriad of other possible examples, each with its special
features and circumstances. Fortunately, many traditional morphospecies
such as P. nicotianae, when subjected to a detailed population biology
analysis, are being shown to comprise a single OSU. Such instances need
little further elaboration, though it should be emphasized that biological
analysis of a suspect morphospecies should be carried out before its status
as a single OSU is accepted.
However, there are also many cases where much complexity is surfac-
ing that so far remains unsolved. There are, for example, many instances
where newly identified population units show a degree of reproductive
isolation, but more information is needed before it is clear whether the
units are OSUs or taxa of a lower rank. A well-advanced scientific case is
that of the worldwide coniferous tree root pathogen, Heterobasidion anno-
sum, a basidiomycete in which mainly pine and spruce attacking sub-
groups designated as the S and P groups were identified in Europe in the
1970s (Korhonen, 1978b; and see Johansson and Stenlid, 1994). The S and P
groups are sympatric, and strongly but not totally reproductively isolated.
Host specialization is an important isolating mechanism in sympatric plant
pathogens (Brasier, 1987) and is almost certainly involved in sustaining the
independence of the S and P groups. Moreover, recent molecular studies
indicate that the S and P groups are widely molecularly divergent
(Karlsson, 1994). A third group, associated with the true firs (Abies sp.) has
recently been identified in southern Europe, and is currently referred to as
the F group (La Porta et al., 1994). The case for the formal recognition of
these groups as separate OSUs seems a powerful one, but they have yet to
be presented in these terms in the literature, and continue to be referred to
by their trivial names.
While the status of the units within Heterobasidion appears to be nearing
resolution, it is less clear for those within Phytophthora capsici, a major
pathogen of perennial crops such as cocoa and pepper across the tropics.
In P. capsici two ubiquitous molecular sub-taxa termed the CAPA and
CAPB groups have been identified that are divergent at about the 40% dis-
similarity level on the basis of isozyme analysis. CAPA and CAPB also
have slightly differently shaped asexual sporangia, and only CAPB forms
chlamydospores (thick-walled asexual resting spores) (Mchau and Coffey,
1995). However, since the host ranges of the two groups are similar, there
remains uncertainty as to whether or not they are conspecific.
P. capsici is an outcrossing species and its sexual stage can be obtained
in culture. In such circumstances there is a strong case for experimentally
investigating the level of reproductive isolation between groups via inter-
group crosses. If two groups prove to be totally reproductively isolated at
the pre-zygotic level, they should be considered separate OSUs. If partial
or no pre-zygotic isolation is found, studies should, if practicable, be
extended to examining the relative fitness of the F^s in order to assess the
degree of post-zygotic isolation. The ecological importance of the latter is
frequently overlooked in fungal biological species tests (cf. Brasier, 1987).
Even where there is full pre-zygotic interfertility, the possibility of post-
zygotic isolation should be investigated if there is reason to suspect that
two subgroups were once geographically or ecologically isolated.
Even the case of Armillaria mellea (Figure 7.3) is less clear-cut when the
sympatrically defined OSUs are considered in allopatric terms. Thus, the
five sympatric OSUs within Europe or the ten within North America were
clearly distinguished on the basis of their total reproductive isolation.
However, the European A. cepistipes shows full interfertility with the
recently discovered North American biological species XI, (not shown in
Figure 7.3) and is therefore conspecific with it, yet it also shows partially
interfertility with the North American species V (A. sinapina) and X
(Guillaumin et a/., 1991). A similar phenomenon occurs in Heterobasidion
annosum, in which North American and European S and P groups are
more interf ertile at the pre-zygotic level than those within North America
or Europe alone (Chase and Ullrich, 1990a,b).
Such examples of varying and somewhat unpredictable interfertility
between geographically isolated taxa are not new to mycology. The prob-
lem was raised during seminal work on fungal breeding units in Fomitopsis
pinicola by Mounce and Macrae (1938) and similar issues were raised with
regard to Auricularia (Duncan and MacDonald, 1967) and Coprinus (Kemp,
1975, 1977). The phenomenon has since been demonstrated for many
closely related allopatric taxa and may reflect the lack of development of
isolating mechanisms following geographic separation of populations
(Brasier, 1987,1995). Even with ecologically distinct sympatric taxa, occa-
sional hybrids may form. A probable example has recently been found for
the S and P types of Heterobasidion in North America (Garbelotto et al,
1995). One way to accommodate such problems, from the point of view of
naming and nomenclature, is to formally name the OSUs and to enhance
scientific awareness of the potential, if not the real prospects, for interspe-
cific hybridization (cf. Purvis, 1997: Chapter 6). Much better use could also
be made of subspecific ranks than at present (cf. Brasier and Rayner, 1987;
Brasier, 1992). In Heterobasidion, for example, the S and P groups could be
named as species and the European versus North American S-types (or
the European versus North American P-types) could, if appropriate, be
named subspecies.
The complex structure of 'P. megasperma' s.l. has been greatly clarified,
but it also remains a case that is still being resolved. In addition to the six
well-defined and one partially defined OSUs (Figure 7.11), 14 additional
unique molecular types have recently been identified (Forster and Coffey,
1993). Two of these, types JJ6 (from apple) and JJ4 (from alfalfa), are rep-
resented in Figure 7.11. The isolates exhibiting these novel molecular
types again tend to be associated with particular hosts, but the sample
sizes examined so far are small and therefore their status is, as yet, uncon-
firmed. They may be additional OSUs. A further possibility is that some of
these additional molecular types are hybrids between other OSUs.
Although the Phytophthora taxa involved are mainly inbreeding
(homothallic), occasional outcrossing in homothallics is likely (Brasier,
1992). Indeed, outcrossing has recently been demonstrated in P. sojae
(Forster et al, 1994; Whisson et at, 1994), a homothallic OSU within the old
morphospecies 'P. megasperma' (Figure 7.11). Such outcrossing may some-
times occur between OSUs, particularly if the pre-zygotic isolating mech-
anisms were weak or the OSUs closely related. The possibility of
hybridization is always a consideration where related and once geo-
graphically isolated taxa have been subsequently intermixed by human
activity (Brasier and Hansen, 1992; Brasier, 1995). Its occurrence, however,
may sometimes be difficult to confirm retrospectively, particularly if
extensive backcrossing were involved.
In traditional morphological taxonomy, species were often perceived as
being of approximately equal weight in the hierarchy (cf. Hawksworth,
1974). The evidence now emerging, however, is that OSUs can occur at
widely differing levels of genetic divergence. For example in 'P. megasper-
ma' (Figure 7.11) OSUs 'ALF, 'CLO' and 'DF' show a much greater mole-
cular relatedness than do, for example, OSUs 'AC' and 'SOY' or indeed
the two outgroup isolates of 'P. cryptogea''. In this instance little is known
of the geographical origins of the OSUs, and it is unclear whether the dif-
ferences in genetic distance relate to differences in their times of appear-
ance or to their exhibiting unequal rates of molecular evolution. It is also
worth noting at this point that isozyme differences between traditional
Phytophthora taxa are often very large (Oudemans and Coffey, 1991) and
would be considered nearer the generic level in some organisms such as
Operational species units in asexual fungi 155
higher animals. While such knowledge of genetic distances does not affect
our ability to identify OSUs as such, it does question the way we have
traditionally viewed the status of 'species' in nomenclature and in the
systematic hierarchy (cf. Brasier and Rayner, 1987). OSUs are units of evo-
lution, speciation is a continuing process, and species may be subject to
near random extinctions over time. The genetic status of modern OSUs
will therefore tend to be unequal, reflecting the occurrence of speciation
events and extinction events in time and space and the varying environ-
mental processes involved in OSU emergence and survival.
7.4 OPERATIONAL SPECIES UNITS IN ASEXUAL FUNGI

A large number of fungi exist which have 'lost' their sexual stage, possibly
as a result of sudden and intense selection events ('episodic selection',
Brasier, 1995). The majority of asexual genera, including Fusarium,
Verticillium and Penicillium are of ascomycete affinity (Table 7.1). In the
absence of sexual structures, morphological characters suitable for system-
atic or phylogenetic use are often even more limited, though in some asex-
ual fungi complex asexual fruiting structures (e.g. pycnidia), are produced.
As in other asexual organisms, repeated mitotic reproduction in asexu-
al fungi leads to the development of clonal lineages or large genets
(Brasier and Rayner, 1987). These lineages may readily become genetical-
ly isolated from each other via the fungus's self/non-self recognition sys-
tem, termed the vegetative incompatibility (vc) system. Vc systems occur
in both sexual and asexual fungi and are analogous to tissue incompati-
bility systems in animals. The vc system allows hyphal fusions to occur
between mycelia carrying the same vc loci (which also tend to be geneti-
cally similar individuals), but prevents fusions between mycelia carrying
different vc loci, thereby restricting gene-flow between 'unlike' genets.
Within some traditional asexual taxa that are conventionally assigned the
status of species, for example the ubiquitous Fusarium oxysporum, numer-
ous vc types occur. These vc types are reproductively isolated micro-units
that could in theory, be considered to function as separate OSUs. When
they are very numerous, however (as in F. oxysporum where some 200
potentially different host-related forms are known), considering each
micro-unit as a separate OSU would clearly present problems! When
assessing the rank of different vc types or widespread genets, other rele-
vant information, such as behavioural differences and their longer-term
genetic stability, must be taken into account. Asexual fungi therefore rep-
resent a special case. They may require a different species concept from
that applied to most sexually reproducing fungi. Alternatively, they might
be included within a general, all-embracing, species concept (see below).
Studies on the molecular structure of asexual fungi have so far often
concentrated on resolving subgroupings within the traditional taxa, such
as the different host-specialized forms within Fusarium oxysporum, rather
than on the overall species structure. For example, there is now evidence
that the different vc types involved in the group of F. oxysporum isolates
specific to musk melon ('f.sp. melonis') are a polyphyletic assemblage
(Jacobson and Gordon, 1990). In Verticillium, molecular studies have
revealed two widely genetically divergent host-related OSUs within the
traditional morphospecies V. albo-atrum (Carder and Barbara, 1991) and
two other widely divergent units, one haploid, the other diploid, within
V. dahliae (Morton et al., 1995). The status of these units has yet to be for-
malized. Within them there are likely to be different vegetative compati-
bility groups.
An important application of molecular phylogenetics is that asexual
fungi are being assigned to their sexual genera, as in the case of asexual
penicillia which are being assigned within the sexual genus Talaromyces
(Figure 7.12; Lobuglio et al., 1993) and fusaria which are being assigned to
the genera Nectria and Gibberella (Guardet et al., 1989). The same two stud-
ies have also revealed that loss of the sexual state in these genera has
occurred repeatedly over time. The placing of asexual taxa in a natural
phylogeny will enhance our understanding of their evolutionary origins
(Brasier, 1995) and hence our understanding of their structure, status and
ecological significance as OSUs.
T. thermophilus
T. trachyspermus
T. gossypii
— P. purpurogenum
r P. dendriticum
r- T. intermedius
r-j j— T. stipitatus
M1- T. flavus
\_i P. minioluteum
' P. funiculosum
— T. purpureus
r T. wortmannii
R variabile
- P. islandicum
— T. mimosinus
i— T. luteus
—— Byssochlamys nivea
Eupenicillium javanicum
Ascosphaera apis
Histoplasma capsulatum
Coccidioides imm it is
Figure 7.12 Assignment of asexual Penicillium species within the sexual genus
Talaromyces on the basis of combined rDNA domain sequence data. The other
genera shown are outgroups. (Redrawn from Lobuglio et al., 1993.)
Criteria and concepts for the future: the way ahead 157
7.5 OPERATIONAL SPECIES UNITS AMONG STERILE MYCELIA
An assemblage of fungi which have been particularly intractable to mor-
phologically based systematics are the non-sporulating 'sterile mycelia'
(Parmeter, 1965). Many sterile mycelia have a basidiomycete affinity, and
some show evidence of sexual recognition responses when paired in cul-
ture, though it is often uncertain whether the sexual state has been lost, or
is simply undiscovered. A well-known sterile fungus of considerable eco-
nomic importance is the common root pathogen Rhizoctonia solani which
is morphologically recognized as a basidiomycete. This has been exten-
sively analysed in terms of host-specialization, inter-sterility groups and
molecular relationships and up to eleven OSUs have now been identified
within it (Parmeter et al, 1969; Vilgalys and Gonzalez, 1990; Sneh et al.,
1992).
Another well-known group of sterile mycelia are the so-called 'snow
moulds', psychrophilic plant pathogens active in the host at or below
freezing point (e.g. under snow cover). By use of combined genetical,
behavioural and molecular criteria, considerable progress is being made in
the elucidation of taxa within these fungi. A recent molecular study of 23
Canadian isolates of cottony snow mould, previously attributed to the
basidiomycete Coprinus psychromorbidus, has revealed at least four highly
genetically divergent population units (Figure 7.13). The units correspond
well to groups with particular host or substrate specificities, and so have
identifiable niches. They also correspond with groups based on sexual
mating responses between isolates in culture (Laroche et al, 1995).
The processes involved in the discovery of these OSUs in the cottony
snow moulds closely parallel the processes involved in the discovery of
the OSUs within Armillaria mellea. Like so many recent examples, they
remain to be formally designated, being for the present referred to by their
trivial names.
7.6 CRITERIA AND CONCEPTS FOR THE FUTURE: THE WAY

AHEAD
7.6.1 An emerging emphasis on operational species units

In fungal systematics traditional taxa defined mainly on morphological
characters form the bulk of the 70 000 or so (Hawksworth, 1991) described
species, and new taxa based on these criteria, many representing a single
collection, continue to be designated routinely in the literature (c. 1700 per
annum; D.L. Hawksworth, personal communication). Where such taxa
are based on multiple morphological or other discontinuities, it is likely
that they are 'good' taxa in the traditional sense. The increasing revelation
of OSUs within such traditional morphological species has, however,
brought into question their utility as accurate or natural units. Indeed, the
Genetic distance
12 6 3
alfalfa, cereals;
no sclerotia
p fruit rots;
no sclerotia
c alfalfa
I L
alfalfa.cereals
with sclerotia
Figure 7.13 Molecular variation among isolates of the cottony snow mould
Coprinus psychromorbidus, based on RAPDs analysis of total DNA, showing the
occurrence of at least four major molecular groups which correspond with groups
defined by mating tests. L, low temperature basidiomycete; F, from Festuca
(fescue); C, Coprinus psychromorbidus; S, from wheat stubble. (Redrawn from
Laroche et al, 1995.)
status of any morphological taxon must be viewed with a degree of cir-

cumspection until it is known whether or not it represents a single OSU.
This is especially the case where a morphological species is known to
exhibit a wide host or substrate range or a wide geographical and climat-
ic distribution. At the same time, since there is such a large number of
described species in the literature, it is improbable that all traditionally
described taxa can ever be investigated using modern analytical methods.
Most are unlikely to receive a detailed behavioural study, particularly
those perceived, rightly or wrongly, as being economically or ecologically
unimportant. This being the case, it is inevitable that systems of tradition-
ally described species and OSUs will occur side by side in the literature
well into the future.
Nonetheless, with an increasing emphasis on the identification of OSUs,
the fungal species concept is in a state of transition. One consequence is that
the 'species unit' is, in broad terms, shifting downwards in the phylo-
genetic hierarchy, with original traditional morphospecies such as Armillaria
mellea approximating in rank to a superspecies or to a genus (Brasier, 1991).
Many traditional 'species' may therefore fall between the species and
generic levels. Another consequence of the discovery of multiple OSUs
within individual traditional taxa is that, in terms of the numbers of taxa
known to exist or remaining to be discovered, the extent of fungal biodi-
versity on our planet has probably been greatly underestimated.
For the practising plant pathologist, fungal ecologist or fungal geneti-
cist, the logical unit of choice is the operational species unit rather than the
traditional morphospecies. OSUs are more precisely biologically defined
entities and should most approximate to the principal population units
occurring in the fungi themselves. They should therefore best meet the
requirements of accurate and unambiguous communication. This require-
ment is paramount in many fields of applied mycology (Table 7.1), where
understanding behavioural and other differences between species units is
the key to good ecological management, effective disease control mea-
sures and clear quarantine and patent specification. It is OSUs, therefore,
that professional systematists must discriminate if they are best to meet
the needs of client applied mycologists.
For similar reasons, it is important that, once their status is clear and
their variation satisfactorily described, newly identified OSUs be formally
designated. The persistence of informal notation and trivial names for
OSUs only promotes confusion in the literature. It also devalues the role
of taxonomy and nomenclature in scientific communication. The nomen-
clatural process 'polices' the system. It gives new names a legitimate basis,
controls the problem of name duplication and serves as the bases of sci-
entific cataloguing (cf. Hawksworth, 1974). Naming OSUs and other pop-
ulation units under the current system of nomenclature is not, however,
ideal. A more flexible system of specific and intra-specific terms fitted to
the special requirements of the fungi may eventually be needed (Brasier
and Rayner, 1987).
7.6.2 Methodology in a new era

The increasing revelation of OSUs also marks the end of the era in fungal
taxonomy in which morphological criteria reigned supreme, and heralds
the beginning of an era in which all biological and ecological properties
are potentially defining. Where morphological differences exist which can
reliably distinguish between OSUs, or where a traditional morphological
taxon is shown to comprise a single OSU, as many are, so much the better.
The morphological characters can be used as a primary aid in identifica-
tion, and there may also be less confusion in the literature. Where useful
morphological differences are absent, however, as with so many fungi in
artificial culture, or where no useful differences have accumulated
between OSUs during their evolutionary history, then there is a need to
apply methods of diagnosis with a more experimental basis. These range
from sexual mating tests to temperature-growth studies and from
isozyme profiles to DNA hybridization and sequence studies. To produce
an accurate diagnosis of a known OSU several different types of test may
sometimes be required.
As already emphasized, the initial discovery of OSUs within tradition-
al taxa is occurring via a number of routes. The demonstration of total pre-
zygotic reproductive isolation via a mating test is often the most rapid and
the least ambiguous means of revelation. However, the occurrence of
strong sterility barriers alone does not in itself fully describe an OSU. It
simply defines its reproductive relationship to another unit. To describe it
adequately, information on its other biological properties must also be
accumulated. Furthermore, many OSUs cannot be revealed by a test for
pre-zygotic isolation. In the cases of partially reproductively isolated taxa,
inbreeding taxa, asexual taxa, and in fungi not amenable to culture such
as the obligate rust parasites, numerous other routes must be undertaken.
The examples of the Dutch elm disease Ophiostomas and Phytophthom
megasperma well illustrate that many of these routes are virtually an evo-
lutionary process in themselves. At some point during the accumulation
of data, it becomes apparent which units are functioning as species, shar-
ing a common gene pool. The initial evidence is often of a common set of
physiological, morphometric and behavioural characters. This may be
supported subsequently by evidence from molecular characters. In other
cases molecular evidence is primary to the whole process. Levels of genet-
ic divergence between units can be inferred from the number, type and
degree of discontinuity of non-molecular characters, and from the extent
of host, substrate or climatic specialization, as well as from the existence of
partial pre- or post-zygotic isolating mechanisms or the degree of molecu-
lar divergence exhibited.
(a) Molecular criteria

Molecular analysis of fungal taxa has become standard practice. A valuable
attribute of a preliminary molecular analysis, or indeed of a preliminary
inter-sterility test, is that it can provide a rapid profile of broad relation-
ships within a group, and can therefore quickly reveal anomalies in tradi-
tional taxonomic systems where they occur. This is well exemplified by
the remarkable complexity, including probable polyphyleticism, being
revealed in some traditional taxa such as P. megasperma.
The sheer power of molecular techniques makes them both exciting
and fascinating. For these very reasons, they can also be seductive! As in
cases where morphological criteria only are used to discriminate taxa, or
where revelation of OSUs occurs solely via the demonstration of inter-
sterility groups, when molecular characters alone are used to discriminate
population units and information on the behaviour and ecological prop-
erties of the units is not afterwards obtained, the interpretation, biological
meaning and taxonomic utility of the units is limited. This problem is
more likely to arise where a molecular analysis is undertaken as an end in
itself, or where the taxonomic question has not been properly defined in
advance. Use of molecular analysis as part of scientific method, for example
to test the validity of OSUs hypothesized from non-molecular characters,
would seem to be an ideal if not always the most practicable or economi-
cally effective approach. Sometimes even the traditional species names
attached to isolates in published dendrograms are incorrect due to prior
misidentification, a problem which is more likely to arise where the mole-
cular analyst has insufficient background in the biology or traditional
taxonomy of the organism being investigated. Where appropriate, it is
recommended that isolates widely used in molecular phylogenetic inves-
tigations be deposited periodically with a specialist taxonomist or with a
major culture collection so their identity may be checked.
Different components of a genome may mutate at different rates (Bruns
ei al, 1991). It is therefore also desirable that more than one region of the
genome (e.g. mitrochondrial versus nuclear DNA) and contrasting meth-
ods (e.g. single gene sequencing versus RAPDs analysis of total DNA) be
used in order to corroborate resultant molecular relationships, and to
reveal both widely divergent and closely related units. An allied aspect of
molecular criteria is that they can lead to the development of rapid tools
for diagnosis of known OSUs, not only in artificial culture or when no
other distinguishing criteria are available, but also even within a host or a
substrate. Laboratories involved in routine fungal identification, particu-
larly applied mycology laboratories, therefore need to be equipped for
molecular diagnostics. Indeed, at the beginning of the 21st century, the
acquisition of molecular methodology by mycology units is likely to be as
necessary as the acquisition of a microscope at the beginning of the 20th
century.
7.6.3 The biological and phylogenetic species concepts

In systematic biology at large, debate has been focused recently on the
value of the biological species concept, encapsulated by Mayr (1942) under
the definition 'groups of actually or potentially interbreeding natural pop-
ulations which are reproductively isolated from other such groups' (see
also the modified definition of Mayr, 1982). The fact that fungal inter-
sterility barriers can often be tested in a simple experimental manner in
culture, and that totally reproductively isolated units identified in this
way form a tidy genetic package, has given added value to the biological
species concept so far as mycology is concerned. Indeed, as a result of a
flood of studies on sterility barriers and breeding units since the 1970s, the
biological species has belatedly become something of a benchmark for a
'species' in fungi (Kemp, 1975; Esser and Hoffman, 1977; Boidin, 1980;
Vilgalys and Miller, 1983; Perkins and Raju, 1986; Brasier and Hansen,
1992).
An obvious limitation of the biological species concept is its inapplicabil-
ity to asexual fungi. So too is its unsuitability for use with sexually uni-
parental or mainly inbreeding fungi. Even with outcrossing taxa, however,
the concept has certain limitations. That the occurrence of total inter-
sterility between two otherwise similar taxa is a valid criterion for considering
them separate species is a point with which few have argued. However,
from studies of complexes within basidiomycete genera, Hallenberg and
colleagues (Hallenberg and Larsson, 1992; Hallenberg et. at, 1994) have pro-
posed that where the genetic distances between totally intersterile units are
very small a sterility barrier may sometimes be a result of a species' propa-
gation strategy. The evolutionary advantages of such a strategy would need
to be clearly demonstrated. It would also be necessary, though difficult, to
confirm that a given case did not actually represent early evolutionary
divergence of a very closely related but adaptively distinct population, i.e.
incipient speciation.
A corollary of the widely held viewpoint, that total intersterility implies
separate species, is that the ability to interbreed implies conspecificity.
This is a far more equivocal matter. Since the genetic basis of inter-
sterility may sometimes be relatively simple (Kemp, 1977; Brasier, 1984;
Chase and Ulrich, 1990a,b; Kile and Brasier, 1990), not all freely inter-
breeding populations will necessarily be conspecific. Inter-sterility groups
in fungi, and particularly in basidiomycetes, could sometimes be para-
phyletic: geographically distinct populations belonging to the same inter-
sterility group could be of a different molecular evolutionary lineage, and
possess different behavioural, ecological and morphological characters. If
sufficiently biologically distinct, they could be different OSUs, though
shown to interbreed in laboratory tests. A possible example of this prob-
lem is given in a study of molecular groups and inter-sterility groups in
the 'oyster mushroom' Pleurotus by Vilgalys and Sun (1994) (and cf.
Hallenberg et al, 1994, 1996). Inter-sterility Group I has North American,
European and Asian representatives. These authors' results suggest that
these allopatric representatives are actually unique evolutionary lineages
and that they may also be paraphyletic to two other inter-sterility groups.
The danger of considering interfertility as evidence for conspecificity is
also applicable in cases of partial interfertility, as between O. ulmi s.s. and
O. novo-ulmi (see above). While these can be induced to breed in the labo-
ratory, in nature they are strongly post-zygotically isolated. Once again it
is apparent that all available characteristics of taxa, including their geog-
raphy, habitat, and potential for post-zygotic isolation, must be taken into
account when assessing their status as OSUs.
In response to the difficulties perceived in the biological species
concept, and in an attempt to satisfy a desire for a unified species concept
fitted to all organisms, a phylogenetic species concept in which all
genealogically monophyletic, geographically distinct clades or lineages
are regarded as taxa, has been proposed (Donaghue, 1985; Mishler and
Brandon, 1987; Cracraft, 1990). Its possible adoption for the fungi has also
been suggested (Vilgalys, 1991; Vilgalys and Sun, 1994). The phylogenetic
species concept has the possible virtue that it would cover the wide diver-
sity of OSUs already identified in fungi such as classical biological species,
partially reproductively isolated units, and OSUs defined mainly on cor-
related behavioural and morphological characters. If different ranking cri-
teria were applied (Donaghue, 1985; and cf. Brasier and Rayner, 1987), the
concept could also be extended to cover widespread genets and vegeta-
tively incompatible micro-units in asexual fungi. The phylogenetic species
concept, is, therefore, something of an all-embracing 'catch all', particular-
ly if every terminal taxon in a phenogram is considered of species rank
(Donaghue, 1985). It would transcend the current shift in the methodolo-
gies used to characterize taxa in fungi, and it would cover the various
types of fungal OSU.
So far as mycology is concerned, an implication of the concept present
in the author's mind, but not a scientific criticism of it, is that if widely
adopted, being so all embracing it might stifle a crucial debate before it has
scarcely begun! This is not intended as an altogether flippant statement. It
has already been mentioned that in the first half of this century tradition-
al fungal systematics lacked sufficient evolutionary and genetical input.
The current application of genetical, ecological and molecular approaches
to defining species in fungi, and the resultant demonstration of functional
operational species units, is bringing about a much-needed reunification
of fungal systematics and fungal genetics. In consequence, fungal system-
atics is in a state of creative tension. A debate on the species concept is a
necessary part of that process, and it would be unfortunate if either a bio-
logical or a phylogenetic species concept were accepted by diffusion, with-
out more thorough discussion. Indeed, in view of both the environmental
and the nomenclatural implications, the issues involved need to be more
widely debated at a formal as well as an informal level. With some excep-
tions, much of the discussion has proceeded so far at the level of the indi-
vidual taxon, individual paper and individual scientist rather than being
driven at the level of national and international mycological societies and
committees. Broadly, there is a need for a greater attempt at synthesis
among those most sympathic to the biological species concept, those
inclined to the phylogenetic concept, and those whose emphasis has been
on morphology.
Whether a species concept can be found that would be acceptable to all
practising mycologists remains to be seen. The biological species concept
probably appeals most strongly to those with a special interest in genetics.
The phylogenetic concept probably appeals most strongly to those involved
in phylogenetic grouping and ranking, since it emphasizes genealogy and
is itself a product of hierarchical phylogenetic analysis. While the latter
might best approximate to a natural system of classification, little has been
done so far to suggest that this concept will promote a better understand-
ing of the biological and dynamic aspects of fungal OSUs. OSUs are first
and foremost behavioural entities. It is their behavioural differences, and
the ecological and genetical processes involved in their emergence and sur-
vival, including the splitting or recombining of their lineages, that are their
most defining properties. Whatever our concepts and definitions, fungi will
continue mainly to exist in such principal functional units, as they have
always done. In a changing world, it is important for us to identify as accu-
rately as possible what the boundaries and behavioural properties of these
units are, if only to assist in the survival of our own environment.
7.7 REFERENCES
Ainsworth, A.M. (1987) Occurrence and interactions of outcrossing and non-out-
crossing populations in Stereum, Phanerochaetae and Coniophora, in Evolutionary
Biology of the Fungi (eds A.D.M. Rayner, C.M. Brasier and D. Moore), Cambridge
University Press, London, pp. 285-99.
Anderson, J.B. and Ullrich, R.C. (1979) Biological species of Armillaria mellea in
North America. Mycologia, 71,402-14.
Barr, D.J.S. (1992) Evolution and kingdoms of organisms from the perspective of a
mycologist. Mycologia, 84,1-11.
Bisby, G.R. (1953) An Introduction to Taxonomy and Nomenclature of Fungi.
Commonwealth Mycological Institute, Kew, Surrey.
Boidin, J. (1951) Sur 1'existence de races intersteriles chez Gloeocystidium tenue
(Pat.); Etude morphologique et comportement nucleaire de leurs cultures.
Bulletin de la Societe Mycologique de France, 66,204-19.
Boidin, J. (1980) La notion d'espece. III. Le critere d'infertilite, ou intercompatibilite:
resultats et problems. Bulletin de la Societe Mycologique de France, 96,43-57.
Brasier, C.M. (1977) Inheritance of pathogenicity and cultural characters in
Ceratocystis ulmi. Hybridization of protoperithecial and non-aggressive stains.
Transactions of the British Mycological Society, 68,45-52.
Brasier, C.M. (1981) The pathogen, in Compendium of Elm Diseases (eds R.J. Stipes
and R.J. Campana), American Phytopathological Society, St Paul, Minnesota,
pp. 11-13.
Brasier, C.M. (1983) Problems and prospects in Phytophthora research, in
Phytophthora, Its Biology, Ecology and Pathology (eds D.C. Erwin, P.H. Tsao and S.
Bartnicki-Garcia), American Phytopathological Society, St Paul, Minnesota, pp.
351-64.
Brasier, C.M. (1984) Inter-mycelial interactions in Ceratocystis ulmi: their physio-
logical properties and ecological importance, in The Ecology and Physiology of the
Fungal Mycelium (eds D.H. Jennings and A.D.M. Rayner), Cambridge
Brasier, C.M. (1987) The dynamics of fungal speciation, in Evolutionary Biology of the
Fungi (eds A.D.M. Rayner, C.M. Brasier and D. Moore), Cambridge University
References 165
Brasier, C.M. (1989) Population dynamics of plant pathogenic 'fungi', in
Proceedings of the Fourth International Mycological Congress Regensberg, Germany
1989, (Abstracts), University of Regensberg, p.lll.
Brasier, C.M. (1990a) China and the origins of Dutch elm disease: an appraisal.
Plant Pathology, 39, 5-16.
Brasier, C.M. (1990b) Current questions on Phytophthora systematics: the role of the
population approach, in Phytophthora (eds J.A. Lucas, R.C. Shattock, D.S. Shaw
and Louise R. Cooke), Cambridge University Press, London, pp. 105-28.
Brasier, C.M. (1991) Ophiostoma novo-ulmi sp. nov., causative agent of the current
Dutch elm disease pandemics. Mycopathologia, 115,151-61.
Brasier, C.M. (1992) A champion thallus. Nature, 356, 382-3.
Brasier, C.M. (1993) The genetic system as a fungal taxonomic tool: gene flow, mol-
ecular variation and sibling species in the 'Ophiostoma piceae - Ophiostoma ulmi'
complex and its taxonomic and ecological significance, in Ceratocystis and
Ophiostoma (eds M.J. Wingfield, K.A. Seifert and J.F. Webber), American
Phytopathological Society, St. Paul, Minnesota, pp. 77-92.
Brasier, C.M. (1994) Missing link in tree disease. Nature, 372,227.
Brasier, C.M. (1995) Episodic selection as a force in fungal microevolution with
special reference to clonal speciation and hybrid introgression. Canadian Journal
of Botany, 73,1213-21.
Brasier, C.M. and Gibbs, J.N. (1976) Inheritance of pathogenicity and cultural char-
acters in C. ulmi. I. Hybridisation of aggressive and non-aggressive strains.
Annals of Applied Biology, 83, 31-7.
Brasier, C.M. and Hansen, E.M. (1992) Evolutionary Biology of Phytophthora. II.
Phylogeny, population structure and speciation. Annual Review of
Phytopathology, 30,173-200.
Brasier, C.M. and Kirk, S.A. (1993) Sibling species in Ophiostoma piceae. Mycological
Research, 97,1416-18.
Brasier, C.M. and Mehrotra, M.D. (1995) Ophiostoma himal-ulmi sp. nov., a new
species of Dutch elm disease fungus endemic to the Himalayas. Mycological
Research, 99, 205-15.
Brasier, C.M. and Rayner, A.D.M. (1987) Whither terminology below the species
level in the fungi? in Evolutionary Biology of the Fungi (eds A.D.M. Rayner, C.M.
Brasier and D. Moore), Cambridge University Press, London, pp. 379-88.
Brayford, D. (1990) Variation in Phomopsis isolates from Ulmus species in the British
Isles and Italy. Mycological Research, 94, 691-7.
Bruns, T.D., White, T.J. and Taylor, J.W. (1991) Fungal molecular systematics.
Annual Review of Ecology and Systematics, 22, 525-64.
Burnett, J.H. (1983) Speciation in fungi. Transactions of the British Mycological Society,
81,1-14.
Carder, J.H. and Barbara, D.J. (1991) Molecular variation and restriction fragment
length polymorphisms (RFLPs) within and between six species of Verticillium.
Mycological Research, 95,935-42.
Cavalier-Smith, T. (1987) The origin of fungi and pseudofungi, in Evolutionary
Biology of the Fungi (eds A.D.M. Rayner, C.M. Brasier and D. Moore), Cambridge
Cavalier-Smith, T. (1993) Kingdom Protozoa and its 18 phyla. Microbial Reviews, 57,
953-94.
Cha, J.Y., Sung, J.M. and Igarashi, T. (1994) Biological species and morphological
characteristics of Armillaria mellea complex in Hokkaido: A. sinapina and two
new species, A. jezoensis and A. singula. Mycoscience, 35,39-47.
Chase, I.E. and UUrich, R.C. (1990a) Genetic bases of biological species in
Heterobasidion annosum: Mendelian determinants. Mycology, 82, 67-72.
Chase, T.E. and Ullrich, R.C. (1990b) Five genes determining intersterility in
Heterobasidion annosum. Mycologia, 82, 73-81.
Claridge, M.F. and Boddy, L. (1994) Species recognition systems in insects and
fungi, in The Identification and Characterization of Pest Organisms (ed. D.L.
Hawksworth), CAB International, Wallingford, UK, pp. 261-74.
Corliss, A.O. (1994) An interim utilitarian ('user-friendly') hierarchical classification
and characterization of the Protozoa. Ada Protozoologia, 33,1-51.
Cracraft, J. (1990) Speciation and its ontology: the empirical consequences of alter-
native species concepts for understanding patterns and processes of differenti-
ation, in Speciation and its Consequences (eds D. Otte and J.A. Endler), Sinauer
Associates, Sunderland, Massachusetts, pp. 28-59.
Donaghue, M.J. (1985) A critique of the biological species concept and recommen-
dations for a phylogenetic alternative. The Bryologist, 88,172-81.
Drechsler, C. (1931) A crown rot of hollyhocks caused by Phytophthora megasperma
n. sp. Journal of the Washington Academy of Science, 21,513-26.
Duncan, E.G. and MacDonald, J.A. (1967) Microevolution in Auricularia auricula.
Mycologia, 59, 803-18.
Esser, K. and Hoffman, P. (1977) Genetic basis for speciation in higher basid-
iomycetes with special reference to the genus Polyporus, in The Species Concept
in Hymenomycetes (ed. H. Clemencon), J. Cramer, Vaduz, pp. 189-214.
Forster, H. and Coffey, M.D. (1991) Approaches to the taxonomy of Phytophthora
using polymorphisms in mitochondrial and nuclear DNA, in Phytophthora (eds
J.A. Lucas, R.C. Shattock, D.S. Shaw and Louise R. Cooke), Cambridge
Forster, H. and Coffey, M.D. (1993) Molecular taxonomy of Phytophthora megasper-
ma based on mitochondrial and nuclear DNA polymorphisms. Mycological
Research, 97,1101-12.
Forster, H., Tyler, B.M. and Coffey, M.D. (1994) Phytophthora sojae races have arisen
by clonal evolution and by rare outcrosses. Molecular Plant-Microbe Interactions,
7, 780-91.
Fries, N. (1985) Intersterility groups in Paxillus involutus. Mycotaxon, 24,403-9.
Garbelotto, M., Ratcliff, A., Cobb, F., Otrosina, W. and Bruns, T. (1995)
Intersterility groups of Heterobasidion annosum in California: a reappraisal of
host specificity, hybridisation and intergroup gene flow. Phytopathology, 85,
1141 (abstract 213).
Guardet, J., Julien, J., Lafay, J.F. and Brygoo, Y. (1989) Phylogeny of some Fusarium
species, as determined by large sub-unit and DNA sequence comparison.
Molecular Biology and Evolution, 6,227-42.
Guillaumin, J.J., Anderson, J.B. and Korhonen, K. (1991) Life cycle, inter-fertility
and biological species, in Armillaria Root Disease (eds C.G. Shaw and G.A. Kile),
United States Department of Agriculture Handbook 691. Washington, DC.
Hall, G. (1993) An integrated approach to the analysis of variation in Phytophthora
nicotianae and a redescription of the species. Mycological Research, 97, 559-74.
References 167
Hallenberg, N. and Larsson, E. (1992) Mating biology in Peniophora cinerea
(Basidiomycetes). Canadian Journal of Botany, 70,1758-64.
Hallenberg, N., Larsson, K.H. and Larsson, E. (1994) On the Hyphoderma praeter-
missum complex. Mycological Research, 98,1012-18.
Hallenberg, N., Larsson, E. and Mahlapuu, M. (1996) Phylogenetic studies in
Peniophora. Mycological Research, 100,179-87.
Hamm, P.B. and Hansen, E.M. (1982) Single-spore isolate variation: The effect on
varietal designation in Phytophthora megasperma. Canadian Journal of Botany, 60,
2931-8.
Hansen, E.M. (1990) Variation in the species of the Phytophthora megasperma com-
plex, in Phytophthora (eds J.A. Lucas, R.C. Shattock, D.S. Shaw and Louise R.
Cooke), Cambridge University Press, London, pp. 148-63.
Hansen, E.M. and Hamm, P.B. (1983) Morphological differentiation of host-
specialized groups of Phytophthora megasperma. Phytopathology, 73,129-34.
Hansen, E.M., Brasier, C.M., Shaw, D.S. and Hamm, P.B. (1986) The taxonomic
structure of Phytophthora megasperma: Evidence for emerging biological species
groups. Transactions of the British Mycological Society, 87, 557-73.
Harrington, T.C. and Wingfield, B.D. (1995) A PCR-based identification method
for species of Armillaria. Mycologia, 87,280-8.
Hawksworth, D.L. (1974) Mycologists Handbook. An Introduction to the Principles of
Taxonomy and Nomenclature in the Pungi and Lichens. Commonwealth
Mycological Institute, Kew, UK.
Hawksworth, D.L. (1991) The fungal dimension of biodiversity magnitude, signif-
icance and conservation. Mycological Research, 95, 641-55.
Hintikka, V. (1973) A note on the polarity of Armillaria mellea. Karstenia, 13, 32-9.
Jacobson, D.J. and Gordon, T.R. (1990) The variability of the mitochondrial DNA as
an indicator of the relationships between populations of Fusarium oxysporum f.
sp. melonis. Mycological Research, 94, 734-44.
Johansson, M. and Stenlid, J. (ed.) (1994) Proceedings of the eighth international confer-
ence on root and butt rots. Swedish University of Agricultural Sciences, Uppsala.
Karlsson, J.O. (1994) Studying genetic fingerprinting in Heterobasidion annosum
using DNA fingerprinting and a mitochondrial DNA ribosomal probe, in
Proceedings of the Eighth International Conference on Root and Butt Rots (eds M.
Johansson and J. Stenlid), Swedish University of Agricultural Sciences,
Uppsala, pp. 276-82.
Kemp, R.F.O. (1975) Breeding biology of Coprinus species in The Section Lanatuli.
Transactions of The British Mycological Society, 65, 375-88.
Kemp, R.F.O. (1977) Oidial homing and the taxonomy and speciation of
Basidiomycetes with special reference to the genus Coprinus, in The Species
Concept in Hymonomycetes (ed. H. Clemencon), J. Cramer, Vaduz, pp. 259-73.
Kemp, R.F.O. (1983) Incompatibility and speciation in the genus Coprinus. Revista
de Biologia, 12,179-86.
Kemp, R.F.O. (1985) Do fungal species really exist? A study of basidiomycete
species with special reference to those in Coprinus section Lanatuli. Bulletin of the
British Mycological Society, 19, 34-9.
Kile, G.A. and Brasier, C.M. (1990) Inheritance and inter-relationship of fitness
characters in progeny of an aggressive x non-aggressive cross of Ophiostoma
ulmi. Mycological Research, 94, 514-22.
Knowlton, N. (1993) Sibling species in the sea. Annual Review of Ecology and
Systematics, 24,189-216.
Korhonen, K. (1978a) Intersterility and clonal size in the Armillaria mellea complex.
Karstenia, 18, 31-42.
Korhonen, K. (1978b) Intersterility groups of Heterobasidion annosum.
Communicationes Instituti Forestalls Fenniae, 94,1-25.
Kuhlman, E.G. (1982) Varieties of Gibberella fujikuroi with anamorphs in Fusarium
section Liseola. Mycologia, 74, 759-68.
Kurtzman, C.P. (1985) Molecular taxonomy of fungi, in Gene Manipulations in Fungi
(eds J.W. Bennett and L.L. Lasure), Academic Press, Orlando, Florida, pp.
35-63.
Kurtzman, C.P. and Robnett, C.J. (1994) Orders and families of ascosporogenous
yeasts and yeast-like taxa compared from ribosomal RNA sequence similarities,
in Ascomycete Systematics: the Problems and Prospects in the Nineties (ed. D.L.
Hawksworth), Plenum Press, New York, pp. 249-58.
La Porta, N., Capretti, P., Kammiovirta, K., Korhonen, K. and Karjalainen, R. (1994)
Genetic variation in F-group isolates of Heterobasidion annosum occurring in
Italy, in Proceedings of the Eighth International Conference on Root and Butt Rots (eds
M. Johansson and J. Stenlid), Swedish University of Agricultural Sciences,
Uppsala, pp. 233-7.
Laroche, A., Gaudet, D.A., Schaalje, G.B., Erickson, R.S. and Ginns, J. (1995)
Grouping and identification of low temperature basidiomycetes using mating,
RAPD and RFLP analyses. Mycological Research, 99,297-310.
Leonian, L.H. (1934) Identification of Phytophthora species. West Virginia
Agricultural Experiment Station Bulletin, 262,1-36.
Lobuglio, K.F., Pitt, J.I. and Taylor, J.W. (1993) Phylogenetic analysis of two ribo-
somal DNA regions indicates multiple independent loss of a sexual Talaromyces
state among asexual Penicillium species in subgenus Biverticillium. Mycologia, 85,
592-604.
Macrae, R. (1967) Pairing incompatibility and other distinctions among
Hirschioporus (Polyporus) abietinus, H. fusco-violaceous and H. laricinus. Canadian
Journal of Botany, 45,1371-98.
Matuo, T. and Snyder, A. (1973) Use of morphology and mating populations in the
identification of formae speciales in Fusarium solani. Phytopathology, 6, 562-5.
Mayr, E. (1942) Systematics and the Origin of Species. Columbia University Press,
New York.
Mayr, E. (1982) The Growth of Biological Thought. Harvard University Press,
Massachusetts.
Mchau, G.R.A. and Coffey, M.D. (1995) Evidence for the existence of two subpop-
ulations in Phytophthora capsici and a redescription of the species. Mycological
Research, 99, 89-102.
Meinhardt, F., Koch, H. and Esser, K. (1984) Compatibility groups in Ascobolus
immersus - an indication of speciation. Theoretical and Applied Genetics, 68,365-7.
Mishler, B.D. and Brandon, R.N. (1987) Individuality, pluralism, and the phyloge-
netic species concept. Biology and Philosophy, 2, 397-414.
Morton, A., Tabrett, A.M., Carder, J.H. and Barbara, D.J. (1995) Sub-repeat
sequences in the ribosomal RNA intergenic regions of Verticillium alboatrum and
V. dahliae. Mycological Research, 99, 257-66.
References 169
Mounce, I. and Macrae, R. (1938) Interfertility phenomena in Fames piernicola.
Canadian Journal of Research, 16, 354-76.
Nelson, R.R. (1965) Assessing biological relationships in fungi. Phytopathology 55,
823-6.
Oudemans, P. and Coffey, M.D. (1990) Relationships between Phytophthora
species: evidence from isozyme analysis, in Phytophthora (eds J.A. Lucas, R.C.
Shattock, D.S. Shaw and Louise R. Cooke), Cambridge University Press,
London, pp. 184-203.
Oudemans, P. and Coffey, M. (1991) A revised systematics of twelve papillate
Phytophthora species based on isozyme analysis. Mycological Research, 95,
1025-46.
Parmeter, J.R. (1965) The taxonomy of sterile fungi. Phytopathology, 55, 826-8.
Parmeter, J.R., Sherwood, R.T. and Platt, W.D. (1969) Anastomosis groupings
among isolates of Thanatephorus cucumeris. Phytopathology, 59,1270-8.
Perkins, D.A. and Raju, N.B. (1986) Neurospora discreta, a new heterothallic species
defined by its crossing behaviour. Experimental Mycology, 10, 323-38.
Perkins, D.D., Turner, B.C. and Barry, E.G.C. (1976) Strains of Neurospora collected
from nature. Evolution, 30, 281-313.
Pipe, N.D., Buck, K.W. and Brasier, C.M. (1995) Molecular relationships between
Ophiostoma ulmi and the BAN and NAN races of O. novo-ulmi determined by
RAPD markers. Mycological Research, 99,653-8.
Purvis, O.W. (1997) Species concepts in lichens, in Species: the Units of Biodiversity
(eds M.F. Claridge, H.A. Dawah and M.R. Wilson), Chapman & Hall, London,
pp. 109-34.
Shear, C.L. and Dodge, B.O. (1927) Life histories and heterothallism of the red
bread-mould fungi of the Monilia sitophila group. Journal of Agricultural Research,
34,1019-41.
Smith, M.L. and Anderson, J.B. (1989) Restriction fragment length polymorphisms
in mitochondrial DNAs of Armillaria: identification of North American biologi-
cal species. Mycological Research, 93,247-56.
Smith, M.L., Bruhn, J.N. and Anderson, J.B. (1992) The fungus Armillaria bulbosa is
among the largest and oldest living organisms. Nature, 356,428-31.
Sneh, B., Burpee, L. and Ogashi, A. (1992) Identification of Rhizoctonia species.
American Phytopathological Society Press: St Paul, MN, USA.
Verrall, A.F. (1937) Variation in Fomes ignarius. Minnesota Technical Bulletin, 117,1-41.
Vilgalys, R. (1991) Speciation and species concepts in the Collybia dryophila com-
plex. Mycologia, 83, 758-73.
Vilgalys, R. and Gonzales, D. (1990) Ribosomal DNA restriction fragment length
polymorphisms in Rhizoctonia solani. Phytopathology, 80,151-8.
Vilgalys, R. and Johnson, J.L. (1987) Extensive genetic divergence associated with
speciation in filamentous fungi. Proceedings of the National Academy of Sciences,
USA, 84, 2355-8.
Vilgalys, R. and Miller, O.K. (1983) Biological species in the Collybia dryophila group
in North America. Mycologia, 75, 707-22.
Vilgalys, R. and Sun, B.L. (1994) Ancient and recent patterns of geographic speci-
ation in the oyster mushroom Pleurotus revealed by phylogenetic analysis of
ribosomal DNA sequences. Proceedings of The National Academy of Sciences, USA,
91.4599-603.
Wainwright, P.O., Hinkle, G., Sogin, M.L. and Shekel, S.K. (1993) Monophyletic
origins of the metazoa: an evolutionary link with fungi. Science, 260, 340-2.
Watling, R., Kile, G.A. and Burdsall, H.H. (1991) Nomenclature, Taxonomy and
Identification, in Armillaria Root Disease (eds C.G. Shaw and G.A. Kile), United
States Department of Agriculture Handbook, 691, Washington, DC.
Webber, J.F. and Gibbs, J.N. (1982) Colonisation of elm bark by Phomopsis oblonga.
Transactions of the British Mycological Society, 82,348-52.
Whisson, S.C., Hewlett, B.J., Liew, E.C.Y., Maclean, D.J., Manners, J.M. and Irwin,
J.A.G. (1993) An assessment of relationships between members of the
Phytophthora megasperma complex and Phytophthora vignae using molecular
markers. Australian Journal of Systematic Botany, 6,295-308.
Whisson, S.C., Drenth, A., Maclean, D.J. and Irwin, J.A.G. (1994) Evidence for out-
crossing in Phytophthora sojae and linkage of a DNA marker to two avirulence
genes. Current Genetics, 27, 77-82.
8
Practical aspects of the species
concept in plants
R. J. Gornall
Contacting address: School of Biological Sciences, Department of Botany, University of
Leicester, University Road, Leicester LEI 7RH, UK
ABSTRACT
As a result of its application to a wide variety of different biological
situations the universally used taxonomic species concept has
evolved from a simple, typological concept based on morphology,
into a more complex, pluralistic entity, deriving its ethos from a vari-
ety of sources. Owing to its perceived lack of theoretical background
and practical difficulties in certain situations, two chief rivals have
emerged: the biological and phylogenetic species concepts, respec-
tively. Although much criticised, especially by botanists, features of
the biological concept have been absorbed into the taxonomic
concept. The relationship between the taxonomic and phylogenetic
concepts appears to be close, although differences exist. In particu-
lar, an appropriate extension of the concept of monophyly to the
species level needs to be formulated, and criteria for the best way to
assign rank are needed if a wider acceptance of the phylogenetic
concept is to be achieved.
8.1 INTRODUCTION
The debate over what constitutes a species has simmered on and off
almost since the start of recorded history. Over the past 50 to 60 years it
has occasionally boiled over in flurries of publications. The burgeoning
literature is replete with frequently confusing terminology and all sorts of
theoretical notions that are untested or even untestable experimentally.
The present symposium concerns species concepts in practice. The prima-
ry requirement of any practical concept is that the species must be identi-
fiable by the working botanist. Since our interpretations of evolution are
172 Practical aspects of the species concept in plants
critically dependent on our operational species concepts, it is also essen-
tial that the latter correspond as closely as possible to theoretical models.
Matching theory with practice inevitably demands compromise and this
is where some of the more apparently intractable problems arise.
As Blackwelder (1967) has pointed out, it is important in any discussion
of this subject to clarify the different usages of the word species. In partic-
ular, we must distinguish the taxon from the category. A species as a taxon
is regarded by many as an individual, an entity made up of component
parts (individual plants), with a beginning, an existence in time and space,
and an end (Ghiselin, 1974); or it can be seen as a class with defining char-
acteristics (Caplan, 1981; Ruse, 1981). Either way, some method of assem-
bly or grouping is required in order to recognize the taxon. The species as
a category is simply a point in the taxonomic hierarchy between genus
and subspecies. The task of the taxonomist is to decide what level of vari-
ation in the plants at hand corresponds to this particular level in the hier-
archy. These two aspects correspond to the assessment of affinity and
rank, the twin problems faced every day by practising taxonomists.
In this review I shall concentrate chiefly on the taxonomic species con-
cept, a term I shall apply to the operational construct used almost univer-
sally by practising plant taxonomists. In doing so I shall highlight some of
the potential problems presented by the plant kingdom, some of which
rarely occur, at least to the same degree, in the animal world. Finally, I
shall briefly consider two alternative concepts that seek to provide both a
working approach and its associated theoretical underpinning. These are
the biological and phylogenetic concepts, respectively.
8.2 THE TAXONOMIC SPECIES

The original species concept which forms the basis of our current
approach was based essentially on morphological criteria. It has its origin
ultimately in folk taxonomy, but can be traced in written form to Aristotle,
who recognized groups of organisms on the basis of resemblances in the
'shapes of their parts, or of their whole body'. Successive subdivision of
these groups produced smaller units - the species. As taxa these were thus
recognized as morphologically well-marked sets of individuals that
shared the same essence. An additional criterion, that of reproductive
cohesion, was added by Ray (1686), who reckoned that species should
breed true within the limits of their variation. This use of the term species
was the one adopted by Linnaeus (1753). The concept was typological in
the sense that both he and Ray believed in the existence of an underlying,
fixed entity created by God, which was overlayed by a certain amount of
variation imposed by nature. It was the job of the taxonomist to identify
the limits of this variation and thereby deduce the essential nature of each
species. However, an appreciation that species were more than types but
were made up of natural populations of individuals soon followed.
Infraspecific variation 173
As a result of the researches of various Russian and German botanists,
e.g. Bunge and Wettstein, in the 19th and early 20th centuries, another
criterion - that of geographical coherence - was added to those of morpho-
logy and breeding behaviour, and together they formed the basis of the
morphological-geographical approach. Du Rietz (1930) summarized the
situation at the time by offering the following formal definition of a
species: 'the smallest natural populations permanently separated from
each other by a distinct discontinuity in the series of biotypes'. Since about
the 1940s onwards, the similarity and discontinuity has usually been
assessed on the basis of evidence from a wide variety of sources, includ-
ing anatomy, chemistry, geography and ecology, as well as breeding
behaviour and cytogenetics, although a key convention has been that any
species must be recognisable on morphological grounds (Davis and
Heywood, 1963). The inclusion of evidence in addition to that from mor-
phology has blurred considerably the original, simple concept and has led
botanists to talk in terms of taxonomic rather than morphological species.
The taxonomic concept of a species is often described as being phenet-
ic, in which recognition of the taxon is on the basis of overall similarity as
perceived by the senses, and its rank is attributed on the basis of the
degree of its discontinuity from related biotypes. This approach in practice
often produces species that are delimited on the basis of certain character
correlations involving at least one (or at least one of a very few) key
characters, even though many other characters - perhaps all available
ones - may have been taken into account. Blackwelder (1964, 1967) has
described this practice as 'omnispective'. Numerical methods are also
available by which to detect pattern and quantify discontinuity and over-
all similarity (Sneath and Sokal, 1973).
Despite this well-established approach to the recognition of species, it is
a frequently quoted complaint that the taxonomic species concept is high-
ly subjective and incapable of precise formulation (Davis and Heywood,
1963). In particular, it has been claimed that the concept does not deal very
well with situations in which there is extensive infra-specific variation, or
uniparental inheritance (through inbreeding or asexual reproduction), or
hybridization, or the existence of cryptic species. All of these are well-
known phenomena in plants, and it is worth considering various aspects
of them, if only briefly, to remind ourselves of the particular problems
which any species concept, not just a taxonomic one, has to face.
8.3 INFRASPECIFIC VARIATION

Mayr (1942,1992) claimed that the phenotypic differences between indi-
viduals of a taxonomic species can exceed those found between such
species, thereby violating the grouping criterion of overall similarity. The
phenomenon occurs chiefly in some morphological species complexes in
which various infraspecific taxa have been recognized. Detailed analysis
(including study of breeding relationships, cytology and morphology),
however, can often help to clarify the situation, as it did in the case of the
Turnera ulmifolia complex (Shore and Barrett, 1985), in which several of the
varieties were shown on further investigation not only to be morphologi-
cally distinct but also reproductively isolated, i.e. good taxonomic and bio-
logical species.
This is not to deny that the occurrence of infraspecific variation can
pose problems when regional or local populations display a phenotype
that snares the characters diagnostic of the species but also possesses addi-
tional distinguishing features. Should these be recognized as separate
species? Inconsistency in the attribution of rank by taxonomists in this
respect when faced with similar situations can lead to a loss of credibility.
This can be remedied to some extent, however, by the adoption of certain
limited requirements in terms of numbers of correlated character differ-
ences. For example, for specific distinction Hedberg (1958) and Davis
(1978) advocated the use of clearly discontinuous variation in at least two
independent characters; some go as far as to allow a single morphological
character together with a difference in a cryptic character, such as chro-
mosome number. Decisions on whether to lump or split may also be made
easier by using appropriate methods of numerical analysis in which mor-
phological variation within and between populations of species can be
partitioned and the patterns compared (Gilmartin et al., 1986). Du Rietz
(1930) has also reviewed and discussed at length the conventions by
which regional variants can be allocated to the rank of subspecies and
local variants to the rank of variety. His proposals have been widely
accepted, if not always strictly followed. Infraspecific variation is especially
hard to handle taxonomically when it is non-hierarchical, as is so often the
case, and its treatment has been discussed at a recent symposium (Styles,
1986).
Generally speaking, however, infraspecific variation of the extent com-
plained about by Mayr is not often a problem. That this is so is illustrated
firstly by a study by Gilmartin (1980), who analysed the partitioning of
morphological variation within and between populations, species and
genera in two families, Asclepiadaceae and Apiaceae. She found signifi-
cant breaks in the pattern of variation at all levels up to but not including
the generic and family levels. Important for the present discussion is the
finding that the extent of variation within species was consistently much
less than it was between species. Secondly, very often the characters dis-
tinguishing species are not those that vary significantly within species.
This has been amply illustrated by Davis and Gilmartin (1985), who
analysed morphological variation in several species of the grass Puccinellia.
They found that variation within species involved characters (e.g. leaf size
and inflorescence branching) that could be interpreted as adaptive, i.e.
they found evidence of ecotypic differentiation; in contrast, an entirely
Uniparental inheritance 175
different set of characters, whose adaptive value was not clear, was found
to distinguish the different species.
8.4 UNIPARENTAL INHERITANCE

The problems surrounding the taxonomic treatment of inbreeders and
apomictic complexes have been presented in detail elsewhere (Davis and
Heywood, 1963; Grant, 1981), and I shall offer only a short description
here. The products of uniparental inheritance are often only slightly dif-
ferent one from another and the term microspecies has gained currency
for these particular taxonomic species. The term microspecies is used to
cover any morphologically uniform set or population of plants which is
slightly different from related uniform sets or populations.
As far as inbreeding is concerned, an array of homozygous pure lines
is likely to be produced where persistent selfing occurs. In many cases
these lines are morphologically distinguishable and it would be possible
to give each a name at specific rank, thereby recognizing them as
microspecies. Where morphological recognition is difficult, the terms
semi-cryptic, cryptic or sibling species are sometimes used. Jordan (1873)
divided a number of inbreeding species in this manner, e.g. in Erophila,
Thlaspi and Iberis. The approach can be defended in cases where the
reproductive isolation mediated by autogamy is complete, and especially
where incompatibility barriers exist between the microspecies, as they do
for example among several species of Gilia in the G. inconspicua complex
(Grant, 1981). In some cases, however, it is often found that products of
the occasional cross-pollination occur which recombine the diagnostic
features of certain of the homozygous lines and consequently do not
belong to any of the microspecies. The gene-pools are still evidently
linked and, in virtually all such cases, the taxonomic problems caused by
inbreeding have been solved more or less satisfactorily by judicious use
of infraspecific ranks, e.g. as in Montia fontana, or by accepting an undi-
vided, but variable species, e.g. Capsella bursa-pastoris.
Apomixis, either through vegetative reproduction or by means of agamo-
spermy, in many respects produces patterns of morphological variation
similar to those seen in persistent inbreeders. Large numbers of morpho-
logically pure lines can be identified. Genetically, however, the situation is
often different. Thus, recent evidence from isozyme studies of agamosper-
mous variants of Taraxacum has shown that the level of heterozygosity is
higher than in some sexual, outbreeding species, but it is fixed (Hughes and
Richards, 1988). This is much as might be predicted from the belief that
many apomicts are of hybrid origin. In terms of genetic variation within
agamospecies, it appears that some may be uniform, apparently consisting
of a single genotype, but others may exhibit various levels of variation, from
the odd plant in a population which differs in one allele at a locus, to region-
al variation where whole populations in different areas may differ by only
one allele or, occasionally, rather more extensively; all three situations occur
in species of Hieracium sect. Alpina (Shi et al., in press). Sometimes this
regional variation is accompanied by consistent but minute morphological
differences, and the question arises as to whether such populations should
be treated as infraspecific taxa or as separate microspecies. The convention
that individual morphological lines should be recognized at specific rank is
defensible, and indeed the current fashion is for each to be accorded such
status, e.g. as in Hieracium, Taraxacum, and the Ranunculus auricomus group
(until recently the various agamospermous variants of the latter in
Scandinavia were treated as subspecies). If the spirit is not willing to dissect
out the apomictic minutiae, there is the alternative of the species aggregate.
Although not of specific rank, use of the aggregate does allow the busy tax-
onomist to acknowledge the variation and give it a name (Heywood, 1963).
This approach has been criticised for somehow perpetuating taxonomic
imprecision, but this is only so when the aggregate is allowed to masquer-
ade as an entity comparable with a sexual species; it cannot be criticised for
not being a proper species - it is not intended to be one (Heslop-Harrison,
1962).
A situation the taxonomic species concept does not cope well with
involves plants with facultative apomixis, whereby new genotypes are pro-
duced by occasional sexual events and each of which can form a new
apomictic lineage. The taxonomic problems that this entails are probably
insoluble, although a modus operandi has been worked out by students of
the Rubus fruticosus aggregate in which this problem occurs (Newton, 1975).
8.5 HYBRIDIZATION
In many plant groups, particularly herbaceous perennials and woody
species, interspecific hybridization can be commonplace. Hybrids may be
sterile, as is usually the case in Potamogeton, or they may exhibit various
degrees of fertility right up to 100%. In the genus Salix, a hybrid has been
made artificially but sexually involving 13 different parental species. In
some families, intergeneric hybrids are not uncommon, e.g. Rosaceae,
Poaceae, and a particularly extreme example is the Orchidaceae where
one horticultural variant has at least eight genera in its parentage. In the
British Isles, 780 hybrids have been reliably reported from a flora of c. 2500
species (Stace, 1989). The phenomenon is widespread in plants and can
lead to speciation if the hybrids become stabilized by sexual means (via
recombination or polyploidy or permanent structural hybridity) or asexu-
al means (via apomixis or clonal growth) (Grant, 1981).
Hybridization can blur the boundaries of taxonomic species.
Nevertheless, it is usually defensible to recognize two species, even though
they hybridize in nature, if the frequency, location or viability of such
hybrids is such that the parental gene pools do not merge, e.g. in species of
Cryptic or sibling species 177
Quercus, Populus and Geum. Grant (1981) has used the term 'semi-species'
for cases where gene pools are incompletely reproductively isolated.
There are also problems with polyploid pillar complexes, where the
taxa belonging to the polyploid superstructure share parts of their
genome both among themselves and with their diploid ancestors. Not sur-
prisingly, the patterns of morphological variation can be complex and mis-
leading when identifying the evolutionary patterns involved.
Nonetheless, taxonomists have laboured to identify the pathways and
have recognized the products as species. In doing this a range of
approaches (not just morphology) is often used, including breeding stud-
ies, cytogenetics and the use of molecular markers. Good examples of this
can be found among the ferns, e.g. in the genera Asplenium, Dryopteris and
Polypodium, in which the complex genome relationships and evolutionary
pathways have been worked out and species recognized on the basis of
the findings (Lovis, 1977). Reliable identification of several of the species
using morphological characters requires use of a microscope and involves
features such as the anatomy of the annulus and paraphyses or the pat-
tern on the surface of the spores; in this sense the species may be regard-
ed as semi-cryptic. Other examples of polyploid pillar complexes have
been discussed by Stebbins (1971) and Grant (1981).
Recent studies of chloroplast DNA variation in plant species have shed
light on another dimension to the phenomenon of hybridization in plants.
It involves the capture of the chloroplast genome of one species by anoth-
er. Following an initial hybridization event the hybrid acts as the maternal
parent in repeated backcrosses to one of the species acting as the paternal
parent (Rieseberg and Soltis, 1991). The result is the capture of the chloro-
plast genome of the maternal parent by the male parent. Such transfer of
cytoplasm has now been documented in a range of different species,
including shrubs, e.g. Salix (Brunsfeld et al.f 1992), herbaceous perennials,
e.g. Heuchera (Soltis et al, 1991), and annuals, e.g. Helianthus (Rieseberg et
al, 1991). Studies of several genera, e.g. Quercus (Whittemore and Schaal,
1991) and Zea (Doebley, 1989), suggest that cytoplasmic gene flow can
occur between species in the absence of significant nuclear gene flow.
Such findings present considerable problems to the taxonomist. What is
one to do about taxa which appear to be good taxonomic species but
which contain an alien cytoplasm, possibly captured in a series of steps via
different species from an ultimate 'donor' that is not closely related? Such
a case is exemplified by Heuchera nivalis which has captured the chloro-
plast of a species in a different section via hybridization with an interme-
diary, H. parvifolia (Soltis et al, 1991).
8.6 CRYPTIC OR SIBLING SPECIES

Reproductive isolation and genetic differentiation can occur without
much associated morphological change. This leads to the formation of
cryptic species. They are also often referred to as sibling species, although
this term implies a relationship which may not necessarily exist. The phe-
nomenon is clearly a problem for the taxonomic species concept. Grant
(1981) cited the example of the Gilia transmontana group which experi-
ments have shown to contain five intersterile but morphologically similar,
though not identical, species.
A related taxonomic problem is presented by autopolyploidy. There are
several well-documented cases now where autoploids that have been
diagnosed on taxonomic, cytological and genetical evidence have been
shown to occupy a substantially different geographical range and ecolog-
ical niche from their diploid parent, e.g. Tolmiea menziesii and species of
Heuchera (Soltis and Soltis, 1989). In these cases the diploid and auto-
tetraploid cytodemes appear to be reproductively strongly isolated, with
little evidence of any hybridization. At least at first, however, there is no
or virtually no differentiation from their diploid parent, either genetically
in terms of isozyme loci or, apart from slight differences sometimes in
pollen or stomata diameter, morphologically (Ness et al., 1989). Current
practice is to accord such autopolyploid populations no formal taxonomic
recognition whatsoever, despite the existence of the barrier to gene flow.
In conclusion, the current approach acknowledges that taxonomic

species are only equivalent by designation, and not by virtue of the nature
or extent of their evolutionary differentiation (Heywood, 1958), and in this
respect the concept embodies an important element of pluralism. It is
sometimes implied that taxonomic species are not delimited on evidence
from cytogenetics or reproductive behaviour, which is usually seen as the
preserve of the 'biological' species concept, but in practice, no such dis-
tinction is drawn and all sources of data are considered fair game. The
practical concept is eclectic in its application. Admittedly, however, what-
ever other evidence there might be, by convention at least one morpho-
logical/anatomical character is required for species recognition. Even at its
most refined, therefore, the taxonomic species concept is essentially an
operational construct with no explicit reference to any particular specia-
tion theory. Nonetheless, it is clear that many botanists believe that, where
possible, the taxa so recognized should represent evolutionary units -
witness the taxonomic treatments of allopolyploid reticulate evolution,
many inbreeders and apomictic complexes, described above, in which
strenuous efforts have been made to delimit putative evolutionary prod-
ucts to be recognized as species (see also Grant, 1981).
It is probably no exaggeration to say that all the c. 300 000 currently
accepted plant species have been diagnosed on this basis (Table 8.1).
There must therefore be a substantial measure of agreement about the
working concept of a plant species (Anderson, 1957). To quote Heslop-
Harrison (1962), 'The success of the morphological-geographical method
Alternative concepts 179
in so many instances indicates with little room for doubt that the varia-
tional units it reveals have a degree of cohesion in nature'. It is the causes
and nature of this cohesion that have prompted fierce debate in the liter-
ature
8.7 ALTERNATIVE CONCEPTS

Aside from the problem areas described earlier, the drive to overturn the
traditional taxonomic concept of a species probably owes as much to its
lack of any well-argued background in speciation theory as it does to its
perceived shortcomings in any particular situation. No causal mechanisms
of the species being recognized are offered as part of the concept and this
produces a vacuum which, its critics say, leads to subjectivity and muddled
thinking. Two alternative concepts, both of which embody hypotheses
about the origin of species, have been offered as potential replacements for
the taxonomic concept. They are the biological and phylogenetic species
concepts, respectively, and are discussed below.
8.7.1 Biological species concept

With the advent in the 1920s and 1930s of the discipline we know as
biosystematics, it was realized that at least one of the processes responsi-
Table 8.1 Number of taxonomic species of living green plants. (Data on non-
vascular plants are taken from Raven et al, 1986; those on vascular plants are from
Mabberley, 1987.)
Group No. of species (approximate)
Chlorophyta (green algae) 7000

Bryophyta
Anthocerotopsida (hornworts) 100
Hepaticopsida (liverworts) 6000
Bryopsida (mosses) 9500
Tracheophyta (vascular plants)
Lycophytina (club-mosses) 1150
Psilophytina (whisk-ferns) 5
Sphenophytina (horsetails) 30
Pterophytina (ferns) 8400
Spermatophytina (seed plants)
Cycadopsida (cycads) 100
Ginkgoopsida (Maidenhair tree) 1
Coniferopsida (conifers) 550
Gnetopsida 70
Angiospermopsida (flowering plants) 240000
ble for the evolution of species was the development of reproductive iso-
lation (Dobzhansky, 1937). This perception of a causal link between
process and product led directly to the so-called biological species con-
cept. Mayr (1942) offered perhaps the most widely-quoted definition as
'groups of actually or potentially interbreeding natural populations which
are reproductively isolated from other such groups'. Blackwelder (1967)
has justifiably criticised the name of this concept on the grounds that it is
no more biological than are any of the other concepts. The idea that plant
species should be delimited at least in part on the basis of breeding behav-
iour in fact has a considerable history, going back, as we have seen, at least
to Ray (1686), and, at least in the sense of true-breeding, is an integral
(though often unacknowledged) part of the traditional taxonomic species
concept. The proposal, however, that reproductive isolation should be
considered the sine qua non of species status has met with considerable
reserve, not to say resistance, by the botanical community. The reasons for
this have been set out in detail by others (Davis and Heywood, 1963;
Raven, 1976; Levin, 1979; Jonsell, 1984; Donoghue, 1985), and will only
briefly be touched on here.
Despite generating much interest, the criterion of actual or potential
reproductive isolation was not widely applied to plants as the sole deter-
minant of species status owing primarily to its general inapplicability in a
group in which, as we have seen, the breeding unit and the morphologi-
cal unit are frequently not correlated and in which uniparental inheri-
tance and lack of or reduced sexuality are not uncommon. Consequently,
gene flow, or lack of it, is not regarded as the sole criterion for specific
status in plants; rather such evidence has been absorbed into the overall
taxonomic approach, to be drawn on as and when considered appropriate.
This is not to deny, however, that there can be an intimate relation-
ship between reproductive isolation and speciation in plants. In some
cases we find what may be described as the ideal situation (at least for
the taxonomist), in which a species is morphologically homogeneous
and well-differentiated from its counterparts, and has a sexual, mixed or
outbreeding mating system but does not hybridize with other species.
Examples of such clear-cut species occur among many annual plants and
can also be found in families such as the Apiaceae and Fabaceae, and in
the genera Allium, Campanula and Sedum (Stace, 1989); Grant (1981) has
listed a selection of others.
There are also now several examples among plants that demonstrate
that shifts in the breeding system from outbreeding to inbreeding, often
mediated by changes in floral morphology, can cause reproductive isola-
tion and rapid speciation (Barrett, 1989). Molecular studies of recent
progenitor-derivative species pairs in Stephanomeria and Clarkia (Gottlieb,
1973, 1974), however, have revealed that although the new species is
morpho-logically distinct it can be no more different at isozyme loci than
are populations of the progenitor species from each other. In other words,
speciation can occur without much associated genetic change, at least in
the genes controlling the fundamental metabolism of the species.
Crawford (1990) has listed other examples.
In his most recent response to the vigorous attacks on the biological
species concept by botanists, Mayr (1992) analysed the local flora of the
township of Concord, Middlesex County, Massachusetts, a well-botanized
area of about 27 square miles. Using a non-dimensional version of the con-
cept, i.e. one to be applied in a given place at a given time, he found he
had trouble dealing with 54 species (6.44% of the flora), all either apomic-
tic or autopolyploid, and by implication regarded this as a successful out-
come. In practice, of course, Mayr used a largely morphological concept
on which to base his inferences about gene flow. It can be seriously doubt-
ed, however, whether gene flow or its absence was necessarily involved in
either causing or preserving any of the phenotypic continuities and dis-
continuities that he found within and between the species (Levin, 1979).
Despite its unpopularity in botanical circles, the soul of the biological
species concept goes marching on in the guise of the recognition species
concept (Paterson, 1985). This essentially turns the biological species con-
cept inside out: the recombinational field of the species is delimited not by
reproductive isolation from other taxa but by shared fertilization mecha-
nisms acting within. Since it is based on reproductive behaviour, it suffers
the same drawbacks in plants as does the biological species concept.
8.7.2 Phylogenetic species concept

Partly as a result of the debate over the biological species concept, and
partly as a result of an increasing understanding of the phylogenetic
approach to taxonomy, it has become increasingly accepted that the
amount and distribution of gene flow is not the only factor influencing the
differentiation or stasis of populations. Other factors include selection and
drift (Carson, 1985), and descent from a common ancestor (Grant, 1980).
The existence of the many cases where morphological discontinuities do
not correspond with ecological discontinuities (Mishler and Donoghue,
1982) has diverted attention away from the role of selection as a cohesive
force and focused it instead on shared evolutionary history as a factor
binding the individuals of a species together (Grant 1980). The mecha-
nisms by which this may occur have been postulated to be of an epistatic
nature (Mishler, 1985).
Over the past 30 years or so several versions of an evolutionary or phy-
logenetic species concept have been proposed (Mishler, 1985). One that
has gained some currency is that proposed by Cracraft (1983:170), accord-
ing to whom a phylogenetic species 'is the smallest diagnosable cluster of
individual organisms within which there is a parental pattern of ancestry
and descent'. This is essentially a refinement of the evolutionary species
concept proposed by Simpson (1961). One of the advantages of the phy-
logenetic (and of the taxonomic) species concept is that it can be applied
to persistent inbreeders and to asexual as well as sexual organisms, and
therefore has a wider applicability among plants than does the biological
species concept. According to Cracraft, any measurable feature can be
used in identifying 'the smallest diagnosable cluster' of individuals. This is
a controversial point because its strict application utilizing, say, DNA data,
could result in the segregation from a single taxonomic species of perhaps
dozens, even hundreds, of cryptic species, each possibly of very few indi-
viduals. The crux of the problem seems to lie in how to delimit the species
from among the welter of individual variants, a position which looks like
square one. Nixon and Wheeler (1990) and Davis and Nixon (1992) have
attempted to resolve this difficulty by suggesting that a phylogenetic
species be recognized as the smallest group of populations (sexual) or lin-
eages (asexual) in which a unique combination of character states is pre-
sent in all comparable individuals, thereby effectively distinguishing
intra- from inter-specific variation. The unique character combination is
then regarded as an autapomorphy. It would also appear to be a require-
ment that the phylogenetic species should be monothetic, in contrast to
the taxonomic species which can be polythetic. Davis and Nixon (1992)
cautioned potential users of this phylogenetic approach about practical
problems regarding the assessment of homology and the effects of under
sampling either characters or populations. Although the phylogenetic
concept makes no reference to gene flow, the way the criteria are used
implies that there should be none, although, as with the taxonomic con-
cept, a limited amount may be allowed if hybrid zones are suspected and
parental taxa do not become merged.
A recent investigation adopting the phylogenetic approach and using
isozyme data showed that populations of the variable western North
American grass Puccinellia nuttalliana could be divided into six species,
each with a unique isozyme profile (Davis and Manos, 1991). Whether
there are any morphological features that can be used by field or herbarium
botanists to identify these isozyme species remains to be seen; it is possi-
ble that some segregate species may correspond. Another test of the phy-
logenetic concept has been conducted by Freudenstein and Doyle (1994)
who examined variation in morphology and plastid DNA in the
Corallorhiza maculata complex in North America. Three phylogenetic
species were recognized by means of unique combinations of morpholog-
ical characters. Preliminary evidence from variation in plastid DNA, how-
ever, showed that one of the species, C. maculata, may have acted as the
ancestor to the other two, i.e. the second species (new and unnamed) may
have evolved from within it, separating Mexican from more northerly
populations, and the third species (C. mertensiana) may have budded off
from the northern populations (Figure 8.1). All four taxa (Mexican and
northern C. maculata, C. mertensiana and the unnamed species) would
qualify as phylogenetic species, each being delimited by a unique combi-
nation of features, molecular and morphological. The two variants of C.
maculata, however, appear to lack any autapomorphies and both they and
the taxonomic species could clearly be regarded as being paraphyletic.
This brings us to a contentious point among proponents of phyloge-
netic (and indeed other) species concepts. By what criterion should the
species be delimited? It has been argued that all species we recognize
should be monophyletic and recognized on the basis of apomorphic char-
acters, e.g. Rosen (1979) and Donoghue (1985). According to Cracraft
(1989), however, the phylogenetic concept does not require that the diag-
nostic features of a species be apomorphic; he nonetheless believed that
most phylogenetic species would turn out to be monophyletic. To what
extent our present taxonomic species are monophyletic is unknown. On
theoretical grounds it may be that only those arising as a result of vicari-
ance processes are of this nature. Other cases, in which phylogenies have
been developed from population studies, have shown that many species
are paraphyletic, in that not all the descendants of the most recent com-
mon ancestor are included in the species. Thus, in studies of cases where
a derivative species originates from its progenitor through rapid specia-
tion, the progenitor is regarded by the analysis as paraphyletic because
not all of its populations have evolved into the new species. Similarly,
when speciation occurs through peripheral isolation and several deriva-
C. trifida
C. maculata (Mexico)
C. sp. nov.
——— C. maculata (North)
C. mertensiana
m
Figure 8.1 Putative phylogenetic tree for American species of the Comllorrhiza
maculata complex, based on morphological (m) and plastid restriction site (p) apo-
morphies. This tree is provisional: further sampling is required to confirm that the
plastid restriction site mutations are truly fixed. (After Freudenstein and Doyle,
1994.)
tives arise from a single progenitor, the latter is regarded as being para-
phyletic, much as in the case of Corallorrhiza maculata discussed earlier.
According to Davis and Nixon (1992), however, phylogenetic species
delimited as the smallest groups that can be recognized on the basis of
unique combinations of attributes can never be monophyletic in the sense
of Hennig (1966) partly because there is no discoverable subordinate hier-
archy, and partly because there are difficulties concerning the concept of
the most recent common ancestor at this level. A properly argued exten-
sion of the Hennigian definition of monophyly to the species level and
below is clearly needed.
Despite the problems with applying the criterion of monophyly to
ancestors, the theoretical arguments for adopting a phylogenetic concept
of some sort are persuasive, perhaps the more so because they comple-
ment and reflect to some extent current practice. For example compare the
definition of the phylogenetic species concept enunciated by Cracraft
(1983) with the taxonomic or morphological concept offered by Du Rietz
(1930). Both require the species to be diagnosable and both attribute
species rank to the least inclusive group so diagnosed. Furthermore, just
as the phylogenetic concept explicitly recognizes that species are evolu-
tionary products involving parental patterns of ancestry and descent, as
already mentioned, so do many practising taxonomists try hard to ensure
that the species they recognize are the products of evolution, although
this is often not made explicit.
Apart from the need to resolve the problems already alluded to of
applying the concept of monophyly at and below the species level, there
are several other issues which need to be addressed before wider agree-
ment is achieved. These include the treatment of the products of hybrid
speciation, the use of cryptic (and also of quantitative) characters for pur-
poses of species recognition, and the treatment of infraspecific variation.
Although not strictly a problem for the phylogenetic species concept
itself, the occurrence of hybridization may present problems in regard to
the cladistic methodology and analysis which the phylogenetic approach
often involves. There is considerable evidence that very many species may
have a hybrid origin and that much plant evolution may be reticulate
rather than cladistic. For example, recombinational speciation at the
diploid level has been shown to be important in the genus Helianthus sec-
tion Helianthus, with three of the stabilized recombinant hybrid derivative
species even sharing the same parents (Rieseberg, 1991). Molecular data
implicate a similar process of homoploid hybrid speciation in other gen-
era, such as Gossypium (Wendel et al, 1991) and Stephanomeria (Gallez and
Gottlieb, 1982). The role of allopolyploidy, a second process involving lin-
eage fusion as a speciation mechanism in plants, is also well-documented,
e.g. as in Spartina anglica, Senecio cambrensis and Tragopogon spp. (Soltis and
Soltis, 1993), and a particularly impressive example of reticulate evolution
involving allopolyploidy is shown by American members of the
Commentary 185
Potypodium vulgare complex (Haufler et at, 1995a,b). It is also notable that
many allopolyploids have evolved more than once, sometimes in well-
separated places, e.g. Senecio cambrensis (Ashton and Abbott, 1992), and
therefore may be described as polyphyletic when considered at a popula-
tion level. Since it has been estimated that 70-80% of angiosperm species
are of polyploid origin (Goldblatt, 1980; Lewis, 1980), it is a mode of speci-
ation which is not only sympatric but also possibly represents the most
common form in flowering plants. Any phylogenetic species concept must
consequently be able to deal the evolutionary pathways and patterns of
variation involved.
One practical difference between the phylogenetic and taxonomic
approaches appears to lie in the nature of the characters distinguishing
the species. In the phylogenetic concept, apparently any character can be
used, whereas in the taxonomic concept, by convention, at least one of
them must be morphological. At least partly on this basis, Freudenstein
and Doyle (1994) proposed to award the Mexican variant of Corallorrhiza
maculata, a diagnosable phylogenetic species, varietal status under C. mac-
ulata because it could not be distinguished reliably on morphological
grounds but only in terms of its plastid genome. The effect on the number
of plant species recognized of abandoning the current morphological
requirement can only be guessed at. The example of Puccinellia nuttalliana
cited earlier may not be a good guide because it is a very variable species,
morphological segregates of which have already been described that may
correspond to some extent with the isozyme species identified by Davis
and Manos (1991). Finally the extent of the problems associated with the
use of quantitative characters in a phylogenetic analysis (Stevens, 1991) at
and below the species level remains to be evaluated.
Another difference between the taxonomic and phylogenetic
approaches lies in the recognition of subspecies and varieties by the tax-
onomic concept. This is not possible under those phylogenetic concepts
in which species are strictly irreducible clusters; no procedures are avail-
able for the recognition of infraspecific taxa because variation below the
species level (as defined by the concept) is not hierarchical. On one
hand, therefore, it may be that many taxonomic subspecies and varieties
would not be recognized at all under a phylogenetic approach that
involved a monothetic species concept because morphological overlap is
allowed, or even required, between them. On the other hand, of course,
many infraspecific variants might be elevated to species status as a result
of the discovery of diagnostic cryptic characters.
8.8 COMMENTARY
If species are regarded as individuals, they appear to be much like genetic
jig-saw puzzles, made up of pieces (individual plants) that fit together
to make a whole. The picture on the upper surface is analogous to the
operational, largely morphological, species concept, with different
species having different pictures of various complexities, corresponding
to their pattern and degree of variability. The prongs and notches of the
individual pieces are analogous to the mechanisms or processes that lock
the individuals of each species into a whole. These processes include not
only reproductive isolation and gene flow, but also genetic drift and nat-
ural selection (Carson, 1985), and epistatic factors (Mishler, 1985) that
result in complex genomic integration and historical (phylogenetic),
developmental and ecological constraints which tie the individuals of a
species together. Templeton (1989) attempted to integrate all these forces
into his cohesion species concept. To what extent this theoretical view of
a species, in which phenotypic variation is delimited by genetical, envi-
ronmental, developmental and phylogenetic cohesion mechanisms, can
actually be translated into an operational construct is not clear. Any
species concept must address not only the theoretical but also the prac-
tical issues posed by the assessment of affinity and rank. What cohesion
mechanisms are the most important in delimiting a particular species,
and on what basis should the latter be recognized - if by the criterion of
monophyly, then how should this be established in practice? Whatever
the cohesion factors, the question of assessing species rank arises: how
continuous should species be? Perhaps not surprisingly, on the one
hand King (1993) regarded the cohesion concept as being no different in
its essentials from the biological species concept, and on the other hand
Endler (1989) regarded it as being close to degenerating into a phenetic
concept.
Where does the analogy of the genetic jig-saw leave us? It is tempting
at this point to invoke some form of uncertainty principle and conclude
that, electron-like, species exist but are impossible to pin down. The
species concept to be adopted must depend on the problem at hand and,
within this constraint, be selected so as to provide maximum insight into
the biological situation rather as the wave and particle properties of an
electron are differentially emphasized depending on context.
Acknowledgement
I should like to thank C.A. Stace for commenting on a draft of this paper.
8.9 REFERENCES
Anderson, E. (1957) An experimental investigation of judgements concerning gen-
era and species. Evolution, 11,260-1.
Ashton, P. and Abbott, R.J. (1992) Multiple origins and genetic diversity in the
newly arisen allopolyploid species, Senecio cambrensis (Compositae). Heredity,
68,25-32.
References 187
Barrett, S.C.H. (1989) Mating system evolution and speciation in heterostylous
plants, in Speciation and its Consequences (eds D. Otte and J.A. Endler), Sinauer
Blackwelder, R.E. (1964) Phyletic and phenetic versus omnispective classification,
in Phenetic and Phylogenetic Classification (eds V.H. Heywood and J. McNeill),
Systematics Association, London, pp. 17-28.
Blackwelder, R.E. (1967) Taxonomy. A text and Reference Book. John Wiley & Sons,
New York.
Brunsfeld, S.J., Solas, D.E. and Soltis, P.S. (1992) Evolutionary patterns and
processes in Salix section Longifoliae: evidence from chloroplast DNA. Systematic
Botany, 17,239-56.
Caplan, A.L. (1981) Back to class: a note on the ontology of species. Philosophy of
Science, 48,130-40.
Carson, H.L. (1985) Unification of speciation theory in plants and animals.
Systematic Botany, 10,380-90.
Cracraft, J. (1983) Species concepts and speciation analysis, in Current Ornithology,
Plenum Press, New York, vol. 1, pp. 159-87.
ation, in Speciation and its Consequences (eds D. Otte and J.A. Endler), Sinauer
Crawford, D.J. (1990) Plant Molecular Systematics, John Wiley & Sons, New York.
Davis, J.I. and Gilmartin, A.J. (1985) Morphological variation and speciation.
Systematic Botany, 10,417-25.
Davis, J.I. and Manos, P.S. (1991) Isozyme variation and species delimitation in the
Puccinellia nuttalliana complex (Poaceae): an application of the phylogenetic
species concept. Systematic Botany, 16,431-45.
Davis, J.I. and Nixon, K.C. (1992) Populations, genetic variation, and the delimita-
tion of phylogenetic species. Systematic Biology, 41,421-35.
Davis, P.H. (1978) The moving staircase: a discussion on taxonomic rank and affin-
ity. Notes from the Royal Botanic Garden, Edinburgh, 36,325-40.
Davis, P.H. and Heywood, V.H. (1963) Principles of Angiosperm Taxonomy,
Edinburgh.
Dobzhansky, T. (1937) Genetics and the Origin of Species, Columbia University Press,
New York.
Doebley, J.E. (1989) Molecular evidence for a missing wild relative of maize and
the introgression of its chloroplast genome into Zea perennis. Evolution, 43,
1555-8.
Donoghue, M.J. (1985) A critique of the biological species concept and recommen-
dations for a phylogenetic alternative. Bryologist, 88,172-81.
Du Rietz, G.E. (1930) The fundamental units of biological taxonomy. Svensk
Botanisk Tidskrift, 24, 333^28.
Endler, J.A. (1989) Conceptual and other problems in speciation, in Speciation and
its Consequences (eds D. Otte and J.A. Endler), Sinauer Associates, Sunderland,
Freudenstein, J.V. and Doyle, JJ. (1994) Plastid DNA, morphological variation, and
the phylogenetic species concept: the Corallorrhiza maculata (Orchidaceae) com-
plex. Systematic Botany, 19,273-90.
Gallez, G.P. and Gottlieb, L.D. (1982) Genetic evidence for the hybrid origin of the
diploid plant Stephanomeria diegensis. Evolution, 36,1158-67.
23, 536-44.
Gilmartin, A.J. (1980) Variation within populations and classification. II. Patterns of
variation within Asclepiadaceae and Umbelliferae. Taxon, 29,199-212.
Gilmartin, A.J., Dobrowolski, J., Soltis, D., Kellogg, E. and Harris, G. (1986)
Variability within and among populations of four grass species. Systematic
Botany, 11, 559-66.
Goldblatt, P. (1980) Polyploidy in angiosperms: monocotyledons, in Polyploidy:
Biological Relevance (ed. W.H. Lewis), Plenum Press, New York, pp. 219-40.
Gottlieb, L.D. (1973) Genetic differentiation, sympatric speciation, and the origin
of a diploid species of Stephanomeria. American Journal of Botany, 60, 545-53.
Gottlieb, L.D. (1974) Genetic confirmation of the origin of Clarkia lingulata.
Grant, V. (1980) Gene flow and the homogeneity of species populations.
Biologisches Zentralblatt, 99,157-69.
Grant, V. (1981) Plant Speciation, 2nd edn, Columbia University Press, New York.
Haufler, C.H., Soltis, D.E. and Soltis, P.S. (1995a) Phylogeny of the Polypodium vul-
gare complex: insights from chloroplast DNA restriction site data. Systematic
Botany, 20,110-19.
Haufler, C.H., Windham, M.D. and Rabe, E.W. (1995b) Reticulate evolution in the
Polypodium vulgare complex. Systematic Botany, 20,89-109.
Hedberg, O. (1958) The taxonomic treatment of vicarious taxa. Uppsala Univ.
Arsskr., 6,186-95.
Hennig, W. (1966) Phylogenetic Systematics, University of Illinois Press, Urbana.
Heslop-Harrison, J. (1962) Purposes and procedures in the taxonomic treatment of
higher organisms, in Microbial Classification (eds G.C. Ainsworth and P.H.A.
Sneath), Cambridge University Press, Cambridge, pp. 14-36.
Heywood, V.H. (1958) The Presentation of Taxonomic Information: A Short Guide for
Contributors to Flora Europaea. Leicester University Press, Leicester.
Heywood, V.H. (1963) The 'species aggregate' in theory and practice. Regnum
Vegetabile, 27, 26-37.
Hughes, J. and Richards, A.J. (1988) The genetic structure of populations of sexual
and asexual Taraxacum (dandelions). Heredity, 60,161-71.
Jonsell, B. (1984) The biological species concept reexamined, in Plant Biosystematics
(ed. W.F. Grant), Academic Press, New York, pp. 159-69.
Jordan, A. (1873) Remarques sur le fait de 1'existence en societe a 1'etat sauvage des
especes vegetales affines. Bulletin Association Frangaise pour I'Avancement des
Sciences, 2.
King, M. (1993) Species Evolution: the Role of Chromosome Change. Cambridge
University Press, Cambridge.
Levin, D.A. (1979) The nature of plant species. Science, 204, 381-4.
Lewis, W.H. (1980) Polyploidy in angiosperms: dicotyledons, in Polyploidy:
Biological Relevance (ed. W.H. Lewis), Plenum Press, New York, pp. 241-68.
Linnaeus, C. (1753) Species Plantarum, Stockholm.
Lovis, J.D. (1977) Evolutionary patterns and processes in ferns. Advances in
Botanical Research, 4, 229-415.
Mabberley, D.J. (1987) The Plant-Book. Cambridge University Press, Cambridge.
References 189
Mayr, E. (1942) Systematics and the Origin of Species. Columbia University Press,
New York.
Botany, 79, 222-38.
Mishler, B.D. (1985) The morphological, developmental, and phylogenetic basis of
species concepts in Bryophytes. Bryologist, 88,207-14.
Mishler, B.D. and Donoghue, MJ. (1982) Species concepts: a case for pluralism.
Systematic Zoology, 31, 491-503.
Ness, B.D., Soltis, D.E. and Soltis, P.S. (1989) Autopolyploidy in Heuchera micrantha
(Saxifragaceae). American Journal of Botany, 76, 614-26.
Newton, A. (1975) Rubus L., in Hybridization and the Flora of the British Isles (ed. C.A.
Stace), Academic Press, London, pp. 200-6.
Nixon, K.C. and Wheeler, Q.D. (1990) An amplification of the phylogenetic species
concept. Cladistics, 6, 211-23.
Raven, P.H. (1976) Systematics and plant population biology. Systematic Botany, 1,
284-316.
Raven, P.H., Evert, R.F. and Eichhorn, S.E. (1986) Biology of Plants, 4th edn, Worth,
New York.
Ray, J. (1686) Historia Plantarum.
Rieseberg, L.H. (1991) Homoploid reticulate evolution in Helianthus: evidence
from ribosomal genes. American Journal of Botany, 78,1218-37.
Rieseberg, L.H., Beckstrom-Sternberg, S.M., Listen, A. and Arias, D.M. (1991)
Phylogenetic and systematic inferences from chloroplast DNA and isozyme
variation in Helianthus sect. Helianthus (Asteraceae). Systematic Botany, 16,50-76.
Rieseberg, L.H. and Soltis, D.E. (1991) Phylogenetic consequences of cytoplasmic
gene flow in plants. Evolutionary Trends in Plants, 5,65-84.
Rosen, D.E. (1979) Fishes from the uplands and intermontane basins of
Guatemala: revisionary studies and comparative geography. Bulletin of the
American Museum. Natural History, 162,267-376.
Ruse, M. (1981) Species as individuals: logical, biological, and philosophical prob-
lems. Brain and Behavioural Sciences, 4, 299-300.
Shi, Y., Gornall, R. J., Draper, J. and Stace, C.A. (in press) Intraspecific molecular
variation in Hieracium L. sect. Alpina (Asteraceae), an apomictic group. Folia
Geobotanica et Phytotaxonomica.
Shore, J.S. and Barrett, S.C.H. (1985) Morphological differentiation and crossabili-
ty among populations of the Turnera ulmifolia L. complex (Turneraceae).
Systematic Botany, 10, 308-21.
Simpson, G.G. (1961) Principles of Animal Taxonomy. Columbia University Press,
New York.
Sneath, P.H.A. and Sokal, R.R. (1973) Numerical Taxonomy, the Principles and Practice
of Numerical Classification. W.H. Freeman, San Francisco.
Soltis, D.E. and Soltis, P.S. (1989) Tetrasomic inheritance in Heuchera micrantha
(Saxifragaceae). Journal of Heredity, 80,123-6.
Soltis, D.E. and Soltis, P.S. (1993) Molecular data and the dynamic nature of poly-
ploidy. Critical Reviews in Plant Sciences, 12,243-73.
Soltis, D.E., Soltis, P.S., Collier, T.G. and Edgerton, M.I. (1991) Chloroplast DNA vari-
ation within and among genera of the Heuchera group (Saxifragaceae): evidence
for chloroplast transfer and paraphyly. American Journal of Botany, 78,1091-112.
Stace, C.A. (1989) Plant Taxonomy and Biosystematics, 2nd edn, Edward Arnold,
London.
Stebbins, G.L. (1971) Chromosomal Evolution in Higher Plants, Edward Arnold,
London.
Stevens, P.P. (1991) Character states, morphological variation, and phylogenetic
analysis - a review. Systematic Botany, 16,553-83.
Styles, B.T. (ed.) (1986) Infraspedfic Classification of Wild and Cultivated Plants.
Clarendon Press, Oxford.
Templeton, A.R. (1989) The meaning of species and speciation: a genetic perspec-
tive, in Speciation and its Consequences (eds D. Otte and J.A. Endler), Sinauer
Wendel, J.F., Stewart, J.M. and Rettig, J.H. (1991) Molecular evidence for homo-
ploid reticulate evolution among Australian species of Gossypium. Evolution, 45,
694-711.
Whittemore, A.T. and Schaal, B.A. (1991) Interspecific gene flow in oaks.
Proceedings of the National Academy of Sciences, USA, 88,2540-4.
Cultivated plant diversity and
taxonomy
/. G. Hawkes
Contacting address: School of Continuing Studies, The University of Birmingham,
Edgbaston, Birmingham B15 2TT, UK
ABSTRACT
Although species diversity in cultivated plants is roughly similar to
that of wild plants, infraspecific diversity is much greater in all but a
few of them. This is undoubtedly due to the hand of humans, by
selecting and hybridizing them and by moving them into environ-
ments far different from their original ones.
The standard taxonomic system used for wild species is also used
for cultivated ones in general outline, but various systems have been
used for infraspecific categories, bearing in mind that their evolution
is largely driven by artificial human-directed selection.
The practice among cultivated plant taxonomists up to the 1940s
and 1950s was to use a microspecies concept. This has now given way,
under the influence of genetical and cytological thought, to the unit-
ing of such microspecies into broadly based large species, to which
breeders and agronomists can more easily relate.
9.1 INTRODUCTION
9.1.1 The basis of diversity and selection

Evolutionary change in cultivated plants is largely due to human selec-
tion, both conscious and unconscious. By unconscious selection is meant
the pressures caused by particular agricultural systems of sowing, culti-
vated and harvesting, and the withdrawal of competition from other
plants and animals in natural ecosystems.
192 Cultivated plant diversity and taxonomy
Agricultural origins took place some 10 000 years ago, and undoubted-
ly conscious selection played a role from very early times. This selection,
long before the advent of scientific plant breeding, involved:
• A search for greater yields.
• Easier and more efficient harvesting, e.g. non-shattering spikes and
spikelets in cereals and non-dehiscing capsules in leguminous crops,
flax, etc.
• Shorter stolons or roots in tuber crops, thus rendering harvesting much
easier.
• A greater range of adaptation to soil, moisture and general climatic con-
ditions as the crop was moved by man into different environments.
• Control of maturity time - long or short growing seasons, uniform
maturity cycle, etc. thus ensuring that all seeds, tubers or fruits matured
at the same time.
It must be recognized that the processes of conscious selection were
minimal or even non-existent during the earlier years of domestication
and farming, but that these gradually became more clearly defined as the
centuries passed.
Over and above this pattern of human selection, the general processes
of mutation and natural selection in cultivated plants also took place, as it
continued to do in their wild ancestors.
As the farmers and their newly domesticated crops moved into other
areas, these crops were no doubt brought into contact with different sub-
species of their wild ancestors and even related wild species. Natural
hybrids and gene flow from such related subspecies and species would
have taken place and these would have added to the genetic diversity of
the crop itself.
As if these processes were not enough to promote morphological and
genetic diversity in the cultivated species, farmers, and later, plant breed-
ers have also been consciously hybridizing and selecting cultivated plants,
thus adding to their diversity.
All these processes have tended to provide much more complex
patterns of variation in cultivated plants than in wild ones. Species concepts
have become even less clear than in some completely wild species, which
are themselves often bedevilled with polyploidy, pillar complexes and
other untidy situations which often render the task of the cultivated plant
taxonomist very difficult.
9.2 POLYPLOIDY
Polyploids occur in cultivated plants under primitive agriculture, and are
not by any means freaks of wide artificial hybridization. It has been stated
on many occasions that polyploidy is more common in cultivated than in
Species concepts in cultivated plants 193
wild species. However, out of 40 widely grown species of field and tree
crops, 20 are completely diploid, while the rest are tetraploid, diploid and
tetraploid, or possess a range of ploidy levels (Table 9.1). Certain fruit
trees, such as apples and pears are basically diploid but possess triploid
cultivars. Several crops possess diploid and tetraploid species (cherries,
coffee, cotton), while others such as wheat, oats, potatoes and plums, pos-
sess a wider range, generally having developed these by ancient natural
hybridization and chromosome doubling. Highly variable ploidies occur
in sugar cane and yams (Hawkes, 1983: 23), but these are exceptions to the
general rule, with the wide ploidy range seeming to be preserved only by
means of vegetative reproduction.
9.3 SPECIES CONCEPTS IN CULTIVATED PLANTS

As with wild species there are some taxonomists of cultivated species who
see and describe new species for every recognizable variant - the so-called
splitters - and on the other hand, there are those who try to group togeth-
er as many variants as they can - the so-called lumpers. Horticulturalists
were certainly more likely in the past to see a new species in every
Table 9.1 Levels of ploidy in certain widely grown species of field and tree crops*
Diploids Polypoloids
Almond Apple (2x, 3x)

Barley Banana (2 x, 3 x, 4 x)
Cabbage Canna (2x, 3x)
Chickpea Cassava (4 X)
Cocoa Cherry (2 x, 4 x)
Coconut Coffee (2 x, 4 x)
Cocoyam Cotton (2 x, 4 x)
Coix Curcuma (2 x, 3 x)
Ginger Oats (2x,4x, 6x)
Lentil Peanut (4 x)
Maize Pear(2x,3x)
Oil palm Plum(2x,4x,6x)
Olive Potato (2x,3x,4x,5x)
Peach Rose(2x,3x,4x,5x,6x)
Phaseolus beans Sugar cane (variable)
Pineapple Sweet potato (6x)
Rice Taro(2x,3x)
Rye Tobacco (4x)
Tea Wheat (2x,4x,6x)
Tomato Yam (variable)
*From The Diversity of Crop Plants (Hawkes, 1983)

variant, particularly if it appeared to breed true. This, of course, can also
happen with groups of wild species, though the temptation to create new
species of horticultural or agricultural plants seems to have been yielded
to with greater abandon in the past.
Thus, de Wet et al. (1986) show that the cultivated sorghums were once
divided by Snowden into 28 species, 156 taxonomic varieties and 521
forms. All of these are now clearly seen to be conspecific, under the single
species Sorghum bicolor. Similarly, N.I. Vavilov and his co-workers in the
1920s and 1930s divided the known wheat species into one diploid
(Triticum monococcum, wild and cultivated), eight tetraploids (of which
one, T. dicoccoides, is wild); and four hexaploids (including T. aestivum). All
these were allotted innumerable varieties and forms, all with their Latin
names (see Vavilov - English translation, 1992, of Russian compilation,
1987). It is now considered best to retain three species T. monococcum (2x),
T. dicoccum (4x), and T. aestivum (6x), with all the diverse agricultural
variants given no more than vernacular names. The same solution has
taken place with other cereals, such as barley, rye and oats.
The cultivated potato has gone through a similar sequence of events.
Linnaeus in 1768 named it Solanum tuberosum, and thus it was continued
to be referred to until 1929 (though S. esculentum Neck, was proposed in
1768, and S. sinense by Blanco in 1837). However, as a result of the Russian
expeditions to South America by S.V. Juzepczuk and S.M. Bukasov in 1927
a polyploid series was discovered, ranging from diploid (2n = 24), triploid
(2n = 36), tetraploid (2n = 48) and pentaploid (2n = 60) - all cultivated
materials. They split the tetraploid S. tuberosum into an Andean and a
Chilean species, and in addition made six diploid, five triploid and one
pentaploid species. After I returned from my first collecting expedition to
South America I regret to say, that with an enthusiasm which I have since
learned to regret, I added four diploid species, five diploid varieties, one
triploid species and one triploid variety, after the fashion of the Russians.
The Russians themselves had described among the Chilean tetraploids
nine varieties and 28 forms, while they split the Andean tetraploids into
two species, 21 varieties and 55 forms (Table 9.2). Ultimately, this sort of
classificatory system became of little or no practical use, since natural
hybridization between clones can lead to an almost limitless number of
genotypes.
The simplified classification shown in Table 9.2 has been helped con-
siderably by the experimental recreation of the polyploid species and even
one of the diploids (see Figure 9.1 and Hawkes, 1990). From this it has
Species concepts in cultivated plants 195
Table 9.2 Russian and other infraspecific classifications of cultivated potatoes
Ploidy Species Previous classification

level now recognized Species Varieties Forms
S. ajanhuiri S. ajanhuiri
S. phureja S. phureja
(two subspecies) S. rybinii
S. boyacense
2x S. kesselbrenneri
(2n = 24) S. cardenasii
S. ascasabii
S. caniarense
S. stenotomum S. stenotomum
(two subspecies) S. goniocalyx
S. churuspi
S. yabari
S. juzepczukii S. juzepczukii
3x S. chaucha S. chaucha
(2n = 36) S. tenuifilamentum
S. mamilliferum
S. coeruleiflorum
S. tuberosum S. tuberosum 17 37
subsp. tuberosum S. esculentum
S. sinense
S. chiloense
S. cultum
S. sabinii
4x S. molinae
S. leptostigma
S. diemii
S. sanmartinense
S. oceanicum
S. ochoanum
S. zykinii
S. tuberosum S. andigena 21 55
4x subsp. andigena S. herrerae
S. apurimacense
5x S. curtilobum S. curtilobum
become fairly clear from experimental studies that the diploid cultivated
species S. stenotomum was the first to be cultivated and that natural
hybridization, followed by unconscious selection gave rise to all the others.
Undoubtedly, environmental pressures and isolation gave rise to sub-
species tuberosum in Chile; and environmental selection for frost resistance
at high altitudes 'fixed' the hybridogenic species S. ajanhuiri, S. juzepczukii
and S. curtilobum.
We have now arrived at a stage where the complex series of varieties
and forms named and described by Bukasov, Juzepczuk and Lechnovich
no longer matter. They played an important role in the 1920s and 1930s by
highlighting the value of the immense diversity of potatoes in the Andean
mountains and in southern Chile. Now, however, they are no longer of
very great importance. In fact, the process of simplification was taken
even further by Dodds (1962) who used a group classification (Table 9.3),
under a single species, S. tuberosum (sensu lato).
As one can see, this corresponds very closely with the present author's
species, classification, apart from subgroup Ila - Amarilla, which is no
more than the northern yellow-fleshed forms of S. phureja.
9.4 CONCLUSIONS
What, then, are the units of biodiversity in cultivated plants and how do
they differ from those of wild plants? To a large extent they depend on the
breeding system. If the species are inbreeding, such as the Old World cere-
als, Linnean systems may generally be useful, though there is a danger of
Wild S. acau/e S. sparsipilum S. leptophyes S. megistacrolobum

species (4x) (2x) (2x) (2x)
i
subsp. andigena
(4x)
(2x)
r
•——— S. futoerosum 4—— S. stenotomum ———*S. ajanhuiri (Yari)
(2x)
i
Cultivated S. tuberosum '—»• S. chaucha •*-
species subsp. tuberosum (3x)
(4x)
S. curt 'lobum —*S. phureja

(5x) (2x)
' (3x)
Figure 9.1 Evolutionary relationships of cultivated potatoes and their ploidy lev-
els. (Adapted from Hawkes, 1990, with kind permission of Belhaven Press.)
Conclusions 197
Table 9.3 K.S. Dodds' classification of S. tuberosum L. (sensu laid) with ploidy lev-
els and present classification added
Dodds' groups' Ploidy Species equivalent
I Group Stenotomum 2x S. stenotomum Juz. et Buk.

S. ajanhuiri Juz. et Buk.
I-A Subgroup Goniocalyx 2x S. stenotomum Juz. et Buk.
subsp. goniocalyx (Juz. et Buk.)
Hawkes
II Group Phureja 2x S. phureja Juz. et Buk.
Ha Subgroup Amarilla 2x Part of S. phureja Juz et Buk.?
(not clarified in text)
III Group Chaucha 3x S. chaucha Juz. et Buk.
IV Group Andigena 4x S. tuberosum L. subsp. andigena
Hawkes
V Group Tuberosum 4x S. tuberosum L. subsp.
tuberosum
VI Solanum x juzepczukii Buk. 3x S. juzepczukii Buk.
VII Solanum x curtUobum Juz. 5x S. curtilobum Juz. et Buk.
et Buk.
*From Dodds (1962: 517-39.)
over-complication when no more than different geographical or altitudi-

nal races are considered to be distinct species. Some attention should be
paid to ploidy levels, and presence or absence of crossability barriers
should be observed in these cases. Sometimes the wild prototypes (as with
Triticum boeoticum/monococcum) need to be considered as the same Linnean
species, with perhaps subspecies differentiation only. In other cases, as
with Zea mays and its wild prototype Zea (or Euchlaena) mexicana, specific
or even generic distinctions need to be made. With Solanum tuberosum,
although this author believes it to be a natural hybrid of Solanum stenoto-
mum x S. sparsipilum, others are less convinced, and consider it to be an
autotetraploid of its ancestral diploid, S. stenotomum. In this instance it is
clearly better to use the Linnean name rather than give it a hybrid desig-
nation.
As I have stressed before (Hawkes, 1986), in cultivated plants the low-
est taxonomic rank may not be the cultivar as some workers have consid-
ered (e.g. Parker, 1978). Under primitive agriculture the cultivar hardly
exists, and its more elemental place is taken by the population (or
'pseudocultivar'). This still applies in many parts of the world today, as for
instance in Ethiopia (Worede, 1991), where in certain regions cultivars do
not exist as such but are preceded by populations, even in predominantly
inbreeding species such as barley (Hordeum) and tetraploid wheats.
Vegetatively propagated crops such as sweet potatoes, cassava and
potatoes are propagated in developed countries as clones and may be clas-
sified as such. However, in Third World countries in and around the cen-
tres of diversity of these crops there may be considerable outcrossing.
Potatoes in any case are mostly allogamous and new seedlings which arise
naturally will clearly possess different gene combinations from their par-
ents. Hence; a clonal classification will clearly be unacceptable. Here, as
we have seen earlier, a broad classification based on general species delim-
itations, as with wild species, is the most appropriate one to adopt.
These broadly based species and subspecies units must surely be the
generally acceptable unit of diversity in most cultivated plants.
Nevertheless, in largely inbreeding or in entirely clonally reproducing
apomictic species, then the 'pseudoclone' or inbred population, or even
the clones themselves may feature as the most easily recognizable units.
9.5 REFERENCES
Dodds, K.S. (1962) Classification of cultivated potatoes, in The Potato and its Wild
Relatives (ed. D.S. Correll), Texas Research Foundation, Renner, Texas.
Hawkes, J.G. (1983) The Diversity of Crop Plants, Harvard University Press,
Cambridge, Massachusetts, USA.
Hawkes, J.G. (1986) Infraspecific classification - the problems, in Infraspecific
Classification of Wild and Cultivated Plants (ed. B.T. Styles), Clarendon Press,
Oxford.
Hawkes, J.G. (1990) The Potato. Evolution, Biodiversity and Genetic Resources,
Belhaven (Pinter) Press, London.
Parker, P.P. (1978) The classification of crop plants, in Essays in Plant Taxonomy (ed.
H.E. Street), Academic Press, London.
Vavilov, N.I. (1992) Origin and Geography of Cultivated Plants. Translated by Doris
Love, Cambridge University Press, from original compilation by V.F. Dorofeyev
and A.A. Filatenko, 1987, Nauka.
de Wet, J.M.J., Harlan, J.R. and Brink, D.E. (1986) Reality of infraspecific taxonom-
ic units in domesticated cereals, in Infraspecific Classification of Wild and Cultivated
Plants (ed. B.T. Styles), Clarendon Press, Oxford.
Worede, M. (1991) An Ethiopian perspective on conservation and utilization of
plant genetic resources, in Plant Genetic Resources of Ethiopia (eds J.M.M. Engels,
J.G. Hawkes and M. Worede), Cambridge University Press, Cambridge.
10
Species of marine invertebrates: a
comparison of the biological and
phylogenetic species concepts
N. Knowlton and L A. Weigt
Contacting address: Smithsonian Tropical Research Institute, Apartado 2072, Balboa,
Republic of Panama
ABSTRACT
Traditional morphological approaches have substantially underesti-
mated species diversity in living marine invertebrates, as assessed by
either the biological or phylogenetic species concepts. The degree to
which these concepts succeed individually or agree with each other
depends on the group and the biogeographic context. In sympatry,
the biological and phylogenetic species concepts should yield the
same result, because reproductive incompatibility implies at least one
diagnostic difference between isolated forms. Hybridization in sym-
patry between partially isolated forms may be a problem for both
species concepts in some groups, although evidence for this in the
field is limited. It is often difficult to find qualitative morphological dif-
ferences between forms that can be unambiguously recognized by
other characters, so that a phylogenetic species concept that depends
on morphological characters will miss many reproductively isolated
forms. In allopatry, the differences between the two species concepts
are potentially much greater, because the phylogenetic species con-
cept has the potential to recognize any diagnosably distinct popula-
tion at the species level, regardless of its triviality. In groups with
extensive dispersal ability, there may be predictable relationships
between genetic and reproductive divergence that allow taxonomic
decisions to be made using either species concept. In groups with lim-
ited dispersal ability and a propensity for founder events and local
extinction, substantial and complex patterns of genetic variation will
prove challenging for both species concepts. The phylogenetic species
concept makes no special distinction between species and higher level
200 Species of marine invertebrates
taxa, while the biological species concept places species at the bound-
ary between reticulation and cladogenesis. This appears to be an
important and well-defined boundary for many marine invertebrates,
and thus merits special taxonomic recognition.
10.1 INTRODUCTION
Morphologically defined species remain the rule in nearly all groups of
marine invertebrates, but recent work has revealed that many so-called
species are in fact complexes of taxa that can be most readily distinguished
using genetic, behavioural or ecological characters (Knowlton, 1993).
Nevertheless, most of these sibling species exhibit subtle (and sometimes
not so subtle) morphological differences that were previously ascribed to
intraspecific variation, based on a priori assumptions of wide geographic
range or extensive non-genetic plasticity.
How much variability should a single species encompass? The taxo-
nomic response to this question will depend upon the nature of the
species concept to be employed. Two classes of options currently domi-
nate the literature (Claridge et al., 1997: Chapter 1). The first class is epito-
mized by the biological species concept (Mayr and Ashlock, 1991). Species
are defined in principle on the basis of reproductive compatibility,
although indirect evidence for the existence of reproductive barriers
marking species boundaries is acceptable with this approach (Avise and
Ball, 1990). Templeton's (1989) cohesion species concept is related in
accepting the biological species concept for those groups for which it
works well, but it also utilizes criteria that can be applied to groups that
are asexual, subdivided, or that hybridize extensively. The second class of
species concepts consists of a group of cladistically based approaches, all
of which have been loosely referred to as the phylogenetic species concept
(Mishler and Theriot, 1997). In its least restrictive usage (Cracraft, 1989,
1997: Chapter 16), a phylogenetic species is simply the minimum diagnos-
able taxonomic unit based on any qualitative character. Some workers,
however, insist that minimum units must be strictly monophyletic and
thus defined by derived characters (reviewed by Smith, 1994), and others
argue that some monophyletic groups are too trivial to merit recognition
at the species level (Mishler and Theriot, 1997).
The purpose of this chapter is to compare the implications of these two
classes of approaches for the definition of species in marine invertebrates
in practice, without getting overwhelmed by the alpha taxonomy of the
species concepts themselves. For simplicity, we will focus primarily on the
biological species concept of Mayr (Mayr and Ashlock, 1991) and
Cracraft's (1989,1997: Chapter 16) version of the phylogenetic species con-
cept. We shall not consider the special problems raised by obligately clon-
al or selfing life-histories, since these are rare in marine invertebrates.
Corals and shrimps as case studies 201
The comparison between the biological and phylogenetic approaches is
clarified by considering sympatric and allopatric taxa separately. In sym-
patry, the two approaches make the same recommendations in principle,
for taxa that are reproductively isolated in sympatry are also by definition
diagnosably distinct in at least the character that generates the isolation.
However, as the title of this symposium (The Units of Biodiversity: Species in
Practice) implied, the translation of theory into practice is not always
straightforward. Reproductively isolated, sympatric taxa can generally be
distinguished by ecological, behavioural, genetic or morphological differ-
ences, but these differences may be quantitative rather than qualitative,
and thus not diagnosable in practice by the criteria of phylogenetic species
concepts. Moreover, reproductive isolation is itself a quantitative charac-
ter when isolation is not complete, and can thus be a problem for both
species concepts in deciding the status of morphs that are partially inter-
fertile in sympatry. In allopatry, the differences between these approach-
es are potentially much more marked, since for the phylogenetic species
concept any geographically isolated population with a diagnosably dis-
tinct characteristic can be recognized at the species level. Founder popu-
lations begun by a small number of individuals may often be diagnosably
distinct at the genetic level, and thus would be species using the approach
favoured by Cracraft (1989), regardless of the triviality of the difference.
10.2 CORALS AND SHRIMPS AS CASE STUDIES

The differences between the biological and phylogenetic species concepts
for practical taxonomy are best evaluated by considering real organisms
rather than hypothetical situations. Below we discuss the implications of
these approaches for corals and snapping shrimp. These two groups share
a bad reputation among practising taxonomists trying to define species,
but the problems they pose for these alternative species concepts are quite
different.
10.2.1 Reef-building corals and the problem of quantitative characters

With few exceptions, alpha taxonomy of corals is entirely based on skele-
tal characters (but see Lang, 1984). Most reef-building corals are colonial,
and characteristics of the skeletal cups secreted by individual polyps,
called corallites, play a major role in defining species boundaries in many
genera (Budd et al, 1994a). Because corallites often vary considerably
within a colony (a genetic individual), species boundaries were tradition-
ally defined to encompass much of this variation (Wood Jones, 1907).
Colony growth form is also known in some cases to exhibit phenotypic
plasticity in response to wave energy and light level (Willis, 1985), the
latter a consequence of the obligate symbiosis between reef-building
corals and photosynthetic dinoflagellates. These features have led most
coral taxonomists to accept large amounts of intraspecific variation at the
level of both the corallite and the colony.
The problems with this approach are well illustrated by the coral
Montastraea annularis. This 'species' is the most abundant and best studied
coral on Caribbean reefs. It exhibits enormous variability in colony mor-
phology, ranging from massive heads to columns to plates. Computer
models and transplant experiments designed to evaluate the effect of
depth on growth form (Graus and Macintyre, 1976; Dustan, 1979) made it
a text-book case of the importance of phenotypic plasticity. Nevertheless,
the different morphologies can be found side by side, and intermediates
between the major types of colony morphology are rare (Graus, 1977).
Recently, a number of features have been found that are strongly con-
cordant with the different patterns of colony growth. These include
corallite morphometrics, skeletal density and growth rate, stable isotope
signatures, aggressive behaviour, allozymes, and reproductive biology
(Knowlton et al, 1992, and in press; Van Veghel and Bak, 1993; Van
Veghel and Kahmann, 1994; Weil and Knowlton, 1994; A.F. Budd and N.
Knowlton, in preparation). However, most of these differences, including
those likely to be useful in identifying skeletal material, are quantitative
rather than qualitative.
The response of coral systematists to these findings has been mixed.
Because numerous and presumably unrelated characters were concor-
dant, Weil and Knowlton (1994) resurrected two previously synono-
mized names to recognize the three common shallow water forms. Van
Veghel and Bak (1993), in contrast, referred to them as morphotypes
rather than species, based on the absence of qualitative differences in the
same set of data. Reviewing this situation, Veron (1995: 33) concluded
that there may be no correct answer because species are arbitrary. The
species level taxonomy of many other corals is comparably problematic
(Veron, 1995).
There are several general and somewhat inter-related explanations for
this failure to find reliable qualitative differences. First, reproductively
isolated species may be genuinely similar in genetic or morphological
characters, either because divergence is slow or because the species are
relatively young. Recently or slowly diverging species may still share the
same polymorphisms across much of the genome, so that finding quali-
tative or fixed differences can be equivalent to searching for a genetic
needle in a haystack (Avise and Ball, 1990). Romano (1995), for example,
found comparatively little genetic divergence even between groups
whose fossil record indicates a long history of isolation, suggesting that
the molecular 'clock' of corals ticks slowly. Some corals have long gener-
ation times, which other things being equal, should act to slow the rate of
divergence (Gillespie, 1991).
Second, coral species, like plants, may be particularly prone to
hybridization (Veron, 1995). The morphological simplicity of corals may
underlie the successful production of hybrids in the laboratory (Wallace
and Willis, 1994), and the annual simultaneous spawning of congenerics
documented for many corals could provide regular opportunities for
interspecific hybridization in the field (Wallace and Willis, 1994; Veron,
1995). Even when hybrids are less fit than the products of intraspecific
matings, selection to avoid such matings is ultimately constrained by the
fact that even a severely inferior hybrid is fitter than an unfertilized egg.
However, it should be emphasized that hard evidence for rampant nat-
ural hybridization in corals is lacking. Indeed, the three described species
of the M. annularis species complex show considerable overlap in the date
of spawning, but appear nevertheless to be reproductively isolated by a
combination of barriers to fertilization or development and differences in
the hour of peak spawning (Knowlton et al., in press). Moreover, in
Acropom, there are several examples of incompatibility between morpho-
logically similar colonies that are supposed to be conspecific (Wallace and
Willis, 1994), suggesting that the importance of breeding barriers may
have been underestimated as well as overestimated.
Molecular techniques have revolutionized the detection of sibling
species in many groups, but in corals, apart from protein electrophoresis,
they have been little used due to technical difficulties. These difficulties
are likely to be resolved in the coming years, and clearly defined species
may emerge from these studies. On the other hand, if closely related
species of corals diverge slowly and hybridize frequently, there may be no
magic molecular bullet for recognizing hybrids or defining species, even
in sympatry.
In summary, corals can be problematic for a variety of species concepts,
even in sympatry, due to the scarcity of diagnostic differences and a poor-
ly understood potential for hybridization. In the absence of a consensus in
sympatry, the situation in allopatry or even in different reef habitats is, of
course, more problematic. In the M. annularis species complex, for exam-
ple, deeper water forms converge on a plate-like morphology, making it
difficult to distinguish species or evaluate relationships between shallow
and deeper water forms. The three basic morphologies can be recognized
throughout the Caribbean, but show differences from site to site which are
also hard to interpret. Veron (1995) described many comparable situations
for Pacific corals.
10.2.2 Snapping shrimp and the problem of allopatric variation

Species of snapping shrimp in the genus Alpheus (Alpheidae) are numer-
ous and often difficult to distinguish using conventional morphological
characters in preserved material. In sympatry, however, many morpho-
logically similar species can be readily identified by their colour patterns
when alive (Knowlton and Mills, 1992), as well as by fixed allozyme dif-
ferences and substantial divergence in mitochondrial DNA (mtDNA)
sequences (Knowlton et al., 1993 and unpublished data). It is also possible
directly to assay reproductive cohesion, because males and females are
often found paired, and are usually aggressive to all individuals except
potential mates. Interspecific pairings are not observed in the field and
are almost impossible to establish in the laboratory, suggesting that
hybridization is uncommon. Thus, in contrast to corals, both the biologi-
cal and phylogenetic species concepts have little difficulty in defining
species boundaries in sympatry when the appropriate characters are
employed.
Neotropical alpheids from opposite sides of the Isthmus of Panama are
particularly useful for exploring the nature of species in allopatry. The
Isthmus now divides what was once a continuous tropical sea, and the
faunas of the eastern Pacific and Caribbean contain numerous examples of
transisthmian sister taxa (Knowlton et al., 1993). Detailed geological and
palaeontological analyses provide strong support for dating the final clo-
sure of the Isthmus at approximately 3 million years ago (Coates et al.,
1992). While the biological history of the region is more complex than
commonly assumed (see below), it nevertheless represents the best
marine example of the classic form of geographic isolation - one large
population being divided into two still large populations that subsequent-
ly slowly diverge through time.
Seven pairs of transisthmian sister taxa have been examined using a
suite of molecular and behavioural techniques (Knowlton et al., 1993).
They show considerable but concordant variation in the degree of diver-
gence exhibited, suggesting that some pairs were isolated more recently
than others. Although the most similar transisthmian pairs show little
aggression when males from one ocean are placed with females from the
other ocean, they almost never produce fertile clutches of eggs. This con-
trasts with the routine production of fertile clutches when conspecifics
from the same ocean are paired under identical laboratory conditions. The
general conclusion to be drawn from these experiments is that 3 million
years is long enough to result in substantial reproductive isolation in these
organisms.
Can we use the information obtained from transisthmian pairs to
interpret other patterns of divergence in allopatry where it has not been
possible to study reproductive compatibility directly? This appears to be a
reasonable procedure for two reasons. First, the relatively large number of
transisthmian pairs studied strengthens our confidence in the relationship
between molecular and reproductive divergence. Second, the classical or
dumb-bell mode of geographic isolation is probably the most sluggish
way of generating new species, based on theory (Templeton, 1981), empir-
ical observations in other groups (Coyne and Orr, 1989), and limited
observations for the alpheids themselves (unpublished data). Thus, we
may underestimate reproductive isolation for a given level of molecular
divergence by using the transisthmian data as a null model, but we are not
likely to overestimate it.
The widely scattered island groups of the Pacific provide an appropri-
ate arena for comparing the biological and phylogenetic species concepts
in a more complex pattern of allopatry. To this end, we here summarize
preliminary data on Alpheus lottini, an obligate symbiont of the corals
Pocillopom and Stylophora that ranges in distribution from the eastern
Pacific to the Red Sea. It is a conspicuous and well-studied snapping
shrimp, but no attempt has previously been made to evaluate variability
across its wide range.
Examination of living material from a variety of sites across the Pacific
immediately revealed two clearly distinct colour patterns: animals were
characterized by either a solid black stripe down the carapace and paired
blotches down the abdomen (type A), or by speckles scattered across both
regions (type B). Limited collections from Panama, Pohnpei, Clipperton
and Moorea contained only type A, those from Hawaii and Guam only
type B, while both types were found in Palau (mated assortatively).
Mitochondrial DNA sequences from types A and B (for the same gene
region as was studied for transisthmian taxa; see caption of Table 10.1 for
methods) had an average corrected sequence divergence of 10-13% (Table
10.2). This corresponds to an estimated time of divergence of 4-5 million
years, based on the transisthmian taxa (Figure 10.1). This degree of
sequence divergence was always associated with reproductive isolation
across the Isthmus, and the finding of assortatively mated pairs at the one
site where they were found together (Palau) also supports an interpretation
of strong reproductive isolation. Thus, both the biological and phyloge-
netic species concepts would recognize types A and B as specifically dis-
tinct.
The molecular data also reveal more subtle allopatric structuring within
these two types, however. The clearest is that between the Hawaiian and
other type B taxa, which differ by about 4.5% in corrected sequence diver-
gence. This is less than that seen between seven transisthmian sister
species (Knowlton et al, 1993), but still substantial, particularly in light of
the fact that even the least divergent of these (6.6%) never successfully
interbred in laboratory experiments. From the perspective of the biological
species concept, these taxa are likely to be specifically distinct, but explicit
tests of their compatibility to further refine the relationship between genet-
ic and reproductive divergence at the low end of the divergence scale are
necessary. For the phylogenetic species concept, however, there is no
ambiguity, because the Hawaiian shrimp can be distinguished genetically
from the other type B shrimp (20 transitions and three silent transversions
Table 10.1 Aligned sequence haplotypes for variable positions within the 564 bp area of analysis of the COI mtDNA gene ofAlpheus
lottini complex (parentheses indicate multiple individuals with same halotype)
00000000000000000001111111111222222222222222223333333333344444444444444555555555555
Position 01122233344566679990011246789012223344577888990013345567700334456778899001122334455
62814703928136950362847042198462581408739258140920324505858581432170958170658140325
B-PALAU1 AATGCCCTAGTCAAGGACGGTAAGATGTCTCK^CGAGAACCCGCT^
B-PALAU2
B-GUAMl ............................... .T. ..... .T. .........................................
B-GUAM2 ............................... .T. .................................................
B-GUAM3(2) ....... .G. ............................. .T. .........................................
B-HAWAIIK2) G. .A. .T.G. . .G.AA. . . .C. .A. .A. ... .TG. . .A. ....... .AT.A.T.GC.G.CT. ........ .T. .... .A. A. .
B-HAWAII2 G. .A. .T.G.C.G.AA. .. .C. .A. .A. ... .TG. . .A. ....... .AT.A.T.QC.G.CT. ........ .T. .... .A.A. .
B- HAWAII 3 G. .A. .T.G. . .G.AA. . . .C. .A. .A. ... .TG. . .A. ....... .AT.A.T.GC.G.CT. ........ .T. .... .A.A. .
A-PALAU1 . GCATAGCCACTGGAAGTA . CGGAGCAATCTG . . . AG . GG . TTCACGA . GACCC . CTG . CCAT . CG . CGCA . GG . . . TAG .

A-PALAU2 . GCATAGCTACTGGA . . TA . CGGAGCAATCTG . . . AG . GG . TTCACGA . GACCC . CTG . CCAT . CGTCGCA . GG . . . TA
A-POHNPEI1 .GCATAGCCACTGGA..TA.CGGAGCAATCTG...AG.GG.TTCACGA.GACCC.CTG.CCAT.CG.CGCA.GG...TAG..A
A- POHNPEI2 . GCATAGCCACTGGA. . TA. CGGAGCAATCTG. . . AG. GG. TTCACGA. GACCC. CTG. CCAT. CG. CGCA. GG. . . TAG. . A
A-POHNPEI3 .GCATAGCCACTGGA.7TAACGGAGCAATCTGT..AG.GG.TTCACGA.GACCC.CTG.CCAT.CG.CGCA.GG...TAG..A
A-MOOREAl .GCATAG.CACT. . . . .TA.CGGAGCAAT.TG. . . .G. .G.TTCAC.A. .ACCC.CT. .CCAT.CG.CGCA.GG. . .TAG. .A
A-MOOKEA2 .GCATAG.CACT.....TA.CGGAGCAAT.TG....G..G.TTCAC.A..ACCC.CT..CCAT.CG.CGCA.GG..T.AC..A
A-MOOREA3 .GCATAG.CACT.....TA.CGGAGCAAT.TG....G..G.TTCAC.A..ACCC.CT..CCAT.CG.CGCA.GGG..TAG..A
A-CLIPPERTON1 .GCATAG.CACT.....TA.CGGAGCAAT.TG....G..G.TTCAC.A..ACCC.CT..CCAT.CG.CGCA.GGG..TAG..A
A-CLIPPERTON2 .GCATAG.CACT.....TA.CGGAGCAAT.TG....G..G.TTCAC.A..ACCC.CT..CCATT.GTCGCA.GGGT.TAG.TA
A-PANAMA1 .GCATAG.CACT..T.GTA.CGGAGCAAT.TG..TAG..G.TTCAC.A..ACCC.CT.TCCAT.CG.CGCA.GGG..TAC..A
A-PANAMA2(2) .GCATAG.CACT..T.GTA.CGGAGCAAT.TG...AG..G.TTCAC.A..ACCC.CT..CCAT.CG.CGCA.GGG..TAG..A
A-PANAMA3 .GCATAG.CACT..T.GTA.CGGAGCAAT.TG...AG..G.TTCAC.A..ACCC.CT..CCAT.CG.CGCA.GG...TAG..A
Dots represent concordance with the top sequence. Mitochondrial DNA methods as in Knowlton et al. 1993 with the following modifications:
primers (5'-3' positions in the amplification product) PCR and sequencing = COI (1-20) and Alpheus specific primer 1 [5' CAT TTA GGC CTA AGA
ACT GTT G 3' (619-640)]; internal sequencing: light strand - ALP7 [TGA CTT GGA ACC CTC CAT GG 3' (304-323)] and ALPS [5' ATT GCY CAC
TGA TTC CCC YTA TT 3' (514-536)]; heavy strand - ALP2 [5' CCR TGG AGG GTT CCR ACT CA 3' (304-323)]. Double stranded sequencing as in
method 2 of Kessing et al. (1989). Sequences in Genbank (u76428-u76455).
Table 10.2 Kimura (1980) corrected percent sequence divergence values* (calculated by Sequencer 3.0; Kessing, 1995) averaged by site
for two types of Alpheus lottini
B-PALAU B-GUAM B-HAWAII A-PALAU A-POHNPEI A-MOOREA A-CLIPPERTON A-PANAMA

1
B-PALAU 0.0
XXX-XXX
8 6
B-GUAM 0.4 0.3
0.2-0.4 0.2-0.5
8 16 6
B-HAWAII 4.7 4.6 0.1
4.6-4.8 4.4-4.8 0.0-0.2
4 8 8 1
A-PALAU 12.7 13.0 11.8 0.7
12.6-12.8 12.8-13.2 11.5-12.0 XXX-XXX
6 12 12 6 3
A-POHNPEI 12.5 12.7 11.7 0.5 0.3
12.4-12.8 12.6-13.0 11.5-11.8 0.4-0.7 0.0-0.4
8 12 12 6 9 3
A-MOOREA 10.3 10.5 10.7 2.4 2.1 0.4
10.2-10.4 10.4-10.8 10.6-11.1 2.2-2.5 1.8-2.6 0.2-0.5
4 8 8 4 6 6 1
A-CLIPPERTON 10.8 11.0 11.4 2.8 2.5 0.7 0.9
10.4-11.1 10.6-11.5 10.9-11.7 2.4-3.3 2.0-3.3 0.0-1.4 XXX-XXX
8 16 16 8 12 12 8 6
A-PANAMA 11.1 11.3 11.5 2.2 2.1 0.8 1.1 0.3
10.8-11.5 11.0-11.9 11.1-11.9 1.8-2.7 1.8-2.5 0.5-1.4 0.5-1.8 0.2-0.5
Top value is number of pairwise comparisons, bottom values indicate range. See text for details.
24 18 12 0
jClipperton1-EP
J1 Moorea3-CP
J L Moorea1-CP
L
J Moorea2-CP
Panama1-EP
Panama2(2)-EP
L
Panama3-EP
Atoff/n/typeA — Clipperton2-EP
|- Palau1-WP
U Pohnpei1-WP
(P Pohnpei2-WP
Pohnpei3-WP
Alpheus lottini complex L- Palau2-WP
r Guam1-WP
P- Guam2-WP
_[L] Palau1-WP
' Palau2-WP
L
A toff/m type B Guam3(2)-WP
j Hawaii 1(2)-CP
__P Hawaii3-CP
L Hawaii2-CP
A formosus A —— Florida 1-WA
transisthmian —— Panama1(2)-WA
Alpheus formosus complex geminates A. panamensis j—— Panama1-EP
~"—— Panama2-EP
A. formOSUS B
—— Panama1(2)-WA
Figure 10.1 UPGMA tree for Alpheus lottini and the A. formosus sister clade
(unpublished data) complex within the genus. The latter includes a pair of gemi-
nate (transisthmian) species that apparently diverged at the final closure of the
Isthmus approximately three million years ago (Knowlton et al., 1993). EP, eastern
Pacific; CP, central Pacific; WP, western Pacific; WA, western Atlantic.
at third position sites; Table 10.1). The Hawaiian individuals also have a
distinct egg colour that would permit their recognition on that basis.
The potential conflict between the two species concepts becomes clear-
er as we consider more similar populations. Within the type A form, max-
imum corrected sequence divergence between sites is only 2.8% (Table
10.2). This value is substantially less than that exhibited across the Isthmus
and is not much more than apparently intraspecific differences between
Panama and Florida or even within Panama (Figure 10.1). Nevertheless,
within the type A clade, fixed third position genetic differences distin-
guish a Palau/Pohnpei taxon (nine transitions), a Panama taxon (one silent
transversion), and a Clipperton-Moorea taxon (two transitions) (Table
10.1). If confirmed by more extensive sampling, this would result in the
elevation of each of these to specific status using the phylogenetic species
concept of Cracraft (1989,1997: Chapter 16), but the slight sequence diver-
gence would argue against such a distinction using the biological species
concept in the absence of other evidence.
Thus, the biological and phylogenetic species concepts provided rather
different results when applied to variability in allopatric Alpheus, although
both support recognition of several taxa that had been missed in tradi-
tional morphological analyses. Their treatment of the transisthmian taxa is
comparable, for Pacific and Caribbean forms are both reproductively iso-
lated and diagnosably distinct. For the pan-Pacific A. lottini, however, the
biological species concept clearly supports the existence of only two or
three species, while the phylogenetic species concept would recognize
five based on the samples available (Figure 10.2).
10.3 OTHER MARINE INVERTEBRATES

In the discussion above we focused on two genera and ignored the vast
majority of marine invertebrates. While we cannot do justice in a brief
review to the diversity represented by over 30 phyla, we can try to assess
the generality of the patterns we highlighted.
10.3.1 Quantitative characters in sympatry

The absence of qualitative morphological differences between closely
related species is a fact of life for many groups of marine invertebrates,
including those that are clearly reproductively isolated (Langan-Cranford
and Pearse, 1995). Multivariate analyses of quantitative morphological
characters (Budd et al., 1994a; Jackson and Cheetham, 1994) often clearly
define taxa, however, and molecular techniques can confirm their status.
This approach is compatible with the phylogenetic species concept sensu
Cracraft (1989) when fixed genetic differences can be found, and concor-
dance of independent characters provides strong indirect evidence for
reproductive isolation (Avise and Ball, 1990).
However, as argued above, there are no easy practical solutions for
species that are either extremely similar across much of the genome or
for species that sporadically hybridize. How common are such cases likely
to be? Although natural hybridization is often inferred from the ability
to produce viable hybrids in the laboratory and the occurrence of 'inter-
mediates', we generally lack the genetic data to test the inference.
Lessios and Pearse (in press) show for Indo-Pacific sea urchins in the
genus Diadema that genetic data do not support the idea of common
hybridization, despite the absence of conspicuous pre- or post-zygotic
isolating mechanisms and the presence of morphologically intermediate
specimens.
We may be fortunate to be doing systematics now, rather than two mil-
lion years ago, when many new species emerged during the turnover
event that created much of the modern tropical marine invertebrate fauna
(Budd et al., 1994b; Jackson, 1994). Since that time, comparatively little
appears to have happened in terms of speciation, thereby increasing the
probability that lineage sorting will have created qualitative genetic dif-
ferences between species (Avise and Ball, 1990). Very recently diverged
species of marine invertebrates (Palumbi and Metz, 1991; Johannesson et
al., in press) seem to be the exception rather than the rule, although the
Other marine invertebrates 211
Moorea1-CP
— Moorea2-CP
Moorea3-CP
Clipperton1-EP
—— Clipperton2-EP
type A
r—Panama1-EP
' Panama2(2)-EP
Panama3-EP
pPalau1-WP
— Pohnpei3-WP
r-Palau2-WP
A. lottini complex Pohnpei1-WP
Pohnpei2-WP
Palau1-WP
- Palau2-WP
Guam1-WP
Guam2-WP
type B - Guam3(2)-WP
Hawaii1(2)-CP
Hawaii3-CP
L
Hawaii2-CP
A. formosus A Florida 1-WA
trans-
C Panama 1 (2)-WA
isthmian Panama!-EP
A. panamensis
A. formosus complex
Panama2-EP
A. formosus B Panama 1(2)-WA
Figure 10.2 One of 144 equally parsimonious trees. (PAUP analysis as in

Knowlton et al., 1993, for the same taxa as in Figure 10.1.)
difficulty of identifying recently diverged species may contribute to our
perception of their rarity.
10.3.2 Divergence in allopatry

While problems with corals may represent an atypically pessimistic view
of species definitions in sympatry, the snapping shrimp case may give an
overly optimistic view of the potential for understanding and categorizing
allopatric variation. The dispersal of most Alpheus larvae is moderate in
range (with a development time of several weeks), and rafting is probably
not a regular feature of their dispersal biology. While many marine inver-
tebrates share these features, many others do not (Jackson, 1986). Of par-
ticular concern are the numerous sessile or sedentary groups with limited
larval dispersal. These taxa as currently defined often have very large geo-
graphic ranges relative to their dispersal ability, and exhibit extensive and
complex patterns of variation.
Several well-studied, small crustaceans provide examples of problems
that are likely to plague many such groups. In the tide pool dwelling cope-
pod Tigriopus californicus, six of nine study sites analysed in the less than
800 km of coastline between San Francisco and San Diego are diagnosably
distinct based on allozymes and DNA sequences (Burton and Lee, 1994;
R.S. Burton, personal communication). Some of these differences appear
to represent divergence times of 7 million years, but there is no predictable
relationship between genetic divergence and post-zygotic compatibility,
or between different molecular measures of divergence (Burton, 1990;
Burton and Lee, 1994; Ganz and Burton, 1995). Similarly complex patterns
have also been reported for the intertidal, beach-dwelling isopod
Excirolana braziliensis, whose current range, based on traditional morpho-
logical criteria, extends from the Gulf of California to southern Chile in the
eastern Pacific, and from the Gulf of Mexico to southern Brazil in the
Atlantic. Just within Panama, populations from 10 of 11 sites could be dis-
tinguished by fixed allozyme differences (Lessios and Weinberg, 1993),
and nearby populations sharing the same morphology were typically
almost as genetically divergent as populations of different morphologies
isolated by the Isthmus of Panama for at least 3 million years (Lessios and
Weinberg, 1994). There were also marked genetic changes at one site over
a 2-year period.
Organisms like these are going to be a problem for both the biological
and phylogenetic species concept. If there are no predictable relationships
between potential for interbreeding and other measures of divergence,
there are going to be no short-cuts such as the Isthmus for Alpheus in
defining species boundaries based on the biological species concept. For
the phylogenetic species concept, sufficiently detailed genetic information
could result in every sample (including samples over time from the same
Broader implications 213
site) being diagnosably distinct. Even the biological species concept may
result in a staggering increase in species level diversity in such taxa.
10.4 BROADER IMPLICATIONS

Taxonomy does not operate in an intellectual vacuum, unconnected with
other fields of science. Thus, the species concept embraced by students of
marine invertebrates is likely to have substantial consequences both with-
in taxonomy and across other disciplines. Here we consider a few exam-
ples.
10.4.1 Nomenclature
Rigorous application of either the phylogenetic or the biological species
concept will inevitably result in the future recognition of numerous sibling
species within marine invertebrates. Regardless of the species concept that
is ultimately employed, use of terms such as 'species complex' or 'species
group' allows one to recognize informally clusters of similar and appar-
ently related forms. This seems a more useful approach than splitting up
speciose genera (see also Mayr and Ashlock, 1991) or designating numer-
ous subgenera without a phylogenetic analysis.
Application of the phylogenetic species concept as defined by Cracraft
(1989,1997: Chapter 16) could result in orders of magnitude increases in
species level diversity in some cases (Avise and Ball, 1990), effectively
replacing a taxonomic entity with a geographic one. Cracraft (1989) has
scorned the 'how many names can you learn' concern, but names that
cannot be readily used by non-specialists will not be used at all. There is
likely to be little sympathy beyond the ranks of a subset of systematists for
a system that gives species names to every genetically distinct population,
regardless of its potential transience and the triviality of its divergence.
Indeed, Mishler and Theriot (in press), while arguing for their version of
the phylogenetic species concept, suggest that monophyletic groups that
are evolutionary trivial, cryptic or poorly supported not be given formal
recognition. This solution, however, negates the chief virtue of Cracraft's
approach, namely its unambiguous universality.
10.4.2 Ecology
Sympatric sibling species, once recognized, generally exhibit ecological
differences that have important implications for understanding commu-
nity structure (Knowlton, 1993). For example, previous conclusions that
corals were generalists (Connell, 1978) are suspect in light of recent taxo-
nomic discoveries which suggest that niche divergence may be more
important than previously realized (Knowlton and Jackson, 1994).
Diversity recognized by adherents of both the biological and phylogenet-
ic species concepts could result in a substantial change in the taxonomic
database upon which ecologists rely.
10.4.3 Biogeography
Both the biological and the phylogenetic species concepts are likely to
result in greater estimates of endemism, particularly with the use of genet-
ic data. For example, only 30% of Hawaii's marine invertebrates are
thought to be endemic (Kay and Palumbi, 1987), but Alpheus lottini is
apparently a false member of the remaining 70%, despite its ability to dis-
perse via planktonic larvae. Based on the studies of Excirolana and
Tigriopus summarized above, essentially all of Hawaii's brooding marine
invertebrates are likely to be endemic.
10.4.4 Conservation
In the era of the Rio Convention on Biodiversity, species boundaries have
important political as well as biological implications. Fine scaled allopatric
splitting, as promoted by the phylogenetic species concept, would facili-
tate national 'ownership' of biodiversity resources. On the other hand, if
such splitting were unwarranted, it could inhibit the development of
international cooperation in the conservation of interconnected popula-
tions whose survival depends on regional approaches (Committee on
Biological Diversity in Marine Systems, 1995).
10.4.5 Palaeontology
While this book focuses deliberately on living species, many marine inver-
tebrate species living today have extensive fossil records. Species concepts
that work in the present but not in the past are particularly problematic for
these groups. Smith (1994), for example, argued that quantitative statistics
are an inappropriate tool for recognizing species, yet they are an essential
and powerful technique for recognizing living species morphologically for
many of the groups with the best fossil records, such as corals, bryozoans
and molluscs. Felsenstein (1988) argued that the biological difference
between quantitative and qualitative characters was more illusory than
real, and that the focus on qualitative characters in parsimony analysis was
primarily a reflection of computational constraints. Thus, there seems to be
little justification for abandoning these fossil species as unrecognizable.
Rather, genetic data for species in the present should be used to evaluate
quantitative morphometric methods for application in the past (Jackson
and Cheetham, 1994).
Conclusions 215
10.4.6 Phylogeny
Phylogenetic analyses have largely ignored the problem of species bound-
aries. A recent study of cheilostome bryozoans, however, showed that
accurate resolution at the species level greatly improved the quality of
phylogenetic analyses in terms of parsimony, consistency, and concor-
dance between morphological and genetic data (Jackson and Cheetham,
1994). For species that are readily distinguished by qualitative characters,
rigorous discrimination of species by either concept will lead to an
improved understanding of relationships. However, in these
cheilostomes, as with the corals discussed earlier, quantitative statistical
methods were required to resolve the species level taxa, so that many of
the species discriminated in this study would not be recognized by most
supporters of the phylogenetic species concept.
10.5 CONCLUSIONS
The traditional, morphological species concept, as applied to marine
invertebrates, has been unduly conservative. In sympatry, application of
both the biological species concept and the phylogenetic species concept
will result in the recognition of numerous new species that reflect previ-
ously undetected reproductive barriers between morphologically similar
forms. However, insistence on qualitative morphological characters by
some advocates of the phylogenetic species concept will make it impossi-
ble to recognize many species, because morphological differences are
often quantitative.
In allopatry both approaches would again result in an increase in the
number of species, but the differences between the two approaches are
more marked because any distinctive population can become a phyloge-
netic species. There are two problems here. First, the difference between
allopatry and sympatry is less clear-cut in the sea because of the variable
and difficult to measure dispersal of larvae. Over what spatial scale does
one assess reproductive cohesion, an essential step if one is 'to avoid
assigning species status to individual organisms, to different sexes and
morphs, or to developmental stages' (Cracraft, 1989)? Second, sporadic
recruitment events are likely to create trivial and sometimes transient
(Lessios and Weinberg, 1994) but diagnosable groups. Mishler and Theriot
(in press) recommend ignoring these minor monophyletic units, but then
we are left with the question, how minor is minor enough? This is no less
subjective than trying to assess how much differentiation is likely to be
associated with reproductive barriers were the allopatric taxa to come
together, one of the primary objections raised against the biological
species concept. Moreover, the subjectivity of the assessment of
reproductive compatibility is declining with the appearance of theory
(Orr, 1995) and empirical studies (Coyne and Orr, 1989; Knowlton et al,
1993) that relate divergence in allopatry to reproductive isolation
(although some groups, like Tigriopus, may prove difficult to evaluate in
this fashion).
Thus, it would appear that all reasonable species concepts require the
use of scientific judgement. If we are going to have to use judgement,
what principle should guide it? The warring species concepts give us two
alternatives from which to choose. The phylogenetic species concept
argues for systematic consistency: species should be defined in a manner
analogous to all higher taxa. The biological species concept, on the other
hand, suggests that species are special, because they lie at the boundary
that divides the realm of reticulation from the realm of cladogenesis. We
strongly prefer the latter approach, because we find the reticulate/cladis-
tic boundary, even if fuzzy (Mishler and Theriot, in press), to be too bio-
logically important to ignore. Most cladists, however, strongly disagree.
Ironically, it is not at all clear where Willi Hennig would stand were he to
participate in today's debate.
Acknowledgements
The symposium to which Nancy Knowlton was invited provided a stimu-
lating environment in which to think carefully about these issues. We
thank A. Budd, A. Cheetham, C. Cunningham, J. Jackson, and S. Palumbi
for comments on the manuscript; B. Mishler and H. Lessios for sharing
their in press manuscripts with us; and E. Gomez for laboratory assistance.
P. Glynn and M. Gleason provided the collections of Alpheus from
Clipperton and Moorea respectively, while J. Jara and M. Gassel assisted
us in the field. Various government agencies graciously granted permis-
sion to collect in their waters. The Smithsonian Institution and the molec-
ular evolution program of the Smithsonian Tropical Research Institute
provided financial support.
10.6 REFERENCES
Avise, J.C. and Ball, R.M., Jr (1990) Principles of genealogical concordance in
species concepts and biological taxonomy. Oxford Surveys in Evolutionary
Biology, 7, 45-67.
Budd, A.F., Johnson, K.G. and Potts, D.C. (1994a) Recognizing morphospecies in
colonial reef corals: 1. Landmark-based methods. Paleobiology, 20,484-505.
Budd, A.F., Stemann, T.A. and Johnson, K.G. (1994b) Stratigraphic distributions of
genera and species of Neogene to Recent Caribbean reef corals. Journal of
Paleontology, 68, 951-77.
Burton, R.S. (1990) Hybrid breakdown in developmental time in the copepod
Tigriopus californicus. Evolution, 44,1814—22.
References 217
Burton, R.S. and Lee, B.N. (1994) Nuclear and mitochondrial gene genealogies and
allozyme polymorphism across a major phylogeographic break in the copepod
Tigriopus californicus. Proceedings of the National Academy of Sciences, USA, 91,
5197-201.
Claridge, M.F., Dawah, H.A. and Wilson, M.R. (1997) Practical Approaches to
species concepts, in Species: the Units of Biodiversity (eds M.F. Claridge, H.A.
Coates, A.G., Jackson, J.B.C., Collins, L.S. et al. (1992) Closure of the Isthmus of
Panama: the near-shore marine record of Costa Rica and western Panama.
Geological Society of America Bulletin, 104, 814-28.
Committee on Biological Diversity in Marine Systems (1995) Understanding Marine
Biodiversity. National Academy Press, Washington, DC.
Connell, J.H. (1978) Diversity in tropical rain forests and coral reefs. Science, 199,
1302-10.
Coyne, J.A. and Orr, H.A. (1989) Patterns of speciation in Drosophila. Evolution, 43,
362-81.
ation, in Speciation and Its Consequences (eds D. Otte and J.A. Endler), Sinauer
Cracraft, J. (1997) Species concepts in systematics and conservation biology: an
ornithological viewpoint, in Species: the Units of Biodiversity (eds M.F. Claridge,
H.A. Dawah and M.R. Wilson), Chapman & Hall, London, pp. 325-39.
Dustan, P. (1979) Distribution of zooxanthellae and photosynthetic chloroplast
pigments of the reef-building coral Montastrea annularis Ellis and Solander in
relation to depth on a West Indian coral reef. Bulletin of Marine Science, 29,79-95.
Felsenstein, J. (1988) Phylogenies and quantitative characters. Annual Review of
Ecology and Systematics, 19,445-71.
Ganz, H.H. and Burton, R.S. (1995) Genetic differentiation and reproductive
incompatibility among Baja California populations of the copepod Tigriopus cal-
ifornicus. Marine Biology, 123, 821-7.
Gillespie, J. H. (1991) The Causes of Molecular Evolution, Oxford University Press,
New York.
Graus, R.R. (1977) Investigation of coral growth adaptations using computer mod-
elling. Proceedings of the Third International Coral Reef Symposium, 2,463-9.
Graus, R.R. and Macintyre, I.G. (1976) Control of growth form in colonial corals:
computer simulation. Science, 193,895-7.
Jackson, J.B.C. (1986) Modes of dispersal of clonal benthic invertebrates: conse-
quences for species' distributions and genetic structure of local populations.
Bulletin of Marine Science, 39, 588-606.
Jackson, J.B.C. (1994) Constancy and change of life in the sea. Philosophical
Transactions of the Royal Society, London, B, 344, 55-60.
Jackson, J.B.C. and Cheetham, A.H. (1994) Phylogeny reconstruction and the
tempo of speciation in cheilostome Bryozoa. Paleobiology, 20,407-23.
Johannesson, K., Rolan-Alvarez, E. and Ekendahl, A. Incipient reproductive isola-
tion between two sympatric morphs of the intertidal snail Littorina saxatilis.
Evolution (in press).
Kay, E.A. and Palumbi, S.R. (1987) Endemism and evolution in Hawaiian marine
invertebrates. Trends in Ecology and Evolution, 2,183-7.
Kessing, B., Croom, H., Martin, A. et al. (1989) The simple fool's guide to PCR, ver-
sion 1.0. Unpublished.
Kessing, B. (1995) Sequencer 3.0: DNA sequence manipulation and analysis.
Program written for the Macintosh.
Kimura, M. (1980) A simple method for estimating evolutionary rate of base sub-
stitutions through comparative studies of nucleotide sequences. Journal of
Molecular Evolution, 16,111-20.
Knowlton, N. and Jackson, J.B.C. (1994) New taxonomy and niche partitioning on
coral reefs: jack of all trades or master of some? Trends in Ecology and Evolution,
9, 7-9.
Knowlton, N. and Mills, D.K. (1992) The systematic importance of color and color
pattern: evidence for complexes of sibling species of snapping shrimp (Caridea:
Alpheidae: Alpheus) from the Caribbean and Pacific coasts of Panama.
Proceedings of the San Diego Society of Natural History, 18,1-5.
Knowlton, N., Weil, E., Weigt, L.A., and Guzman, H.M. (1992) Sibling species in
Montastraea annularis, coral bleaching, and the coral climate record. Science, 255,
330-3.
Knowlton, N., Weigt, L.A., Solorzano, L.A., Mills, D.K., and Bermingham, E. (1993)
Divergence in proteins, mitochondrial DNA, and reproductive compatibility
across the Isthmus of Panama. Science, 260,1629-32.
Knowlton, N., Mate, J. L., Guzman, H.M., Rowan, R. and Jara, J. Direct evidence
for reproductive isolation among the three species of the Montastraea annularis
complex in Central America (Panama, Honduras). Marine Biology (in press).
Lang, J.C. (1984) Whatever works: the variable importance of skeletal and non-skele-
tal characters in scleractinian taxonomy. Palaeontographica Americana, 54,18-44.
Langan-Cranford, K.M. and Pearse, J.S. (1995) Breeding experiments confirm
species status of two morphologically similar gastropods (Lacuna spp.) in cen-
tral California. Journal of Experimental Marine Biology and Ecology, 186,17-31.
Lessios, H.A. and Pearse, J.S. Hybridization and introgression between Indo-
Pacific species of the sea urchin Diadema. Marine Biology (in press).
Lessios, H.A. and Weinberg, J.R. (1993) Migration, gene flow and reproductive iso-
lation between and within morphotypes of the isopod Excirolana in two oceans.
Heredity, 71, 561-73.
Lessios, H.A. and Weinberg, J.R. (1994). Genetic and morphological divergence
among morphotypes of the isopod Excirolana on the two sides of the Isthmus of
Panama. Evolution, 48, 530-48.
Mayr, E. and Ashlock, P.D. (1991). Principles of Systematic Zoology, 2nd edn,
McGraw-Hill, New York.
Mishler, B.D. and Theriot, E. (1997) Monophyly, apomorphy, and phylogenetic
species concepts, in Species Concepts and Phylogenetic Theory: A Debate (eds Q. D.
Wheeler and R. Meier), Columbia University Press, New York (in press).
Orr, H.A. (1995) The population genetics of speciation: the evolution of hybrid
incompatibilities. Genetics, 139,1805-13.
Palumbi, S.R. and Metz, B.C. (1991) Strong reproductive isolation between closely
related tropical sea urchins (genus Echinometra). Molecular Biology and Evolution,
8,227-39.
References 219
Romano, S.L. (1995) A Molecular Phylogenetic Analysis of Reef-Building Corals, PhD
Thesis, University of Hawaii.
Smith, A.B. (1994) Systematics and the Fossil Record: Documenting Evolutionary
Patterns, Blackwell Scientific, Oxford.
Templeton, A.R. (1981) Mechanisms of speciation - a population genetic approach.
Annual Review of Ecology and Systematics, 12,23^48.
Templeton, A.R. (1989) The meaning of species and speciation: a genetic perspec-
tive, in Speciation and Its Consequences (eds D. Otte and J. A. Endler), Sinauer
Van Veghel, M.L.J. and Bak, R.P.M. (1993) Intraspecific variation of a dominant
Caribbean reef building coral, Montastrea annularis: genetic, behavioral and
morphometric aspects. Marine Ecology Progress Series, 92, 255-65.
Van Veghel, M.L.J. and Kahmann, M.E.H. (1994) Reproductive characteristics of
the polymorphic Caribbean reef building coral Montastrea annularis: II.
Fecundity and colony structure. Marine Ecology Progress Series, 109, 221-7.
Veron, J.E.N. (1995) Corals in Space and Time: The Biogeography and Evolution of the
Scleractinia, UNSW Press, Sydney.
Wallace, C.C. and Willis, B.L. (1994) Systematics of the coral genus Acropora: impli-
cations of new biological findings for species concepts. Annual Review of Ecology
and Systematics, 25, 237-62.
Weil, E. and Knowlton, N. (1994) A multi-character analysis of the Caribbean coral
Montastraea annularis (Ellis and Solander, 1786) and its two sibling species, M.
faveolata (Ellis and Solander, 1786) and M. franksi (Gregory, 1895). Bulletin of
Marine Science, 55,151-75.
Willis, B.L. (1985) Phenotypic plasticity versus phenotypic stability in the reef
corals Turbinaria mesenterina and Pavona cactus. Proceedings of the Fifth
International Coral Reef Congress, Tahiti, 4,107-12.
Wood Jones, F. (1907) On the growth-forms and supposed species in corals.
Proceedings of the Zoological Society of London, 1907, 518-56.
11
Nematode species: concepts and
identification strategies
exemplified by the Longidoridae,
Steinernematidae and
Heterorhabditidae
D. J. Hunt
Contacting address: International Institute of Parasitology, 395A, Hatfield Road, St Albans
AL4 OXU, UK
ABSTRACT
The differing approaches to the practical identification of phytopara-
sitic nematodes in the family Longidoridae and entomopathogenic
nematodes in the families Steinernematidae and Heterorhabditidae
are compared and contrasted. Both groups contain economically
important species: the former as parasites of the root systems of plants
and, in certain species, with the potential for transmission of plant
nepoviruses; the latter as possible biocontrol agents of insect pests in
horticulture and agriculture.
The taxonomy of the Longidoridae, a group displaying an abun-
dance of relatively well-defined and stable characters, is almost exclu-
sively concerned with the classical approaches of morphology and
morphometrics. However, in one species complex within the genus
Xiphinema, namely the X. americanum-group, molecular approaches
are being applied with increasing frequency. This new approach has
largely been precipitated by the exceptionally close morphological
and morphometric similarity of the members of this substantial group
of species and the practical difficulties involved in specific determina-
tion; this being of particular relevance bearing in mind the importance
of certain members in vectoring economically important plant viruses.
In the Steinernematidae and the Heterorhabditidae, all the mem-
bers are obligate parasites of the insect haemocoel and exhibit few reli-
able morphological or morphometric characters. This has led to con-
fusion as to the value and status of the differences which have been
observed, together with concomitant problems in establishing the spe-
cific identity of isolates. In these families the major progress to date
has been via molecular techniques involving PCR products, a tech-
nique which can rapidly yield reliable results from a single infective
juvenile.
A synthesis of classical and molecular methodologies is advocated
as the most productive approach in delimiting and recognizing the
species taxon in these families.
11.1 INTRODUCTION
The Phylum Nematoda comprises a diverse assemblage with tens of thou-
sands of nominal species. Estimates of the total number of species are
speculative, but range into the hundreds of thousands. Representatives
may be parasitic in vertebrates and/or invertebrates, parasitic on plants
and/or fungi, predators, or free-living microbivores in soil or water. They
are found throughout the world's seas and continents, often in colossal
numbers, and may be the dominant biomass in selected habitats. They are
even found colonizing such extreme niches as the abyssal ocean depths
and the lacunary system on the undersurface of Arctic sea ice (Tchesunov
and Rieman, 1995). Despite this ecological diversity they are usually rather
similar in their gross morphology, being essentially cuticle-bounded tubes
supported by a hydrostatic skeleton. The internal organs essentially con-
form to a tubular pattern and lie within a fluid-filled space. Those nema-
todes parasitizing plants or insects tend to be small, typically less than 1
mm in length, although exceptionally over 50 cm in some insect parasites
from the Mermithida. They are usually transparent so that the internal
organs are readily visible with the light microscope, even at low magnifi-
cations. This phenomenon has led to a conflict between features which,
although useful as key characters, are not necessarily of equivalent value
in systematics. Similarly, some characters which are of great systematic
importance are visually obscure and of little practical use in routine iden-
tification procedures.
What is a species? This fundamental question continues to spawn a
confusing plethora of erudite concepts (biological, evolutionary, morpho-
logical, phylogenetic, etc.) which are often championed by vehement
protagonists, but a unified theory is lacking, perhaps unattainable. The
conflict between theoretical and operational constraints implicit in any
concept inevitably results in compromises. Cronquist (1977) expressed
the morphological species concept thus: 'Species are the smallest groups
that are consistently and persistently distinct, and distinguishable by
ordinary means'. Mayr, an ardent exponent of the major alternative
viewpoint, has long advocated the use of a biological species concept
Introduction 223
with its implication of cryptic species (Mayr, 1969). Both concepts are
used in plant and insect nematology, the emphasis being on the phenet-
ic approach as nematodes are inadequately known at the biological level
to facilitate widespread use of the latter. Philosophical conflict between
proponents of the morphological or biological species concept is frequent,
but sterile when indulged to excess. The theories are surely complemen-
tary in overall purpose and deserve due recognition as viable perspec-
tives with differing strengths and weaknesses. A more pragmatic, less
adversarial approach is required, one which recognizes the unpalatable
fact that species are not completely definable and will not conform neat-
ly to the tenets of any one concept. Our idea of what constitutes a species
demands continual reappraisal; it must evolve with increasing knowl-
edge, but carries the proviso that it is unlikely to be consistently applica-
ble over the whole gamut of Life.
A common artifice is to regard the opinion of a competent taxonomist
as the determining factor as to whether an organism qualifies for specific
status or not. Despite provoking additional comment on the definition of
taxonomic competence, this circular argument has considerable function-
al appeal, particularly in view of the lack of consensus on a more objective
alternative. As the species is the fundamental taxon within systematics,
any imprecision in definition inevitably invites problems by allowing sub-
jective interpretation of the degree of plasticity acceptable before the per-
mitted variation of one species crosses some imaginary boundary and
becomes another entity, distinct, at least, in name. One could argue, with
little difficulty, that a species, as such, does not really exist as a discrete,
definable object. It is merely a convenient pigeon hole with which we seek
to define and categorize the variability of an expression of life (the organ-
ism) in question. A species may exhibit a core or suite of polythetic char-
acter states, but the margins of phenotypic variability are usually diffuse
and ill-defined and more indicative of our state of knowledge than any-
thing more germane. The specific attribution of an organism displaying a
phenotype at the periphery of variation is therefore probabilistic, i.e. taxa
are fuzzy, not discrete entities.
The purposes and/or uses of species categorization may be manifold,
but the desire to label and catalogue is surely a prime mover (labelling is
often more comforting than informative - it is always easier to label than
to define!). The need for a species concept is, however, inescapable and
cataloguing can be a reasonable compromise, particularly if the 'entities'
are then arranged into patterns which shed light on their inter-relation-
ships. We should be careful not to assume too readily that these named
life forms are imbued with validity or relevance outside our own philoso-
phy - the species concept is not as plastic as the organisms it purports to
define. As we often have little idea of the potential variability within one
of our 'species' it follows that the species concept itself, although possessing
a set of core criteria, must of necessity be diffuse and imprecise if it is to be
anything other than an exercise in semantics. This imprecision is itself not
constant and may be influenced by prevailing fashion, type of organism at
hand and/or environmental parameters.
The International Code of Zoological Nomenclature conveniently defines a
species by a nomen attached, in the ultimate resort, to a single specimen, the
holotype. Such a nomenspecies concept, although necessary for nomen-
clatural stability, offers little solace to the person faced with the phenotypic
(and genotypic) plasticity of populations. The interpretation of what consti-
tutes a species sensu lato is intuitive. In this respect the term 'unit of diversi-
ty' expresses rather more of a sense of precision, of being in control, than is
warranted. Despite semantic and philosophical quibbling, there is an irre-
ducible requirement to identify and name organisms as a prerequisite to
many areas of biological research. A species concept, if interpreted sensibly,
provides an acceptable means of categorizing and packaging variability in
order to achieve this purpose - even if our convenient units do prove a tri-
fle leaky and bleed characters across boundaries which are often a reflection
of ignorance, rather than anything more profound.
Which species concept should be applied? The choice may be eclectic
and involve factors such as the predilection of the user; the group being
studied and overall purpose. For example, a phylogenetic species concept
may be robust in highly visible, well-studied groups, such as birds
(Cracraft, 1997: Chapter 16), but suffers overwhelming practical disad-
vantages in cryptic, poorly known groups such as nematodes where the
necessary criteria are more elusive. Although the ultimate (and laudable)
purpose of systematists may be to produce an entirely natural scheme
reflecting evolutionary lineages, this is not the only raison d'etre for clas-
sification. Cain (1959) wrote: 'Taxonomists have forgotten for too long
that they are the name makers for all zoologists and botanists, pure and
applied, and are under an obligation never to impose an unnecessary
burden on others'. With the current widespread interest in biodiversity,
the user list could be expanded to include politicians and the public at
large. The role of the taxonomist in facilitating a practical, reliable and
consistent approach to problems inherent in the identification process is
crucial. Biosystematists no longer work in an esoteric backwater.
Consequently, they need to assume broader responsibilities than previ-
ously and perhaps exercise more caution and restraint in the pursuit of
their art - internecine squabbles over semantic abstractions risk misinter-
pretation and ultimately detract from the wider cause of biosystematics.
The problem of correctly naming a plant or insect parasitic nematode
species has practical considerations in that a proportion of species are eco-
nomically important, perhaps causing severe damage to agricultural crops
either directly, as a result of their feeding activities, indirectly by facilitat-
ing attack by pathogenic fungi or, as in the case of the longidorids and
Phytoparasitic nematodes of the family Longidoridae 225
trichodorids, by vectoring plant viruses. Species with biocontrol potential,
such as entomopathogenic nematodes, may have a restricted host range
or require specific environmental parameters to facilitate invasion and
control of the target organism. Correct specific attribution in such
instances carries considerable practical implications, not to mention
responsibilities.
In modern nematology, the traditional phenetic species concept com-
bining morphology and morphometrics still holds sway, regardless of
discipline. Apomixis, a phenomenon of common occurrence in the
Nematoda, poses awkward questions for both morphological and bio-
logical species concepts and such problems have yet to be satisfactorily
resolved. Biochemical and molecular approaches, although late starters,
are now well established in many aspects of nematology and show par-
ticular potential where classical techniques lack the necessary precision
to discriminate closely related taxa or host races. The seminal value of
molecular techniques is, in the absence of fossil nematodes of any geo-
logically significant age, likely to prove particularly potent in elucidating
phylogenetic relationships.
Two contrasting paradigms of the methodology and exigent problems
concerning the concept and characterization of the species taxon within
the Nematoda, exemplified here by the phytoparasites in the Longidoridae
and the entomopathogenic nematodes in the Steinernematidae and
Heterorhabditidae, are now discussed in greater depth.
11.2 PHYTOPARASITIC NEMATODES OF THE FAMILY

LONGIDORIDAE (WITH PARTICULAR EMPHASIS ON THE
GENUS XIPHINEMA COBB, 1913)
11.2.1 Introduction
The Longidoridae, a family of phytoparasitic nematodes in the Order
Dorylaimida, comprises a handful of genera, representatives of which are
found more or less worldwide. Longidorids are large nematodes, ranging
in length from about 1.5 mm to over 12 mm. They are all characterized by
the form of the oesophagus and by the possession of an attenuated,
needle-like odontostylet with which plant roots are pierced in order to feed
on the cell contents. The degree to which the odontostylet is developed,
the form of the junction between the distal portion (the odontostyle) and
the proximal section (the odontophore) and the location and relative size
of the nuclei of the oesophageal glands together allow the family to be
subdivided into subfamilial groups. Furthermore, these characters, in
combination with others such as the form of the amphid, a sense organ
opening on the cephalic region, also serve to define the genera. A brief
systematic scheme is as follows:
Order Dorylaimida
Superfamily Dorylaimoidea
Family Longidoridae
Subfamily Longidorinae
Genus Longidorus Micoletzky, 1922 (Filipjev, 1934)
Genus Longidoroides Khan, Chawla & Saha, 1978
Genus Paralongidorus Siddiqi, Hooper & Khan, 1963
Subfamily Xiphidorinae
Genus Xiphidorus Monteiro, 1976
Genus Paraxiphidorus Coomans & Chaves, 1995
Subfamily Xiphinematidae
Genus Xiphinema Cobb, 1913
The Longidoridae currently contains about 400 valid species, over half
belonging to Xiphinema with Longidorus (about 100 species), Paralongidorus
(about 50 species), Longidoroides (about 15 species) and Paraxiphidorus and
Xiphidorus (nine species) making up the total. The two genera of most eco-
nomic importance are Longidorus and Xiphinema, both genera being
known to contain species capable of vectoring plant viruses in addition to
causing direct root damage by their feeding activities.
The morphological and morphometric characters used to define species
within the Longidoridae are similar across all genera, although Xiphinema
exhibits appreciably more intrinsic variation, particularly with regard to
the anatomy of the female genital apparatus and the form of the tail.
Indeed, attempts, albeit unsuccessful, have been made to use such char-
acter set variation as justification for subgenera (Cohn and Sher, 1972). To
date, the only genus in the Longidoridae where species have been stud-
ied using biochemical or molecular techniques is Xiphinema. This is main-
ly a result of the economic and quarantine importance of an intractable
species aggregation containing virus vectors - the so-called americanum-
group - where accurate specific determination is paramount.
The genus Xiphinema, partly because of the comparatively large size
(1.5-8 mm) of its members, attracts considerable interest from nematolo-
gists wishing to describe a new species. While this carries a positive aspect,
it inevitably opens the door to less than adequate species descriptions
which a more rigorous approach would ascribe to existing taxa.
Geographically widespread taxa, such as X. brasiliense Lordello, 1951, X.
elongatum Schuurmans Stekhoven and Teunissen, 1938, X. ensiculiferum
(Cobb, 1893) and X. radicicola T. Goodey, 1936 are particularly prone to this
malaise, a fact reflected in their depressing list of synonyms. Xiphinema,
species-rich by phytoparasitic nematode standards, currently contains
over 200 valid species (Hunt, 1993), all members being ectoparasites of
plant roots and feeding deep within the tissues by means of a protrusible
odontostylet which may be over 350 jxm long in some species. In addition
to the plant damage (cell death, galling, disruption of root function)
caused by their direct feeding activities, some species achieve consider-
ably greater importance by virtue of their ability to transmit plant
nepoviruses such as tobacco ringspot virus, tomato ringspot virus and
grapevine fanleaf virus.
Species identity within Xiphinema has traditionally relied upon the
assessment of phenotypic characters. With the ever-increasing number of
nominal species and the close similarity of many of the monosexual
species, the situation is becoming more difficult and a group which was
once reasonably accessible to the non-expert is now posing considerable
problems even to the cognoscenti.
Problems in Xiphinema taxonomy are exacerbated by the fact that, as
with many soil nematodes, a substantial proportion of the nominal taxa
are monosexual. Most of these species appear to reproduce by obligate
meiotic parthenogenesis. Apparently functional males do occur rarely in
some species, thus implying the existence of a facultative parthenogenet-
ic condition which may be more widespread than the data reflect.
Monosexual taxa display a frustrating tendency to radiate into clusters of
closely similar forms or morphospecies. The more populations of a form
that are studied the better are the chances that the spectrum of variation
can be adequately assessed, thus leading to a more representative blend
of character states and a stable nomenclature. The ability to delimit such
forms in a meaningful way such that other workers can recognize and
identify them is therefore crucial. The manifest failure of morphological
characters to fulfil this function in perplexingly speciose complexes such
as the X. americanum-group was pointed out by Heyns (1983). The inter-
vening years have seen this situation become more refractory and an
alternative methodology is urgently required to assist in systematically
untying, rather than cutting, the Gordian knot. This alternative approach
may now be on the threshold of realization with the advent of readily
accessible and reliable PCR (polymerase chain reaction) based tech-
niques. The characters used to define species of Xiphinema will now be
examined in greater detail.
11.2.2 Characterization
(a) Classical morphology and morphometrics

A varied suite of reliable phenetic characters for species determination in
Xiphinema has been developed over the past 35 years (Luc and Dalmasso,
1975a,b; Loof and Luc, 1990). These mainly concentrate on the female sex
because many species are not amphimictic, males being either entirely
absent or but rarely recorded. The major morphometric and morphologi-
cal characters currently used to define species of Xiphinema are as follows:
Morphometric characters
• Body length (L)
• Odontostylet length
• Tail length
• Vulva position (V - vulval position from the anterior end of the body
expressed as a percentage)
• Ratio a (body length divided by maximum body width)
• Ratio c (body length divided by tail length)
• Ratio c' (tail length divided by anal body width)
Morphological characters
• Shape of cephalic extremity (continuous, offset or expanded from body
contour)
• Habitus of heat-relaxed nematode
• Number and development of the female genital tracts
• Presence or absence of Z-organ
• Presence or absence (and type) of pseudo Z-organ
• Tail shape
• Presence or absence of blind terminal canal on tail
• Shape of juvenile tail
• Monosexual or amphimictic
When applied cautiously, these are all good characters. Problems arise
when the material is distorted or shrunken by bad fixation or when the
coverslip of the slide mount is not supported properly. This causes the
nematode to become squashed and therefore alters the appearance of
both head contour and tail shape, not to mention changes in the value of
the various morphometric ratios. Even in properly supported mounts,
specimens can still flatten over a relatively brief time-span (Heyns, 1983)
and thus elicit misleading responses when re-examined.
The presence or absence in the female genital tracts of a Z-organ or
pseudo Z-organ is an important and reliable character (Luc and Dalmasso,
1975a,b), but if an old description does not mention such a structure (par-
ticularly the pseudo Z-organ) as being present, the assumption, a priori,
that it is absent, can be only too false. The original author may have over-
looked the feature entirely, may have seen it, yet not considered the fact
important, or misinterpreted it as something inconsequential. Re-
examination of type material for this type of structure is absolutely essen-
tial if mistakes are to be avoided and 'new species' proposed unnecessar-
ily merely on presence versus purported absence of such a character.
Particular problems are caused by the X. americanum-growp. This bloated
complex contains over 40 nominal species, many of which are taxonomi-
cally adjacent in their morphology with credence given to differences
which would not be entertained or countenanced in other members of
the genus. Resolving this chaotic situation is made more urgent by the
economic importance of the group as certain species can vector damaging
plant viruses. The problem of how to identify taxa with virus vector
potential from non-virus vectors is pressing and is particularly pertinent
in quarantine procedures where nematodes from this group are regularly
intercepted in the rhizosphere of imported plants. The taxonomy of the
group is made more intransigent by the predominance of an apparently
obligate parthenogenetic reproductive strategy. This phenomenon
removes male morphology from the characterization/identification
process and also results in genetically isolated populations which effec-
tively operate outside the usual biological species concept as clones or
morphospecies. Heyns (1983) commented on these problems and stressed
the need for an alternative approach to morphological characterization
within the group to satisfactorily distinguish the nominal species. Since
his paper, the problems have been exacerbated, not diminished, by many
more taxa being proposed.
Problems of species delimitation within the X. americanum-group have
long been recognized (Tarjan, 1969). It was, however, the paper by
Lamberti and Bleve-Zacheo (1979) which precipitated many of the prob-
lems, when they proposed and differentiated 15 new species in the com-
plex by utilizing a combination of minute differences in head and tail
shape and small differences in morphometric characters such as odon-
tostyle length. Such parameters may well be discriminatory, but unless
other taxonomists can see and are prepared to accept such nuances as
being reliable and meaningful criteria at the species level, the proposed
species will be disputed or risk being dismissed in a perfunctory manner
out of sheer exasperation. Indeed, the situation has become so abstruse
that the authors of otherwise comprehensive multiple-entry or polyto-
mous keys to the genus (Loof and Luc, 1990,1993) declined to tackle the
X. americanum-group and Loof et at (1993) disputed the validity of the sim-
plistic dichotomous key approach to the complex proposed by Lamberti
and Carone (1991).
(b) Scanning electron microscopy (SEM)

Although SEM techniques have proved valuable in some groups of phy-
toparasitic nematodes such as the Heteroderidae, Meloidogynidae and
Pratylenchidae, the longidorids have not been extensively studied. There
are several good reasons for this. Longidorids, being members of the Order
Dorylaimida, do not have the outer body cuticle annulated or otherwise
adorned as is the norm in many phytoparasitic genera from the Order
Tylenchida. Furthermore, they lack features such as longitudinal cuticular
incisures in the lateral field and their cephalic structure is generally less
complex, although the recent discovery (Swart and Heyns, 1987) of small,
flap-like cephalic lobes in certain species of the genus Paralongidorus is a
notable exception which may yet validate the related genus Siddiqia, cur-
rently regarded as a synonym, or at best a subgenus, of Paralongidorus
(Hunt, 1993). SEM has, however, been used to elucidate the form of
dissected-out internal organs such as the cuticular odontostylet with which
the nematode pierces the plant root. In conclusion, it must be said that
SEM has little immediate value in the determination of longidorids to
species level, although its merit in other groups of nematodes sporting
more complex cuticular and/or cephalic structures is indisputable.
(c) Biochemical and molecular techniques

Although biochemical and molecular techniques have been successfully
applied to other phytoparasitic nematodes, comparatively little work has
been done on longidorids. The reason for the lack of biochemical method-
ologies is probably due to the difficulty in obtaining sufficient nematode
material as longidorid species, in marked contrast to members of the
Heteroderidae, Meloidogynidae or Pratylenchidae, are difficult to culture
and usually very slow to reproduce, life-cycles of several years being report-
ed for some of the taxa where detailed biological studies have been made.
With the availability of relatively cheap and reliable PCR techniques, it
is now possible to clone enough DNA from a single specimen to facilitate
molecular studies. Vrain and Wakarchuk (1989), using cloned rDNA
probes and Vrain et al (1992), using rDNA restriction fragment length
polymorphism (RFLP) techniques, studied several species in the X. ameri-
canum-group from North America. Vrain et al. (1992) designed two primers
to amplify the ITS region using PCR and compared the 1.5 kb amplified
product from each of 16 populations for RFLPs. Fourteen restriction
enzymes were used to digest the PCR product. The DNA products were
separated by electrophoresis, the data being used to calculate dissimilari-
ty coefficients and to construct dendrograms by cluster analysis. This
enabled Vrain et al. (1992) to separate the original 16 populations into five
clusters and show that two populations of X. rivesi, one of the better-
defined morphological species, were well separated from other X. ameri-
canum-group populations. The results also implied the affinity of uniden-
tified populations with existing species. Vrain et al. (1992) concluded that
the molecular approach could be used to separate species in the X. ameri-
canum-group and thereby assist classical taxonomic techniques in gauging
the usefulness of morphological and morphometric characters for species
determination.
De Giorgi et al. (1994) attempted to use amplification of a segment of
mtDNA from X. index as an identification aid. They employed a subunit of
the cytochrome oxidase gene and concluded that mtDNA offered possi-
bilities as a species specific identification tool, but cautioned that much
more data on the gene organization and nucleotide sequences in plant
nematodes were required to assess intraspecific variation in the mito-
chondrial genome.
11.2.3 Discussion
Although plant nematology has been traditionalist in its approach, there is
an increasing awareness that alternative methodologies are necessary,
particularly in more problematic groups where the definition of specific or
subspecific taxa demands increased precision and objectivity. This is partic-
ularly so for those species reproducing via meiotic or mitotic parthenogen-
esis. A variety of techniques have been tried of which the most promising to
date have been biochemical or molecular. Burrows (1990) reviewed some of
these approaches, but the intervening years have witnessed rapid progress
[see the compilation edited by Lamberti et al. (1994) for a recent appraisal].
Most biochemical and molecular studies have concentrated on com-
plex, economically important groups, such as the Heteroderidae and
Meloidogynidae, although other damaging genera, such as Radopholus,
have also been studied (Hahn et al, 1994; Kaplan, 1994) while species in
the notoriously difficult genus Aphelenchoides have been proposed partly
on the results of esterase and PCR techniques (Hooper and Ibrahim, 1994;
Ibrahim and Hooper, 1994; Ibrahim et al, 1994a,b). Isozyme electrophore-
sis has been used successfully in identifying Meloidogyne spp. on a routine
basis (Esbenshade and Triantaphyllou, 1990; Cenis et al, 1992), yet suffers
from the disadvantage that the technique is restrictive in the nematode
material required and does not operate satisfactorily for single infective
juveniles, a problem which PCR is able to overcome. Powers and Harris
(1993) showed that PCR could differentiate juveniles from a number of
root knot (Meloidogyne) species by employing differences in mtDNA and
Williamson et al. (1994), using this technique, claimed a high success rate
in obtaining an amplified band for some 80% of single juveniles studied.
The randomly amplified polymorphic DNA (RAPD) technique allows a
broad spectrum of markers to be developed from single primers and can
be a useful tool, if currently less consistent, than PCR products from spe-
cific regions of the genome DNA. Williamson et al (1994) commented on
the potential of RAPDs in assisting research into nematode genetics and
systematics and presented results of studies on Heterodera spp. and
Meloidogyne spp. They concluded that RAPDs could be a useful tool in dis-
tinguishing species and populations, but felt that the technique was not
yet robust enough for a role in general identification practices. Fargette et
al (1994), using both RFLP and RAPD patterns, distinguished various
species of Meloidogyne, although one species in particular, M. arenaria, was
polymorphic. The dendrograms produced by these two techniques were
reasonably congruent and allowed grouping of resistance breaking lines
of the nematodes.
Heterologous cloned DNA probes have been used to distinguish
between races of Meloidogyne incognita, an important phytoparasitic
nematode throughout the tropics and subtropics (Chacon et al, 1995).
Such accuracy is fundamental with nematode species displaying host
races as non-chemical control strategies often employ non-host or toler-
ant varieties to minimize crop damage. Here, the molecular approach can
facilitate host race determination more conveniently, for example, than
differential host tests (Sasser and Carter, 1985).
Although classical taxonomic techniques have served well and justifiably
dominate within the Longidoridae, a group enjoying a rich and varied suite
of morphological characters, there is little doubt that the application of mol-
ecular techniques will prove to be of immense value. The resulting synthe-
sis, if sensibly implemented, should help to resolve problems posed by
groups of closely related, parthenogenetic species as exemplified by the X.
americanum-group, where the molecular approach (Vrain and Wakarchuk,
1989; Vrain et al, 1992) may dismiss or validate emergent species based on
minute phenotypic differences (see, for example, Cho and Robbins, 1991;
Lamberti and Carone, 1991). From the limited information available, there
now seems real hope that the X. americanum-growp will yield to molecular
techniques and that the validity or otherwise of the nominal species will be
resolved, thus assisting not only nematode taxonomists, but also quarantine
authorities and virologists concerned with the nematode transmission of
plant nepoviruses. The use of molecular techniques, not only on other
species of Xiphinema (De Giorgi et al., 1994), but also on other genera within
the family, may also illuminate phylogenetic relationships within the group
and thus confirm or refute the current, somewhat speculative, theories con-
cerning the higher systematics.
11.3 ENTOMOPATHOGENIC NEMATODES OF THE FAMILIES

STEINERNEMATIDAE AND HETERORHABDITIDAE
11.3.1 Introduction
Entomopathogenic nematodes are characterized by their ability to carry
specific pathogenic bacteria which are released into the insect haemocoel
after penetration of the insect host has been achieved by the infective
stage of the nematode. Such nematodes have been known since the early
part of the century (Steiner, 1923) and one species, Steinernema glaseri,
was used as a biocontrol agent of a scarabaeid grub as early as the 1930s
(Glaser and Farrell, 1935). There are two families of major importance, the
Steinernematidae and the Heterorhabditidae. Both fall within the Order
Rhabditida. Both carry symbiotic entomopathogenic bacteria;
Xenorhabdus Thomas & Poinar, 1979 in Steinernema, and Photorhabdus
Boemare, Akhurst & Mourant, 1993 in Heterorhabditis, and both share a
Entomopathogenic nematodes 233
broadly similar life-cycle. Despite these shared attributes, they are other-
wise remarkably distinct in the morphology of the male tail and copula-
tory apparatus and possibly have a diphyletic origin, their similarities
arising from convergent evolution (Poinar, 1990, 1993). The
Steinernematidae comprises two genera: Steinernema (syn. Neoaplectana),
the type genus, and Neosteinernema. The Heterorhabditidae is monotypic,
represented by the genus Heterorhabditis (syn. Chromonema). The system-
atic problems encountered in this group when applying the phenetic
approach arise because the adult nematodes feed and reproduce in the
protected environment of the insect haemocoel. Such specialized, but
essentially similar modus operandi, implies a considerable degree of simi-
larity in the morphological features expressed in combination with exces-
sive morphometric variability attributable to density-dependent nutri-
tional factors. Classical techniques have therefore concentrated on the
free-living infective stage which, although lacking considerable gross
morphological variation as a result of being a non-feeding stage, does
show enhanced morphometric consistency. These problems have result-
ed in considerable confusion as to the status of the nominal species, a
confusion which has enormous practical importance now that the nema-
todes have attracted commercial interest as potential biocontrol agents. In
addition, accurate identification is often demanded by quarantine regu-
lations stipulating that only indigenous species/isolates can be released as
part of a biocontrol programme. Partly in response to this stricture, more
extensive surveys for the infective, soil-dwelling stage have been carried
out both in temperate and tropical regions with the result that a large
number of isolates, many of which appear to be new species, have been
found and cultured. This explosion of data has further stressed the
already ill-defined and unstable systematics at a juncture when precision
in attributing identity is most needed to facilitate the exploitation of such
potentially useful biocontrol agents.
11.3.2 Bionomics
The typical life-cycle of a steinernematid is as follows. The infective third
stage juvenile (IJ3) occurs in the soil and, although non-feeding, can sur-
vive for a considerable period due to extensive food reserves. The IJ3 car-
ries a specific bacterium of the genus Xenorhabdus in the intestine and
either penetrates an insect directly via the cuticle, spiracles or anus, or is
ingested. Whichever mode of entry is effected, the IJ3 then penetrates to
the haemocoel where it moults and releases the bacteria which rapidly
produce a fatal septicaemia within 24 to 48 hours. The nematodes feed on
the bacterial soup produced by the breakdown of the host tissues and
moult to the first generation adults. These mate and subsequently the sec-
ond generation adults occur. These are smaller than the first generation
and may show morphological differences. Typically, the progeny from the
second generation cease development at the third stage juvenile as by this
time virtually all the food reserves have been exhausted and the cadaver
is reduced to a nematode-filled skin. Eventually the nematodes escape
from the remains of the host and migrate into the soil as the IJ3.
Heterorhabditids show a basically similar life-cycle except for the crucial
difference that the first generation is hermaphroditic and not amphimic-
tic. The IJ3 also carries a different genus of bacterium, namely
Photorhabdus, so-called because of its ability to bioluminesce.
11.3.3 Systematics
The families Steinernematidae and Heterorhabditidae are regarded as
being closely related by some authorities, but rather more distant by
others (Poinar, 1990,1993). Certainly the male tail region is very different
- heterorhabditids have a bursa supported by caudal rays whereas the
steinernematids lack a bursa completely, have differently shaped copula-
tory spicules and possess numerous copulatory papillae. Such features are
considered fundamental in the Rhabditida. Both, however, share a similar
life-cycle and depend on the same symbiotic relationship with related
genera of insect pathogenic bacteria. Regardless of the higher systematics,
the genera are grouped as follows:
Family Steinernematidae
Genus Steinernema Travassos, 1927
Genus Neosteinernema Nguyen & Smart, 1994
Family Heterorhabditidae
Genus Heterorhabditis Poinar, 1976
Steinernema currently contains 15 valid species; Neosteinernema is mono-

typic and Heterorhabditis is represented by five recognized species.
According to A.P. Reid (personal communication) there are currently at
least another 20 isolated species of Steinernema, the status of which is clear-
ly indicated by molecular data. These await specific designation and formal
description via a combination of phenotypic and genotypic characteriza-
tion. In contrast to Steinernema, the genus Heterorhabditis appears to exist as
species complexes and a biological species concept is often used.
11.3.4 Characterization
(a) Classical morphology and morphometrics

Initial attempts at characterizing the species within these two families
involved the tried and tested approaches used on other soil-dwelling or
phytoparasitic nematodes, i.e. morphometrics and morphology (for a
review, see Poinar, 1990). The difficulties in employing morphometric
techniques are profound, as virtually all the stages in the life-cycle occur
bathed in their food supply in the protected environment of the host
haemocoel and parameters such as body length, width, etc., depend very
largely on extrinsic factors such as host species and nutrient availability. In
addition, as there are usually two generations within the host, there are
marked differences in nematode size between the generations, differences
which do not appear to be purely density-dependent. In an attempt to
overcome these difficulties, much attention has been focused on the infec-
tive third stage juvenile (the IJ3). This stage is the survival/infective aspect
of the life-cycle in both genera and is found in the soil. Being non-feeding,
its morphometric characters are more stable. Conversely, being a non-
feeding stage also means that its internal morphology is degenerate and
does not provide a wealth of useful characters. Some of the measurements
widely used in IJ3 identification are listed below:
Total body length (L)
Oesophagus length
Distance from cephalic extremity to the excretory pore
Tail length
Ratio E (head to excretory pore divided by tail length)
Ratio D (head to excretory pore divided by oesophagus length)
Ratios E and D are often represented in percentage form.
Body length, although variable, provides a useful parameter, dividing
the species of Steinernema, for example, into broad groups. Oesophagus
length is often difficult to determine with accuracy due to the degenerate
nature of the structure, but excretory pore position relative to the cephal-
ic extremity appears to be a fairly stable character in all the genera and has
an acceptable coefficient of variation. For this reason, the ratio E is often
employed in distinguishing species, as tail length is also reasonably con-
sistent.
Morphological characters include the tail shape and whether the ter-
minus contains a refractive spine or not, and the number and form of the
incisures in the lateral fields (Kozodoi and Spiridonov, 1988; Mracek and
Bednarek, 1991). Although the lateral field characters appear to work in a
reasonably satisfactory manner with the currently described species (and
isolates), their reliability will increasingly come under scrutiny now that
the groups potential as biocontrol agents has been realized, the resulting
attention throwing up a great many different isolates and not a few new
species. The great number of new isolates has caused particular problems
by increasing the perceived phenotypic variation in character profile of a
'species'. Attempts to counteract this tendency by inventing new ratios
(Nguyen and Smart, 1992) and by turning once again to the parasitic
stages for inspiration have yet to be overburdened by success. Male
characters, such as the shape (or even colour) of the copulatory spicules
and associated gubernaculum have been particularly used in Steinernema
where the presence or absence of a terminal mucron on the tail is anoth-
er, if somewhat more variable, character. While the spicule is undeniably
a potentially useful character, great care needs to be exercised as the
shape varies not only within a generation, but even more markedly
between generations in the same host. In addition, the spicule structure
is not confined to one plane and under the light microscope it is possible
to interpret different shapes which are not substantive - they merely
reflect relative observation angle. Variations in number and disposition of
copulatory papillae in Steinernema (typically numbering 21 or 23) offer
another possibility.
(b) Scanning electron microscopy

Use of SEM first contributed to the synonymy of Neoaplectana under
Steinernema on the basis of the number of cephalic papillae on the adults
(Mracek and Weiser, 1979, 1981). Subsequent work has shown that the
number and spacing of the incisures in the lateral field of the IJ3 can, if
characterized in a standard way, be a useful diagnostic character (Kozodoi
and Spiridonov, 1988; Mracek and Bednarek, 1991), although the tech-
nique has yet to be rigorously tested on the profusion of new isolates
and/or new species. It seems likely that, although the technique should
prove useful in dividing the species into a number of groups, there will be
insufficient variation in the lateral field structure to characterize specifi-
cally all of the new material awaiting examination. Other uses of SEM
include the morphology of the male copulatory spicules and gubernacu-
lum and the disposition of the copulatory papillae on the tail region of
Steinernema. These features are often difficult to see or interpret under the
light microscope, the spicule shape being particularly difficult to ascertain
as the outline of this three-dimensional structure varies considerably with
differing angles of view, a factor which is not always appreciated.
(c) Cross-breeding
Cross-breeding putative species using virgin females from one and males
from the other (together with the reverse cross) and assessing the viabili-
ty of the offspring is currently regarded as the acid test of reproductive
isolation and hence specific status of isolates. Although this technique
works relatively easily with Steinernema, an amphimictic genus,
Heterorhabditis has an hermaphroditic first generation and so the bisexual
second generation adults must be used. Two isolates currently grouped as
H. bacteriophora have been reported not to produce viable offspring (Dix et
al, 1991) when crossed.
(d) Host specificity
There is some evidence that certain species, such as S. kushidai (Mamiya,
1988) and S. scapterisci (Nguyen and Smart, 1990) reproduce much better
on specific insect hosts rather than on the commonly employed laborato-
ry insects. Most other species appear to be non-host-specific, at least under
laboratory conditions. There is also the complicating factor that our
knowledge of host specificity may be incomplete as most isolations from
soil samples are done by using a baiting technique employing the larvae
of the wax moth, Galleria mellonella. Thus, host specific isolates could well
be discriminated against in the collection/survey process and consequent-
ly go unrecorded.
(e) Biochemical and molecular techniques

Attempts to characterize these taxa reliably have involved a range of
techniques including allozyme electrophoresis (Akhurst, 1987), isoelec-
tric focusing (Kozodoi et al, 1986; Griffin et al., 1994; Joyce et al, 1994b)
and various techniques for analysing DNA (Curran and Webster, 1989;
Smits et al, 1991; Curran and Driver, 1994; Joyce et al, 1994a,b; Reid,
1994). Molecular techniques involving DNA banding patterns appear
particularly promising in the Steinernematidae (Curran, 1990; Reid,
1994) and probably represent the most rapid, accurate and reliable first
step with any unknown isolate. Reid and Hominick (1992, 1993, 1994)
and Reid (1994) reported on the use of RFLPs in the taxonomy of
Steinernema. The technique involves oligonucleotide primers flanking
the ITS region of the ribosomal DNA (rDNA) repeat unit and allows the
amplification of the DNA of a single IJ3 via PCR. Subsequently, the PCR
products are subjected to restriction digestion by a suite of enzymes and
the fragments separated via agarose gel electrophoresis. The resulting
library of restriction maps based on the rDNA repeat unit, together with
the comparative location of enzyme sites, allows phylogenetic compar-
isons and relationships between individual species within a single fami-
ly and also between the Steinernematidae and Heterorhabditidae to be
constructed. The topology of the resulting phylogenetic tree (Reid, 1994)
implied that the Steinernematidae and Heterorhabditidae were more
closely related than had been suggested by the diphyletic theory of
Poinar (1990,1993) and showed that the maps of the rDNA repeat units
of Steinernema revealed greater heterogeneity than those of
Heterorhabditis. Interestingly, two species (S. anomali and S. glaseri) dif-
fering only by their failure to interbreed (Poinar, 1990) were linked by
the tree as were two other morphologically similar species (S. affinis and
S. intermedia), the IJ3 of which is differentiated by the presence or
absence of a refractive spine in the tail tip. Joyce et al (1994c) discussed
a RAPD technique able to identify entomopathogenic nematodes from
either purified DNA or crude lysates of single IJ3s and it would appear
that this technique will be increasingly employed.
In the Heterorhabditidae, several workers report reliable results with
isoelectric focusing (IEF) of proteins (Griffin et at, 1994; Joyce et at, 1994b)
whereas others (Curran and Driver, 1994; Joyce et al, 1994a) employ
RFLPs in the rDNA intergenic spacer region (IGS) of Heterorhabditis iso-
lates. The latter technique was successful in allocating the isolates to broad
groupings concordant with the phenotypic species. Intra-group variation
was identified by the use of various restriction endonucleases and the
authors concluded that molecular markers would be useful in species
recognition, although simultaneously advocating caution in the interpre-
tation of such preliminary data and acknowledging the need for addi-
tional studies to complement their findings (Curran and Driver, 1994).
11.3.5 Discussion
In contrast to the Longidoridae, the most operational approach to the
entomopathogenic nematodes seems to be to use molecular methodolo-
gy as the preliminary investigative technique on new isolates. The preci-
sion of this more objective approach provides an indispensable tool not
only at the subspecific level where accurate identification of isolates can
be critical for biocontrol purposes, but also as a first approach at the
species level. If significant genomic differences are detected by reference
to a restriction map library, phenotype characterization of the putative
new species can then proceed with cross-breeding studies to determine
genetic isolation as the final arbiter of specific status. To date,
Heterorhabditis species and isolates appear to show much less variability
in their banding profiles than Steinernema species. This phenomenon may
be related to the life-cycle where the first generation of Heterorhabditis is
hermaphroditic, not amphimictic, thus resulting in a less variable geno-
type. Molecular techniques seem set to dominate in the initial characteri-
zation of entomopathogenic nematodes because of their ability to bypass
phenotypic variability by direct analysis of the genotype and to discrimi-
nate at all systematic levels.
11.3.6 Conclusion
Any identification system must not only be capable of working with the
existing species, but must be sufficiently soundly based to accommodate
new material without risk of undue compromise or, worse still, collapse.
One of the major problems in delimiting similar species on minute differ-
ences in phenotypic characters is the difficulty of incorporating a new
population into the scheme. Such a population, particularly in monosexual
species where non-lethal mutations can more readily increase in frequen-
cy and become fixed, is likely to be subtly different from the other nomi-
nal species and so the researcher is faced with the option of expanding the
range of an existing species, a process which, if oft repeated, leads to the
obliteration of supposed differences and synonymy of adjacent species.
This reinvokes the broad species concept which precipitated the dilemma
in the first place. The alternative approach of creating yet another name
for the new, intermediate, population risks virtually every population
being attributed to a different species - a less than enticing prospect, if one
which may have to be faced. In botany, this concept results in several hun-
dred nominal microspecies of, for example, the genus Rubus or the
apomictic hawkweeds and dandelions. Such zealous partitioning, even
when well-founded, has considerable practical implications in denying
access to any but the most determined esotericist.
The Longidoridae and the genus Xiphinema have, with the notable
exception of the X. americanum-group, been mostly well served by the clas-
sical approach of judiciously combining morphometrics with morpholo-
gy. The main problems in the taxonomy of the genus have arisen with the
parthenogenetic species, an area where the 'traditional' species concept is
always likely to be in difficulties. Heyns (1983) pointed out some of the
problems of using morphological characters in longidorid taxonomy. Such
factors include: subjective terminology when describing head or tail
shapes; differing interpretations of the same structure; post-mortem
changes due to fixation, processing techniques and temporal changes in
mounted specimens due to a gradual flattening allied with concomitant
changes in body shape, morphometric parameters, such as width and
length, and calculated ratios. Such problems are not restricted to the
Longidoridae - fixation and mounting methods may have profound
effects on morphology in other groups (Curran and Hominick, 1981).
Subtle differences may well be readily apparent and have great signifi-
cance to a specialist intimately familiar with a group of species, but tend to
be overlooked, viewed with suspicion or disregarded by others. It is partly
a result of relying on such potentially subjective characters, that the delim-
itation of species within the X. americanum-group has proved so controver-
sial. Yet there are clearly valid species within the group matrix and there
are undoubtedly others, perhaps many others, which are also valid, but
which may require the support of a non-morphological technique to
unambiguously express and categorize their distinctness. Miniscule phe-
notypic characters are not necessarily invalid, but there are very real prob-
lems in communicating such subtle differences via words and/or conven-
tional illustrations to a broader, and (perhaps rightly) cynical audience.
Taxonomy advances by consensus and muddying the water is not a step to
be taken lightly. What is needed is an alternative methodology with a dif-
ferent perspective; one which can provide a more profound view of
species and their relatedness one to another. Such an approach could be,
as with the entomopathogenic nematodes, DNA profiling based on PCR
methodologies. Vrain et al. (1992) showed the potential of such techniques,
although many more of the nominal species in the X. americanum-group
need to be scrutinized before substantive conclusions can be drawn.
By way of contrast, the classical approach has stumbled when faced
with the confusingly variable morphometrics and morphologically con-
served entomopathogenic nematodes of the Steinernematidae and
Heterorhabditidae. Although the classical approach is still employed in
describing new species, there is a fundamental requirement for an alterna-
tive, independent diagnostic technique which can readily handle and cat-
egorize the influx of new isolates from around the world. Such a technique
is molecular taxonomy. The ramifications and ultimate potential of, for
example, PCR-based techniques as a taxonomic tool can only be guessed at,
but molecular taxonomy has gone a long way in a relatively short time
span and its full potential has yet to be realized. Molecular techniques have
an intrinsic elegance; they must surely have a secure future in helping to
resolve problems in nematode identification and phylogenetic relation-
ships. As such, they should be welcomed and judged by classical taxono-
mists on their merits rather than being viewed with suspicion.
This is neither to advocate nor endorse a total reliance on molecular
characterization - the methodology offers an alternative insight into the
problem, a view which then needs to be tested with data from other
approaches before a consensus on speciation is reached. Molecular tax-
onomy is not a panacea; it is a powerful tool, but one which needs to be
handled with awareness and understanding. While DNA studies are
undeniably efficient at producing differing genomic profiles, we are
entitled to ask what do these mean? We need to clarify whether such dif-
ferences reflect the evolution of a particular gene, the organism, or both.
In the biosystematics of phytoparasitic and entomopathogenic nema-
todes, both the classical and molecular methodologies are valid
approaches. The two are complementary, offering different perspectives
on the same problem. The onus is on biosystematists to integrate the
methodologies and to achieve a synthesis, a synthesis with the ability to
delimit, recognize and identify an organism as its ultimate purpose. As
advocated by Hyman and Powers (1991) and Ferris and Ferris (1992), a
judicious combination of classical and molecular approaches can only
enhance our understanding of nematode speciation. In conclusion, the
following quote from Powers and Adams (1994) is particularly apt:
'...when we make the conceptual step from genetic marker to genet-

ic relatedness, we must be careful not to bias our view of nematode
relationships with expectations of well-defined species boundaries.
It would be convenient if nematode species existed at the terminal
branches of a well-pruned phylogenetic tree. The reality may be that
References 241
some nematode groups exist at the terminal branches of a tangled
hedge, with rather indistinct boundaries'.
Acknowledgements
I thank Alex Reid and Bernie Briscoe for helpful discussions on the char-
acterization of the Steinernematidae and Heterorhabditidae and Julie
Oliver and Janice Sheldon for assistance.
11.4 REFERENCES
Akhurst, R.J. (1987) Use of starch gel electrophoresis in the taxonomy of the genus
Heterorhabditis (Nematoda: Heterorhabditidae). Nematologica, 33,1-9.
Burrows, P.R. (1990) The use of DNA to identify plant parasitic nematodes.
Nematological Abstracts, 59,1-8.
Cain, A.J. (1959) The post-Linnaean development of taxonomy. Proceedings of the
Linnaean Society of London, 170, 23d 44.
Cenis, J.L., Opperman, C.H. and Triantaphyllou, A.C. (1992) Cytogenetic, enzy-
matic and restriction fragment length polymorphism variation of Meloidogyne
spp. from Spain. Phytopathology, 82, 527-31.
Chacon, M.R., Parkhouse, R.M.E., Burrows, P.R. and Garate, T. (1995) The use of
heterologous cloned DNA probes to distinguish between races of Meloidogyne
incognita. Nematologica, 41, 251-7.
Cho, M.R. and Robbins, R.T. (1991) Morphological variation among 23 Xiphinema
americanum populations. Journal of Nematology, 23,134 44.
Cohn, E. and Sher, S.A. (1972) A contribution to the taxonomy of the genus
Xiphinema Cobb, 1913. Journal of Nematology, 4, 36-65.
Coomans, A. and Chaves, E. (1995) Paraxiphidorus michelluci n. g., n. sp. from
Argentina (Nematoda: Longidoridae). Fundamental and Applied Nematology, 18,
303-6.
Cracraft, J. (1997) Species concepts in systematics and conservation biology - an
Cronquist, R. (1977) Once again, what is a species? in Beltsville Symposia in
Agricultural Research. 2. Biosystematics in Agriculture (ed. J.A. Romberger), Wiley,
New York, pp. 3-20.
Curran, J. (1990) Molecular techniques in taxonomy, in Entomopathogenic Nematodes
in Biological Control (eds R. Gaugler and H.K. Kaya), CRC Press, Boca Raton, pp.
63-74.
Curran, J. (1992) Molecular taxonomy of nematodes, in Nematology from Molecule to
Ecosystem (eds F.J. Gommers and P.W.Th. Maas), Proceedings, Second
International Nematology Congress, 11-17 August, 1990, Veldhoven, The
Netherlands, European Society of Nematologists, pp. 83-91.
Curran, J. and Driver, F. (1994) Molecular taxonomy of Heterorhabditis, in COST 822
Biotechnology: Genetics of entomopathogenic nematode-bacterium complexes,
Proceedings of Symposium and Workshop, St. Patrick's College, Maynooth, Co. Kildare,
Ireland (eds A.M. Burnell, R.-U. Ehlers, and J.P. Masson), Luxembourg,
European Commission, DG XII, pp. 41-8.
Curran, J. and Hominick, W.M. (1981) Effect of mounting methods on taxonomic
characters of adult male mermithids (Nematoda: Mermithidae). Nematologica,
26 (1980), 455-66.
Curran, J. and Webster, J.M. (1989) Genotypic analysis of Heterorhabditis isolates
from North Carolina. Journal of Nematology, 21,140-5.
De Giorgi, C, Sialer, M.F. and Lamberti, F. (1994) The use of mtDNA for the iden-
tification of plant parasitic nematodes, in Advances in Molecular Plant Nematology
(eds F. Lamberti, C. De Giorgi and D. Mck. Bird), vol. 268, NATO ASI Series,
Series A: Life Sciences, Plenum Press, pp. 109-18.
Dix, I., Burnell, A.M., Griffin, C.T., Joyce, S.A., Nugent, M.J. and Downes, M.J.
(1991) The identification of biological species in the genus Heterorhabditis
(Nematoda: Heterorhabditidae) by cross-breeding second-generation amphim-
ictic adults. Parasitology, 104, 509-18.
Esbenshade, P.R. and Triantaphyllou, A.C. (1990) Isozyme phenotypes for the
identification of Meloidogyne species. Journal of Nematology, 22,10-15.
Fargette, M., Blok, V.C., Phillips, M.S. and Trudgill, D.L. (1994) Genetic variation in
tropical Meloidogyne species, in Advances in Molecular Plant Nematology (eds F.
Lamberti, C. De Giorgi and D. Mck. Bird), vol. 268, NATO ASI Series, Series A:
Life Sciences, Plenum Press, pp. 91-6.
Ferris, V.R. and Ferris, J.M. (1992) Integration of classical and molecular approach-
es in nematode systematics, in Nematology from Molecule to Ecosystem (eds FJ.
Gommers and P. Th. Maas), Proceedings, Second International Nematology
Congress, 11-17 August, 1990, Veldhoven, The Netherlands. European Society
of Nematologists, pp. 92-100.
Glaser, R.W. and Farrell, C.C. (1935) Field experiments with the Japanese beetle and
its nematode parasite. Journal of the New York Entomological Society, 43, 345-71.
Griffin, C.T., Joyce, S.A., Dix, I., Burnell, A.M. and Downes, M.J. (1994)
Characterization of the entomopathogenic nematode Heterorhabditis
(Nematoda: Heterorhabditidae) from Ireland and Britain by molecular and
cross-breeding techniques and the occurrence of the genus in these islands.
Fundamental and Applied Nematology, 17,245-53.
Hahn, M.L., Burrows, P.R., Gnanapragasam, N.C., Bridge, J., Vines, N. and Wright,
D.J. (1994) Molecular diversity amongst Radopholus similis populations from Sri
Lanka detected by RAPD analysis. Fundamental and Applied Nematology, 17,
275-81.
Heyns, J. (1983) Problems of species delimitation in the genus Xiphinema, with spe-
cial reference to monosexual species, in Concepts in Nematode Systematics (eds
A.R. Stone, H.M. Platt and L.F. Khalil), vol. 22, Systematics Association Special
Volume, pp. 163-74.
Hooper, D.J. and Ibrahim, S.K. (1994) Aphelenchoides nechaleos n. sp. and A.
paranechaleos n. sp. (Nematoda: Aphelenchoididae) from rice plants.
Fundamental and Applied Nematology, 17,153-60.
Hunt, D. J. (1993) Aphelenchida, Longidoridae and Trichodoridae: Their Systematics and
Bionomics. CAB International, Wallingford, UK.
Hyman, B. and Powers, T. (1991) Integration of classical and molecular approach-
es in nematode taxonomy. Annual Review of Phytopathology, 29, 89-107.
Ibrahim, S.K. and Hooper, D.J. (1994) Aphelenchoides varicaudatus sp. n. (Nematoda:
Aphelenchoididae). Afro-Asian Journal of Nematology, 4,210-14.
References 243
Ibrahim, S.K., Perry, R.N. and Hooper, D.J. (1994a) Use of esterase and protein pat-
terns to differentiate two new species of Aphelenchoid.es on rice from other
species of Aphelenchoides and Ditylenchus angustus and D. myceliophagus.
Nematologica, 40, 267-75.
Ibrahim, S.K., Perry, R.N., Burrows, P.R. and Hooper, D.J. (1994b) Differentiation
of species and populations of Aphelenchoides and of Ditylenchus angustus using a
fragment of ribosomal DNA. Journal of Nematology, 26,412-21.
Joyce, S.A., Burnell, A.M. and Powers, T.O. (1994a) Characterization of
Heterorhabditis isolates by PCR amplification of segments of mtDNA and rDNA
genes. Journal of Nematology, 26, 260-70.
Joyce, S.A., Griffin, C.T. and Burnell, A.M. (1994b) The use of isoelectric focusing
and polyacrylamide gel electrophoresis of soluble proteins in the taxonomy of
Heterorhabditis (Nematoda: Heterorhabditidae). Nematologica, 40, 601-12.
Joyce, S.A., Reid, A., Driver, F. and Curran, J. (1994c) Application of polymerase
chain reaction (PCR) methods to the identification of entomopathogenic nema-
todes, in COST 812 Biotechnology: Genetics of entomopathogenic nematode-bacterium
complexes, Proceedings of Symposium and Workshop, St. Patrick's College, Maynooth,
Co. Kildare, Ireland (eds A.M. Burnell, R.-U. Ehlers and J.P. Masson),
Luxembourg, European Commission, DG XII, pp. 178-87.
Kaplan, D.T. (1994) Molecular characterization of the burrowing nematode sibling
species, Radopholus citrophilus and R. similis, in Advances in Molecular Plant
Nematology (eds F. Lamberti, C. De Giorgi and D. Mck. Bird), vol. 268, NATO
ASI Series, Series A: Life Sciences, Plenum Press, pp. 77-83.
Kozodoi, E.M. and Spiridonov, S.E. (1988) Cuticular ridges on lateral fields of inva-
sive larvae of Neoaplectana (Nematoda: Steinernematidae). Folia Parasitologica,
35, 359-62.
Kozodoi, E.M., Voronov, D.A. and Spiridonov, S.E. (1986) [The use of an elec-
trophoretic method to determine the species association of Neoaplectana speci-
mens]. Trudy Gel' minthologicheskoi laboratorii (Voprody biotsenologii gel' mintov),
34, 55-9.
Lamberti, F. and Bleve-Zacheo, T. (1979) Studies on Xiphinema americanum sensu
lato with descriptions of fifteen new species (Nematoda, Longidoridae).
Nematologia mediterranea, 7, 51-106.
Lamberti, F. and Carone, M. (1991) A dichotomous key for the identification of
species of Xiphinema (Nematoda: Dorylaimida) within the X. americanum-group.
Nematologia Mediterranea, 19, 341-8.
Lamberti, P., De Giorgi, C. and Bird, D. Mck. (eds) (1994) Advances in Molecular
Plant Nematology. NATO ASI Series, Series A: Life Sciences, vol. 268, Plenum
Press.
Loof, P.A.A. and Luc, M. (1990) A revised polytomous key for the identification of
species of the genus Xiphinema Cobb, 1913 (Nematoda: Longidoridae) with
exclusion of the X. americanum group. Systematic Parasitology, 16, 35-66.
Loof, P.A.A. and Luc, M. (1993) A revised polytomous key for the identification of
species of the genus Xiphinema Cobb, 1913 (Nematoda: Longidoridae) with exclu-
sion of the X. americanum group. Supplement 1. Systematic Parasitology, 24,185-9.
Loof, P.A.A., Luc, M. and Coomans, A. (1993) The Xiphinema americanum group
(Nematoda: Dorylaimida). 1. Comments upon the key to species published by
Lamberti and Carone (1992). Fundamental and Applied Nematology, 16, 355-88.
Luc, M. and Dalmasso, A. (1975a) A 'lattice' for the identification of species of
Xiphinema Cobb, 1913, in Nematode Vectors of Plant Viruses (eds F. Lamberti, C.E.
Taylor and J.W. Seinhorst), Plenum Press, London, New York, pp. 53-70.
Luc, M. and Dalmasso, A. (1975b) Considerations on the genus Xiphinema Cobb,
1913 (Nematoda: Longidoridae) and a 'lattice' for the identification of species.
Cahiers de I' Office de la Recherche Scientifique et Technique Outre-Mer, Biologie, 10,
303-27.
Mamiya, Y. (1988) Steinernema kushidai n. sp. (Nematoda: Steinernematidae) asso-
ciated with scarabaeid beetle larvae from Schizuoka, Japan. Applied Entomology
and Zoology, 23, 313-20.
Mayr, E. (1969) Principles of Systematic Zoology, McGraw-Hill, New York.
Mracek, Z. and Bednarek, A. (1991) The morphology of the lateral fields of infec-
tive juveniles of entomogenous nematodes of the family Steinernematidae
(Nematoda). Nematologica, 37, 63-71.
Mracek, Z. and Weiser, J. (1979) The head papillae of the Steinernematidae. Journal
of Invertebrate Pathology, 34, 310-11.
Mracek, Z. and Weiser, J. (1981) Head and cuticular structures of some species in
the family Steinernematidae (Nematoda). Nematologica, 27,443-8.
Nguyen, K. B. and Smart, G. C. (1990) Steinernema scapterisci n. sp. (Rhabditida:
Steinernematidae). Journal ofNematology, 22,187-99.
Nguyen, K. B. and Smart, G. C. (1992) Steinernema neocurtillis n. sp. (Rhabditida:
Steinernematidae) and a key to the species of the genus Steinernema. Journal of
Nematology, 24,463-77.
Poinar, G. O. (1990) Taxonomy and biology of Steinernematidae and
Heterorhabditidae, in Entomopathogenic Nematodes in Biological Control (eds R.
Gaugler and H.K. Kaya), CRC Press, Boca Raton, pp. 23-61.
Poinar, G. O. (1993) Origins and phylogenetic relationships of the entomophilic
rhabditids, Heterorhabditis and Steinernema. Fundamental and Applied Nematology,
16, 333-8.
Powers, T.O. and Adams, B. (1994) Nucleotide sequences in nematode systemat-
ics, in Advances in Molecular Plant Nematology (ed. F. Lamberti, C. De Giorgi and
D. Mck. Bird), NATO ASI Series, Series A: Life Sciences, vol. 268, Plenum Press,
pp. 99-108.
Powers, T. O. and Harris, T. S. (1993) A polymerase chain reaction method for
identification of five major Meloidogyne species. Journal ofNematology, 25,1-6.
Reid, A.P. (1994) Molecular Taxonomy of Steinernema, in COST 812 Biotechnology:
Genetics of entomopathogenic nematode-bacterium complexes, Proceedings of
Symposium and Workshop, St. Patrick's College, Maynooth, Co. Kildare, Ireland (ed.
A.M. Burnell, R.-U. Ehlers and J.P. Masson), Luxembourg, European
Commission, DG XII, pp. 49-58.
Reid, A.P. and Hominick, W.M. (1992) Restriction fragment length polymorphisms
within the ribosomal DNA repeat unit of British entomopathogenic nematodes
(Rhabditida: Steinernematidae). Parasitology, 105, 317-23.
Reid, A.P. and Hominick, W.M. (1993) Cloning of the rDNA repeat unit from a
British entomopathogenic nematode (Steinernematidae) and its potential for
species identification. Parasitology, 107, 529-36.
Reid, A.P. and Hominick, W.M. (1994) Molecular taxonomy and phylogeny of
entomopathogenic nematodes. Paper presented at the 22nd International
Nematology Symposium, Gent, Belgium, 7-12 August, 1994; abstract p. 107.
References 245
Sasser, J.N. and Carter, C.C. (1985) Overview of the International Meloidogyne
Project 1975-1984, in An Advanced Treatise on Meloidogyne. Biology and Control
(eds J.N. Sasser and C.C. Carter), North Carolina State University, pp. 19-24.
Smits, P.H., Groenen, J.T.M. and De Raay, G. (1991) Characterization of
Heterorhabditis isolates using DNA restriction fragment length polymorphism.
Revue de Nematologie, 14, 445-53.
Steiner, G. (1923) Aplectana kraussei n. sp., einer in der Blattwespe Lyda sp. para-
sitierende Nematodenform, nebst Bemerkungen her das Seitenorgan der par-
asitischen Nematoden. Zentralbl. Bakteriol. Parasitenkd. Infektionskr. Hyg. Abt. 2,
59,14.
Swart, A. and Heyns, J. (1987) Comparative morphology of the head regions of
some longidorid nematodes from South Africa using the SEM. Phytophylactica,
19, 99-101.
Tarjan, A.C. (1969) Variation within the Xiphinema americanum group (Nematoda:
Longidoridae). Nematologica, 15,241-52.
Tchesunov, A.V. and Rieman, F. (1995) Arctic sea ice nematodes (Monhysteroidea),
with descriptions of Cryonema crassum gen. n., sp. n. and C. tenue sp. n.
Nematologica, 41, 35-50.
Vrain, T.C. and Wakarchuk, D.A. (1989) Construction of a set of cloned rDNA
probes for the specific differentiation of the Xiphinema americanum group.
Journal of Nematology, 21, 593-4.
Vrain, T.C., Wakarchuk, D.A., Levesque, A.C. and Hamilton, R.I. (1992)
Intraspecific rDNA restriction fragment length polymorphism in the Xiphinema
americanum group. Fundamental and Applied Nematology, 15, 563-73.
Williamson, V.M., Caswell-Chen, E.P., Wu, F.F. and Hanson, D. (1994) PCR for
nematode identification, in Advances in Molecular Plant Nematology (eds F.
Lamberti, C. De Giorgi and D. Mck. Bird), NATO ASI Series, Series A: Life
Sciences, vol. 268, Plenum Press, pp. 119-27.
12
Species in insect herbivores and
parasitoids - sibling species, host
races and biotypes
M. E Claridge, H. A. Dawah and M. R. Wilson
Contacting address: School of Pure and Applied Biology, University of Wales Cardiff,
P.O. Box 915, Cardiff CF1 3TL, Wales, UK
ABSTRACT
The species diversity of terrestrial ecosystems is dominated by insects.
Of these, the most speciose groups are those that feed either on green
plants as herbivores, or as parasitoids on other insects, including the
herbivores. Generally these insects are characterized by extreme levels
of host specificity and low levels of morphological differentiation. The
genetic and taxonomic status of populations from different hosts is
often very difficult to determine. Species taxonomy is thus a major
problem. Particular difficulties with host-associated populations are
shared with all parasitic organisms. The terms host race, biotype, etc.
have been widely, but often uncritically, used.
In many families differentiation of the male external genitalia pro-
vides morphological markers widely used by taxonomists to discrimi-
nate species. Most groups of Hymenoptera Parasitica, which include
the greatest diversity of insect parasitoids, do not exhibit such diversi-
fication of male genitalia so that species characterization is even more
difficult.
So-called biological concepts have been used widely for sexually
reproducing species. The investigation of specific mate recognition
systems as factors maintaining reproductive isolation have revealed
sibling species in a variety of taxa. Enzyme electrophoresis and DNA
technologies are being used widely to provide species markers.
In general, biological and phylogenetic concepts will provide simi-
lar solutions to species problems in these insects.
248 Species in insect herbivores and parasitoids
12.1 INTRODUCTION
Insects dominate the diversity of life on earth. They comprise more than
56% of the described species of all living organisms and best recent esti-
mates suggest that about 64% of all existing species are insects
(Hammond, 1992). Paradoxically, in morphology the insects are a relative-
ly conservative group so that species and higher taxon differentiation has
to be based on relatively small differences.
Of living insect species it is estimated that about 50% are herbivores,
mostly associated with vascular plants (Strong et al., 1984). Of the remain-
ing 50%, most are thought to be parasitoids, largely attacking insect her-
bivores as their primary hosts. Thus, the diversity of modern insects is
dominated by these two essentially parasitic life-styles (Price, 1980). Both
insect herbivores and parasitoids are characterized by extreme specializa-
tion. In particular most species are extremely specific in host utilization,
typically being restricted to one, or a few related, host species. More
polyphagous species are unusual and typically fewer in number (Figure
12.1). Clearly, a precise understanding of the nature of species and the
species status of particular populations is important in understanding
such phenomena.
Since the first descriptions by Linnaeus (1759), the species taxonomy of
insects has been almost entirely dependent on external morphological
characters. The generally hardened nature of insect cuticle means that
samples are still most frequently preserved dry in museum collections and
this dependence on external morphology still continues. Indeed, the his-
tory of insect taxonomy may be seen as one of seeking for more and more
refined methods for the morphological differentiation of species. In the
early years of the 20th century entomologists realized that the structure of
the external genitalia, particularly but not exclusively of males, provided a
new and very diverse array of characters that were extremely valuable in
discriminating between otherwise very similar, or indeed apparently
identical, species. The result is that for very many groups of insects from
most orders, characters of the genitalia are essential for species delimita-
tion. For example, in the large leafhopper subfamily Typhlocybinae
(Cicadellidae), species are mostly small and rather uniform in appearance.
The discovery of distinctive features of the male genitalia very greatly
increased the number of recognized species. For example, the genus
Edwardsiana (=Typhlocyba in part) in Europe includes small tree-living
leafhoppers, mostly of a uniform pale yellow colour. Ribaut (1936) first
made detailed studies widely on male genitalia of these insects. He found
in particular that the terminal branching of the aedeagus was very vari-
able (Figure 12.2) and then went on to recognize 19 species from France on
this basis. Ossiannillsson (1981), using the same methodology, recognized
24 species from Scandinavia. With few exceptions, Edwardsiana females
show no obvious morphological differentiation. However, some
Number of species
g 8
confirmation of the biological reality of the male-based differentiation is
provided by analyses of host plant-associated distributions of the species
so recognized (Claridge and Wilson, 1976, 1981). The realization of the
importance of genitalial characters in species taxonomy has been repeated
for most major groups of insect herbivores, including many Hemiptera,
Lepidoptera, Diptera, Coleoptera, Hymenoptera and most smaller orders.
The biological significance of such species characteristics has been
reviewed by Eberhard (1985). Most insect parasitoids are Hymenoptera
Parasitica. In general, male genitalia in this group are relatively simple in
structure and do not provide reliable morphological species markers; con-
sequently, species taxonomy in these families is notoriously difficult.
There are some notable, but rare, exceptions to this generalization, for
example egg parasites of the genus Trichogramma (Nagarkatti and
Nagaraja, 1971).
Most insect herbivores and parasitoids are biparental, sexually repro-
ducing organisms. Thus, a biological species concept may be applied to
host-associated populations and the determination of species status will
ultimately depend on the demonstration of high levels of reproductive
isolation in the field. The morphological characters generally used for con-
venience in the study of these insects are thus markers indicating lack of
gene flow. In many groups, refined biochemical markers, primarily
obtained as a result of using enzyme gel electrophoresis, but also more
often now also by DNA technology, have been used to establish levels of
gene flow and reproductive isolation. In some cases, reviewed below,
attempts have been made to study the specific mate recognition systems
which result in the observed reproductive isolation between species. Also
such studies may help in the determination of the genetic status of
allopatric populations. In addition, mate choice experiments - if
adequately controlled - can provide essential information on the species
status of related populations (Claridge and Morgan, 1987; Claridge, 1988).
12.2 STATUS OF HOST-ASSOCIATED POPULATIONS - HOST

RACES AND BIOTYPES
In all parasitic organisms a major factor, if not the most important one, is
the relationship with their hosts which provide not only a food source but
also frequently the bulk of their environment. This is certainly true of insect
herbivores and parasitoids. The interpretation of the taxonomic and genet-
ic status of populations from different, often related, hosts has long been an
area of controversy. Where such populations differ clearly in morphology
there has normally been no problem about recognizing them as distinct
species, whatever the author's preferred species concept. However, where
host-related populations show no clear-cut diagnostic differences in mor-
phology, but do show some differences in survival and development on a
avellanae
bergmani
crataegi
spinigera
Figure 12.2 Variation in the male aedeagus as shown in seven species of Edwardsiana (Cicadellidae, Typhlocybinae). (After Ribaut,
1936.)
particular host, difficulties have arisen. This is a longstanding problem first
clearly enunciated by Benjamin Walsh (1864) in his classic paper on
'phytophagic varieties' and 'phytophagic species' (see also Bush, 1995).
Not surprisingly, such problems have been most widely studied in pests of
agriculture and forestry. The phytophagic races of Walsh have more usu-
ally been termed biological races in recent years (see Thorpe 1930 for early
review; and Jaenike, 1981) or biotypes (Claridge and den Hollander, 1983;
Diehl and Bush, 1984). The interpretation of such populations, which
apparently differ primarily in adaptation to survival and reproduction on
particular hosts, is difficult, but has often been made more so by exclusive
support for particular theories of speciation. Thus, strict adherence to the-
ories of allopatric speciation (Mayr, 1963; Paterson, 1985) would suggest
that host-associated differentiation and speciation will occur only when
populations are isolated in space. Intermediate, partially differentiated
host races or biotypes should then only be found in allopatry. However, if
sympatric speciation is prevalent (Bush, 1975,1993,1994) then all interme-
diate host race stages between freely interbreeding parasitic populations
and fully isolated biological species should be found reasonably frequent-
ly in sympatry. The arguments between the two general schools of thought
continue. However, for parasitic organisms the determination of allopatry
and sympatry may be difficult. Ultimately, differentiating spatial from host
isolation may itself be impossible and have little meaning; indeed, perhaps
this is no longer a useful approach. More important is an understanding of
the nature of the genetic differences between host-associated populations
and the degree and regularity of gene flow between them.
In practice, morphological differentiation is still most widely used in
attempts to characterize host associated populations of insects. The contin-
uing development of modern multivariate techniques for handling quanti-
tative data derived from morphometric studies have greatly expanded the
possibilities for analysis (Foottit and Sorenson, 1992). However refined the
statistical methods available, it is essential to differentiate between those
characteristics of a population or individual that are a direct result of
induced responses to feeding and living on particular hosts and those that
represent real genetic differences between the insect populations. The
ideal way to achieve this is to transfer insects between hosts and repeat the
original morphometric measurements (Claridge and Gillham, 1992).
Unfortunately this is often not easy for purely practical reasons, including
difficulties in growing suitable host plants and culturing particular herbi-
vores and parasitoids. However, a few detailed examples are available.
12.2.1 The leafhopper Alnetoidia alneti

Alnetoidia alneti (Dahlbom) is a common and easily recognized species in
Europe. In Britain it feeds and reproduces in the field on 17 species from
Status of host-associated populations - host races and biotypes 253
13 different genera of trees and shrubs (Claridge and Wilson, 1981).
Individuals reared from some different hosts differ in size and colour. This
prompted Ribaut (1936) to recognize two species -A. alneti found on alder
and elm, and a smaller and paler form, A. coryli from hazel. Claridge and
Gillham (1992) and Gillham and Claridge (1994) reported multivariate
analyses of populations of A. alneti from 13 different host plants in South
Wales and southern England based on 27 different body measurements.
Canonical variates analyses (CVA) were performed on these data and
demonstrated significant differences between most populations. The most
extreme were those from alder and hazel. Populations sampled in differ-
ent years and from different parts of South Wales and southern England
showed very little within host-plant group variation. First and second
instar nymphs collected from a single alder tree in South Wales were
reared on the foliage of three alder trees and similar nymphs collected
from one hazel tree were reared on three hazels. In addition, a sample of
the same nymphs from alder were reared on hazel foliage. CVAs of adult
measurement data for these samples show very clearly that those reared
on alder and hazel respectively were all similar among themselves, but
quite distinct from each other. That transferred from alder to hazel, how-
ever, grouped with the hazel insects (Figure 12.3). Thus, the clear differ-
ences between these samples, which included features of both size and
shape, must be interpreted as being host plant-induced. There is then no
support in these data to suggest that the host plant associated populations
of A. alneti are genetically distinct and certainly not that they are separate
biological species.
12.2.2 Biotypes of the brown planthopper, Nilaparvata lugens

Nilaparvata lugens (Stal) is a major pest of cultivated rice, Oryza sativa,
throughout Asia. Populations of N. lugens identified by screening against
cultivars of rice which incorporate different genes for resistance to feeding
have been termed biotypes. They are most obviously categorized by their
abilities in mass screening trials to damage particular rice cultivars charac-
terized by different genotypes (Table 12.1). Saxena and co-workers sug-
gested that these biotypes were equivalent to host races and were therefore
intermediate stages in a speciation process. They used CVA techniques to
demonstrate significant differences between the biotypes numbered 1, 2
and 3 reared on the rice cultivars to which they were adapted (Saxena and
Rueda, 1982). However, when the same biotypes were reared on the same
susceptible cultivar the significant differences between them disappeared
(Claridge et al., 1984) (Figure 12.4). These findings, together with the ease
with which these biotypes hybridize and the viability of Fj and F2 hybrids
(Claridge et al., 1985a), suggest that genetic differentiation between them is
slight.
4-
2-
1-2-
-4-
I r
-4 -2
Discriminant function 1
Figure 12.3 Canonical variate plots of space circumscribed for adults of the
leafhopper Alnetoidia alneti reared on hazel, Corylus avellana, (grouped to the left)
and alder, Alnus glutinosa (grouped to the right). Shaded areas represent adults
derived from early instar nymphs originally collected from alder. (After Claridge
and Gillham, 1992.)
(a) (b)
4-
2-
V o•
«00»
ooooo ooo •
*2 o- o oo» 3oo
03
I -2-
o
1 1 I I I I I
-4 -2 2 4 6 -A -2
Discriminant function 1
Figure 12.4 Canonical variate plots for biotypes 1,2 and 3 of Nilaparvata lugens. (a)
reared respectively on cultivars TNI, Mudgo and ASD7; (b) all reared on TNl.
(After Claridge et al, 1984.)
Biological species, specific mate recognition and sibling species 255
Table 12.1 The possible 'biotypes' of Nilaparvata lugens on rice showing some tra-
ditional and modern high-yielding cultivars susceptible to each and associated
nomenclature. The dominance status of different genes in the plants is also indi-
cated. (After Claridge and den Hollander, 1982.)
Biotype Resistance gene Rice variety

Traditional Improved
I None identified TB1 IR8

2 Bph 1 (dominant) Mudgo IR26
3 bph 2 (recessive) ASD7 IR36
IR42
4 Bph 3 (dominant) Rathuheenati IR56
5 bph4 (recessive) Babawee -
Populations attributable morphologically to N. lugens, but feeding

exclusively on the grass Leersia hexandra, show no greater morphometric
differentiation than do the rice-associated biotypes and were indeed orig-
inally designated as a further biotype by Saxena et al. (1983). However, we
now know them to represent different biological species (Claridge et al.,
1985b, 1988, and see below).
These examples demonstrate the need for care in interpreting differen-
tiation between host-associated populations in the absence of experimen-
tal and genetic data.
12.3 BIOLOGICAL SPECIES, SPECIFIC MATE RECOGNITION AND

SIBLING SPECIES
A solution to the problems of interpreting the different degrees of host-
associated variation discussed above in both insect herbivores and
parasitoids is to use the biological species concept with its emphasis on
reproductive isolation in the field, achieved through distinct mate recogni-
tion systems, and thus lack of significant gene flow between species
(Claridge et al., 1997: Chapter 1). This is possible since most insects are
biparental. The morphological characters normally used in taxonomy are
markers for presumed genetic differentiation of the species under study.
More direct methods involve the use of a variety of biochemical and molec-
ular techniques (Loxdale and den Hollander, 1989; Hawksworth, 1994;
Symondson and Liddell, 1996). In particular, enzyme gel electrophoresis has
provided simple techniques for the estimation of gene flow between popu-
lations. The identification of diagnostic and unique enzyme gene loci in host-
associated populations clearly establishes their biological species status
(Menken and Ulenberg, 1987; Menken, 1989). These techniques have now
been applied widely in many groups of insect herbivores and parasitoids.
A particularly thorough study has been made of the small ermine
moths, Yponomeuta species, in Europe (summary in Menken et al, 1992).
Most differ by small, but clear morphological characters, particularly of
the male genitalia. The larvae of these insects feed on the foliage of a vari-
ety of mostly trees and shrubs. Some species show minor host plant dif-
ferentiation but no clear-cut morphological characterization. For example,
Y. malinellus attacks Malus and Pyrus species and Y. padellus a variety of
other Rosaceae, including Cmtaegus and Prunus, but not Malus or Pyrus.
These two were originally cited as examples of host races of one species
(Thorpe, 1929), but Menken (1980) showed that, despite high levels of
genetic identity that allow laboratory hybridization between them, inter-
breeding does not occur in the field and at least one completely diagnos-
tic locus has been identified.
Similar techniques have been used to determine species status in a wide
variety of insect groups. Reviews of many such examples are provided in
the volume edited by Loxdale and den Hollander (1989). The diversity of
genetic techniques available now provides a refined set of markers by
which we may establish the existence of reproductive isolation between
sympatric populations. However, biological species are characterized by
specific mate recognition systems (Paterson, 1985) which result in the
observed reproductive isolation between species. Thus, the study of spe-
cific mate recognition systems (SMRS) will enable us to identify biological
species with greater certainty.
Mate finding and courtship in insects is usually very complex and
involves a sequence of signal exchanges between males and females
before mating and successful fertilization take place. This complete
sequence of signal exchange forms the SMRS of any species. Insects have
a wide diversity of sense organs and are capable of discriminating signals
in different modalities. Any complete SMRS is likely to involve different
signals and receptors, including the chemical, visual, mechanical and
auditory senses, and will be unique for each species. Unfortunately, rather
few studies have been made of complete SMRS sequences. However,
since the chemical senses predominate in many aspects of insect life it is
probable that specific chemical signals - pheromones - are usually impor-
tant. These are particularly well known in the Lepidoptera where in many
species specific distance communication is achieved in mate finding by
the release of a sex pheromone by virgin females. This, known as 'calling'
behaviour, brings conspecific receptive males to the stationary females,
often over very long distances. In many species further pheromones are
released by both sexes at closer range as further elements of the SMRS
sequence. At all stages the potential for chemical discrimination is very
great. The pheromones in many species consist of cocktails of several com-
pounds with specificity being conferred by the precise proportions of the
different compounds involved. Similar chemical systems of communication
dominate mate finding and courtship in many major groups of herbivores
and parasitoids, including many Coleoptera, Diptera and Hymenoptera
(Phelan, 1992).
Few detailed studies on pheromone systems have been made on
groups of related insect herbivores or parasitoids. Probably the best
worked are the nine species of small ermine moths, Yponomeuta, in Europe
(Menken et al., 1992). Virgin females 'call' as described above. Males
respond maximally to the female pheromone of their own species
(Hendrikse, 1979, 1986). When in close proximity males emit their own
pheromone, accompanied usually by characteristic wing fanning behav-
iour (Hendrikse et al., 1984). The result is that attempts at interspecific cop-
ulations are very rare. These chemical interactions are clearly central to the
SMRS of Yponomeuta species.
Unfortunately, chemical signals and their role in courtship are difficult
to study and at least require complex analytical facilities. Difficulties of
chemical analysis mean that such systems have rarely been used to deter-
mine the nature of SMRS sequences. Most of what we know derives from
studies on individual pest species where pheromones may be used in
traps for monitoring populations and even in systems of control by dis-
rupting normal mating in the field. However, there is little doubt that fur-
ther analyses of pheromone systems will lead to the recognition of groups
of sibling species presently thought to be one. Published examples of sup-
posed pheromone polymorphisms should be treated with suspicion in
this context.
Visual and acoustic systems of communication and mate finding are
rarer than chemical ones, but easier for the human observer to study. In
particular, acoustic systems have been widely studied, mainly among the
Orthoptera and Hemiptera. The Homoptera Auchenorrhyncha is a major
group exclusively of herbivores. All species so far as known use acoustic
signals in the initial stages of mate finding and courtship (Claridge,
1985a,b). There is no evidence of chemical communication in a sexual con-
text before physical contact is made between potential mates. In the larg-
er cicadas (Cicadoidea) long-distance communication is achieved by loud,
high-intensity male calling to which virgin females are attracted. In close
contact other calls may be made and pheromones may also be important.
In all other groups, including the species-rich Delphacidae and
Cicadellidae, low-intensity acoustic signals are produced by males and
usually also by females. These calls are transmitted through the plant sub-
strate on which the insects live. Calls of both sexes are usually species spe-
cific and form essential elements of the SMRS sequence and have now
been used to establish biological species status in a number of different
genera (Claridge, 1985a,b; Claridge and de Vrijer, 1994).
One of the best examples of the analysis of a group of sibling species is
provided by the work of Wood (1993) on the complex of morphologically
258 Species in insect herbivores and pamsitoids
almost identical species of treehoppers (Membracidae) in North America,
known as Enchenopa binotata. In a series of elegant and detailed elec-
trophoretic studies Wood has shown that £. binotata consists of at least
nine different host-specific and reproductively isolated biological species.
Recently, Hunt (1994) has shown that these insects also use acoustic sig-
nals in courtship and mating behaviour. Further studies on all of the
species will provide exciting evidence in the debate about species and spe-
ciation in these insects.
12.3.1 Species of Nilaparvata - a herbivore case study

The planthopper genus Nilaparvata includes 16 morphological species, dif-
ferentiated primarily on features of the male genitalia and mostly distrib-
uted in the tropical and subtropical regions of the world. N. lugens (Stal),
the Brown Planthopper, is a major pest of rice in Asia and thus well stud-
ied (Claridge and Morgan, 1987; Wilson and Claridge, 1991).
(a) Nilaparvata lugens

N. lugens (Figure 12.5) is clearly differentiated from all other species on a
basis particularly of the male aedeagus and paramere structure. It is wide-
ly distributed through much of Asia and northern Australia where it feeds
only on wild and cultivated rices, Oryza (Figure 12.5). Over this enormous
geographical range it shows no obvious variation in morphology.
Like other planthoppers, N. lugens uses primarily acoustic signals trans-
mitted through the plants on which they live in mate finding and
courtship (Claridge and de Vrijer, 1994). Receptive adult males move
actively from plant to plant producing characteristic amplitude-modulat-
ed signals (Figure 12.6). If a receptive virgin female is within the acoustic
field she may respond with her own simple call consisting of a train of reg-
ularly repeated pulses (Figure 12.6). In a typical sequence the male and
female continue to exchange calls. While the female remains stationary,
the calling male moves actively over the plant surface until contact is
made. Calling continues and copulation usually ensues. There is no doubt
that acoustic signals are vitally important in species recognition in these
insects. Indeed, appropriate behaviour from both males and females may
be elicited simply by the use of pre-recorded signals in the absence of an
insect (Claridge and de Vrijer, 1994).
Populations of N. lugens from different parts of Asia and Australia show
geographical variation in quantitative features of both male and female
calls, in particular the rates at which pulses are produced - pulse repeti-
tion frequency (PRF) - in main sections of their calls. Most extreme are the
differences between populations from Australia and those collectively
from Asia. Correlated with these quantitative differences in calls is the
Figure 12.5 Adult Nilaparvata lugens on rice plants. Brachypterous female (left)
and macropterous male (right).
relative difficulty of obtaining laboratory hybrids between such popula-

tions (Claridge et al., 1985a). Mate choice experiments between Asian and
Australian populations show significant preferences for homogametic
matings, though not total reproductive isolation. The relative status of
populations from Asia and Australia is thus doubtful.
Populations morphologically attributable to N. lugens living and feed-
ing on the grass Leersia hexandra are now known to occur widely over Asia
and Australia where they are frequently sympatric with rice-associated
ones (Figure 12.5) (Claridge et al, 1985b, 1988). Calls of both males and
females from sympatric populations differ significantly in PRF. However,
as in the rice-feeding populations, those from Leersia show similar geo-
graphical variation (Claridge et al., 1985b, 1988). Mate choice experiments
for sympatric populations from both the Philippines and Australia
showed very significant preferences for homogametic matings.
Laboratory-produced hybrids between the two populations from the
Philippines were intermediate in call characters between the two parental
o N. lugens (Leersia)
• N. lugens (rice)
* N. bakeri (Leersia)
Figure 12.6 Sketch map of Asia to show localities from which Claridge and co-workers have sampled Nilaparvata lugens from rice and
Leersia hexandra, and N. bakeri from Leersia hexandra only.
types and significantly different from both. Thus, if hybrids were com-
monly produced in field samples they would be detectable. However, no
field-caught individuals with such intermediate calls were found by
Claridge et al (1985b, 1988) from any of the many regions in Asia and
Australia that were sampled (Figure 12.5).
Playback of pre-recorded calls to both males and females of each host-
associated population showed very significant preferential responses for
the calls of their own type (Table 12.2). Thus, it is clear that the populations
of N. lugens from rice and Leersia respectively in both Asia and Australia
represent different biological species and not just different biotypes or
host races, as suggested by Saxena and Barrion (1985). The status of the
populations in Australia by comparison with those in Asia is difficult to
determine. However, some recent preliminary molecular data suggest
that Australian populations should be regarded as specifically distinct
from those in Asia (Jones et al., 1996).
(b) Nilaparvata bakeri

Of the remaining morphological species of Nilaparvata only N. bakeri has
been studied intensively. It is a distinct morphological species differing
from N. lugens in obvious features of both the male aedeagus and para-
meres. In Asia, N. bakeri is widely sympatric with both the rice- and Leersia-
associated species of N. lugens, but it has not yet been found with certain-
ty in Australia. It is a specific feeder associated only with Leersia.
The calls of N. bakeri males and females show some similarity to those
of N. lugens, but differ significantly, particularly in the males (Figure 12.7)
(Claridge and Morgan, 1993). Neither males nor females respond to the
calls of N. lugens from either rice or Leersia. Despite many efforts neither
have we been able to obtain laboratory hybrids between them. They
Table 12.2 Percentages of males and females of rice- and Leersw-associated pop-
ulations of Nilaparvata lugens responding to playback of prerecorded calls of
female and male calls respectively. Actual numbers of insects responding out of
total numbers given in parentheses. (After Claridge et al., 1985b.)
Pecentage responding to
Rice male call Leersia male call
Female response
Rice females 93 (14/15) 13 (2/15)
Leersia females 27 (4/15) 87 (13/15)
Rice female call Leersia female call
Male response
Rice males 77 (23.30) 30 (6/20)
Leersia males 17 (5/30) 90 (18/20)
Male
Rice
Leers/a
1s
Female
Rice
Leers/a
1s
Figure 12.7 Oscillograms of sections of individual male and female calls from rice-
and Leersw-associated species of Nilaparvata lugens from the Philippines. (After
Claridge et al, 1985b.)
undoubtedly represent quite distinct species, whatever species concept is

employed.
Like both species attributed to N. lugens, N. bakeri shows no obvious
geographical variation in morphology. However, populations from Luzon
(Philippines), Bali (Indonesia), West Bengal (India) and Kandy (Sri Lanka)
show dramatic differences in male calls (Claridge and Morgan, 1993). In
particular, calls from Philippine males are strikingly different to those
from Bali to the human observer (Figure 12.8). However, mate choice
experiments showed no preference for homogametic matings between
these two very distinct call types. Indeed, there is no indication that the
striking call differences to the human observer between these allopatric
Herbivore/parasitoid food webs - grass-feeding chalcid wasps 263
populations would have any role in maintaining reproductive isolation
between them. Thus, despite clearly diagnostic differences in calls they
cannot be regarded as different species.
Unfortunately, most of the other morphological species of Nilaparvata
are little known. Distinctive calls have been recorded for N. muiri from
Japan and N. maeander from West Africa, but little more information is
available (Claridge and Morgan, 1987).
12.4 HERBIVORE/PARASITOID FOOD WEBS - GRASS-FEEDING

CHALCID WASPS
It is estimated that between 20% and 25% of all insect species are para-
sitoids of other insects (Godfray, 1994: 16). These are dominated by the
Hymenoptera Parasitica and the large Dipterous family Tachinidae.
Feeding interactions between parasitoid and host species are often spe-
cialized, with several parasitoids attacking the same hosts. Similar prob-
lems concerning difficulties of determining the status of host-associated
populations occur for parasitoids as for herbivores. Because of the diffi-
culties of Hymenoptera taxonomy, there has been a tendency to restrict
species in these groups to morphologically diagnosable entities. Indeed,
the frustration with the often rudimentary state of taxonomic knowledge
has often led ecologists to use artificial minimally recognizable units. For
example, Memmott and Godfray (1993: 229) suggested that morphological
species or 'morphotype' determination rather than detailed biological
species analyses are adequate to determine the structure of parasitoid
dominated food webs. However, such procedures undoubtedly underes-
timate the degree of host specificity and thus of compartmentation in such
Luzon
Ball
1s
Figure 12.8 Oscillograms of sections of individual male calls of Nilaparvata bakeri

from Luzon, Philippines, and Bali, Indonesia. (After Claridge and Morgan, 1993.)
webs. Sometimes the scale of such underestimation may be dramatic
(Claridge and Dawah, 1994; Dawah et at, 1995).
Relatively few studies have been made in an attempt seriously to apply
biological species concepts to groups of parasitoids. However, where it has
been done, morphological species have been shown to consist of two or
more sibling species, usually with more limited host ranges and other sig-
nificant differences in behaviour and life histories (Table 12.3).
A good example is provided by the herbivorous chalcid wasps of the
genus Tetramesa (Eurytomidae) and their mostly also chalcid parasitoids
associated with grasses and cereals (Claridge and Dawah, 1994; Dawah et
al, 1995). Species discrimination in Tetramesa has been based on small,
Table 12.3 Some examples of the application of biological species concepts in

Hymenoptera Parasitica using various techniques, including behavioural,
pheromones, electrophoresis and molecular methods
Species complex Host Reference
Braconidae
Aphidius spp. Various aphid Pungerl, 1986; Unruh et al, 1986;
host Holler, 1991; Castanera et al, 1983
Asobara spp. Drosophilidae Vet and Janse, 1984; Vet et al, 1984
Eulophidae
Pediobius eubis complex Tetramesa sp.
Tetrastichus spp. Crioceris spp.
Eurytomidae
Tetramesa sp. Phytophagous Dawah, 1987
Eurytoma appedigaster Tetramesa sp. Dawah, 1988b
group Eurytoma sp.
Trichogrammatidae
Trichogramma sp. Various Lepidoptera Hung, 1982; Pintureau and
Voegele, 1980; Pinto et al, 1992,
1993; Kostadinov and Pintureau,
1991; Landry et al, 1993
Mymaridae
Anaphes sp. Listronotus spp. Landry et al, 1993
Aphelinidae
Aphytis spp. Diasipididae Khasimuddin and DeBach, 1976
Pteromalidie
Chlorocytus spp. Tetramesa sp. Dawah, 1989
Eurytoma sp.
Spalangia spp. Musca domestica Propp, 1986
Muscidifurax spp. Stomoxys calcitrans Propp, 1986; Assem and
Povel, 1973
Nasonia spp. Muscidae Assem and Werren, 1994
Discussion and conclusions 265
often difficult to appreciate, and variable characters of the adult females
(Walker, 1832; Thomson, 1875; Hedicke, 1920; Phillips, 1920; Phillips and
Poos, 1922; Claridge, 1961; Zerova, 1976). In fact, great emphasis has usu-
ally been placed on host plant records and features of life-histories. It is
only recently that biological species boundaries have been determined
more precisely by mate choice experiments and gel electrophoresis to
determine evidence for lack of gene flow and therefore of reproductive
isolation between supposed species (Dawah, 1987). These studies have
confirmed the previously suspected generally extreme host specificity of
Tetramesa species. Only one species is known to attack host plants from
more than one genus of grasses.
Tetramesa are attacked by a series of characteristic parasitoids dominat-
ed particularly by the Eurytomidae - Eurytoma (Figure 12.9) and Sycophila,
Eulophidae - Pediobius, and Pteromalidae - Chlorocytus and Homopoms. All
of these show extreme host specificity and very limited morphological dif-
ferentiation, as confirmed particularly by enzyme gel electrophoresis
(Dawah 1988a,b, 1989). For Pediobius, Dawah (1988a) was able to identify
nine different biological species in Britain alone on a basis of mate choice
experiments and electrophoresis in the P. eubius (Walker) complex.
Previously these had been variously regarded as either only one (Boucek,
1965) or three (Graham, 1963) species. Once real species limits were deter-
mined, careful and detailed microscopic work revealed very small differ-
ences which make it possible normally to identify dead adult females
(Dawah, 1988a).
Thus, it is clear that the application of biological species approaches to
these groups of Hymenoptera reveals much more significant biological
variation than would otherwise have been suspected. The real structure of
food webs can only be determined following such studies.
12.5 DISCUSSION AND CONCLUSIONS

Most insect herbivores and parasitoids are biparental, sexually reproduc-
ing forms and it is clear that the application of a broadly biological species
concept leads to the recognition of more species than a traditional purely
morphological approach. Such species represent important ecological
entities about which useful generalizations concerning host ranges and
patterns of exploitation may be made. In particular, the widespread occur-
rence of groups of sibling species can only be demonstrated by a biologi-
cal approach.
For insect herbivores and parasitoids in practice, application of either the
biological species or the phylogenetic species (in the sense of Cracraft, 1997:
Chapter 16) will almost always produce the same results, though the exam-
ple of Nilaparvata bakeri illustrates the more precise resolving power of the
biological concept. Only in very well-worked groups is the problem of
266 Species in insect herbivores and pamsitoids
Figure 12.9 Courting male and female of Eurytoma pollux, a parasitoid of Tetmmesa
calamagrostidis in Calamagrostis epigejos.
determining the status of allopatric populations an important one. On the

whole if allopatric populations are clearly diagnosably distinct and, in the
absence of contrary evidence, then it is preferable to give them specific rank.
Our major concern about the phylogenetic species as applied to the
insects under consideration here is a procedural one. The biological
species approach lays emphasis on specific mate recognition systems and
reproductive isolation so that sibling species should be readily detected.
The phylogenetic species approach gives no incentive to expose the exis-
tence of sibling species if no obvious diagnosable differences are apparent
between the populations under investigation.
Thus, in practice, for specialist insect herbivores and parasitoids a bio-
logical species approach has the best potential for revealing species
diversity and is thus to be preferred. However, we see no difficulty in
reconciling the two approaches for these organisms.
Ackno wle dgements

We thank all those colleagues who have helped us in discussion of species
problems in insects. In particular our special thanks are due to Arthur Cain
for reading a draft of this chapter, John Morgan for preparing some of the
References 267
figures, David Windsor for the photograph in Figure 12.9, and Rosemary
Jones for preparing the manuscript.
12.6 REFERENCES
Alphen, J.J.M. van (1980) Aspect of the foraging behaviour of Tetrastichus asparagi
Crawford and Tetrastichus spec. (Eulophidae), gregarious egg parasitoids of the
asparagus beetles Crioceris asparagi L. and C. duodecimpunctata L.
(Chrysomelidae). Netherlands Journal of Zoology, 30, 307-25.
Assem, J. van den and Povel, G.D.E. (1973) Courtship behaviour of some
Muscidifurax species (Hymenoptera: Pteromalidae): a possible example of a
recently evolved ethological isolating mechanism. Netherlands Journal of Zoology,
23,465-87.
Assem, J. van den and Werren, J.H. (1994) A comparison of the courtship and mat-
ing behaviour of three species of Nasonia (Hymenoptera: Pteromalidae). Journal
of Insect Behaviour, 7, 53-66.
Boucek, Z. (1965) Studies of European Eulophidae IV: Pediobius Walk, and two
allied genera (Hymenoptera). Ada Entomologica Musei Nationalis Prague, 36,
5-90.
Bush, G.L. (1975) Modes of animal speciation. Annual Review of Ecology and
Systematics, 6, 339-64.
Bush, G.L. (1993) A reaffirmation of Santa Rosalia, or why are there so many kinds
of small animals? in Evolutionary Patterns and Processes (eds D.R. Lees and D.
Edwards), Academic Press, London, pp. 229-49.
Bush, G.L. (1994) Sympatric speciation. Trends in Ecology and Evolution, 9, 285-8.
Bush, G.L. (1995) Species and speciation. Trends in Ecology and Evolution, 10, 38
Castanera, P., Loxdale, H.D. and Nowak, K. (1983) Electrophoretic study of
enzymes from cereal aphid populations. 11. Use of electrophoresis for identify-
ing aphidiid parasitoids (Hymenoptera) of Sitobion avenae (F.) (Hemiptera:
Aphididae). Bulletin of Entomological Research, 73,659-65.
Claridge, M.F. (1961) A contribution to the biology and taxonomy of some
Palaearctic species of Tetramesa Walker (=Isosoma Vfalk.,—Harmolita
Motsch)(Hymenoptera: Eurytomidae), with particular reference to the British
fauna. Transactions of the Royal Entomological Society of London, 113,175-216.
Claridge, M.F. (1985a) Acoustic signals in the Homoptera: behaviour, taxonomy
and evolution. Annual Review of Entomology, 30,297-317.
Claridge, M.F. (1985b) Acoustic behaviour of Leafhoppers and Planthoppers:
species problems and speciation, in The Leafhoppers and Planthoppers (eds L.R.
Nault and J.G. Rodriquez), John Wiley, New York, pp. 103-25.
Claridge, M.F. (1988) Species concepts and speciation in parasites, in Prospects in
Systematics (ed. D.L. Hawksworth), Clarendon Press, Oxford, pp. 92-111.
Claridge, M.F. and Dawah H.A. (1994) Assemblages of herbivorous chalcid wasps
and their parasitoids associated with grasses-problems of species and specifici-
ty, in Plant Galls (ed. M.A.J. Williams), Clarendon Press, Oxford, pp. 313-28.
Claridge, M.F. and den Hollander, J. (1982) Virulence to rice cultivars and selection
for virulence in populations of the brown planthopper Nilaparvata lugens.
Entomologia experimenatalis et Applicata, 32, 213-21.
Claridge, M.F. and den Hollander, J. (1983) The biotype concept and its application
to insect pests of agriculture. Crop Protection, 2, 85-95.
Claridge, M.F. and de Vrijer, P. (1994) Reproductive behaviour: the role of acoustic
signals in species recognition and speciation, in Planthoppers Their Ecology and
Management (eds R.F. Denno and TJ. Perfect), Chapman & Hall, New York, pp.
216-33.
Claridge, M.F. and Gillham, M. (1992) Variation in populations of leafhoppers and
planthoppers (Auchenorrhyncha): bio types and biological species, in
Ordinations in the study of Morphology, Evolution and Systematics of Insects:
Application and Quantitative Genetic Rationales (eds R. Foottit and J.T. Sorenson),
Elsevier, Amsterdam, pp. 241-59.
Claridge, M.F. and Morgan, J.C. (1987) The Brown Planthopper, Nilaparvata lugens
(Stal), and some related species: a biotaxonomic approach, in Proceedings of the
2nd International Workshop on Leafhoppers and Planthoppers of Economic Importance
(eds M.R. Wilson and L.R. Nault), CAB Institute of Entomology, London, pp.
19-32.
Claridge, M.F. and Morgan, J.C. (1993) Geographical variation in acoustic signals
of the planthopper, Nilaparvata bakeri (Muir), in Asia: species recognition and
sexual selection. Biological Journal of the Linnean Society, 48,267-81.
Claridge, M.F. and Wilson, M.R. (1976) Diversity and distribution patterns of some
mesophyll-feeding leafhoppers of temperate woodland canopy. Ecological
Entomology, 1,231-50.
Claridge, M.F. and Wilson, M.R. (1981) Host plant associations, diversity and
species-area relationships of mesophyll-feeding leafhoppers of trees and
shrubs in Britain. Ecological Entomology, 6, 217-38.
Claridge, M.F. and Wilson, M.R. (1982) Insect herbivore guilds and species-area
relationships: leaf miners on British treas. Ecological Entomology, 7,19-30.
Claridge, M.F., den Hollander, J. and Haslam, D. (1984) The significance of mor-
phometric and fecundity differences between the' biotypes' of the brown plan-
thopper, Nilaparvata lugens. Entomologia Experimentalis et Applicata, 36,107-14.
Claridge, M.F., Den Hollander, J. and Morgan, J.C., (1985a) Variation in courtship
signals and hybridization between geographically definable populations of the
rice brown planthopper Nilaparvata lugens (Stal). Biological Journal of the Linnean
Society, 24, 35-49.
Claridge, M.F., den Hollander, J. and Morgan, J.C. (1985b) The status of weed-
associated populations of the brown planthopper, Nilaparvata /wgens(Stal) - host
race or biological species? Zoological Journal of the Linnean Society, 84, 77-90.
Claridge, M.F., den Hollander, J. and Morgan, J.C. (1988). Variation in hostplant
relations and courtship signals of weed-associated populations of the brown
planthopper, Nilaparvata lugens (Stal), from Australia and Asia: a test of the
recognition species concept. Biological Journal of the Linnean Society, 35, 79-93.
Claridge, M.F., Dawah, H.A. and Wilson, M.R. (1997) Practical approaches to species
concepts for living organisms, in Species: the Units of Biodiversity (eds M.F.
Claridge, H.A. Dawah and M.R. Wilson), Chapman & Hall, London, pp. 1-15.
Cracraft, J. (1997) Species concepts in systematics and conservation-an ornitholog-
ical viewpoint, in Species: the Units of Biodiversity (eds M.F. Claridge, H.A.
Dawah and M.R. Wilson), Chapman & Hall, London, pp. 325-39
Dawah, H.A. (1987) Biological species problems in some Tetramesa (Hymenoptera:
Eurytomidae). Biological Journal of the Linnean Society, 32,237-45.
References 269
Dawah, H.A. (1988a) Taxonomic studies on the Pediobius eubius complex
(Hymenoptera: Chalcidoidea; Eulophidae) in Britain, parasitoids of
Eurytomidae in Gramineae. Journal of Natural History, 22,1147-71.
Dawah, H.A. (1988b) Differentiation between Eurytoma appendigaster group
(Hymenoptera: Eurytomidae) using electrophoretic esterase patterns. Journal of
Applied Entomology, 105,144-8.
Dawah, H.A. (1989) Separation of four Chlorocytus species (Hymenoptera:
Pteromalidae), parasitoids of stem-boring Hymenoptera (Eurytomidae and
Cephidae) using enzyme electrophoresis. The Entomologist, 108,216-22.
Dawah, H.A., Hawkins, B.A. and Claridge, M.F. (1995) Structure of the parasitoid
communities of grass-feeding chalcid wasps. Journal of Animal Ecology, 64,708-20.
Diehl, S.R. and Bush, G.L. (1984) An evolutionary and applied perspective of insect
biotypes. Annual Review of Entomology, 29,471-504.
Eberhardt, W.G. (1985) Sexual Selection and Animal Genitalia, Harvard University
Press, Cambridge, Massachusetts.
Foottit, R. and Sorenson, J.T. (eds) (1992) Ordinations in the Study of Morphology,
Evolution and Systematics of Insects: Application and Quantitative Genetic Rationales.
Elsevier, Amsterdam.
Gillham, M.C. and Claridge, M.F. (1994) A multivariate approach to host plant
associated morphological variation in the polyphagous leafhopper, Alnetoidia
alneti (Dahlbom). Biological Journal of the Linnean Society, 53,127-51.
Godfray, H.C.J. (1994) Parasitoids: Behavioral and Evolutionary Ecology, Princeton
University Press, Princeton, New Jersey.
Graham, M.W.R. de V. (1963) Key to the British genera and species of Elachertinae,
Eulophinae, Eutodontinae and Enderinae (Hymenoptera: Chalcidoidea).
Transactions of the Society for British Entomology, 13,169-204.
Hammond, P.M. (1992) Species inventory, in Global Biodiversity: Status of the Earth's
Living Resources (ed. B. Groombridge), Chapman & Hall, London, pp. 17-39.
Hawksworth, D.L. (ed.) (1994) Identification and Characterization of Pest Organisms,
CAB International, Wallingford.
Hedicke, H. (1920) Beitrage zu eine Monographic der palaarktischen Isosominen
(Hymenoptera: Chalcidoidea). Archives Naturgeschichtes, 86,1-165.
Hendrikse, A. (1979) Activity patterns and sex pheromone specificity as isolating
mechanisms in eight species of Yponomeuta (Lepidoptera: Yponomeutidae).
Entomologia Experimentalis et Applicata, 25,172-80.
Hendrikse, A. (1986) Intra- and interspecific sex-pheromone communication in the
genus Yponomeuta. Physiological Entomology, 11,159-69.
Hendrikse, A., van der Laan, C.E. and Kerkhof, L. (1984) The role of male abdom-
inal brushes in the sexual behaviour of small ermine moths (Yponomeuta Latr.,
Lepidoptera). Mededelingen van de Faculteit Landbouwwetenschappen
Rijksuniversiteit Gent, 49, 719-29.
Holler, C. (1991) Evidence for the existence of a species closely related to the cere-
al aphid parasitoid Aphidius rhopalosiphi De stefani Perez based on host ranges,
morphological characters, isoelectric focusing, banding patterns, cross-breed-
ing experiments, sex pheromones specificities (Hymenoptera: Braconidae:
Aphidiinea). Systematic Entomology, 16,15-28.
Hung, A.C.F. (1982) Chromosome and isozyme studies in Trichogramma
(Hymenoptera: Trichogrammatidea). Proceedings of the Entomological Society of
Washington, 4, 791-6.
Hunt, R.E. (1994) Vibrational signals associated with mating behaviour in the tree
hopper, Enchenopa binotata Say (Hemiptera: Homoptera: Membracidae) Journa
of the New York Entomological Society, 102, 266-70.
Jaenike, L. (1981) Criteria for ascertaining the existence of host races. The American
Naturalist, 117, 830-4.
Jones, P., Gacesa, P. and Butlin, R. (1996) Systematics of brown planthopper anc
related species using nuclear and mitochondrial DNA, in The Ecology o
Agricultural Pests: Biochemical Approaches (ed. W.O.C. Symondson and J.E
Liddell), Chapman & Hall, London, pp. 133-48.
Khasimuddin, S. and DeBach, P. (1976) Biosystematic studies of three allopatrii
population of Aphytis maculicornis (Hymenoptera: Aphelinidae). Entomophaga
21, 81-92.
Kostadinov, D. and Pintureau, B. (1991) A possibility to discriminate females of
three closely related sp. of Trichogramma (Hymenoptera: Trichogrammatidae)
with special purpose analysis of the type T. evanescens Westwood. Annales
Societe Entomologique de France, 27, 393-400.
Landry, B.S., Dextraze, L. and Boivin, G. (1993) Random amplified polymorphic
DNA markers for DNA fingerprinting and genetic-variability assessment of
minute parasitic wasp species (Hymenoptera, Mymaridae and
Trichogrammartidae) used in biological-control programs of phytophagous
insects. Genome, 36, 580-7.
Linnaeus, C. (1759) Systema Naturae, 10th edn, Stockholm.
Loxdale, H.D. and Den Hollander, J. (eds) (1989) Electrophoretic Studies on
Agricultural Pests, Clarendon Press, Oxford.
Mayr, E. (1963) Animal Species and Evolution, Harvard University Press, Cambridge,
Massachusetts.
Memmott, J. and Godfray, H.C.J. (1993) Parasitoids' webs, in Hymenoptera and
Biodiversity (eds J. LaSalle and I.D. Gauld), CAB International, Wallingford, pp.
217-34.
Menken, S.B.J. (1980) Inheritance of allozymes in Yponomeuta. II. Interspecific
crosses within the padellus-complex and reproductive isolation. Proceedings of
the Koninklyke Nederlandse Akademic van Weterschapper, Series C, 83,424-31.
Menken, S.B.J. and Ulenberg, S.A. (1987) Biochemical characters in agricultural
entomology. Agricultural Zoology Reviews, 2, 305-60.
Menken, S.B.J., Herrebout, W.M. and Wiebes, J.T. (1992) Small Ermine moths
(Yponomeuta): their host relations and Evolution. Annual Review of Entomology,
37,41-66.
Nagarkatti, S. and Nagaraja, H. (1971) Redescription of some known species of
Trichogramma (Hymenoptera: Trichogrammatidae) showing the importance of
the male genitalia as a diagnostic character. Bulletin of Entomological Research, 61,
13-31.
Ossiannilsson, F. (1981) The Auchenorrhyncha (Homoptera) of Fennoscandia and
Denmark. Part 2: The families Cicadidae, Membracidae and Cicadellidae (excluding
Deltocephalinae). Fauna Entomologica Scandinavica, 7. Scandinavian Science Press
Ltd, Klampenborg.
References 271
Phelan, P.L. (1992) Evolution of sex pheromones and the role of asymmetric track-
ing, in Insect Chemical Ecology: An Evolutionary Approach (eds B.D. Roitberg and
M.B. Ismam), Chapman & Hall, London, pp. 265-314.
Phillips, WJ. (1920) Studies on the life history and habits of the jointworm flies of
the genus Harmolita (Isosoma), with recommendation for control. United States
Department of Agriculture Bulletin, 808,1-27.
Phillips, W.J. and Poos, F. (1922) Five new species belonging to the genus Harmolita
Motschulsky (Isosoma Walker et auctt.). Kansas University Science Bulletin, 14,
349-59.
Pinto, J.D., Kazmer, D.J., Platner, G.R. and Sassaman, C.A. (1992) Taxonomy of the
Trichogramma minutum complex (Hymenoptera: Trichogrammatidae):
AUozymic variation and its relationship to reproductive and geographic data.
Annals of the Entomological Society of America, 85,413-22.
Pinto, J.D., Platner, G.R. and Sassaman, C.A. (1993) Electrophoretic study of two
closely related sp. of North American Trichogramma: T. Pretiosum and T. deion
(Hymenoptera: Trichogrammatidae). Annals of the Entomological Society of
America, 86, 702-9.
Pintureau, B. and Voegele, J. (1980) Une nouvelle espece proche de Trichogramma
evanescens et de T. maidis (Hymenoptera: Trichogrammatidae). Entomophaga, 25,
431-40.
Price, P.W. (1980) Evolutionary Biology of Parasites, Princeton University, New Jersey.
Propp, G.D. (1986) Characterization of intra- and interspecific isoenzyme in
Spalangia spp. and Muscidifurax spp. (Chalcidoidea: Pteromalidae), parasitoids
of synathropic Diptera, in Biological Control of Muscoid Flies (eds R.S. Patterson
and D.A. Rutz), Miscellaneous Publications of the Entomological Society of
America, No. 61, pp. 164-74.
Pungerl, N.B. (1986) Morphometric and electrophoresis study of Aphidius
(Hymenoptera: Aphidiidae) reared from a variety of Aphid hosts. Systematic
Entomology, 11, 327-54.
Ribaut, H. (1936) Faune de France: Homopteres Auchenorrhynques, 1: (Typhlocybidae).
Paul Lechevalier, Paris.
Saxena, R.C. and Barrion, A.A. (1985) Biotypes of the brown planthopper
Nilaparvata lugens (Stal) and strategies in deployment of host plant resistance.
Insect Science and its Application, 6, 271-89.
Saxena, R.C. and Rueda, L.M. (1982) Morphological variation among three bio-
types of the brown planthopper Nilaparvata lugens in the Philippines. Insect
Science and its Application, 6,193-210.
Saxena, R.C., Velasco, M.V. and Barrion, A.A. (1983) Morphological variations
between brown planthopper biotypes on Leersia hexandra and rice in the
Philippines. International Rice Research Newsletter, 18, 3.
Strong, D.R., Lawton, J.H. and Southwood, R. (1984) Insects on Plants: Community
Patterns and Mechanisms, Blackwell, Oxford.
Symondson, W.O.C. and Liddell, J.E. (eds) (1996) The Ecology of Agricultural Pests:
Biochemical Approaches, Chapman & Hall, London.
Thomson, C.G. (1875) Hymenoptera Scandinaviae, 4 (1). Lund, Sweden.
Thorpe, W.H. (1929) Biological races in Hyponomeuta padella L. Zoological Journal of
the Linnean Society, 36, 621-34.
Thorpe, W.H. (1930) Biological races in insects and allied groups. Biological Reviews,
5,177-212.
Unruh, T.R., White, W., Gonzalez, D. and Luck, R.F. (1986) Electrophoresis studies
of parasitic Hymenoptera and implications for biological control, in Biological
Control of muscoid flies (eds R.S. Patterson and D.A. Rutz), Miscellaneous
Publications of the Entomological Society of America, No. 61, pp. 150-63.
Vet, L.E.M. and Janse, C.J. (1984) Fitness of two sibling species of Asobara
(Braconodae: Alysiinae), larval parasitoids of Drosophilidae in different micro-
habitats. Ecological Entomology, 9, 345-54.
Vet, L.E.M., Janse, C, Achterberg, C. van and Alphen, J.J.M. van (1984)
Microhabitat location and niche segregation in two sibling species of
Drosophilidae parasitoids: Asobara tabida (Nees) and A. rufescens (Foerster)
(Braconidae: Alysiinae). Oecologia, 61,182-8.
Walker, F. (1832) Monographia Chalcidum. Entomological Magazine, 1,12-29.
Walsh, B.J. (1864) On phytophagic varieties and phytophagic species. Proceedings
of the Entomological Society of Philadelphia, 3,403-30.
Wilson, M.R. and Claridge, M.F. (1991) Handbook for the Identification of Leafhoppers
and Planthoppers of Rice, CAB International, Wallingford.
Wood, T. K. (1993) Speciation of the Enchenopa binotata complex (Insecta:
Homoptera: Membracidae), in Evolutionary Patterns and Processes (eds D.R. Lees
and D. Edwards), Academic Press, London, pp. 289-317.
Zerova, M.D. (1976) Hymenoptera Chalcidiodea family Eurytomidae, Subfamilies
Rileyinae and Harmolitinae. (in Russian). Fauna SSSR, 7,1-23, Leningrad.
13
The species concept in blood-
sucking vectors of human diseases
R. Lane
Contacting address: Department of Entomology, The Natural History Museum, Cromwell
ABSTRACT
Blood-sucking insects have attracted much attention from biologists
because some species are vectors of devastating human and animal
diseases. The need to determine which species transmit pathogens or
parasites in any one location has put great demands on systematics on
the one hand, but made considerable resources available on the other.
The classical biological species concept underpins the systematics
of these insects, although most of the 14 500 or so species of blood-
sucking insect are still defined on morphological criteria alone, i.e. dis-
continuities in morphological variation.
Species complexes of morphologically indistinguishable but bio-
logically distinct and often sympatric species have been discovered in
mosquitoes, simuliids and sandflies but not yet in other groups.
Surprisingly, there is less proof than might be expected for the exis-
tence of species complexes from the obvious route - experimental
genetics. Much evidence is inferential rather than experimentally
proved and has necessitated the use of many characters and tests of
reproductive distinctness. Chromosome markers, isozymes and labo-
ratory cross-mating tests are the most commonly used criteria for
establishing the specific status of taxa in species complexes. The chem-
istry of sex pheromones has even been used in one sandfly complex.
The techniques of molecular biology are only just beginning to
make an impact on defining species although DNA probes have been
very successful in the identification of wild-caught mosquito sibling
species.
274 The species concept in blood-sucking vectors of human diseases
13.1 INTRODUCTION
Insects are the most species-rich group of organisms, with more species
known already than all other organisms combined. However, within this
vast group only a relatively few - some 14 500 described species - are
members of families which suck mammalian blood and are therefore
potential vectors of parasites and pathogens to humans. These blood-
sucking insects do not form a 'natural' but a polyphyletic assemblage shar-
ing one major life-history characteristic - feeding on vertebrate blood.
Blood-feeding on homiotherms is found throughout the Insecta, but it
occurs mainly in the Diptera (Lane and Crosskey, 1993).
Although, as a group, haematophagous insects are structurally diverse,
they are not as genetically diverse as the parasites they transmit: viruses,
bacteria, spirochetes, rickettsia, protozoa and nematodes. What they lack
in relative diversity they make up in sheer numbers, so that the number
of proven or suspected vector species is always much greater than the
number of parasite species they transmit; for example, there are four
human malaria parasites (Plasmodium spp.) but more than 70 species of
Anopheles are implicated in their transmission. As a general principle, one
species of parasite can be transmitted by several vector species but rarely
the other way around. This relationship between the number of parasites
and vectors is one reason why there is a considerable difference in the
practical criteria used to define species.
An important feature distinguishing the systematics of blood-sucking
insects from most other insect groups is the considerable interest it attracts
from non specialists - interest shown because of the insects' ability to
transmit disease. It is often recognized that accurate delimitation and sub-
sequent identification of species is essential for effective vector control.
Perhaps this recognition is not as often as the systematists would like, but
in comparison with many other areas of applied entomology the impor-
tance of systematics has become axiomatic for effective vector control. To
this end, the development of biting-insect taxonomy has been driven by a
quest to explore the 'structure' of species at an ever-increasing resolution
and this has had an effect not only on the theoretical concepts of species
but on the practical application of these ideas.
13.2 THEORETICAL SPECIES CONCEPTS

Several species concepts are currently debated in the systematics litera-
ture, although relatively little debate on the broader theoretical aspects
has taken place in the medical or veterinary entomology literature. The
most well-known of the modern species concepts, is the biological species
concept attributed to Dobzhansky (1937) subsequently promulgated by
Mayr (1966) in which species are considered 'groups of actually or poten-
Theoretical species concepts 275
tially interbreeding natural populations which are reproductively isolated
from other such groups'. The concept is biological because the means of
differentiating species is biological (reproductive ability) in contrast to the
earlier concepts based on intrinsic or essential qualities used by Linnaeus
and his contemporaries.
There are a number of problems, mainly operational but some theoret-
ical, associated with the application of the biological species concept and
this has generated a number of alternative concepts. One of the earliest
was the evolutionary species concept of Simpson (1951) in which species
are 'a lineage evolving separately from others and with its own unitary
evolutionary role and tendencies'. This idea emphasized the notion that
populations rather than whole species evolve, as is generally accepted, but
there are significant difficulties in the practical application of this concept.
In response to the emphasis on isolating mechanisms inherent in the BSP,
Paterson (1985) focused attention on the unifying reproductive elements
of species and therefore in his concept species share a common mate
recognition system. It is of note that Paterson is one of the few medical
entomologists to have engaged seriously in the debate on species con-
cepts. While the differences between species in mate recognition mecha-
nisms are known and can be characterized accurately in a few species, for
example, species-specific sex pheromones of tsetse flies (Carlson et al.,
1984) or sandflies (Ward et al, 1988; Ward, 1989), these characters have not
been used primarily as the criteria for recognizing species. In the examples
given species were defined originally by morphological discontinuities or
cross-mating studies rather than by the mate recognition signals.
One of the most recent species concepts to be introduced has been the
phylogenetic or diagnostic species concept (Cracraft, 1983) in which
species have been defined as the 'smallest diagnosable cluster of individ-
ual organisms within which there is a parental pattern of ancestry and
decent'. While this concept mirrors the pragmatic emphasis many taxon-
omists have on detecting the features which diagnose species and is use-
ful for those primarily interested in resolving historical relationships, it
has been rejected by others because of its lack of theoretical depth. This
concept has not been used yet in an overt manner in medical entomology.
Although not always explicitly stated, the biological species concept of
Mayr and Dobzhansky underlies the vast majority of studies in blood-
sucking insect systematics. However, the criteria used by systematists to
distinguish most species are substantially different from those that directly
test adherence to such a theoretical concept (Lane and Crosskey, 1993). As
in most groups of organisms, very few species of blood-sucking insects
have been tested in relation to this or any other theoretical species concept.
Even though all species reproduce sexually, parthenogenicity is known
only in the bat bugs (Polyctenidae), thus making it at least theoretically
possible to test the degree of reproductive distinctness between taxa.
While recognizing its weaknesses, the strength of the biological species
concept (BSP) for many biologists interested in biting insects is that it is
closely tied to the process of speciation and to genetic differences between
populations. These genetic differences, including differences in parasite or
pathogen susceptibility, or host preferences are important factors in
assessing the vectorial competence of species, or even in the vectorial
capacity (a quantitative measure of disease transmission; Dye, 1990). For
example, in the mosquito Aedes aegypti, vector competence varies between
mosquito populations for yellow fever and dengue viruses and hence the
transmission of these pathogens (Gubler et al., 1982; Tabachnick, 1991).
There are many factors which influence the ability of a vector to transmit
a parasite or pathogen and often there is ambiguity over exactly what is
meant by a vector. On one hand it represents vector competence (the
physiological ability to transmit) and on the other vectorial capacity (the
epidemiological ability to transmit). The above examples notwithstanding,
it is not surprising that given the complexity of transmission dynamics
there are few really good data on this aspect for many of the important
vector complexes, including the Anopheles gambiae complex.
13.3 SPECIES COMPLEXES

If there is one characteristic of biting-insect systematics that stands out
from the rest of insect systematics it is the notion of species complexes.
Species complexes were originally discovered when ecological inconsis-
tencies were observed in the field between populations of the same mor-
phologically defined species. They have received much attention in the
medical entomology literature because several major vectors of human
disease are members of species complexes and a full understanding of the
variation of parasites, vectors and the human population is essential to the
correct interpretation of the epidemiology of disease; this is especially the
case for malaria and onchocerciasis in Africa (Wright and Pal, 1967; World
Health Organization, 1977; Steiner et al, 1982; Service, 1988).
Species complexes consist of sibling species, which are defined as mor-
phologically indistinguishable but reproductively isolated species and
frequently are sympatric. Among insects, sibling species are particularly
common in the Diptera of medical importance (Simuliidae, Culicidae) but
they are not confined to the group, or even to insects (Knowlton, 1993).
Once it became known that species complexes existed, the genetic structure
of many morphospecies (species defined on morphological criteria only)
were examined and further complexes discovered over the past 25 years.
Sibling species are, by definition, isomorphic or virtually so, and other
means than morphology have therefore been used to detect the existence
of a complex. In practice, this usually means using cytological attributes of
chromosomes (cytotaxonomy) or isozyme variation (chemotaxonomy).
Species complexes 277
The Diptera are cytologically unusual in that they have fewer pairs of
chromosomes than is usual in insects and often have massive polytenized
chromosomes, of a type unknown in other insect orders and formed by
lengthwise multiplication of DNA. Polytene chromosomes are often most
developed (or at least easy to visualize) in cells undergoing high levels of
protein synthesis, e.g. adult ovarian nurse cells in mosquitoes and larval
salivary glands in blackflies. When stained they show conspicuous bands
and other micromorphological landmarks characteristic of each species.
Although these banding sequences can be used to reconstruct phyloge-
nies (cytophylogenies), in practice their prime significance in systematics
is to detect speciation events within a morphospecies. The term
'cytospecies' is frequently used for species, usually sibling species within a
complex, and is recognized primarily on chromosomal characters. A cyto-
type or cytoform is a chromosomally recognized constituent of a mor-
phospecies whose status is still uncertain: it might be a sibling species or
only a polymorphic variant. The principal chromosomal differences found
between sibling species (and for other species) include fixed sequential
differences in inversions and interchanges within populations, sex chro-
mosome differences and polymorphic (floating) inversions in which alter-
native banding sequences are seen in one of a complementary chromo-
some pair. If the putative species being compared are sympatric the
absence of hybrids is prima facie evidence for the absence of cross-mating.
Similarly, in isozyme studies an absence of hybrids or a deficiency in
the proportions of heterozygotes expected by the Hardy-Weinberg
Equilibrium indicates the absence of gene flow. The analysis of isozymes,
with its very direct genetic interpretation, has been particularly useful in
those vector groups (Culex and Aedes mosquitoes, tsetse flies, phle-
botomine sandflies and ticks) which do not have well-developed polytene
chromosomes (Miles and Paterson, 1979; Hunt and Hilburn, 1985; Caillard
et al, 1986; Munstermann, 1988; Kreutzer et al, 1990; Lanzaro et al, 1993).
Tabachnick and Black (1995) have argued that the tremendous interest in
recognizing sibling species might well lead to a revival of the old typologi-
cal concept of species. Given the broad biological framework in which many
systematic studies are now made this seems unlikely; however, the risk of
narrow-mindedness remains, as it does for any approach to systematics.
It is important to stress that sibling species within species complexes are
not conceptually different from other species - they are only different on
operational grounds, i.e. in the sort of characters used to distinguish them.
Species complexes are an anthropocentric concept in which human per-
ceptions of readily observable morphological variation are used. Sibling
species are frequently presumed to be more recently evolved than mor-
phologically more distinct species but this need not necessarily be the
case, morphological distinctness might well be a function of the adaptive
significance of the features being examined. However, this caveat
notwithstanding, it is likely that species complexes are indeed groups of
species of recent, common ancestry. They are probably still in the process
of evolving into separate species by reduction of gene flow between pop-
ulations through both intrinsic and extrinsic mechanisms. If members of
species complexes in particular are currently separating into new species
it is probably impossible to determine which of the taxa recognized are
species in an unambiguous way. There is, theoretically at least, no funda-
mental difference between sibling species and morphologically discrete
species in this respect; it is only a matter of differences in the nature of the
characters being assessed and the population scale at which many of these
sibling species studies are made.
One important aspect in using genetic data to define species is that
there are no hard and fast rules on the minimum threshold for how much,
or little, gene flow occurs (or is permitted in the definition) between
species. The crux of the issue is estimating gene flow - the genetic varia-
tion within a species should be considerably less than the variation
between species. Interestingly, this is the genetic analogy of the pheneti-
cist's approach inherent in multiple discriminant function analysis.
13.4 MOLECULAR DATA AND SPECIES CONCEPTS

All reliable taxonomic characters are genetically controlled (environmen-
tally induced variation being of little taxonomic use) and therefore DNA
studies should provide some of the most important evidence for species
recognition. However, molecular techniques have only just begun to have
an impact on recognizing and identifying sibling species, and for deter-
mining genetic relationships between species.
Much of the impetus for the development of molecular tools has been
the need to find a simple and effective means of identifying field-caught
samples using DNA probes (Gale and Crampton, 1988; Hill and
Crampton, 1994 for Anopheles; Post and Crampton, 1988; Brockhouse et al.,
1993 for Simulium, Ready et al., 1988 for Phlebotomus; Adamson et al., 1991;
Ready et al., 1991 for Lutzomyia) or the polymerase chain reaction (PCR)
(Paskewitz and Collins, 1990; Adamson et al., 1993; Paskewitz et al, 1993;
Scott et al., 1993). In the development of DNA probes, most studies do not
use sequences of known function, and certainly not genes which are
known to be involved in species differentiation. Thus, in many cases, ran-
dom discriminating sequences are used which give rather limited infor-
mation on the relatedness of taxa (including populations) by their patterns
of hybridization to probes. If they contain sequences that are present in
both sexes, DNA probes will identify all stages, adults and immatures,
which is a considerable advantage over other techniques.
Nomenclature and species concepts 279
These tools are particularly useful where the currently used techniques
for identifying species are technically limiting (e.g. isozyme electrophore-
sis requires material frozen from the field), or the technique is confined to
one life-stage only (e.g. chromosomes from the salivary glands of mature
larvae in Simulium). Adult females have polytenized chromosomes but
they are frequently difficult to interpret and are often only accessible to
analysis within a few hours of a blood meal.
To date, relatively few studies have used DNA methods to explore the
genetic variation within species and their constituent populations
although the potential is enormous. The deployment of the Randomly
Amplified Polymorphic DNA techniques (Kambhamptai et al., 1992;
Adamson et al., 1993; Wilkerson et al., 1993; Favia et al, 1994) is likely to be
particularly useful in this respect.
Some molecular studies have used sequence data, usually of mitochon-
drial DNA in reconstructing phylogenies of blood-sucking insects, which
often confirm hypotheses based on morphological data (Xiong and
Kocher, 1991).
It is unlikely that the introduction of molecular techniques will sub-
stantially alter our theoretical concepts of species but they will greatly
enhance our means of exploring the nature of species.
13.5 NOMENCLATURE AND SPECIES CONCEPTS

One complication - but operationally a quite different aspect of studying
species at fine levels of resolution - is nomenclature, or the naming of taxa.
At present, the binomial system of Linneaus is used for naming an organ-
ism in which the basic assumption is made that all individuals belong to
discrete and mutually exclusive sets. Even though Linneaus and his con-
temporaries were essentialists it has been possible for most cases to recon-
cile this now outmoded philosophical standpoint and the pragmatic rules
of nomenclature with modern biology.
However, the fundamental problem with the binomial system is that it
does not take account of populations (or species) in the process of specia-
tion; many populations might not be unambiguously assignable to one
species or another simply because the two species are not yet distinct. In
attempting to come to terms with the conflict between the needs of com-
munication which nomenclature seeks to serve on the one hand, and bio-
logical reality on the other, many taxonomists have effectively dropped
out of the binomial system by developing an informal nomenclature using
categories particular to the taxonomic group being studied, e.g. biotypes,
populations, cytoforms. The relationship between some of these informal
terms and formal nomenclatural categories are given in Lane and
Marshall (1981).
The essential need to have a unified and universally agreed set of
nomenclatural rules which conflict at times with biological reality is a
pragmatic compromise, but unfortunately it adds further confusion to the
complicated debate on species concepts. For some viewpoints at least, the
discussion of complicated issues is not possible without the appropriate
language.
13.6 PRACTICAL SPECIES CONCEPTS

While there are potentially several theoretical species concepts which
could be applied to the study of blood-sucking insects, in reality there is
only one regularly used - the biological species concept. As in other areas
of systematics, there is a substantial gap between the a theoretical model
followed by a practising systematist and the criteria that the systematist
uses on a day-to-day basis for recognizing and defining species.
Before embarking on a review of the way in which species concepts are
approached in blood-sucking insects, it is important to distinguish
between the systematics of a group as a whole and that of the taxa con-
taining vector species - the level of sophistication can be very different.
For example, contrary to popular belief the higher classification of the
mosquitoes is not well resolved, or even of the genus Anopheles (Harbach,
1994) but some species groups containing important vector species are
very well studied, e.g. Anopheles gambiae complex.
To review the working concepts used in the systematics of medically
important insects the following scheme has been developed in which
practical concepts have been arranged in a sequence of levels of increas-
ing resolution and closeness to testing the Biological Species Concept
(Table 13.1). It is important to note that all the techniques and criteria used
to distinguish species give relative data, i.e. non-absolute data, and in this
way they mirror the relational nature of the biological species concept.
The sequence of criteria for recognizing species given below, do not
necessarily reflect the chronological order in which studies are made; i.e.
systematists do not work up the levels in an orderly way. However, level
1, in which qualitative morphological data are used, always precedes the
others. For example, in the early studies on the Anopheles gambiae complex
the primary method of recognizing constituent sibling species was by
cross-mating tests (Davidson, 1964). This labour-demanding technique
was the impetus to discover new methods of discriminating species such
as analysing the banding patterns on chromosomes. The various contri-
butions to Lane and Crosskey (1993) give an overview of the characters
used to distinguish taxa in the various groups of medically important
insects and arachnids.
The criteria levels for recognizing species do not simply describe an
increasing sophistication in the technology or characters used, although
superficially it looks that way. The techniques employed are a consequence
Table 13.1 Criteria for recognizing species of blood-sucking insects
Total Level 1 Level 2 Level 3 Level 4 Level 5 Level 6 Species Species described on
no. of Morphology Morphology Non- Genetic Cross- Measuring complexes non-morphological
species qualitative quantitative morphological mating gene-flow known characters
Culicidae 3450 • • • • • • • •
(mosquitoes)
Simuliidae 1580 • • • • • •
(blackflies)
Phlebotominae 700 • • • AJ AJ AJ
(sandflies)
Ceratopogonidae 1400 • • AJ
(biting midgs) (5000 total)

Tabanidae 4000 • • *J
(horse-flies)
Glossinidae 23 • • *J
(tsetse-flies)
Blood-sucking muscids 50 •
(stable-flies) (5000)
Triatomine bugs 118 • AJ
Cimicidae 91 •
(bed bugs)
Anoplura 490 • • *J
(sucking lice)
Siphonaptera 2500 •
(fleas)
of attempting to base conclusions on ever-more biologically informative
characters. Often, experimental design is more important than the sophis-
tication of the techniques used in delimiting and identifying species (Lane,
1994).
13.6.1 Criteria for recognizing species

Level 1: Qualitative morphology
Like most insects, the majority of blood-sucking insects are distinguished
at the species level by morphological means, i.e. by recognizing disconti-
nuities in the variation of morphological characters. Many discontinuities
can be very overt, such as the patterns on tabanid wings, patterns on mos-
quito legs and wings or variations in the shape of parts of the genitalia. In
this context, morphology is being used as a surrogate for genetic differ-
ences and in the vast majority of cases gives quite acceptable results.
Sometimes, special techniques are required to prepare the specimens
for examination, especially for microscopical characters such as sensilla or
internal structures, or the conditions for examination need to be specified,
e.g. the patterns on the thorax of Simulium adults are generated by the
reflectance from microtrichia arrays and the direction of the incident light
greatly affects the pattern (Lowry and Shelley, 1990).
Level 2: Quantitative morphology

Often, the need to resolve species at a finer level (perhaps based on eco-
logical or geographical evidence) has meant that there is a need to look for
very fine quantitative differences between species. Discontinuities are still
being sought, as at Level 1, but samples are extensive and the concept of
a population, at least in the statistical sense, is introduced. There are
numerous examples of this approach to defining species scattered
throughout the medically important groups (Busvine, 1978; Lane and
Ready, 1985). At this level there are no experimental or genetic data to
confirm these working concepts of species beyond maintaining biological
homogeneity of samples.
Level 3: Non-morphological data

The use of non-morphological data, usually chemical data, to replace mor-
phological characters is usually used where the evidence from morpholo-
gy is ambiguous or it conflicts with field observations. Often there is no
underlying theoretical reason to justify the use of these characters or why
chemical data should be intrinsically more predictive that morphology.
The majority of the occasions on which cuticular hydrocarbons have been
used would fit this description (Carlson and Service, 1979; Hamilton and
Practical species concepts 283
Service, 1983; Milligan et al., 1986; Kamhawi et al., 1992), although it is now
known that some of these hydrocarbons are contact pheromones (Carlson
et al., 1984). If these pheromones are in fact important factors in pre-
mating mate choice (whether defined in terms of a mating barrier in the
Mayr/Dobzhansky concept, or common mate recognition systems in the
Paterson concept) they could be much more effective in defining species
than the whole-body extracts accompanied by multivariate analysis that is
the usual approach to the use of hydrocarbons. In other taxa where cutic-
ular hydrocarbons have been used, in phlebotomine sandflies, simuliids,
and mosquitoes there is no such underlying evidence for their species-
specific function.
The recent use of the randomly amplified polymorphic DNA (RAPD)
technique (Adamson et al., 1993; Wilkerson et al., 1993) would fit into this
category since the genetic basis for these characters is not yet known,
although for RAPDs most are thought to be dominant. It is likely that with
the rapid development of this technique a much clearer understanding of
its genetic significance will be forthcoming and therefore its power
increased.
Level 4: Field data to test genetic models

Studies of this type involve sampling from the field and laboratory analy-
sis to test differences according to a predetermined genetic model. This is
the level at which most species complexes have been investigated, some
classic studies being those of Davidson and Hunt (1973), Mahon et al.
(1976), Miles (1979) and Coluzzi et al. (1985) on An. gambiae and Vajime and
Dunbar (1975) on Simulium.
The usual model is the Hardy-Weinberg Equilibrium to test for pan-
mixis (random mating). Data can be from either isozymes or frequencies
of chromosomal inversions, although it is usually the former. Usually,
cross-sectional data collected from several sites or localities are used, with
the sample pooled within a locality (rather than collected in each locality
over a period of time and the changing gene frequencies measured)
(Green et al, 1992).
Frequently, this approach is used where there are difficulties in assess-
ing the significance of morphological variation (e.g. in Lutzomyia yucumen-
sis, Caillard et al., 1986; or Anopheles culicifaecies, Green and Miles, 1980).
There are substantial theoretical difficulties in testing the differences in
gene frequencies between allopatric populations since the geographical
separation invalidates the prerequisite of free access between individuals.
Level 5: Experimental tests

At this level, data are generated, usually in the form of cross-mating stud-
ies to indicate reproductive distinctness between populations or species
(Davidson, 1964; Davidson and Hunt, 1973; Ward et al, 1988). This usually
requires colonization of the samples - a major constraint both biologically
and in the resources required. These experiments, while appearing to be
the 'gold standard', can be difficult to interpret - some species will not
mate in laboratory cages (stenogamy) or the offspring cannot be reared eas-
ily to determine Fj sterility. The usual method for conducting these exper-
iments is to compare the fertility of reciprocal crosses to within-sample
crosses. Usually, the insects are given the choice between an experimental
insect or nothing, rather than an experimental mate and a putative con-
specific mate. Sometimes these experiments produce crossing artefacts not
present, or rarely found in the field and therefore they express what is
physiologically possible rather than what happens in nature.
Level 6: Measurement of gene flow

This is the level at which species can be most comprehensively tested for
their integrity and adherence to the BSC. It requires detailed genetic data
from extensive field studies. Some of the most comprehensive studies
have been those of Coluzzi and colleagues on Anopheles gambiae in Africa,
in which both geographic and seasonal shifts in gene frequencies have
been found (Coluzzi et al, 1979), and other tropical species such as the
Anopheles dirus complex (Green et al., 1992). The gene flow between popu-
lations in relation to dispersal has been investigated by several studies, e.g.
Tabachnick (1991) foiAedes aegypti, Munstermann (1985) for Ae. triseriatus,
Narang et al. (1991) for An. albimanus, Cheng et al. (1982) for Culex pipiens,
and Tabachnick (1992) for Culicoides variipennis.
13.7 IDENTIFICATION - A REVERSE LOOK AT SPECIES

It is essential to distinguish between initial recognition of a species and the
subsequent identification of samples. Often the same characters are used
for both these functions. However, in the case of many cryptic or sibling
species, much more detailed data might be used to establish the status of
species and then simpler, more economical methods developed for subse-
quent identification are found. For example, recognition of species by
chromosomes or electrophoresis and the subsequent identification by
morphometrics (Wilson et al., 1993), establishing species status by breed-
ing then identification by simple morphology (Gebre-Michael and Lane,
1993), defining species by isozymes and chromosomes then identification
by DNA probes (Post and Crampton, 1988).
13.8 CONCLUSIONS
The pressure from non-systematists to have an ever-more refined classifi-
cation of blood-sucking insects which is both a succinct summary of the
References 285
diversity of these insects as well as providing a predictive framework, has
meant that a variety of techniques have been deployed to understand the
nature of species. In the future, the ability to measure gene flow in the
field with molecular tools to a greater degree of sensitivity and conve-
nience than is currently available will undoubtedly push the theoretical
concepts to the limit. Whether the biological species concept, the most
widely used concept in medical entomology, will withstand this
onslaught remains to be seen.
13.9 REFERENCES
Adamson, R.E., Chance, M.L., Ward. R.D., Feliciangeli, D. and Maingon, R.D.C.
(1991) Molecular approaches to the analysis of sympatric sandfly populations in
endemic areas of western Venezuela. Parassitologia, 33,45-53.
Adamson, R.E., Ward, R.D., Feliciangeli, D. and Maingon, R. (1993) The application
of random amplified polymorphic DNA for sandfly species identification.
Medical and Veterinary Entomology, 7,1-5.
Brockhouse,C.L., Vajime, C.G., Marin, R. and Tanguay, R.M. (1993) Molecular
identification of onchocerciasis vector sibling species in black flies (Diptera:
Simuliidae). Biochemical and Biophysical Research Communications, 194,628-34.
Busvine, J.R. (1978) Evidence from double infestations for the specific status of
human head and body lice (Anoplura). Systematic Entomology, 3,1-8.
Caillard, T., Tibayrenc, M., Le Pont, P., Dujardin, J-P., Desjeux, P. and Ayala, F.J.
(1986) Diagnosis by isoenzyme methods of two cryptic species, Psychodopygus
carrerai and P. yucumensis (Diptera: Psychodidae). Journal of Medical Entomology,
23, 489-92.
Carlson, D.A. and Service, M.W. (1979) Differentiation between species of the
Anopheles gambiae Giles complex (Diptera: Culicidae) by analysis of cuticular
hydrocarbons. Annals of Tropical Medicine and Parasitology, 73, 589-92.
Carlson, D.A., Nelson, D.R., Langley, P.A., Coates, T.W., Davis, T.L. and
Leegwater-van der Linden, M.E. (1984) Contact sex pheromones in the tsetse
fly Glossina pallidipes (Austen): identification and synthesis. Journal of Chemical
Ecology, 10, 429-50.
Cheng, M.L., Racket, C.S., Pryor, S.C., Ferrell, R.E. and Kitto, G.B. (1982) The eco-
logical genetics of the Culex pipiens complex in North America, in Recent
Developments in the Genetics of Insect Disease Vectors (eds W.W.M. Steiner, W.J.
Tabachnick, K.S. Rai and S. Narang), A Symposium Proceedings, Stipes,
Champaign, Illinois, pp. 581-627.
Coluzzi, J., Petrarca, V. and Di Deco, M.H. (1985) Chromosomal inversion intergra-
dation and incipient speciation in Anopheles gambiae. Bolletino di Zoologica, 52,
45-63.
Coluzzi, M., Sabatini, A., Petrarea, V. and Di Deco, M.A. (1979) Chromosomal differ-
entiation and adaptation to human environments in the Anopheles gambiae com-
plex. Transactions of the Royal Society of Tropical Medicine and Hygiene, 73,483-97.
vol. 1, Plenum Press, New York, pp. 159-87.
Davidson, G. (1964) The five mating types of the Anopheles gambiae complex. Rivista
Malariologica, 13,167-218.
Davidson, G. and Hunt, R.H. (1973) The crossing and chromosome characteristics
of a sixth species in the Anopheles gambiae complex. Parassitologia, 15,121-8.
Dobzhansky, T. (1937) Genetics and the Origin of Species, Colombia University Press,
New York.
Dye, C. (1990) Epidemiological significance of vector parasite interactions.
Parasitology, 101,409-15.
Favia, G., Dimopoulos, G. and Louis, C. (1994) Analysis of the Anopheles gambiae
genome using RAPD markers. Insect Molecular Biology, 3,149-57.
Gale, K. and Crampton, J.M. (1988) DNA probes for species identification of mos-
quitoes in the Anopheles gambiae complex. Medical and Veterinary Entomology, 1,
127-36.
Gebre-Michael, T. and Lane, R.P. (1993) Distinguishing the females of Phlebotomus
(Synphlebotomus) martini and P. (S) celiae (Diptera: Phlebotominae), vectors of
visceral leishmaniasis in southern Ethiopia. Bulletin of Entomological Research, 83,
353-60.
Green, C.A. and Miles, S.J. (1980) Chromosomal evidence for sibling species of the
malaria vector Anopheles (Cellia) culicifaecies. Journal of Tropical Medicine and
Hygiene, 83, 75-8.
Green, C.A., Munstermann, L.E., Tan, S.G., Panyim, S. and Baimai, V. (1992)
Population genetic evidence for species A, B, C and D of the Anopheles dims
complex in Thailand and electromorphs for their identification. Medical and
Veterinary Entomology, 6,29-36.
Gubler, D.J., Novak, R. and Mitchell, CJ. (1982) Arthropod vector competence -
epidemiological, cytogenetic and biological considerations, in Recent
Developments in the Genetics of Insect Disease Vectors (eds W.W.M. Steiner, W.J.
Tabachnick, K.S. Rai and S. Narang), A Symposium Proceedings, Stipes,
Champaign, Illinois, pp. 343-78.
Hamilton, R.J. and Service, M.W. (1983) The value of cuticular and internal hydro-
carbons for the identification of larvae of Anopheles gambiae Giles, An. arabiensis
Patton and An. melas Theobald. Annals of Tropical Medicine and Parasitology, 77,
203-10.
Harbach, R. (1994) Review of the internal classification of the genus Anopheles
(Diptera: Culicidae): the foundation for comparative systematics and phyloge-
netics research. Bulletin of Entomological Research, 84, 331-42.
Hill, S.M. and Crampton, J.M. (1994) Synthetic DNA probes to identify members of
the Anopheles gambiae complex and to distinguish the two major vectors of
malaria within the complex, An. gambiae s.s. and An. arabiensis. American Journal
of Tropical Medicine and Hygiene, 50, 312-21.
Hunt, L.M. and Hilburn, L.R. (1985) Biochemical differentiation between species of
tick (Acari: Ixodidae). Annals of Tropical Medicine and Parasitology, 79, 525-32.
Kambhamptai, S., Black, W.C. and Rai, K.S. (1992) Random amplified polymorphic
DNA of mosquito species and populations (Diptera: Culicidae). Techniques,
statistical analysis and applications. Journal of Medical Entomology, 29,939-45.
Kamhawi, S., Lane, R.P., Cameron, M.M., Phillips, A., Milligan, P. and Molyneux,
D. (1992) Variation in the cuticular hydrocarbons of Phlebotomus argentipes from
two field populations and a subsequent laboratory colony. Bulletin of
Entomological Research, 82,209-12.
References 287
Kreutzer, R.D., Palau, M.T., Morales, A., Ferro, C, Feliciangeli, D. and Young, D.G.
(1990) Genetic relationships among phlebotomine sandflies (Diptera:
Psychodidae) in the verrucarum series. Journal of Medical Entomology, 27,1-8.
Lane, R.P. (1994) Morphometric discrimination of pest species: experimental
design, in Identification and Characterisation of Pest Organisms (ed. D.L.
Hawksworth), CAB International and Systematics Association, pp. 195-205.
Lane, R.P. and Crosskey, R.W. (eds) (1993) Medical Insects and Arachnids, Chapman
& Hall, London.
Lane, R.P. and Marshall, J. (1981) Geographic variation, races and subspecies, in
The Evolving Biosphere (ed. P. Forey), British Museum (Natural History),
London, pp. 9-19.
Lane, R.P. and Ready, P.O. (1985) Multivariate discrimination between Lutzomyia
wellcomei, a vector of mucocutaneous leishmaniasis, and Lu. complexus. Annals of
Tropical Medicine and Parasitology, 78,469-72.
Lanzaro, G.C., Ostrovska, K., Herrero, M.V., Lawyer, P.G. and Warburg. A. (1993)
Lutzomyia longipalpis is a species complex: genetic divergence and interspecific
sterility among three populations. American Journal of Tropical Medicine and
Hygiene, 48, 839-47.
Lowry, C.A. and Shelley, A.J. (1990) Studies on the scutal patterns of three South
American Simulium species (Diptera: Simuliidae). Canadian Journal of Zoology, 68,
956-61.
Mahon, R.J., Green, C.A. and Hunt, R.H. (1976) Diagnostic allozymes for routine
identification of adults of the Anopheles gambiae complex (Diptera: Culicidae).
Bulletin of Entomological Research, 66, 25-31.
Massachusetts.
Miles, S.J. (1979) The Anopheles gambiae complex: a biochemical key. Journal of
Medical Entomology, 15,297-9.
Miles, S.J. and Paterson, H.E. (1979) Protein variation and systematics in the Culex
pipiens group of species. Mosquito Systematics, 11,187-202.
Milligan, P.J.M., Phillips, A., Molyneux, D.H., Subbarao, S.K. and White, G.B. (1986)
Differentiation of Anopheles culicifaecies Giles (Diptera: Culicidae) sibling species
by analysis of cuticular components. Bulletin of Entomological Research, 76,529-37.
Munstermann, L.E. (1985) Geographic patterns of genetic variation in the treehole
mosquito Aedes triseriatus, in Ecology of Mosquitoes: Proceedings of a Workshop (eds
L.P. Lounibos, J.R. Rey and J.H. Frank), Florida Medical Entomology
Laboratory, Vero Beach, Florida, pp. 327-43.
Munstermann, L.E. (1988) Biochemical systematics of nine Nearctic Aedes mosqui-
toes (subgenus Ochelerotatus annulipes group B), in Biosystematics of
Haematophagous Insects (ed. M.W. Service), Clarendon Press, Oxford, pp. 133-47.
Narang, S.K., Seawright, J.A. and Svarez, M.F. (1991) Genetic structure of natural
populations of Anopheles albimanus in Colombia. Journal of American Mosquito
Control Association, 7,437-45.
Paskewitz, S.M. and Collins, F.H. (1990) Use of polymerase chain reaction to iden-
tify mosquito species of the Anopheles gambiae complex. Medical and Veterinary
Entomology, 4, 367-73.
Paskewitz, S.M., Wesson, D.M. and Collins, F.H. (1993) The internal transcribed
spacers of ribosomal DNA in five members of the Anopheles gambiae complex.
Insect Molecular Biology, 2,247-57.
288 The species concept in Hood-sucking vectors of human diseases
(ed. E.S. Vrba), Transvaal Museum Monograph No. 4, Pretoria, South Africa,
pp. 21-9.
Post, R.J. and Crampton, J.M. (1988) The taxonomic use of variation in repetitive
DNA sequences in the Simulium damnosum complex, in Biosystematics of
Ready, P.O., Smith, D.F. and Killick-Kendrick, R. (1988) DNA hybridisations on
squash-blotted sandflies to identify both Phlebotomus papatasi and infecting
Leishmania major. Medical and Veterinary Entomology, 2,109-16.
Ready, P.O., Lainson, R., Shaw, JJ. and Souza, A.A. (1991) DNA probes for distin-
guishing Psychodopygus wellcomei from Psychodopygus complexus (Diptera:
Psychodidae). Memorias Instituto Oswaldo Cruz, 86,41-9.
Scott, J.A., Brogdon, W.G. and Collins, F.H. (1993) Identification of single speci-
mens of the Anopheles gambiae complex by the polymerase chain reaction.
American Journal of Tropical Medicine and Hygiene, 49, 520-9.
Service, M.W. (1988) New tools for old taxonomic problems in bloodsucking
insects, in Biosystematics of Haematophagous Insects (ed. M.W. Service), Clarendon
Press, Oxford, pp. 325-45.
Simpson, G.G. (1951) The species concept. Evolution, 5, 285-98.
Steiner, W.W.M., Tabachnick, W.J., Rai, K.S. and Narang, S. (eds) (1982) Recent
Developments in the Genetics of Insect Disease Vectors. A Symposium Proceedings,
Stipes, Champaign, Illinois.
Tabachnick, W.J. (1991) Evolutionary genetics and arthropod-borne disease: the
yellow fever mosquito. American Entomologist, 37,14-24.
Tabachnick, W.J. (1992) Genetic differentiation among populations of Culicoides
variipennis (Diptera: Ceratopogonidae), the North American vector of blue-
tongue virus. Annals of the Entomological Society of America, 85,140-7.
Tabachnick, W.J. and Black, W.C. (1995) Making a case for molecular population
genetic studies of arthropod vectors. Parasitology Today, 11, 27-30.
Vajime, G.C. and Dunbar, R.W. (1975) Chromosomal identification of eight species
of the subgenus Edwardsellum near and including Simulium (Edwardsellum)
damnosum Theobald (Diptera: Simuliidae). Tropenmedizin und Parasitologie, 26,
131-8.
Ward, R.D. (1989) Some aspects of the biology of phlebotomine sandflies. Advances
in Disease Vector Research, 91-126.
Ward, R.D., Phillips, A., Burnet, B. and Marcondes, C.M. (1988) The Lutzomyia
longipalpis complex: reproduction and distribution, in Biosystematics of
Wilson, M.D., Post, R.J. and Gomulski, L.M. (1993) Multivariate morphotaxonomy
in the identification of adult females of the Simulium damnosum Theobold com-
plex (Diptera: Simuliidae) in the Onchocertiasis Control Programme area of
West Africa. Annals of Tropical Medicine and Parasitology, 87,65-82.
Wilkerson, R.C., Parsons, T.J., Albright, D.G., Klein, T.A. and Braun, M.J. (1993)
Random amplified polymorphic DNA (RAPD) markers readily distinguish
cryptic mosquito species. Insect Molecular Biology, 4,205-11.
World Health Organization (1977) Species complexes in insect vectors of disease
(blackflies, mosquitoes, tsetse flies). Mimeographed document
WHO/VBC/77.656, World Health Organization, Geneva.
References 289
Wright, J.W. and Pal, R. (eds) (1967) Genetics of Insect Vectors of Disease, Elsevier,
Amsterdam.
Xiong, B. and Kocher, T.D. (1991) Comparison of mitochondrial DNA sequences of
seven morphospecies of black flies (Diptera: Simuliidae). Genome, 34,306-11.
14
Recognition of parthenogenetic
insect species
R. G. Foottit
Contacting address: Eastern Cereal and Oilseed Research Centre, Research Branch,
Agriculture and Agri-Food Canada, K.W. Neatby Bldg., Central Experimental Farm,
Ottawa, Ontario, K1A OC6, Canada
ABSTRACT
Parthenogenesis is a common phenomenon in the Animal Kingdom. It
is estimated that there are over 1000 obligately parthenogenetic species
situated over a broad range of taxa and over 15 000 species reproduc-
ing by cyclic parthenogenesis. In such insect groups as aphids, thrips,
and gall midges, such factors as complex alternation of generations
(with numerous different morphs) and the presence of persistent clon-
al, parthenogenetic populations, make the recognition of species a dif-
ficult task. Parthenogenetic insects require a practical species definition
for proper taxonomic and biological handling of species. This paper
examines the nature and extent of parthenogenetic insects, their evo-
lution and diversity, and makes recommendations for the taxonomic
treatment of species. Topics such as methodologies, degree of discrete-
ness among species, patterns of variation and past treatment under dif-
ferent species concepts are discussed. It is suggested that there is no
one comprehensive definition for parthenogenetic species. They can
best be handled by concepts and practices that interpret pattern along
with biological reality and which incorporate a genealogical perspec-
tive at the clonal, population and species level.
14.1 INTRODUCTION
Whether one is a taxonomist or evolutionary biologist, describing or
analysing biodiversity, or one who studies pest or beneficial organisms
for practical purposes, the species is the fundamental unit of diversity
(Wilson, 1992). There have been many proposals for the conceptual and
practical handling of species and among these, the biological species con-
cept (BSC; Mayr, 1942 ) has been the most thoroughly promoted and
292 Recognition of parthenogenetic insect species
most widely accepted. One of the more frequently stated deficiencies of
the BSC is its inapplicability to uniparental organisms. This fact has been
considered at length by Mayr, the main proponent of the BSC (Mayr,
1963, 1982, 1987) and by many critics of the BSC and advocates of other
viewpoints and solutions to the taxonomic handling of asexual and
parthenogenetic organisms.
In terms of sheer numbers, parthenogenesis is a common phenomenon
in the Animal Kingdom. It has been estimated that there are over 1000
obligately parthenogenetic species situated over a broad range of taxa
and over 15 000 species which reproduce by cyclic parthenogenesis
(White, 1978; Bell, 1982). However, obligate parthenogenesis can be con-
sidered to be rare when viewed in the relation to the total number of ani-
mal species. This phenomenon is estimated to occur in approximately
1% of insect species (Bell, 1982) and 0.1% of the total animal kingdom
(White, 1978). The BSC, with its condition of reproductive isolation
between interbreeding populations of different species, produces diffi-
culties when one is dealing with completely asexual or parthenogenetic
organisms; that is, where no interbreeding takes place. However, these
asexual organisms show a level of integrity that is commonly recognized
by specialist taxonomists.
Workers differ in their opinions about the need to resolve the issue of
species concepts and practical taxonomic handling in asexual and
parthenogenetic organisms. Mayr (1963: 27) has stated that It is too early
for a definitive proposal concerning the application of the species concept
to asexually or uniparentally reproducing organisms'. Scudder (1974) has
pointed out that this stance hardly satisfies those who study groups where
the BSC does not apply. Hull (1970) noted that uniparental organisms
adapt, invade new ecological niches and evolve; thus, they form species
and criteria are needed to delimit these species. Nevertheless, the task of
determining species in completely, partially or cyclically parthenogenetic
animals presents a practical challenge. Characteristics associated with the
parthenogenetic mode of reproduction, such as the complex alternation of
generations with numerous different morphological forms, and the pres-
ence of persistent clonal, parthenogenetic populations among cyclically
parthenogenetic populations, can make the practical recognition of
species operationally difficult. A recent workshop (Hawksworth, 1994)
identified again the pressing need to resolve the disorder of differing or
even obscure species concepts associated with asexual and partheno-
genetic organisms.
This chapter will review the nature of parthenogenesis and examine
the theoretical and practical aspects of species recognition in partheno-
genetic insects. As will be shown, parthenogenetic insects are diverse,
important to society and require a practical definition for proper taxo-
nomic and biological consideration. The problem of parthenogenetic
Nature and extent of parthenogenetic reproduction 293
species is not unique to any particular species concept; the impact of var-
ious species concepts, from phenetic to phylogenetic interpretations, will
be considered.
14.2 NATURE AND EXTENT OF PARTHENOGENETIC

REPRODUCTION
General reviews of the nature, taxonomic distribution and ecological and
evolutionary significance of asexual and parthenogenetic reproduction
are provided by Bell (1982), Hughes (1989), Oliver (1971) and Suomalainen
et al. (1987). The following is a brief survey of asexual and parthenogenet-
ic reproduction with an emphasis on insects.
Many taxonomic groups of organisms include species that reproduce
uniparentally or clonally, that is, the result is an assemblage of individuals
that is genetically identical by descent (Bell, 1982). This occurs in animals
in two different ways. Asexual (agametic) reproduction involves fragmen-
tation, fission or budding and occurs in organisms of rather simple struc-
ture. Parthenogenesis, which includes both the meiotic and the ameiotic
formation of reproductive cells, includes all breeding systems in which the
eggs develop into new individuals without fertilization by gametes pro-
duced by another individual (Bell, 1982).
Parthenogenesis is a varied mechanism with a complex terminology
associated with it. There are two main categories of parthenogenesis that
are of interest, namely, arrhenotoky and thelytoky. In arrhenotoky, a com-
mon mode of reproduction in rotifers, some mite groups and several insect
orders, the female undergoes meiosis, resulting in haploid eggs. The egg
may develop with (producing a female) or without fertilization (producing
a male). There are two types of thelytoky, the development of diploid
females from unfertilized eggs. Automictic (meiotic) thelytoky involves
meiosis and some mechanism of fusion of meiotic products. Depending on
how diploidy is restored, this mechanism could bring about some degree
of genetic mixing and result in the formation of diverse, distinct clones
with subsequent adaptive radiation over time. In apomictic (ameiotic) the-
lytoky, a mechanism of cloning, the eggs develop without meiosis; this
does not result in genetic mixing. It is typically found in groups with cycli-
cal parthenogenesis, such as the aphids, and obligate parthenogenetic
groups that have lost the capacity for bisexual reproduction.
Extensive surveys of the distribution of parthenogenesis in insects are
provided by Bell (1982), Suomalainen et al. (1976,1987) and White (1978).
The occurrence of parthenogenesis in insects is sporadic throughout the
order; it involves a wide range of cytological mechanisms (Suomalainen et
al., 1976), but is, in general, concentrated in certain taxa and is entirely
lacking in others. Parthenogenetic species often occur sporadically among
bisexual species.
Facultative or obligate thelytoky is present in over 200 genera from over
80 families of insects (Bell, 1982; Suomalainen et al, 1987). For example, it
is common in the Collembola, Psocoptera, curculionid beetles and the
sternorhynchous Homoptera (aphids, scales and whiteflies). Arrhenotoky
is found in the order Thysanoptera; there are also some secondarily the-
lytokous thrips. Parthenogenesis, including both arrhenotoky and thely-
toky, is the characteristic mode of reproduction in the Hymenoptera.
Cyclical parthenogenesis is a common reproductive pattern found in
groups such as trematodes, cladocerans, rotifers, aphids, cynipid gall
wasps, cecidomyiids and the beetle, Micromalthus. There is alternation
between parthenogenesis (usually apomictic) and sexual reproduction,
thus combining the ecological and evolutionary advantages of both modes
of reproduction. There are from one to many parthenogenetic generations
followed by a sexual generation. Among these cyclical species there may
also be purely thelytokous (anholocyclic) populations that do not return to
the sexual phase of the life-cycle. Through biogeographical analysis, it has
been shown that in one instance an aphid-host plant association and cycli-
cal life history has persisted for 48 million years (Moran, 1989).
14.3 PAST TREATMENT OF ASEXUAL AND PARTHENOGENETIC

SPECIES
Species concepts have long been an important topic of discussion in
systematics and evolutionary biology. Often, in these many treatments,
there has been discussion of asexual and parthenogenetic species.
Many viewpoints and solutions to the conceptual and practical han-
dling of these organisms have been proposed. The botanist G. E. Du
Rietz (1930), one of the early workers to discuss comprehensively the
nature of species concepts, acknowledged the problems for species
concepts presented by asexual populations and noted the difficulties in
determining the correspondence of rank between sexual species and
units of populations of asexual organisms. He cited Turesson's (1929)
use of the term agamospecies, 'An apomict-population the constituents
of which, for morphological, cytological or other reasons, are to be con-
sidered as having common origin', but dismissed this definition as too
vague to be of practical value, noting that it could be applicable at a
number of taxonomic ranks. In fact, Du Rietz stated (1930) that it was a
rather hopeless task to apply a species concept to groups of what he
termed apomictic microspecies. He advocated distinguishing sexual and
asexual species and recommended applying the corresponding sexual
species rank to the asexual microspecies (populations of apomicts) as
such, noting that the presence of discontinuity is the most important
consideration when determining species, not the relative size of the pop-
ulations. As will be seen, use of a number of terms such as agamospecies
Past treatment of asexual and parthenogenetic species 295
and microspecies, have arisen in discussions by zoologists on a number
of occasions.
Dobzhansky (1937) took issue with Du Rietz's conclusions, claiming that
agamic complexes do not comprise species. He defined species by the abil-
ity actually or potentially to exchange genes. He noted that asexual groups
are aggregations of distinct, constantly reproducing genotypes, clustered
around adaptive peaks and that these aggregates are arranged in an hier-
archical order analogous to that in sexual forms (i.e. species, sub-genera,
genera, etc.). Dobzhansky felt that assigning rank was arbitrary, that is,
there is no potential ability to develop a species category in asexual organ-
isms that is more fixed and less arbitrary than other ranks; all criteria of
species distinction break down. Later, Mayr (1957, 1963) also discussed a
similar concept. Dobzhansky (1972) suggested that species in asexual or
parthenogenetic organisms should be labelled as pseudospecies, to distin-
guish their biological differences from sexual species. He noted that, as
there are different strategies of evolutionary adaptation, there are different
kinds of species.
In his treatment of animal species, Cain (1954) again used the term
agamospecies to account for those forms to which the biological species
concept cannot apply because there is no sexual reproduction. He regard-
ed agamospecies as being relatively recent phenomena, exhibiting little
variability or long-term adaptability, not part of the main evolutionary
pathway, although they may be successful in the relative short term.
Unlike Dobzhansky, Cain felt that a morphological species concept could
be applied equally well to agamospecies as to sexually reproducing
species, and that agamospecies could be placed in a natural hierarchy and
that they should be recognized.
Meglitsch (1954) proposed an evolutionary concept of the species and
argued that it could be applied to both uniparental and biparental organ-
isms and that, while the genetic processes are different, the outcome in
both cases is the evolution of species. This was a view that was later devel-
oped by Simpson (1961) who also noted that while the evolution of uni-
parental and biparental populations is very different, this does not mean
that species are not formed. However, as argued by some (Sokal and
Crovello, 1970), the evolutionary species definition will fit most situations
but is so vague as to be operationally impossible, particularly in terms of
the criterion of evolutionary role.
While taking an operational approach and noting that thelytokous
clones of parasitic Hymenoptera are of importance in biological control
and that these entities need to be named for purposes of information
retrieval, De Bach (1969) proposed an ethological-ecological species con-
cept for the recognition of species of uniparental organisms. He pointed
out that the proposed criteria of an operational approach to species defin-
ition would delimit parthenogenetic species, even if these forms were
morphologically identical to other bisexual forms. He provided a series of
taxonomic tests for comparison with closely related morphological
biparental species which would determine if the uniparental group
behaved like a biological species; if it had a significantly different etholog-
ical or ecological role it should be considered as a valid species.
The basic tenet of the biological species concept is discontinuity
among species due to reproductive isolation (Mayr, 1957). In many of his
works (1957,1963,1969,1982,1987) Mayr has dealt with asexuality as the
fundamental impediment of the BSC. This obstacle is due to the fact that
in these organisms, reproductive isolation cannot be tested and that
asexual organisms do not exist in populations in the same sense as sexu-
al organisms, that is, as interbreeding groups. Mayr (1957,1963,1969) has
argued that the most satisfactory solution in taxonomic terms is to use
the biological concept of species in sexual organisms and a morphologi-
cal concept in asexual organisms. This is justified biologically because of
the relationship between reproductive isolation and morphological dif-
ferences, that is, they are the products of the same degree of genetic dif-
ference and are thus correlated. Mayr suggested the same adaptive peak
situation as had Dobzhansky, and later (Mayr, 1982) extended this con-
cept to include the fact that morphological discontinuities give indica-
tions of the specific ecological niches that asexual organisms occupy and
thus provide inferences about their species status. Mayr (1957) also
argued that while many previous workers would refer to asexual or uni-
parental populations, every individual and its descendants are repro-
ductively isolated and that the term population does not apply.
However, it would appear that workers, such as Simpson (1961), are
really discussing the diversity, the spread and the differential selection
of clones.
Realizing that the BSC does not reflect the range of kinds of species and
speciation processes that exist, some have argued for a wide, pluralistic
species concept which would apply to organisms with different modes of
reproduction (Mishler and Donoghue, 1982; Mishler and Brandon, 1987).
Some have attempted to resolve the issue by arguing that asexual organ-
isms, by definition, cannot form species (Ghiselin, 1987). A counter argu-
ment, presented by Mishler and Brandon (1987) is that the inability of the
BSC to apply to uniparental organisms is not a reason to deny that asexu-
al species exist, but that it is simply a fault of the BSC. In several papers,
Mayr (1963, 1992) has simply minimized the impact of asexual and
parthenogenetic reproduction on the BSC.
In many cases, the empirical evidence of species analysis reveals complex
patterns of variation in morphological, genetic and ecological variation
while the species concepts are too generalized to represent these situations.
The criteria recommended for determining meaningful discontinuities
within biological variation are very much dependent upon the biology of
Evolutionary considerations 297
specific groups (Mishler and Donoghue, 1982). Scudder (1974) noted that it
is an impossible task to develop a single, precise, meaningful species defin-
ition that is applicable to all organisms and acceptable to both taxonomists
and evolutionists. While there are many published arguments to the con-
trary, some of which have been discussed above, particularly from an oper-
ational viewpoint, I feel that it is both practical and biologically accurate to
recognize that there are different kinds of species as a result of different
kinds of speciation processes.
14.4 EVOLUTIONARY CONSIDERATIONS
14.4.1 Variability
Many writers, while discussing the difficulties of systematically handling
groups of parthenogenetic organisms within existing species concepts,
often on the same occasion dismissed the importance of these groups in
ecological and evolutionary terms, perhaps in an attempt to reduce the
problem. Due to presumed limited variability, parthenogenesis has often
been thought to be restricted to narrow, specific environmental conditions
in space and time. Many workers (Du Rietz, 1930; Dobzhansky, 1937;
Cain, 1954; Mayr 1957, 1963) have generalized that parthenogenetic
groups are relatively recent, secondarily derived phenomena and thus are
impossible to designate as species in a non-arbitrary way, are too similar
to their sexual relatives to distinguish as separate species, or simply do not
require an essentially different species category.
Parthenogenesis has also been considered by many theoreticians to be
an evolutionary dead-end due to a lack of genetic variability and the
steady accumulation of deleterious mutations in clonal lineages (Muller's
Ratchet; Felsenstein, 1974) (White, 1978; Bell, 1982; Maynard-Smith, 1986).
Clones are believed to be unable to diversify rapidly enough to meet
changing environmental conditions and to overcome average rates of
extinction. However, there are indications that these traditional views are
now considered less general and acceptable (Suomalainen et al., 1976;
Ghiselin, 1988). Studies by Suomalainen (1961) showed a large amount of
morphological variation among populations of parthenogenetic weevils.
Subsequently, allozyme analyses provided further evidence of substantial
clonal variation in many instances. For example, Saura et al. (1976)
revealed 76 clones of a weevil and Mitter et al. (1979) identified 36 clones
of a parthenogenetic moth. This diversity could be the result of different
origins or the result of within-clone evolution.
While there have been evolutionary arguments about just how rela-
tively long- or short-term the evolution of parthenogens is, the genetic
variability that does exist in parthenogens is usually considered to be a rel-
atively finite phenomenon; variation is ultimately selected out through
clonal selection or driven to homozygosity by automixis. There is growing
evidence that clones do become extinct and only persist successfully in sit-
uations where the conditions are very favourable for them (Hughes, 1989).
Single gene mutations may result in the development of partheno-
genetic lineages within sexual populations. However, the majority of
parthenogenetic taxa are the result of hybridization (White, 1978).
Polyploidy is a frequent and important correlate of parthenogenesis (Bell,
1982); it is particularly common in those groups with apomictic partheno-
genesis. As a result, in the short term, parthenogenetic lineages are able to
occupy broad niches and be ecologically successful.
Increasingly, data show that the genetic structure of parthenogenetic
populations can be complex (Hebert, 1987) and that the evolutionary
potential of these groups in not as limited as has been thought. Finston et
al. (1995), in a study of genome size variation in aphids, have postulated
that shifts in small genome sizes, associated with short generation times in
parthenogenetic aphids, provide a means for saltational change in charac-
ter states and may be important in the evolutionary diversification of this
group.
14.4.2 Diversity
Many consider that the taxonomic pattern where most parthenogens have
close sexual relatives also indicates that they are successful only in the evo-
lutionary short term. There is taxonomic and phylogenetic evidence that,
in some groups, there has been considerable evolutionary radiation in
some parthenogenetic lineages. The most widely cited example is the
bdelloid rotifers, a thelytokous, species-rich group of over 300 relatively
easily recognized species in four families (Hutchinson, 1967). As an expla-
nation for this diversity, it has been postulated that these discrete species
represent scattered adaptive peaks that are the survivors from a larger
array of produced clonal forms and that, due to their protected aquatic
habitats, they have low rates of extinction (Stanley, 1975).
In the oribatid mites, parthenogenesis is estimated to occur in 8-9% of
the known species (Norton and Palmer, 1991). Included is the largest
group of animals that reproduce solely by thelytokous parthenogenesis,
the Desmonomata, an early derivative taxon comprising about 400 species
in more than 30 genera and seven families (Palmer and Norton, 1992). It
has been proposed (Norton and Palmer, 1991) that many families of orib-
atid mites have speciated in the absence of sexual reproduction through
meiotic thelytoky from ancestors who were also parthenogenetic.
Among the insects, another possible example of parthenogenetic radia-
tion can be found in the aphid genus Trama. This group, in which no
males have been found, consists of about 30 species that feed exclusively
Evolutionary considerations 299
on the roots of Compositae and which have an extensive distribution pri-
marily in Europe and Asia. Again, their comparatively protected environ-
ment may have resulted in a low rate of extinction (Eastop, 1953).
14.4.3 Cyclical parthenogenesis

Cyclical parthenogenesis has been a source of evolutionary change in
parthenogens. For example, the Aphidoidea exhibit a wide range of varia-
tion in cyclical parthenogenesis (Moran, 1992). The complete life-cycle
(holocycle) consists of a single sexual generation which occurs during the
colder part of the year and series of summer, parthenogenetic generations.
There are many previously holocyclic aphids, that are now confined to a
single host plant and that are without sexual reproduction in their life-cycle
(anholocycle). These obligately parthenogenetic lineages could be the result
of accumulated mutations within clones or could have come about through
the multiple evolution of thelytoky from different sexual genotypes.
Among the aphids, complex life-cycles involving cyclical partheno-
genesis and winter and summer host plant alternation occur frequently.
For example, the agricultural pest, Myzus persicae, now spread through-
out the world, is likely to have originated in China on its primary, winter
host plant, Primus persica. In cold climates M. persicae undergoes the com-
plete holocycle from winter, primary to summer, secondary hosts while
in tropical climates it loses the bisexual part of the life-cycle, remaining
completely parthenogenetic on secondary hosts (anholocycly). In inter-
mediate climates, this aphid can be holocyclic, anholocyclic or even
androcyclic, where the parthenogenetic females eventually produce
functional males (Blackman, 1981,1985). In addition to heritable life-cycle
variation, some clones of M. persicae exhibit insecticide resistance, variable
ability to transmit plant viruses, general and specific and geographic host
plant preferences and interclonal variation in the ability to produce sex-
ual morphs. Related species, such as M. dianthicola and M. ascalonicus
appear to consist entirely of anholocyclic forms.
There are a number of examples of incipient speciation among
parthenogens. The Spotted Alfalfa Aphid, Therioaphis trifolii, first appeared
as anholocyclic populations on alfalfa in the southern United States in the
1950s. It was subsequently determined that this aphid had been intro-
duced into eastern North America over 70 years before and named the
Yellow Clover Aphid. The Spotted Alfalfa Aphid subsequently exhibited
insecticide resistance which spread to other areas. As its range extended
into the northern United States, the aphid started to produce the sexual
stage and overwintering eggs. It appears that these two entities represent
separate introductions from rather different genotypes of Old World T.
trifolii (Blackman, 1981).
14.4.4 Geographical variation
Parthenogens show considerable geographic variation which indicates a
capacity to evolve. Even in situations where parthenogenesis is a rela-
tively recent, secondarily derived situation, there is evidence of evolu-
tionary change. Geographical analyses have shown that parthenogenetic
organisms have distributions that are different from those of their bisex-
ual relatives. It has also been noted that if there is a parthenogenetic and
a bisexual form of the same species, the successful parthenogenetic form
usually has a wider distribution. These situations have been called geo-
graphical parthenogenesis (Lynch, 1984). There are many examples in
Europe where the bisexual form is confined to a limited area and the
parthenogenetic form has spread extensively (Suomalainen et at, 1976).
This situation can be explained by the existence of general-purpose geno-
types which occupy a wide range of habitats, that is, there is strong selec-
tion pressure on obligate parthenogens to produce general-purpose
genotypes, able to tolerate a wide range of conditions (Lynch, 1984). The
alternate explanation would be that species consist of a set of distinct
genotypes, each specialized for a narrow subset of the habitat range.
Suomalainen (1962) and Suomalainen et al. (1976) have explained geo-
graphical parthenogenesis in weevils by invoking a glacial refugium model.
Glaciation during the last Ice Age has been the selective force involved in
determining the distribution of species of parthenogenetic insects. Bisexual
forms of the weevil, Otiorrhynchus scaber, survived in isolated, ice-free areas.
As the glaciers retreated, these populations spread and interbred, produc-
ing increased heterozygosity, which became fixed in the parthenogenetic
forms. Subsequent polyploid clones were able to adapt to new habitats
which were created as the glacier retreated (Saura et al., 1976).
Cyclically parthenogenetic Adelgidae alternate between a primary host
which is always a species of spruce (Picea) and a secondary host which is
another species of conifer. As is the case with the closely related true
aphids (Aphididae), the adelgids often have a truncated life-cycle with
associated anholocyclic forms. In the Adelgidae these may occur on either
the primary or the secondary host plant. Many of these persistent
parthenogenetic lineages are considered to be distinct species by taxo-
nomic specialists. Anholocycly has come about either through mutational
elimination of the male line or by glacial displacement of the primary host
plant. Steffan (1961, 1963) has, for example, shown that the anholocyclic
species, Sacciphantes abietis, on spruce and S. segregis on larch have split off
from the holocyclic S. viridis which migrates from spruce to larch. Steffan
(1964) has used a concept of agamospecies for these parthenogenetic lin-
eages. He has also used the term microspecies, as components of a larger
artenkreis or superspecies, in this case the A. abietis-A. segregis-A. viridis
combination.
Practical considerations 301
A number of adelgid agamospecies are important forest pests; the abili-
ty to produce overwintering eggs enables them to persist in otherwise
harsh climates. The balsam woolly aphid, Adelges piceae, is indigenous to
Europe, where it has lost its primary host and is found only on species of
Abies. This adelgid has been introduced into North America on nursery
stock on a number of occasions and into different regions. This completely
parthenogenetic species has shown considerable biological variability in its
physiological effects on its host. Using multivariate morphometrics, mor-
phologically distinct groupings of population samples, associated with
three different areas of introduction in North America have been demon-
strated (Foottit and Mackauer, 1980). These groupings were subsequently
distinguished as geographical subspecies (Foottit and Mackauer, 1983).
14.5 PRACTICAL CONSIDERATIONS

Claridge (1995) has rightly pointed out that species concepts are necessary
in order to describe patterns of biological diversity and that different
species concepts will have different consequences for the analysis of spe-
ciation processes. Parthenogenetic insects may not exist as populations or
form species in the same manner as do sexual species. However, they do
adapt and evolve and they have unique ecological and evolutionary roles.
Clonal diversity in obligate and cyclical parthenogens is real, generated by
life-history dynamics, vicariant processes and selection pressure. In bio-
logical terms, it is necessary to recognize parthenogenetic species. A cate-
gory for parthenogenetic insects is necessary for practical purposes, such
as the needs of applied biologists.
14.5.1 Populations
The use of the term population, in the context of parthenogenetic organ-
isms, needs clarification. It is often stated that obligate parthenogens only
form clones, not populations in the sense of the interbreeding group of
evolutionary biology. As has been outlined above, parthenogenetic
species are far from uniform; they are genetically variable and clonally
diverse. Mutations that are selectively advantageous spread throughout a
species. Parthenogenetic species do form populations of co-adapted
clones and it is in that sense that the term population should be used.
14.5.2 Analytical methods

As discussed above, there has been extensive consideration of how to
accommodate asexual and parthenogenetic organisms within the various
species concepts. Nevertheless, at the operational level, the main
approach has been that of the descriptive morphospecies and comparison
with closely related sexual species; biological data are incorporated into
the taxonomic process, if the data are available. In fact, many workers
(Mayr, 1963; De Bach, 1969) have essentially recommended a holomor-
phological approach to delimiting parthenogenetic species (Enghoff,
1976). Morphological gaps are viewed as representative of the degree of
genetic isolation resulting from natural selection and are indicative of the
species-specific ecological niche that is occupied (Mayr, 1982).
Recognition of the extent of parthenogenetic clones and species can be
a relatively local problem or it may require a broad perspective over wide
species, host and geographical ranges. Theory has provided much
towards the understanding of the evolution and systematics of partheno-
genetic organisms. Further understanding and the practical recognition of
parthenogenetic taxa will come about through the empirical comparison
of clonal lineages in a phylogenetic framework. A number of approaches,
particularly DNA sequencing techniques, allozyme analysis and multi-
variate morphometrics, will prove useful in parthenogenetic species reso-
lution as these methods not only delimit taxa they also provide inferential
information about the ecology and evolution of these species.
Mitochondrial DNA, which is maternally inherited, and which exhibits
a rapid rate of sequence divergence, will provide the data necessary to
associate parthenogenetic lineages and species with their most closely
related sexual species. Allozyme analysis, particularly indications of fixed
gene differences (Hales, 1991), provides additional evidence in support of
morphological differences and may provide indications of the existence
of lineages or species in cases where differences are morphologically
cryptic. Parthenogenesis, particularly cyclical parthenogenesis and asso-
ciated progenetic development, results in a reduction in available mor-
phological characters. Multivariate morphometrics is necessary for the
resolution of subtle population and species limits in these cases.
Additionally, and based on the assumption that multivariate ordinations
evaluate partitioned variance and reflect the polarity of character states,
these methods can be used to provide phylogenetic resolution (Foottit
and Sorensen, 1992).
14.5.3 Recommendations: handling parthenogenetic lineages

Many absolute statements have been made about what is a variable phe-
nomenon. I doubt that we will obtain a single, comprehensive species con-
cept or species definition for all parthenogenetic insects, given the variety
of genetic and evolutionary mechanisms involved. However, species of
parthenogenetic insects can be best handled taxonomically by concepts
and practices that interpret pattern along with biological reality and
which incorporate a genealogical perspective, at the clonal, population
Practical considerations 303
and species level, where possible. Sound taxonomic judgement, as always,
is an important requirement.
The BSC should be supplemented through a pluralistic approach
(Mishler and Donoghue, 1982). The recognition of agamospecies (asexual
or parthenogenetic species; species without sexual reproduction) as pro-
posed by a number of workers, mentioned above, is a minimally pluralis-
tic solution to the problem of parthenogenetic species (Mayr, 1992). This
recognizes that parthenogenetic species have an integrity that is main-
tained by selection for gene combinations that are adaptive and that these
mechanisms are different from those of sexual species. Cyclical
parthenogens can be accommodated in a strict sense with the BSC,
although even in these cases, geographically limited, infrequent sexual
reproduction will result in patterns similar to obligate parthenogenesis
(Mishler and Budd, 1990; Hebert et al, 1991).
Blackman (1995) has recently pointed out that some of the difficulty of
the species problem is in fact confusion between the theoretical issues of
species concepts and the practical issues of working definitions of species.
The biological diversity of parthenogenetic organisms, as reviewed above,
clearly exists; these parthenogenetic species exist, evolve and require tax-
onomic handling. The practical issues of a working definition can best be
addressed through the use of the agamospecies definition, as above.
Clonal populations in various circumstances, if clearly distinctive and
diagnosable, can be designated, in a taxonomically neutral fashion, as uni-
parental clonal lineages, that is, as parthenogenetic ancestral-descendant
sequences of clones that are unique historical entities.
As was recommended by Enghoff (1976) and Suomalainen et al. (1987),
when a parthenogenetic lineage is derived from a known bisexual species,
it should not be designated as a separate species, it should be designated
as the parthenogenetic form of species Aus bus. Known hybrid forms
should be identified by a hybrid designation (e.g. Aus bus x Bus CMS).
These measures will prevent unnecessary naming of species where there
is inadequate data. However, if there is evidence that clonal populations
are evolving and occupying a habitat and perpetuating, they constitute a
historically unique species and should be distinguished as agamospecies.
Criteria for agamospecies status would be differences from bisexual rela-
tives in terms of host plant preferences, chromosome number, etc., that is,
in meaningful differences, not just taxonomic convenience.
While the phylogenetic species concept (Mishler and Brandon, 1987) is
an attractive conceptual framework for emphasizing phylogenetic recon-
struction, strict adherence to this concept would result in the recognition
of all monophyletic, geographically distinct, uniparental clonal lineages as
formal taxa. This increase in the number of binomials would not be prac-
tically advantageous. Nevertheless, each historical group of organism lines
(tokogenetic arrays; Hennig, 1966) could be handled in a phylogenetic
framework, assuming appropriate cladistic evidence, but only using infor-
mal names for purposes of communication, unless there was convincing,
biologically significant evidence for species status. A phylogenetic frame-
work would serve the important function of providing a hierarchical
means of grouping clonal lineages.
14.6 CONCLUSIONS
There is increasing evidence that while the distribution of parthenogenesis
is sporadic it is not random (Bell, 1982). In many taxa, parthenogens are not
evolutionary dead-ends, but they are genetically diverse and ecologically
widely adapted and have the ability to adapt to changing environments
(Suomalainen et al., 1987). Parthenogenesis is most common in situations of
disturbance, such as fire, drought, glaciation and grazing and ephemeral or
fluctuating habitats, as are prevalent in agricultural situations. For exam-
ple, there is considerable evidence for host plant-associated genotypes in
parthenogenetic insects; host plant adaptations play an important role in
the pest status of agriculturally important insects. It is therefore very
important from the perspective of effective pest management that we have
operational species definitions which allow workers to handle partheno-
genetic insect species.
14.7 REFERENCES
Bell, G. (1982) The Masterpiece of Nature. The Evolution and Genetics of Sexuality,
University of California Press, Berkeley and Los Angeles.
Blackman, R.L. (1981) Species, sex and parthenogenesis in aphids, in The Evolving
Biosphere (ed. P. Forey), Cambridge University Press, Cambridge, pp. 75-85.
Blackman, R.L. (1985) Aphid cytology and genetics, in Evolution and Biosystematics
of Aphids, Proceedings of the International Aphidological Symposium at Jablonna, 1981,
Wydawnictowa Polska Akademia Nauk, Warsaw, pp. 171-237.
Blackman, R.L. (1995) What's in a name? Species concepts and realities. Bulletin of
Entomological Research, 85,1-4.
Cain, AJ. (1954) Animal Species and their Evolution, Hutchinson, London.
Claridge, M.F. (1995) Species and speciation. Trends in Ecology and Evolution, 10,38.
De Bach, P. (1969) Uniparental, sibling and semi-species in relation to taxonomy
and biological control. Israel Journal of Entomology, 4,11-28.
Dobzhansky, Th. (1937) Genetics and the Origin of Species, Columbia University
Press, New York.
Dobzhansky, Th. (1972) Species of Drosophila. New excitement in an old field.
Science, 177, 664-9.
Botanisk Tidskrift, 24, 333-428.
Eastop, V. (1953) A study of the Tramini (Homoptera-Aphididae). Transactions of
the Royal Entomological Society of London, 104, 385-413.
References 305
Enghoff, H. (1976) Taxonomic problems in parthenogenetic animals. Zoologica
Scripta, 5,103-4.
Felsenstein, J. (1974) The evolutionary advantage of recombination. Genetics, 78,
737-56.
Finston, T.L., Hebert, P.D.N. and Foottit, R.G. (1995) Genome size variation in
aphids. Insect Biochemistry and Molecular Biology, 25,189-96.
Foottit, R.G. and Mackauer, M. (1980) Morphometric variation between popula-
tions of the balsam woolly aphid, Adelges piceae (Ratzeburg) (Homoptera:
Adelgidae), in North America. Canadian Journal of Zoology, 58,1494-503.
Foottit, R.G. and Mackauer, M. (1983) Subspecies of the balsam woolly aphid,
Adelges piceae (Homoptera: Adelgidae), in North America. Annals of the
Entomological Society of America, 76, 299-304.
Foottit, R.G. and Sorensen, J.T. (1992) Ordination methods: Their contrast to clus-
tering and cladistic techniques, in Ordination in the Study of Morphology,
Evolution and Systematics of Insects: Applications and Quantitative Genetic Rationales
(eds J. T. Sorensen and R.G. Foottit), Elsevier, Amsterdam, pp. 1-10.
Ghiselin, M.T. (1987) Species concepts, individuality, and objectivity. Biology and
Philosophy, 2,127-43.
Ghiselin, M.T. (1988) The evolution of sex: A history of competing points of view,
in The Evolution of Sex. An Examination of Current Ideas (eds R. E. Michod and R.
B. Levin), Sinauer Associates, Sunderland, Massachusetts, pp. 7-23.
Hales, D. (1991) Use of allozyme electrophoresis in distinguishing morphological-
ly similar species of Aphididae (Homoptera: Aphidinea). Entomologia Generalis,
16,193-200.
Hawksworth, D. (ed.) (1994) The Identification and Characterization of Pest Organisms,
CAB International, Wallingford, UK.
Hebert, P.D.N. (1987) Genotypic characteristics of cyclic parthenogens and their
obligately asexual derivatives, in Evolution of Sex and its Consequences (ed. S.J.
Stearns), Birkhauser, Berlin, pp. 175-95.
Hebert, P.D.N., Finston, T.L. and Foottit, R. (1991) Patterns of genetic diversity in
the sumac gall aphid, Melaphis rhois. Genome, 34, 757-62.
Hennig, E. (1966) Phylogenetic Systematics, University of Illinois Press, Urbana.
Hughes, R.N. (1989) A Functional Biology of Clonal Animals, Chapman & Hall, New
York.
Hull, D. (1970) Contemporary systematic philosophies. Annual Review of Ecology
and Systematics, 1,19-54.
Hutchinson, G.E. (1967) When are species necessary? in Population Biology and
Evolution (ed. R.C. Lewontin), Syracuse University Press, pp. 177-86.
Lynch, M. (1984) Destabilizing hybridization, general-purpose genotypes and geo-
graphic parthenogenesis. The Quarterly Review of Biology, 39, 257-90.
Maynard-Smith, J. (1986) Contemplating life without sex. Nature, 324, 300-1.
Mayr, E. (1942) Systematics and the Origin of Species from the Viewpoint of a Zoologist,
Mayr, E. (1957) Difficulties and importance of the biological species, in The Species
Problem (ed. E. Mayr), American Association For the Advancement of Science,
Massachusetts.
Mayr, E. (1982) The Growth of Biological Thought. Diversity, Evolution, and Inheritance,
Harvard University Press, Cambridge, Massachusetts.
Mayr, E. (1987) The ontological status of species: Scientific progress and philo-
sophical terminology. Biology and Philosophy, 2,145-66
Botany, 79, 222-38.
Meglitsch, P.A. (1954) On the nature of the species. Systematic Zoology, 3,49-65.
Mishler, B.D. and Brandon, R.N. (1987) Individuality, pluralism, and the phyloge-
Mishler, B.D. and Budd, A.F. (1990) Species and evolution in clonal organisms -
Introduction. Systematic Biology, 15, 79-85.
Mishler, B.D. and Donoghue, M.J. (1982) Species concepts: a case for pluralism.
Mitter, C, Futuyma, D.J., Schneider, J.C. and Hare, J.D. (1979) Genetic variation
and host plant relations in a parthenogenetic moth. Evolution, 33, 777-90.
Moran, N.A. (1989) A 48-million-year-old aphid-host plant association and com-
plex life cycle: biogeographic evidence. Science, 245,173-5.
Moran, N.A. (1992) The evolution of aphid life cycles. Annual Review of Entomology,
37,321^8.
Norton, R.A. and Palmer, S.C. (1991) The distribution, mechanisms, and evolu-
tionary significance of parthenogenesis in oribatid mites, in The Acari:
Reproduction, Development and Life-History Strategies (eds R. Schuster and P.W.
Murphy), Chapman & Hall, London, pp. 107-36.
Oliver, J.H. (ed.) (1971) Parthenogenesis. American Zoologist, 11,239-398.
Palmer, S.C. and Norton, R.A. (1992) Genetic diversity in thelytokous oribatid
mites. Biochemical Systematics and Ecology, 20,219-31.
Saura, A., Lokki, J., Lankinen, P. and Suomalainen, E. (1976) Genetic polymor-
phism and evolution in parthenogenetic animals. III. Tetraploid Otiorrhynchus
scaber (Coleoptera: Curculionidae). Hereditas, 82, 79-100.
Scudder, G.G.E. (1974) Species concepts and speciation. Canadian Journal of Zoology,
52,1121-34.
Simpson, G.G. (1961) Principles of Animal Taxonomy, Columbia University Press,
New York and London.
Sokal, R.R. and Crovello, T.J. (1970) The biological species concept: a critical evalu-
ation. American Naturalist, 104,127-53.
Stanley, S.M. (1975) Clades versus clones in evolution: why we have sex. Science,
190, 382-3.
Steffan, A.W. (1961) Die Stammes- und Siedlungsgschichte des Artenkreis
Sacciphantes viridis (Ratzeburg 1843) (Adelgidae: Aphidoidea). Zoologica
(Stuttgart), 109,1-113.
Steffan, A.W. (1963) Zur systematischen und phylogenetischen Stellung der
Agamospecies in den Adelgidae- Genera (Homoptera). Verhandlungen der
Deutschen Zoologischen. Gesellschaft in Wein, 1962, pp. 640-55.
Steffan, A.W. (1964) Problems of evolution and speciation in Adelgidae
(Homoptera: Aphidoidea). Canadian Entomologist, 96,155-7.
Suomalainen, E. (1961) On morphological differences and evolution of different
polyploid parthenogenetic weevil populations. Hereditas, 47, 309-41.
References 307
Suomalainen, E. (1962) Significance of parthenogenesis in the evolution of insects.
Annual Review of Entomology, 7, 349-66.
Suomalainen, E., Saura, A. and Lokki, J. (1976) Evolution of parthenogenetic
insects. Evolutionary Biology, 9,209-57.
Suomalainen, E., Saura. A. and Lokki, J. (1987) Cytology and Evolution in
Parthenogenesis, CRC Press, Boca Raton.
Turesson, G. (1929) Zur Natur und Begrenzung der Arteinheiten. Hereditas, 12,
323-34.
White, M.J.D. (1978) Modes of Speciation, W.H. Freeman and Company, San
Francisco.
Wilson, E.O. (1992) The Diversity of Life, Harvard University Press, Cambridge,
Massachusetts.
15
The species in terrestrial non-
insect invertebrates (earthworms,
arachnids, myriapods, woodlice
and snails)
A. Minelli and D. Foddai
Contacting address: Universita di Padova, Dipartimento di Biologia, Via Trieste 75,135121
Padova, Italy
ABSTRACT
In current practice, species discrimination in these diverse groups
often rests on showy, complex morphological traits, such as the repro-
ductive apparatus of gastropods, the complex gonopods of males in
juliform millipedes, and the not less complex palps of many spider
males. Problems with species delimitation are often caused by
parthenogenesis, well studied in oligochaetes, but common in most
groups of invertebrates.
Perceptions of species barriers are affected by patterns of geo-
graphic distribution, in turn reflecting dispersal power. Allopatric taxa
are usually easy to recognize, but their species status is usually less
obvious. Easily distinguishable and easily dispersed species include
anthropochorous isopods, centipedes and millipedes, well-documented
in the faunas of islands. Very interesting patterns of intensive specia-
tion characterize insular biotas. Large species swarms are known
among Macaronesian millipedes (Cylindroiulus, Dolichoiulus, Acipes)
and Hawaiian spiders (Tetragnathidae). More widespread, however,
are the patterns of continental insularity, for example in the extensive
cave systems along the Southern margin of the Alps, but also on
mountain tops.
Most taxonomic methods have been used, including karyology
(good tradition and good results in earthworms), allozymes (relative-
ly few, but for some snails and earthworms), isozymes, and more
recently also mtDNA. Allozymes are sometimes better than morphol-
ogy (earthworms such as Hormogaster species; sibling species of
Macrocheles mites), but quite often they do not correlate with mor-
phology. Attitudes of specialists towards the value of Nei's D esti-
mates as a cue to species status are very divergent.
For this sample of more than 100 000 named species from a diver-
sity of phyla, species are neither easier nor more difficult to recognize
than in most other animal groups.
15.1 INTRODUCTION
The animal groups surveyed in this paper are very diverse and the taxo-
nomic traditions are very different. Also some of the key biological fea-
tures of these animals, such as dispersal power and reproductive systems,
are very different. Thus, these groups offer a variegated sample, both in
terms of species circumscription and identification. Some groups, such as
centipedes, millipedes and many arachnid taxa, are among the less-
studied groups of terrestrial invertebrates. Others, however, have often
been the subject of critical discussion of species boundaries and a few have
even played the role of textbook examples in discussions about species and
speciation. Obvious examples include snails of the genera Cepaea, Partula,
Albinaria and Cerion. To date, very few critical reviews of 'the species in
practice' have been devoted to these groups, exceptions being snails (Giusti
and Manganelli, 1992) and spiders (Blandin, 1977; Brignoli, 1988).
Overall, more than 100 000 living species from these groups have been
described: in crude figures, 12 000 myriapods, some 3500 woodlice, 1000
terrestrial oligochaetes, 35 000 spiders, nearly 40 000 non-aquatic mites,
9000 further arachnids and some 35 000 land snails.
A first reflection of the current state of affairs may be gleaned from a
few examples of the levels of synonymy in these groups. We may expect
that groups where species are more difficult to identify will have a higher
average level of synonymy than groups where species are easier to cir-
cumscribe (Holman, 1987). Recent estimates of the number of known liv-
ing gastropod species range between 40 000 and over 100 000 (Bieler,
1992). To this uncertainty, marine, freshwater and terrestrial forms are
likely to contribute in similar ways. It is not too difficult to find species
with a few dozen synonyms each, when browsing through recent revi-
sionary works, especially those dealing with land snails from the
European/Mediterranean area.
For spiders, very striking are the numbers of nominal taxa recently put
into synonymy under other species as the effect of revisionary works on
some genera of orb-web spiders (Araneidae). For example, Levi (1983)
synonymized 17 nominal species with Argiope aetherea (Walckenaer), 14 of
them described between 1911 and 1915; and Grasshoff (1986), in revising
the African species of Neoscona, transferred 19 nominal taxa to N. subfusca
Diagnostic characters versus SMRS 311
(C.L. Koch), 13 to N. penicillipes (Karsch) and 11 to N. triangula (Keyserling).
The most impressive case, however, is that of the pisaurid spider,
Thalassius spinosissimus (Karsch), of which Sierwald (1987) identified not
less than 38 synonyms, 23 among them having been described in a single
paper by Roewer in 1954!
Just to add one example from earthworms, let us cite Aporrectodea calig-
inosa caliginosa (Savigny) with 18 synonyms and A. c. trapezoides (Duges)
with seven synonyms (Mrsic, 1991).
Some groups, especially land snails, experienced waves of excessive
splitting, then followed by strong, if not equally excessive, lumping.
However, these very high levels of synonymy are not simply a reflection
of the sloppy taxonomy of the past. One must seriously ask, whether the
pattern of variability we discover in nature actually allows, today, an easy
circumscription of species. As we shall try to demonstrate, animals do
actively contribute to taxonomists problems.
15.2 DIAGNOSTIC CHARACTERS VERSUS SPECIFIC MATE

RECOGNITION SYSTEMS (SMRS)
In the current practice, species discrimination in these groups often rests
on showy, complex morphological traits, such as the reproductive appa-
ratus of gastropods, the gonopods of male helminthomorph millipedes
and the palps of many male spiders. In most examples, however, there is
no direct evidence as to a function of these traits as specific mate recogni-
tion systems (SMRS). One of the few studies demonstrating the specifici-
ty of millipede gonopod structure is that of Barnett and Telford (1996). In
a laboratory experiment, Costa and Francescoli (1991) found anomalous
mating behaviour between interspecific couples of two sympatric sibling
species of the wolf spiders, Lycosa thorelli (Keyserling) and L. carbonelli
Costa and Capocasale, probably involving mechanical incompatibility
between the genitalia of the two species. Males of L. carbonelli proved to be
more discriminative between potential partners than male L. thorelli.
Apparently no post-copulatory isolating mechanism exists, however, as
demonstrated by the easy production of Fj hybrids.
In the recent malacological literature, there has been discussion over
the possible role of chirality (the left- or right-handedness of the shell) in
the process of speciation. This role has been affirmed by Gittenberger
(1988), but rejected by Johnson et al. (1990). Field evidence is not clear-cut.
In Partula suturalis (Pfeiffer) populations respectively monomorphic for
right- and left-handed shells are separated by steep clines (Johnson, 1987).
Johnson et al. (1987) studied geographic variation of allozymes at 23 loci
through the range of this species. Some correlation was found between
allozymic variation and direction of coil, but the geographic patterns of
allozymic variability did not closely overlap the distribution of the two
shell types.
In the animal groups we are discussing here, chemical and behaviour-
al SMRS are probably widespread, but have been very seldom exploited
by taxonomists, except for spiders (see below).
Sex pheromones have been demonstrated at work in spiders (Krafft
and Roland, 1980). In ixodid ticks, contact sex pheromones produced by
Dermacentor variabilis (Say) and D. andersoni (Stiles) are species-specific and
enable the courting males to discriminate conspecific females, with which
they mate, from heterospecific, which they ignore (Sonenshine, 1984).
Takeda and Tsuruoka (1979) described a head gland of a terrestrial snail
(Euhadra), consisting of fleshy tissue lying close to the base of the eye
stalks. This gland increases in size during the breeding season and mating
snails touch each other's gland during courtship. Tests with ethanol
extracts succeeded in eliciting sexual behaviour in conspecifics. Further
circumstantial evidence concerning the possible role of sex pheromones in
land snails was discussed by Croll (1983). A study of competitive interac-
tions between two introduced European land snails in South Australia
(Smallridge and Kirby, 1988) suggests that species-specific inhibitory sub-
stances in the mucus may aid in keeping the species apart.
Discriminating characters and potential SMRS are often differently
developed in otherwise similar, related groups. Within Opilionida, for
example, the species of Trogulidae are character-poor, even lacking those
conspicuous glandular organs that are present on the chelicerae in other
families of harvestmen, as in Ischyropsalididae and Nemastomatidae. In
these families these organs seem to play a key role in mating behaviour,
and hence in the delimitation of biological species (Martens, 1969a,b).
Other groups with uniform, less informative morphology include scorpi-
ons, pseudoscorpions, centipedes, symphylids, pauropods, and even two
groups of diplopods, i.e. Pselaphognatha and Pentazonia. Sometimes,
sizeable differences in usually showy traits are lacking in some species
groups, diagnostic characters being provided by other, usually more triv-
ial and less reliable characters, such as colour patterns and even size.
Examples are found for example in millipedes.
In spiders, behavioural traits are sometimes accorded prominent
importance in the delimitation of species (Blanke, 1986). Den Hollander
and Dijkstra (1974), for example, described a new species of wolf spider,
Pardosa vlijmi, on the basis of its courtship display. This behavioural char-
acter is, in practice, the only ground for distinguishing this ethospecies
from its closest relative, P. proxima (C.L. Koch).
Jumping spiders (Salticidae) are a good group in which to test the
importance of behavioural characters in species discrimination. Their
showy displays are well known and extensively documented. Other
Pitfalls of morphological evidence 313
spiders, however, also merit close attention. In Ontario, the spider
Philodromus rufus Walckenaer (Philodromidae) is represented by two sym-
patric forms, differing in the way their males use their legs when
approaching females. Only one of these forms exhibits conspicuous wav-
ing movements of the legs. An attentive morphometric study revealed
subtle differences between the two forms in the size of the cephalothorax,
in the relative size of the second pair of femora, in colour and pattern, as
well as in genitalia. When given a choice, these spiders mate preferential-
ly with individuals of the same form. According to Dondale (1964) these
differences in mating behaviour justify recognizing a separate species, Ph.
vibrans Dondale, but matters are probably more intricate, as shown by sub-
sequent papers of the same author.
Stridulatory organs used in courtship have been described in spiders
(Uetz and Stratton, 1982; Grabner and Thaler, 1986). Finally, construction-
al features of spider webs have been sometimes regarded as taxonomic
characters, but these seem to be useful only within local species-poor
faunas, not in a comprehensive survey of a whole genus (Eberhard, 1990).
15.3 PITFALLS OF MORPHOLOGICAL EVIDENCE: SEXUAL

DIMORPHISM, FUZZY DEVELOPMENTAL SCHEDULES
Practical problems sometimes arise because of sexual dimorphism. Quite
often, only one sex, generally the male, offers good taxonomic characters
for species identification. Examples are among the juliform millipedes
with their very complex male gonopods. In other examples, both sexes
exhibit useful diagnostic characters, but the sexual dimorphism is so
strong, that putting together male and female of the same species may
prove difficult for the museum taxonomist. Examples in spiders are
referred to by Coddington and Levi (1991).
Another problem, which does not occur in some other major animal
groups, such as pterygote insects, derives from the difficulty of differenti-
ating juveniles from adults. Sometimes, especially in groups with rampant
heterochrony, such as lithobiomorph centipedes (Minelli et al, 1996) the
differences between juveniles and adults of one and the same species are
comparable to those differentiating closely related species. Therefore,
many nominal species have been created for the juveniles of other
already-named species.
In still other examples, mature specimens of the same species are
sometimes intrinsically inhomogeneous because of the plasticity of devel-
opmental schedules. Coddington and Levi (1991) have stressed the
propensity of spiders to mature in any of several moults. Also julidan mil-
lipedes often may mature in any of several moults (e.g. in Enghoff et al.,
1993). However, this problem is lessened by the developmental
constraints acting in the morphogenesis of adult genitalia, whose dimen-
sions vary, within a given population, much less than overall body
dimensions, or proportions, despite the occurrence of allometry.
This problem of the taxonomic implications of relative developmental
control has been discussed in centipedes by Minelli and Bortoletto (1990)
in respect to a character widely used in the taxonomy of geophilomorph
centipedes, that is the (modal) number of body segments.
15.4 MORPHOLOGICAL AND MOLECULAR EVIDENCE:

CHROMOSOMES, ALLOZYMES, ISOZYMES AND mtDNA
Most kinds of taxonomic approaches have been tried, including karyolo-
gy (good tradition and good results in earthworms), allozymes (relatively
few, but for snails and earthworms), isozymes; and more recently also
mtDNA.
The usefulness of mtDNA for species discrimination is open to discus-
sion. On the one hand, coherent geographic trends in variation of mtDNA
polymorphisms have been confirmed at least for some species in one
group of snails, for example by Murray et al (1991) for Partula suturalis and
P. taeniata from the Society Islands. On the other hand, the diagnostic
value of mtDNA data is lessened by the confirmed presence of common
genotypes in what we would plainly regard as different species. One case
is discussed by Murray et al. (1991) in the same paper. One wonders how
far this overlapping of mtDNA variants between closely related species is
due to hybridization or introgression and how often it simply documents
an ancestral similarity surviving after cladogenesis.
Allozymes are sometimes better than morphology in discriminating
species, for example in earthworms of the genus Hormogaster (Cobolli
Sbordoni et al, 1992), in the lithobiomorph centipedes of the Eupolybothrus
fasciatus (Newport) group (Zanazzo et al., 1994) and among the sibling
species of Macrocheles mites (Pomponi et al, 1988). Quite often, however,
the biochemical evidence does not correlate with morphology. Examples
of more or less extensive mismatch between these different kinds of evi-
dence are provided by the land snails Medora (Giusti et al, 1986),
Solatopupa (Boato, 1988) and Albinaria (Kemperman and Degenaars, 1992).
We should be cautious, however, that different authors also have differ-
ent attitudes towards the data, sometimes according electrophoretic data
undue preference in assessing taxonomic relationships.
Speaking of genetic distance as determined from protein electrophoret-
ic data, attitudes of specialists towards the value of Nei's D estimates as a
clue to species status differ greatly. Some still regard as conspecific two
earthworm populations with a D distance up to 0.7, whereas D = 0.4 is ade-
quate for others to discriminate species of oribatid mites, and much less is
sometimes accepted as discriminating for snail species. Some examples fol-
low:
Morphological and molecular evidence 315
1. For Barros et. al. (1992), the most distantly related strains within the
Allolobophora molleri Rosa complex (earthworms) belong to one and
same species, in spite of D values larger than 0.88!
2. According to Bernini et al. (1988), in the oribatid genus Steganacarus, D
is about 0.4 in comparisons between related species, but only 0.00 to
0.036 between populations of S. magnus (Nicolet).
3. The values of D are very low (at most 0.03) between species of the land
snails, Samoana, from the Society Islands (Johnson et al, 1986).
Such figures can be profitably discussed in relation to Thorpe's (1983)
overall averages of D of about 0.40 (range 0.03 to >1.00) in interspecific
congeneric comparisons (900 cases) and D <0.10 in 98% of 7000 intraspe-
cific comparisons.
The study of allozymic variation in some species of the Australian
genus Cristilabrum of camaenid snails by Woodruff and Solem (1990) is
very illuminating. These species have extreme, narrowly parapatric
ranges. Nei's D, as calculated between C. primum Solem and its southern
neighbour C. grossum Solem, was not significantly different from zero
(range 0.026-0.051). Nevertheless, the authors rejected considering them
conspecific because of strong morphological differences (shell and geni-
talia) and the absolute lack of intermediates. Moreover, they point to the
lack of a necessary link between speciation and genetic differentiation. It
is worth noting here that C. primum is more differentiated genetically from
the other, northern neighbour, C. monodon Solem, with D = 0.17.
Even more impressive is evidence of genetic differentiation among
semi-terrestrial (amphibious) snails of the genus Novisuccinea by Hoagland
and Davis (1987). The widespread N. ovalis and the localized N. chittenan-
goensis (Pilsbry) cannot be distinguished electrophoretically, but they are
clearly distinct in terms of shell morphology, anatomy and ecology. On
the contrary, within what is a morphologically uniform N. ovalis, two pop-
ulation groups can be separated easily by electrophoresis.
In the rock-dwelling clausiliid snail, Solatopupa, allozymic data match
morphology in distinguishing four out of the five conventional species,
whereas the fifth, S. similis (Bruguiere), splits into at least two cryptic, but
genetically distinguishable species (Boato, 1988).
In the Ionian island of Kephallinia (Greece) there are several different
forms of the clausiliid land snail Albinaria. Traditional taxonomy, based on
morphological and biogeographical evidence, distinguishes four species
with a total of 12 subspecies. Two of the species are endemic to
Kephallinia. Kemperman and Degenaars (1992) studied allozymes in 42
populations representative of 11 of the 12 subspecies. The subspecies dis-
tinguished within the two non-endemic species are hardly differentiated
genetically. However, allozymic differences were found between sub-
species of the two endemic species. Moreover, each subspecies of these
endemic species showed unexpectedly conspicuous genetic resemblance
to different non-endemic species. These authors obviously question the
validity of the traditional taxonomic arrangement, but do not feel confi-
dent, for the time being, to suggest an alternative systematization.
Allozymes sometimes suggest that a morphologically distinguishable
species is nested within the diversity of its morphologically uniform but
cladistically paraphyletic closest relative. An example is the large Italian
earthworm, Hormogaster samnitica Cognetti, clearly derived from within
what is traditionally called H. pretiosa Michaelsen (Cobolli Sbordoni et. at,
1992). This species could be better regarded as a cluster of cryptic species,
estimates of divergence times within it being up to 3 800 000 years!
In the land snail Partula, the evolution of the morphological, elec-
trophoretic and mitochondrial phenotypes occur at variable rates, and
independently of one another (Murray et al, 1991). That is possibly a
widespread phenomenon.
15.5 HYBRIDS
Good evidence for natural hybrids is not extensive. For example, for all
myriapod taxa studied to date, the only substantiated case involves
Rhymogona cervina (Verhoeff) and the related Rh. silvatica (Rothenbuehler)
(Diplopoda). Hybrids between them have been demonstrated morpho-
logically, and verified electrophoretically, by Pedroli-Christen and Scholl
(1990). However, it is important to note that it was in a study of two natu-
rally hybridizing land snails (Cerion stevensoni and C. fernandina) that
Woodruff (1989) first noted the unique electromorphic variants
(hybrizymes) which sometimes characterize hybrid populations. A further
interesting example has been recently illustrated by Schilthuizen and
Gittenberger (1994).
15.6 DISPERSAL, DISTRIBUTION AND SPECIES

Geography, i.e. the spatial distribution of variability, offers perhaps the
biggest challenge to the (in principle, non-dimensional) biological species
concept. Our groups of terrestrial non-insect invertebrates do not escape
from this stricture. Sometimes, what is a couple of well-differentiated and
distinguishable species in a part of the common range turns out to be an
undifferentiated continuum within a polymorphic species in another part
of the same range. That is, at least, what seems to emerge from years of
investigations on the European land snails known as Cochlicopa lubrica
(O.F. Miiller), C. lubricella (Porro), C. nitens (Gallenstein) and C. repentina
Hudec. Apparently only one species of this group occurs in Italy (Giusti
and Manganelli, 1992), but it splits into two or more divergent phenotypes
without intermediates in Central Europe.
Dispersal, distribution and species 317
Our perception of species barriers is affected by patterns of geographic
distribution, itself reflecting dispersal power. Critical taxa in different
groups are often co-occurring in the same areas (e.g. the Western and
Eastern ends of the Alps). Non-insect terrestrial invertebrates include
extremely low dispersers, for example many with a subterranean lifestyle.
Among earthworms there are a few deep-burrowing species (e.g. Eophila
tellinii Rosa). There are many subterranean species-rich taxa of isopods, a
few spiders, many pseudoscorpions, centipedes and millipedes. These
allopatric taxa are usually easy to recognize, but their species status is
often not so obvious. At the opposite end of the spectrum, easily dispersed
and seemingly coherent species include anthropochorous isopods, cen-
tipedes and millipedes, well-documented in the faunas of islands, e.g.
Iceland, St. Helena, Hawaii, etc.
Very interesting patterns of species multiplication are revealed by insu-
lar biotas. Insular species swarms are known among Macaronesian milli-
pedes of the genera Cylindroiulus, Dolichoiulus and Acipes (Enghoff 1982,
1983a,b, 1992a,b; Enghoff and Baez, 1993) and Hawaiian spiders of the
family Tetragnathidae (Okuma, 1988; Gillespie, 1991, 1992, 1993, 1994).
Some 29 species of Cylindroiulus occur on Madeira; 27 of them are endem-
ic to that island, one is common to the Azores and another to the Desertas.
Together with one species from Porto Santo and another from the
Canaries, these species form a monophyletic group within this large
genus. As many as 15 species occur at a single location, but with clear
microhabitat differentiation. Most microhabitats, such as logs or the inter-
face between soil and litter, have more than one characteristic species but,
in most cases, such narrowly coexisting species are of different sizes. Also
in terms of cladogenesis, size is the most conspicuous trait differentiating
pairs of sister species. In particular, each of the pairs gemellus/madeirae and
julipes/lundbladi includes two species essentially similar except for size.
Dolichoiulus is a millipede genus with 52 known species. Total lengths of
mature specimens range between 7 and 49mm, with diameter of their
cylindrical body between 0.56 and 3.65 mm. The species differ in details of
eye number, leg number and size, and sculpture, but particularly in the
shape of the complex male gonopods. Of these 52 species, 46 live in the
Canary Islands: 20 on Tenerife, 10 on Gran Canaria, nine on La Gomera and
one to three species each on the remaining islands of the archipelago.
Nearly all species are confined to one island. On certain islands, the distrib-
ution of the individual species is all but random: closely related species
occur in micro-allopatry. The so-called archipelago effect (cf. Carlquist, 1974)
does not seem to have helped in generating this large number of endemics.
From the Canary Islands large numbers of congeneric species are also
known for three spider genera - Dysdera (Dysderidae) with 46 species,
Spermophorides (Pholcidae) with 23 species and Oecobius (Oecobiidae) with
34 species (Wunderlich, 1992).
More widespread are the patterns of continental insularity, e.g. in the
extensive cave systems along the Southern margin of the Alps (with many
pseudoscorpions, millipedes and spiders), in the lava tubes of the Canary
Islands (spiders: Dysdera), but also on mountain tops (e.g. arctic-alpine
spiders). Extreme examples are the camaenid land snails in north-west
Australia, with 27 species in three genera along the 52-km-long limestone
Ningbing Range. The average range for one species is 1 km2 and species
ranges are mostly allopatric.
15.7 THE TEMPORAL DIMENSION

Many studies approaching the time dimension of species have been made
in terrestrial snails. Sometimes (e.g. Mandarina from the Bonin Islands)
there is evidence of polymorphism and hybridization patterns extending
through time since the Late Pleistocene (Chiba, 1993).
For the groups we are discussing, evidence from fossils is restricted, in
practice, to terrestrial snail shells, which are common at many
Quaternary sites. These shells can be exploited for palaeoecological
reconstructions, but the limited morphological evidence they provide
has but scanty information content, in comparison with what can be
obtained from many Quaternary insect remains (Elias, 1994). We cannot
hope to get from these materials any sound evidence as for delimitation
of species. However, in individual cases this fossil evidence may prove
crucial in choosing among alternative hypotheses. Gould and Woodruff
(1990) studied an example of an 'area effect' within the land snail Cerion
columna Pilsbry and Vanatta on Great Inagua, Bahamas. A local popula-
tion of this snail exhibits unusually squat and flat-topped shells, very dif-
ferent from those of the common kind, occurring both east and west of
this area. Gould and Woodruff explained that area effect as the result of
introgression from propagules of a related species, C. dimidiatum
(Pfeiffer) from Cuba. This explanation, already suggested by multivari-
ate morphometry, was most convincingly supported by the finding of
fossils of C. dimidiatum cemented into soil crusts within the region inhab-
ited by the unusual C. columna.
15.8 UNIPARENTALS
Animals with uniparental reproduction offer particularly intractable prob-
lems to the taxonomist. Parthenogenesis in earthworms is common, as is
uniparental reproduction in land snails both as parthenogenesis and as
autogamy. Further examples of parthenogenesis are known among
isopods (e.g. Trichoniscus pusillus Brandt), centipedes (some Lamyctes), mil-
lipedes (e.g. Nemasoma varicorne C.L. Koch) and scattered groups of arach-
nids, from scorpions to mites.
Uniparentals 319
In the opinion of Giusti and Manganelli (1992: 158) in genera such as
Vallonia, Lauria, Columella, Pagodulina, Chondrina, Abida, etc., 'the distinc-
tion between an ecophenotype and a subspecies, or a subspecies and a
species, is nearly always left to the personal philosophical or practical con-
victions of the individual researchers'. In Vallonia, for example, shell mor-
phology allows the recognition of about seven European species. Of these
some are well defined; others, however, live in mixed populations where
intermediate phenotypes occasionally occur. To disentangle these messy
groups, malacologists cannot resort to the usual strong morphological evi-
dence of genitalia, because nearly all of these Vallonia are completely
devoid of the penial complex and reproduce through parthenogenesis or
self-fertilization. In other genera such as Abida and Chondrina, however,
things are not so bad, in spite of the variability of some species (E.
Gittenberger, personal communication).
Bisexual and parthenogenetic forms coexist in what are currently
called Nemasoma varicorne and Polyxenus lagurus, two widespread
European millipedes.
Variability in parthenogens has been extensively studied in terrestrial
oligochaetes. In Eiseniella tetraedra and Dendrobaena octaedra, two obligate-
ly parthenogenetic earthworms, there is extensive morphological varia-
tion, not correlated with enzyme patterns (Terhivuo et al, 1994; Terhivuo,
1988). In other earthworms, such as the obligate parthenogenetic
Octolasion tyrtaeum (Orley) and O. cyaneum (Savigny) the number of clones
occurring in populations at the northern edge of the range (Eastern
Fennoscandia) is high (24 clones found in 238 individuals from eight local-
ities) in the former, spontaneously dispersing species, but low (only two
clones found in 134 individuals from four localities) in the other, anthro-
pochorous species. But polymorphism has been documented in other
obligate parthenogens, such as the enchytraeid worm Fridericia striata
(Levinsen), whose sexual ancestor is unknown. In this animal, the diploid
chromosome number is restored by terminal fusion, i.e. through the
fusion of the products of the second meiotic division. Of 27 loci studied
electrophoretically by Christensen et al. (1989), 13 were polymorphic, all
but one in homozygotes. Two clones even differed in all polymorphic loci,
probably due to the polyphyletic origin of this parthenogenetic agamo-
species. But what is an agamospecies? According to Ghiselin (1984)
agamospecies are 'heaps of leaves that have fallen off the tree that gave
rise to them'. These clones, recognizable within a morphologically uni-
form parthenogen, may differ in subtle habitat requirements, for example
in the woodlouse Trichoniscus pusillus and the oligochaetes Dendrobaena
octaedra and Fridericia galba (Hoffmeister) (Christensen and Noer, 1986;
Terhivuo and Saura, 1990; Christensen et al, 1992).
In one group, the Desmonomata, a taxon of 'lower' oribatid mites, the-
lytokous parthenogenesis is perhaps the only manner of reproduction,
thus presenting us with a taxonomic and evolutionary puzzle similar to
that of the better-known bdelloid rotifers. According to conventional tax-
onomy, this group consists of seven families, 32 genera and about 400
nominal species (Palmer and Norton, 1992). Males may occur, in some of
these mites, but they seem to be non-functional. Variability of these
diploid parthenogens is low, but no population found comprised only one
clone. No close sexual relative of these uniparental mites is known point-
ing to a remote origin of parthenogenesis. By implication, Palmer and
Norton (1992) believe that genetic variability in these forms did arise after
the loss of sex.
15.9 EPILOGUE
To sum up, difficulties with species limits are common among terrestrial
non-insect invertebrates. However, these groups, also offer many oppor-
tunities for demonstrating the usefulness of biological species concepts.
We hope not to join the ranks of those authors, who, according to
Brignoli (1988) are not too rare among the taxonomists of spiders and
other groups, whose species identification criteria heavily rely on topo-
graphical data of museum labels, typological numerical values, and the
good luck of the taxonomist!
Acknowledgements
We are grateful to Michael F. Claridge for inviting us to the Systematics
Association Symposium on 'The Units of Biodiversity: Species in Practice'.
We are also very indebted towards Henrik Enghoff (Copenhagen),
Edmund Gittenberger (Leiden) and Konrad Thaler (Innsbruck) for useful
comments on a previous draft. The work of one of the authors (A.M.) has
been partly supported by grants of the Italian National Research Council
(CNR) and the Italian Ministry of University and Scientific and
Technological Research (MURST).
15.10 REFERENCES
Barnett, N. and Telford, S.R. (1996) Sperm competition and the evolution of milli-
pede gentalia. Memoires du Museum national d'Histoire naturelle, Paris, 169,331-9.
Barros, F., Carracedo, A. and Diaz Cosin, D.J. (1992) Karyological and allozymic
study of Allolobophora molleri complex and related taxa. Initial results. Soil
Biology and Biochemistry, 24,1241-5.
Bernini, F., Avanzati, A.M. and Petrucci, R. (1988) Genetic variation in the soil
phthiracarid mite Steganacarus (S.) magnus (Nicolet, 1855) (Acarida, Oribatida)
(Notulae Oribatologicae XL). Zeitschrift fur zoologische Systematik und
Evolutionsforschung, 26,104—13.
Bieler, R. (1992) Gastropod phylogeny and systematics. Annual Review of Ecology
and Systematics, 23, 311-38.
References 321
Blandin, P. (1977) Le probleme de 1'espece chez les araignees, in Les problemes de I'e-
spece dans le regne animal. Tome II (Memoire n. 39) (eds C. Bocquet, J. Genermont
and M. Lamotte), Societe Zoologique de France, Paris, pp. 13-56.
Blanke, R. (1986) Homologien in Fortpflanzungsverhalten von Krenzspinnen
(Araneae, Araneidae) und deren Interpretation im Kontext von Systematik und
der Existenz von Artbarrieren. Adas do X Congreso Internacional de Aracnologia,
Jaca/Espana, 1, 69-94.
Boato, A. (1988) Micro evolution in Solatopupa landsnails (Pulmonata:
Chondrinidae): generic diversity and founder effects. Biological Journal of the
Unman Society, 34, 327-48.
Brignoli, P.M. (1988) Ragni, specie e filosofia, in // problema biologico della specie,
Mucchi, Modena, pp. 293-307.
Carlquist, S. (1974) Island Biology, Columbia University Press, New York and
London.
Chiba, S. (1993) Modern and historical evidence for natural hybridization between
sympatric species in Mandarina (Pulmonata: Camaenidae). Evolution, 47,
1539-56.
Christensen, B. and Noer, H. (1986) Spatial and temporal components of genetic
variation in triploid parthenogenetic Trichoniscus pusillus (Isopoda: Crustacea).
Hereditas, 105, 277-85.
Christensen, B., Hvilson, M.M. and Pedersen, B.V. (1989) On the origin of clonal
diversity in parthenogenetic Fridericia striata (Enchytraeidae, Oligochaeta).
Hereditas, 110, 89-91.
Christensen, B., Hvilson, M. and Pedersen, B.V. (1992) Genetic variation in coex-
isting sexual diploid and parthenogenetic triploid forms of Fridericia alba
(Enchytraeidae, Oligochaeta) in a heterogeneous environment. Hereditas, 117,
153-62.
Cobolli Sbordoni, M., De Matthaeis, E., Alonzi, A., Mattoccia, M., Omodeo, P. and
Rota, E. (1992) Speciation, genetic divergence and palaeogeography in the
Hormogastridae. Soil Biology and Biochemistry, 24 (12) (special issue), 1213-21.
Coddington, J.A. and Levi, H.W. (1991) Systematics and evolution of spiders
(Araneae). Annual Review of Ecology and Systematics, 22, 565-92.
Costa, F.G. and Francescoli, G. (1991) Analyse experimental de 1'isolement repro-
ductif entre deux especes jumelles et sympatriques d'araignees: le Lycosa thorel-
li (Keyserling) et le Lycosa carbonelli Costa et Capocasale. Canadian Journal of
Zoology, 69,1768-76.
Croll, R.P. (1983) Gastropod chemoreception. Biological Reviews, 58, 293-319.
den Hollander, J. and Dijkstra, H. (1974) Pardosa vlijmi sp. nov., a new ethospecies
sibling of Pardosa proxima (C.L. Koch, 1848) from France, with description of
courtship display (Araneae: Lycosidae). Beaufortia, 22,57-65.
Dondale, C.D. (1964) Sexual behaviour and its application to a species problem in
the spider genus Philodromus (Araneae: Thomisidae). Canadian Journal of
Zoology, 42,817-27.
Eberhard, W.G. (1990) Function and phylogeny of spider webs. Annual Review of
Ecology and Systematics, 21, 341-72.
Elias, S.A. (1994) Quaternary Insects and their Environment, Smithsonian Institution
Press, Washington and London.
Enghoff, H. (1982) The millipede genus Cylindroiulus on Madeira - an insular
species swarm. Entomologica Scandinavica, suppl., 18,1-142.
Enghoff, H. (1983a) Acipes - a Macaronesian genus of millipedes (Diplopoda,
Julida: Blaniulidae). Steenstrupia, 9,137-79.
Enghoff, H. (1983b) Adaptive radiation of the millipede genus Cylindroiulus on
Madeira: habitat, body size and morphology (Diplopoda, Julida: Julidae). Revue
d'Ecologie et Biologie du Sol, 20, 403-15.
Enghoff, H. (1992a) Dolichoiulus - a mostly Macaronesian multitude of millipedes
with the description of a related new genus from Tenerife, Canary Islands
(Diplopoda, Julida, Julidae). Entomologica Scandinavian, suppl, 40,1-158.
Enghoff, H. (1992b) Macaronesian millipedes (Diplopoda) with emphasis on
endemic species swarms on Madeira and the Canary Islands. Biological Journal
of the Linnean Society, 46,153-61.
Enghoff, H. and Baez, M. (1993) Evolution of distribution and habitat patterns in
endemic millipedes of the genus Dolichoiulus (Diplopoda: Julidae) on the
Canary Islands, with notes on distribution patterns of other Canarian species
swarms. Biological Journal of the Linnean Society, 49, 277-301.
Enghoff, H., Dohle W. and Blower G. (1993) Anamorphosis in millipedes
(Diplopoda) - the present state of knowledge with some developmental and
phylogenetic considerations. Zoological Journal of the Linnean Society, 109,103-234.
Ghiselin, M.T. (1984) Narrow approaches to phylogeny: a review of nine books on
cladism, in Oxford Surveys in Evolutionary Biology (eds R. Dawkins and M.
Ridley), Oxford University Press, Oxford, pp. 209-22.
Gillespie, R.G. (1991) Hawaiian spiders of the genus Tetragnatha: I. Spiny leg clade.
Journal ofArachnology, 19,174-209.
Gillespie, R.G. (1992) Hawaiian spiders of the genus Tetragnatha: 2. Species from
natural areas of windward East Maui. Journal ofArachnology, 20,1-17.
Gillespie, R.G. (1993) Biogeographic pattern of phylogeny in a clade of endemic
Hawaiian spiders (Araneae: Tetragnathidae). Memoirs of the Queensland
Museum, 33, 519-26.
Gillespie, R.G. (1994) Hawaiian spiders of the genus Tetragnatha: 3. Tetragnatha
acuta clade. Journal ofArachnology, 22,161-8.
Gittenberger, E. (1988) Sympatric speciation in snails: a largely neglected model.
Giusti, F. and Manganelli, G. (1992) The problem of the species in malacology after
clear evidence of the limits of morphological systematics. Proceedings of the ninth
International Malacological Congress, pp. 153-72.
Giusti, F., Grappelli, C, Manganelli, G., Fondi, R. and Bullini, L. (1986) An attempt
of natural classification of the genus Medora in Italy and Jugoslavia on the basis
of conchological, anatomical and allozymic characters (Pulmonata:
Clausiliidae). Lavori della Societtd Italiana di Malacologia, 22, 256-341.
Gould, SJ. and Woodruff, D.S. (1990) History as a cause of area effects: an illustra-
tion from Cerion on Great Inagua, Bahamas. Biological Journal of the Linnean
Society, 40, 67-98.
Grabner, R. and Thaler, K. (1986) Bin weiteres Stridulationsorgan bei manchen
Arten der mughi-Gruppe der Gattung Lepthyphanthes (Araneae, Linyphiidae)?
Mitteilungen der Schweizer Entomologischen Gesellschaft, 59,15-21.
Grasshoff, M. (1986) Die Radnetzspinnen-Gattung Neoscona in Afrika (Arachnida:
Araneae). Zoologische Wetenschaapen, 250,1-123.
References 323
Hoagland, K.E. and Davis, G.M. (1987) The succineid snail fauna of Chittenango
Falls, New York: taxonomic status with comparisons to other relevant taxa.
Proceedings of the Academy of Natural Sciences, Philadelphia, 139,465-526.
Holman, E. (1987) Recognizability of sexual and asexual species in rotifers.
Johnson, M.S. (1987) Adaptation and rules of form: chirality and shape in Partula
suturalis. Evolution, 41,672-5.
Johnson, M.S., Murray, J. and Clarke, B. (1986) High genetic similarities and low
heterozygosities in land snails of the genus Samoana from the Society Islands.
Malacologia, 27, 97-106.
Johnson, M.S., Murray, J. and Clarke, B. (1987) Independence of genetic subdivi-
sion and variation for coil in Partula suturalis. Heredity, 58, 307-13.
Johnson, M.S., Clarke, B. and Murray, J. (1990) The coil polymorphism in Partula
suturalis does not favor sympatric speciation. Evolution, 44,459-64.
Kemperman, T.C.M. and Degenaars, G.H. (1992) Allozyme frequencies in Albinaria
(Gastropoda: Pulmonata: Clausiliidae) from the Jonian Islands of Kephallinia
and Ithaka. Malacologia, 34, 33-61.
Krafft, B. and Roland, Ch. (1980) Quelques remarques au sujet de la communica-
tion chimique chez les araignees. Comptes Rendues du V Colloque d'Arachnologie,
9,129-34.
Levi, H.W. (1983) The orb-weaver genera Argiope, Gea and Neogea from the
Western Pacific region (Araneae: Araneidae, Argiopinae). Bulletin of the Museum
of Comparative Zoology, Harvard, 150, 247-338.
Martens, J. (1969a) Die Sekretdarbietung wahrend des Paarungsverhaltens von
Ischyropsalis C.L. Koch (Opiliones). Zeitschrift fur Tierpsychologie, 26, 513-23.
Martens, J. (1969b) Die Abgrenzung von Biospezies auf biologisch-ethologischer
und morphologischer Grundlage am Beispiel der Gattung Ischyropsalis C.L.
Koch 1839 (Opiliones, Ischyropsalididae). Zoologische Jahrbucher, Abteilung fur
Systematik, 96,133-264.
Minelli, A. and Bortoletto, S. (1990) Segmentation in centipedes and its taxonomic
implications, in Proceedings of the 7th International Congress of Myriapodology (ed.
A. Minelli), E.J. Brill, Leiden, pp. 81-8.
Minelli, A., Negrisolo, E. and Fusco, G. (1996) Developmental trends in the post-
embryonic development of lithobiomorph centipedes. Memoires du Museum
National d'Histoire Naturelle, Paris, 169,351-8.
Mrsic, N. (1991) Monograph on earthworms (Lumbricidae) of the Balkans, Slovenska
Akademija Znanosti in Umetnosti, Ljubljana.
Murray, J., Stine, O.C. and Johnson, M.S. (1991) The evolution of mitochondrial
DNA in Partula. Heredity, 66,93-104.
Okuma, C. (1988) Redescriptions of the Hawaiian spiders of Tetragnatha described
by Simon (Araneae, Tetragnathidae). Journal of the Faculty of Agriculture, Kyushu
University, 33, 77-86.
Palmer, S.C. and Norton, R.A. (1992) Genetic diversity in thelytokous oribatid
mites (Acari; Acariformes: Desmonomata). Biochemical Systematics and Ecology,
20,219-31.
Pedroli-Christen, A. and Scholl, A. (1990) Ecological and genetic studies on parap-
atric Rhymogona silvatica (Roth.) and R. cervina Verh. (Diplopoda:
Craspedosomatidae) with special reference to hybrid populations in a zone of
contact. Revue Suisse de Zoologie, 97, 349-59.
Pomponi, G., Cicolani, B. and Bullini, L. (1988) Ricerche elettroforetiche su cinque
specie del genere Macrocheles (Acarina: Mesostigmata). Atti del XV Congresso
nazionale italiano di Entomologia, L'Aquila 13-17 giugno 1988, pp. 233-7.
Roewer, C.F. (1954) Araneae Lycosaeformia I. (Agelenidae, Hahniidae, Pisauridae).
Exploration du Pare National de I'Upemba, Bruxelles, 30,1-420.
Schilthuizen, M. and Gittenberger, E. (1994) Parallel evolution of an sAat-
'hybrizyme' in hybrid zones in Albinaria hyppoliti (Boettger). Heredity, 73,244-8.
Sierwald, P. (1987) Revision der Gattung Thalassius (Arachnida, Araneae,
Pisauridae). Verhandlungen des naturwissenschaftlichen Vereins Hamburg, 29,51-142.
Smallridge, M.A. and Kirby, G.C. (1988) Competitive interactions between the land
snails Theba pisana (Miiller) and Cernuella virgata (Da Costa) from South
Australia. Journal ofMolluscan Studies, 54,251-8.
Sonenshine, D.E. (1984) Tick pheromones, in Current Topics in Vector Research (ed.
K.F. Harris), Praeger, New York, pp. 225-63.
Takeda, N. and Tsuruoka, H. (1979) A sex pheromone secreting gland in the ter-
restrial snail Euhadra peliomphala. Journal of Experimental Zoology, 207,17-26.
Terhivuo, J. (1988) Morphological and morphometric variation of the partheno-
genetic earthworm Dendrobaena octaedra (Sav.) (Oligochaeta, Lumbricidae) in
Eastern Fennoscandia. Annales Zoologici Fennici, 25, 303-20.
Terhivuo, J. and Saura, A. (1990) Allozyme variation in parthenogenetic
Dendrobaena octaedra (Oligochaeta: Lumbricidae) populations of Eastern
Fennoscandia. Pedobiologia, 34,113-39.
Terhivuo, J., Saura, A. and Hongell, K. (1994) Genetic and morphological variation
in the parthenogenetic earthworm Eiseniella tetraedra (Sav.) (Oligochaeta:
Lumbricidae) from South Finland and North Norway. Pedobiologia, 38, 81-96.
Thorpe, J.P. (1983) Enzyme variation, genetic distance and evolutionary diver-
gence in relation to levels of taxonomic separation, in Protein Polymorphism:
Adaptive and Taxonomic Significance (eds G.S. Oxford and D. Rollinson),
Academic Press, London, pp. 131-52.
Uetz, G.W. and Stratton, G.E. (1982) Acoustic communication and reproductive
isolation in spiders, in Spider Communication: Mechanisms and Ecological
Significance (eds P.N. Witt and J.S. Rogner), Princeton University Press,
Princeton, pp. 123-59.
Woodruff, D.S. (1989) Genetic anomalies associated with Cerion hybrid zones: the
origin and maintenance of new electromorphic variants called hybryzymes.
Biological Journal of the Linnean Society, 36,281-94.
Woodruff, D.S. and Solem, A. (1990) Allozyme variation in the Australian
camaenid land snail Cristilabrum primum: a prolegomenon for a molecular phy-
logeny of an extraordinary radiation in an isolated habitat. The Veliger, 33,
129-39.
Wunderlich, J. (1992) Die Spinnen-Fauna der makaronesischen Inseln. Taxonomie,
Okologie, Biogeographie und Evolution. Beitrage zur Araneologie, 1(1991).
Zanazzo, G., Filippucci, M.G., Boato A., and Minelli, A. (1994) Allozymic and kary-
ological evidence concerning Italian populations of the Eupolybothrus fasciatus-
group (Chilopoda: Lithobiomorpha). Zoologischer Anzeiger, 232, 77-95.
16
Species concepts in systematics
and conservation biology - an
ornithological viewpoint
/. Cracraft
Contacting address: Department of Ornithology, American Museum of Natural History,
Central Park West at 79th Street, New York 10024, USA
ABSTRACT
The biological species concept (BSC) has dominated within ornitholo-
gy since the 1930s, but the past decade has seen increased application
of the phylogenetic species concept (PSC). The central role of species
concepts is to delineate the units of nature and thus provide the essen-
tial framework for understanding biological diversity. The PSC does
this more objectively than the BSC. Many conservation biologists, par-
ticularly those who manage in situ and ex situ breeding programmes,
have recognized that the BSC is inappropriate for this task. Their solu-
tion, 'evolutionary significant units (ESUs)', has gained wide support
within the conservation community, yet it has significant problems.
There is no general support on how to define ESUs nor apply the con-
cept objectively. Perhaps more important, ESUs have no status within
formal taxonomy, hence they have no standing within those legal
instruments designed to conserve and use sustainably biological
diversity. Phylogenetic species, as basal diagnosable units, are effec-
tive functional equivalents of ESUs, have standing in formal taxono-
my, and have many advantages over biological species when applied
to conservation and management problems. It is suggested that the
concept of ESU be abandoned and that the PSC become the taxonom-
ic currency of conservation biology.
16.1 INTRODUCTION
The discipline of ornithology has had a large influence on the debates over
species concepts. Ornithologists such as O. Kleinschmidt and E. Hartert,
326 Species concepts - an ornithological viewpoint
and later Erwin Stresemann and Bernard Rensch, were among the first to
see the importance of species concepts and begin the shift toward the so-
called polytypic, biological concept. In their midst was a young systema-
tist, Ernst Mayr, who became the leading advocate for the biological
species concept (BSC) and who remains an active partisan over 50 years
later (Mayr, 1942,1963,1970,1992,1993). Since Mayr's influential work of
1942, many avian systematists whose primary interest was geographic
variation and speciation analysis adopted the BSC (Short, 1969; Mayr and
Short, 1970; Selander, 1971; Bock, 1987; Haffer, 1992).
In recent years support for the BSC within ornithology has waned as
systematists have adopted a phylogenetic species concept (PSC; Cracraft,
1983,1989; McKitrick and Zink, 1988). Although the latter is not strictly a
cladistic concept - in fact, there has been debate within cladistics over
species concepts - the notion of phylogenetic species has found strong
support among those avian systematists who see hypotheses of the histo-
ry of taxic differentiation as being at the conceptual centre of the analysis
of geographic variation and speciation.
The purpose of this chapter is not to revisit the debates over species
concepts within ornithology (see Haffer, 1992, for a review from the per-
spective of a supporter of the BSC). The reasons for adopting the phylo-
genetic species concept within ornithology are compelling and have been
discussed elsewhere (Cracraft, 1983, 1989; McKitrick and Zink, 1988; see
also many papers cited below). The first section, instead, touches on some
conceptual and linguistic arguments that tend to obfuscate understanding
of the differences and implications of adopting alternative species con-
cepts, no matter which group of organisms might be considered. The sec-
ond section attempts to dispel several myths about the phylogenetic
species concept. Following this, a section is devoted to a brief comparison
of the BSC and PSC to illustrate their different implications for systematic
and evolutionary biology. Finally, the last section examines the role of
species concepts within conservation biology. Biologists have begun to
recognize the implications of alternative species concepts when consider-
ing the units underlying conservation action and sustainable develop-
ment, yet systematists have played little role in shaping this discussion.
16.2 SPECIES IN THEORY AND PRACTICE
16.2.1 The language of discourse

One underlying reason for the ongoing debates over species concepts
and definitions, and what they mean theoretically and empirically, is the
structure of the discourse itself. How species might be recognized
becomes confused with how species might be defined. The evidence
used in recognizing species is often taken to be more or less equivalent
to the definition of a species. Biologists speak of morphological species,
Species in theory and practice 327
genetic species, genotypic species, or behavioural species as if they are
conceptually on the same level as biological species, phylogenetic
species, or evolutionary species. The latter, more theoretical, definitions
are often conflated with the evidence perceived to be at the heart of
applying those definitions. Thus, biological species are taken to be genet-
ic whereas phylogenetic species are characterized as being morphologi-
cal. Inasmuch as genetic, morphological, behavioural, or other kinds of
character data can be used to delimit species boundaries within the context
of any view of species, such language serves only to confuse the dia-
logue over alternative species concepts. Notions of genetic species, geno-
typic species, or morphological species are inherently confusing and
often non-sensical.
Equally unhelpful is the use of pejorative language in characterizing
some species concepts as being typological, non-populational, or non-
biological as is sometimes done by advocates of the BSC (Mayr, 1992;
Haffer, 1992). Proponents of all species concepts use populational think-
ing inasmuch as they would not knowingly consider placing different
sexes, morphs, or different stages of a life-cycle in separate species.
Mistakes can happen under the guise of any species concept, but the bio-
logical species concept, simply because the word 'biological' appears in its
name, is not inherently more biological, populational, or genetic than is
the phylogenetic species concept. Likewise, the latter is not more inher-
ently morphological. It is also not useful to characterize some concepts as
being purely theoretical, in contrast to others that might be said to be prac-
tical or empirical. All definitions, no matter how empirical they may
sound, are theory laden and rely on some conceptual understanding of
other terms.
16.2.2 The structure of species definitions

Unless species concepts are used to individuate real, discrete entities in
nature, they will have little or no relevance for advancing our under-
standing of the structure and function of biological phenomena involving
those things we call species (Cracraft, 1987). Debates over species concepts
rarely include much discussion about this. Perhaps systematists just take
for granted that species are discrete entities, but this is not evident in some
of the debates (nor has it been true in practice, especially within palaeon-
tology where a nominalist-like view of species has long prevailed). If
species are not considered to be discrete real entities - and holding this
view does not mean that delimiting boundaries will always be straight-
forward or that the boundaries cannot be fuzzy depending on the spa-
tiotemporal scale of the observer - then it implies that evolutionary and
systematic biology would be based largely on units that are fictitious,
whose boundaries, if drawn, are done so arbitrarily. It would also mean
that most, if not all, of the processes that we ascribe to species are concoctions
of the mind and have no objective reality. Entities of postulated processes
must be real and discrete if those processes are to have much meaning.
Having a theory about the behaviour of electrons, for example, makes no
sense if electrons do not exist as discrete entities; the same is true for
species. The notion of discreteness, clearly, must be contextualized with
respect to a certain spatiotemporal frame of reference, and different per-
ceptions of the latter can lead biologists to see species as being discrete,
real things, on the one hand, or as arbitrary segments of evolutionary con-
tinua, on the other (Cracraft, 1987).
Unless a species concept can be used to individuate real-world entities,
that concept will have limited utility for systematists getting on with their
task of sorting out and understanding biological diversity. Both the BSC and
PSC are meant to be guides to the practice of recognizing species, but they
do not do so in the same way or equally effectively. Some biologists think
the primary task of species concepts is to understand process-level phe-
nomena. Important though that task may be, the view taken here is that the
central role of species concepts is to delineate the units of nature and thus
provide the essential framework for understanding life's diversity.
Species definitions should have three key elements if they are going to
be useful for systematic practice:
1. They need to mention or imply reproductive cohesion (be popu-
lational) in order to provide a conceptual basis for including males and
females of the same population in the same species.
2. They must have some notion of diagnosability so that populations
or groups of populations can be distinguished one from another.
3. They must include some criterion for ranking these populations at
the species-level as opposed to some other level of the Linnean hierarchy.
The definition of biological species - a group of interbreeding natural
populations that is reproductively isolated from other such groups (Mayr,
1963) - has the first two of these key elements. 'Interbreeding natural
populations' implies reproductive cohesion, and the phrase 'reproduc-
tively isolated' provides for some basis of diagnosability (if one assumes
there are character-based differences that lead to reproductive incompati-
bility). Yet the definition is inherently flawed because it does not provide
any specific framework for ranking. Proponents would like to use the con-
junction of 'interbreeding' and 'reproductively isolated' to rank an entity
at the species-level. The definition, however, sets no bound on where
species limits might be drawn, only that it will stop at some point of
'reproductive isolation'. Yet, few supporters of the BSC would claim that
all populations capable of interbreeding should be included in the same
biological species, and considerable difference of opinion exists among
supporters of the BSC over how much (or little) interbreeding is necessary
and/or sufficient to justify uniting populations in the same species. In fact,
Phylogenetic species concept: myths and misrepresentations 329
it is the difficulty in reconciling these two components of the BSC that has
led many biologists over the years to abandon it in practice.
In contrast, the definition of phylogenetic species - the smallest popu-
lation or group of populations within which there is a parental pattern of
ancestry and descent and which is diagnosable by unique combinations of
character-states (Eldredge and Cracraft, 1980; Nelson and Platnick, 1981;
Cracraft, 1983; Nixon and Wheeler, 1990) - has all three key elements. The
phrase 'parental pattern of ancestry and descent' implies reproductive
cohesion over time; the element of diagnosability is specifically men-
tioned; and the statement referring to the smallest population establishes
the basis for ranking (that is, the boundary to species limits is the smallest
population or group of populations that is diagnosably distinct).
16.3 THE PHYLOGENETIC SPECIES CONCEPT: SOME MYTHS

AND MISREPRESENTATIONS
Growing support for the PSC within systematics has resulted in various
reactions from those - primarily non-systematists - whose traditional alle-
giance has been to the BSC. This reaction has included a significant
amount of misunderstanding about the PSC: what it is, how it might be
applied, and what might be the consequences of using it broadly within
systematics and evolutionary biology. Several of the most important mis-
understandings deserve discussion.
As already noted, linguistic characterization can go far in casting asper-
sions on a particular species concept without actually facilitating a rational
dialogue about it. Describing the PSC as being purely morphological or
non-populational (Mayr, 1992), and therefore typological, for example,
does little to further scientific discourse about species concepts because
the description is patently false. The PSC is defined in terms of popula-
tions and their diagnosability, and nowhere in the definition does the
word morphological appear. Indeed, systematists will want to use all rel-
evant data when determining species limits under the PSC. The fact
remains that most of the information available to systematists - no matter
what definition they adhere to - is morphological, yet this does not make
a definition, including even the BSC, non-biological or non-populational.
A strength of the PSC is that it can be applied using only morphological
data. The PSC is centred around the notion of diagnosability, which can
be inferred more or less directly from morphological data; the BSC, in con-
trast, relies on an understanding of reproductive isolation, which can be
inferred only very indirectly from morphological data.
The PSC is sometimes mistakenly interpreted as being a cladistic con-
cept (Haffer, 1992; Mallet, 1995), and although cladists have been among
the strongest proponents of the PSC within systematics, over the years a
number of species concepts have been used by cladists. One perception
that has arisen from the association of the PSC with cladism is that phylo-
genetic species are defined by apomorphies, or derived characters (Mallet,
1995), that they are (or must be) monophyletic (Haffer, 1992), or that they
are somehow based on phylogeny (Mallet, 1995). Whereas it is true that
some have supported defining phylogenetic species in terms of the small-
est populations having apomorphic characters (Rosen, 1978, 1979) and
that phylogenetic species should be monophyletic (McKitrick and Zink,
1988), others have argued that not all diagnosably distinct populations
may have characters inferred to be derived yet those populations still
deserve taxonomic recognition [Nelson and Platnick, 1980; Cracraft, 1983
(contra Mallet, 1995: 298), 1989]. Although the populations included with-
in a phylogenetic species that is diagnosed only in terms of primitive char-
acters may actually represent more than one phylogenetic species, such a
mistake is simply a matter of available evidence. All designated species,
whether delimited by apomorphic characters or not, are hypotheses sub-
ject to revision when new evidence arises. Inasmuch as phylogenetic
species are basal, application of the concept of monophyly is superfluous
and unnecessary. In addition, even though the PSC is the clear choice
when the goal is to reconstruct phylogeny accurately, this does not imply
that the definition of phylogenetic species is dependent upon any
assumptions of cladistics or, contrary to Mallet (1995), is dependent on
phylogeny. The PSC is a mechanism for sorting and interpreting charac-
ter (broadly interpreted) variation within and among populations in order
to recognize basal diagnosable taxa, and is largely independent of
assumptions about process.
Critiques of the PSC have argued that a consequence of defining
species as the smallest, diagnosably distinct population will be a prolifer-
ation of species taxa, since each small population can be found to be dis-
tinct for some character, particularly at the molecular level. This criticism
can be carried to extremes when not thinking in terms of taxa: 'with
detailed morphology or modern molecular techniques, one can find apo-
morphies for almost every individual' (Mallet, 1995: 298; see also Avise
and Ball, 1990; Avise, 1994; among others).
Two misunderstandings are hidden in this criticism. First, it is argued
that each individual organism is distinct, therefore the PSC is inapplicable.
It may well be the case that each individual organism can be distinguished
using one method or another -who would be surprised at this? - but such
a finding is irrelevant because the PSC is not about the diagnosability of
individuals but of populations. Such a criticism forgets that species con-
cepts are populational concepts that are used to delimit basal taxa. The
concept of apomorphy, moreover, has no meaning at the level of an indi-
vidual but only at the level of taxa. Delimiting species taxa is a problem for
systematics not population biology (Nixon and Wheeler, 1990; Wheeler
and Nixon, 1990), although obviously population-level data are relevant.
Phylogenetic species concept: myths and misrepresentations 331
Seeing species as taxa, and as being different from populations and indi-
vidual organisms, is fundamental for describing and understanding bio-
logical diversity. Some critics of the PSC have seemingly not understood
this distinction.
A second misunderstanding related to the first is the belief that the PSC
will lead to an inordinate inflation of species names. The belief that having
too many names would be inconvenient or pernicious has had a long his-
tory within systematics, and the consolidation of names was given as one
reason why the BSC was a particularly useful innovation: with the intro-
duction of the BSC the 'total number of [bird] species to be memorized by
the taxonomist has thus been cut by two-thirds [from 27 000 to 8500; but see
below]. The practical advantage of this simplification is so obvious that
nothing more needs to be said' (Mayr, 1942: 127). One's memory capacity
aside, such views have little scientific relevance for our attempts to describe
taxic diversity accurately. It is essential - for many reasons - to individuate
diagnosably distinct taxa in nature, including those that are basal (species)
and if nature has a multitude of these, so be it.
There are several reasons for thinking that the PSC will not lead to an
extraordinary proliferation of names. First, though they may not specifi-
cally acknowledge the use of a PSC, most systematists already apply the
notion of a basal diagnosable taxon when delimiting species. This is true
within most of entomology and those disciplines concerned with the
highly speciose groups of non-vertebrates. As a consequence, extending
the PSC to all groups of organisms, most of which are far less diverse, will
not make a significant difference in global species numbers.
Second, even in those groups such as birds in which the BSC has had
considerable influence, virtually all diagnosable taxa have already been
described - as subspecies - and thus application of the PSC will not great-
ly affect the number of names, just potentially their ranking. Within the
birds-of-paradise (Paradisaeidae), for example, applying the PSC to a
group whose taxonomy has been dominated by the BSC increased the
number of recognized species by only about two-fold (Cracraft, 1992).
These results can possibly be generalized across all birds. When one ran-
domly samples the world's biological species of birds, evaluates those
species and their included subspecies as to whether they are basal diag-
nosable taxa, the estimated number of phylogenetic species in the world
is again only about twice that of the current number (9000 or so) of puta-
tive biological species (G. F. Barrowclough, J. Cracraft and R. M. Zink,
unpublished data).
There is no question that improved methods of resolving variation will
increase our ability to recognize more diagnosably distinct taxa. Critics of
the BSC seem to bemoan this fact; biologists interested in seeing nature
described and interpreted with accuracy and precision will, in fact,
applaud the use of the PSC.
16.4 TWO SPECIES CONCEPTS: A COMPARISON
The BSC and PSC provide very different lenses with which to see the world
of systematic and evolutionary biology. How species are defined influences
one's interpretations of patterns and processes (Table 16.1; see Cracraft,
1989, for additional discussion). Under the BSC, diagnosably distinct popu-
lations will sometimes be recognized as separate, monotypic species, but
often those populations are united together under a single species name if
the diagnosable differences are not judged to be significant. Within the con-
text of the PSC, on the other hand, diagnosably distinct populations would
always be accorded specific status. As a consequence, some biological
species, those consisting of multiple diagnosable taxa of uncertain relation-
ships to one another, will confound an accurate reconstruction of history if
used as terminal taxa in a phylogenetic or biogeographic analysis (Rosen,
1978,1979; Cracraft, 1983,1989; Frost and Hillis, 1990; among many others).
Phylogenetic species, in contrast, are appropriate terminal taxa for such
analyses and cannot, in and of themselves, lead to a misrepresentation of
history.
In principle, there is no gene flow among biological species because they
are assumed to be reproductively isolated. In practice, however, propo-
nents of the BSC will recognize two entities as biological species, even
though there may be gene flow, as long as it is judged to be minor. Because
phylogenetic species are defined in terms of diagnosability and not repro-
ductive isolation, it is not uncommon, especially in plants, for there to be
extensive gene flow across species borders. If the two entities are diagnos-
ably distinct, they will be recognized as phylogenetic species even though
there may be a hybrid zone within which gene flow might be extensive. In
almost all cases the latter situation would result in the recognition of a sin-
gle biological species. Under both concepts, however, populations that are
reproductively isolated will also be distinct taxonomic entities.
Finally, the two concepts have markedly different implications for bio-
geography. If species are taken to be the units of analysis for recognizing
areas of endemism, the use of biological species unless they are all mono-
typic and thus equivalent to phylogenetic species will always lead to a less
precise classification of those areas than will be the case with the use of
phylogenetic species. Furthermore, adoption of the BSC can sometimes
confound the reconstruction of history and consequently lead to inaccu-
rate depictions of the relationships among areas of endemism.
16.5 TAXONOMIC UNITS AND CONSERVATION BIOLOGY
16.5.1 The concept of evolutionary significant unit should be abandoned

As already noted, there is a growing awareness on the part of conserva-
tion biologists that species concepts have relevance for determining what
to protect, how to protect it, and how to facilitate the sustainable use of
Taxonomic units and conservation biology 333
Table 16.1 Criteria for recognition and some comparisons between the biologi-
cal and phylogenetic species concepts
Criterion Biological Phylogenetic

species concept species concept
Diagnosably distinct Sometimes Always

populations recognized as separate
species
Species unit often includes Yes Never
diagnosable allopatric (subspecies concept (subspecies concept
populations widely applied) not relevant)
Species represent terminal taxa Sometimes Always
that can be used in phylogenetic
and biogeographic analysis
Inherently capable of Yes No
misleading historical analysis
Gene flow among species Rarely Sometimes
Reproductively isolated Always Always
populations recognized as
separate species
Extensively hybridizing, Rarely Almost always
diagnosably distinct
populations recognized as
separate species
'Potential' interbreeding Yes Never
of allopatric populations
important for establishing
species status
Delimitation of areas of Coarse, less precise Fine, more precise
endemism
biodiversity (Ryder, 1986; Avise, 1989, 1994; Rojas, 1992; Moritz, 1994a,b,
1995; Vogler and DeSalle, 1994; Grant, 1995; Barrowclough and Flesness,
1996). Many of these discussions, however, have not taken into account
the full implications of the debates about species concepts within the sys-
tematic literature over the past 10 or 15 years. Instead, much of the dia-
logue has centred on the relevance of 'evolutionary significant units'
(ESUs) and their use in conservation studies.
Discussion at the 1985 meeting of the American Association of
Zoological Parks and Aquariums (AAZPA) sharpened the debate over the
issue of the units of conservation (Ryder, 1986). With the introduction of
the term ESU for those taxonomic entities having a distinct evolutionary
history, the AAZPA aimed to identify groups that were most in need of
conservation action, particularly those in captive breeding programs.
Subsequent writers have given tacit support to the importance of ESUs in
conservation (Dizon et al, 1992; Moritz, 1994a,b, 1995; Vogler and DeSalle,
1994; Barrowclough and Flesness, 1996).
The theoretical and practical goals of the ESU concept are important: to
provide an objective basis for the definition and recognition of manage-
ment units in conservation activities. A primary difficulty is that no clear
agreement has been reached on what constitutes an ESU. Thus, Vogler
and DeSalle (1994: 356) see ESUs as 'clusters of organisms that are evolu-
tionarily distinct and hence merit separate protection'. And, Moritz
(1994b: 373; see also 1994a, 1995) argues that to be an evolutionarily 'sig-
nificant' unit implies that 'the set of populations has been historically iso-
lated and, accordingly, is likely to have a distinct potential'. Moritz goes
on to note (1994b: 373) that 'ESUs should be reciprocally monophyletic for
mtDNA alleles and show significant divergence of allele frequencies at nuclear
loci' [italics in original]. The confusion and ambiguity over how to define
an ESU is so great - ranging from populations that are significantly differ-
entiated, to biological species, to phylogenetic species - that its objective
use is virtually precluded (see Grant, 1995, for a description of the confu-
sion).
Both Vogler and DeSalle (1994) and Moritz (1994a,b, 1995) recognize
earlier arguments [Cracraft, 1991; Barrowclough and Flesness, 1996 (in a
paper circulated since 1993)] that ESUs are essentially equivalent to phy-
logenetic species. Despite this acceptance, Vogler and DeSalle (1994) and
Moritz (1994a,b) would maintain the concept of ESU, even though the cri-
teria used to recognize historically distinct and significant units are large-
ly those that individuate phylogenetic species. Yet, the ambiguities
implied by the various definitions of the ESU remain, thus leading one to
conclude that it should be abandoned by conservation biologists.
Such a solution is also strongly supported by another consideration.
The use of ESUs within conservation biology undercuts the scientific
foundation and results of that discipline. ESUs have no scientific status
within systematics, and it is systematics that provides the linguistic and
historical framework for the study of biodiversity. The results of formal
taxonomy - as reflected in species-level taxa - are now codified in an enor-
mous series of national and international legal instruments. As Geist (1992:
274) remarks: 'courts and solicitors' offices are allowed to rule on taxono-
my. Judges may now decide on matters such as the definition of species
or subspecies, the criteria for establishing taxa, which taxa are valid, and
which populations can be legally protected. The implications for conser-
vation, but also for biology in general, are profound and worrying'.
Because ESUs have no formal systematic status, they will rarely have
any legal status. The Convention on Biological Diversity will be the chief
international legal instrument affecting the conservation and sustainable
use of biodiversity for the foreseeable future. The use of ESUs in meeting
the goals of the Convention would be difficult at best because of the lack
of international standards of scholarship and formal nomenclatural rules
Taxonomic units and conservation biology 335
over those units. Despite the fact that there may be arguments over
species concepts, species-level taxonomy and its rules of nomenclature
have broad acceptance within the systematic community and among
those biodiversity sciences that use taxonomic information. No such
framework exists for ESUs.
16.5.2 Phylogenetic species are the most relevant units for conservation
biology
A comprehensive programme of conservation and sustainable use of bio-
logical diversity will depend upon having all taxonomically distinct, diag-
nosable populations identified and named. A comparison of the BSC with
the PSC demonstrates that phylogenetic species, not polytypic biological
species, are the most appropriate units for conservation (Table 16.2).
Table 16.2 Possible implications of different species concepts for conservation

and sustainable development
Implication Biological Phylogenetic

species concept species concept
Equivalent to Sometimes Essentially always
'evolutionary
significant units' of
conservation biology
Estimates of diversity Underestimates numbers Yields accurate estimates
of differentiated taxa of basal taxa given
available data
Delineation of areas of Broadly defined; Narrowly defined on basis
of endemism
underestimates numbers basal taxa; more finely
of areas; confounds their resolved;historical
historical relationships relationships not
confounded
Demographic analysis May tend to overestimate Accurately estimates
population sizes of population sizes of
endangered taxa basal taxa
Apportionment of Confounded Accurately apportioned
genetic variation
Captive breeding of Units of programme Units of programme are
endangered taxa (and ambiguous; increased basal taxa; less risk of
reintroductions) danger of mixing distinct mixing distinct
evolutionary units evolutionary units
Trade in endangered Enforcement loosely Enforcement stringently
taxa interpreted; fewer taxa interpreted; more taxa
protected; higher risk due protected; lower risk due
to trade to trade
Political impact for Less More
endangered taxa
Phylogenetic species meet the objectives envisioned for evolutionary
significant units. The desire of conservation biologists to have a unit that
is relevant for management purposes is fulfilled by the PSC but not the
BSC. As numerous authors have noted, many biological species are com-
posites of diagnosably distinct as well as arbitrarily demarcated races or
subspecies. As judged by the managers of captive breeding programmes,
this inconsistency will lead to innumerable problems in assigning conser-
vation priorities (Ryder, 1986).
Different species concepts can result in different estimates of species
diversity. The use of the BSC will underestimate the numbers of diagnos-
ably (evolutionarily) distinct units as compared with the PSC. Comparisons
of diversity are important for countries in setting conservation priorities
and are of interest to funding agencies that support activities to implement
those priorities.
Many countries and conservation organizations are currently concerned
with creating protected areas and ecosystem management zones using
measures of endemism to set priorities. If species limits are drawn differ-
ently under different species concepts, then so too will their areas of
endemism. The use of biological species results in an underestimate of the
numbers of areas of endemism, tends to overestimate their size, and con-
founds the analysis of their historical interrelationships. All these problems
are avoided by the use of the PSC.
Species concepts have important implications for demographic analy-
ses. Inasmuch as different concepts can allocate populations to species
taxa in opposing ways, numbers of individuals within populations of
those species will be estimated differently. Many decisions about conser-
vation priorities and actions depend on demographic information.
Phylogenetic species allow for an accurate count of individuals within
basal taxa, whereas biological species do not necessarily identify basal taxa
and can lead to inappropriate or inaccurate estimates. If all the popula-
tions of a polytypic biological species are included in a demographic
assessment of that species, the numbers of individuals within endangered
basal taxa may be overestimated or ignored altogether.
For the same reasons, the use of phylogenetic species will allow genet-
ic variation to be apportioned within and between taxa more accurately
than will biological species. If diagnosably distinct basal taxa are united
into a single biological species, there is a risk of ignoring the levels of
genetic variation within those distinct taxa in favour of creating an esti-
mate of variation within the biological species as a whole, a situation that
is likely to compromise effective management of biodiversity.
The concept of the ESU grew out of attempts by conservation biologists
to provide an effective framework for managing in situ and ex situ breed-
ing programmes for endangered taxa. The ESU was born because the BSC
failed to provide that framework. The PSC, in contrast, provides the for-
mal taxonomic context for managing breeding programmes.
References 337
The BSC also does not provide an effective basis on which to manage
trade in endangered taxa. To be effective, management programmes will
need to monitor and regulate trade of all distinct taxa. If distinct and
endangered basal taxa are lumped with more common forms under the
BSC, the danger is that sufficient protection will not be extended to the
former. This problem is avoided with the PSC because each distinct basal
taxon would be ranked at the species-level and therefore assume a height-
ened legal status over unnamed populations or subspecies.
Finally, although some national and international legal instruments
extend protection to taxonomic units below the level of species, many do
not. Taxa of species rank are still the primary currency for conserving and
managing biological diversity. Endangered, distinct populations ranked at
subspecies, or not given formal rank at all, will generally carry less politi-
cal and conservation significance than those ranked at the level of species.
Because the PSC provides for specific rank for all diagnosable populations,
the importance of the latter is magnified relative to lumping those popu-
lations into a larger biological species.
Acknowledgements
I should especially like to thank Professor Michael Claridge for inviting
me to participate in this symposium. I am also grateful for the helpful
comments of Drs Rob DeSalle and Robert Zink on the manuscript.
16.6 REFERENCES
Avise, J.C. (1989) A role for molecular genetics in the recognition and conservation
of endangered species. Trends in Ecology and Evolution, 4, 279-81.
Avise, J.C. (1994) Molecular Markers, Natural History and Evolution, Chapman & Hall,
New York.
Avise, J.C. and Ball, R.M. (1990) Principles of genealogical concordance in species
concepts and biological taxonomy. Oxford Surveys Evolutionary Biology, 7,45-67.
Barrowclough, G.F. and Flesness, N.A. (1996) Species, subspecies, and races: the
problem of units of management in conservation, in Wild Mammals in Captivity
(eds M. Alien and H. Harris), University Chicago Press, Chicago, pp. 247-54.
Bock, W.J. (1987) Species concepts, speciation, and macroevolution, in Modern
Aspects of Species (eds D. Iwatsuki, P. Raven, and W.J. Bock), University Tokyo
Press, Tokyo, pp. 31-57.
Cracraft, J. (1987) Species concepts and the ontology of evolution. Biology and
Philosophy, 2,329-46.
ation, in Speciation and its Consequences (eds D. Otte and J. A. Endler), Sinauer
Cracraft, J. (1991) Systematics, species concepts, and conservation biology.
Abstract, 5th Annual Meeting, Society of Conservation Biology, Madison,
Wisconsin.
Cracraft, J. (1992) The species of the birds-of-paradise (Paradisaeidae): applying
the phylogenetic species concept to complex patterns of diversification.
Cladistics, 8,1-43.
Dizon, A.E., Lockyer, C, Perrin, W.F., Demaster, D.P. and Sisson, J. (1992)
Rethinking the stock concept. Conservation Biology, 6,24-36.
Eldredge, N. and Cracraft, J. (1980) Phylogenetic Analysis and the Evolutionary
Process, Columbia University Press, New York.
Frost, D.R. and Hillis, D.M. (1990) Species in concept and practice: herpetological
applications. Herpetologica, 46, 87-104.
Geist, V. (1992) Endangered species and the law. Nature, 357,274-6.
Grant, W.S. (1995) Multi-disciplinary approaches for defining evolutionarily sig-
nificant units for conservation. South African Journal Science, 91, 65-7.
Haffer, J. (1992) The history of species concepts and species limits in ornithology.
Bulletin British Ornithologists' Club Centenary Supplement, 112A, 107-58.
Mayr, E. (1942) Systematics and the Origin of Species, Columbia University Press,
New York.
Massachusetts.
Mayr, E. (1970) Populations, Species, and Evolution, Harvard University Press,
Botany, 79, 222-38.
Mayr, E. (1993) Fifty years of progress in research on species and speciation.
Proceedings of the California Academy of Sciences, 48,131^40.
Mayr, E. and Short, L.L. Jr (1970) Species taxa of North American birds, a contri-
bution to comparative Systematics. Publications Nuttall Ornithological Club, 9,
1-127.
McKitrick, M.C. and Zink, R.M. (1988) Species concepts in ornithology. Condor, 90,
1-14.
Moritz, C. (1994a) Applications of mitochondrial DNA analysis in conservation: a
critical review. Molecular Ecology, 3, 401-11.
Moritz, C. (1994b) Defining ' evolutionarily significant units' for conservation.
Trends in Ecology and Evolution, 9, 373-5.
Moritz, C. (1995) Uses of molecular phylogenies for conservation. Philosophical
Transactions Royal Society London, 349B, 113-18.
Nelson, G. and Platnick, N.I. (1981) Systematics and Biogeography: Cladistics and
Vicariance, Columbia University Press, New York.
Nixon, K.C. and Wheeler Q.D. (1990) An amplification of the phylogenetic species
Rojas, M. (1992) The species problem and conservation: what are we protecting?
Conservation Biology, 6,170-8.
Rosen, D.E. (1978) Vicariant patterns and historical explanation in biogeography.
References 339
Guatemala: revisionary studies and comparative geography. Bulletin American
Museum Natural History, 162,267-376.
Ryder, O.A. (1986) Species conservation and systematics: the dilemma of sub-
species. Trends in Ecology and Evolution, 1, 9-10.
Selander, R.K. (1971) Systematics and speciation in birds, in Avian Biology (eds D.
Farner and J. R. King), vol. 1, Academic Press, New York, pp. 57-147.
Short, L.L. (1969) Taxonomic aspects of avian hybridization. Auk, 86, 84-105.
Vogler, A.P. and DeSalle, R. (1994) Diagnosing units of conservation management.
Conservation Biology, 8, 354-63.
lem: a reply to de Quieroz and Donoghue. Cladistics, 6, 77-81.
17
The species in mammals
G. B. Corbet
Contacting address: Little Dumbarnie, Upper Largo, Leven, Fife KY8 6JQ, Scotland, UK
ABSTRACT
The biological species concept has been widely used in the discrimi-
nation of mammalian species and continues to provide a valuable the-
oretical framework. Sympatric and parapatric cryptic species should
be recognized as such however subtle the differences, provided that
these are likely to be stable rather than ephemeral; that there is good
evidence of lack of hybridization; and that sampling has been ade-
quate to extrapolate from specimens examined to the entire ranges of
the species. Allopatric taxa should be considered conspecific if they
differ only in ways that are analogous to those found within inter-
breeding populations and there is no other evidence of reproductive
incompatibility. The use of subspecies is valuable when they can be
diagnosed and shown to have objective geographical boundaries.
Phylogenetic species concepts are not more objective since they are
equally subject to errors in extrapolation from sample to natural pop-
ulation, and depend upon interpretation of differences in terms of
hypothetical history. Data from karyology and molecular techniques
can play a valuable part in detecting species limits provided they are
integrated with other data and that geographical sampling is ade-
quate to extrapolate conclusions to the whole species.
17.1 INTRODUCTION
The delimitation of species in mammals shows many parallels to the situ-
ation in birds, and as in birds the problem is not complicated by asexual
reproduction. The species-level taxonomy of mammals has been consid-
erably influenced by that of birds in which much of the variation is more
visible and sample size can more easily be enhanced by field observation.
In these respects some groups of mammals closely resemble birds, e.g.
among the diurnal and colourful primates and squirrels. However, most
342 The species in mammals
mammals are small, communicate by scent, are nocturnal or otherwise
elusive, and the collection of adequate samples has often been difficult.
An important factor in determining the nature of species is dispersive
ability. In this respect mammals show a considerable range that can be
reduced to the following main categories, although these are by no means
discrete and each includes considerable variation:
1. Flying species (i.e. bats) capable of colonization across considerable
stretches of water.
2. Large and highly mobile terrestrial species, especially those occupying
the more continuous habitats, such as hares (Lepus) in the steppe and
savannah zones.
3. Smaller, less mobile terrestrial species and those specialized for more
discrete habitats (the majority).
4. Subterranean species with minimal dispersive ability, e.g. moles,
gophers and mole-rats.
5. Oceanic species with good dispersive ability, e.g. most whales and dol-
phins.
6. Coastal marine species with more limited dispersive ability, e.g. most
pinnipedes and some dolphins and porpoises.
The limitations of dispersive ability mean that most species of mammals
have a moderately stable range, i.e. each species has a distinctive spatial
pattern and shape.
17.2 HISTORICAL REVIEW

Linnaeus (1758), and most of his successors during the following hundred
years, recognized species intuitively as groups of individuals within
which clear-cut divisions could not easily be made but which appeared to
be clearly definable in relation to each other. Species were recognized as
being variable, but any variation tended to be interpreted as deviation
from a fixed 'essence' or 'type'. Linnaeus himself in the 10th edition of his
Systema Naturae (1758) described varieties within some species, but main-
ly domestic ones such as the dog, with eleven varieties, and the sheep,
with seven.
From the total of 184 species of mammals described by Linnaeus in 1758
the number of known and described species rose rapidly, partly as a result
of geographical exploration but helped enormously by the ease with
which the Linnean system allowed additions to be named and added to
the inventory in an orderly way. By the time Darwin's Origin of Species was
published in 1859 the concept of the species as a discrete natural entity
was already wavering because of the sheer number of new species being
described that appeared to be progressively filling the gaps between hith-
erto readily distinguishable species. The burst of interest in evolution and
Historical review 343
the 'transmutation of species' further eroded the concept of species as dis-
crete entities.
The main development in species-level taxonomy of mammals during
the latter half of the 19th century came however not from the new evolu-
tionary theories but from a growing interest in field natural history.
Travelling became easier and this led to a shift from classifications that
were concerned only with museum collections to classifications that were,
it was hoped, applicable to the living populations from which the muse-
um samples came. The increasing intensity of geographical sampling
made it clear that in certain groups the 'gap-filling' between described
'species' demonstrated a continuum of variability, with a spatial compo-
nent. Geographical subspecies, conceived as subdivisions of species and
designated by trinomials, were first used for birds about 1845. In mammals
J.A. Alien was one of the first to study geographical variation, for example
of North American mammals (Alien, 1876). He argued, for instance, that
the American grizzly and brown bears and the Eurasian brown bears were
best considered as races of one species, Ursus arctos, a view now generally
held although the intervening years have seen a great proliferation of
species described within this group.
Towards the end of the 19th century a great boost was given to the col-
lection and description of mammals by the work of C. Hart Merriam as
part of the Biological Survey of North America, organized by the US
Department of Agriculture. Merriam introduced the technique of prepar-
ing dry 'study skins' and skulls of small mammals in a standard way and
building up series with precise geographical data in order to take intraspe-
cific variation into account in comparing species. Merriam was emulated
at the British Museum (Natural History) by Oldfield Thomas, who over
the period from 1890 to the 1920s organized and encouraged similar col-
lecting, especially of small mammals, throughout almost the entire world
other than North America.
The only comprehensive enumeration of mammalian species around
this period was by Trouessart (1897-1905) who listed 9381 species of which
about 60% were fossil, leaving about 3750 living species.
A common perception of species at that time was expressed by Miller
(1901), mammalogist at the US National Museum, who wrote: 'Species
and subspecies, to use for the present these obsolete terms in the absence
of the single word needed to replace them, are not invariably separated
from their allies by characters easy to describe'. Between then and 1940 the
description of species, and increasingly of subspecies, continued, but with
little progress towards any kind of consensus as to how these groups were
to be defined.
During the period 1920 to 1940 a fresh concept of the species as an
objective entity began to emerge, especially in ornithology where field
observation could more easily supplement museum investigation in
detecting the overall patterns of variability in bird populations, but also
helped by advances in the fields of genetics, ethology and ecology. The
biological species concept that emerged was epitomized by Mayr's (1940)
definition of a species as 'a group of populations which replace each other
geographically or ecologically and of which the neighbouring ones inter-
grade or hybridize wherever they are in contact, or which are potentially
capable of doing so in those cases where contact is prevented by geo-
graphical or ecological barriers'. In other words, stress was laid upon the
gaps separating species and the limitations imposed upon variability with-
in species by the retention of interfertility.
The relevance of this concept to mammals was expounded by Hall
(1943) and put into practice in his revision of North American weasels
(Hall, 1951), in which he recognized three 'polytypic' species in the sub-
genus Mustela in place of the 25 species listed by Miller (1924). This con-
cept was also applied in The mammals of North America (Hall and Kelson,
1959) and in the checklists of Palaearctic and Indian mammals (Ellerman
and Morrison-Scott, 1951) and of southern African mammals (Ellerman et
a/., 1953). In general terms it has stood the test of time, although since the
1950s there has been a tendency towards the recognition of a larger num-
ber of more narrowly defined species, for several quite separate reasons.
These can be categorized as: (i) the refinement of morphological analysis;
(ii) the development of karyology; (iii) the development of molecular biolo-
gy; and (iv) improvement of geographical sampling. All have contributed to
the discovery of sibling and cryptic species. When these have been sym-
patric there is little difficulty in applying the biological species concept and
this has generally been done. Allopatric taxa show a wider range of cryptic
differences that has led to many cases of differing opinion as to how many
species should be recognized. Examples are discussed below.
17.3 SYMPATRIC CRYPTOSPECIES
17.3.1 Well-resolved examples

The first task of the taxonomist in delimiting sympatric species is to rec-
ognize those differences that are correlated with reproductive isolation
amongst the noise created by other kinds of variability, e.g. that related to
sex, age and season, and polymorphisms that do not interfere with repro-
duction. Examples from the European fauna of closely related pairs of
species that were resolved on purely morphological grounds are the mice
Apodemus sylvaticus and A. flavicollis, and the shrews Crocidura russula and
C. suaveolens. However, although these are easily discriminated in the
well-studied parts of their sympatric ranges in western Europe, they still
give trouble in interpreting variability on the fringes of their ranges,
beyond the sympatric zone, where character displacement tends to occur.
Sympatric cryptospecies 345
Satisfactory resolution of two pairs of sympatric species of European
bats, Plecotus auritus/P. austriacus, and Myotis mystacinus/M. brandti, was
first achieved by the discovery of differences in the male genitalia previ-
ously overlooked (Corbet, 1964; Gaukler and Kraus, 1970).
All the above examples are sufficiently differentiated to coexist over
large parts of their ranges and presumably have different ecological
requirements. Many other examples of sibling pairs have only been satis-
factorily discriminated by karyological study, usually followed by bio-
chemical work and the elucidation of subtle morphological differences.
Most such cases that have been well studied are of basically parapatric
pairs or series but with overlapping sympatric areas forming minor parts
of the geographical ranges. Well-studied examples are mice of the genus
Mus, the voles Microtus arvalis/M. rossiaemeridionalis, and shrews of the
Sorex araneus complex in Europe. The last can be used as an example.
The common shrew, Sorex araneus, was first found to show Robertsonian
chromosome polymorphism in Britain (Sharman, 1956). Subsequent kary-
ological work has shown three sibling species in continental Europe, with
substantial zones of sympatry: S. araneus (north and east), S. coronatus
(France, etc.), and S. granarius (Iberia), reviewed by Hausser et al. (1985).
These species differ by three pericentric inversions, three paracentric inver-
sions and one reciprocal translocation, plus a few secondary translocations.
Allozyme analysis has shown these three species to form a discrete group,
with little differentiation between them - a mean genetic distance (Nei,
1972) of only 0.055. Attempts at morphological discrimination have failed to
find differences allowing unequivocal identification of all individuals, but
indices based upon cranial measurements allow about 96% to be identified
(Hausser et al, 1985; Handwerk, 1987).
Geographical sampling of this complex in western Europe has probably
now been sufficient to demonstrate a pattern of ranges that is likely to
remain reasonably stable and these three species have been accepted as
such in the Handbuch der Saugetiere Europas (Niethammer and Krapp,
1990). However, S. araneus extends east to central Siberia. Of the 20 or so
other currently defendable species of Sorex in Eurasia, 15 are clearly dis-
tinct from the S. araneus complex (Dannelid, 1991). Since sampling has
been less intensive in eastern Europe and Asia it is quite probable that fur-
ther cryptic species will be found associated with eastern S. araneus.
17.3.2 Partially resolved examples

Many cases where karyological work has indicated the presence of sibling
species are still unresolved, some in spite of considerable study. A partial-
ly resolved example concerns the African mice of the genus Mastomys.
Intensive work in southern Africa, summarized by Skinner and Smithers
(1990), has revealed two widely sympatric species, M. natalensis and
M. concha, differing in chromosome number (2n = 32 and 36 respectively)
and in haemoglobin (by electrophoresis). Although further differences
have been detected by multivariate analysis of cranial measurements, in
small details of penis and spermatozoa, and in reproductive behaviour
and ultrasounds, no easy means of identification by morphology have
been found, making it difficult to use existing museum collections (which
are very extensive) and to extend resolution of the species to the rest of
Africa, although similar sibling species have been resolved in West Africa.
A further example, also of considerable economic and medical impor-
tance, involves the rats of the Rattus rattus complex in south-eastern Asia.
Careful morphological work in recent decades has demonstrated that sev-
eral forms previously included in R. rattus are in fact specifically distinct,
e.g. R, argentiventer and R. tiomanicus, both important agricultural pests
(reviewed by Corbet and Hill, 1992). This still leaves about a hundred
named forms within R. rattus s.l. Although karyological work has indicat-
ed the apparent coexistence of sibling chromosome species (Yosida, 1980),
this has rarely been accompanied by any indication of the morphological
characters nor by the preservation of material for morphological study. A
recent study in southern India (F. Catzeflis et al, unpublished results) has
demonstrated a good correlation between chromosomal, genie and cranial
characters, allowing the recognition of two sympatric species, provisional-
ly called R. rattus (with 2n = 38) and R. satarae (2n = 42). However, both
species are polychromatic in pelage characters, they cannot be confident-
ly identified by external appearance, and geographical variation makes it
difficult to apply the results far beyond the study area. Forms with 2n =
42 occur throughout south-east and east Asia and have been equated with
R. tanezumi Temminck, 1844 from Japan (Musser and Carleton, 1993).
However, so few populations have been sampled for chromosome data
that there is a long way to go before these two species can be satisfactori-
ly diagnosed and their ranges determined.
17.4 PARAPATRIC SERIES: SUBSPECIES OR SPECIES?

Several more or less well-studied examples of series of parapatric cryptic
taxa are known, especially among subterranean mammals. The best docu-
mented is the group of mole-rats of the genus Nannospalax (or Spalax) in
Israel and adjacent areas, reviewed by Nevo (1989). Four chromosomal
forms of the N. ehrenbergi complex form a parapatric series with very lim-
ited hybridization in zones varying from 320 to 2800 m in width. They are
not distinguishable by morphology in spite of intensive study and show
very slight genetic differences (maximum 0.012). Although usually
referred to as 'chromosome species' these have never been separately
named. Since they can easily be referred to as, for example, 'Nannospalax
ehrenbergi 2n = 52', this is a practical solution.
Parapatric series 347
It seems likely that similar situations are repeated elsewhere in the
spalacine mole-rats (Savic and Nevo, 1990), the North American gophers
of the genus Thomomys (Thaeler, 1985) and the South American tuco-tucos
of the genus Ctenomys (Ortello, 1995).
Problems with the resolution of parapatric series are not confined to
inconspicuous small mammals. There are many cases where parapatric
pairs or series of taxa that have been recognized as distinct on the grounds
of morphology or colour are variously treated as species or subspecies.
The gibbons (Hylobates) include two series of parapatric taxa, with consid-
erable instability in the number of species recognized (summarized by
Corbet and Hill, 1992). In the subgenus Hylobates these are H. pileatus in
Eastern Thailand and Cambodia; H. lar from Burma to peninsular Malaya
and Northern Sumatra; and H. agilis in parts of peninsular Malaya and
Southern Sumatra. These abut without intergradation and with only rare
interbreeding, and there appears to be a consensus that they represent
three distinct species. The situation on Borneo is less certain. The gibbons
on most of Borneo are generally agreed to constitute a distinct species, H.
muelleri. In the south is a parapatric form, albibarbus, which, on morpholo-
gy, would appear to be H. muelleri, but is distinguished by a quite distinct
territorial song that suggests that it represents H. agilis of Sumatra. The
two forms are separated by major rivers such as the Barito but a small
degree of hybridization takes place at the headwaters (Marshall and
Sugardjito, 1986).
In the colourful leaf monkeys of the genus Presbytis recognition of three
parapatric species on Sumatra has been based largely on vocalization
although the situation is confused by considerable geographical variation
in colour and pattern within species (Aimi and Bakar, 1992; Corbet and
Hill, 1992).
On a larger scale the baboons, Papio, form a series of more or less para-
patric forms spanning the length of Africa. Traditionally treated as about
five distinct species, ranging from P. hamadryas in Ethiopia to P. ursinus in
South Africa, the most recent review recognizes only one species (Groves,
1993). An equally controversial example concerns the sportive lemurs of
Madagascar, Lepilemur, where the recognition of six parapatric species has
been advocated on the basis of karyological data (Ishak et al, 1992)
although all have been treated as conspecific.
The African elephants, Loxodonta, have most often been treated as a sin-
gle species, L. africana. However, the status of the smaller forms in the
forests of west and central Africa has always been uncertain. Claims of two
species have inevitably been hampered by small sample size and sparse
geographical data (Eisentraut and Bohme, 1989; Bohme and Eisentraut,
1990). On the basis of multivariate analysis of cranial measurements
Groves et al. (1993) supported the recognition of two species in spite of
about six probable hybrids among 45 from the contact zone. More
integrated presentation of the data and detailed mapping of the contact
zone will be needed to clarify the situation.
In the case of pelagic cetaceans the problem of discriminating species is
exacerbated by their mobility as well as the difficulty of collecting and pre-
serving adequate samples. The genus Delphinus, for example, has been
variously considered to include one species, the Common dolphin, D. del-
phis, or two: a short-beaked species, D. delphis and a longer-beaked one, D.
capensis (or D. tropicalis) (van Bree and Gallagher, 1978). These have
recently been shown to be discrete species in Californian waters (Heyning
and Perrin, 1994) but there are difficulties in extrapolating the results to
the rest of the world.
17.5 ALLOPATRY
Allopatric populations that have been shown to differ slightly, whether in
morphology, colour, karyotype or in biochemical characters, present all
the problems and uncertainties seen in parapatric forms, with the addi-
tional problem that there is no opportunity to determine the degree of
hybridization, if any, under natural conditions. Many examples of insular
populations that were originally distinguished from their continental rela-
tions as separate species have subsequently been treated as conspecific.
This may be on the ground that the differences (often in size and colour)
are little or no greater than those seen among clinal variation within con-
tinental populations, or on the basis of experimental attempts at inter-
breeding. Examples among small mammals of the British Isles are the
Orkney voles, Microtus arvalis orcadensis and M. a. westrae, which are inter-
fertile with continental M. arvalis; and the Skomer vole, Clethrionomys
glareolus skomerensis which is easily distinguishable from but fully interfer-
tile with C. glareolus of mainland Britain.
In many cases an abundance of data seems to confuse rather than clar-
ify the question of where the species line should be drawn. Two species of
chimpanzee, Pan troglodytes and P. paniscus, are generally recognized on
the basis of a suite of morphological and behavioural characters. These are
effectively allopatric, being separated by the Congo River. Studies of mito-
chondrial DNA gene sequences in all the great apes and humans has
shown greater divergence within the gorilla, Gorilla gorilla, than between
the two species of Pan, and even greater differences between the two
island populations of orang utan, Pongo pygmaeus, on Borneo and Sumatra
(Ruvolo et al, 1994).
A more problematical example concerns the right whales of the genus
Eubalaena (sometimes included in Balaena). Three allopatric populations
are found in the temperate north Atlantic (glacialis), the temperate north
Pacific (sieboldi or japonica) and the south temperate ocean (australis). In
spite of earlier recognition as separate species more recent studies have
Discussion 349
failed to detect differences between the two northern populations
(Omura, 1958; Omura et al, 1969). Most recent authors continue to treat
northern and southern populations as separate species (£. glacialis and £.
australis) but even in the most comprehensive recent handbook
(Cummings, 1985) no diagnostic differences are given. One cranial differ-
ence has been reported, concerning the shape and extent of the alisphe-
noid bone which was described as 'strikingly different' between the two
forms by Muller (1954) but on the basis of only four specimens of each
form. Using mitochondrial DNA Schaeff et al. (1991) reported a genetic
distance of 1.82% 'suggesting that the two diverged between c. 0.9 to 1.8
million years ago'. Since the reproductive cycle of the two populations is
6 months out of phase these authors concluded that they must be repro-
ductively isolated and therefore considered them separate species.
A similar uncertainty exists in the case of the minke whales Balaenoptera
acutirostrata s.l. of the northern and southern oceans which have been con-
sidered separate species on the basis of allozyme analysis (Wada and
Numachi, 1991).
17.6 DISCUSSION
17.6.1 The biological species concept

The biological species concept has, I believe, stood the test of time as the
best theoretical framework for the description and diagnosis of species in
living mammals (and indeed in most other sexually reproducing animals).
We should accept that different species concepts are needed in organisms
with very different breeding systems and therefore very different patterns
of variation. It can easily be added to in order to take account of species in
time (see section 17.6.5).
For living mammals a useful definition is: 'A species is a set of contem-
porary individuals that are capable of interbreeding to an extent that pre-
cludes the coexistence in the same place of discrete subsets'. This remains
a useful concept, even when no direct data are available on interfertility or
reproductive behaviour, since there are sufficient well-studied examples
to enable us to use morphological, ethological, karyological, biochemical
or other characters as indicators of the probable degree of reproductive
isolation involved. Even in the best-studied examples we can only exam-
ine a tiny sample of individuals so that our conclusions must always carry,
explicitly or not, a degree of probability that our sample is representative
of the species.
It is probably no accident that many of the most intractable problems of
defining mammalian species involve taxa of economic or medical impor-
tance - in other words the closer we look, the more complexity we find.
Example are Mus musculus, etc., Rattus rattus, etc., and Pan spp.
Ideally, any judgement in borderline cases should be based on a suite
of different kinds of data, e.g. on morphology, ethology, karyology,
hybridization and genetics. Where results from a number of sources are
congruent a judgement based on the biological species concept is relative-
ly easy, especially in the case of sympatric and parapatric taxa. Where only
morphological data are available we must make the best of it and assess
the probable relationship by analogy with better-known taxa. Where data
on karyotype and/or biochemistry are available these should be integrat-
ed with existing morphological or other data before decisions are made
with regard to the rank of the taxa concerned. Consideration also needs to
be given to the question whether the samples subjected to analysis are
adequate to allow extrapolation of the results to the species as a whole,
given its known geographical range.
Keeping these limitations in mind, recommendations can be made
according to the following categories (Figure 17.1):
• Sympatric forms (Figure 17.1 (a)): where two forms coexist with little or
no hybridization they should be treated as species, however subtle the
differences. However, formal naming is best delayed until it is reason-
ably certain that the differences are not ephemeral, and care needs to
be taken to avoid extrapolation beyond the area from which samples
have been examined. Examples of moderately well-resolved cryptic
species in this category are the shrews Sorex araneus, S. coronatus and S.
granarius in western Europe; and the multimammate mice Mastomys
natalensis and M. concha in southern Africa.
(a) (b)
H- H-
(c) (d)
H+ H±
(e) (f)
Figure 17.1 Spatial and phenetic relationships between sister taxa. H: ± hybridize
tion. See text.
Discussion 351
• Diagnosable parapatric forms with minimal hybridization (Figure
17.1(b)): these should be treated as separate species if they are diagnos-
able on the basis of a suite of differences, or even on a single character
provided that sampling has been sufficiently intense for us to be rea-
sonably certain that it is diagnostic. Example: the hedgehogs Erinaceus
europaeus and £. concolor in Europe.
• Diagnosable parapatric forms with substantial hybridization or
intergradation (Figure 17.1(c)): these are best treated as subspecies, e.g.
the house mice Mus musculus musculus and M. m. domesticus in Europe
and the African elephants Loxodonta africana africana and L. a. cyclotis.
• Parapatric forms with minimal differences and some hybridization, e.g.
the mole-rat Nannospalax ehrenbergi in Israel (Figure 17.1(d)): these are
better treated as subspecies, diagnosed by karyotype, since they are
very unlikely to be able to coexist and such situations are difficult to
detect without intensive study.
• Allopatric forms diagnosable on the basis of differences of a kind that
are commonly found within interbreeding populations (Figure 17.1(e)):
subspecies, e.g. many insular rodents such as the voles Microtus arvalis
orcadensis and Clethrionomys glareolus skomerensis,
• Allopatric forms diagnosable on a suite of characters or on a single char-
acter difference that is unlikely to occur within an interbreeding popu-
lation (Figure 17.1(f)): species, e.g. the chimpanzees Pan troglodytes and
Pan paniscus, and possibly the right whales Eubalaena glacialis and E. aus-
tralis (but variability of the sole diagnostic character in the latter case is
unknown and may prove not to justify specific rank).
17.6.2 The phylogenetic species concept

The phylogenetic species concept defines a species as, for example, 'the
smallest possible group of a sexually reproducing organism that possesses
at least one diagnostic character that is present in all group members but
is absent from all close relatives of the group' (Quicke, 1993). This has both
theoretical and practical shortcomings. It assumes monophyly of the diag-
nostic characters, and assumes a degree of sampling intensity within the
taxa concerned that can rarely be achieved other than in birds and a very
few equally conspicuous taxa. It also depends upon the recognition of
hypothetical past branching patterns that are notoriously difficult to
establish with a high degree of confidence, especially at low taxonomic
levels. The application of parsimony may enable the most probable clado-
gram to be found, but only among many other almost equally probable
ones. Although parsimony is a valuable concept in phylogenetic recon-
struction, evolution has had plenty of time to be messy.
In practice, application of the phylogenetic species concept would lead
to many forms now recognized as insular subspecies being raised to
species rank on the basis of characters that are unique in the samples avail-
able from the island but that are not yet known from the continental range
of the species, even where these insular populations are known to be com-
pletely interfertile with the continental population. The consequence
would be greater instability than results from application of the biological
species concept, although both are vulnerable to the problems of inade-
quate sampling. There is no reason why 'phylogenetic' subspecies should
not be treated as parts of 'biological' species.
To be practicable, the phylogenetic species concept needs the addition
of a scale, other than 'smallest possible group'; the choice of an appropri-
ate scale could make it virtually identical to the biological species concept
in practice.
17.6.3 Domestic taxa

Domesticated forms of mammals do not fit comfortably into definitions of
species or subspecies based upon wild forms, whichever concept is used,
and their classification and nomenclature have been correspondingly
unstable. Proposals to improve stability were made by Corbet and
Glutton-Brock (1984) who advocated that domesticated forms that are
readily recognizable as such, e.g. the domestic dog, should be treated as
species separate from their wild ancestors (in this case the wolf, Canis
lupus). Strict adherence to either biological or phylogenetic species con-
cepts would be both theoretically irrelevant and grossly disruptive to
longstanding nomenclature.
17.6.4 Subspecies
Subspecies have been widely used in mammals but also widely misused
- the great majority of subspecific names used in the literature are virtu-
ally meaningless in that they cannot be related to discrete diagnosable
taxa. The recorded ranges of contiguous continental subspecies can be
quite spurious because of the chance siting of type localities (Corbet,
1966: 7-9; 1970).
Many versions of the biological species concept have also, unnecessar-
ily, encouraged a spurious view of subspecific variation by describing
species as, for example, 'groups of actually or potentially interbreeding
natural populations' (Mayr, 1963). Most studies of variation in species
occupying continuous habitat have shown a pattern of incongruent clines
rather than discrete, definable 'populations' or subspecies, e.g. in the
American marten, Maries americana (Hagmeier, 1958).
Nevertheless, I believe that the concept of subspecies is a valuable one
when applied to allopatric taxa that differ only to a degree that is com-
monly found within interbreeding populations or that have been demon-
Conclusion 353
strated experimentally to be highly interfertile; and to parapatric taxa with
a considerable degree of hybridization.
17.6.5 The dimension of time
Many small mammals are effectively annuals. Major changes in gene fre-
quency can therefore take place very quickly, e.g. through founder
effects (Corbet, 1975). Consideration therefore needs to be given to
whether observed patterns of variation are ephemeral or stable. This
may be particularly applicable in the case of complexes of allopatric or
parapatric chromosomal forms, as in Mus musculus in the Italian Alps
(Corti et al, 1986).
Species of course need to be definable in palaeontological time. The
biological species concept has the advantage that when applied to living
animals it reflects the contemporary pattern and is not dependent upon
any hypothetical reconstruction of past events. Nevertheless, it can be
extended to include those ancestors that are not separated from the liv-
ing species by any discrete gaps or episodes of rapid change. A workable
definition to accommodate ancestors is:
'A species is a set of contemporary individuals that are capable of inter-

breeding to an extent that precludes the coexistence in the same place of
discrete subsets, along with their descendants and ancestors that are
demonstrably part of the same lineage and are not separated from the set
by a discrete episode of rapid change'.
17.7 CONCLUSION
Although recent techniques of karyology, molecular biology and cladistic
analysis have significantly increased the objectivity of classification, most
studies still suffer from fundamental problems of inadequate sampling,
making it difficult to answer such questions as: how representative of the
entire genome are the few genes sequenced?; or how representative of the
entire species are the individuals examined? It will often be difficult to
answer such questions but at least they should always be asked.
Whatever species concept is used, there is a need to bridge the gap
between observed sample and the living population. We need to weigh
up the probability that two morphologically distinct forms are interfertile;
or that two diagnosable samples represent two equally diagnosable taxa in
nature. Mathematical techniques may assist in the assessment of such
probabilities, but will never be a complete substitute for human judge-
ment based on a wide knowledge of the taxonomy, ecology and behav-
iour of the organisms concerned.
New data relative to the discrimination of mammalian species are
appearing at an accelerating rate, much of it from sources unconnected
with traditional taxonomy. However, unless these data are integrated
with the existing body of taxonomic knowledge, following the disciplines
learned over the past 250 years, much will not effectively enhance our
understanding of biodiversity.
17.8 REFERENCES
Aimi, M. and Dakar, A. (1992) Taxonomy and distribution of Presbytis melalophos
group in Sumatra, Indonesia. Primates, 33(2), 191-206.
Alien, J.A. (1876) Geographical variation among North American mammals espe-
cially in respect to size. Bulletin of the U.S. Geological and Geographical Survey of the
Territories, 2 (4).
Bohme, W. and Eisentraut, M. (1990) Zur weiteren Dokumentation des
Zwergelefanten. Zeitschrift Kolner Zoo, 33(4), 153-8.
Catzeflis, F., Corbet, G.B., Furano, A., Orth, A., Usdin, K. and Volobuev, V.
Morphological, chromosomal and genie differences between Rattus rattus and
a sibling, sympatric species of South India (unpublished).
Corbet, G.B. (1964) The grey long-eared bat Plecotus austriacus in England and the
Channel Islands. Proceedings of the Zoological Society of London, 143,511-15.
Corbet, G.B. (1966) The Terrestrial Mammals of Western Europe, Foulis, London.
Corbet, G.B. (1970) Patterns of subspecific variation. Symposia of the Zoological
Society of London, 26,105-16.
Corbet, G.B. (1975) Examples of short- and long-term changes of dental pattern in
Scottish voles (Rodentia: Microtinae). Mammal Review, 5,17-21.
Corbet, G.B. and Clutton-Brock, J. (1984) Appendix: taxonomy and nomenclature, in
Evolution of Domesticated Animals (ed. I.L. Mason), Longman, London, pp. 434-8.
Corbet, G.B. and Hill, J.E. (1992) The Mammals of the Indomalayan Region: a Systematic
Review, Oxford University Press, Oxford.
Corti, M., Capanna, E. and Estabrook, G.F. (1986) Microevolutionary sequences in
house mouse chromosomal speciation. Systematic Zoology, 35,163-75.
Cummings, W.C. (1985) Right whales Eubalaena glacialis (Muller, 1776) and
Eubalaena australis (Desmoulins, 1822), in Handbook of Marine Mammals (eds S.H.
Ridgeway and R. Harrison), vol. 3, Academic Press, London, pp. 275-304.
Dannelid, E. (1991) The genus Sorex (Mammalia, Soricidae) - distribution and evo-
lutionary aspects of Eurasian species. Mammal Review, 21(1), 1-20.
Eisentraut, M. and Bohme, W. (1989) Gibt es zwei Elefantenarten in Afrika?
Zeitschrift Kolner Zoo, 32(2), 61-8.
Ellerman, J.R. and Morrison-Scott, T.C.S. (1951) Checklist of Palaearctic and Indian
Mammals, British Museum (Natural History), London.
Ellerman, J.R., Morrison-Scott, T.C.S. and Hayman, R. (1953) Southern African
Mammals 1758 to 1951: a Reclassification, British Museum (Natural History),
London.
Gaukler, A. and Kraus, M. (1970) Kennzeichnen und Verbreitung von Myotis
brandti (Eversmann, 1845). Zeitschrift filr Saugetierkunde, 35,113-24.
References 355
Groves, C.P. (1993) Order Primates, in Mammalian Species of the World/ a Taxonomic
Review, 2nd edn (eds D.E. Wilson and D.M. Reeder), Smithsonian Institution
Press, Washington, DC, pp. 243-77.
Groves, C.P., Grubb, P. and Bricknell, S. (1993) Taxonomy of African elephants,
Abstracts, 65th International Theriological Congress, p.120.
Hagmeier, E.M. (1958) Inapplicability of the subspecies concept to North American
marten. Systematic Zoology, 7,1-7.
Hall, E.R. (1943) Criteria for vertebrate subspecies, species and genera: the mam-
mals. Annals of the New York Academy of Sciences, 44,141-4.
Hall, E.R. (1951) American weasels. University of Kansas Publications, Museum of
Natural History, 4,446 pp.
Hall, E.R. and Kelson, K.R. (1959) The Mammals of North America, Ronald Press,
New York.
Handwerk, J. (1987) Neue Daten zur Morphologic, Verbreitung und Okologie der
Spitzmause Sorex araneus und S. coronatus im Rheinland. Banner Zoologische
Beitrage, 38,273-97.
Hausser, J., Catzeflis, F., Meylan, A. and Vogel, P. (1985) Speciation in the Sorex ara-
neus complex (Mammalia: Insectivora). Acta Zoologica Fennica, 170,125-30.
Heyning, J. and Perrin, W.F. (1994) Evidence for two species of common dolphin
(genus Delphinus) from the eastern North Pacific. Contributions in Science,
Natural History Museum Los Angeles County, 442,35 pp.
Ishak, B., Warter, S., Dutrillaux, B. and Rumpler, Y. (1992) Chromosomal
rearrangements and speciation of sportive lemurs (Lepilemur species). Folia
Primatologia, 58,121-30.
Linnaeus, C. (1758) Systema Naturae, (10th edn), Stockholm.
Marshall, J. and Sugardjito, J. (1986) Gibbon systematics, in Comparative Primate
Biology, Systematics, Evolution and Anatomy (eds D.R. Swindler and J. Erwin), vol.
1, Alan R. Liss, New York, pp. 137-85.
Mayr, E. (1940) Speciation phenomena in birds. American Naturalist, 74, 249-78.
Massachusetts.
Miller, G.S. (1901) [Review of A Synopsis of the Mammals of North America and the
Adjacent Seas, by D.G. Elliot], Science, NS, 14,25-7.
Miller, G.S. (1924) List of North American Recent mammals. Bulletin of the United
States National Museum, 128, xvi + 673 pp.
Muller, J. (1954) Observations on the orbital region of the skull of the Mystacoceti.
Zoologische Mededeelingen, 32(23), 279-90.
Musser, G.G. and Carleton, M.D. (1993) Family Muridae, in Mammal Species of the
World: A Taxonomic Review (eds D.E. Wilson and D.M. Reeder), Smithsonian
Institution Press, Washington, DC, pp. 501-755.
Nei, M. (1972) Genetic distance between populations. American Naturalist, 106,
283-92.
Nevo, E. (1989) Modes of speciation: the nature and role of peripheral isolates in
the origin of species, in Genetics, Speciation and the Founder Principle (eds L.V.
Giddings, K.Y. Kanshiro and W.W. Anderson), Oxford University Press, New
York, pp. 205-36.
Niethammer, J. and Krapp, F. (ed.) (1990) Handbuch der Sa'ugetiere Europas, band 3/1,
Insektenfresser-Insectivora/ Herrentiere - Primates, Aula-Verlag, Wiesbaden.
Omura, H. (1958) North Pacific right whale. Scientific Reports of the Whales Research
Institute, Tokyo, 13,1-52.
Omura, H., Ohsumi, S., Nemoto, T., Nasu, K. and Kasuya, T. (1969) Black right
whale in the North Pacific. Scientific Reports of the Whales Research Institute,
Tokyo, 21,1-78.
Ortello, M.O. (1995) Phylogenetic analysis of G-banded karyotypes among the
South American subterranean rodents of the genus Ctenomys. Biological Journal
of the Linnean Society, 54,43-70.
Quicke, D.L.J. (1993) Principles and Techniques of Contemporary Taxonomy, Blackie A
& P, London.
Ruvolo, M, Pan, D., Zehr, S., Goldberg, T., Disotell, T.R. and Dornum, M. (1994)
Gene trees and hominid phylogeny. Proceedings of the National Academy of
Science, USA, 91(19), 8900-4.
Savic, J.R. and Nevo, E. (1990) The Spalacidae: evolutionary history, speciation and
population biology, in Evolution of Subterranean Mammals at the Organismal and
Molecular Levels. Progress in Chemical and Biological Research, 335,129-53.
Schaeff, C, Kraus, S., Brown, M., Perkins, J., Payne, R., Gaskin, D., Boag, P. and
White, B. (1991) Preliminary analysis of mitochondrial DNA variation within
and between the right whale species Eubalaena glacialis and Eubalaena australis.
Reports of the International Whaling Commission, Special Issue, 13,217-23.
Sharman, G.B. (1956) Chromosomes of the common shrew. Nature, 177, 941-2.
Skinner, J.D. and Smithers, R.H.N. (1990) The Mammals of the Southern African
Subregion, University of Pretoria, Pretoria.
Thaeler, C.S. (1985) Chromosome variation in the Thomomys talpoides complex. Acta
Zoologica Fennica, 170,15-18.
Trouessart, E.L. (1897-1905) Catalogus Mammalium tarn Viventum quam Fossilium,
Friedlander, Berlin,
van Bree, P.J.H. and Gallagher, M.D. (1978) On the taxonomic status of Delphinus
tropicalis van Bree, 1971. Beaufortia, 28,1-8.
Wada, S. and Numachi, K. (1991) Allozyme analysis of genetic differentiation
among the populations and species of the Balaenoptera. Reports of the
International Whaling Commission, Special Issue, 13,125-54.
Yosida, T.H. (1980) Cytogenetics of the Black Rat, University of Tokyo Press, Tokyo;
University Park Press, Baltimore.
18
The ideal species concept - and
why we can't get it
D. L Hull
Contacting address: Department of Philosophy, Northwestern University, Evanston,
IL 60208, USA
ABSTRACT
Ideally scientists would like their concepts to be as general, applica-
ble and theoretically significant as possible. Unfortunately, these
goals tend to conflict with each other, that is, one goal can be realized
only at the expense of other, equally desirable goals. For example,
theoretical significant species concepts tend not to be very opera-
tional. Attempts to make them more operational result in their being
theoretically less significant. Recent suggestions for improving the
species concept can be seen as attempting to realize one or more of
these three goals. The test cases for each of the species concepts
examined are asexual organisms, hybrid species and the inclusion of
males and females with each other and their offspring in the same
species.
18.1 INTRODUCTION
Cracraft (1997: Chapter 16) compares the search for the ideal species con-
cept to the quest for the Holy Grail, and the comparison is not far off the
mark. The temptation has always been to hope that, if we can only for-
mulate just the right definition, all our problems will be solved. Enough
time has passed and enough energy expended to convince quite a few of
us that no magic bullet exists for the species concept (Mishler and Theriot,
1997). Any species concept, no matter which one we choose, will have
some shortcoming or other. Either it is only narrowly applicable, or if
applicable in theory, not in practice, and so on. One problem is that dif-
ferent systematists have different goals for their species concepts, but even
those systematists who agree in principle on what a species concept
358 The ideal species concept - and why we can't get it
should do frequently prefer different species concepts. The trouble is that
we have several criteria that we would like an ideal species concept to
meet, and these criteria tend to conflict. Most importantly, if a species con-
cept is theoretically significant, it is hard to apply, and if it is easily applic-
able, too often it is theoretically trivial (for additional general treatments of
these issues, see Slobodchikoff, 1976; Otte and Endler, 1989; Ereshefsky,
1992; and Sterelny, 1994 critical notice).
In this chapter I set out three of the most common criteria that concepts
are supposed to meet in science and see how well various species concepts
meet these criteria. These criteria are universality (or generality), applica-
bility, and theoretical significance. Scientists want their concepts to be as
general as possible. For example, physicists intend their concept of physi-
cal element to encompass all physical substances, not just a subset.
Biologists have had a more difficult time formulating a species concept
that encompasses all organisms. The two phenomena that have proved to
be the most intractable for species definitions are asexual reproduction
and hybridism. One sign that scientists really do value generality is the
increasing attention that these two problem cases are receiving (Cracraft,
1983: 171; Donoghue, 1985: 179; Templeton, 1989: 8,10,11; Echelle, 1990:
111; Nixon and Wheeler, 1990: 219; McDade, 1990, 1992; Vrana and
Wheeler, 1992: 70; Mishler and Theriot, 1997). In this chapter I treat these
two phenomena as test cases for proposed species concepts.
An issue closely connected to generality is monism. In its most extreme
form monism is the view that a single way exists for dividing up the world
into kinds and organizing these kinds into a single hierarchy of laws.
Although no such monistic explanation of the empirical world currently
exists, it is the goal toward which many scientists strive. Since Einstein
physicists have attempted to produce a unified field theory. Right now
monism is decidedly out of favour in certain philosophical circles. The
sophisticated, not to mention politically correct, position is pluralism.
Pluralism comes in a spectrum of forms from promiscuous pluralism to
more moderate positions. The major claim that pluralists make is that no
unified picture can be presented of nature. We can view living things from
a variety of perspectives, and each of these perspectives is legitimate. We
can view living organisms from a genealogical perspective and classify
them accordingly. Or we can produce ecological classifications that ignore
genealogy. Or we can organize organisms into diagnosable units without
any attention to genealogy, ecology, embryology, etc. Pluralists maintain
that some of these ways may be preferable to others for certain purposes,
others may be preferable for other purposes, but none of these perspec-
tives is any more fundamental than any other (Kitcher, 1984,1989; Dupre,
1993).
Applicability and theoretical significance tend to be in opposition to
each other. The more theoretically significant a concept is, the more diffi-
Introduction 359
cult it is to apply. A repeated theme in the taxonomic literature is that
species concepts should be as operational as possible. Species should be
defined with an eye to the sorts of data available to systematists. Since the
most easily available and widespread data are patterns of morphological
variation, species should be defined in terms of some sort of morphologi-
cal similarity. Other systematists insist that such characters are not ends in
themselves but are evidence to be used to infer something else - some-
thing a good deal more theoretically important than morphological simi-
larity (Endler, 1989 uses theoretical versus operational as one of his four
criteria for evaluating species definitions).
Temporal considerations tend to come into play at this juncture. Both
observation and theory are important in science, but a belief common
among scientists is that, to be genuinely scientific, scientists should begin
with observations and only much later proceed to speculate about more
theoretical issues. Thus, in evaluating alternative species concepts, theo-
retical significance poses two problems: theoretical input and theoretical
output. Epistemically conservative systematists will allow that classifica-
tions can be legitimately used as a basis for theoretical speculations but
that no theoretical considerations should be allowed into the formulation
of these classifications. Other systematists insist that no such thing as a
theory-neutral classification exists. Theoretical considerations should and
do enter into classification right from the start.
The epistemically conservative position is reflected in the chapters by
several authors in this volume. For example, Gornall (1997: Chapter 8)
recommends using a working definition of species that is devoid of
much, if any theoretical background. Hawkes (1997: Chapter 9) thinks
that systematists tend to agree with each other, except when they are
encumbered by theoretical baggage. Claridge et al. (1997: Chapter 1) con-
cur, complaining that objectivity may often be clouded by adherence to
particular theories of speciation. Most authors in this volume see no rea-
son why our knowledge of the empirical world - all of it - cannot be used
in constructing our classifications.
Classifications in terms of observable characters are not only more cer-
tain, so epistemically conservative authors claim, but in addition they are
also more practical or useful. Systematists frequently justify their contin-
ued existence by reference to practical uses of classifications, e.g. identify-
ing insect vectors. Such justifications have some point. For example, in the
United States we have an Endangered Species Act. Lumpers recognize
widespread species that are unlikely to be endangered, while splitters
identify not only more species but more restricted species that are likely to
be more vulnerable to extinction. Thus, as Mayden (1997: Chapter 19) has
argued, which species concept we adopt has very direct and important
effects on something as practical as the mass extinction currently under
way. More importantly, if systematists give the impression that species
recognition is largely an arbitrary affair, the justification for attempting to
save endangered species is weakened. Thus, even the most philosophical
exercises in science can have very practical effects (Eisner et al., 1995).
But all the effort devoted to honing our species concepts also has effects
on science in general. For example, the title of this volume is Species: The
Units of Biodiversity, but as John and Maggs (1997: Chapter 5) point out,
accurate estimates of biodiversity require sound species concepts. We
often hear about the error of comparing apples and oranges, but if differ-
ent workers use different species concepts in estimating biodiversity, then
the results of all their efforts will not be comparable. They will have truly
been comparing apples and oranges. Typically, higher taxa are not very
comparable. Patterns of distribution of the same organisms differ for gen-
era, families and orders (Signer, 1985). If species are no more comparable
than higher taxa, then we are all in real trouble. To put the point more
strongly, since species are not comparable, then we are in deep trouble if
we do not take this fact into account (for a more general evaluation of sys-
tematics and the species category than provided here, see Frost and
Kluge, 1994).
18.2 PRESENT-DAY SPECIES CONCEPTS

In this chapter I propose to evaluate the most salient species concepts in
use today on the three criteria listed above to see how well each of these
concepts does. I classify these species concepts into three basic kinds. First
one group of species concepts can be lumped together because they all
require similarity of some sort. Similarity concepts include traditional mor-
phological species concepts, the phenetic species concept, as well as cer-
tain molecular concepts. In each~case, systematists look for some sort of
overall similarity and/or gaps in character distributions. The chief differ-
ence between the phenetic species concept and earlier morphological
species concepts is that early systematists tended to be typologists in the
sense that each species (not to mention each higher taxon) must be
defined in terms of universally covarying characteristics, while the phe-
neticists treat all Operational Taxonomic Units as being polythetic, that is,
they can be defined only in terms of statistically covarying characteristics
(see Van Regenmortel, 1997: Chapter 2).
When the notion of polythetic definitions of taxa was first introduced
into systematics, it looked very promising (Hull, 1965). However, it has
turned out to have some major drawbacks. Chief of these is that any
group of organisms can be classified in indefinitely many ways using
various clustering techniques, and no reasons internal to these methods
exist for choosing among these classifications (see also de Queiroz and
Good (1997) for a more extensive critique of phenetic clustering mecha-
nisms).
Present-day species concepts 361
The second groujj of species concepts that I evaluate includes Mayr's
(1969) biological species concept, Simpson's (1961) evolutionary species
concept, especially as it has been reworked by Wiley (1981), Paterson's
(1981) species mate recognition concept, and Templeton's (1989) cohe-
sion concept. What all these concepts have in common is theoretical
commitment, especially a commitment to evolutionary theory. Because
the evolutionary process operates in a particular way, species are to be
individuated accordingly. Each of these theoretically committed species
concepts is designed to highlight some significant level of organization
in the evolutionary process.
I term my third group of species concepts phylogenetic to indicate their
historical connection to Hgnnig's phylogenejic__SYst£matics (Mishler and
Theriot, 1996). The two most prominent phylogenetic species concepts are
the monophyletic (or autapomorphic) species concept and the diagnostic
species concept. The monophyletic species concept as developed by Mishler
and Donoghue (1982), Mishler and Brandon (1987), and de Queiroz and
Donoghue (1988,1990) among others, has its origins in Hennig as well as
Rosen (1978, 1979). According to the monophyletic species concept, a
species is the least inclusive monophyletic group definable by at least one
autapomorphy. The reference to autapomorphies explains why this con-
cept is also termed the autapomorphic species concept.
The diagnostic species concept also has its historical roots in the work of
Hennig and of Rosen (1978, 1979), but it has taken its own direction; e.g.
Platnick (1977), Eldredge and Cracraft (1980), and Nelson and Platnick
(1981). The most influential formulation of the diagnostic species concept
was first formulated by Cracraft (1983) and then further developed by
McKitrick and Zink (1988), Nixon and Wheeler (1990), Wheeler and Nixon
(1990), and Vrana and Wheeler (1992) among others. According to Cracraft
(1983:170), a species is the smallest diagnosable cluster of individual organ-
isms within which there is a parental pattern of ancestry and descent.
Cracraft includes reference to parental pattern of ancestry and descent (or
cohesion) to avoid one of the prevailing problems of defining species pure-
ly in terms of similarity- the failure sometimes to include males and females
in the same species with each other as well as with their progeny.
By now, the reader should have noticed that my classification of species
concepts involves a degree of cross-classification. The monophyletic
species concept is as theoretically committed as are the concepts of Mayr,
Simpson, Wiley, Paterson and Templeton. However, I have separated it
from these earlier concepts because of its connection to cladistics.
Conversely, the diagnostic species concept could have easily been classi-
fied with the morphological and phenetic species concepts because of its
concern to be as operational as possible. Once again, I have classified it
along with the monophyletic species concept to highlight its connection to
Hennig.
Although the phylogenetic species concepts were developed by
Hennig's intellectual descendants, they form two quite distinct conceptu-
al transformation series (Platnick, 1979; Nixon and Wheeler, 1990). They
both start with concerns important to Hennig but then take paths of their
own. For example, the diagnostic species concept eventually came to look
very similar to pre-Darwinian morphological species concepts. In general,
phylogenetic cladists such as Wiley and Donoghue remain truer to
Hennig's original system than do such pattern cladists as Nelson and
Platnick. In fact, pattern cladists have reworked Hennig's system so radi-
cally that it begins to look like the systems of ideal morphology that
Hennig was concerned to refute. But in the identification of conceptual
transformation series, what counts is descent, not similarity. As far as sim-
ilarity is concerned, the diagnostic species concept may look suspiciously
like the species concepts of Hennig's opponents, but as far as descent is
concerned, it belongs in an Hennigian transformation series.
The most famous of the theoretically committed species concepts is
Mayr's biological species concept. According to Mayr (1969: 26), species
are groups of interbreeding natural populations that are reproductively
isolated from other such species. Mayr's species concept is designed to
distinguish between those groups of organisms in which gene flow is
significant and those in which some isolating mechanism or other has
interrupted gene flow.
One common criticism of the biological species concept is that it is not
sufficiently operational and will be discussed in the next section. Critics
have also alleged that each species according to Mayr must possess an iso-
lating mechanism capable of insulating it against all possible species. To
the contrary, each species needs to have a mechanism that isolates it only
from those species with which it happens to come into contact. If it pos-
sesses such a mechanism, it remains distinct from this other species. If not,
the two merge into a single species (Horvath, in press). The biological
species concept is also non-dimensional. It applies only to organisms dur-
ing a relatively short time span. A biological species is a time-slice of an
evolving lineage. It also, quite obviously, applies only to organisms that
reproduce sexually. (The notion of the temporal dimensionality of species
has not been given the attention it properly deserves. For now, all it indi-
cates is some unspecified but relatively short period of time.)
Simpson's evolutionary species concept extends the biological species
concept through time. According to Simpson (1961: 153) evolutionary
species evolve separately from other such lineages and possess their own
unitary roles and tendencies. Wiley (1981: 25) reworked Simpson's evolu-
tionary species concept, replacing unitary roles and tendencies with evo-
lutionary tendencies and historical fate. Wiley's interpretation of species
also differs from Simpson's interpretation in that Wiley is concerned to
make his species concept compatible with the phylogenetic system of
Present-day species concepts 363
Hennig. For example, Simpson was willing to subdivide a gradually
evolving lineage into successive chronospecies while Wiley is not.
Paterson (1981) caused quite a stir with his specific mate recognition
concept. According to Paterson, each species has its own fertilization sys-
tem which enables organisms belonging to that species to mate. The chief
virtue of this definition according to Paterson is that it defines each species
in terms of what binds it together rather than in terms of indefinitely
many isolating mechanisms that are supposed to keep it separate from
indefinitely many other species (Horvath, in press). For this reason, he
terms the biological species concept the isolation concept. One common
response to Paterson's species concept is that the isolation and recognition
concepts are just two sides of the same coin (Templeton, 1989: 7), a con-
tention which obviously does not sit all that well with Paterson (1981: 26).
One point to notice is that the species mate recognition concept, like the
biological species concept, applies only to organisms that reproduce sexu-
ally at least sometimes.
Finally, Templeton formulated his cohesion species concept in order to
synthesize the preceding theoretically committed species concepts.
According to Templeton (1989:12), a species is the most inclusive popula-
tion of individuals having the potential for phenotypic cohesion through
intrinsic cohesion mechanisms. Among these intrinsic cohesion mecha-
nisms are gene flow and natural selection as well as ecological, develop-
mental and historical constraints. The cohesion species concept is more
general than either the biological species concept or the species mate
recognition concept because it applies to asexual organisms via their adap-
tation to particular ecological niches.
According to Hennig (1966) his principles of phylogenetic systematics
come into play only when reticulation gives way to splitting (phylogenet-
ic relations). This is also the point, according to Hennig, at which species
are to be recognized. According to Hennig, all higher taxa must be strictly
monophyletic. Species to the contrary cannot be monophyletic because
monophyly does not even apply to species. However, a group of Hennig's
intellectual descendants have argued that monophyly can be extended far
below the species level. For example, a series of populations can exhibit a
tree-like structure. The result is the monophyletic species concept. Species
are monophyletic groups of organisms characterized by at least one
autapomorphy.
In addition, the monophyletic species concept is synchronous with
respect to its grouping criterion because monophyly in the sense of sister-
group relations has no temporal dimension. All taxa are arrayed along the
top of a cladogram. Some may be older than others, some may even be an
ancestor of one or more of the other taxa included in the cladogram, but
the only relationship indicated in a cladistic classification is that of sister-
group relations and they are synchronous. However, the monophyletic
species concept assumes at least a minimal time-dimension because its
ranking criteria do take at least some time to function.
The monophyletic species concept is, in principle, monistic with
respect to its grouping criterion: monophyly and only monophyly mat-
ters. However, advocates of the monophyletic species concept also allow
the naming of metaspecies, the smallest unresolved groups that have not
yet been shown to be monophyletic, just as long as they are distin-
guished symbolically from true monophyletic species (Donoghue, 1985:
179). The monophyletic species concept is also pluralistic in a weak sense
with respect to its ranking criteria because it makes reference to more
than one ranking criterion. Pluralism in the strongest sense requires that
quite different and incompatible species definitions be considered equal-
ly good in their own domain. Here, the authors are providing a single
definition but include more than one way of fulfilling the requirements
of this definition.
The diagnostic species concept differs from the monophyletic species
concept in that the diagnostic characters that define species need not be
autapomorphies. Instead, any unique combination of primitive and
derived characters that have become fixed in a group will do (Cracraft,
1983: 170). In a later reformulation, Cracraft (1989: 34-35) modifies the
diagnostic species concept to refer to a phylogenetic species as being an
irreducible (basal) cluster of organisms which are diagnosably distinct
from other such clusters (see also Nixon and Wheeler, 1990). Diagnostic
characters need not be present in all the organisms that belong to a species
throughout their development. Instead they can be present in just males,
just females, or at some particular stage in their development. However,
with these provisions in mind, diagnostic characters must be universally
distributed, although not uniquely derived characters (autapomorphies).
As the preceding brief characterizations of theoretically committed
species concepts have indicated, some are dimensional with respect to
time; that is, species are explicitly treated as lineages extended through
time (e.g. the evolutionary species concepts of Simpson and Wiley).
Terming the concepts of Mayr, Paterson and Templeton non-dimensional
is a bit too strong. The processes by which they define species take at least
some time to operate. Gene flow, for instance, is not instantaneous. In
addition, in various places, advocates of these species concepts, explicitly
state that their non-dimensional species are intended to be time-slices of
evolving lineages (Templeton, 1989: 20; Nixon and Wheeler, 1990: 218,
220; Mishler and Theriot, 1997).
The phenetic species concept, however, is radically non-dimensional.
Phenetic species are not intended to be lineages or time-slices of lineages.
Time simply does not enter into the equation. Phenetic species are defined
by means of phenetic clusters of characters with no attention whatsoever
to time. The importance of the distinction between dimensional and non-
Universality and monism 365
dimensional species concepts is attested to by Endler (1989) selecting it to
be one of his four criteria of evaluation.
Because species are evolving lineages according to the evolutionary
species definition, this definition has both retrospective and prospective
dimensions. What counts as a species is determined in part by what has
happened to it in the past, in part by what happens to it in the future.
Reference to evolutionary fates seems to require that we can predict the
future. Thousands of times a population begins to bud off from the main
body of its species only to be reabsorbed later. However, every once in a
while, one of these populations becomes established and is transformed
into a new species. In retrospect, that founding population is considered
part of this new species. If speciation had not been successful, it would
simply be included as an ordinary population with the rest of its species
(for details, see Kornet, 1993 and Graybeal, 1995).
But, critics note, one thing that we cannot do in biological evolution is
predict the course of particular lineages. Of all the species of fruit flies,
which will go extinct first, which will last the longest without going
extinct, and which will speciate? I think I am safe in saying that no one
knows. We can make some general observations, e.g. that small, geneti-
cally homogeneous species are more likely to go extinct than their larger,
more heterogeneous relatives, but that is all. It is difficult enough to
decide when a particular lineage has its own evolutionary role and ten-
dencies. How can we possibly tell which lineages will retain their own
fate? The answer is that we cannot. But this is the price that we pay for
treating species as lineages. Lineages can be recognized only in retrospect
(Sober, 1984: 339; de Queiroz and Donoghue, 1988: 330).
O'Hara (1993: 238-242) argues that the evolutionary species concept is
clearly a lineage concept, but even such apparently non-dimensional
species concepts as the biological, recognition, monophyletic and diag-
nostic species concepts include notions of fate, temporariness and perma-
nence. As a result, they too cannot be applied with certainty in the present
because they all depend on the future. However, he does not regard
future dependence as a flaw of these species concepts; it is simply an une-
liminable characteristic of them. I might add that the retrospective charac-
ter of evolving species is only to be expected if species are individuals
(Ghiselin, 1974; Hull, 1976).
18.3 UNIVERSALITY AND MONISM

Given the brief descriptions of current species concepts provided above,
the question becomes how well each of these species concepts does
according to the criteria commonly used to evaluate scientific terms -
universality, applicability and theoretical significance. As I mentioned
previously, scientists would like their concepts to be as general as possible.
For example, any definition of element in physics should apply to all mat-
ter, not just a restricted subset. Physicists would be less than pleased if
their element concept applied only to metals or to non-radioactive sub-
stances. One virtue of atomic number is that all material substances are
made up of atoms, and every atom has some atomic number or other.
Atomic number applies across all substances and divides them into strict-
ly comparable kinds. The question is whether a species definition can be
found in biology that does the same for biological kinds. Does a single
level of organization exist across all organisms that deserves to be recog-
nized as the species level? Anyone answering the preceding question in
the affirmative is a monist.
The choice between monism and pluralism is not easy, especially if
attention is paid to self-reference. Scientists who are monists with respect
to their own subject matter typically are also monists with respect to the
philosophical issue of monism. For example, Mayr has argued for
decades that the biological species concept and only the biological species
concept is adequate for understanding the living world. And each subject
matter must be explained in one and only one way. The evolutionary
perspective is basic. Pluralists surprisingly also tend to be monists when
it comes to pluralism. Pluralists argue that scientists do and should pro-
duce a variety of incommensurable classifications from a variety of per-
spectives. Any discomfort that scientists may feel about such conceptual
riches simply stems from their inexcusable bias toward monism.
However, as monistic as it may sound, pluralists insist that pluralism is
the only acceptable view of science (Kitcher, 1984,1989; Dupre, 1993). In
general, investigators find pluralism attractive in someone else's disci-
pline but tend to be monists with respect to their own (see also Mishler
and Brandon, 1987; de Queiroz and Donoghue, 1988; and Sterelny, 1994
for criticisms of Kitcher-type pluralism).
With respect to the monism/pluralism issue, several of the species con-
cepts discussed above appear to be strictly monistic. For example, the
phenetic species concept refers to phenetic similarity and only phenetic
similarity. It is also universal because it applies to all organisms regardless
of any other considerations. It applies to asexual organisms as readily as
to those that reproduce sexually. Because Sokal and Sneath (1963: 174)
are not attempting to represent phylogeny in their phenetic classifica-
tions, the only problem that interspecific hybrids pose for them is a pos-
sible increase in borderline cases between parental clusters.
As long as the pheneticists assume that some one thing exists out there
in nature properly termed overall similarity, then the phenetic species con-
cept is monistic. Species are the smallest group exhibiting the appropriate
degree of overall similarity. However, quite early on, pheneticists began to
have their doubts about overall similarity and the general-purpose classifi-
cations that supposedly reflect overall similarity. For example, Ehrlich and
Ehrlich (1967) concluded that there is no such thing as overall similarity or
general-purpose classifications.
Such critics as Johnson (1968) and Ghiselin (1969) added their voices to
these suspicions. For a group like the pheneticists, who view themselves
as being hard-headed and highly empirical, overall similarity eventually
came to look very much like a metaphysical concept. Even Sneath (1995)
seems to have given up the notion of overall similarity. In defence of phe-
netic taxonomy, he argues that the power of overall similarity measures to
construct taxonomic groups, to determine evolutionary relationships, and
for identification, has been amply borne out, but he adds, even if some-
what different forms of similarity may be needed for different purposes.
From the start, one of the most serious problems with respect to the
phenetic species concept was that no reasons could be given to choose one
level of phenetic similarity over any other to term the species level. Even
so, all that mattered was phenetic similarity. In this respect the pheneti-
cists were monists. But now that phenetic similarity comes in different
forms, pheneticists have joined the pluralist camp in a weak sense of plu-
ralism. However, the phenetic species concept remains universal. To the
extent that it can be applied at all, it applies to all organisms.
In the past, certain limitations to prevailing species concepts, though
acknowledged, tended to be relegated to the last paragraph or two in a
chapter on the species concept. Such is the case with asexual reproduction
and species of hybrid origin. Authors tended either to argue these prob-
lem cases away or to claim that they occur very rarely, if at all. For exam-
ple, advocates of the biological species concept attempt to minimize this
problem by observing that most organisms reproduce sexually, at least
some times, but it should be kept in mind that very little in the way of
gene exchange occurred during the first half of life on Earth, and meiosis
evolved even later. According to the biological species concept, no species
existed for at least the first half of life on Earth. Evolution occurred but in
the absence of species. As strange as this observation may sound, it is no
stranger than the recognition that multicellularity also did not evolve until
quite late. We can no more insist that all organisms form biological species
than we can insist that all organisms be multicellular.
Species of hybrid origin do not present as serious a problem to the bio-
logical species concept as do asexual organisms. According to the Mayr,
reproductive isolation builds up very slowly. If the biological species con-
cept is applied sequentially through time, a single unproblematic species
will be found to give way to a series of problem cases until two distinct
species emerge. However, if reproductive isolation can build up, it might
also break down. The breakdown of reproductive isolation provides the
same problem as does its original build up. In this case, two unproblemat-
ic species either gradually merge into one or produce hybrid individuals
through mechanisms such as allopolyploidy.
When Mayr (1969: 32) addresses the issue of hybridism, he does his best
to downgrade its pervasiveness and importance. He notes that the taxo-
nomic literature records a number of instances in which two parental
species have supposedly merged into a single new species but regards
these instances as not being established unequivocally. Introgression is a
populational affair, but hybrids can arise at the level of organisms as well
(e.g. via allopolyploidy). Mayr is unable to dismiss these examples so read-
ily. Life would be easier if reproductive isolation, once established, never
broke down, but it does so, especially in plants.
Two points are at issue with respect to interspecific hybrids: one is
empirical while the other is conceptual. The empirical issue is whether or
not hybrid species actually occur in nature. As inconvenient as they may
be, it seems as if they do. The conceptual issue is how to classify hybrids if
they do occur. Right from the start, the formal similarity between evolu-
tion and classification seemed deceptively obvious. On the one hand, evo-
lution is primarily a matter of successive splittings. On the other hand, the
Linnaean hierarchy consists of a series of successive subdivisions. Hence,
one should be easily mapped onto the other.
If Hennig and his descendants have done anything for the systematic
community, it has been to show that this impression is seriously mistaken.
It is easy enough to draw one line splitting in two; it is not so easy to rep-
resent such an occurrence in a Linnaean classification. Classifying com-
mon ancestors with their progeny in an unambiguous way has proven to
be a major challenge to systematists. It is also easy to draw two lines merg-
ing into one, but such an occurrence poses exactly the same problem for
Linnaean classifications as does the classification of common ancestors
(Hull, 1979; Wiley, 1981). Although speciation is much more common than
hybridism, the two processes pose the same problem for the comparative
method. Their character distributions cannot be resolved unequivocally
into a series of dichotomous divisions. Instead they produce unresolvable
trichotomies.
Echelle (1990) and McDade (1990, 1992) show that the effects of intro-
ducing hybrid species into a classification are not quite as catastrophic as
some early in-principle arguments implied. McDade (1992: 1329) con-
cludes that hybrids are unlikely to cause breakdown of cladistic structure
unless they are between distantly related parents. However, her results
also indicate that cladistics may not be especially useful in distinguishing
hybrids from normal taxa. In any case, advocates of more recent species
concepts all take the ability of their favoured species concept to handle
hybridism as a decided advantage (Cracraft, 1983: 171; Donoghue, 1985:
179; Templeton, 1989: 10; Nixon and Wheeler, 1990: 220; and Mishler and
Theriot, 1997).
One reason for claiming that the biological species concept and the
species mate recognition concept are two sides of the same coin is that
asexual reproduction and species of hybrid origin pose the same problems
for both. Asexual organisms possess no system for recognizing mates
because asexual organisms do not have mates, and just as such recogni-
tion systems can build up, they can break down. Thus, like the biological
species concept, the species mate recognition system is not universal. It is
more clearly monistic in the sense that each species presumably has one
and only one mate recognition system.
Templeton's cohesion species concept is monistic in the sense that all
that matters is cohesion, but cohesion can be produced by several differ-
ent mechanisms acting independently and in consort. One of these mech-
anisms (adaptation to a particular ecological niche) is designed primarily
to accommodate asexual organisms. As Templeton (1989: 8) remarks, the
asexual world is for the most part just as well (or even better) subdivided
into easily defined biological taxa as is the sexual world. Species of hybrid
origin pose the same problems for the cohesion species concept as they do
for the other theoretical significant concepts.
Advocates of the monophyletic species concept take pains to point out
that asexual organisms form monophyletic groups, whether at the level of
individual organisms or at the level of more inclusive clones (Mishler and
Donoghue, 1982: 491; Donoghue, 1985: 179; Mishler and Brandon, 1987:
406; Budd and Mishler, 1990; Mishler and Theriot, 1997). The mono-
phyletic species concept is universal in that it applies to all organisms. It is
also monistic with respect to its grouping criterion monophyly. Whether
or not one wants to rank all least inclusive monophyletic groups as species
is quite another matter. Mishler and Brandon (1987: 406) argue for a vari-
ety of criteria for ranking, depending on the causal agent judged to be
most important in producing and maintaining distinct lineages in the
groups in question. Hence, the monophyletic species concept is pluralistic
with respect to ranking criteria. However, these authors make it clear that
their sense of pluralism is quite distinct from the more radical views of
Kitcher and Dupre.
Similarly, the diagnostic species concept is universal in that it applies to
all organisms. One or more diagnostic characters and a parental pattern of
ancestry and descent are all that is needed, and asexual organisms fulfil
these requirements (Nixon and Wheeler, 1990: 219). This concept is also
monistic in that no appeal to alternative mechanisms is made. Asexual
organisms form a strictly branching pattern, while sexual organisms form
a messy reticulation. Hybridism only adds to this reticulation at higher
levels of organization. This is as pluralistic as the diagnostic species con-
cept gets.
In sum, the authors discussed in this section acknowledge that general-
ity is a virtue in a species concept. One of the most persuasive arguments
for the phenetic species concept is that it applies equally to all organisms.
No issues of splitting and merger arise because these notions play no role
in phenetic taxonomy. Both Simpson and Wiley give as one reason for
accepting an evolutionary species concept that it applies to sexual and
asexual organisms alike. Cracraft (1997: Chapter 16) states explicitly that
one reason why his diagnostic species concept is superior to other species
concepts is that it is applicable to all groups of organisms.
Thus, we are left with the question with which we began. Is there a sin-
gle level of organization across all organisms that deserves to be recog-
nized as the species level? If the papers in this anthology do anything,
they cast doubt on the monistic reply to this question. Some of the species
concepts that I have discussed do apply to all organisms. For example, the
monophyletic species concept is designed to mark the place where reticu-
lation gives way to successive branchings. This transition is certainly
important and is worth marking. The trouble is that this level varies from
single organisms in asexual organisms, to single populations in some sex-
ual species, to more traditional species, and finally to much larger groups
where interspecific hybridization occurs. From the perspective of the
replacement of reticulation with splitting, this transition forms a single
level. However, from a more common-sense perspective, it wanders from
level to level, depending on the circumstances; for a detailed discussion of
reticulation in a particular group (corals), see Veron (1995).
The diagnostic species concept also applies to all organisms, but it sub-
divides organisms into smaller, more numerous species than systematists
are used to recognizing. If a local population exhibits one or more diag-
nostic characters, it must be considered a species. However, several stud-
ies have shown that this in-principle problem is not as pervasive as one
might think. For example, Cracraft estimates that the 8500 to 9000 species
of birds recognized under the biological species concept would be
expanded to about 18 000 under his diagnostic species concept (see also
Cracraft, 1983: 173; Echelle, 1990: 110; Ereshefsky, 1989: 90; Nixon and
Wheeler, 1990: 219)
A more fundamental problem is that the diagnostic species concept can
divide up organisms into overlapping and incompatible species, depend-
ing on which characters are picked to be diagnostic. Cladistic characters
must nest. Otherwise, they are not genuine characters. But the diagnostic
characters of the diagnostic species concept need not be autapomorphies.
As such, they need not nest.
Some species concepts (e.g. the biological and mate recognition con-
cepts) apply to only a limited number of organisms. Where they apply,
they do seem to mark important groups of organisms. They cut nature
at her joints. But since these definitions do not apply to all organisms,
there is some tendency not to use them even for the organisms to which
they do apply. Pluralists would complain that this response is too monis-
tic. Although I have a strong preference toward monism, I am forced to
agree.
Applicability and theoretical significance 371
18.4 APPLICABILITY AND THEORETICAL SIGNIFICANCE
Scientists are strongly predisposed to make their concepts as operational

as possible. Theoreticians are much less concerned. They formulate
extremely abstract theories about space, time and energy or gene
exchange and density-dependent selection and lesser mortals can worry
about testing these theories. Philosophers tend to side with the theoreti-
cians. The philosophy of science deals primarily with theories and their
development. So does the history of science. The names of theoreticians
that turned out to be right or at least wrong in big ways are chronicled at
great length. We find it difficult to recall the names of more empirically
minded scientists - or 'fact grubbers' as they are termed - with no fond-
ness. We are all familiar with the distinction between batesian and mul-
lerian mimicry, and most of us have heard about the paradigm example of
batesian mimicry (monarch and viceroy butterflies) but few of us are like-
ly to be able to recall the names of the authors who recently put this dis-
tinction to the test with disconcerting results (Ritland and Brower, 1991).
The philosophical arguments against operationism are decisive.
Philosophers do not agree on much, but two propositions that have
gained wide acceptance among philosophers of science are that no theo-
retically significant terms can be totally operationally defined, and that no
terms are totally theory neutral. Although at one time I argued with all the
enthusiasm of a graduate student that these propositions are well-taken
(Hull, 1968), I now think that the attention that scientists pay to formulat-
ing operational criteria for theoretical concepts is appropriate. My early
prejudice led me to be more critical of the operational tendency with
respect to species concepts than I now think I should have been. Maybe
philosophers don't have to apply scientific concepts, but scientists do.
In the previous section I discussed difficulties in applying concepts. In
these discussions of applicability the issue was theoretical limitations, e.g.
the biological and mate recognition concepts do not apply to asexual
organisms. It is not that we lack techniques for applying these concepts
but that they do not apply even in principle. The issues addressed in this
section concern difficulties in applying concepts in practice to cases to
which we know that they apply in principle. For example, as sure as we
know anything, we know that some extinct organisms mated with each
other and thus formed biological species, but such relations must now be
inferred via various sorts of evidence, and sometimes these inferences are
none too secure. In this section, it is this sort of applicability that I will be
discussing.
The importance of applicability can be seen in the terms 'practical' or
'practice' appearing in the titles of five of the contributions to this volume.
The term practice has a nice ring to it (Pickering, 1984). No fancy theories
or idle speculations here. Just the facts and nothing but the facts. From our
present-day perspective it is difficult to recall that the biological species
concept was introduced to make species recognition more practical in the
sense of objective. Early in the history of systematics, species recognition
was largely intuitive. Experts in a particular group developed a knack for
recognizing species in their group. I do not mean to denigrate such intu-
itive reasoning on the part of practising systematists. It still plays an
important role in systematics. If I had limited money and time and need-
ed a good rough-and-ready classification, I would surely turn to an expe-
rienced systematist, regardless of his or her methodological persuasion.
However, three problems plague intuitive species recognition and classi-
fication. One is that systematists who group their organisms according to
implicit criteria on the basis of their individual expertise tend not to agree
with each other, and there are no objective standards by which these dis-
agreements can be arbitrated. Second, such intuitive methods are difficult to
pass on to later generations. The only way to do so is the apprentice system.
Extensive personal contact is necessary, and frequently even it is not
enough. Disciples frequently decide that their master was all wrong and
proceed to rework the classification. Finally, the most destructive feature of
the intuitive species concept is that just when systematists get really good at
what they are doing, they have the unfortunate habit of dying and with
them go all their intuitive abilities. One reason why the biological species
concept looked so attractive initially was that it provided a criterion for dis-
tinguishing between species - gene flow and its interruption.
Later workers take for granted the increased objectivity afforded by the
biological species concept and proceed to raise their standards. They want
concepts that are even more operational. In fact, a major motivation for
nearly all later species concepts is the desire to make them more opera-
tional. The question then becomes how operational must a concept be in
order to be scientifically acceptable? Certain species concepts are very the-
oretical; others are as operational as possible. But even those systematists
who opt for theoretical species concepts acknowledge the need to provide
operational criteria to facilitate their application. A second issue is whether
or not these operational criteria must be included in the definition or
whether they can just be appended to it.
Neither the philosophical nor the scientific literature provides a clear
answer to the first question. Philosophers have argued conclusively that
no theoretically significant term can be completely defined by means of
the techniques used to apply it. Conversely, terms which lack any means
of application are of doubtful use in science. As positivists used to put the
point, such terms are metaphysical. But how about all the terms that fall
somewhere in between these two extremes? Unfortunately, philosophers
have not addressed this question in any detail, in large measure because
the process of operationalizing concepts is highly particularistic. One can
list example after example, but no general pattern emerges. The process of
Applicability and theoretical significance 373
operationalizing concepts seems to be just one damned thing after anoth-
er. Nor do scientists say much about this process. They just do it. For any
student of science who is getting a bit bored with the Covering Law Model
of Scientific Explanation or incommensurability, the topic of operational-
izing concepts is wide open.
Answers to the second question seem more clear-cut. Scientists in
general see no real need to include their criteria of application in their
theoretical definitions. For example, physicists provide definitions of
such theoretical terms as mass, force and length, but they do not include
in these definitions all the different ways that they have developed to
operationalize these concepts. Numerous ways exist for determining
length. Including all these methods in the definition of length would be
unduly cumbersome. Similarly, Mayr refers to isolation in his definition,
but he does not include in his definition all possible isolating mecha-
nisms or ways of determining their presence or influence. The same can
be said for the evolutionary concepts of Simpson and Wiley. The appro-
priate criticism (if valid) is that they do not provide sufficiently detailed
descriptions of how evolutionary roles and tendencies or evolutionary
tendencies and historical fate are to be operationalized.
One virtue of Templeton's cohesion species concept is the lengths to
which he goes to spell out all of the processes that can contribute to cohe-
siveness. Cracraft also refers to cohesion but does not state in his diagnostic
species concept what he means by this term. He expands on this notion in
the body of his text. Similarly, monophyly and autapomorphy are referred
to in the monophyletic species concept but neither are defined or opera-
tionalized in that definition. One major feature of definitions is that they are
relatively short. Putting too much terminological explanation or empirical
data within the body of the definition would be counter-productive.
Certainly one of the major motivations for the phenetic species concept
was to make it as operational as possible. All the taxa in phenetic taxono-
my are to be operational. That is why Sokal and Sneath termed them
Operational Taxonomic Units. In phenetic taxonomy, all elements used
must be spelled out explicitly: the clustering technique employed, the list
of characters used, and the level of similarity to be considered specific. As
little as possible is left to personal judgement. However, pheneticists
thought that this increased objectivity had to be purchased at the expense
of theoretical significance. No general knowledge of the empirical world
can be used in producing phenetic classifications. The fate of operational
taxonomic units among systematists gives some indirect evidence for the
value of such notions. Although quite a few systematists voice approval of
making systematic method as operational as possible, they do not seem
very enthusiastic about using the results of this process.
The authors of both the monophyletic and diagnostic species concepts
prefer to call them phylogenetic, signalling their connection to Hennigian
systematics. One would think, given the central role of autapomorphies in
phylogenetic systematics, that no cladist would have any reservations
about using autapomorphies to define species, but such pattern cladists as
Nixon and Wheeler do. The reason revolves once again around epistemo-
logical issues. Nixon and Wheeler (1990: 217) argue that species concepts
based on topological (cladogram) knowledge are problematic, because
they cannot be implemented prior to cladistic analysis. The diagnostic
species definition is precisely what is needed for cladistic analysis because
diagnostic species can be recognized prior to a cladistic analysis, so that
the analysis can be undertaken.
Previously, I mentioned the difference between phylogenetic and pat-
tern cladists. The chief difference between the two groups is that pattern
cladists want their conclusions to be more certain and their methods and
definitions to be more operational than do the phylogenetic cladists. Thus,
pattern cladists have been compared with the pheneticists in wanting
classificatory process to be theory-free. Systematists must start with obser-
vations. This position is somewhat surprising given all the attention that
pattern cladists have paid to Karl Popper, the author of the view that no
terms in science, even the most observational, can be totally theory-free.
Nor can scientists begin their investigations with observations and noth-
ing but observations. In any case, the diagnostic species concept is more
operational than the monophyletic species concept because its diagnostic
characters need not be autapomorphies.
18.5 SUMMARY AND CONCLUSION

Several criteria are used commonly to evaluate scientific concepts. Others
apply more narrowly to species concepts. How do the seven species con-
cepts that I have discussed in this paper score on these criteria? As far as
universality is concerned, the phenetic species concept is the most gener-
al concept because it does not purport to reflect very much about the nat-
ural world. Males and females need not be included with each other or
with their progeny in the same lowest level operational taxonomic unit. At
the other extreme, the biological and mate recognition concepts are the
least general because they apply only to species that reproduce sexually.
The other four species concepts are arrayed somewhere in between these
two. The diagnostic species concept is closest to the phenetic species con-
cept except that the phenetic species does require reproductive cohesion
while the diagnostic species does not. The other three have more empiri-
cal content but not so much that any organisms are excluded.
Connected to universality is the distinction between monism and plu-
ralism. Monism in the strong sense is the view that a particular concept is
the correct concept. The supporters of all the concepts that have been dis-
cussed in this chapter are monists in this sense. They all insist that their
Summary and conclusion 375
concept is the correct concept. Monism and pluralism also come in weak-
er forms. A definition that includes alternative criteria or alternative ways
of meeting a single criterion are pluralist in a weak sense. The biological,
evolutionary, cohesion and monophyletic species concepts are pluralist in
the sense that more than one force or mechanism can bring about the state
specified in their respective definitions. For example, Templeton lists sev-
eral forces that can increase cohesion. The monophyletic species concept
is monistic with respect to its grouping criterion (monophyly) but plural-
istic when it comes to its ranking criteria (not all minimally monophyletic
taxa need to be ranked as species). The most monistic species concepts in
this sense are the phenetic, diagnostic and mate recognition concepts.
Pluralism, in the sense of alternative forces and mechanisms, makes a
concept look more universal than it actually is. For example, one might
argue that all species are cohesive, and the only mechanism for cohesion
is gene exchange. If this concept were actually applicable to all organisms,
it would be universal in a very strong sense. However, one might argue
that all species are cohesive and then go on to specify alternative mecha-
nisms that promote cohesion. Instead of all species being X, all species are
X or Y or Z. This concept is universal in the sense that it applies to all
organisms but only by listing a series of alternative criteria. This is univer-
sality on the cheap.
Universality is generally considered a positive feature of a concept by
both philosophers and scientists. As I mentioned earlier, some difference
of opinion exists among philosophers about the virtues of monism and
pluralism. However, among those systematists whose species concepts we
have examined, monism is clearly valued quite highly. They push their
preferred species concept for all its worth. Pluralism begins to look attrac-
tive only when it appears as if one's own preferred concept is losing in the
battle over priority.
A special feature of species concepts is their dimensionality with
respect to time. The evolutionary species concept is explicitly temporal.
Species are lineages extended in time (space-time worms), while the phe-
netic species concept is as non-dimensional as you can get. Advocates of
the other species concepts do not always explicitly state that their species
concepts are designed to individuate time-slices of evolving lineages, but
enough of them do to conclude that this is their intention. The pheneti-
cists would certainly conclude that inclusion of temporal considerations in
a species concept is a mistake and detracts from the value of this concept.
However, advocates of other species concepts would disagree.
Dimensionality is a virtue.
Applicability is a universally recognized virtue among philosophers
and scientists alike, although scientists tend to have higher standards of
applicability than do philosophers. Instead of the species concept dis-
cussed in this chapter falling into two or three groups, they are arrayed
linearly with respect to how applicable (or operational) they happen to be.
The phenetic species concept is clearly the most operational, since this
after all was the main motivation for its formulation. The diagnostic
species concept is the next most applicable concept. A systematist must
know who tends to mate with whom, and what the results of these unions
turn out to be, but little else save character covariation. The monophylet-
ic species concept is as operational as the methods of cladistic analysis
allow. The goal of cladistic analysis is the individuation of characters so
that they nest perfectly. The literature on the difficulties involved in such
an enterprise is huge. The mate recognition, biological, and cohesion con-
cepts are even more difficult to apply because the forces and mechanisms
that they specify are more difficult to discern. Finally, the evolutionary
species concept is the most difficult to apply because it explicitly specifies
that species are extended through time. At any one time it is impossible to
say for sure that a particular group of organisms is a species, because
species status can be determined only in retrospect, a characteristic of all
historical entities.
With respect to theoretical significance, only the phenetic species con-
cept is designed to be theory-neutral or theory-free. The diagnostic species
concept assumes only some very low-level, unproblematic theories, while
all the others are openly theoretical in their content. Whether or not one
chooses to accept one or more of these theories turns on some very
abstruse empirical issues, such as how effective gene flow is in producing
cohesion. Once again, a difference of opinion exists among scientists, not
to mention philosophers, on the virtues of theory-dependence. Most
philosophers nowadays accept some version of theory-dependence.
Theory neutral concepts are impossible, but even if they were possible,
they would be undesirable. The pheneticists and pattern cladists have
made ambiguous claims about the role of process theories in systematics.
On the surface at least, they seem strongly opposed to anything that
might be termed a theory to enter into the classificatory process, at least in
the early stages. Advocates of the other species concepts discussed in this
paper are more accepting of theories. They think that theories, especially
process theories, are absolutely central to the science of systematics from
beginning to end.
If the readers of this chapter so choose, they can make up a chart com-
paring the seven species concepts discussed on the criteria listed. One
problem is that far from total agreement exists on which characteristics of
species definitions are considered virtues and which vices. Scientists at
least value universal, monistic, applicable concepts. The issues of tempo-
ral dimensionality, operationism, and theoretical significance are more
controversial. However, I have made up charts on each of the possible
permutations, and the depressing result is that all seven species concepts
score about the same. This outcome may well explain why the species con-
troversy continues unabated. Even if everyone agreed on what counts as
References 377
a good species concept, no one concept is clearly superior to all the others
- and not everyone agrees on what general criteria characterize a good
species concept.
Acknowledgements
I would like to thank Michael Donoghue, Marc Ereshef sky, Chris Horvath,
Rick Mayden, Brent Mishler and Ed Wiley for reading and commenting
on early drafts of this paper.
18.6 REFERENCES
Budd, A.F. and Mishler, B.D. (1990) Species and evolution in clonal organisms -
summary and discussion. Systematic Botany, 15,166-71.
Claridge, M.F., Dawah, H.A. and Wilson, M.R. (1997) Practical approaches to
species concepts, in Species: the Units of Biodiversity (eds M.F. Claridge, H.A.
ation, in Speciation and Its Consequences (eds D. Otte and J.A. Endler), Sinauer
Cracraft, J. (1997) Species concepts in systematics and conservation biology: an
de Queiroz, K. (1990) Phylogenetic systematics and species revisited. Cladistics, 6,
83-90.
de Queiroz, K. and Donoghue, M.J. (1988) Phylogenetic systematics and the
species problem. Cladistics, 4,317-38.
de Queiroz, K. and Good, D.A. (1997) Phenetic clustering in biology: a critique,
Quarterly Review of Biology (in press).
Donoghue, M.J. (1985) A critique of the BSC and recommendations for a phyloge-
netic alternative. The Bryologist, 83,172-81.
Dupre, J. (1993) The Disorder of Things: Metaphysical Foundations of the Disunity of
Science, Harvard University Press, Cambridge, Massachusetts.
Echelle, A.A. (1990) In defense of the phylogenetic species concept and the onto-
logical status of hybridogenetic taxa. Herpetologica, 46,109-17.
Ehrlich, P.R. and Ehrlich, A.H. (1967) The phenetic relationships of the butterflies.
Eisner, T., Lubchenco, J., Wilson, E.O., Wilcove, D.S. and Bean, M.J. (1995).
Building a scientifically sound policy for protecting endangered species.
Science, 268,1231-2.
Eldredge, N. and Cracraft, J. (1980) Phylogenetic Patterns and the Evolutionary Process,
Endler, J.A. (1989) Conceptual and other problems in speciation, in Speciation and
Its Consequences (eds D. Otte and A. Endler), Sinauer Associates, Sunderland,
Ereshefsky, M. (1989) Where's the species? Comments on the phylogenetic species
concept. Biology and Philosophy, 4, 89-96.
Ereshefsky, M. (ed.) (1992) The Units of Evolution: Essays on the Nature of Species, MIT
Press, New York.
Frost, D.R. and Kluge, A.G. (1994) A consideration of epistemology in systematic
biology, with special reference to species. Cladistics, 10, 259-94.
Ghiselin, M.T. (1969) The principles and concepts of systematic biology, in
Systematic Biology (ed. C. G. Sibley), National Academy of Sciences,
23, 536-44.
Gornall, R.J. (1997) Practical aspects of the species concept in plants, in Species: the
Graybeal, A. (1995) Naming species. Systematic Biology, 44,237-50.
Hawkes, J. (1997) Cultivated plant diversity and taxonomy, in Species; the Units of
Biodiversity (eds M.F. Claridge, H.A. Dawah and M.R. Wilson), Chapman &
Hall, London, pp. 191-8.
Hennig, W. (1966) Phylogenetic Systematics, University of Illinois Press, Chicago.
Horvath, C. Phylogenetic species concepts: pluralism, monism, and history.
Biology and Philosophy (in press).
Horvath, C. Some questions about identifying individuals: failed intuitions about
organisms and species. Philosophy of Science (in press).
Hull, D.L. (1965) The effect of essentialism on taxonomy. British Journal for the
Philosophy of Science, 15, 314-326; 16,1-18.
Hull, D.L. (1968) The operational imperative - sense and nonsense in opera-
tionism. Systematic Zoology, 17,438-57.
Hull, D. (1976) Are species really individuals? Systematic Zoology, 25,174-91.
Hull, D. (1979) The limits of cladism. Systematic Zoology, 28, 416-40.
John, D.M. and Maggs, C.A. (1997) Species problems in Eukaryotic algae: a mod-
ern perspective, in Species: the Units of Biodiversity (eds M.F. Claridge, H.A.
Johnson, L.A.S. (1968) Rainbow's end: the quest for optimal taxonomy, in
Proceedings of the Linnean Society of New South Wales, 93,1—45; reprinted 1970,
Kitcher, P. (1984) Against the monism of the moment: a reply to Elliott Sober.
Philosophy of Science, 51, 616-30.
Kitcher, P. (1989) Some puzzles about species, in What Philosophy of Biology Is:
Essays Dedicated to David Hull (ed. M. Ruse), Kluwer, Dordrecht, pp. 183-208.
Kornet, D. (1993) Reconstructing Species: Demarcations in Genealogical Networks,
Hortus Botanicus, Leiden.
Mayden, R.L. (1997) A hierarchy of species concepts: the denouement in the saga
of the species problem, in Species: the Units of Biodiversity (eds M.F. Claridge,
McDade, L. (1990) Hybrids and phylogenetic systematics I. Patterns of character
expression in hybrids and their implications for cladistic analysis. Evolution, 44,
1685-700.
References 379
McDade, L. (1992) Hybrids and phylogenetic systematics II. The impact of hybrids
on cladistic analysis, Evolution, 46,1329-346.
McKitrick, M.C. and Zink, R.M. (1988) Species concepts in ornithology. Condor, 90,
1-14.
Mishler, B.D and Brandon, R.N. (1987) Individualism, pluralism, and the phyloge-
Mishler, B. and Donoghue, M.J. (1982) Species concepts: a case for Pluralism.
Mishler, B. and Theriot, E. (1997) Monophyly, apomorphy, and phylogenetic
species concepts, in Species Concepts and Phylogenetic Theory: A Debate (eds Q.D.
Wheeler and R. Meier), Columbia University Press, New York, (in press).
Nelson, G. and Platnick, N. (1981) Systematics and Biogeography: Cladistics and
Nixon, K.C. and Wheeler, Q.D. (1990) An amplification of the phylogenetic species
concept. Cladistics, 6,211-23.
O'Hara, R.J. (1993) Systematic generalizations, historical fate, and the species prob-
lem. Systematic Biology, 42, 231-46.
Otte, D. and Endler, J.A. (eds) (1989) Speciation and Its Consequences, Sinauer
Associates, Sunderland, Massachusetts.
Paterson, H.E.H. (1981) The continuing search for the unknown and unknowable:
a critique of contemporary ideas on speciation. South African Journal of Science,
77,113-19.
Picketing, A. (1984) Constructing Quarks, University of Chicago Press, Chicago.
Platnick, N. (1977) Review of Concepts of Species (1976) (ed. C.N. Slobodchikoff).
Platnick, N. (1979) Philosophy and the transformation of cladistics. Systematic
Zoology, 28, 537-46.
Ritland, D.B. and Brower, L.P. (1991) The viceroy butterfly is not a batesian mimic.
Science, 350, 497-8.
Rosen, D. (1978) Vicariant patterns and historical explanation in biogeography.
Rosen, D. (1979) Fishes from the uplands and intermontane basins of Guatemala.
Bulletin of the American Museum of Natural History, 162,267-376.
Signer, P.W. Ill (1985) Real and apparent trends in species richness through time,
in Phanerozoic Diversity Patterns: Profiles in Macroevolution (ed. J. Valentine),
Princeton University Press, Princeton, pp. 129-50.
New York.
Slobodchikoff, C.N. (ed.) (1976) Concepts of Species, Hutchinson & Ross,
Stroudsburg, Penn., Dowden.
Sneath, P.H.A. (1995) Thirty Years of Numerical Taxonomy. Systematic Biology, 44,
281-98.
Sober, E. (1984) Discussion: Sets, Species, and Evolution: comments on Philip
Kitcher's Species. Philosophy of Science, 51, 334-41.
Sokal, R.R., and Sneath, P.H.A. (1963) The Principles of Numerical Taxonomy, W.H.
Freeman, San Francisco.
Sterelny, K. (1994) The nature of species; review of Ereshefsky (1992). Philosophical
Books, 35, 9-20.
Templeton, A.R. (1989) The meaning of species and speciation, in Speciation and Its
Van Regenmortel, M.H. (1997) Viral species, in Species: the Units of Biodiversity (eds
M.F. Claridge, H.A. Dawah and M.R. Wilson), Chapman & Hall, London, pp.
17-24.
Veron, J.E.N. (1995) Corals in Space and Time, Comstock, Cornell University Press.
Vrana, P. and Wheeler, W. (1992) Individual organisms, as terminal entities: laying
the species problem to rest. Cladistics, 8, 67-72.
lem: a reply to de Queiroz and Donoghue. Cladistics, 6, 7-81.
Wiley, E.O. (1981) Phylogenetics: The Theory and Practice of Phylogenetic Systematics,
John Wiley and Sons, New York.
19
A hierarchy of species concepts:
the denouement in the saga of the
species problem
R. L May den
Contacting address: Department of Biological Sciences, P.O. Box 0344, University of
Alabama, Tuscaloosa, AL 35487, USA
ABSTRACT
At least 22 concepts of species are in use today and many of these are
notably incompatible in their accounts of biological diversity. Much
of the traditional turmoil embodied in the species problem ultimate-
ly derives from the packaging of inappropriate criteria for species
into a single concept. This results from a traditional conflation of
function of concepts with their applications, definitions with con-
cepts, taxonomic categories with groups, and the ontological status of
real species with teleological approaches to recover them. Analogous
to classifications of supraspecific taxa, our forging inappropriate and
ambiguous information relating to theoretical and operational dis-
cussions of species ultimately results in a trade-off between conve-
nience, accuracy, precision, and the successful recovery of natural
biological diversity. Hence, none of these expectations or intentions
of species or classifications is attainable through composite, and pos-
sibly discordant, concepts of biological diversity or its descent.
Reviewing and evaluating the concepts of species for their theo-
retical and operational qualities illustrates that a monistic, primary
concept of species, applicable to the various entities believed to be
species, is essential. This evaluation reveals only one theoretical con-
cept as appropriate for species, the Evolutionary Species Concept.
This conceptualization functions as a primary concept and is essential
in structuring our ideas and perceptions of real species in the natural
world. The remaining concepts are secondary, forming a hierarchy of
definitional guidelines subordinate to the primary concept, and are
essential to the study of species in practice. Secondary concepts
382 A hierarchy of species concepts
should be used as operational tools, where appropriate, across the
variance in natural diversity to discover entities in accord with the
primary concept. Without this theoretical and empirical structuring
of concepts of species our mission to achieve reconciliation and
understanding of pattern and process of the natural world will fail.
19.1 INTRODUCTION
'I believe that the analysis of the species problem would be consid-
erably advanced, if we could penetrate through such empirical terms
as phenotypic, morphological, genetic, phylogenetic, or biological, to
the underlying philosophical concepts. A deep, and perhaps widen-
ing gulf has existed in recent decades between philosophy and
empirical biology. It seems that the species problem is a topic where
productive collaboration between the two fields is possible'.
(Mayr, 1957)
Little has changed with regard to the species problem since Mayr com-
posed this piece. Some researchers argue for a particular concept of diver-
sity known as species, while others prefer a pluralistic approach (Mishler
and Donoghue, 1982). Today, the controversy continues over the concep-
tualization of species. This volume reflects some of this diversity of
thought across multiple taxonomic groups. This seemingly timeless
debate has generated a heterogeneous proliferation of concepts, most
hoping to capture the operational and/or theoretical qualities of a good
concept. The search has been for a concept-definition that is biologically
relevant and meaningful, one that is easily applied, and one that encom-
passes natural biodiversity. That is, a concept of real species assisting in
and ensuring their recognition and our understanding of them in nature.
This goal has not been achieved for several reasons.
The 20th century history of biological classification illustrates why this
so-called silver bullet species concept, one that will attend to all our per-
ceived needs, has not yet been achieved. In phylogenetic systematics (or
cladistics) the Linnaean classification scheme represents a hierarchical sys-
tem of categories coordinate with a phylogenetic tree of named taxa.
Represented in the classification is the idea of monophyly of taxa, or sister
group (genealogical) relationship. Classifications are information retrieval
systems about genealogical relationships. In evolutionary systematics the
classification is purported to represent sister group relationship and evo-
lutionary distinctiveness. Paradoxically, while this may be viewed as an
expedient method to group information in a retrieval system, under this
method one can never be sure which criteria are optimized at any part of
a classification. Thus, confusion is inherent in an ambiguous information
retrieval system. The ultimate trade-off of combining too many desired
Methodology 383
functions into a convenient method is that it is not always possible to iso-
late any one function (e.g. genealogy versus distinctiveness).
Much of the turmoil embodied in the species problem ultimately
derives from our packaging inappropriate criteria for species into a single
concept. This results from a traditional conflation of function of concepts
with their applications, definitions with concepts, taxonomic categories
with groups, and ontological status of real species with teleological
approaches to recover them. Analogous to classifications of supraspecific
taxa, our forging inappropriate and ambiguous information relating to
theoretical and operational discussions of species ultimately results in a
trade-off between convenience, accuracy, precision, and the successful
recovery of natural biological diversity. None of these expectations or
intentions of species or classifications is attainable through composite, and
possibly discordant, concepts of biological diversity or its descent.
With this in mind can one tease apart the theoretical concepts and
operational definitions of species and develop a primary concept applica-
ble to the various entities believed to be species? I think this is possible
through a hierarchical view of species concepts and their definitions.
Below, I review the various species concepts and propose a hierarchical
classification for them. Each of these concepts is briefly evaluated relative
to their consequential qualities thought to be important in a concept (Hull,
1997: Chapter 18). This evaluation reveals only one appropriate primary
and theoretical concept of species. The remaining definitions are sec-
ondary concepts, forming a hierarchy of definitional guidelines subordi-
nate to this primary concept. The secondary concepts are engaged only as
operational tools, where appropriate, across the variance in natural diver-
sity to discover entities in accord with the primary concept.
19.2 METHODOLOGY
Probably more is written about species than any other topic in evolution-
ary biology. There are many opinions and studies addressing this ques-
tion. Hence, an exhaustive survey of these is impossible. Concepts are
ideas or intuitions uniquely developed in the minds of every person.
Definitions of these concepts are the only form with which one can com-
pare them. Sometimes, these definitions may be poorly developed or mis-
interpreted, ultimately leading to miscommunication of ideas. Regardless,
I have endeavoured to understand the arguments on the various species
concepts (Table 19.1), and compare and evaluate them. In section 19.7 I
have also made an effort to identify synonyms of concepts; these are list-
ed by assigned standard abbreviations or full titles. Where concepts were
formerly identified as synonymous, credit is provided; in part refers to the
observation that portions of concepts are equivalent.
Table 19.1 Species concepts and standardized abbreviations
1. Agamospecies (ASC) 14. Morphological (MSC)

2. Biological (BSC) 15. Non-dimensional (NDSC)
3. Cohesion (CSC) 16. Phenetic (PhSC)
4. Cladistic (CISC) 17. Phylogenetic (PSC)
5. Composite (CpSC) 1. Diagnosable Version (PSCj)
6. Ecological (EcSC) 2. Monophyly Version (PSC2)
7. Evolutionary Significant Unit (ESU) 3. Diagnosable and Monophyly
8. Evolutionary (ESC) Version (PSC3)
9. Genealogical Concordance (GCC) 18. Polythene (PtSC)
10. Genetic (GSC) 19. Recognition (RSC)
11. Genotypic Cluster Definition (GCD) 20. Reproductive Competition (RCC)
12. Hennigian (HSC) 21. Successional (SSC)
13. Internodal (ISC) 22. Taxonomic (TSC)
19.3 IMPORTANT QUALITIES TO CONSIDER

Few concepts can be viewed as more fundamental to the natural sciences
than that of the species. Species as individuals (Hull, 1976) represent a
unique level of organization of the natural world; they are self-organizing
entities or particulars. This level of universality is the upper-most limit to
involve tokogenetic relationships and the lower-most level participating
in phylogenetic relationships. They are purported to be the highest level
of integration to participate in natural processes while being spatiotempo-
rally constrained. Thus, they are essentially fundamental units of evolu-
tion. In plain English, this means that for all disciplines using any species
in pure or applied research, education, management, conservation, etc.,
success at accomplishing identified missions or deriving informative
answers to particular questions is inextricably tied to a basic assumption
that the species involved are real by-products of natural processes and not
misguided fabrications of our own invention.
One method to compare and contrast the various concepts of species is
through three criteria that have traditionally been employed to evaluate sci-
entific concepts. Hull (1997: Chapter 18) provides such a review for seven of
the most frequently used concepts for their theoretical significance, gener-
ality (or universality), and applicability (or operationality). Following a
review of definitions, metaphysical topics, and species concepts, I extend a
similar evaluation to all 22 of the various concepts of species. The relation-
ships of the concepts within and among the three criteria, together with the
question of monism versus pluralism, reveals a hierarchy of species con-
cepts that should finally put the species problem to rest.
19.4 CONCEPTS, DEFINITIONS, GROUPS, CATEGORIES AND

NAMES: THE UNFORTUNATE CONFLATION OF TERMINOLOGY
Important in the discussion and resolution of the species problem is the
correct usage and understanding of ideas and terms related to the species
Concepts, definitions, groups, categories and names 385
issue. Historically, discussions of species have involved the use of four
critical terms: concept, definition, group and category. These terms have
been central to the fundamental melange that both scientists and philoso-
phers have encountered with the species problem.
Communication of ideas or concepts in science is of utmost importance
and hinges upon statements or definitions developed by persons formu-
lating or discussing the concepts. Critically important to exact communi-
cation of ideas embodied in concepts demands that we do not obfuscate
the terms or words used in definitions or statements of the concepts. Here,
we confront difficulties both in the logical treatment and the evaluation of
definitions.
Concepts of biological systems serve as fundamental links between pat-
tern and process in nature, are employed in every discipline, and help
guide our perception of natural systems. They are formulated by individ-
ual persons through observation, study and synthesis (impressions and
imagination) of both theory and empirical data. Concepts may be real or
abstract. Real concepts are those representing easily agreed discrete
objects. Abstract concepts are those representing hypothetical and tran-
sient phenomena. A concept may be relayed from one person to one or
more other persons by adapting it into a statement or definition, either
verbally, in writing, or graphically. Such a definition may or may not
induce the same concept in the mind of the other persons, depending
upon the appropriateness, precision, and accuracy of the words used in the
definition and the level of understanding of the other person. With real
concepts (e.g. round versus square) one may compare statements devel-
oped by different observers with discrete objects to see if they agree. With
abstract concepts (evolution, natural selection, species as taxa) it is difficult
to know for sure if statements represent the same transient or hypothetical
things, and the respective definitions can only be compared using previ-
ously agreed definitions of words used in the statement. One may also
observe the effectiveness of such a concept through direct examination.
The term 'group' refers to a collection of objects or things. In the inter-
section of natural sciences, taxonomy, and systematics the term taxon is
often used synonymously with the term group. A group can be real and
have objective reality if it corresponds to qualities that are real and exclu-
sive to it, and if it consists of things that have material existence. They may
be arranged hierarchically, either as non-reticulate or reticulate groups.
They may be represented at various levels of universality from groups of
things to more inclusive groups of things, etc. They may be of any size and
arise on the basis of intrinsic attributes and/or extrinsic decisions.
Organisms can be members of any number of groups so long as they pos-
sess the attributes of the said groups. Groups, however, are not like con-
cepts. Groups develop from sense impressions of concepts and can be
agreed upon and definite if the statements about them are unambiguous
and decisive.
Groups and categories are distinctly different and there is no real con-
nection between them. The tradition in codes of nomenclature artificially
forces the use of taxonomic categories in a hierarchy for groups. A biolog-
ical classification is a contrived system of categories used for the storage
and retrieval of information about biological diversity, taxa, or groups.
The concept 'category' is a class and has no separate existence from its use
in organizing objects or thoughts; categories have no reality. Unlike
groups, categories have no attributes; things or objects are not members of
categories, but are parts of groups; and organisms are not members of any
taxonomic category. For example, Cyprinidae is a proper name given to a
group of fishes possessing certain attributes. By taxonomic convention the
-idae ending denotes a traditional level of universality in the zoological
hierarchy. The group Cyprinidae can be a part of many other groups
(Cypriniformes, Ostariophysi, Teleostei), but is only a member of one tax-
onomic category, Family. Because categories have no reality the Family
Cyprinidae is not a member of any other more-inclusive categories, but
Cyprinidae is.
Multiple classifications may exist for the same group of organisms,
depending upon criteria being optimized in the classification. The
Linnaean hierarchy imparts information regarding relationship, descent
from hypothesized immediate common ancestors. Supraspecific cate-
gories are more inclusive than the species category. Groups are assigned
to categories; assignment is based on the definition of the category.
Historically, categories of this hierarchy were defined on the basis of dis-
tinctiveness; that is, distinctiveness of the group assigned to Genus was
less than distinctiveness at the Family, and so on.
In the modern-day hierarchical system a distinction between the
supraspecific categories and the species category is a dichotomy between
phylogenetic and tokogenetic processes and relationships (Hennig, 1966;
Frost and Kluge, 1994; Wiley and Mayden, 1997). Supraspecific categories
are defined only as monophyletic groups having phylogenetic relation-
ships and historical cohesion. The categorical rank assigned to a group sat-
isfying this criterion is only a by-product of its level of inclusiveness.
Assignment of a group to a supraspecific category is definite because of its
historical existence, and our discovery and recognition of this previous
existence requires demonstration of monophyly. Demonstration that a
group shared an immediate common ancestor is by way of synapomor-
phies, one or more features inherited from and evolved in the immediate
common ancestor to all known descendants. Only the concepts of mono-
phyly and historical cohesion apply to all groups assigned to supra-
specific categories.
The species category shares some, but not other qualities with supraspe-
cific categories. The groups assigned to this category are different. Like
other categories, the category species is an artificial construct used for orga-
The twin meanings of species 387
nization of information. Unlike groups assigned to supra-specific cate-
gories those assigned to the species category have tokogenetic relation-
ships (sexuals) or are tokogenetic vectors (clones), and may or may not be
definite. Because species being classified today are potential future ances-
tors of groups to be placed in supraspecific categories of tomorrow, their
existence does not necessitate demonstration via synapomorphies. These
groups (species) are the types of entities once existing in historical com-
munities (as ancestors) that modern-day systematists endeavour to docu-
ment today by way of synapomorphies. Unlike supraspecific categories,
there is a real connection between the species category and the groups
assigned to this category. While the species category is a class construct like
supraspecific categories, it is ontologically distinct from the class construct
supraspecific categories. The species category is unique and contains only
those groups of things conforming to the concept of this category. Herein
lies the nucleus of the species problem. What is the appropriate concept for
species and the species category?
Names applied to categories should not be confused with the names
applied to the groups included in the categories. For example, there is a
categorical level of the hierarchy to which is assigned the arbitrary name
genus. A group (or group of groups) may be referred to this level of the
hierarchy and referred to as a genus (see discussion below on twin mean-
ings of species). The group of organisms is given a proper name distinct
from the category name. For example, there is a group of fishes named
Cyprinella and this group is assigned to a level of the hierarchy, namely
genus. Problems arise when one defines the named category (genus) and
confuses this definition with or extends this definition to the group of par-
ticulars (Cyprinella) being placed into the category.
19.5 THE TWIN MEANINGS OF SPECIES

The term species has two different meanings that, when not clearly dif-
ferentiated, will result in confusion and misunderstanding relevant to
real species, species concepts, species category and speciation. The term
species is used to represent both a taxonomic category and those natu-
rally occurring particulars that we discover, describe, and order into our
classification system. Confusion of these terms is most detrimental to
elucidating and understanding the importance of species concepts
because they each have a very different ontological status (Hull, 1976).
While this may seem obtuse, irrelevant, or strictly metaphysical to the
working biologist, naturalist, or general scientist, it is not. Much of the
confusion over species concepts relates directly to the conflation of these
critically different meanings.
The two species terms are aligned in two different philosophical cate-
gories, Class and Individual (not to be confused with the formal taxonomic
category or single organism, respectively). The taxonomic category species
is a class. A category is spatiotemporally unbounded, lacks cohesion, is not
self replicating, does not participate in any natural processes, has members,
and can be defined. Members of a class may be classes, or not. A class can
exist anywhere in the universe, so long as there is a definition for member-
ship. Members of a class also can exist anywhere, if the definition applies
(Hull, 1976).
Those tokogenetic and cohesive entities discovered and described in
nature, referred to species, that we place in the category species, are indi-
viduals (or things, particulars). Individuals are spatiotemporally bounded,
have intrinsic cohesion, are self-replicating, participate in natural process-
es, have part-whole relationships, but cannot be defined. Individuals
change over time and can only be described. Individuals exist throughout
the universe. Because individuals have no definitions, they do not have
members; rather, they exist as parts of wholes. Parts of individuals may be
other individuals (Homo sapiens, organisms, organs, tissues, cells, mito-
chondria, etc.), also resulting from various natural processes.
Thus, the twin meanings of species refer to two, radically different and
basic metaphysical categories, classes and individuals, that when con-
fused generates elementary problems for understanding. As a class, the
category species is temporally unbounded, has a definition, and only
those things fitting this definition can be included. Species as taxa change
with time, have no definitions, they can only be described, identified,
pointed to, etc. In our discussions of 'What is a species?', we reference
Linnaean category species, the class concept, and what we decide should
and should not be included in this category through a definition. In dis-
cussions of species as they exist in nature we reference the individual with
a unique origin and no definition. These are particulars. If an organism is
found on Mars that looks, acts, and speaks like Homo sapiens on Earth, it is
not H. sapiens unless it descended from H. sapiens on Earth.
19.6 SPECIES CONCEPTS VERSUS EMPIRICAL DATA

It is not uncommon to find in discussions of species and species concepts
researchers confusing empirical data used in the operation of recognizing
a species with a conceptualization or definition of species. Empirical data
can include such things as anatomy, morphology, genetics (DNA, pro-
teins), behaviour, etc., all possibly evaluated and analysed in a variety of
ways and with a variety of methods. Our abilities to gather these data are
artificially constrained by technological advances; that is, we can only col-
lect data that current technology permits. These artificial constraints on
our ability to perceive variation in nature should not be confused with our
desires, objectives, or attempts to illuminate natural variation. For exam-
ple, it is often said that a particular group of organisms represents a dis-
The species concepts 389
tinct species based on a morphological or genetic species concept. If
empirical methods are confused with concepts of species the logical out-
come will be confusion. That is, species that fit the definition of a mor-
phological species concept may not fit the definition of a genetic species
concept, and vice versa. Assuming that we are mainly interested in identi-
fying natural diversity resulting from historical processes encoded in
genomes, empirically driven concepts are untenable. Given that any
research study is grossly limited in the type and amount of empirical data
available from all that is technologically possible, universally applicable
concepts of species should not be bound by or confused with empirical
evidence.
19.7 THE SPECIES CONCEPTS

The taxonomic, systematic, and evolutionary literature reveals that at least
22 concepts have been developed to characterize diversity (Table 19.1).
Developed by researchers to suit individual needs, some are operational-
ly or empirically motivated, some are galvanized by theoretical necessity,
while some are motivated by peculiarities of organisms studied. Not all
concepts have been equally well characterized or explicitly defined. Some
have been essentially bequeathed to academic descendants of particular
fields (e.g. population genetics, taxonomy, entomology, mammalogy,
etc.), together with an awareness of the requisite qualities of species and
necessary operations to be employed by researchers embracing them.
Only abbreviated discussions of the various concepts are presented below
alphabetically. Included are synonyms, definitions, discussion, and a syn-
opsis as to the suitability of the concept.
19.7.1 Agamospecies concept (ASC)
Synonyms
Microspecies, Paraspecies, Pseudospecies, Semispecies.
Discussion
This concept refers specifically to taxa that do not fit the biparental, sexu-
ally reproducing mode. It serves as a general umbrella concept for all taxa
that are uniparental and reproduce via asexual reproduction; often these
species are the result of interspecific or intergeneric hybridization. These
species may produce gametes but there is often no fertilization, except via
hybridization. Ghiselin (1984a: 213) refers to these species as 'heaps of
leaves that have fallen off the tree that gave rise to them'. Agamospecies
may be part of a species complex wherein there also exist bisexually repro-
ducing species. In these cases the agamospecies maybe facultative or oblig-
ate apomicts. Obligate apomicts are sometimes referred to as microspecies.
In reality the composite of individual organisms of the species may often
be polyphyletic, resulting from multiple crosses between parental, bisexu-
al species. These taxa are most often diagnosed by features related to either
morphology or chromosomes. Often, these species have very restricted
ranges. Some authors only recognize them as species if their range includes
at least 20 km diameter (Weber, 1981).
Synopsis
Because of the limited application of the ASC to asexually reproducing
species the ASC should serve as a primary concept.
19.7.2 Biological Species Concept (BSC)

'A biological species is an inclusive Mendelian population; it is inte-
grated by the bonds of sexual reproduction and parentage'.
(Dobzhansky, 1970: 354)
'...groups of actually or potentially interbreeding natural popula-
tions which are reproductively isolated from other such groups'.
(Mayr, 1940)
'A species is a group of interbreeding natural populations that is
reproductively isolated from other such groups'. (Mayr and Ashlock,
1991)
Synonyms
GSC, Isolation Species Concept (Paterson, 1993), Second Species Concept
(Mayr, 1957), Speciationist Species Concept (Blackwelder, 1967),
Discussion
This concept has been reviewed by its strongest proponent, Mayr, in sev-
eral publications and by several other authors (see Mayden and Wood,
1995). As recently espoused by Mayr and Ashlock (1991: 26-27) and Mayr
(1997), species consist of reproductive communities wherein there is both
an ecological and genetic unit. Individuals of a species seek and recognize
one another for mating and thereby maintain an intercommunicating gene
pool that, 'regardless of the individuals that constitute it, interacts as a unit
with other species with which it shares its environment'. For Mayr (1997)
'each biological species is an assemblage of well balanced, harmonious
genotypes and... indiscriminate interbreeding of individuals, no matter
how different genetically, would lead to an immediate breakdown of these
harmonious genotypes. As a result, there was a high selective premium for
the acquisition of mechanisms, now called isolating mechanisms, that
would favour breeding with conspecific individuals and inhibit mating
with non-conspecifics. This consideration provides the true meaning of
species. The species is a device for the protection of harmonious, well inte-
grated genotypes. It is this insight on which the biological species concept
is based'. Central to this concept, and the sole criterion for the reality of a
species, is thus the idea of reproductive isolation of species from other such
species. 'A species is a protected gene pool' that is 'shielded by its own
devices (isolating mechanisms) against unsettling gene flow from other
gene pools' (Mayr and Ashlock, 1991). The word interbreeding in the defi-
nition above 'indicates a propensity; a spatially or chronologically isolated
population, of course, is not interbreeding with other populations but may
have the propensity to do so when the extrinsic isolation is terminated'
(Mayr, 1997). Accordingly, speciation is the process of achieving reproduc-
tive isolation (Mayr, 1963: 502; 1970: 288).
The BSC specifically excludes uniparental species even though they are
known to exist, and some have relegated diversity of this type to pseu-
dospecies (Dobzhansky, 1970). The concept also is viewed as being an
operational definition in that 'taxa of the species category can be delimit-
ed against each other by operationally defined criteria, for example, inter-
breeding versus non-interbreeding of populations' (Mayr and Ashlock,
1991: 27). This concept is relational because 'A is a species in relation to B
and C because it is reproductively isolated from them'. Finally, it is a non-
dimensional concept that 'has its primary significance with respect to
sympatric and synchronic populations..., and these are precisely the situ-
ations where the application of the concept poses the fewest difficulties.
The more distant two populations are in space and time, the more difficult
it becomes to test their species status in relation to each other but the more
biologically irrelevant this status becomes'.
At least ten elements of this concept are viewed by Mayden and Wood
(1995) as counter-productive toward discovering and understanding bio-
diversity. The BSC has received substantial criticism in recent years for
issues dealing with: (1) the absence of a lineage perspective; (2) its non-
dimensionality; (3) erroneous operational qualities as a definition; (4) its
exclusion of non-sexually reproducing organisms; (5) indiscriminate use
of a reproductive isolation criterion; (6) confusion of isolating mechanisms
with isolating effects; (7) implicit reliance upon group selection; (8) its rela-
tional nature; (9) its teleological overtones; and (10) its employment as a
typological concept, no different from the frequently criticized morpho-
logical species concept.
Synopsis
The nature of the unfavourable attributes inherent in the BSC preclude it
from being considered a primary species concept.
19.7.3 Cladistic Species Concept (CISC)

'...that set of organisms between two speciation events, or between
one speciation event and one extinction event, or that are descend-
ed from a speciation event'. (Ridley, 1989)
Synonyms
ISC (in part; Kornet, 1993), CSC (in part; Kornet, 1993).
Discussion
Ridley (1989) proposed this minimalistic lineage concept of species where-
in species are treated as individuals, not classes. As subtheories, discussion
of this concept incorporates the BSC and EcSC (within cladistic frame-
work) to provide a more complete theory for understanding species.
Ridley is one of the few authors discussing species that makes a clear dis-
tinction between theoretical and practical concepts. A species is a lineage
and speciation produces two or more lineages via splitting. By definition,
species cannot be paraphyletic, even if individual organisms of one or
more of the descendant species are genealogically more closely related to
individuals of one or more other descendant species. Rather, ancestral
species necessarily become extinct following a speciation event. This con-
cept is free from operational constraints of necessary defining attributes,
typical of concepts treating species as Classes.
Synopsis
In some ways, this concept of species could serve as primary concept for
biological diversity. It is a lineage concept, treats species as individuals,
and places no constraints on necessary attributes that a species must pos-
sess in order to be validated. In this sense it is similar to the CpSC, ESC,
ISC, and some versions of the PSC. However, there are important differ-
ences that preclude all of these concepts, except the ESC, from being con-
sidered a primary theoretical concept. With respect to the CISC, ancestral
species, by definition, become extinct following a speciation event and
hence cannot be considered paraphyletic with respect to the organisms of
ancestral and descendant species. Descendant species, by definition, are
monophyletic; ancestral species, by definition, go extinct following speci-
ation. This concept is criticized by Wilkinson (1990) for lack of specificity
with regard to speciation, an issue related to the enforced monophyly of
species. Kornet and McAllister (1993) compare the CISC and CpSC and
argue that discussions concerning the monophyly of species are inappro-
priate, but that organisms forming species involved in a speciation event
will, in all probability, be paraphyletic relative to one another. Thus, the
CISC is inappropriate as a primary concept.
19.7.4 Cohesion Species Concept (CSC)

'...the most inclusive population of individuals having the potential
for phenotypic cohesion through intrinsic cohesion mechanisms'.
(Templeton, 1989:12)
'...the most inclusive group of organisms having the potential for
genetic and/or demographic exchangeability'. (Templeton, 1989: 25)
Discussion
Templeton (1989) developed this concept following a review and short cri-
tique of the ESC, BSC and RSC. Of these concepts only the ESC did not
exclude known biological diversity. The BSC and RSC were rejected
because of substantial inadequacies for all living organisms. The argument
emphasized was that both concepts underscore a certain level of sexual
reproduction and, as such, fail in recovering naturally occurring diversity.
This was considered highly significant given that a whole spectrum of
known and valid diversity never, or only rarely, employs sexual repro-
duction or many have too high of levels of sexuality to be validated using
either the RSC or BSC.
Borrowing positive aspects of all three concepts reviewed, especially
the ESC, Templeton (1989) provides specific guidelines representing
mechanisms of cohesion (1989: 13, table 2) to be used in understanding
species. Cohesion of a species includes various aspects classified as either
genetic or demographic exchangeability. As demonstrated, there is no
clear break between sexual and asexual reproduction in terms of mecha-
nisms and its ultimate outcome to a population. As such, this concept
accepts all reproductive modes, and species are evaluated and validated
on the basis of cohesion, not isolation.
Synopsis
Templeton (1989: 5) noted that the ESC 'is not a mechanistic definition',
and favoured the CSC because it was developed with operational mecha-
nistic qualities in mind. While this criticism is valid when one seeks an
operational concept of species, because the CSC provides extensive oper-
ational details and guidelines for recognizing species it must be specifical-
ly excluded as a primary concept of species. However, the comprehensive
operational nature of the CSC makes it an important practical surrogate
(secondary) for a primary concept.
19.7.5 Composite Species Concept (CpSC)

'... all organisms belonging to an originator internodon, and all
organisms belonging to any of its descendant internodons, exclud-
ing further originator internodons and their descendant intern-
odons'. (Kornet and McAllister, 1993: 78)
Synonyms
PSC (in part).
Discussion
This concept has its origin in the ISC as formalized by Kornet (1993).
Fundamental problems with the ISC, recognized by Kornet and McAllister
(1993), lead to the formulation of this concept. An internodon is defined
by Kornet and McAllister (1993: 78) as 'a set of organisms such that, if it
contains some organism x, it contains all organisms which have the INT
relation with x, and no other organism'. While any internodon could tech-
nically be referred to as a species under the ISC, these authors view
species as historical conglomerates formed of internodons that are perma-
nently isolated and morphologically divergent. A species is a 'set of organ-
isms belonging to several consecutive internodons in the phylogenetic
succession, identified and grouped together by some procedure' (page:
66). Composite species begin with the evolution of an 'originator intern-
odon' possessing 'a morphological property shown by the internodons'
member organisms' (page: 67). They close with the extinction of the latest
internodon that is a descendant of the originator, that is not an originator
itself, and where another originator between it and its originator has not
evolved. These species may endure permanent splits in the network
wherein no morphological property shown by member organisms may be
detectable or permanent reproductive isolation has not evolved. In other
words, composite species are diagnosable or reproductively isolated enti-
ties that, using the terminology traditionally applicable to phylogenetic
trees, may be either monophyletic or paraphyletic groups of internodons.
When permanent rifts in networks are not accompanied with anagenesis
of morphology or changes in reproductive abilities, then these rifts will
not be permanent or detectable. That species are largely paraphyletic
groups of internodons is viewed as an essential element of this concept
because species must be mutually exclusive entities. If species, as groups
of internodons, were required to be monophyletic then species would not
be mutually exclusive, but nested sets. This concept also sanctions the
recognition of successional species (page: 84-85) and 'superposed' species
(page: 85), two or more species evolving within a single internode of a
composite species and defined by at least one morphological fixation each.
Synopsis
These authors provide a lucid comparison of the appropriateness of the
terms monophyly, paraphyly, and polyphyly which refer directly to
groupings of things (species, internodons, or organisms), but not to
species as Individuals. In a practical sense, this concept is essentially insep-
arable from PSCj. Theoretically the CpSC and PSCs are very different.
Kornet and McAllister (1993) recognize that under the CpSC species are
like higher taxa; that is, historical entities that cannot interbreed, lacking
interspecies cohesion. Treatment of species as classes, the recognition of
successional or superposed species, and the intolerance for unisexuals
precludes the CpSC as a primary concept. Like the PSC, it may be a useful
operational surrogate assisting with discovering some species diversity.
19.7.6 Ecological Species Concept (EcSC)

'... species is a lineage (or a closely related set of lineages) which
occupies an adaptive zone minimally different from that of any
other lineage in its range and which evolves separately from all lin-
eages outside its range' (Van Valen 1976: 233).
Synonyms
ESC (sensu Simpson; Stuessy, 1990; Minelli, 1993).
Discussion
This concept views species as ecological units forming lineages through
time in a competitive environment. It is an operational definition where-
in differences in ecology constitute different, independently evolving
species. It is tolerant of both bisexual and unisexual species, species that
evolve via hybridization, and the species that exchange genes, so long as
ecological distinction is maintained in the lineage. The equivalence of the
Evolutionary Species Concept (ESC) and EcSC (Stuessy, 1990; Minelli,
1993) is inaccurate. These concepts are distinct, in that the ESC does not
necessitate or outline any ecological divergence between sympatric
species. Only in the original ESC of Simpson (1961) was species referred to
in an evolutionary and ecological context.
Synopsis
There is no doubt that the possession of divergent ecologies among sym-
patric lineages warrants their recognition as distinct species. While a toler-
ant lineage concept, as an operational concept it cannot serve as a primary
concept.
19.7.7 Evolutionary Species Concept (ESC)

'... a lineage (an ancestral-descendant sequence of populations)
evolving separately from others and with its own unitary evolution-
ary role and tendencies'. (Simpson 1961: 153)
'...a single lineage of ancestor-descendant populations which main-
tains its identity from other such lineages and which has its own
evolutionary tendencies and historical fate'. (Wiley, 1978)
'... an entity composed of organisms which maintains its identity
from other such entities through time and over space, and which has
its own independent evolutionary fate and historical tendencies'.
(Wiley and Mayden, 1997)
Synonyms
ESU (in part; Mayden and Wood, 1995).
Discussion
This concept was championed originally by Simpson (1951,1961) out of a
general dissatisfaction with the non-dimensionality of the BSC. Wiley
(1978, 1981) developed the concept further and argued for its general
application to biological systems. Unlike other definitions reviewed herein,
the ESC largely was ignored, until recently. Frost and Hillis (1990), Frost
and Kluge (1994), and Wiley and Mayden (1997) reviewed or further
developed the concept. These authors argue that the ESC is the only avail-
able concept with the capacity to accommodate all known types of bio-
logically equivalent diversity. Contrary to the perception of some (Minelli,
1993: 66-9) the ESC does not consider species as Classes or focus on
species as ecological entities. The ESC is not equivalent to the EcSC. While
Simpson (1961) advocated a lineage concept to species and ecological and
evolutionary divergence, he also condoned the delineation of artifactual
successional species. Thus, the logical corollaries of Simpson's ESC and
Wiley's ESC are quite different.
The ESC is not an operational concept. However, it is a lineage concept
that is non-relational. Thus, the attributes and patterns of species can be
correctly interpreted with respect to their unique descent. The ESC accom-
modates uniparentals, species formed by hybridization, and ancestral
species. It does not require knowledge of, nor specific changes in, a
Specific Mate Recognition System (SMRS, see RSC, section 19.7.22). There
is no threshold for particular attributes needed for the existence of a
species. Finally, reproductive isolation, is considered a derived attribute
from the plesiomorphic status of reproductive compatibility; reproductive
success is thus largely uninformative.
Synopsis
The ESC is the most theoretically significant of the species concepts; it
accommodates all 'types' of species known to date and thus has the great-
est applicability. As such, the ESC can serve as a primary concept.
19.7.8 Evolutionary Significant Unit (ESU)

'... a population (or group of populations) that 1) is substantially
reproductively isolated from other conspecific population units, and
2) represents an important component in the evolutionary legacy of
the species'. (Waples, 1991)
Synonyms
BSC (in part), ESC (in part; Mayden and Wood, 1995).
Discussion
The reliance upon criteria such as 'substantially reproductively isolated'
and 'evolutionary legac/ incorporates attributes traditionally viewed as
qualities of species from other concepts. It combines the isolation or mate
recognition system of the non-dimensional BSC and RSC, and invokes the
evolutionary lineage perspective of the CISC, CSC, PSC, and ESC. These
components are nothing more than the 'identities' of cohesive groups of
organisms through time and over space, possessing their own indepen-
dent evolutionary fate and historical tendencies advocated in the ESC.
While the ESU has been proposed as a concept targeted at revealing 'dis-
tinct' populations within species (Waples, 1991, 1996), the distinction
between 'distinct' populations and species as natural, evolutionary enti-
ties is not made clear.
Synopsis
This concept excludes known biodiversity, thereby unduly biasing our
perception of process. Incorrect assumptions about diversity targeted for
protection, brought about by misconceived formulations, only obstructs
efforts to understand and preserve it. While basically a lineage concept,
its emphasis on genetics and isolation preclude its use as primary con-
cept (see Mayden and Wood, 1995).
19.7.9 Genealogical Concordance Concept (GCC)

'... population subdivisions concordantly identified by multiple inde-
pendent genetic traits should constitute the population units worthy
of recognition as phylogenetic taxa'. (Avise and Ball, 1990: 52)
Synonyms
BSC (in part), CISC (in part), PSC (in part).
Discussion
Faced with the impending abandonment of the BSC for the PSC, the GCC
is asserted by Avise and Ball (1990: 46) to be from the 'better elements of the
PSC and BSC'. They proposed that the general principles of the new con-
cept 'derive most easily from the theories and observations in molecular
evolution, but can also be applied to hereditary, morphological, behaviour-
al and other phenotypic attributes traditionally studied by systematists'
(page: 46). They noted three problems with the exclusive use of the PSC.
These include: (1) the number of species depends upon resolving power of
analytical tools available, (2) unless persistent extrinsic (geographic) or
intrinsic RBs [reproductive barriers] are present different gene genealogies
will usually disagree in the boundaries of 'species' under the PSC, and (3)
shared ancestry in sexually reproducing organisms implies historical mem-
bership in a reproductive community.
The arguments generated by Avise and Ball (1990: 45) between the
PSC and the BSC are deceptive; they build a strawman argument of the
PSC and portray it inaccurately. For example, Cracraft (1983) never
required monophyly of species, only that species be diagnosable.
Furthermore, in no discussion by proponents of the PSC has it been
restricted to uniparental species, or the possibility that eventually
individual organisms will qualify as species under the PSC. Monophyly
as part of the PSC was a criterion developed after Cracraft's hypothesis
(de Queiroz and Donoghue, 1988; McKitrick and Zink, 1988) and well
before the GCC. The GCC specifies at least two or more apomorphies of
a species, while the PSC does not. The specification of at least two apo-
morphies is no less arbitrary than is the specification of one, three, or
more. The PSC does not advocate that one can find apomorphies for
almost every individual, any more than one would by employing the
BSC. Thus, there is essentially no difference between the GCC and mono-
phyly formulations of the PSC, criticized by these authors. Other prob-
lems associated with the GCC are those identified with PSCj or PSC2.
While the GCC is defined and titled as a genealogical characterization of
species, the criteria used by Avise and Ball (1990) for species recognition
actually range from monophyly to geographic concordance to genetic
differences without relevance to genealogy.
Synopsis
Avise and Ball (1990) emphasize that a problem with the PSC is resolution
with available tools. There is no question that this is a limitation, but this
limitation extends to all operational definitions of species, including the
GCC. Emphasis on differences relegates the GCC to a concept that
ignores differences between primitive and derived attributes and uses
diagnosability as an operational guideline. Genetic differences can exist
with respect to plesiomorphies that provide no relevant information on
genealogy, making this essentially a typological concept. The general phi-
losophy promulgated in the GCC is largely inseparable from that of the
BSC. Thus, the GCC adopts with it all of the misgivings of the BSC, mak-
ing it inappropriate as a primary concept of species.
19.7.10 Genetic Species Concept (GSC)

'... group of organisms so constituted and so situated in nature that
a heredity character of any one of these organisms may be transmit-
ted to a descendant of any other.' (Simpson, 1943)
'... the largest and most inclusive reproductive community of sexual
and cross-fertilizing individuals which share in a common gene
pool'. (Dobzhansky, 1950)
'... members of a species form a reproductive community. The indi-
viduals of a species of animals recognize each other as potential
mates and seek each other for the purpose of reproduction... The
species, finally, is a genetic unit consisting of a large, intercommuni-
cating gene pool [and these] properties raise the species above the
typological interpretation of a 'class of objects'. (Mayr, 1969: 23)
Synonyms
BSC (Grant, 1981), PhSC, MSC.
Discussion
This concept is similar to the morphological species concept except that the
method used to delineate species is a measure of generic differences, pre-
sumed to reflect reproductive isolation and evolutionary independence. As
a phenetic concept genetic distances and similarities are used to identify dif-
ferent species. Genetic independence is assessed using methods varying
from chromatography, to protein electrophoresis, to sequencing.
While apparently operational, one of the basic problems with the GSC is
that for the vast majority of diversity there exists no genetic information.
Because divergence for any particular gene may not be at a uniform rate,
in all likelihood there will never exist a standard distance for species. This
concept rests on the assumption that for every speciation event there will
be particular changes in all genes. If the researcher examines 200 genes and
they are all identical between two species they would be considered the
same species. However, the next gene may show tremendous revolution
between sister species as a result of the speciation event. One divergent
gene out of 200 monoallelic genes will result in a trivial genetic distance.
On a linear scale, such a divergence will be trivial to a species comparison
where five of 20 genes are divergent. Yet, in this heuristic example both
species pairs are existing as evolutionarily independent and genetically
independent species.
Synopsis
The GSC is essentially a surrogate, operational concept developed out of
the BSC. A particular degree of genetic divergence is assumed to warrant
species recognition. However, this operational definition, lacks any guid-
ance for researchers as to how much difference is enough? This is largely
because divergence for particular genes or across multiple genes is impos-
sible to predict either within or between taxonomic groups. Using this
non-evolutionary concept researchers are also misled to believe that the
lack of divergence in genes that are merely available because of technolo-
gy negates the reality of divergence that may be present for any other
characters. As such, the reality of species with divergent and heritable
morphologies may be naively questioned if divergence at readily accessi-
ble genes or proteins is wanting. While this concept has served as a tradi-
tional method for identifying species it is fatally flawed as a primary con-
cept. The general paucity of data, combined with the enormous genetic
variability observed between sister species, the questionable validity of
relying exclusively upon genetic divergence for species validation, and
the deficiency of a phylogenetic perspective in interpreting variation pre-
cludes the GSC from serving as a primary concept.
19.7.11 Genotypic Cluster Definition (GCD)
'... clusters of monotypic or polytypic biological entities, identified
using morphology or genetics, forming groups of individuals that
have few or no intermediates when in contact'. (Mallet, 1995)
Synonyms
ASC, BSC, GSC, HSC, MSC, NDSC, PhSC, PtSC, PSCj, SSC (in part), TSC.
Discussion
Mallet (1995) argues that a preferred alternative to the BSC is the GCD.
While not stated directly, the GCD recognizes those clusters of monotyp-
ic or polytypic biological entities, identified using morphology or genetics,
forming 'groups of individuals that have few or no intermediates when in
contact'. This is a non-dimensional, polythetic, and phenetic concept of
diversity serving largely as a surrogate of the BSC.
Synopsis
There are several evidential, philosophical, empirical and theoretical prob-
lems associated with this definition, precluding its use as a primary con-
cept for species. Problems associated with the BSC, GSC, HSC, MSC,
NDSC, PhSC, PtSC, SSC, and TSC hold true for this species concept.
19.7.12 Hennigian Species Concept (HSC)

'... involving tokogenetic relationships"; a (potential) reproductive
community'. (Hennig, 1950: 45-46)
'... reproductively isolated natural populations or groups of natural
populations [that] originate via the dissolution of the stem species in
a speciation event and cease to exist either through extinction or spe-
ciation'. (Meier and Willmann, 1997)
Synonyms
BSC.
Discussion
This concept is a derivative of Hennig's (1950) earlier notion of species. It
has been further developed by Willmann (1985a,b) and Meier and
Willmann (1997). Importantly, however, the version advocated by these
latter authors only incorporates some of Hennig's view of species. Their
concept is an operational concept, and by their own admission, is 'identi-
cal to the biological species concept if absolute [reproductive] isolation is
adopted as the criterion for contemporaneous populations, and the origin
of the isolation of two sister species is used to delineate species boundaries
in time'. However, they do view this concept as different from the Mayr's
BSC because 'he failed to provide a criterion that specifies how and when
biospecies originate and cease to exist (if not by extinction)'. Intertwined
in their discussion is the species concept issue and the significance of stem
(ancestral) species. Logically following from this extreme version of the
isolation concept (BSC) is that unisexuals are not species but are agamo-
taxa (sensu ASC), taxa not to be considered equivalent to bisexual species.
The HSC is rejected as an appropriate characterization of entities par-
ticipating in speciation for many of the same reasons the BSC is rejected.
The HSC should neither be employed for systematic questions nor issues
of biodiversity. For some points, however, it is apparent that Meier and
Willmann are more cognisant than Mayr of the fact that a concept of
species is important to people other than just a 'cataloguer and curator of
collections'. Thus, the HSC is characterized to be a dimensional concept to
be used for allopatric or allochronic questions, and unlike the BSC, it
acknowledges the importance of comparisons between sister taxa.
Synopsis
Regardless of any positive attributes over the BSC, the HSC is viewed as
inappropriate for biological systems and developed out of a limited view
of natural systems. Important problematic issues of this concept include
the exclusion of some biological diversity, its relational nature, its heavy
reliance upon operational criteria, its artificial advocation of isolation as a
non-arbitrary demarcation of species, and its artificial contrivance of stem
species.
19.7.13 Internodal Species Concept (ISC)

'... individual organisms are conspecific in virtue of their common
membership of a part of the genealogical network between two per-
manent splitting events or between a permanent split and an extinc-
tion event'. (Kornet, 1993: 28)
Synonyms
CISC and HSC (in part; Kornet, 1993), PSC (in part).
Discussion
Formalized in philosophical and analytical detail by Kornet (1993), this
concept identifies species solely on the basis of genealogical relationship.
No criteria exist for conspecificity (e.g. morphological similarity, inter-
breeding) other than that species are mutually exclusive groups of organ-
isms that derive from a permanent rift in genealogical connections.
Permanence refers to separation of a lineage into two or more lineages
that are never reunited by any level of interbreeding. This concept also
precludes the origin of taxa via hybridization because such an event
would terminate the independence of the lineages. While similar in some
ways to the CISC and HSC, Kornet and McAllister (1993: 64) advocate
modified versions of each but admit that the concept has 'very limited
practical value'.
Synopsis
The strict reliance upon permanent splits in genealogical networks, with
no possibility for future exchange, and the non-acceptance of species of
hybrid origin are unrealistic restrictions for a primary concept of species.
Such a concept would eliminate many taxa that either maintain their
independence through various mechanisms in spite of the fact that they
freely interbreed with relatives or are divergent lineages of hybrid origin.
This concept also confuses the phylogenetic lineages of species with the
life spans of individual organisms in tokogenetic arrays, such that the
death of one family unit would constitute a permanent split in the net-
work and hence speciation. Thus, this 'concept does not approximate at
all closely to our intuitions about the life span of species' (Kornet and
McAllister, 1993: 64).
19.7.14 Morphological Species Concept (MSC)

'Species are the smallest groups that are consistently and persistently
distinct, and distinguishable by ordinary means'. (Cronquist, 1978:15)
'Species may be defined as the easily recognized kinds of organisms,
and in the case of macroscopic plants and animals their recognition
should rest on simple gross observation such as any intelligent per-
son can make with the aid only, let us say, of a good hand-lens'.
(Shull, 1923: 221)
'The smallest natural populations permanently separated from each
other by a distinct discontinuity in the series of biotypes'. (Du Rietz,
1930: 357)
'A species is a community, or a number of related communities,
whose distinctive morphological characters are, in the opinion of a
competent systematist, sufficiently definite to entitle it, or them, to a
specific name'. (Regan, 1926: 75)
Synonyms
Classical Species Concept, Linnaean Species Concept, Morphospecies
Concept, PhSC, TSC. (Sokal, 1973; Grant, 1981; Stuessy, 1990).
Discussion
This is probably considered the most sensible and commonly used
method of species definition by taxonomists, general biologists, and
laypersons alike. Because in the vast majority of situations involving
allopatric populations little or no information is available regarding repro-
ductive independence, morphological distinctiveness serves only as a sur-
rogate to lineage independence. This concept also bridges a decided gap
inherent in some other concepts between sexual and asexual species, so
long as morphological distinctiveness is heritable and is representative of
lineage independence. Given that humans are a vision-oriented species, it
is readily appealing as an operational concept. Kornet (1993) considers
morphology in its widest sense wherein 'similarity between organisms
may thus be perceived in macromorphology as well as in gene-structure,
and may range from shared "sets of independent characters" for classical
taxonomists to shared "unique combinations of character states" for pat-
tern cladists'. In this case, some may consider the MSC to be synonymous
with the PSCj.
The only real problem with a morphological concept involves instances
of sibling or cryptic species, or the retention of plesiomorphic morpholo-
gies. Here, little or no morphological divergence has accompanied the
acquisition of lineage independence and two or more different species
may appear similar. In such cases a morphological concept of species will
underestimate biological diversity. Another potential problem with this
concept is the inherent tendency to require an arbitrary level of morpho-
logical divergence. By employing such a criterion the researcher assumes
that all morphological traits, especially those traditionally employed in a
taxon, evolve at a constant rate of divergence. This is an unjustified
assumption and is falsified by the observation that even within a taxo-
nomic group morphological divergence is largely random.
Synopsis
This is a non-dimensional concept that treats species as classes, defining
them on the basis of particular essential morphological attributes.
Possession of these essential attributes provides for membership in the
species. As such it does not allow the researcher to treat species as historical
entities forming lineages. As individuals, the definition of every species will
necessarily change as the essential attributes of a species at tj will be differ-
ent from t2 through descent. While this concept has served as a traditional
method for identifying species it is fatally flawed as a primary concept.
19.7.15 Non-dimensional Species Concept (NDSC)
Synonyms
BSC, GSC, MSC, Palaeontological Species Concept, SSC, TSC.
Discussion
Several traditional concepts of species qualify as NDSCs, the most popu-
lar being the BSC. Concepts of this type have limited spatial and no tern-
poral dimension of species in question. Thus, there is no evolutionary,
phylogenetic, or lineage perspective with which one can view, perceive,
or interpret descent of the taxa or their attributes (e.g. shared plesiomor-
phies or apomorphies, distances), including the ability or propensity to
interbreed. Concepts of this nature may appear to be more operational
than those incorporating temporal and geographic components.
However, this convenience compromises both the accuracy and precision
with which we are able to identify, quantify, and understand biodiversity.
Finally, in this lack of accuracy we also lose our abilities to discover and
understand the processes responsible for the evolution, functions, and
maintenance of biodiversity.
Synopsis
Thus, while the non-dimensional species concept has been argued by
some as a preferred operational concept of diversity, it has actually been
a hindrance to the advancement of comparative and evolutionary biolo-
gy. Concepts of this type should not be considered as primary concepts
for species. Interestingly, in some areas of science (medicine) the non-
dimensional concept has been perceived as grossly inferior to concepts
incorporating spatial and temporal dimensions in discovering diversity
(Paterson, 1993).
19.7.16 Phenetic Species Concept (PhSC)

'... the species level is that at which distinct phenetic clusters can be
observed'. (Sneath, 1976: 437)
Synonyms
BSC (in part), GCC (in part), GSC, GCD, MSC, NDSC, Palaeontological
Species Concept, SSC, PtSC, TSC.
Discussion
This is a non-dimensional and strictly operational concept that may be
likened to any concept where overall similarity is the primary criterion for
the existence of species. Operationally, where variation in a set of charac-
ters is less within a group than between groups the entity is recognized as
a distinct taxon. Species are treated as Classes under this concept; they do
not exist as lineages and, if a species changes through descent, then the
diagnosis will have to be revised.
Synopsis
While essentially the methodology employed by taxonomists, the barren
theoretical nature of this concept precludes its use as a primary concept.
19.7.17 Phylogenetic Species Concept (PSC)
Currently at least three different concepts of species are identified as phy-
logenetic. These definitions represent an outgrowth of phylogenetic sys-
tematics and a general need among some researchers for an operational,
lineage definition of species that is process-free. Some argue that with the
growing popularity of phylogenetics it is critical to have a definition to
identify the smallest units suitable for analysis (boundary between toko-
and phylogenetic processes). For some, species is the smallest unit appro-
priate for analysis, and infraspecific units are inappropriate in this context
(Nixon and Wheeler, 1990; Wheeler and Nixon, 1990). This same perspec-
tive holds that species diversity must be understood before a phylogenet-
ic analysis is performed. Others defend the position that hierarchical pat-
terns exist within species and phylogenetic methods are appropriate (de
Queiroz and Donoghue, 1988,1990; McKitrick and Zink, 1988).
Common to PSCs is an attempt to identify the smallest biological enti-
ties that are diagnosable and/or monophyletic. Species are thus the bio-
logical entities and unit product of natural selection and descent.
Consequently, subspecies, fraught with ambiguities between convenience
and naturalness, is not an appropriate evolutionary unit and has no onto-
logical status (Cracraft, 1983; McKitrick and Zink, 1988; Warren, 1992). The
different PSCs form three general Classes; one emphasizing monophyly,
one emphasizing diagnosability, and one emphasizing both. Many simi-
larities exist with the ISC, CISC, CpSC and the PSC.
19.7.18 Diagnosable Version (PSCj)

'... a diagnosable cluster of individuals within which there is a
parental pattern of ancestry and descent, beyond which there is not,
and which exhibits a pattern of phylogenetic ancestry and descent
among units of like kind'. (Eldredge and Cracraft, 1980: 92)
'... the smallest diagnosable cluster of individual organisms within
which there is a parental pattern of ancestry and descent'. (Cracraft,
1983: 170)
'... simply the smallest detected samples of self perpetuating organ-
isms that have unique sets of characters'. (Nelson and Platnick,
1981: 12)
'... the smallest aggregation of populations (sexual) or lineages (asex-
ual) diagnosable by a unique combination of character states in com-
parable individuals (semaphoronts)'. (Nixon and Wheeler, 1990)
'... the smallest aggregation of (sexual) populations or (asexual) lin-
eages diagnosable by a unique combination of character states'.
(Wheeler and Platnick, 1997)
Synonyms
CISC (in part), CpSC (in part; Kornet and McAllister, 1993), GSD, ISC (in
part; Nixon and Wheeler, 1990; Kornet, 1993), PhSC, PtSC, TSC.
Discussion
This Class of definitions emphasizes the a priori diagnosability of species,
irrespective of a criterion of monophyly. There are two purported benefits
of this perspective. First, process is not invoked before pattern is observed.
Second, phylogenetic methodologies are argued to be applicable only to
genealogical relationships of species and supraspecific taxa, not below the
level of integration of species wherein tokogenetic relationships of infra-
specific entities are the norm (sensu Wheeler and Nixon, 1990; Nixon and
Wheeler, 1990). To conduct a phylogenetic analysis below the level of
species would confuse the reticulate tokogenetic relationships with the
usual non-reticulate phylogenetic relationships.
For proponents of this concept, monophyly, paraphyly, and polyphyly
apply only at a level of organization above species. Species are delimited by
the distributions of fixed, diagnostic characters across populations. Where
variability exists in an attribute within the taxon this attribute is considered
inappropriate for that level of analysis where only tokogenetic, not phylo-
genetic, relationships exist. However, the operation(s) necessary for the
practical delineation of tokogenetic and phylogenetic relationships is not
developed explicitly by those favouring this concept. Without knowing if
you are dealing with one or more species a priori, one is not likely to know
if phylogenetic methods are appropriate. Likewise, the difference is
unclear between the theoretical inapplicability of phylogenetic methods in
tokogenetic systems versus using the same methods for resolving relation-
ships of species derived via hybrid origin. Both contain reticulate patterns
of history.
19.7.19 Monophyly Version (PSC2)

'...a geographically constrained group of individuals with some
unique apomorphous character, is the unit of evolutionary signifi-
cance'. (Rosen, 1978:176)
Synonyms
Apomorphy Species Concept (Wheeler and Platnick, 1997), CISC (in part),
ISC (in part; Kornet, 1993).
Discussion
For Rosen (1978,1979) and de Queiroz and Donoghue (1988,1990) species
have reality if they are monophyletic and supported by autapomorphies.
Any biological entity possessing a uniquely derived character, of any type,
magnitude, or quantity, qualifies as a species. Those not possessing
autapomorphic attributes do not constitute a species, as traditionally
viewed, but are referred to as metaspecies by some. The application of this
concept necessitates a phylogenetic analysis. A lucid discussion is offered
in papers by de Queiroz and Donoghue.
19.7.20 Diagnosable and Monophyly Version (PSC3)
Synonyms
CISC (in part), CpSC (in part; Kornet and McAllister, 1993), ISC (Nixon
and Wheeler, 1990; Kornet, 1993), SSC.
Discussion
The PSC of McKitrick and Zink (1988) is a modification of the PSC pro-
vided by Cracraft (1983) but incorporates the criterion of monophyly for
species. While a definition was not provided by McKitrick and Zink (1988),
they identified a species as the smallest diagnosable cluster of individual
organisms forming a monophyletic group within which there is a parental
pattern of ancestry and descent. Because in this conceptualization all rec-
ognized monophyletic taxa are diagnosable, this definition, the methods
for the discovery of species, and any associated practical and theoretical
limitations are equivalent to aspects of the PSCj and PSC2-
Synopsis
Several positive aspects of the phylogenetic concepts make them particu-
larly attractive as operations in discovering biodiversity, and resolving
some of the perceived problems with other concepts (Mayden and Wood,
1995). In all versions the PSC is an operational definition, whether one
uses diagnosability or monophyly. The set of operations necessary to dis-
cover diversity associated with species are clearly outlined. The concepts
incorporate the notion of lineage(s), making them appropriate for recon-
structing descent and interpreting evolution of attributes. The ability to
interbreed is viewed as a shared-primitive attribute and not of conse-
quence in the recognition of species as taxa. These concepts also have the
ability to recognize both biparental and uniparental species, and possess
no implied modes of selection nor speciation. Finally, in the execution of
these concepts there is no inherently arbitrary divergence or distinction
between species or subspecies in a polytypic species (Cracraft, 1983;
Warren, 1992); subspecies have no ontological status.
There are some problems with the use of these concepts and these are
reviewed by Mayden and Wood (1995). However, while there are prob-
lems with the exclusive use of any of the Classes of the PSC, there are also
important positive operational aspects to these concepts over some others.
I concur with the conclusions of Warren (1992: 34) in that the PSC serves
as an excellent operational surrogate to a concept of species not implicat-
ed with as many variables limiting our potential to discover biodiversity.
Yet, none of the versions of the PSC should serve as a primary concept.
19.7.21 Polythetic Species Concept (PtSC)
Synonyms
BSC (in part), GCD, MSC, NDSC, PSC (in part), PtSC, SSC, TSC.
Discussion
This concept derives essentially from what philosophers call cluster con-
cepts. That is, species are defined by the statistical covariance of characters
deemed important. A given individual belongs to a particular species if it
possesses enough of the important characters for the species. This statisti-
cal and practical definition treats species as classes, not individuals. Often,
species are delimited by their possession of a unique combination of char-
acters, and these are usually phenotypic. Most individuals of a species
may possess attribute A, while those not possessing A will still have attrib-
utes B, C or D, all features also viewed as characteristic of the species.
Treated as natural kinds, species are not viewed as lineages.
Synopsis
While this concept may serve as a very useful operational recipe for the
delineation of species, especially in situations with complex patterns of
variability of characters, it has no theoretical basis for being considered a
primary concept. Because species are both individuals and lineages, their
diagnoses will necessarily have to be modified over time as their diagnos-
tic attributes become modified through descent.
19.7.22 Recognition Species Concept (RSC)

'A species is that most inclusive population of individual, biparental
organisms which share a common fertilization system'. (Paterson,
1993:105)
Synonyms
BSC (Mayr, 1988).
Discussion
This concept was introduced Paterson (collective writings in Paterson,
1993). It was developed from a dissatisfaction with the BSC, a definition
considered inadequate and inaccurate of natural patterns or processes,
and inhibiting progress towards related goals. For Paterson the biological
limits to the field for gene recombination are determined by the mate
recognition system, more precisely, a specific mate recognition system
(SMRS), a series of coadapted signals and releasing properties exchanged
between partners through complementary systems. The system is func-
tional across a broad array of conceivable signal-reception methods from
elaborate behaviours, including chemicals and pheromones, to cellular
recognition by gametes. This coadapted complex is maintained by strong
stabilizing selection as long as the species inhabits its natural habitat; this
changes when the natural habitat for the species (perhaps ancestral) is
changed through geographic or temporal disjunctions. At this point the
coadapted complex of signals exchanged between partners may become
altered via directional selection in the new habitats occupied by the
descendant groups of daughter populations (or species). Paterson (1993:
33) argues that 'a new SMRS, derived in this way, determines a new gene
pool and, hence, a new species. According to the recognition concept,
species are populations of individual organisms which share a common
specific-mate recognition system. Species are, thus, incidental effects of
adaptive evolution'.
The RSC does not invoke a major role for selection in the evolution of
positive assortative mating, the development of isolating mechanisms,
and does not require sympatry and evolutionary reinforcement to com-
plete speciation. The fallacy that selection is responsible for producing
adaptations that, by design, are responsible for the isolation of gene pools
is obvious from the observation that in large part the documented cases of
speciation are the direct result of total allopatry, a speciation model that
does not involve secondary contact and/or reinforcement of isolating
mechanisms (see May den and Wood, 1995). Thus, if isolating mechanisms
are products of descent they are the result of chance rather than design.
The general question for the RSC is not what are the characters and
mechanisms that have evolved in the recognition or reproductive systems
of a species that prevents successful matings and resulting ontogenetic
development between sympatric species? Rather, what are the characters
and mechanisms that have evolved in species that ensure effective syn-
gamy, development, and future generations within a population occupy-
ing its preferred or natural habitat? (Paterson, 1993: 33).
Synopsis
While there are important positive theoretical and applied aspects to this
concept permitting the identification of species in a largely process-free
environment, there are important problems with a universal application
of the RSC. These include: (i) strict reliance upon and knowledge of the
SMRS; (ii) lack of a lineage perspective; and (iii) exclusion of uniparental
species and species with retained-primitive SMRSs. These are reviewed
by Mayden and Wood (1985). Thus, the RSC should not be viewed as a
primary concept of species.
19.7.23 Reproductive Competition Concept (RCC)
'... the most extensive units in the natural economy such that repro-
ductive competition occurs among their parts'. (Ghiselin, 1974: 538)
Synonyms
BSC (in part; Ridley, 1989), Hypermodern species concept (Platnick, 1976).
Discussion
This is a non-dimensional and non-operational conceptualization of species.
It is essentially limited to sexually reproducing species because of its focus
on the intra- and interspecies competition for mates species. In its formal-
ization, Ghiselin (1974, 1984a) likens species and evolutionary theory to
firms, corporations, small businesses, craftsmen, etc. and economic theory.
Synopsis
The restriction of this concept to sexually reproducing organisms pre-
cludes its use as a primary concept of species. Should this restriction be
eliminated, this concept could serve as a primary theoretical concept.
However, competition for mates in reproduction is difficult to entertain
for entities generally termed uniparentals.
19.7.24 Successional Species Concept (SSC)
Synonyms
Palaeospecies concept (Simpson, 1961), ESC (in part; Simpson, 1961),
Chronospecies concept (George, 1956).
Discussion
This concept was devised as a surrogate for estimating divergence
through time by researchers studying fossil taxa. Often these researchers
have only fragmentary data both in specimens and through time to eval-
uate anagenesis and divergence.
In reality, the distinctions between successional species is an arbitrary
delineation in time or strata based on divergent morphologies or gaps in
morphologies or time. With anagenetic change within a lineage and only
remnants surviving for study there is potentially an unlimited number of
chronospecies throughout the history of what was once only a single self-
integrating lineage behaving evolutionarily as a single species. The SSC is
an operational concept, largely of convenience, to allow researchers of fos-
sil taxa to communicate equivalent geological strata. Species identified
using this concept should not be misconstrued as being biologically equiv-
alent to species identified using most other concepts. This is not to say that
there are not valid species that have been identified using this concept.
However, in general, palaeospecies are usually temporal forms of a single
species' lineage. While Simpson's (1961) ESC did extend the non-dimen-
sional BSC through time and provide much more of a lineage perspective
to species, Simpson would argue for subdividing a single lineage into mul-
tiple chronospecies. The ESC of Wiley (1981) and Wiley and Mayden
(1985,1997), however, does not advocate chronospecies.
Synopsis
Because of the arbitrary and non-evolutionary nature of this concept it
should not be considered a primary concept.
19.7.25 Taxonomic Species Concept (TSC)

'...a species consists of all the specimens which are, or would be,
considered by a particular taxonomist to be members of a single
kind as shown by the evidence or the assumption that they are as
alike as their offspring or their hereditary relatives within a few
generations. When there is no evidence of the hereditary relation-
ship, the taxonomist will rely on distinctions that have been found
to be effective in segregating species among other [groups]'.
(Blackwelder, 1967: 164)
Synonyms
ASC, GCD, MSC, PhSC (Sokal, 1973; Sneath, 1976), PSC (in part), PtSC.
Discussion
As described by Blackwelder (1967), 'these are the species of the taxono-
mist; they are not necessarily the species of the geneticist or the evolu-
tionist'. This concept is probably used by most practising taxonomists as a
working definition to segregate individual organisms in different taxa. It
relies primarily on morphological attributes in the delineation of species
because many other character bases have traditionally not been readily
available to taxonomists. In practice, it is non-dimensional, treats species
as classes, and lacks a lineage perspective.
Synopsis
The traditional character-based limitations for those in the field of taxon-
omy are less real in modern science. Many different types of characters are
become increasingly more available and should be used in the delineation
of taxa. However, given that humans are a vision-oriented species, the
more convenient morphological attributes will probably remain the most
used characters in deciphering taxonomic diversity. This truism, however,
need not negate the existence of taxa identified using other types of char-
acters (ecology, proteins, behaviour, sequences, etc.).
19.8 DISCUSSION
'An ideal species concept should meet the various intuitions that we
have about species.... It is tempting to try to define a fully satisfying
species concept which meets all the intuitions mentioned by some-
how combining the definitions which address the different intuitive
requirements. But part of the species problem originates in the fact
that any attempt to combine [different] definitions into a more
embracing concept, in which their criteria are given equal weight, is
doomed to fail. This is because their criteria for conspecificity are
incompatible; i.e., two organisms which are conspecific on the crite-
rion of one concept are not necessarily so on that of another'.
(Kornet, 1993: 29)
'To do justice to the intuition that species are historical entities, we
required a species concept which defines species as entities with con-
tinuity in time between their origin and end'. (Kornet, 1993: 32)
'The species problem has often been approached with the presup-
position that a single kind of entity exists in nature that corresponds
to a species concept, just because the word 'species' exists in the lan-
guage of biology. If this presupposition is dropped then the tradi-
tional species problem could be answered, at least in principle, by
enumerating a heterogeneous list of the general characteristics that
have been thought to bestow specific status to clusters of organisms'.
(Wilkinson, 1990: 445)
In this discussion the following theories are taken as having reality in
the natural world:
1. The notion of descent with modification is a unifying theory of natural
sciences. Descent operates from kin-groups or populations to species as
groups. Descent involves differential change in attributes or qualities
originating through a variety of processes over time (generations) and
space (geography).
2. Speciation results in the production of new species over time and
space, a direct result of (1).
3. Classes have definitions, are spatiotemporally unrestricted, lack cohe-
sion, and do not participate in natural processes.
4. Individuals lack definitions, are spatiotemporally restricted, have cohe-
sion, and participate in processes.
There are at least five consequential factors that have fuelled the long-
standing controversy over the species problem. These include: (i) a tradition
of occupation; (ii) formalized rules of nomenclature; (iii) misunderstanding
of terms; (iv) a persistent desire by humans for working definitions; and (v)
the unique nature of those things that we hope to understand, i.e. species
as taxa (or groups). Traditionally, the job of discovering and identifying
Discussion 413
diversity was left to the occupation of the taxonomist. For many of these
researchers their responsibilities were viewed as finding different species
and detailing attributes important for their identification (sensu TSC). In
many ways this mode of operation relegates species as taxa to classes with
essential features. While convenient for a user hoping to distinguish
between the different things, this treatment ultimately leads to great diffi-
culties with operationality and theory when species are known to partici-
pate in processes and evolve as either ancestors or descendants.
The formalized rules of nomenclature have reinforced the view of
species as classes. The recognition of species requires not only their
description and the designation of a type, but also its diagnostic features.
For many, a diagnosis entails a listing of defining features, a prescription
easily misunderstood as equivalent to essential traits. Thus, the opera-
tional necessity and emphasis on a diagnosis may be viewed as treating
species taxa as classes. This, in concert with the traditional TSC, fosters
great difficulties in the reconciliation of species as individuals.
This occupational and operational legacy has resulted in the confusion
of the ontological categories classes and individuals (more recently
Historical Groups). While some may view this aspect of the problem as
purely metaphysical and without significant bearing on the issue, such a
perspective is absolutely wrong and continually generates difficulties in
resolving the controversy. As discussed herein, it is not merely an argu-
ment to distinguish between species as category and species as taxa. The
delineation between classes and individuals is necessary, but not suffi-
cient to resolve the problem. Species as taxa are individuals; species as cat-
egory are classes. The former have no defining properties and can only be
described; the latter can be defined through a series of desired properties
for its members (species assigned to categories). Both diagnoses mandat-
ed by nomenclatural formalities and most species concepts treat species as
taxa as if they are classes and immutable. In reality, they function only as
operational guidelines or surrogate concepts for the discovery of those
individual-like things that we think to be species. This exercise is of great
necessity because the individual-type things are fuzzy and can only be
diagnosed retrospectively. Most species concepts are functional constructs
or definitions (class) employed link to our notion or concept of the species
as taxon (individual).
19.8.1 Concepts and definitions of species and supraspecific categories

It is clear that the various categories used in biological classification are
class concepts. In the way that we use these categories they are intimate-
ly linked to both theoretical and operational concepts. While there exists
an infinite number of ways to organize groups of groups and assign them
to supraspecific categories, the definition for Hennig (1950,1966) and most
others today is any historically formed group wherein the ancestor and all
its descendants are included in a phylogenetic nexus, also known as
monophyly. There is nothing operational about this definition; it is strict-
ly a notion thought to be in harmony with the theory of descent with
modification and our empirical observations of the end products of this
process. Yet, most agree that while non-operational it is the appropriate
theoretical concept for groups assigned to supraspecific categories. It is
impossible to observe monophyly either historically or in real time
because we do not witness the evolution of ancestors and their descen-
dants, and the theoretical definition of monophyly provides no opera-
tional guidelines. Without direct observation, we are incompetent in our
abilities to locate monophyly, unless we retain bridging principles
through some type of operational concept. The concept that we now use
as a suitable surrogate to monophyly is that of synapomorphy. That is,
operationally we recognize supraspecific taxa or monophyletic groups
through character analysis and the discovery of shared-derived charac-
ters inferred to have evolved in a common ancestor and retained or exist-
ing as homologues in immediate descendants. There are, however, other
possible concepts of monophyly (e.g. percent similarity, ability to
hybridize, etc.) that could be employed. These are rejected because of
known inconsistencies between alternative concepts and empirical and
theoretical observations of the real world. Families are no longer defined
by a prescribed level of similarity greater than that expected of orders, but
less than that of genera, because similarity does not always denote close
relationship. Likewise, the ability to reproduce is an ancestral feature of
lineages and is retained unless there is anagenesis of some attribute(s)
closely linked to reproductive success.
Interestingly, the species problem can be seen to parallel the theoretical
and operational issues traditionally associated with supraspecific taxa and
categories. These separate problems are very similar if one is conscious of
the ontological differences between species as taxa and species as catego-
ry. Everyone has their own notion of what species as taxa really are. In
many instances we use this concept to guide our perceptions, observa-
tions, and understanding about theoretical and empirical aspects of the
natural world (diversity, character evolution, speciation, ecology, physiol-
ogy, etc.). The class concept species as category used in the assignment of
groups from nature thought to fit its definition. Is there, or should there
be, a definition of species that is both theoretical and operational analo-
gous to the relationship that exists for supraspecific taxa and categories?
Should the concept of the species category be operational or should it be
theoretical like the concept of monophyly for supraspecific groups? Either
way, which concepts should be used? If operational, which of the avail-
able operations should be employed, if any? If theoretical, which theory
should be used? Unlike the abundant discussions following from the long-
Discussion 415
standing issue of concepts and criteria for recognizing supraspecific taxa,
there is limited discussion clearly focusing on these issues for species. To
date, no progress has issued from the multitude of pages debating the
species problem. It is my opinion that the difficulties associated with
species in theory and species in practice derives directly from the confla-
tion of these notions. Much to the disappointment of many, a coveted
denouement in this long standing controversy will never follow until
these questions are addressed with a sound metaphysical framework.
19.8.2 Evaluating concepts for important qualities

Hull (1997: Chapter 18) advocates a comparison of species concepts on
three traditional criteria: theoretical significance, generality, and applica-
bility (or operationality). I agree that this is an appropriately unbiased
method of comparison. Herein, I regard applicability and operationality as
separate criteria because a concept can be fully operational but not at all
applicable to a problem, and vice versa. How do the various concepts fair
in evaluations of these criteria?
(a) Theoretical significance

While difficult to measure across the various concepts, there are important
theoretical differences between them. The most significant ingredient is
the treatment of species as individuals rather than classes. Excluding the
ESC, all concepts treat species as classes. Such concepts preclude sound
interpretations of speciation, character evolution, etc. because species can-
not be perceived as lineages. The deficiency of a lineage perspective even-
tually leads researchers to view all attributes and geographic locations of
species as proximal and causal explanations. For example, if one of three
taxa has an array of traits appearing intermediate between two other adja-
cent taxa, what explanation other than hybridization or introgression,
could account for such a pattern? Phylogenetic intermediacy represents
only one of several when species are viewed as lineages (Mayden and
Wood, 1995).
(b) Generality
Several theoretical and empirical elements of species concepts, relative to
species as taxa, may be considered under this criterion, including their tol-
erances of divergent lifestyles, modes of reproduction, modes of specia-
tion, genetic exchange, distributional and character information, and
finally, diagnoses. Not all concepts view evidentiary information perti-
nent to these elements equally. Informative comparisons of generality
require some estimate of baseline diversity to be recovered, or things that
we currently envisage as behaving like species. Tolerance limits for each
concept must be compared with this baseline of diversity.
What is our working baseline of diversity? First, we know that species
exist that encompass the entire gamut between sexual and asexual repro-
duction, with numerous intermediate conditions (Templeton, 1989).
Numerous speciation modes have been hypothesized for organismic
diversity, ranging from complete allopatry to complete sympatry (Wiley,
1981). Numerous examples exist wherein species exchange genetic infor-
mation either in current communities or historical communities without
condemnation of identities. In fact, some hypothesized historical genetic
exchange between groups may be responsible for the evolutionary success
of the involved groups, each going on to produce diverse clades (Mayden
and Wood, 1995). Finally, the types of character information traditionally
used to discover species is heterogeneously distributed across taxonomic
groups. When viewed across all taxonomic groups all types of data from
DNA and RNA sequences and similarity, to behaviour and ecology, pro-
tein variability, morphology, and other traits, are standard markers used
to reveal species diversity.
Some concepts are basically intolerant of gene exchange between
species and require sympatry before species can be validated (HSC, BSC,
ISC, CpSC). Under the BSC, taxa in allopatry are sometimes considered
semispecies. Because gene exchange is not tolerated, speciation via
hybridization is also not a valid form of speciation under some concepts.
Some of these concepts are also intolerant of uniparental reproduction.
Some are intolerant of groups of individual organisms that may be para-
phyletically related to one another relative to one or more descendants;
that is, all surviving ancestral species (PSC, GCC, CISC). Some usually
only recognize species wherein there has been divergence at the mor-
phological level (MSC, TSC, PhSC). Likewise, some demand divergence
at the ecological (EcSC) or recognition system (RSC) level. One concept,
the ESC demands only that speciation and evolution are natural process-
es involving lineages that maintain cohesion and have unique identities.
The ESC has thus the greatest generality. All other concepts are less gen-
eral and exclude real diversity.
(c) Operationality
This is one quality consistently argued in discussions of species, either
implicitly or explicitly. That is, anyone should be able to follow a pre-
scribed set of identifiable and repeatable operations and at the end of
these operations be able to tell (with a certain level of confidence) if they
have a species. The requirement of such an execution places limits on
what is recognizable, defined by criteria of the operational concept. While
this may be more convenient, convenience is not a criterion that should be
Hierarchy of concepts: species in theory and practice 417
optimized when attempting to discover and understand pattern and
process in the natural world. Operationalism is a fundamental fault of any
species concept adopting it. What is operational is determined strictly by
the perceived reality of the viewer. If the viewer's senses perceive only a
portion of reality and these are expressed in an operational definition of
what reality consists of, then we will never know otherwise. If, however,
the viewer is capable of perceiving or conceptualizing all of reality, then
all of diversity can be discovered without placing limits on what can be
recognized with an operational concept. For instance, it is a mistake for
someone who is red-green colour blind to mandate a concept of species
based on the operational criterion of colour. Anyone discussing species
diversity of hummingbirds, flowering plants, or darters, with this person
would continually be frustrated with what is reality.
Excluding the ESC, all of the other concepts are operational at some
level. That is, with all of them one can conduct certain experiments and
extract pertinent information about the criterion emphasized. Some are
more operational than others but with this increasing operationality one
necessarily sacrifices an ability to account for diversity. For example, the
ASC, MSC, PhSC, SSC, or TSC are probably the most operational con-
cepts guiding the discovery of species. These concepts, however, will nec-
essarily exclude equally valid species that can and will be recognized
using other concepts. The next most operational concepts would include
CISC, CpSC, EcSC, GCC, GCD, NDSC, versions of the PSC, and RSC. The
BSC, HSC, and RCC are all minimally operational. The ESC is unique in
not being an operational concept, a consequential quality for a primary
concept. Nothing in the ESC, other than evolution produces species as
lineages with identities and cohesion, is operational; this, however, is
extremely difficult to apply without bridging principles.
(d) Applicability
Given that the various concepts were all formulated from research on
patterns of diversity across a diversity of temporal and geographical sit-
uations using varied technologies, each attempting to unveil processes
associated with descent, they are all applicable as concepts of species as
taxa. However, applicability extends from those having lesser applicabil-
ity and embracing only a subset of natural diversity, to those with
greater applicability wherein the concept embraces most or all of diver-
sity. The ESC has the greatest applicability because it is consistent with
and embraces all known species diversity that has evolved through cur-
rently understood processes descent. All other concepts have lesser
applicability because as class constructs they are capable only of embrac-
ing a lesser portion of natural diversity by excluding some forms of
species (e.g. asexuals, ancestors, etc.).
19.9 HIERARCHY OF CONCEPTS: SPECIES IN THEORY AND
PRACTICE
When Mayr (1957) discussed species concepts and definitions, he
mentioned - but did not dwell on - a need for two different levels of con-
cepts for species, these being primary and secondary concepts. 'All our
reasoning in discussions of "the species" can be traced back to the stated
three primary concepts. As concepts, of course, they cannot be observed
directly, and we refer to certain observed phenomena in nature as
"species", because they conform in their attributes to one of these concepts
or to a mixture of several concepts. From these primary concepts, just dis-
cussed, we come thus to secondary concepts, based on particular aspects
of species' (Mayr, 1957:16).
As primary concepts, Mayr is referring to the typological, second and
third species concepts discussed in that paper. From the discussion it is clear
that he recognized that all species as taxa would fit one or a combination of
these concepts, but that secondary concepts are those used to identify
species and employ differences in morphology, genetics, behaviour, etc. to
infer diversity consonant with primary concepts. It is unfortunate that more
scientists from all disciplines had not read this passage in 1957 and
bequeathed this philosophy to their academic descendants.
As fundamental links bridging observable patterns and inferred
processes, concepts are employed in every discipline, assisting to guide our
understanding, perception, and disclosure of natural systems. Given that
descent with modification and speciation are undeniable processes of
diversification and that individual species are the highest level of organi-
zation capable of participating in these processes, a monistic notion of
species is not only natural but is logical. Descent and speciation are
processes occurring in lineages. Individual organisms to populations, each
with spatiotemporal cohesion, and only lineages with this type of integrity
uniquely participate in speciation. A primary concept of species is funda-
mental to the whole of biological sciences, particularly for understanding
species as taxa. Currently, the multiple concepts in operation are decided-
ly inconsistent with one another as to what constitutes diversity (also see
Hull, 1997: Chapter 18) and most are inconsistent with the range of diver-
sity acknowledged as species in different disciplines. Without a primary
concept as a working hypothesis and to serve as a bridge between pattern
and process, it is untenable that we can advance on many fronts.
Heretofore, much of our effort has been expended struggling with the con-
ceptualization of species. Many adopt only operational concepts that will
produce contrived species diversity; unfortunately those searching for pat-
tern and processes associated with this diversity may be deceived.
What then are the criteria we should be looking for in a primary con-
cept? It should be consistent with current theoretical and empirical knowl-
Hierarchy of concepts: species in theory and practice 419
edge of diversification. It should be consistent with the ontological status
of those entities participating in descent and other natural processes; that
is, species as taxa must be referenced as individuals. Finally, it should be
general enough to encapsulate all types of biological entities considered
species as taxa by researchers working with supraspecific taxa. Only the
ESC is suitable as primary concept, guiding our quest for species as taxa
and our search for natural order. This concept is robust theoretically and
is unique in its global generality. One drawback is that it is not opera-
tional. While this may be viewed as a possible shortcoming, it is not so for
a primary concept. The ESC is maximally applicable because everything
we currently understand about descent, speciation, and species are com-
patible with the intent of the ESC.
While the ESC is the most appropriate primary concept, it requires
bridging concepts permitting us to recognize entities compatible with its
intentions. To implement fully the ESC we must supplement it with more
operational, accessory notions of biological diversity - secondary con-
cepts. Secondary concepts include most of the other species concepts.
While these concepts are varied in their operational nature, they are
demonstrably less applicable than the ESC because of their dictatorial
restrictions on the types of diversity that can be recognized, or even
evolve. However, they serve as surrogates for the ESC and, together, fur-
ther our understanding of descent, both anagenesis and cladogenesis, by
recognizing any entity consistent with the primary concept. They repre-
sent the practical or applied definitions, guidelines, or tools, used by
investigators to discover hypothesized real particulars, entities, individu-
als, or things that we accept as species. These secondary concepts can
account for nearly all species diversity, with the possible exception of
ancestral species, either surviving or extinct. Because the ISC and SSC are
both capable of delineating diversity beyond real species, these definitions
must be used with caution.
Together, the primary and secondary concepts form a hierarchical sys-
tem displaying both their operational and theoretical inter-relationships
(Figure 19.1). The primary concept is the ESC. Relationships among the
secondary concepts may be envisioned in multiple forms; I have illustrat-
ed only one such system. The one criterion emphasized most throughout
discussions of secondary concepts is sex. I have chosen to use this as a first
level criterion among secondary concepts. Reproduction, similarity-
dissimilarity, monophyly, diagnosability, apomorphy, and tolerances for
gene exchange are other criteria used to further reveal relationships
among secondary concepts (Figure 19.1). Some concepts (BSC, GSC, RCC)
terminate at multiple locations in the hierarchy either because of different
uses or ambiguities in the concepts. However, this is acceptable and one
may view other concepts as having similar results.
Primary concepts
nary species
1 1
^^VVSAAAA^SA/vVVVVVSAVW^
<X Secondary concepts $g<
0<X> Sexual or asexual
X< Only sexual reproduction 8?s<
|
No N ^ Minor ^ x\Vx\\N\\\\\> NX\\\\\\\VxX
^Reproduction \ sxMonophyyxN ^Difference-similarity^
,x
interspecific^ ^interspecific^; X\xx-v\\\x\xX>
^ gene \ ^ gene ^
^
exchange N ^ exchange ^ f
tolerated x ^ tolerated ^ BSC* ^Evidence of ^ ^Reciprocal ^ ASC NdSC
', monophyly ;/ ^monophyly//
1 1 1 1 .
GCS RCC / > CSC PhSC
1
HSC
1
BSC
^ No gene /
/exchange^
', tolerated /
Minor gene^
: exchange ',
tolerated ',
GCC
1
EcSC
1
PtSC
1 1 CSC CISC
1
GCD
1ssc
RSC GSC
ISC
1
PSC2
1
GSC
. 1PSC!
1
MSC
1TSC
PSC3
Figure 19.1 A. hierarchy of primary and secondary species concepts. The non-operational Evolutionary Species Concept serves as the
primary concept of species. The operational secondary concepts form a hierarchy below this primary concept based on their tolerances
or requirements for modes of reproduction, gene exchange, monophyly, and diagnosability. Because some concepts represent hybrid
versions of other concepts (mixed criteria) they may be depicted more than once in the hierarchy. Species concepts are listed alpha-
betically within any grouping. Asterisk denotes a version of BSC modified for asexual species. See Table 19.1 for concept abbreviations.
References 421
There are extraordinary advantages to accepting the premise of monistic
primary and pluralistic secondary concepts of species. First, a primary con-
cept of species that can be continually evaluated in light of new information
ensures that all things behaving as species are potentially recoverable, given
unavoidable constraints associated with available technology and extinc-
tion of taxa never observed. Second, with the possible exception of the ISC
and SSC, all currently employed secondary concepts theoretically compati-
ble with the primary concept can be mutually applicable in the discovery of
species and the elucidation of pattern and process. While some concepts
conflict in their intentions, they are all equally valid. When viewed togeth-
er as guidelines in the detection of species they ensure that natural species
diversity is neither unrecognized nor misunderstood. Thus, patterns
observed in the natural world can be used by all disciplines to reveal natur-
al processes in an uninhibited manner.
Our classification system for supraspecific taxa is analogous to the out-
lined system of primary and secondary concepts of species. Classifications
are theories about the organization of biological diversity. What groups
should be placed in the various supraspecific categories, and how and why
should just these groups be recognized over other possible groups? One
may choose to optimize various information in a classification, from overall
similarity, ecological guilds, or modes of reproduction, to genealogical rela-
tionships, just to mention a few. In the current system the concept adopted
for supraspecific categories is a particular genealogical relationship, specifi-
cally monophyly. Other criteria have been rejected as primary concepts
because of ambiguity, inconsistencies or artificiality. Thus, we employ
monophyly as a primary concept to bridge to natural groups in the classifi-
cation of supraspecific taxa. Because we are unable to observe descent we
adopt secondary concepts or definitions, particular homologies compatible
with the intentions of the primary concept. Through character evaluation
secondary concepts permit the organization of diversity into such groups.
Inferences derived from phylogenetic systematics can either corroborate or
falsify hypotheses of groups suspected to meet criteria outlined in our pri-
mary concept monophyly. That is, a secondary, operational concept, the
discovery of synapomorphy, permits continual re-evaluation of monophy-
ly of groups and our theory of descent represented through our classifica-
tion. This is all done within the context of a theory that there is a history of
descent, that characters are modified and inherited through this descent,
and that pattern and process is recoverable. Here we have a primary con-
ceptual basis for the type of supraspecific taxa that we wish to recognize in
classifications. This concept is necessary and sufficient in our search for
them. The concept of monophyly, like the ESC, is applicable but is in no
way operational. Secondary concepts for both species and supraspecific
categories are requisite in our discovery of species and supraspecific
groupings, respectively.
19.10 REFERENCES
Avise, J.C. and Ball, R.M. Jr (1990) Principles of genealogical concordance in species
concepts and biological taxonomy, in Oxford Surveys in Evolutionary Biology (eds
D. Futuyma and J. Antonovics), Oxford University Press, Oxford, pp. 45-67.
Blackwelder, R.E. (1967) Taxonomy: A Text and Reference Book, John Wiley & Sons,
New York.
Cronquist, A. (1978) Once again, what is a species? in BioSystematics in Agriculture
(ed. L. V. Knutson), Allenheld Osmun, Montclair, New Jersey, pp. 3-20.
de Queiroz, K. and Donoghue, M.J. (1988) Phylogenetic systematics and the
species problem. Cladistics, 4, 317-38.
de Queiroz, K. and Donoghue, M.J. (1990) Phylogenetic systematics and species
revisited. Cladistics, 6, 83-90.
Dobzhansky, T. (1950) Mendelian populations and their evolution. The American
Naturalist, 74, 312-21.
Dobzhansky, T. (1970) Genetics of the Evolutionary Process, Columbia University
Press, New York.
Botanisk Tidskrift, 24, 333-428.
Eldredge, N. and Cracraft, J. (1980) Phylogenetic analysis and the Evolutionary Process,
Frost, D.R. and Hillis, D.M. (1990) Species concepts and practice: herpetological
applications. Herpetologica, 46, 87-104.
Frost, D.R. and Kluge, A.G. (1994) A consideration of epistemology in systematic
biology, with special reference to species. Cladistics, 10,259-94.
George, T.N. (1956) Biospecies, chronospecies and morphospecies, in The Species
Concept in Paleontology (ed. P.C. Sylvester-Bradley), Systematics Association,
London, pp. 123-37.
23, 536-44.
Ghiselin, M.T. (1984a) Narrow approaches to phylogeny: a review of nine books
on cladism, in Oxford Surveys in Evolutionary Biology (eds R. Dawkins and M.
Ridley), Oxford University Press, Oxford, pp. 209-22.
Ghiselin, M.T. (1984b) 'Definition/ 'Character/ and other equivocal terms.
Grant, V. (1981) Plant Speciation, 2nd edn, Columbia University Press, New York.
Hennig, W. (1950) Grundzuge einer Theorie der Phylogentischen Systematik, Aufbau
Verlag, Berlin.
Hennig, W. (1966) Phylogenetic Systematics, University Illinois Press, Urbana.
Hull, D.L. (1976) Are species really individuals? Systematic Zoology, 25,174-91.
Hull, D.L. (1997) The ideal species concept - and why we can't get it, in Species: the
Kornet, D. (1993) Permanent splits as speciation events: a formal reconstruction of
the internodal species concept. Journal of Theoretical Biology, 164,407-35.
References 423
Kornet, DJ. and McAllister, J.W. (1993) The composite species concept, in
Reconstructing Species: Demarcations in Genealogical Networks. Unpublished PhD
dissertation, Institute for Theoretical Biology, Rijksherbarium, Leiden, pp.
61-89.
Mayden, R.L., and Wood, R.M. (1995) Systematics, species concepts, and the evo-
lutionarily significant unit in biodiversity and conservation biology. American
Fisheries Society Symposium, 17, 58-113.
Mayr, E. (1940) Speciation phenomena in birds. The American Naturalist, 74,249-78.
Mayr, E. (1957) Species concepts and definitions, in The Species Problem (ed. E.
Mayr), American Association for Advancement of Science, Washington, DC,
pp. 1-22.
Massachusetts.
Mayr, E. (1970) Populations, Species and Evolution, Belknap Press of Harvard
University, Cambridge, Massachusetts.
Mayr, E. (1988) The why and how of species. Biology and Philosophy, 3,431^1.
Mayr, E. (1997) The biological species concept, in Species Concepts and Phylogenetic
Theory: A Debate (eds Q.D. Wheeler and R. Meier), Columbia University Press,
New York (in press).
Mayr, E. and Ashlock, P.D. (1991) Principles of Systematic Zoology, McGraw-Hill,
New York.
McKitrick, M.C. and Zink R.M. (1988) Species concepts in ornithology. The Condor,
90,1-14.
Meier, R. and Willmann, R. (1997) The Hennigian species concept, in Species
Concepts and Phylogenetic Theory: A Debate (eds Q.D. Wheeler and R. Meier),
Columbia University Press, New York (in press).
Minelli, A. (1993) Biological Systematics: The State of the Art, Chapman & Hall, New
York.
Mishler, B.D. and Donoghue M.J. (1982) Species concepts: a case for pluralism.
Nelson, G. and Platnick N.I. (1981) Systematics and Biogeography: Cladistics and
Nixon, K.C. and Wheeler Q.D. (1990) An amplification of the phylogenetic species
Paterson, H.E.H. (1993) Evolution and The Recognition Concept of Species, Collected
Writings ofH. E. H. Paterson (ed. S. F. McEvey), Johns Hopkins University Press,
Baltimore, Maryland.
Platnick, N.I. (1976) Are monotypic genera possible? Systematic Zoology, 26,198-9.
Regan, C.T. (1926) Organic Evolution. Report British Association for Advancement
of Science, 1925, pp. 75-86.
Ridley, M. (1989) The cladistic solution to the species problem. Biology and
Philosophy, 4,1-16.
Rosen, D.E. (1978) Vicariant patterns and historical explanation in biogeography.
Guatemala: Revisionary studies and comparative biogeography. Bulletin
American Museum of Natural History, 162,267-376.
Shull, G.H. (1923) The species concept from the point of view from the geneticist.
American Journal of Botany, 10,221-8.
Simpson, G.G. (1943) Criteria for genera, species, and subspecies in zoology and
paleontology. Annals New York Academy of Science, 44,145-78.
Simpson, G.G. (1951) The species concept. Evolution, 5,285-98.
New York.
Sneath, P.H.A. (1976) Phenetic taxonomy at the species level and above. Taxon, 25,
437-50.
Sokal, R.R. (1973) The species problem reconsidered. Systematic Zoology, 22,360-74.
Stuessy, T.F. (1990) Plant Taxonomy. The Systematic Evaluation of Comparative Data,
Templeton, A.T. (1989) The meaning of species and speciation: a genetic perspec-
tive, in Speciation and Its Consequences (eds D. Otte and J. A. Endler), Sinauer
Van Valen, L. (1976) Ecological species, multispecies, and oaks. Taxon, 25,233-9.
Waples, R.S. (1991). Pacific salmon, Oncorhynchus spp., and the definition of
'species' under the Endangered Species Act. Marine Fisheries Review, 53,11-22.
Waples, R.S. (1996) Evolutionary significant units and the conservation of biologi-
cal diversity under the Endangered Species Act. American Fisheries Society
Symposium, 17 (in press).
Warren, M.L. (1992) Variation of the spotted sunfish, Lepomis punctatus complex
(Centrarchidae): meristics, morphometrics, pigmentation and species limits.
Bulletin Alabama Museum of Natural History, 12,1-47.
Weber, H.E. (1981) Sonderb. Naturwiss. Ver., Hamburg, 4,1-229.
lem: a reply to de Queirox and Donoghue. Cladistics, 6, 77-81.
Wheeler, Q.D. and Platnick N.I. (1997) The argument for phylogenetic species, in
Species Concepts and Phylogenetic Theory: A Debate (eds Q.D. Wheeler and R.
Meier), Columbia University Press, New York (in press).
Wiley, E.O. (1978) The evolutionary species concept reconsidered. Systematic
Zoology, 27,17-26.
Wiley, E.O. (1981) Phylogenetics: The Theory and Practice of Phylogenetic Systematics,
John Wiley & Sons, New York.
Wiley, E.O. and Mayden R.L. (1985) Species and speciation in phylogenetic
Systematics, with examples from the North American fish fauna. Annals
Missouri Botanical Garden, 72, 596-635.
Wiley, E.O. and Mayden R.L. (1997) The evolutionary species concept, in Species
Concepts and Phylogenetic Theory: A Debate (eds Q.D. Wheeler and R. Meier),
Columbia University Press, New York (in press).
Wilkinson, M. (1990) A commentary on Ridley's cladistic solution to the species
problem. Biology and Philosophy, 5,433-46.
Willmann, R. (1985a) Die Art in Raum und Zeit, Paul Parey Verlag, Berlin.
Willmann, R. (1985b) Reproductive isolation and the limits of the species in time.
Cladistics, 2, 336-8.
Index
Page numbers appearing in bold refer to figures; page numbers appearing in

italic refer to tables.
Abida 319 see also Leafhopper

Abies 152,301 coryli 253
Abstract species classes 20 Alnus glutinosa 254
Acarospora 116,122 Alpheidae 203
alberti 123 Alpheus 203,207, 209,212, 216
immersus 143 formosus 209
isortoquensis 123 lottini 205,206,208,209,214
smaragdula 123 Alpina 176
undata 123 American marten 352
Acipes 317 Amphibious snails 315
Acropora 203 Amycolatopsis methanolica 47
Actinomycetes 40,41,43, 70, 72, 73, 74 Anopheles 274, 278, 280,283
Acytotype 277,279 albimanus 284
Adelgidae 300 culicifaecies 283
Aedes 277 dims complex 284
aegypti 276, 284 gambiae complex 276, 280,283, 284
triseriatus 284 Aphididae 300
Aeromonas 40 Aphidoidea 299
African elephants, see Loxodonta Aphids 294, 300
Agamospecies 1, 8-9,12,294-5, 303, Apiaceae 180
319 Apodemus
Agamospecies concept 389,400-1,411, flavicollis 344
417 sylvaticus 344
Agromyzidae 249 see also Mice
Ahnfeltiopsis 98 Apomictic microspecies 294,295
Ahnfeltiopsis devoniensis 86 Aporrectodea
Albinaria 310,314, 315 caliginosa caliginosa 311
see also Clausiliid snail caliginosa trapezoides 311
Algae 84,110 Aquatic algae 91
Allium 180 Arachnids 280, 310, 314-15, 318
Allolobophora molleri 315 Araneidae 310
Allozyme analysis 212, 345, 349, 279, Archaea 26, 63, 67, 71
302, 311, 314-16 Archaebacteria 66, 67
Alnetoidia Archipelago effect 317
alneti 252, 253, 254 Argiope aetherea 310
426 Index
Aristotelian classes 19, 20 Biotype 252-3,261,279
Armillaria 141,142 Birds-of-paradise 10, 331
cepistipes 153 Birds 341, 343-4, 351, 370
gallica 140 Biting insects 276
mellea 140,141,153,157-8 Blackflies 227
ostoyae 140 Blood-sucking insects 274-5, 279-80,
sinapina 153 282, 284
Arrhenotoky 293 Blue-green alga 84, 111, 115
Asclepiadaceae 174 Brown algae 91
Ascobolus immersus 143 Brown planthopper 253
Aspicilia 122 see also Nilaparvata lugens
Asplenium 177 Buellia pulverea 123
Auchenorrhyncha 257 Burkholderia vietnamiensis 47
Audouinella 88
see a/so Red algae Calamagrostis epigejos 266
Auricularia 153 Calling behaviour 256
Aus feus 303 Camaenid snails 315
Autapomorphic species concept 361 Campanula 180
Avrainvillea 88-9, 93 Canis lupus, see Wolf
few's 89 Capsella bursa-pastoris 175
see also Green algae Catillaria
albocincta 119
Baboons 347 pulverea 119
see also individual species Cecidomyiids 294
Bacillaceae 27 Centipedes 310, 312, 314,317-18
Bacillus 30, 32, 35-6, 40 see also individual genera
amylolyticus 37 Cepaea 310
circulans 29, 37 Cephalodia 115
lautus 37 Cerion 310
pabuli 37 columna 318
sphaericus 29, 37 dimidiatum 318
validus 37 fernandina 316
Bacteria 26 stevensoni 316
Bacterial species concept 26, 363 see also Land snails
Baeomyces rufus 119 Chaenotheca chlorella 115
Balaenoptera acutirostrata, see Minke Chalcid 264
whales Chara 86
Balsam wooly aphid 301 Chemosystematics 31, 32,45-7, 230
Bat bugs 275 Chemotype 124
Batrachospermum 88 Chiffchaff 4
see also Red algae Chimpanzee 348, 351
Behavioural species 327 see also individual species
Bellemerea 128 Chlamydomonas 96, 99
Bilimbia tornense 118 moewusii 95
Binomial system 3,279 reinhardtii 96
Biological species concept 5-9, 20, 29, Chlorocytus 265
85, 91, 96,129,161-3,172,178-81, Chondrina 319
200-1,204-5, 209, 212, 214-16,250, Chondrus crispus 91
264-5, 280,285, 292, 296, 316, 320, see also Red algae
326-9, 331-3, 335, 336, 344, 349, Chromonema 233
350, 352-3,361, 362-3, 365-72, see also Heterorhabditis
374-6, 390, 392-3, 395-401,403^, Chromosome species 346
408, 410, 416-17,419 Chrysophyte 87
Index 427
Cicadas 257 Costaria costata 99
Cicadellidae 248, 249, 251, 257 Crataegus 256
Cicadoidea 257 Cristilabrum
Cladina 127 grossum 315
Cladistic methods 8 monodon 315
Cladistic species concept 391, 396-7, primum 315
407, 416-17 Crocidura
Cladocerans 294 russula 344
Cladonia 118,127 suaveolens 344
chlorophaea 125 see also Shrews
furcata 113 Ctenomys, see Tuco-tucos
mitts 128 Culexpipiens 284
rangiferina 127,128 Culicidae 276
squamosa 122,127,128 Culicoides variipennis 284
Cladophora 90,92,97 Curculionids 294
see also Green algae Cyanobacteria, see Blue-green alga
Clarkia 180 Cylindroiulus 317
Classical species concept 402 gemellus 317
Clausiliid snail 315 julipes 317
Clethrionomys 348 lundblandi 317
glareolus skomerensis 348, 351 madeirae 317
Closterium ehrenbergii 95, 96 Cynipid gall 294
Clostridium 35 Cytophylogenies 277
Clustering techniques 360 Cytospecies 277
Cochlicopa
lubrica 316 Delphacidae 257
lubricella 316 Delphinus 348
nitens 316 capensis or tropicalis 348
repentina 316 de/pfes 348
see a/so Land snails Dendriscocaulon 116
Cohesion species concept 361, 363, Dendrobaena octaedra 319
369, 373, 375-6, 392-3, 396 Dengue viruses 276
Coleoptera 250, 257 Dermacenfor
Collembola 294 andersoni 312
Collybia dryophila 143 tropicalis 348
Columella 319 variabilis 312
Common dolphin 348 see a/so Ticks
see also individual species Desmids 90
Common shrew 345 Desmonomata 319
Compositae 299 Diadema 210
Composite species concept 392-4, Diagnosable and monophyly version
405-7, 416-17 407
Conferva 86 Diagnosable version species concept
Coniophora puteana 143 400, 403, 405, 407
Coprinus 143,153 Diagnostic species concept 361-2, 364,
psychromorbidus 157,158 365, 369-70, 373-6
Corallorhiza Diatom 84,88, 90, 93^, 99,101
maculata 182,183,184-5 Dictyochloropsis splendida 115
mertensiana 182-3 Dimensional species concept 365
Corals 201, 212, 213, 214 Diplopods 312
Corollites 201,202 Diploschistes muscorum 118
Corylus avellana 254 Diptera 250, 257, 274, 276-7
Cosmarium botrytis 95 Dog 342
428 Index
Dolichoiulus 317 Evolutionary lineages 12
Dolphins 342 Evolutionary significant unit 395-7
Dorylaimida 225,226 Evolutionary species concept 361-2,
Dorylaimoidea 226 364-5, 370,375-6, 392-3,395-7,
Dryopteris 177 410,, 417-19,421
Dutch elm disease 143 Exdrolana 214
see also Ophiostoma ulmi braziliensis 212
Dysdera 317-18
Dysderidae 317 Fabaceae 180
Festuca 158
Earthworms 311, 319 Flavabacterium 35, 36
Ecological species concept 392, 394-6, Fomitopsis pinicola 153
416-17 Fridericia
Edwardsiana 248,251 galba 319
Eiseniella tetraedra 319 striata 319
Electrophoresis 30,203,250, 255, 265, Fucus devoniensis 86
279, 284, 346 Fusarium 155
Elephants 347, 351 oxysporum 155
Enchenopa binotata 258 Fuscidea 128
English elm 147
Enterobacteriaceae 27,37 Galleria mellonella, see Wax moth
Enteromorpha 90 Gasterosteus aculeatus,see Three spined
see also Green algae stickleback
Entomopathogenic nematodes 225, Gastropod 310, 311
238, 240 Gelidium 100
Eophila tellinii 317 attenuatum 100
Erinaceus kantvilas 121
concolor 351 see also Red algae
europaeus 351 Genealogical concordance concept
Escherichia coli 37, 68 397-8,404,416-17
Ethological-ecological species concept Genetalial characters 250-1
295 Genetic species 327
Ethospecies 312 Genetic species concept 389, 398-400,
Eubacteria 66, 67 403^, 419
Eubalaena 348 Genomic species 36, 41
australis 348-9,351 Genomovars 48
glacialis 348-9, 351 Genotypic cluster definition 400, 404,
japonica or sieboldi 348 408, 411,417
Euchlaena 197 Genotypic species 327
Eudorina 95 Geographical parthenogenesis 300
Euhadra 312 Geographical subspecies 301
Eukarya 26 Geophilomorph centipedes 314
Eulophidae 265 Geum 177
Euphila tellinii 317 Gibberella 156
Eupolybothrus fasciatus 314 fujikoroi 143
European bats 345 Gibbons 347
see also individual species see also individual species
European white elm 147 Cilia 175
Eurytoma 265 inconspicua 175
pollux 266 transmontana 178
Eurytomidae 264-5 Golden age of plant taxonomy 86
see also Tetramesa Gonium pectorale 95
Evernia 127 Gophers 342, 347
Index 429
Gordona 41 flgz'/z's 347
amarae 46 tar 347
see also Nocardia amarae muelleri 347
Gorilla 348 pileatus 347
Gossypium 184 Hymenelia 116
Gracilaria 100 Hymenoptera 250, 257, 265, 294
pacifica 100 Hymenoptera parasitica 263
robusfa 100
verrucosa 100 /ken's 175
see a/so Red algae Insect parasitic nematode 224
Gracilariopsis 100 Internodal species 12
Gram-negative bacteria 30 Internodal species concept 392-4, 405,
see a/so individual genera 416, 419, 421
Gram-positive bacteria 30 lonaspis 116
see also Bacillus Ishyropsalididae 312
Graphis scripta 128 Isopods 317-18
Green algae 90-3, 95-6, 98-9 see also individual species
see also individual genera Isozyme 182,276,283-4
Gymnogongrus 98 Italian earthworm 316
devoniensis 98
Julidan millipedes 313
Haematophagous insects 274 Jumping spiders 312
Haemophilus 30
Hardy-Weinburg Equilibrium 277, 283 Karyological 345-51, 353
Hares 342
Hedgehogs 351 Lamyctes 318
see also individual species Land snails 310-12,314-16
Helianthus 177 see also individual species
Helianthus section Helianthus 184 Lauria 319
Hemiptera 250,257 Leafhopper 248, 252
Hennigian species concept 400-2, see also Alnetoidia alneti
416-17 Leafmining 249
Herbivores 257,263 Leaf monkeys 347
Heterobasidion 153-4 Lecanactis
annosum 152-3 abietina 114
Heterodera 231 subabietina 114
Heteroderidae 229,231 Lecanora 114
Heterorhabditidae 225, 232^, 237-40 conizaeoides 123
Heterorhabditis 232-4, 236-8 vinetorum 123
bacteriophora 236 Lecidea 110,114,128
Heuchera 177-8 Leersia 259, 261, 262
nivalis 177 hexandra 255, 259, 260
parvifolia 177 Legionella 40
Hieracium 176 Lepidoptera 256
Homogametic mating 259, 262 Lepilemur 347
Homoporus 265 Lepus, see Hares
Homoptera 257 Lichen 110, 111
Homo sapiens 388 Linnaean species concept 402
Hordeum 197 Lithobiomorph centipedes 313-14
Hormogaster 314 Lobaria 116
pretiosa 316 Longidoridae 225-6, 232, 238-9
samnitica 316 Longidoroides 226
Hylobates 347 Longidorus 226
430 Index
Loxodonta 347 Molecular systematic methods 7, 32,
africana 347 45-6, 62, 64, 66, 73-4, 97-100,130,
africana africana 351 142,148,151,156,158,160-1,180,
africana cyclotis 351 203-5, 212,230-2,237-8,240,250,
Lumpers 193,359 261, 278-9, 283,284-5, 302, 314-16,
Lutzomyia 278 348, 349, 353
yucumensis 283 Mole-rats 342, 346, 351
see also individual species
carbonelli 311 Moles 342
311 Monism 358, 366, 374-5, 384
Monophyletic species concept 361,
Macaronesian millipedes 317 363-5, 369-70, 373-6
see also individual genera Monophyly version species concept
Macrocheles 314 406-7
Madeirae 317 Montastraea annularis 202-3
Malaria 276 Montia fontana 175
Mains 256 Morphological-geographical method
Mammals 341-4, 347, 349, 352-3 178
Mandarina 318 Morphological species concept 4, 7,
Martes americana, see American marten 84-6, 87, 92,100,186,215,295, 327,
Mastomys 345 361-2, 389, 399,400, 402-4,408,
concha 345-6, 350 411, 416-17
natalensis 345,350 Morphometric analysis 150, 226, 227,
see also Mice 229, 234, 235, 239-40, 284, 301,
Medora 314 319
see also Land snails Morphospecies 100,130,173,181,186,
Meloidogyne 231 215, 216, 263, 264
arenaria 231 Morphotype 91,100, 263
incognita 232 Mosquitoes 283
Meloidogynidae 229,231 Mus 345
Membracidae 258 musculus 349, 353
Menegazzia 120, 121 musculus domesticus 351
eperforata 120,121 musculus musculus 351
nothofagi 120,121 see also Mice
prototypica 120, 121 Mustela 344
Mice 344, 345, 351 Mycobacterium 32
Micrasterias thomasiana 95 Myotis
Microalgae 88-90 brandti 345
Micromalthns 294 mystacinus 345
Microspecies 175, 300 Myriapod 310-16
Microthamnion 99 Myzus
denigrata 114 ascalonicus 299
Microthrix 74 dianthicola 299
parvicella 74 persicae 299
Microtus
arvalis 345,348 Nannospalax 346
arvalis orcadensis 348, 351 ehrenbergi 346, 351
arvalis westrae 348 see also Mole-rats
rossiaemeridionalis 345 Nectria 156
Millipedes 310-11, 313, 317, 318 haematococca 143
see a/so individual genera Neisseria 30
Minke whales 349 gonorrhoea 64
Mites 318 meningitidis 64
Index 431
Nemasoma varicorne 318 Oribated mites 314-15
see also Millipedes Orkney voles 348
Nemastomatidae 312 see also individual species
Nematoda 222,225 Ornithology 325, 326
Neoaplectana 233, 234, 236 Orthoptera 257
see also Steinernema Oryza 258
Neoscona 310 sativa 253
penicillipes 311 Otiorrhynchus scarber 300
subfusca 310
triangula 311 Paenibacillus validus 47
Neosteinernema 233-4 Pagodulina 319
Neurospora 143,148 Paleontological species concept 403—4
Nilaparvata 258,263 Paleospecies 12
bakeri 260, 261, 262, 263, 265 Paleospecies concept 410
lugens 253, 254, 255, 258, 259, 260, Pan 348-9
261, 262 paniscus 348, 351
maeander 263 troglodytes 348,351
muiri 263 Pandorina morum 96
Nocardia Papio 347
amarae 41 hamadryas 347
see also Gordona amarae ursinus 347
pinensis 41, 46 Paradisaeidae, see Birds-of-paradise
pseudobrasiliensis 47 Paralongidorus 226, 229-30
Nomenspecies 29, 48 Parapatric crytic taxa 346
Non-aquatic mites 310 Paraphysomonas 87
Non-dimensional species concept 400, Parasitic Hymenoptera 295
403^, 408, 417 Parasitoids 257,264,265
Nostoc 115 Paraxiphidorus 226
Novisuccinea 315 Pardosa
chittenangoensis 315 proxima 312
o^fl/z's 315 vlijmi 312
Numerical analysis 31, 32, 34, 47, 63, Parmelia 127,113,118
173^ Parmeliopsis
ambigua 123
Obligate thelytoky 294 hyperopta 123
Octolasion Parthenogenetic species 301^4
cyaneum 319 Partula 310,316
tyrtaeum 319 suturalis 311, 314
Oecobiidae 317 tieniata 314
see also individual species Pauropods 312
Oecobius 317 Paxillus involutus 143
Oligochaetes 319 Pediobius 265
see also individual species eubius complex 265
Onchocerciasis 276 Peltigera 115-16,119
Opegrapha 114 aphthosa 120
Ophiostoma dadactyla 119
himal-ulmi 148 erumpens 119
novo-ulmi 145,148,162 hazslinszkyi 119
piceae 143 leucophlebia 117
ulmi 143,144,145,146,147,148,162 spwnfl 119
Orangutan 348 Penicillium 155,156
Orb-web spiders 310 Pentazonia 312
Orchidaceae 176 Phaeocystis 99
432 Index
Phaeophyscia orbicularis 126 Polyphasic species 50
Phenetic species concept 360-1, 364, Polypodium 177
366-7, 369, 373-6, 399-400, 402, vulgare 185
404, 411, 416-17 Polyporus abietinus 143
Phenetic system 7 Polysiphonia 94
Philodromidae 313 fibrillosa 94
Philodromus harveyi 94
rufus 313 violacea 94
vibrans 313 Polytene chromosome 277
Phlebotomus 278 Polytenized chromosome 279
Phlyctis argena 115 Polythetic classes 21
Pholcidae 317 Polythetic species concept 21, 400,404,
Phomopsis oblonga 143 408,411
Photorhabdus 232, 234 Polyxenus lagurus 319
Phylloscopus Pongo pygmaeus, see Orang utan
collybita, see Chiffchaff Populus 177
trochilus, see Willow warbler Porpoises 342
Phylogenetic species concept 8-9, 96, Potamogeton 176
129,161-3,172,181-5, 200-1, Potato blight fungus 149
204-5, 209-10, 212-16, 265-6, 303, Potyviridae 22
326-33, 334-7, 361, 362, 392, 394, Potyvirus 22
396-8, 405,407-8,411,416-17 Pratylenchidae 229
Physcia 123,126,127 Presbytis 347
tenella 127 Primates 341
Phytophagic species 252 Prokaryotes 38,47, 61, 62-6
Phytophagic varieties 252 Prunus 256
Phytophthora 154 persica 299
capsici 153 Pselaphognatha 312
cryptogea 151,152,154 Pseudendoclonium 92
drechsleri 151 see also Green algae
infestans 149 Pseudocyphellaria 116
megasperma 148,149,150,151,154, Pseudomonas 32, 36
160 Pseudoscorpions 312, 317-18
nicotianae 152 Pseudotrebouxia 126
sojae 154 Psilolechia 110
Picea, see Spruce leprosa 123
Pilayella littoralis, see Brown algae Psocoptera 294
Pinnipedes 342 Psoroma durietzii 118
Placopsis Pteromalidae 265
gelida 123 Puccinellia 174
lambii 123 nuttalliana 182,185
Planthopper 258 Pulse repetition frequency 258
Plasmodium 274 Pyrenula 128
Plecotus Pyrus 256
auritus 345
austriacus 345 Quasispecies 23
Pleurotus 162 Quercus 177
ostreatus 143
Pluralism 178, 346, 366, 374-5, 384 Ramalina
Poaceae 176 cuspidata 114,125
Pocillopora 205 siliquosa 114,125
Polyctenidae, see Bat bugs Ranunculus auricomus 176
Polyphasic approach, taxonomy 46-7, Rats 346
50, 64-5, 69 see also individual species
Index 433
Rattus Snow moulds 157
argentiventer 346 Solanum
rattus 346 ajanhuiri 196
tanezumi 346 curtilobum 196
tiomanicus 346 esculentum 194
Recognition species concept 5-6, 365, juzepczukii 196
393, 396, 408-9, 416-17 phureja 196
Red algae 88, 91, 98,100 sinense 194
see also individual species sparsipilum 197
Reproductive communities 8 stenotomum 196,197
Reproductive competition concept tuberosum 194,196,197
410, 417, 419 Solatopupa 314-15
Rhizobium mediterraneum 47 similis 315
Rhizocarpon geographicum 128 Solorina crocea 118
Rhizoctonia solani 143 Sorex
Rhymogona araneus 345, 350
cervina 316 coronatus 345, 350
silvatica 316 granarius 345, 350
Rice 253 Sorghum bicolor 194
Rosaceae 176,256 Spalax 346
Rubus 239 Spartina anglica 184
fruticosus 176 Species concept in virology 18
Specific mate recognition system 4, 6,
Sacciphantes 7-8,10, 255-7, 266, 275, 283,
abietis 300 311-12, 360-1, 363,368-71, 374-6,
segregis 300 396, 409
viridis 300 Spermophorides 317
Sfl/zx 176,177 Spiders 310, 312-13, 317-18
Salticidae 312 Spider webs 313
Samoana 315 Spirogyra 91
see also Land snails see also Green algae
Sandflies 275, 277 Splitters 193,359
Scales 294 Spotted alfalfa aphid 299
Scenedesmus 90 Spruce 300
Scorpions 312,318 Squirrels 341
Serfum 180 Staurothele 115
Sellaphora 95 Steganacarus 315
pupula 95 magnus 315
Senecio cambrensis 184—5 Steinernema 232-8
Serratia 44 fljQfmis 237
Sex pheromones 256, 283, 293 anomali 237
Sheep 342 glaseri 237
Shrews 344-5, 350 intermedia 237
Siddiqia. 230 kushidai 237
Simuliidae 276 scapterisci 237
Simulium 278-9,282-3 see a/so Neoaplectana
Skeletonema Steinernematidae 225, 232-4,237,
costatum 99 240
pseudocostatum 99 Stenogamy 284
Skomervole 348 Stephanomeria 180,184
see a/so individual species Stereoecaulon 115,118
Small ermine moths 256 Sterrnorrynchous Homoptera 294
Snails 310, 312, 314, 318 see a/so Aphids
Snapping shrimp 201, 203, 205,212 Stichococcus 115
434 Index
Sticta 116 Treponema 74-5
canariensis 117,118 Trichogramma 250
dufourii 118 Trichoniscus pusillus 318-19
felix 116 Triticum 197
Stigeoclonium 90 aestivum 194
Streptococcus 30 boeoticum 197
Streptomyces 27, 32 dicoccoides 194
Stylophora 205 monococcum 194,197
Successional species concept 400, Trogulidae 312
403^, 408,410, 417, 419, 421 Tsetse flies 275, 277
Superspecies 8, 300 Tsukamurella 41
Sycophila 265 inchoniensis 47
Symphylids 312 Tuco-tucos 347
Turner a ulmifolia 174
Tachinidae 263 Tylenchida 229
Talaromyces 156 Typhlocyba 248
Taraxacum 175-6 Typhlocybinae 248, 249, 251
Taxonomic species concept 172-3,176,
178,180,184-5, 400,402-4,408, Ulmus
411, 416-17 arctos 343
Taxospecies 31, 33, 48 laevis, see European white elm
Terrestrial oligochaetes 310 procera, see English elm
Territorial song 347 Ultrasounds 346
Tetragnathidae 317 Ulva 86
Tetramesa 264, 265 Umbilicaria 114,125
calamagrostidis 266 Uniparental 292,410
Thalassius spinosissimus 311 Ursus arctos 343
Thelotrema 116 Usnea 119
Thelytoky 293,295,298, 299
Theory of reinforcement 11 Vallonia 319
Therioaphis trifolii, see Spotted alfalfa Vector competence 276
aphid Vectorial capacity 276
Thlaspi 175 Verticillium 155-6
Thomomys 347 albo-atrum 156
see also Gophers dahliae 156
Three spined stickleback 6 Vezdaea 114
Thrips 294 Vibrio 70
Thysanoptera 294 Voles 345, 351
Ticks 277, 312, 214 see also individual species
Tigriopus 214,216
californicus 212 Wax moth 237
Tokogenetic relationships 8 Weasels 344
Tolmiea menziesii 178 Whales 342, 348, 351
Toninia 128 see also individual species
leucophaeopsis 118 Whiteflies 294
Tragopogon 184 Willow warbler 4
Trama 298 Wolf 352
Trebouxia 92,115-16,127 Wolf spiders 311, 312
irregularis 118 see also individual species
showmanii 118 Woodlouse 310, 319
see also Green algae see also individual species
Treehoppers 258
see also individual species Xanthomonad 47
Trent epohlia 115-16 Xanthomonas 36, 37
Index 435
Xanthoria 123,126 rivesi 230
parietina 126,127
Xenorhabdus 232, 233 Yellow clover aphid 299
Xiphidorus 226 Yellow fever 276
Xiphinema 226-7,232,239 Yponomeuta 256-7
americanum-group 226-30,232, 239, malinellus 256
240 padellus 256
brasiliense 226
ensiculiferum 226 Zea mays 197
z'ndoc 230 Zefl (or Euchlaena) mexicana 197
radicicola 226
Systematics Association Publications
1. Bibliography of key works for the identification of the British fauna and
flora, 3rd edition (1967)f
Edited by G.J. Kerrich, R.D. Meikle and N. Tebble
2. Function and taxonomic importance (1959)+
Edited by A.]. Cain
3. The species concept in palaeontology (1956)+
Edited by P.C. Sylvester-Bradley
4. Taxonomy and geography (1962)f
Edited by D. Nichols
5. Speciation in the sea (1963)+
Edited by J.P. Harding and N. Tebble
6. Phenetic and phylogenetic classification (1964)+
Edited by V.H. Hey wood and J. McNeill
7. Aspects of Tethyan biogeography (1967)f
Edited by C.G. Adams and D.V. Ager
8. The soil ecosystem (1969)+
Edited by H. Sheals
9. Organisms and continents through time (1973)"1"
Edited by N.F. Hughes
10. Cladistics: a practical course in systematics (1992)
P.L. Forey, C.J. Humphries, I.J. Kitching, R.W. Scotland, D.J. Siebert and DM.
Williams
tPublished by the Association (out of print)
Systematics Association Special Volumes

1. The new systematics (1940)
Edited by J.S. Huxley (reprinted 1971)
2. Chemotaxonomy and serotaxonomy (1968)*
Edited by J.G. Hawkes
3. Data processing biology and geology (1971)*
Edited by J.LCutbill
4. Scanning electron microscopy (1971)*
Edited by V.H. Hey wood
Out of print
5. Taxonomy and ecology (1973)*
Edited by V.H. Hey wood
6. The changing flora and fauna of Britain (1974)*
Edited by D.L. Hawksworth
Out of print
Systematics Association Publications 437
7. Biological identification with computers (1975)*
Edited by R.J. Pankhurst
8. Lichenology: progress and problems (1976)*
Edited by D.H. Brown, D.L. Hawksworth and R.H. Bailey
9. Key works to the fauna and flora of the British Isles and north-western
Europe, 4th edition (1978)*
Edited by G.J. Kerrich, D.L. Hawksworth and R.W. Sims
10. Modern approaches to the taxonomy of red and brown algae (1978)
Edited by D.E.G. Irvine andJ.H. Price
11. Biology and systematics of colonial organisms (1979)*
Edited by G. Larwood and B.R. Rosen
12. The origin of major invertebrate groups (1979)*
Edited by M.R. House
13. Advances in bryozoology (1979)*
Edited by G.P. Larwood and M.B. Abbott
14. Bryophyte systematics (1979)*
Edited by G.C.S. Clarke andJ.G. Duckett
15. The terrestrial environment and the origin of land vertebrates (1980)
Edited by A.L. Pachen
16. Chemosystematics: principles and practice (1980)*
Edited by F.A. Bisby, J.G. Vaughan and C.A. Wright
17. The shore environment: methods and ecosystems (2 volumes) (1980)*
Edited by J.H. Price, D.E.G. Irvine and W.F. Farnham
18. The Ammonoidea (1981)*
Edited by M.R. House and J.R. Senior
19. Biosystematics of social insects (1981)*
Edited by P.E. House and J.-L. Clement
20. Genome evolution (1982)*
Edited by G.A. Dover and R.B. Flavell
21. Problems of phylogenetic reconstruction (1982)*
Edited by K.A. Joysey and A.E. Friday
22. Concepts in nematode systematics (1983)*
Edited by A.R. Stone, H.M. Platt and L.F. Khalil
23. Evolution, time and space: the emergence of the biosphere (1983)*
Edited by R.W. Sims, J.H. Price and P.E.S. Whalley
24. Protein polymorphism: adaptive and taxonomic significance (1983)*
Edited by G.S. Oxford and D. Rollinson
25. Current concepts in plant taxonomy (1983)*
Edited by V.H. Hey wood and DM. Moore
26. Databases in systematics (1984)*
Edited by R. Allkin and F.A. Bisby
27. Systematics of the green algae (1984)*
Edited by D.E.G. Irvine and D.M. John
28. The origins and relationships of lower invertebrates (1985)$
Edited by S. Conway Morris, J.D. George, R. Gibson and H.M. Platt
29. Infraspecific classification of wild and cultivated plants (1986)$
Edited by B.T. Styles
438 Systematics Association Publications
30. Biomineralization in lower plants and animals (1986)$
Edited by B.S.C. Leadbeater and R. Riding
31. Systematic and taxonomic approaches in palaeobotany (1986)$
Edited by R.A. Spicer and B.A. Thomas
32. Coevolution and systematics (1986):}:
Edited by A.R. Stone and D.L. Hawksworth
33. Key works to the fauna and flora of the British Isles and north-western
Europe, 5th edition (1988)*
Edited by R.W. Sims, P. Freeman and D.L. Hawksworth
34. Extinction and survival in the fossil record (1988)*
Edited by G.P. Larwood
35. The phylogeny and classification of the tetrapods (2 volumes) (1988)*
Edited by M.J. Benton
36. Prospects in systematics (1988)*
Edited by D.L. Hawksworth
37. Biosystematics of haematophagous insects (1988)*
Edited by M.W. Service
38. The chromophyte algae: problems and perspective (1989)*
Edited J.C. Green, B.S.C. Leadbeater and W.L. Diver
39. Electrophoretic studies on agricultural pests (1989)*
Edited by Hugh D. Loxdale and J. den Hollander
40. Evolution, systematics, and fossil history of the Hamamelidae (2 volumes)
(1989)*
Edited by Peter R. Crane and Stephen Blackmore
41. Scanning electron microscopy in taxonomy and functional morphology
(1990)*
Edited by D. Claugher
42. Major evolutionary radiations (1990)*
Edited by P.D. Taylor and G.P. Larwood
43. Tropical lichens: their systematics, conservation and ecology (1991)*
Edited by G.J. Galloway
44. Pollen and spores: patterns of diversification (1991)*
Edited by S. Blackmore and S.H. Barnes
45. The biology of free-living heterotrophic flagellates (1991)*
Edited by D.J. Patterson and J. Larsen
46. Plant-animal interactions in the marine benthos (1992)*
Edited by D.M. Johns, S.J. Hawkins and J.H. Price
47. The Ammonoidea: environment, ecology and evolutionary change (1993)*
Edited by M.R. House
48. Designs for a global plant species information system*
Edited by F.A. Bisby, G.F. Russell and R.J. Pankhurst
49. Plant galls: organisms, interactions, populations*
Edited by MichSle A.]. Williams
50. Systematics and conservation evaluation^
Edited by P.L. Forey, C.J. Humphries and R.l. Vane-Wright
51. The Haptophyte algae*
Edited by J.C. Green and B.S.C. Leadbeater
Systematics Association Publications 439
52. Models in phylogeny reconstruction^
Edited by R. Scotland, D.J. Siebert and DM. Williams
53. The Ecology of Agricultural Pests: biochemical approaches**
Edited by W.O.C. Symondson andJ.E. Liddell
54. Species: the units of diversity**
Edited by M.F. Claridge, H.A. Dawah and M.R. Wilson
'Published by Academic Press for the Systematics Association

f
Published by the Palaeontological Association in conjunction with Systematics Association
^Published by the Oxford University Press for the Systematics Association
"'Published by Chapman & Hall for the Systematics Association
The Systematics Association Special Volume Series 54
Species
The units of biodiversity
Edited by M.F. Claridge, H.A. Dawah and M.R. Wilson
The number of species of organisms inhabiting our planet has, in recent

years, taken on great significance, as the importance of biological diversity to
the survival of mankind and the sustainable use of our natural resources has
been realized. It has therefore become vital that the nature and definition of
'species' should be clarified.
This Systematics Association Special Volume brings together an international
team of experts to give an account of their ideas on the species concept for
the particular group of organisms of their interest. The contributors represent
expertise on a wide diversity of living organisms, and chapters are included
on viruses through to mammals, in addition to general reviews of species
concepts.
This important book should be read by all biologists and is of special interest
to those working in the fields of taxonomy, Systematics and biodiversity, ecology
and conservation.
Michael Claridge is Professor of Entomology and Hassan Dawah is an

Insect Taxonomist and Ecologist in the School of Pure and Applied Biology,
University of Wales, College of Cardiff, UK. Michael Wilson is the Assistant
Keeper of Zoology, National Museum & Gallery of Wales, Cardiff, UK
Also available from Chapman & Hall

Global Biodiversity
Status of the Earth's living resources
World Conservation Monitoring Centre
Hardback (0 412 47240 6), 586 pages
Biodiversity
Measurement and estimation
Edited by D.L.'Hawksworth
Paperback (0 412 75220 4), 140 pages
Biological Sytematics
The state of the art
A. Minelli
ISBN 0 - 4 1 2 - 6 3 1 2 0 - 2
Paperback (0 412 62620 9), 388 pages
CHAPMAN & HALL 9"780412"631207">

London • Weinheim • New York • Tokyo • Melbourne • Madras

(The Systematics Association Special Volume Series) M.F. Claridge, A.H. Dawah, M.R. Wilson - Species - The Units of Biodiversity-Springer (1997)

Uploaded by

Copyright:

Available Formats

(The Systematics Association Special Volume Series) M.F. Claridge, A.H. Dawah, M.R. Wilson - Species - The Units of Biodiversity-Springer (1997)

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

(The Systematics Association Special Volume Series) M.F. Claridge, A.H. Dawah, M.R. Wilson - Species - The Units of Biodiversity-Springer (1997)

Uploaded by

Copyright:

Available Formats

What is the Systematics Association and what does it do?

What is the Systematics Association and what does it do?

What publications has the Systematics Association produced?

What publications has the Systematics Association produced?

Secies

M.R Claridge, HA Dawah

CHAPMAN & HALL Systerriatics

JOIN US ON THE INTERNET VIA WWW, GOPHER, FTP OR EMAIL

Department of Botany, The Natural History Museum, Cromwell Road,

The Systematics Association provides a forum for discussing systematic

Forthcoming titles in the series:

CHAPMAN & HALL

Chapman & Hall, 2-6 Boundary Row, London SE1 8HN, UK

First edition 1997

Library of Congress Catalog Number: 96-70872

^-^ Printed on acid-free text paper, manufactured in accordance with

1 Practical approaches to species concepts for living

3 Towards a practical species concept for cultivable bacteria 25

4 Species in practice: exploring uncultured prokaryote

5 Species problems in eukaryotic algae: a modern perspective 83

7 Fungal species in practice: identifying species units in fungi 135

8 Practical aspects of the species concept in plants 171

9 Cultivated plant diversity and taxonomy 191

10 Species of marine invertebrates: a comparison of the

11 Nematode species: concepts and identification strategies

12 Species in insect herbivores and parasitoids - sibling

13 The species concept in blood-sucking vectors of human

15 The species in terrestrial non-insect invertebrates

16 Species concepts in systematics and conservation biology

17 The species in mammals 341

18 The ideal species concept - and why we can't get it 357

19 A hierarchy of species concepts: the denouement in the

The idea of bringing together a wide diversity of specialists on different

1.2 LINNAEUS AND EARLY SPECIES CONCEPTS

Protoza Fungi Bacteria Viruses Insects

Figure 1.1 Estimates of proportions of species of major groups of organisms con-

1.3 BIOLOGICAL SPECIES CONCEPTS

Figure 1.2 Diagrammatic representation of the exchange of signals and respon-

distinct SMRSs which result in the reproductive isolation observed

With the development and widespread acceptance of cladistic methods by

• Phylogenetic species can be applied to non-sexual and parthenogenetic

Diagnosably distinct allopatric populations will therefore be regarded as

1.5 SPECIES CONCEPTS AND SPECIATION: PATTERN AND

1.6 SPECIES IN PRACTICE - AN EVOLUTIONARY SYNTHESIS

2.3 CONTINUITY VERSUS DISCONTINUITY AND THE PROBLEM

2.4 SPECIES AS POLYTHETIC CLASSES

2.5 SPECIES FUZZINESS

2.6 SPECIES OR QUASISPECIES

3.2 EARLY SPECIES CONCEPTS

Genus Bergey's Manual Approved lists

1974 1980 1995

A. Medically important bacteria

Cowan (1968) recognized the subjective nature of the traditional

3.3 THE NEW BACTERIAL SYSTEMATICS

3.3.1 Numerical taxonomy

Figure 3.1 Major steps in numerical classification and identification.

3.3.2 Molecular systematics

(a) DNA base composition

(b)DNA: DNA pairing