Biology, Ecology and Behaviour of Aphidophagous Insects - PDF Room

Ecology of Aphidophaga:
Biology, Ecology and Behaviour of

Aphidophagous Insects
Proceedings of the 8th International Symposium
Ponta Delgada 1-6 September 2002
António O. Soares, Maria A. Ventura,

Vasco Garcia & Jean-Louis Hemptinne (Eds)
Arquipélago - Life and Marine Sciences

Supplement 5
2003
ARQUIPÉLAGO
Life and Marine Sciences. Bulletin of the University of the Azores.
PUBLISHER
University of the Azores, Rua Mãe de Deus, Apartado 1422
PT - 9501-801 Ponta Delgada, Azores, Portugal.
GUEST EDITORS FOR SUPPLEMENT 5

António O. Soares, Maria A. Ventura, Vasco Garcia & Jean-Louis Hemptinne
INFORMATION ON THE INTERNET

Abstracts of all papers as well as instructions to authors are available on the Internet at:
http://www.horta.uac.pt/editions/arquipelago.html
SCIENTIFIC COMMITTEE FOR THIS SUPPLEMENT

Professor Ivo Hodek Professor Edward W. Evans
Professor Anthony F. G. Dixon Professor Daniel Coderre
Professor Manfred Mackauer Professor Jean-Louis Hemptinne
Professor Pavel Kindlmann Professor Hironori Yasuda
EDITORIAL SECRETARIAT
Helen Rost Martins, João M. Gonçalves, Rogério R. Ferraz, Horta.
Indexed in:
Aquatic Sciences and Fisheries Abstracts (ASFA), BIOSIS, Current Awareness in Biological Sciences
(CABS), Marine Science Content Tables (MSCT), Zoological Record
Biology, Ecology and Behaviour of Aphidophagous Insects
For purpose of citation, this supplement should be cited as follows:
A.O. SOARES, M.A. VENTURA, V. GARCIA & J.-L. HEMPTINNE (Eds) 2003. Proceedings of the 8th International
Symposium on Ecology of Aphidophaga: Biology, Ecology and Behaviour of Aphidophagous Insects.
Arquipélago. Life and Marine Sciences. Supplement 5: x + 112 pp.
i
8th International Symposium on Ecology of Aphidophaga
University of the Azores, Ponta Delgada, 1-6 September 2002
ORGANIZATION
SPONSORS
ii
TABLE OF CONTENTS
PAG.
PREFACE v
LIST OF PARTICIPANTS vi
GROUP PHOTO ix
DIXON, A.F.G. & J.-L. HEMPTINNE

Ladybirds and the biological control of aphid populations. 1
WIESSER, W.W.
Additive effects of pea aphid natural enemies despite intraguild predation. 11
BUENO, V.H.P., A.B. CARNEVALE & M.V.SAMPAIO

Host preference of Lysiphlebus testaceipes (Cresson) (Hymenoptera: Aphidiidae) for Myzus persicae
(Sulzer) and Aphis gossypii Glover (Hemiptera: Aphididae). 17
SCHELT, J.- V. & F. WÄCKERS.

The biological control of Aulacorthum solani (Kaltenbach) (Homoptera: Aphididae) in greenhouse
grown pepper; research on a tri-trophic system. 21
MAGRO, A., J-L. HEMPTINNE, A. NAVARRE & A.F.G.DIXON

Comparison of the reproductive investment in Coccidophagous and Aphidophagous ladybirds
(Coleoptera: Coccinellidae). 29
HEMPTINNE, J.-L., A.F.G. DIXON & E. WYSS

Biological control of the rosy apple aphid, Dysaphis plantaginea (Passerini) (Homoptera: Aphididae):
learning from the ecology of ladybird beetles. 33
SERPA, L., H. SCHANDERL, C. BRITO & A. O. SOARES

Fitness of five phenotypes of Harmonia axyridis Pallas (Coleoptera: Coccinellidae). 43
SOARES, A.O., D. CODERRE & H. SCHANDERL

Influence of prey quality on the reproductive capacity of two phenotypes of Harmonia axyridis Pallas
(Coleoptera: Coccinellidae). 51
JAROŠÍK, V., J. POLECHOVÁ, A.F.G. DIXON & A. HONċK

Developmental isomorphy in ladybirds (Coleoptera: Coccinellidae). 55
KINDLMANN, P. & A.F.G. DIXON.

Past and future models of predator-prey population dynamics: the importance of generation time ratio. 65
FRÉCHETTE, B., C. ALAUZET & J.-L. HEMPTINNE

Oviposition behaviour of the two-spot ladybird beetle Adalia bipunctata (L.) (Coleoptera:
Coccinellidae) on plants with conspecific larval tracks. 73
GODEAU, J.-F., J.-L. HEMPTINNE & J.-C. VERHAEGHE

Ant trail: a highway for Coccinella magnifica Redtenbacher (Coleoptera: Coccinellidae). 79
iii
RģŽIýKA, Z. & R. ZEMEK

Effects of conspecific and heterospecific larval tracks on mobility and searching patterns of Cycloneda
limbifer Say (Coleoptera: Coccinellidae) females. 85
BASKY, Z.
Predators and parasitoids on different cereal aphid species under caged and no caged conditions in
Hungary. 95
FÉLIX, S. & A. O. SOARES

Intraguild predation among the aphidophagous ladybird beetles Harmonia axyridis Pallas and
Coccinella undecimpunctata L. (Coleoptera: Coccinellidae): characterization of the direction and
symmetry. 103
SOARES, A.O., R. B. ELIAS & A. RAIMUNDO

Approach to the knowledge of Coccinellidade (Coleoptera) species diversity of Madeira and Porto
Santo Islands. 107
iv
PREFACE
Aphids and their natural enemies were central stage from 1st to 6th of September, 2002, at the
University of the Azores in Ponta Delgada. Aphids were recorded as major pests early in the XXth
Century (GAUMONT 1977) and continue to threaten field and greenhouse crops. As a consequence,
pesticides are used regularly and intensively, which hampers the move towards sustainable agriculture.
In theory, biological control of aphids is an alternative to chemical control but the record of biological
control is not good (DIXON 2000). This does not mean that biological control should be abandoned. On
the contrary, we should attempt to improve it, and this is the main objective of the IOBC Working
Group “Ecology of aphidophaga”.
Traditionally the majority of the communications presented at the meetings of this Working Group
are on ladybird beetles. This was also the case at this meeting but the prominence of ladybirds was not a
deliberate attempt to restrict the scientific scope of the meeting. History partly explains the domination
of ladybirds. The outstanding success of Rodolia cardinalis in California many years ago still influences
these conferences. This is not the only explanation. Ladybirds are a good model organism for
improving our understanding of the role of natural enemies in regulating herbivore populations.
Parasitoids have, and continue to be used as models for studying predator prey dynamics. However, the
biology of parasitoids differs fundamentally from that of predators (DIXON 2000) and it is therefore
unlikely that studies on parasitoids will shed light on all the theoretical and practical aspects of the
interactions between prey and natural enemies. Studies on ladybirds complement those on parasitoids
and give a better understanding of why the biological control of aphids sometimes fails. This was
central to the interesting communications on predator-prey models, guild structure, fundamental
biology and applications presented at the meeting.
For practitioners confronted with aphid outbreaks, however, theoretical studies may seem a waste of
time. In the face of heavy economic losses immediate action often appears to be the only solution.
KAREIVA (1996) has highlighted the risks of a “hit and miss” approach to biological control. However,
a “search and wait” strategy, in which all efforts are directed to research while farmers sit and wait for a
solution, is not practical. The future is a trade-off between research and practice. In theory there is an
infinity of possible trade-offs, but which is the best? In Ponta Delgada the Scientific Committee decided
that the Working Group should provide opportunities for practitioners of biological control and
scientists to meet and discuss mutual problems. Through exchange of ideas and discussion of field
results the right trade off will be approached and the gap between academics and practitioners of
biological control progressively reduced.
The meeting in Ponta Delgada was fruitful in both scientific and strategic terms. I thank the
University of the Azores for hosting the congress and the Local Committee for all its efforts and
dedication. Muito obrigado pela sua hospitalidade!
The next meeting will be held in Japan at Yamagata University in September 2005. The first
announcement is already on our web site (http://www.bf.jcu.cz/tix/strita/aphidophaga/main.html).
REFERENCES
DIXON, A.F.G. 2000. Insect predator-prey dynamics. Ladybird beetles & biological control. Cambridge University
Press. 257 pp.
GAUMONT, L. 1977. L’importance économique des aphides (Texte inédit publié à l’occasion du centenaire de sa
naissance). Annales de Zoologie-Ecologie animale 9: 173-180.
KAREIVA, P. 1996. Contribution of ecology to biocontrol. Ecology 77: 1964-1965.
v
LIST OF PARTICIPANTS
Basky, Zsuzsa Duelli, Peter

Plant Protection Institute of the Hungarian Swiss Federal Research Institute WSL,
Academy of Sciences, P.O.B. 102 HU-1525 Zuercherstr. 111, CH-8903 Birmensdorf/Zürich,
Budapest, Hungary. E-mail: h10433bas@ella.hu Switzerland. E-mail: peter.duelli@wsl.ch
Bianchi, Félix Evans Edward W.

Alterra Green World Research, Department of Biology, Utah State University,
Droevendaalsesteeg 3, P.O.Box 47, NL-6700 AA Logan, UTA 84322-5305, USA. E-mail:
Wageningen, Netherlands. E-mail: ewevans@biology.usu.edu
f.j.j.a.bianchi@alterra.wag-ur.nl Félix, Sandra
Boertjes, Bertin Department of Biology, University of the Azores,
Applied Plant Reseach, Linnaeuslaan 2a, NL- Rua da Mãe de Deus, Apartado 1422. PT-9501-
1431 JV Aalsmeer, Netherlands. E-mail: 801 Ponta Delgada, Azores, Portugal. E-mail:
b.c.boertjs@ppo.dlo.nl felixs@mail.telepac.pt
Fréchette, Bruno
Branquart, Etienne Université Paul Sabatier. Laboratoire d'écologie
Belgian Biodiversity Platform / CRNFB Avenue terrestre. UMR 5552. 118 route de Narbonne (4
Maréchal Juin, 23 BE-5030 Gembloux, Belgium. R 3). FR-31062 Toulouse cedex 04. France.
E-mail: E.Branquart@mrw.wallonie.be Email: frechette_bruno@yahoo.ca
Brito, Carlos. Garcia, Vasco
Department of Biology, University of the Azores, Department of Biology, University of the Azores,
Rua da Mãe de Deus, Apartado 1422. PT-9501- Rua da Mãe de Deus, Apartado 1422. PT-9501-
801 Ponta Delgada, Azores, Portugal. E-mail: 801 Ponta Delgada, Azores, Portugal. E-mail:
brito@notes.uac.pt vgarcia@notes.uac.pt
Bueno, Vanda H.P. Godeau, Jean-François
Department of Entomology, Federal University of Laboratoire De Zoologie (Université de Mons-
Lavras, P.O.Box 37, CEP-37200-000 Lavras, Hainaut) Avenue du Champs de Mars, 6 BE-7000
MG, Brazil. E-mail: vhpbueno@ufla.br Mons, Belgium. E-mail: jean-francois.godeau@
umh.ac.be
Coderre, Daniel
Département des Sciences Biologiques, Gonçalves, Ana M. S.
Université du Québec à Montréal, C.P. 8888 University Catholique de Louvan, Laboratoire de
Succ. Centre-ville, Montréal, Canada, H3C 3P8. Chimie des Interfaces, Croix du Sud 2 (bte 18)
E-mail: coderre@uqam.ca BE-1348 Louvan-la-Neuve, Belgium. E-mail:
dossantos@cifa.ucl.ac.be
Cunha, Regina.
Department of Biology, University of the Azores, Grassewitz, Tessa
Rua da Mãe de Deus, Apartado 1422. PT-9501- Harper Adams University College, Edgmond,
801 Ponta Delgada, Azores, Portugal. E-mail: Newport Shropshire, TF 10 8NB, UK. Email:
rcunha@notes.uac.pt tgrasswitz@harper-adams.ac.uk
Dixon, Anthony F.G. Hemptinne, Jean-Louis
School of Biological Sciences, University of East Ecole Nationale de Formation Agronomique, B.P.
Anglia, Norwich, NR4 7TJ, UK. E-mail: 87, FR-31326 Castanet-Tolosan, France. E-mail:
a.f.dixon@uea.ac.uk jean-louis.hemptinne@educagri.fr
vi
Hindayana, Dadan British-Colombia V5A 156, Canada. E-mail:

Institute of Plant Pests and Diseases, Bogor mackauer@sfu.ca
Agricultural University (IPB) Jl. Kamper,
Kampus IPB Darmaga Bogor, ID-16680 Bogor, Magro, Alexandra
Indonesia. E-mail: hindayana@ipb.ac.id Ecole Nationale de Formation Agronomique,
Jeune Equipe 000271JE1 Laboratoire d’Agro-
Hodek, Ivo écologie, BP 87, FR-31326 Castanet-Tolosan,
Academy of Sciences of Czech Republic, France. E-mail: alexandra.magro@educagri.fr
Branišovská 31, CZ-370 05 ýeské BudČjovice,
Czech Republic. E-mail: hodek@entu.cas.cz Martinkova, Zdenka
Research Institute of Crop Production, Drnovska
HonČk, Alois 507 CZ-16106 Praha 6-Ruzyne, Czech Republic.
Research Institute of Crop Production, Drnovská E-mail: martinkova@vurv.cz
507, CZ-161 06 Prague 6 - RuzynČ, Czech
Republic. E-mail honek@vurv.cz Martins, António P.
Department of Biology, University of the Azores,
Jarošík, Vojtech Rua da Mãe de Deus, Apartado 1422. PT-9501-
Department of Zoology, Charles University, 801 Ponta Delgada, Azores, Portugal. E-mail:
Viniþná 7, CZ-128 44 Prague 2, Czech Republic. martins@notes.uac.pt
E-mail jarosik@mbox.cesnet.cz
Ng, Sook M.
Kindlmann, Pavel Department of Zoology, University of Malaya,
Faculty of Biological Science, University of Kuala Lumpur, Malaysia.
South Bohemia, Branišovská 31, CZ-370 05
ýeské BudČjovice, Czech Republic. E-mail: Nedved, Oldrich
pavel@entu.cas.cz Institute of Entomology, University of South
Bohemia, Branišovská 31, CZ-370 05 ýeské
Kontodimas, Dimitris
BudČjovice, Czech Republic. E-mail: nedved@
Benaki Phytopathological Institue, 8 Delta Str.
bf.jcu.cz
GR-14561 Kifissia (Athenes), Greece. Email:
dckontodimas@hotmail.com Oliveira, Luisa.
Lourenço, Paula Department of Biology, University of the Azores,
Department of Biology, University of the Azores, Rua da Mãe de Deus, Apartado 1422. PT-9501-
Rua da Mãe de Deus, Apartado 1422. PT-9501- 801 Ponta Delgada, Azores, Portugal. E-mail:
801 Ponta Delgada, Azores, Portugal. E-mail: ziza@notes.uac.pt
lourenco@notes.uac.pt Resendes, Roberto
M’Rabet, Salima M. Department of Biology, University of the Azores,
University Catholique de Louvan, Laboratoire de Rua da Mãe de Deus, Apartado 1422. PT-9501-
Chimie des Interfaces, Croix du Sud 2 (bte 18), 801 Ponta Delgada, Azores, Portugal. E-mail:
BE-1348 Louvan-la-Neuve, Belgium. E-mail: rresendes@notes.uac.pt
machkourmrabet@cifa.ucl.ac.be
Robert, Pierre-Emmanuel
Machado, Ricardo Imperial College at Wye, University of London,
Department of Biology, University of the Azores, c/o MCR, Imperial College at Wye, Ashford, TN
Rua da Mãe de Deus, Apartado 1422. PT-9501- 25 5AH, UK. E-mail: Pierre-emmanuel.robert@
801 Ponta Delgada, Azores, Portugal. E-mail: ic.ac.uk
rmachado@linus.uac.pt
RĤžiþka, Zdenik
Mackauer, Manfred Institute of Entomology, Academy of Sciences,
Department of Biological Sciences, Simon Fraser Branišovská 31, CZ-370 05 ýeské BudČjovice,
University, 8888 University Drive, Burnaby, Czech Republic. E-mail: ruzicka@entu.cas.cz
vii
Sato, Satoru Yamagata JP-997-8555 Japan. E-mail:

United Graduate School of Agricultural Science, takizawa@tds1.tr.yamagata-u.ac.jp
Iwate University, Wakaba-cho 1-23 Tsuruoka
Yamagata JP-997-8555 Japan. E-mail: hyasuda@ Vanhaelen, Nicolas
tds1.tr.yamagata-u.ac.jp Agricultural University of Gembloux, Unité de
Zoologie Générale et Appliquée, 2 Passage des
Schelt, Jeroen van Déportés, BE-5030 Gembloux, Belgium. E-mail:
Koppert B.V., Veilingweg 17, NL-2650 AD vanhaelen.n@fsagx.ac.be
Berkel en Rodenrijs, The Netherlands;
Netherlands Institute of Ecology (NIOO-CTE), Ventura, Anunciação.
Boterhoeksestraat 22, NL-6666 GA Heteren, The Department of Biology, University of the Azores,
Netherlands. E-mail: jvschelt@koppert.nl Rua da Mãe de Deus, Apartado 1422. PT-9501-
Serpa, Luis 801 Ponta Delgada, Azores, Portugal. E-mail:
Department of Biology, University of the Azores, mateus@notes.uac.pt
Rua da Mãe de Deus, Apartado 1422. PT-9501-
801 Ponta Delgada, Azores, Portugal. E-mail: Werf, Wopke van der
lserpa@sapo.pt Group Crop & Weed Ecology, Wageningen
University, P.O.Box 430, NL-6700 AK
Soares, António O. Wageningen, Netherlands. E-mail: wopke.
Department of Biology, University of the Azores, vanderwerf@cwe.dpw.wau.nl
Rua da Mãe de Deus, Apartado 1422. PT-9501-
801 Ponta Delgada, Azores, Portugal. E-mail: Weisser, Wolfgang W.
onofre@notes.uac.pt Institute of Ecology, Friedrich-Schiller-
Stathas, George University, Dornburger Str. 159, DE-07743 Jena,
Benaki Phytopathological Institue, 8 Delta Str., Germany. E-mail: wolfgang.weisser@uni-jena.de
GR-14561 Kifissia (Athenes), Greece. E-mail:
georgestathas@hotmail.com Yasuda, Hironori
Department of Animal Ecology, Department of
Takizawa, Tadashi Agriculture, Yamagata University, 1-23,
United Graduate School of Agricultural Science, Tsuruoka, Yamagata JP-997-8555, Japan. E-mail:
Iwate University, Wakaba-cho 1-23 Tsuruoka hyasuda@tds1.tr.yamagata-u.ac.jp
viii
1 - António Martins 14 - Peter Duelli 27 - Tessa Grasswitz

2 - Vanda Bueno 15 - Felix Bianchi 28 - Bertin Boertjes
3 - Paula Lourenço 16 - Tadashi Takizawa 29 - Salima Machkour M’Rabet
4 - Jeroen van Schelt 17 - Jean-François Godeau 30 - Zsuzsa Basky
5 - Alois HonƟk 18 - Anthony F.G. Dixon 31 - Hironori Yasuda
6 - Zdenka Martinkova 19 - Wopke van der Werf 32 - Dadan Hindayana
7 - Vojtech Jarošík 20 - Satoru Sato 33 - George Stathas
8 - Anunciação Ventura 21 - Dimitris Kontodimas 34 - Voula Statha
9 - Oldrich Nedved 22 - António Onofre Soares 35 - Theodora June Dixon
10 - Wolfgang W. Weisser 23 - Bruno Fréchette 36 - Pierre-Emanuel Robert
11 - Sandra Félix 24 - Jean-Louis Hemptinne 37 - João Hemptinne
12 - Roberto Resendes 25 - Alexandra Magro 38 - Edward W. Evans
13 - Zdenek RĤžiþka 26 - Sook Ming Ng 39 - Pavel Kindlmann
ix
x
LADYBIRDS AND THE BIOLOGICAL CONTROL OF APHID POPULATIONS
A.F.G. DIXON & J.-L. HEMPTINNE
DIXON, A.F.G. & J.-L. HEMPTINNE 2003. Ladybirds and the biological control of
aphid populations. Pp. 1-10 in A.O. SOARES, M.A. VENTURA, V. GARCIA & J.-L.
HEMPTINNE (Eds) 2003. Proceedings of the 8th International Symposium on
Ecology of Aphidophaga: Biology, Ecology and Behaviour of Aphidophagous
Insects. Arquipélago. Life and Marine Sciences. Supplement 5: x + 112 pp.
Although Rodolia and other ladybirds have been successfully used to control pest coccids
they have not proved effective in classical biological control programmes against aphids. A
better understanding of the foraging behaviour of ladybirds and a more realistic theory of
insect predator- prey dynamics are beginning to reveal the reason for this.
Aphidophagous ladybirds exploit patches of aphid prey for feeding and reproduction. As
suitable nurseries for their offspring patches of aphid prey generally only persist for about
the same period of time as it takes the larvae of these ladybirds to complete their
development. This is the case even in the absence of natural enemies. Thus aphids become
scarce within a patch just when the food requirements of the ladybirds are greatest. Optimal
foraging theory predicts that ladybirds should lay a few eggs early in the development of a
patch and empirical data indicates that ladybirds appear to forage optimally.
There have been several studies on the cues used by ladybirds when selecting patches of
prey for oviposition. This review will consider how the responses shown by ladybirds may
have shaped what has become known as the "egg window", how cannibalism may regulate
the number of ladybirds within a patch, and the consequences of this for classical biological
control.
Anthony F.G. Dixon (e-mail: a.f.dixon@uea.ac.uk), School of Biological Sciences,

University of East Anglia, Norwich, NR4 7TJ, UK; Jean-Louis Hemptinne, Ecole Nationale
de Formation Agronomique, B.P. 87, FR-31326 Castanet-Tolosan, France.
THEORY (PRICE et al. 1980; Fig 1B). Predators are

considered to be part of a plant's defence. When
In classical insect predator-prey population attacked by herbivores some plants emit volatiles
dynamics organisms in two trophic levels that are attractive to natural enemies, which has
interact; prey and predator (Fig. 1A). resulted in them being likened to "body guards"
A plant through it's morphology and and the use of emotive phraseology like " the
chemistry can directly affect the well being of enemy of my enemy is my ally" (DICKE &
herbivores, and they similarly can affect SABELIS 1988; SABELIS et al. 2001). That
predators. That is, in addition to their effects on ladybirds respond to these volatiles is supported
one another's abundance a plant can have a direct by technically elegant studies in which gas
effect on a herbivore, which can have a direct chromatography of plant volatiles was directly
effect on a predator, and vice versa. In addition to coupled with recordings from the olfactory organs
these direct effects there is a growing literature of a ladybird. Herbivore damaged plants emit (Z)-
that claims predators and parasitoids are attracted jasmone, which is attractive to adult Coccinella
by volatiles emitted by plants under attack by septempunctata (BIRKETT et al. 2000; NINKOVIC
herbivores. This is regarded as a mutualism, in et al. 2001). The central tenet of the mutualism
which the effectiveness of the searching hypothesis is that herbivore-induced plant
behaviour of the natural enemy is enhanced and volatiles enable natural enemies to more easily
the herbivore pressure on the plant reduced find their prey and so reduce herbivore pressure.
1
Claims that such signals are so used by to aphid infestation? Similarly, is the synomone
parasitoids was scrutinized by VINSON (1999) and produced by a plant in response to being eaten by
VAN DER MEIJDEN & KLINKHAMMER (2000), who lepidopterous larvae different from that produced
found no field evidence for this. when infested with aphids? Therefore, in addition
to determining whether the odour originates
directly from the prey (prey pheromone
hypothesis) or indirectly - after feeding by the
prey - from the plant (plant synomone hypothesis)
there is an urgent need to determine whether the
signals are prey specific and how they affect
predators' searching behaviour. It is well
documented that bark beetles aggregate in
response to volatiles produced by trees and
attractant pheromones produced by the beetles,
Fig. 1. The direct and indirect effects on one another and so overcome the host's defences by a mass
of plants, herbivores and predators in classical insect
attack, but avoid heavily attacked trees, when the
population dynamics (A) and plant predator
mutualisms (B). beetles present produce deterrent pheromones
(WOOD 1982; RAFFA 2001). That is, if chemical
Although there is no doubting that the signaling by plants significantly influences
volatiles (synomones - DICKE & SABELIS 1988) ladybird foraging then it is likely the signal is
released by plants when attacked by herbivores complex, as in bark beetles.
are attractive to predators and parasitoids, the way Alternatively one can ignore plants when
in which they affect their searching behavior and considering predator-prey interactions, which is
the distance over which they operate still needs to the case in most mathematical models of
be resolved. Discussions of this problem (e.g. population dynamics. These have been widely
JANSSEN et al. 2002) tend to follow PRICE et al. used to predict the behavior of predator-prey
(1980) and only consider the adaptive systems, in particular their stability and the
significance of herbivore-induced plant volatiles outcome of introducing natural enemies on the
in terms of plant fitness. It is generally assumed it abundance of pests (BEDDINGTON et al. 1976,
is advantageous for natural enemies to respond to 1978; HASSELL 1978; MURDOCH 1994). In spite
such signals. However, it is pertinent to ask - of the great and long-standing interest in these
What advantages would a predator gain by models, there has been little success in using
responding to these signals? Here we consider them to account for why insect predators,
only ladybird beetles, but the principles are likely compared to parasitoids, have generally not been
to apply to all natural enemies. very effective in suppressing the numbers of their
It seems likely that the quantity of volatile prey (DEBACH 1964).
material released by a plant depends on the Our studies on the way insect predators, and
intensity of herbivore attack, i.e., density- ladybirds in particular, forage, led to an
dependent. If this is true then aphid-infested appreciation of the ecological significance of the
plants are likely to be at their most attractive for difference in mobility of juvenile and adult
ladybirds when aphids are most abundant. insects; the latter can fly while the former cannot
However, at this stage in the infestation it is (Fig. 2). That larvae generally stay within a prey
highly likely that ladybird larvae will already be patch while adults may not was incorporated into
present. Therefore, responding to a strong cue a model. Patch in this sense means the space that
that a plant is under attack by aphids is not a larva can explore by walking, usually one or
necessarily advantageous. In addition, as not all only a few adjacent plants, or even only part of an
aphids are equally suitable as prey for ladybirds individual plant as in the case of trees. Three
(RANA et al. 2002) it is relevant to ask: - Is the factors are likely to determine the reproductive
synomone emanating from a plant specific for a strategy of ladybirds to a much greater extent than
particular species of aphid or a general response availability of food, which is the usual
2
assumption of models of predator-prey systems: about 1% of the eggs laid in a patch survive
(1) Ladybird developmental time is much longer (DIXON 2000). Cannibalism may be selected for
than that of its aphid prey and comparable with (see below) and even sibling cannibalism may
the average duration of a patch of prey (Fig. 3; have a selective advantage, if prey becomes
HEMPTINNE et al. 1990; HEMPTINNE & DIXON scarce (OSAWA 1992). To avoid cannibalism,
1991). Thus it is maladaptive for a ladybird to lay adults should avoid patches of aphids where
eggs in an old prey patch, as they are unlikely to ladybird larvae are already present.
complete their development before the aphids
disappear. (2) As shown by KINDLMANN &
DIXON (1993), there should be a selective
advantage in optimizing the number of eggs laid
in a patch. This is because - as stated above -
ladybird developmental time is similar to the
duration of a patch of aphids. If many eggs are
laid, the ladybird larvae may reduce the rate of
increase of the aphids, cause an earlier decline in
Fig. 2. Aphidophagous ladybirds quickly leave patches
aphid abundance, and thus food may become
where aphids are scarce (A) but oviposit in patches
scarce well before the larvae complete their where prey is abundant (B). The larvae (D) that hatch
development (Fig. 3). (3) Cannibalism is common from the eggs (C) are confined to the patch, and have to
in aphidophagous ladybirds and in nature often pursue and subdue the aphids they need for their
reduces juvenile survival dramatically, as only development.
Fig. 3. Graphical presentation of the components of the ladybird-aphid interaction: temporal changes in the
abundance of aphids and relative developmental time of the ladybird, and the outcome if (A) the eggs are laid late,
(B) a few eggs are laid early, or (C) many eggs are laid early.
3
Assuming that the proportion of conspecifics relative abundance of the predator and therefore
in the diet of ladybirds is proportional to their likely to be density dependent.
relative abundance then if prey abundance is kept Consideration of the above leads to the
constant the incidence of cannibalism increases prediction that there should be a strong selection
with increase in predator abundance. That is, for ladybirds to lay eggs only in patches in the
cannibalism is likely to act as a density dependent early stages of development and avoid those
mortality factor. Alternatively if the number of containing conspecific larvae (KINDLMANN &
predators is kept constant and that of their prey is DIXON 1993; DOSTALKOVA et al. 2002). Thus in
varied the incidence of cannibalism decreases assessing the potential effectiveness of a predator
with increase in the abundance of prey (Fig. 4). for biological control one should take into
account that selection maximizes predator fitness,
not its effectiveness as a biocontrol agent
(KINDLMANN & DIXON 1999a). In aphidophagous
ladybirds the major determinant of their
reproductive strategy is that their prey develops
much faster then they do (DIXON et al. 1995;
DIXON & KINDLMANN 1998; KINDLMANN &
DIXON 1999b). Therefore, the potential fitness of
an adult depends mainly on the future trends in
resource availability for its larvae, which unlike
the adult are confined to a patch (Fig. 2). This
leads to the following predictions. In arthropod
predator-prey systems in which the predator has a
long generation time relative to that of its prey
(ladybird/aphid systems), predator reproduction
should be correlated with the age of a prey patch
rather than the numbers of prey present, and top-
down regulation is unlikely. However, in
ladybird/ coccid systems, where both prey and
predator have similar developmental times,
ladybird reproduction is likely to be correlated
with prey abundance and top-down regulation is
possible (KINDLMANN & DIXON 2001). In
Fig. 4. The predicted (A) increase in cannibalism with addition there is evidence that specificity may
increase in predator density, 50 and 150, and (B) the
also be an important attribute of a biological
decrease in cannibalism with increase in aphid density
when predator density is kept constant assuming that: control agent. The coccidophagous ladybirds that
f(x, y) = ay/(x+y), where x is the number of prey, y is feed on Margarodidae, the group of coccids that
the number of ladybirds and a is a scaling constant. includes Icerya, are generally more specific than
those that feed on other groups of coccids. In
terms of successful control ladybirds have been
This is referred to as the "meet and eat" used 20 times more successfully to control
hypothesis and accounts for the incidence of Margarodidae than other groups of coccids
cannibalism in time (DIXON 2000). However, it is (DIXON 2000).
just as plausible that the latter is due to the
occurrence in time of certain vulnerable stages -
eggs/hatchling larvae and pre-pupae/pupae, which EXPERIMENTAL EVIDENCE FOR OPTIMAL
are unable to avoid or defend themselves against FORAGING IN LADYBIRDS
active larvae. Whatever the reason for the
temporal incidence of cannibalism the outcome is What evidence is there that selection maximizes
the same: cannibalism is proportional to the predator fitness? Below is presented the results of
4
studies undertaken to assess this in the case of In the presence of conspecific larvae and/or
aphidophagous ladybirds. In particular, this will their tracks gravid females of Adalia bipunctata,
be done by examining the evidence for an egg Coccinella septempunctata, Cycloneda limbifer,
window, mechanisms for avoiding cannibalism Harmonia axyridis, and Semiadalia
and the proposed consequences for aphid undecimnotata become very active and if
abundance. prevented from leaving the area refrain from
laying eggs for a few hours (HEMPTINNE et al.
1992; DOUMBIA et al. 1998; YASUDA et al. 2000;
Egg Window RģŽIýKA 2001b). Similar responses are observed
when females of A. bipunctata are placed on
Experimental and field studies indicate there is a plants in the field experimentally infested with
density below which ladybirds are unlikely to lay aphids and contaminated with larval tracks.
eggs (DIXON 1959; WRATTEN 1973; HONċK (Fréchette, unpublished). Although some species
1978). In addition, in the field ladybirds tend to of ladybird respond to the tracks left by larvae of
lay their eggs well before aphid populations peak other species the response is generally statistically
in abundance (Fig. 5; HEMPTINNE et al. 1992). insignificant and much weaker than that to
That is, there is a window in the development of a conspecific larvae or their tracks (HEMPTINNE et
patch of aphids when ladybirds are most likely to al. 1992; YASUDA et al. 2000; RģŽIýKA 1997b,
lay their eggs. The opening of the window is 2001a, b). This is expected because the greatest
possibly determined by the minimum density of threat to the survival of a ladybird in its preferred
aphids required for the survival of the first instar habitat, where it is likely to be the most abundant
larvae (DIXON 1959). The closing of the window ladybird, are individuals of the same species. In
appears to be initiated by adults responding to the addition, ladybirds appear to be well defended
presence of conspecific larvae (HEMPTINNE et al. chemically against intraguild predation
1992). (AGARWALA & DIXON 1992; HEMPTINNE et al.
2000). The deterrent effect of larval tracks is
density dependent and mediated via a pheromone
present in the tracks. In the case of A. bipunctata
the cue consists of a cocktail of alkanes, which
spread easily on the hydrophilic cuticle of plants
and so leave a large signal. In addition the
oviposition-deterring pheromone is very stable
lasting for at least 10 days (DOUMBIA et al. 1998;
HEMPTINNE et al. 2001).
In summary, there is good field evidence that
aphidophagous ladybirds, as predicted by theory,
lay their eggs early in the development of patches
of aphids, and laboratory and field experiments
reveal the possible mechanisms by which this is
achieved.
Cannibalism
Fig. 5. Distribution in time, relative to peak aphid Cannibalism is widely recorded for
abundance of the laying of eggs by Adalia bipunctata
on lime trees. Development of aphid populations
aphidophagous ladybirds, but rarely mentioned in
expressed in weeks before and after the recorded peak the literature on coccidophagous species. Theory
in aphid abundance in each year. (After HEMPTINNE et predicts that it should occur when the relative
al. 1992) abundance of ladybirds is high and/or is
5
associated with an asymmetry between cannibal ladybird H. axyridis indicates it prefers to eat
and victim. The victim is usually at a vulnerable conspecifics (GAGNÉ et al. 2002). Thus
stage in its development (AGARWALA & DIXON cannibalism would appear to have been selected
1992), i.e., in the egg or pupal stage, or is smaller for in the individuals of H. axyridis used in this
or about to moult or pupate. That is, cannibalism study.
should be highest in the egg and pupal stages, and Not only does the high probability of egg
in the fourth instar larval stage when prey is cannibalism make it advantageous for ladybirds
likely to be scarce, and decrease with increase in to avoid ovipositing in patches of prey already
aphid abundance (Fig. 4). Life table studies done occupied by conspecific larvae field, but evidence
on field populations and laboratory studies (Fig. indicates that cannibalism, as predicted by theory,
6) support these predictions (AGARWALA & serves subsequently to regulate the numbers of
DIXON 1992; YASUDA & SHINYA 1997). ladybird larvae within a patch (Fig. 7).
Fig. 7. The relationship between egg cannibalism and

the number of eggs of Adalia bipunctata per unit area
of lime foliage in relation to aphid abundance in the
field (After MILLS 1982)
Fig. 6. The incidence of cannibalism in the laboratory That is, cannibalism is strongly density
of clutches of eggs (A) and larvae (B) of Adalia dependent and capable of regulating the
bipunctata in relation to aphid abundance (After
abundance of ladybird larvae within patches
AGARWALA & DIXON 1992)
(KINDLMANN & DIXON 2001).
In summary, there is good field evidence that
In the grain beetle Tribolium there are strains
cannibalism is widespread and an important
that show either a high or a low level of
mortality factor potentially capable of regulating
cannibalism, which is genetically determined
the abundance of aphidophagous ladybird larvae
(STEVENS 1992). This has also been shown for H.
in a patch.
axyridis (WAGNER et al. 1999). Thus, selection
should favour an optimum level of cannibalism in
a given environment. That is, a species may be Aphid abundance
more or less cannibalistic than one would expect
on the basis of the predicted frequency of The prediction that ladybirds that forage
encounters between conspecifics outlined above. optimally have little affect on aphid abundance
Is there any evidence for this? Clearly some (KINDLMANN & DIXON 1993) is the most
species are more difficult to rear collectively contentious. The implied altruism on the part of
because they show higher levels of cannibalism the ladybirds and criticism of biological control
than other species (unpublished results). A recent practice has greatly impeded the general
study of cannibalism in the aphidophagous acceptance of this supposedly counterintuitive
6
idea. There is good evidence that ladybirds forage Unlike in other studies (e.g. ELLIOT &
in a way similar to that predicted by optimal KIECKHEFER 2000) the shrubs were not caged, so
foraging theory and they achieve this by the patches in effect were open to both
behaviour that is clearly adaptive at the individual immigration and emigration of both aphids and
level. The fact that cannibalism is adaptive and ladybirds as in natural ecosystems. That is, as
strongly density dependent indicates that ladybird predicted by theory these predators do not have a
numbers are likely to be strongly auto-regulated. negative effect on the peak numbers of aphids in
Therefore, the prediction that ladybirds should nature.
have little affect on aphid abundance is in reality In summary, although well based theoretically
also not counterintuitive. and supported by a rigorous field experiment, the
This prediction was tested by monitoring the prediction that aphidophagous ladybirds have
numbers of the aphid, Aphis gossypii, on 34 two little affect on aphid abundance is likely to be
metre high shrubs of Hibiscus syriacus in the subject to further critical experimentation before
field. All the eggs of Coccinella septempunctata it is generally accepted.
brucki were removed from 8 of the shrubs, all
those of Harmonia axyridis from another 8, all
CONCLUSIONS
the eggs of both ladybirds from another 12 and no
eggs were removed from the remaining 6 shrubs
(control). Sticky bands were placed around the Although the idea of a mutualism between plants
base of the stem of each shrub to prevent the and ladybirds is an attractive one there are no
immigration of larvae on to the shrubs from compelling theoretical reasons for, or field
surrounding plants. The results were very variable evidence of, such a relationship. Classical
but clearly indicate that the presence of predator-prey models do not account for why
aphidophagous predators on the shrubs did not insect predators are generally less effective in
significantly affect the peak number of aphids suppressing the abundance of pests than
(Fig. 8). parasitoids. A model that includes the essential
features of the foraging behavior of larvae and
adults and the reproductive behavior of adult
ladybirds predicts the patterns observed in the
field. The major determinant of abundance in this
system is the relative developmental times of the
predator and prey - generation time ratio (GTR)
hypothesis. If that of the predator is considerably
longer than that of the prey, as in aphid/ladybird
systems, than top down regulation of prey
abundance is unlikely, whereas when it is of
similar length, as in coccid/ladybird systems, then
top down regulation is possible. The cues used by
aphidophagous ladybirds to assess the quality of
patches of prey have been identified and
rigorously assessed. That is, in the last ten years
there has been a great advance in our
understanding of the patterns and processes in
ladybird-prey interactions.
Fig. 8. The peak number of Aphis gossypii on Hibiscus The GTR model should apply to all insect
shrubs in the field when aphid numbers were monitored predators. However, as far as aphidophaga are
in the presence of all the naturally occurring natural
enemies (control), and when all the Harmonia axyridis
concerned it makes a prediction: those that have
(H.a) or Coccinella septempuntata brucki (C.s.) or both longer generation times than aphids should
species of ladybird (H.a. + C.s.) were removed at the behave similarly to ladybirds. Although this has
egg stage. not been studied intensively many are known to
7
be cannibalistic and show similar reproductive DIXON, A.F.G. & P. KINDLMANN 1998. Generation ratio
behaviour. For example, the adults of some and the effectiveness of ladybirds as classical
cecidomyids, chrysopids and syrphids are biological control agents. Pp. 315-320 in M.P.
deterred from ovipositing by the presence of ZALUCKI, R.A.I. DREW & G.G. WHITE (Eds.) Pest
Management - Future Challenges 1. University of
conspecific larvae or their tracks (HEMPTINNE et Queensland Printery.
al. 1993; RģŽIýKA 1994, 1996, 1997a, 1998; DOSTÁLKOVÁ, I., P. KINDLMANN & A.F.G. DIXON 2002.
RģŽIýKA & HAVELKA 1998). Thus, it is likely Are classical predator-prey models relevant to the
that the GTR hypothesis holds for all insect real world? Journal of Theoretical Biology 218:
predators. At present the best support for this 328-330
comes from studies on aphidophagous insects. DOUMBIA, M., J.-L. HEMPTINNE & A.F.G. DIXON 1998.
Assessment of patch quality by ladybirds: role of
larval tracks. Oecologia 113: 197-202.
REFERENCES ELLIOT, N.C. & R.W. KIECKHEFER 2000. Response of
coccinellids to spatial variation in cereal aphid
AGARWALA, B.K. & A.F.G. DIXON 1992. Laboratory density. Population Ecology 42: 81-90.
study of cannibalism and interspecific predation in GAGNÉ, I., D. CODERRE & Y. MAUFFETTE 2002. Egg
ladybirds. Ecological Entomology 17: 303-309. cannibalism by Coleomegilla maculata lengi
BEDDINGTON, J.R., C.A. FREE & J.H. LAWTON 1976. neonates: preference even in the presence of
Concepts of stability and resilience in predator-prey essential prey. Ecological Entomolgy 27: 285-291.
models. Journal of Animal Ecology 45: 791-816. HASSELL, M.P. 1978. The dynamics of arthropod
BEDDINGTON, J.R., C.A. FREE & J.H. LAWTON 1978. predator-prey systems. Princeton University Press,
Characteristics of successful natural enemies in Princeton.
models of biological control of insect pests. Nature HEMPTINNE, J.-L., A.F.G. DIXON & A. MACKENZIE
273: 513-519. 1990. Adaptations du cycle biologique des
BIRKETT, M.A., C.A.M. CAMPBELL, K. CHAMBERLAIN, prédateurs aphidiphages aux fluctuations
E. GUERRIERI, A.J. HICK, J.L. MARTIN, M. démographiques de leaurs proies. Pp. 101-104 in
MATTHES, J.A. NAPIER, J. PETTERSSON, J.A Ed. INRA. Régulation des cycles saisonniers chez
PICKETT, G.M. POPPY, E.M. POW, B.J. PYE, L.E. les Invertébrés. Les Colloques de L'INRA 52.
SMART, G.H. WADHAMS, L.J. WADHAMS & C.M. HEMPTINNE, J.-L. & A.F.G. DIXON 1991. Why
WOODCOCK 2000. New roles for cis-jasmone as an ladybirds have generally been so ineffective in
insect semiochemical and in plant defense. biological control. Pp. 149-157 in L. POLGAR, R.J.
Proceedings National Academy of Science, USA CHAMBERS, A.F.G. DIXON & I. HODEK (Eds).
97: 9329-9334. Behaviour and Impact of Aphidophaga.
DEBACH, P. 1964. Biological Control of Insect Pests HEMPTINNE, J.-L., A.F.G. DIXON & J. GOFFIN 1992.
and Weeds. Chapman & Hall, London. Attack strategy of ladybird beetles (Coccinellidae):
DICKE, M. & M.W. SABELIS 1988. How plants obtain factors shaping their numerical response.
predatory mites as body guards. Netherlands Oecologia 90: 238-245.
Journal of Zoology 38: 131-138. HEMPTINNE, J.-L., A.F.G. DIXON, J.-L. DOUCET & J.-E.
DIXON, A.F.G. 1959. An experimental study of the PETERSEN 1993. Optimal foraging by hoverflies
searching behaviour of the predatory coccinellid (Diptera: Syrphidae) and ladybirds (Coleoptera:
beetle Adalia decempunctata (L.). Journal of Coccinellidae): Mechanisms. European Journal of
Animal Ecology 28: 259-281. Entomology 90: 451-455.
DIXON, A.F.G. 2000. Insect Predator-Prey Dynamics: HEMPTINNE, J.-L., A.F.G. DIXON & C. GAUTHIER 2000.
Ladybirds & Biological Control. Cambridge Nutritive cost of intraguild predation on eggs of
University Press, Cambridge. Coccinella septempunctata and Adalia bipunctata
DIXON, A.F.G., J.-L. HEMPTINNE & P. KINDLMANN (Coleoptera: Coccinellidae). European Journal of
1995. The ladybird fatasy - prospects and limits to Entomology 97: 559-562.
their use in biocontrol of aphids. Hemptinne, J.-L., G. Lognay, M. Doumbia & A.F.G.
Züchtungsforschung 1: 395-397. Dixon 2001. Chemical nature and persistence of the
DIXON, A.F.G., J.-L. HEMPTINNE & P. KINDLMANN oviposition deterring pheromone in the tracks of
1997. Effectiveness of ladybirds as biological the larvae of the two spot ladybird, Adalia
control agents: patterns and processes. bipunctata (Coleoptera: Coccinellidae).
Entomophaga 42: 71-83. Chemoecology 11: 43-47.
8
HONċK, A. 1978. Trophic regulation of post diapause RģŽIýKA, Z. 1996. Oviposition-deterring pheromone in
ovariole maturation in Coccinella septempunctata chrysopids: Intra- and interspecific effects.
[Col.: Coccinellidae]. Entomophaga 23: 213-216. European Journal of Entomology 93: 161-166.
JANSSEN, A., M.W. SABELIS & J. BRUIN 2002. Evolution RģŽIýKA, Z. 1997a. Persistence of the oviposition-
of herbivore-induced plant volatiles. Oikos 97: 134- deterring pheromone in Chrysopa oculata (Neur.:
138. Chrysopidae). Entomophaga 42: 109-114.
KINDLMANN, P. & A.F.G. DIXON 1993. Optimal RģŽIýKA, Z. 1997b. Recognition of oviposition-
foraging in ladybird beetles (Coleoptera: deterring allomones by aphidophagous predators
Coccinellidae) and its consequences for their use in (Neuroptera: Chrysopidae, Coleoptera:
biological control. European Journal of Coccinellidae). European Journal of Entomology
Entomology 90: 443-450. 94: 431-434.
KINDLMANN, P. & A.F.G. DIXON 1999a. Strategies of RģŽIýKA, Z. 1998. Further evidence of oviposition-
aphidophagous predators: lessons for modelling deterring allomone in chrysopids (Neuroptera;
insect predator-prey dynamics. Journal of Applied Chrysopidae). European Journal of Entomology
Entomology 123: 397-399. 95: 35-39.
KINDLMANN, P. & A.F.G. DIXON 1999b. Generation RģŽIýKA, Z. 2001a. Response of chrysopids
Time Ratios - Determinants of prey abundance in (Neurptera) to larval tracks of aphidophagous
insect predator-prey interactions. Biological coccinellids (Coleoptera). European Journal of
Control 16: 133-138. Entomology 98: 283-285.
KINDLMANN, P. & A.F.G. DIXON 2001. When and why RģŽIýKA, Z. 2001b. Oviposition responses of
top-down regulation fails in arthropod predator- aphidophagous coccinellids to tracks of ladybirds
prey systems. Basic and Applied Ecology 2: 333- (Coleoptera: Coccinellidae) and lacewings
340. (Neuroptera: Chrysopidae) larvae. European
MILLS, N.J. 1979. Adalia bipunctata (L.) as a generalist Journal of Entomology 98: 183-188.
predator of aphids. Ph.D. Thesis, University of RģŽIýKA, Z. & J. HAVELKA 1998. Effects of
East Anglia, UK
oviposition-deterring pheromone and allomone on
MILLS, N.J. 1982. Voracity, cannibalism and Aphidoletes aphidimyza (Diptera: Cecidomyidae).
coccinellid predation. Annals of Applied Biology European Journal of Entomology 95: 211-216.
101: 144-148.
SABELIS, M.W., A. JANSSEN & M.K. KANT 2001.
MURDOCH, W.W. 1994. Population regulation in theory
Perspective: The enemy of my enemy is my ally.
and practice. Ecology 75: 271-287.
Science 291: 2104-2105.
NINKOVIC, V., S. AL ABASSI & J. PETTERSON 2001. The
STEVENS, L. 1992. Cannibalism in beetles. Pp. 156-175
influence of aphid-induced plant volatiles on
in ELGAR, M.A. & B. J. CRESPI (Eds). Cannibalism,
ladybird beetle searching behavior. Biological
Ecology and Evolution among Divers Taxa.
Control 21: 191-195.
Oxford: Oxford University Press.
OSAWA, N. 1992. Sibling cannibalism in the lady beetle
Harmonia axyridis: fitness consequences for VAN DER MEIJDEN, E. & P.G.L. KLINKHAMMER 2000.
mother and offspring. Researches in Population Conflicting interests of plants and natural enemies
Ecology 34: 45-55. of herbivores. Oikos 89: 202-208.
PRICE, P.W., C.E. BOUTON, P. GROSS, B.A. MCPHERON VINSON, S.B. 1999. Parasitoid manipulation as a plant
J.N. THOMSON & A.E. WEIS 1980. Interaction defense strategy. Annals of Entomological Society
among three trophic levels: Influence of plants on of America 92: 812-828.
interaction between insect herbivores and natural WAGNER, J.D., M.D. GLOVER, J.B. MOSELEY & A.J.
enemies. Annual Review of Ecology and MOORE 1999. Heritability and fitness consequences
Systematics 11: 41-65. of cannibalism in Harmonia axyridis. Evolutionary
RAFFA, K.F. 2001. Mixed messages across multiple Ecology Research 1: 375-388.
trophic levels: the ecology of bark beetle chemical WOOD, D.L. 1982. The role of pheromones,
communication systems. Chemoecology 11: 49-65. kairomones, and allomones in the host selection
RANA, J.S., A.F.G. DIXON & V. JAROSIK 2002. Costs behaviour of bark beetles. Annual Review of
and benefits of prey specialization in a generalist Entomology 27: 411-446.
insect predator. Journal of Animal Ecology 71: 15- WRATTEN, S.D. 1973. The effectiveness of the
22. coccinellid beetle, Adalia bipunctata (L.), as a
RģŽIýKA, Z. 1994. Oviposition-deterring pheromone in predator of the lime aphid, Eucallipterus tiliae (L.).
Chrysopa oculata (Neuroptera: Chrysopidae). Journal of Animal Ecology 42: 785-802.
European Journal of Entomology 91: 361-370. YASUDA, H. & K. SHINYA 1997. Cannibalism and
9
interspecific predation in two predatory ladybirds oviposition behaviour of the predatory ladybird,
in relation to prey abundance in the field. Harmonia axyridis (Coleoptera: Coccinellidae).
Entomophaga 42: 155-165. European Journal of Entomology 97: 551-553.
YASUDA, H., T. TAKAGI & K. KOGI, K. 2000. Effect of
conspecific and heterospecific larval tracks on the Accepted 31 May 2003.
10
ADDITIVE EFFECTS OF PEA APHID NATURAL ENEMIES DESPITE

INTRAGUILD PREDATION
WOLFGANG W. WEISSER
WEISSER, W.W. 2003. Additive effects of pea aphid natural enemies despite
intraguild predation. Pp. 11-15 in A.O. SOARES, M.A. VENTURA, V. GARCIA & J.-
L. HEMPTINNE (Eds) 2003. Proceedings of the 8th International Symposium on
Intra-guild predation (IGP) has been shown for many arthropod pest - natural enemy
systems but the importance of IGP for biological control is still unclear. While theoretical
results and some experiments suggest that IGP can disrupt long-term control of pest
populations, few studies have investigated short-term consequences of IGP. An experiment
was performed in which up to two larvae of the predatory ladybird Coccinella
septempunctata and up to two females of the parasitoid Aphidius ervi were released in
colonies of pea aphids, Acyrthosiphon pisum. Ladybirds kill parasitoid larvae when they
consume parasitized aphids. After one week, aphid population size in the control treatment
was higher than in treatments with natural enemies. The effect of predators and parasitoids
on aphid population size was additive, and the greatest effect on aphid population size was
found with two predator larvae and two parasitoid females. Thus, in this experiment, a
release of multiple natural enemies was beneficial and IGP did not interfere with the short-
term control of the pest. It is suggested that the effects of IGP on biological control will
depend on the desired aims of the control program. If short-term control is desired, the
disadvantages of IGP for long-term control may not be important.
Wolfgang W Weisser (e-mail: wolfgang.weisser@uni-jena.de), Institute of Ecology,

Friedrich-Schiller-University, Dornburger Str. 159, DE-07743 Jena, Germany.
INTRODUCTION presence of generalist predators can reduce the

number of herbivores parasitized by parasitoid
Most herbivorous arthropods are attacked by a Hymenoptera and Diptera (references in SNYDER
variety of natural enemies. In the case of & IVES 2001). However, the occurrence of IGP
generalist predators, these predators may not only does not necessarily imply that biological control
feed on the herbivore but also on other natural of a pest species is interrupted. Only a few studies
enemies. Predation on other natural enemies, an have tested whether IGP has a measurable effect
interaction referred to as intraguild predation on the population growth of the pest species, as
(IGP), appears to be common among generalist predicted by theory (POLIS et al. 1987). For
predators, in both natural (POLIS et al. 1989; example, SNYDER & IVES (2001) found that
YASUDA & KIMURA 2001) and agricultural control of pea aphids by the parasitoid Aphidius
(ROSENHEIM et al. 1993, 1995; ROSENHEIM 1998) ervi Haliday was disrupted when predatory
systems. For biological control, IGP is important carabid beetles were present in field cages. This
as it may interfere with the efficient control of effect, however, only appeared after more than
pest species (ROSENHEIM et al. 1995). Specialist one parasitoid generation, presumably because
natural enemies, in particular parasitoids, are the IGP Predator mostly consumed prey in the
most often the victims of IGP by generalist first generation after introduction. On short
predators (SNYDER & IVES 2001; LUCAS et al. plants that could be climbed by carabids, the
1998). A number of studies have shown that the immediate effect of predators was a strong
11
reduction in aphid densities (SNYDER & IVES laboratory. For the experiment, 2nd and 3rd-instar
2001). Thus, while long-term control of the pest larvae were used. Mummies of the parasitoid A.
was hindered by the presence of generalist ervi were obtained from Katz Biotech Services,
predators, IGP did not appear to have hindered Industriestr. 38, 73642 Welzheim. Females were
short-term reduction of pest densities. For fed for 1-3 days with honey and allowed to mate
biological control of insect pests, in particular in before use in the experiment.
glasshouse situations, a short-term reduction of For the experiment, aphid colonies of a fixed
pest densities may often be more desirable than a age-structure were obtained by placing on each
long-term control of the herbivore. If short-term plant nine 1st or 2nd instar, eight 3rd instar, six
reduction of pest densities is the aim, a joint 4th instar and seven adults, in total 30 aphids.
release of natural enemies might be useful despite Clone RG2 has an intrinsic growth rate of about
of the occurence of IGP. r=0.4 (Braendle, unpubl.) and the age
Pea aphids, Acyrthosiphon pisum Harris, are composition of the colony mimicked its stable
pests of legumes (BLACKMAN & EASTOP 2001). age distribution.
Among the many natural enemies that attack the All aphid colonies were assembled on the
pea aphid are the aphid parasitoid A. ervi and the same day. To start the experiment, zero, one or
predatory coccinellid, Coccinella decempunctata two females of A. ervi and zero, one or two larvae
L. In pea aphid-ladybird-parasitoid systems, IGP of C. septempunctata were introduced into the
occurs when ladybirds feed on parasitized aphids, colony. The treatment with no predators and no
killing both the aphid and the developing parasitoids served as a control. In total, there
parasitoid larvae (e.g. SNYDER & IVES 2001). were three (parasitoid treatments) x three
Both natural enemies are regularly used as (ladybird treatments) x 15 replicates = 135
biological control agents. The experiment replicates. To prevent the escape of aphids and
reported here was designed to test whether the natural enemies, plants were caged using micro-
effects of the parasitoid and the predator are perforated polypropylen bags (230x420mm).
additive in reducing aphid population growth over Replicates were blocked in groups of nine (one
a short time scale of one week. Specifically, I was replicate for each treatment) to control for the
interested in testing the hypothesis that a joint effects of microclimatic differences within the
release of both natural enemies leads to a more climate chamber on aphid population growth.
efficient control of pea aphids in a situation when After one week, all aphids were carefully
colonies of A. pisum are still small and a swift brushed from the plant into Petri dishes and
reduction of pest density is desired. frozen for counting. Parasitoid mummies found
on plants were counted separately.
Data was analysed using three-way ANOVAs
MATERIAL & METHODS
using the software package SPSS (SPSS 1993).
Data was tested for normality prior to the analysis
For all experiments, the pea aphid clone RG2 was and transformed as necessary. Block was treated
used. This clone was collected in Riehen, as a random effect.
Switzerland, in 1998 and reared in the laboratory
in low densities on a dwarf form of broad bean,
Vicia faba L. (variety The Sutton, Nickerson- RESULTS
Zwaan Ltd, Roswell, Lincolnshire LN7 6DT,
UK), potted in 10cm () pots. Aphids and plants Both ladybirds and parasitoids significantly
were kept in constant temperature chambers reduced the number of aphids counted after one
under long-day conditions (L:D 16:8) at 20±1°C. week (Table 1). The interaction between
Ladybird larvae, C. septempunctata, were predators and parasitoids was not significant. The
collected around Jena in 1999 and bred in the block effect was not significant (Table 1).
12
Table 1.
Results of a 3-way ANOVA on aphid population size (ln-transformed). SS-sum of squares, df - degrees of
freedom, MS - mean square. Error terms were calculated by SPSS.
Source SS df MS F Significance
Intercept Hypothesis 2108.2 1 2108.2 507.1 <0.001
Error 16.6 4 4.2
Ladybirds Hypothesis 390.8 2 195.4 74.3 <0.001
Error 21.0 8 2.6
Parasitoids Hypothesis 103.7 2 51.8 31.1 <0.001
Error 13.4 8 1.7
Block Hypothesis 16.6 4 4.2 1.6 0.307
Error 11.4 4.5 2.5
Ladybirds*Parasitoids Hypothesis 13.9 4 3.5 2.0 0.151
Error 28.4 16 1.8
Ladybirds*Block Hypothesis 21.0 8 2.6 1.5 0.239
Error 28.4 16 1.8
Parasitoids*Block Hypothesis 13.4 8 1.7 0.9 0.511
Error 28.4 16 1.8
*Block Error 171.0 90 1.9
Only few parasitoid mummies were found on population size were additive. The interaction
the plants (Fig. 1). While the parasitoid effect was term was far from being significant, emphasising
significant, the ladybird effect was marginally not that IGP was not strong enough to interfere with
significant. None of the interactions or the block the control of aphids by the natural enemies. The
effect were significant (Table 2 next page). same results were obtained when data were not
log-transformed (data not shown). Thus, for any
2.0 given number of ladybird larvae, the addition of
parasitoid females led to a better control of the
aphid population. Over the time scale of one
Number of mummies
1.5
week, therefore, the control was better the more

1.0 natural enemies were used. The control was
dramatic: instead of on average more than 600
0.5 aphids on the control plant, aphid population size
was often reduced to zero when two ladybird
0.0
larvae and two parasitoids were released on the
0 1 2 plants. A mean aphid population size smaller than
Number of ladybird larvae
30, the initial size of the aphid populations, was
Fig. 1. The effect of the numbers of predators and only obtained when at least two ladybird larvae
parasitoids on the number of parasitoid mummies. No and at least one parasitoid female were released in
mummies were counted in the absence of parasitoids. the aphid colony. Densities of aphids as high as
Open bars: one parasitoid female, black bars: two observed in the control lead to a fast death of the
parasitoid females. host plant (Weisser, unpublished). A fast
reduction of pea aphid populations is therefore
DISCUSSION necessary to limit the damaging effects on the
host plant. Thus, in this experiment, efficient pest
The main result of the experiment was that the control was possible when both natural enemies
effects of the parasitoid and the predator on aphid were used.
13
Table 2.
Results of a 3-way ANOVA on the number of parasitoid mummies (ln-transformed). SS-sum of squares, df -
degrees of freedom, MS - mean square. Error terms were calculated by SPSS.
Source SS df MS F Significance
Intercept Hypothesis 3.7 1 3.7 12.1 0.025
Error 1.2 4 0.3
Ladybirds Hypothesis 1.2 2 0.6 4.1 0.060
Error 1.2 8 0.1
Parasitoids Hypothesis 2.2 2 1.1 11.4 0.005
Error 0.8 8 0.1
Block Hypothesis 1.2 4 0.3 2.8 0.244
Error 0.3 2.5 0.1
Error 2.1 16 0.1
Ladybirds*Block Hypothesis 1.2 8 0.1 1.1 0.402
Error 2.1 16 0.1
Parasitoids*Block Hypothesis 0.8 8 0.1 0.7 0.673
Error 2.1 16 0.1
*Block Error 14.0 90 0.2
The number of parasitoid mummies found effective in preventing the outbreak of an aphid
after one week was very low because of the population despite the adverse effects of the
duration of parasitoid development (SEQUEIRA & predator on parasitoid reproduction.
MACKAUER 1992). As a consequence, the effect More generally, the results presented here
of ladybird presence on the number of mummies suggest that the presence of IGP does not
was marginally non-significant. It is likely that a necessarily interfere with biological pest control.
later sampling date would have resulted in a In the past years, an increasing number of studies
significant ladybird effect. Similarly, the apparent have found that IGP commonly occurs in pest-
decrease in mummy number when two rather than natural enemy systems. The implicit assumption
one parasitoid females were released is likely due in many of the studies is that if IGP can be
to chance effects given the very low numbers of documented in feeding trials in the laboratory or
mummies. in the field, it must have negative consequences
The experimental situation was artificial in the for biological control. Theoretical studies that
sense that natural enemies and aphids were not focus on long-term equilibrium conditions
able to emigrate from the host plant. support this view (e.g. POLIS et al. 1989).
Nevertheless, it would be premature to conclude However, as shown in this study, it depends not
that the results do not apply to a greenhouse or only on the biology of the system but also on the
field situation. SNYDER & IVES (2001) also found aims of a biological control program whether or
a strong reduction in aphid numbers by generally not IGP prevents efficient pest control. If a short
less effective carabid beetles in large field term reduction of pest densities is desired because
enclosures. Thus, it is well possible that the a harvest in the immediate future is endangered,
results reported here also hold in a greenhouse long-term negative effects of IGP may not be
scenario. Because of the high rates of increase of important. Similarly, if the aim is to quickly
aphid populations, control has to be fast to limit reduce pest density populations to prevent an
the impact on the host plants. The results outbreak, then a release of multiple natural
presented in this paper suggest that the joint enemies may still be advisable despite the
release of parasitoids and ladybirds can be occurrence of IGP. To assess the importance of
14
IGP for a biological control program, experiments POLIS G.A., C. MYERS & R. HOLT 1989. The ecology
should be performed that are closely linked to the and evolution of intraguild predation: Potential
aims of this program. competitors that eat each other. Annual Review of
Ecology and Systematics 20: 297- 330.
ROSENHEIM, J.A. 1998. Higher-order predators and the
ACKNOWLEDGEMENTS regulation of insect herbivore populations. Ann.
Rev. Entomol. 43: 421-447.
ROSENHEIM, J.A., H.K. KAYA, L.E. EHLER, J.J. MAROIS
I thank Ingrid Jakobi, Grit Kunert and the
& B.A. JAFFEE 1995. Intraguild predation among
students of course “GPI-WS01/02”, in particular biological control agents: Theory and evidence.
Ms. Dagmar Lorch, for help with the Biological Control 5: 303-335.
experiments. John J. Sloggett kindly reared the ROSENHEIM, J. A., L.R. WILHOIT & C.A. ARMER 1993.
ladybird larvae and Jens Schumacher gave Influence of intraguild predation among generalist
statistical advice. Parasitoid mummies were insect predators on the suppression of an herbivore
donated by Katz Biotech Services. This work was population. Oecologia 96: 439-449.
supported by grant Wei 2618/2-1 of the Deutsche SEQUEIRA, R. & M. MACKAUER 1992. Nutritional
Forschungsgemeinschaft (DFG). ecology of an insect host-parasitoid association: the
pea aphid-Aphidius ervi system. Ecology 73: 183-
189.
REFERENCES SPSS for Windows: Base System User's Guide,
Release 6.0, 1993. SPSS Inc. Headquarters, 233 S.
Wacker Drive, 11th floor, Chicago, Illinois 60606.
BLACKMAN, R. L. & V.G. EASTOP 2001. Aphids on the YASUDA, H. & T. KIMURA 2001. Interspecific
world's crops: an identification and information interactions in a tri-trophic arthropod system:
guide. John Wiley & Sons, New York, USA. 476 effects of a spider on the survival of larvae of three
pp. predatory ladybirds in relation to aphids.
LUCAS, E., D. CODERRE & J. BRODEUR 1998. Intraguild Entomologia Experimentalis et Applicata 98: 17-
predation among aphid predators: characterization 25.
and influence of extraguild prey density. Ecology
79: 1084-1092. Accepted 31 May 2003.
15
16
HOST PREFERENCE OF Lysiphlebus testaceipes (CRESSON) (HYMENOPTERA:

APHIDIIDAE) FOR Myzus persicae (SULZER) AND Aphis gossypii GLOVER
(HEMIPTERA: APHIDIDAE).
V.H.P. BUENO, A.B. CARNEVALE & M.V. SAMPAIO
BUENO, V.H.P., A.B. CARNEVALE & M.V. SAMPAIO. 2003. Host preference of
Lysiphlebus testaceipes (Cresson) (Hymenoptera: Aphidiidae) for Myzus persicae
(Sulzer) and Aphis gossypii Glover (Hemiptera: Aphididae). Pp. 17-20 in A.O.
SOARES, M.A. VENTURA, V. GARCIA & J.-L. HEMPTINNE (Eds) 2003. Proceedings
of the 8th International Symposium on Ecology of Aphidophaga: Biology,
Ecology and Behaviour of Aphidophagous Insects. Arquipélago. Life and Marine
Sciences. Supplement 5: x + 112 pp.
The acceptance of a host by a parasitoid should indicate that the host must have appropriate
characteristics for oviposition. The majority of Aphidiid parasitoids attack a range of host
species, but these hosts may differ in visual and gustatory cues and/or in the quality for
parasitoid progeny. Such differences may lead to evolution of host preference. We studied
the host preference of L. testaceipes for A. gossypii and M. persicae in choice and non-
choice tests. The searching behavior of L. testaceipes in a non-choice test and in a choice
test with A. gossypii and M. persicae as hosts showed that the numbers of encountered
hosts, probes with the ovipositor, ovipositions, and the parasitoid larvae found after host
dissection were higher in A. gossypii than in M. persicae. Based on the number of hosts
accepted for oviposition we conclude that the parasitoid L. testaceipes preferred A. gossypii
both in the non-choice and in choice tests.
Vanda H.P. Bueno (e-mail: vhpbueno@ufla.br), A.B. Carnevale & M.V. Sampaio,
Department of Entomology, Federal University of Lavras, P.O.Box 37, CEP-37200-000
Lavras, MG, Brazil.
INTRODUCTION explain the final parasitization rates of the host

population.
Whether an immature parasitoid will successfully Aphidiidae are solitary parasitoids of aphids
develop in its host, depends on the host choice by (HAGEN & BOSCH 1968). Lysiphlebus testaceipes
the adult parasitoid female. Host acceptance by together with Aphidius colemani stand out as the
parasitoids is composed of three steps: host dominant species in South America with several
habitat finding, host finding, and host acceptance host species (STARÝ & CERMELI 1989) including
(VINSON & IWANTSCH 1980). Success of Myzus persicae (Sulzer) and Aphis gossypii
development of the parasitoid is determined by Glover (STARÝ et al. 1993). Parasitism of L.
the suitability of the host. After host finding, the testaceipes is low in M. persicae (CARVER 1984;
parasitoid begins its evaluation of the host by STEENIS 1993; CARNEVALE 2002) when
antennal tapping and ovipositor probing compared to other parasitoids species such as
(MACKAUER et al. 1996). Physiological and Aphidius colemani (STEENIS 1993; SAMPAIO et al.
nutritional conditions of the host are inspected by 2001a, b), or to L. testaceipes parasitism on A.
the parasitoid during its evaluation (VINSON gossypii (RODRIGUES & Bueno 2001;
1997), but behavioral responses are also CARNEVALE 2002) in laboratory conditions.
important in host acceptance/rejection (CHAU & We studied the host preference of L.
MACKAUER 2001). The understanding of testaceipes for M persicae and A. gossypii
parasitoid behavior is accomplished by studying through comparison of the searching and host
the steps in the host selection process and helps acceptance behavior of the parasitoid.
17
MATERIAL & METHODS species were placed in a Petri dish, and 15 L.

testaceipes females were tested per treatment (N=
Insect rearing: Rearing and experiments were 2x15= 30 replicates). In the choice test each
carried out at 25±1qC, 70 ±10% RH and 12 hr female of L. testaceipes was exposed to 10 aphids
photophase. M. persicae and A. gossypii were of each species; 28 females were tested (N= 28
reared on sweet pepper and cotton seedlings, replicates). In the non-choice test the counts were
respectively. Twenty apterous females of each transformed as ¥x+0.5, and a t-test at 5% of
aphid species were placed in Petri dishes (15 cm probability was used to test for differences
Ø) with a 1 cm layer of 1% agar-water medium between aphid species. To test for differences
and leaves of sweet pepper and cotton according between aphid species in the choice test, the F2
to the aphid species to be studied. These dishes test was used.
were placed in climatic chambers and adult
females removed after two days, with nymphs of RESULTS & DISCUSSION
2nd and 3rd instars left for the studies with the
parasitoids.
In the non-choice tests, the mean number of
L. testaceipes was kept on Schizaphis
encounters (t = 3.49; p< 0.05), ovipositor probes
graminum (Rondani) on sorghum leaves. Only 24
(t = 2.97; p< 0.05), and ovipositions (t = 8.57; p<
hr old mated females without previous
0.05) in the 15min period and the number of
oviposition experience were used in this study.
parasitoid larvae (t = 6.32; p< 0.05) of L.
Preference in non-choice and choice tests:
testaceipes found after host dissection were all
Petri dishes (5 cm Ø) with a 1 cm layer of agar-
significantly higher in A. gossypii (mean±SD,
water medium and 4 cm Ø sweet pepper leaf
37.7±3.60, 18.7±1.83, 18.0±1.28, and 10.5±1.44
disks with 20 M. persicae nymphs and/or A.
respectively) than in M. persicae (mean±SD,
gossypii were used. A single female of L.
21.2±3.78, 10.7±2.28, 5.1±0.79, and 2.0±0.83
testaceipes was released in each dish and
respectively) (Fig. 1). In the choice test we found
observations were performed under a stereoscopic
the same differences as in the non-choice test: the
microscope during 15 minutes. The number of
number of encounters of L. testaceipes was
encounters of the parasitoid with hosts, the
significantly higher in A. gossypii (537) than in
number of ovipositor probings and the number of
M. persicae (457) (F2 = 6.44; p< 0.05), as well as
accepted hosts were recorded. A. gossypii was
transferred to cotton leaves in a new Petri dish ovipositor probing (286 and 168, respectively) (F2
after the experiment while M. persicae was kept = 30.67; p< 0.01), the number of ovipositions
on sweet pepper leaves. The number of larvae of (277 and 65, respectively) (F2 = 131.41; p< 0.01),
the parasitoid in the hosts was determined by and the total number of parasitoid larvae found at
dissecting of aphids three days after parasitism. dissection (132 and 29, respectively) (F = 65.89;
In the non-choice test, 20 aphids of one of a p< 0.01) (Table 1).
50
M y z u s p e r s ic a e
A p h is g o s s y p ii
a
40
Avarege number
30
b
a a
20
b a
10
b
b
0
E n c o u n te r s P ro b in g O v ip o s itio n s L arv ae
Fig. 1. Average number of antennal tapping (encounters), ovipositor probing, oviposition and total of parasitoid
larvae (mean r standard error) of Lysiphlebus testaceipes on Myzus persicae and Aphis gossypii in non-choice test.
18
Table 1.
Number of encounters, ovipositor probes, ovipositions and parasitoid larvae of Lysiphlebus testaceipes on Aphis
gossypii and Myzus persicae in choice test. Significant at 5% (*) and 1% (**) probability, F2 test.
Hosts Total number
Species Number Encounters Probes Ovipositions Larvae
A. gossypii 280 537* 286** 277** 132**
M. persicae 280 457* 168** 65** 29**
Olfactory and visual stimuli may have great without need of physical contact with the host.
importance in the mechanism of locating and The low number of M. persicae hosts accepted by
recognition of hosts at short distances, besides L. testaceipes directly influences the low
offering information of direction and distance of percentage parasitism of this host.
the hosts (MACKAUER et al. 1996). Observations
of the number of encounters in the choice and
ACKNOWLEDGEMENTS
non-choice test differed statistically between the
two host species. Thus, it was demonstrated that
the host preference of L. testaceipes was The first and third author wish to thank CNPq, the
apparently determined before antennal contact. second author wishes to thank CAPES/ Brazil, for
For host preference of other species of scholarships for their studies. CNPq is also
aphidiids in choice tests, Chow & MACKAUER thanked for the financial support of equipment
(1991) and SAMPAIO et al. (2001b) found and materials.
differences only in the number of accepted hosts,
demonstrating that the preference is defined only REFERENCES
after the touch with the ovipositor. On the
contrary, L. testaceipes preferred A. gossypii in CARNEVALE, A.B. 2002. Adequabilidade de Myzus
this study without need of host touching with the persicae (Sulzer) e Aphis gossypii Glover
ovipositor. (Hemiptera: Aphididae) como hospedeiros de
In this study, L. testaceipes clearly preferred Lysiphlebus testaceipes (Cresson) (Hymenoptera:
A. gossypii over M. persicae. Surprisingly, it Aphidiidae). (MS Thesis). UFLA, Lavras, 47 pp.
seems that both hosts are suitable for CARVER, M. 1984. The potential host ranges in
development of L. testaceipes. The only Australia of some imported aphid parasites (Hym.:
difference we found was that parasitoids reared Icheneumonoidea: Aphidiidae). Entomophaga 38:
351-359.
on A. gossypii had a higher longevity (5 days)
CHAU, A. & M. MACKAUER 2001. Host-instar selection
than the ones reared on M. persicae (4 days) in the aphid parasitoid Monoctonus paulensis
(CARNEVALE 2002). It is generally proposed that (Hymenoptera: Braconidae, Aphidiinae): assessing
parasitoids prefer hosts that guarantee optimum costs and benefits. Canadian Entomologist 133:
conditions for development and growth in 549-564.
immature stages (GODFRAY 1994). In some cases, CHOW, A. & M. MACKAUER 1991. Patterns of host
however, the opposite was found: MESSING & selection by four species of aphidiid
RABASSE (1995) showed that Aphis citricola van (Hymenoptera) parasitoids: influence of host
der Goot (unsuitable) was preferred by Aphidius switching. Ecological Entomology 4: 403-410.
colemani instead of M. persicae (suitable). In GODFRAY, H.C.J. 1994. Parasitoids, behavioral and
evolutionary ecology. Princeton, New Jersey:
other cases the preferred host is the one more Princeton University Press. 473 pp.
easily attacked by the parasitoid female, even if it HAGEN, K.S. & R. VAN DEN BOSCH 1968. Impact of
is less suitable (CHAU & MACKAUER 2001). pathogens, parasites, and predators on aphids.
The parasitoid L. testaceipes preferred the Annual Review of Entomology 13: 325-384.
species A. gossypii in all tests we carried out, MACKAUER, M., J.P. MICHAUD, & W. WÖLKL 1966.
although this host does not seem to be more Host choice by aphidiid parasitoid (Hymenoptera:
suitable than M. persicae. In the choice and in the Aphidiidae): host recognition, host quality, and
non-choice test the preference was defined host value. Canadian Entomologist 6: 959-980.
19
MESSING, R.H., & J.M. RABASSE 1995. Oviposition REMAUDIÈRE 1993. Environmental research on
behavior of the polyphagous aphid parasitoid aphid parasitoid biocontrol agents in Chile (Hym.,
Aphidius colemani Viereck (Hymenoptera: Aphidiidae; Hom., Aphidoidea). Journal of Applied
Aphidiidae). Agriculture, Ecosystems and Entomology 115: 292-306.
Environment 52:13-17. STARÝ, P., & M. CERMELI 1989. Parasitoides de áfidos
RODRIGUES, S.M.M., & V.H.P. BUENO 2001. Parasitism en plantas cultivadas de Venezuela. Boletín de
rate of Lysiphlebus testaceipes (Cresson) (Hym.: Entomología venezolano 5: 77-80.
Aphidiidae) on Schizaphis graminum (Rond.) and STEENIS, M.J. VAN 1993. Suitability of Aphis gossypii
Aphis gossypii Glover (Hem.: Aphididae). Glov., Macrosiphum euphorbiae (Thom.) and
Neotropical Entomology 30: 625-629. Myzus persicae Sulz. (Hom.: Aphididae) as host for
SAMPAIO, M.V., V.H.P. BUENO, & R. PÉREZ-MALUF several aphid species (Hym.: Braconidae).
2001a. Parasitismo de Aphidius colemani Viereck IOBC/WPRS 26: 157-160. (Bulletin).
(Hymenoptera: Aphidiidae) em diferentes VINSON, S.B. 1997. Comportamento de seleção
densidades de Myzus persicae (Sulzer) (Hemiptera: hospedeira de parasitóides de ovos, com ênfase na
Aphididae). Neotropical Entomology 30: 81-87. família Trichogrammatidae. Pp. 67-119 in J.R.P.
SAMPAIO, M.V., V.H.P. BUENO, & J.C. VAN LENTEREN PARRA, & R.A. ZUCHI (Eds.). Trichogramma e o
2001b. Preferência de Aphidius colemani Viereck controle biológico aplicado. Piracicaba, FEALQ.
(Hymenoptera: Aphidiidae) por Myzus persicae 324 pp.
(Sulzer) e Aphis gossypii Glover (Hemiptera: VINSON, S.B., & G.F. IWANTSCH 1980. Host suitability
Aphididae). Neotropical Entomology 30: 655-660. for insect parasitoids. Annual Review of
STARÝ, P., M. GERDING, H. NORAMBUENA, & G. Entomology 25: 397-419.
Accepted 31 May 2003.
20
THE BIOLOGICAL CONTROL OF Aulacorthum solani (KALTENBACH)

(HOMOPTERA: APHIDIDAE) IN GREENHOUSE GROWN PEPPER; RESEARCH
ON A TRI-TROPHIC SYSTEM
J.V. SCHELT & F. WÄCKERS
SCHELT, J.V. & F. WÄCKERS 2003. The biological control of Aulacorthum solani
(Kaltenbach) (Homoptera: Aphididae) in greenhouse grown pepper; research on a
tri-trophic system. Pp. 21-27 in A.O. SOARES, M.A. VENTURA, V. GARCIA & J.-L.
HEMPTINNE (Eds) 2003. Proceedings of the 8th International Symposium on
To improve the control of the foxglove aphid (Aulacorthum solani) in greenhouse sweet
pepper, several trials on three trophic levels have been conducted. Two different lines of
Aphelinus abdominalis were compared on flight capacity, parasitation and predation. One
line was significantly better in parasitation of A. solani (6.3 vs 0.2 mummies/female/day).
This line performed also better at lower temperatures (15-18º C.) in flight capacity.
Predation was the same for both lines (2 aphids/female/day).
The honeydew of A. solani was tested as a food source on Aphidius spp. The life span of
Aphidius on this honeydew was 4 days which was equal to water and half of sucrose. On
100 ha. of commercially grown sweet pepper a banker plant system (wheat, Sitobion avena,
A. abdominalis) was used. In general growers were able to reduce the number of chemical
corrections from 7 to around 3.
In sweet peppers grown under glass there is a natural oscillation of 4-5 weeks in flowering
and fruit set. We observed large differences in aphid growth depending on this vegetative or
fruiting phase. The practical considerations how to adjust the biological control still has to
be investigated.
Jeroen van Schelt (e-mail: jvschelt@koppert.nl), Koppert B.V., Veilingweg 17, 2650 AD
Berkel en Rodenrijs, Netherlands & F. Wäckers, The Netherlands Institute of Ecology
(NIOO-CTE), Boterhoeksestraat 22, NL-6666 GA Heteren,The Netherlands.
INTRODUCTION years. Because A. solani is a very polyphagous

and a cosmopolitan species, it has become a
The integrated control of aphids in glasshouses problem in most pepper growing regions in North
has become common practice in tomatoes, West Europe and Canada. The aphids are
peppers, cucumbers, and aubergines. (VAN generally found in the lower parts of the plant and
SCHELT 1999; KLAPWIJK 1999; MULDER et al. are easily overlooked at the beginning of an
1999). Aphis gossypii and Myzus spp. are infestation. Already at low densities the plants
controlled with Aphidius colemani; Macrosiphum can react very strongly to the saliva of the aphid.
euphorbiae with Aphidius ervi. Often generalist On the leaves yellow necrotic spots can be found
predators like Aphidoletes aphidimyza and lady and often the top is showing malformations. In
beetles are used in conjunction with the the period 1997-2000 biological control with A.
parasitoids to control high aphid numbers. ervi and the gall midge A. aphidimyza was
However the control of Aulacorthum solani in practised on several hundreds of hectares by
peppers is still very problematic. commercial growers in the Netherlands. Results
A. solani has become a very common pest in were unsatisfactory because too often chemical
sweet pepper in greenhouses over the last three corrections with Pirimicarb had to be applied.
21
The low percentage parasitation may be the beneficials have to fly for an optimal dispersal
explained by the foraging behaviour of this in the greenhouse. Flight propensity was assessed
parasite (SCHWORER & VÖLKL 2001). If within a at different temperatures for both lines.
certain time no aphids are found than A. ervi has
the tendency to disperse over large distances. Material and Methods
The potential use of another parasite,
Aphelinus abdominalis, was explored. Two lines The parasitism and predation rate was determined
were compared on their dispersal capacity and by introducing individual females of A.
fecundity. The use of banker plant systems
abdominalis (n=30) for 24 hours on a sweet
(winter wheat with Sitobion avenae) was pepper leaf disc on agar with 20 first and second
developed to enhance the numbers of A. instar A.solani. After 24 hours the female was
abdominalis before aphid growth in the crop.
removed and the number of aphids that were fed
Research on the aphid itself was carried out upon by the female, was determined. Because of
by NIOO-CTE. We speculated that the aphid‘s slow detoriation of the leaf, the aphids were
honeydew could be toxic or unsuitable as a food
transferred to a fresh leaf after 8 days. After
source for the beneficials used. Finally we looked fourteen days the number of parasitized aphids
more in detail at the physiological status of the (mummies) was assessed.
plant and its influence on aphid growth.
The French line was tested again on A. solani
after being reared for one generation on this host.
SELECTING LINES OF A. abdominalis The experiments were conducted in a climate
cell at 21 ± 1°C, 70% RH, 16L:8D.
Introduction Flight propensity was determined by putting
100 adult wasps, less than 24 hrs old, in an open
50 ml. plastic bottle. The bottle was placed on a
Selecting of lines within a species can be an
small concrete platform in the middle of a bowl
option to improve the control capacity of
( 25 cm.) with water containing a drop of
beneficials.
detergent.
Aphelinus abdominalis is normally associated
This set up was put in a cage of 30 by 30 cm.
with Macrosiphum euphorbiae and cereal aphids
and 50 cm. in height. The opening of the bottle
as Sitobion avenae. HÖLLER & HAARDT (1993)
was 6.5 cm. above the surface of the water. The
compared a uniparental German line and a
cage was put in a climate box with fluorescent
biparental French line. Both lines showed a high
tubes at all sides. Experiments were done at 15,
fecundity on S. avenae in the laboratory but failed
18, 21, 24 and 27 °C and 75% RH. After 24 hours
in the field. Both lines however had similar
the number of parasitic wasps that remained in
biological characteristics, but were only tested on
the bottle, that had drowned, and that had crossed
S. avenae.
the water were counted. Before releasing the
In this study we compared two lines of A.
parasitic wasps, they were acclimatised for one
abdominalis. One line came from a German
hour to the ambient temperature. All tests were
producer, the other line was provided by
done twice.
INRA/Antibes (France). Both lines were reared
separately on Macrosiphum euphorbiae.
Results
They were compared with respect to their host
feeding behaviour and their fecundity with A. Parasitation and Predation
solani as a host.
Because there is a tendency to save energy in Amount of host feeding was 2.33 aphid/day for
greenhouse systems by reducing temperature in the French line, 2.03 for the German line (n.s.
the winter, the dispersal capacity of beneficials MWU-test).
under cool conditions (15-18 °C) is important. Parasitization was significantly different: 6.3
Moreover if the leaves of the plants are not mummie/female for the French line, 0.23 for the
touching each other in the beginning of the season German line (MWU, P<0.001)
22
No differences could be seen between the Discussion

parasitation capacity of the French line that had
been reared on M. euphorbiae (6.3 From these results it was concluded to use the
mummie/female) and parasites of the same line French strain of A. abdominalis in the mass
that had been reared for one generation on rearing and for commercial use. In theory the
A.solani (7.7 mummie/female) (n.s. MWU). German line could be improved by rearing them
for several generations on A. solani, but the
Flight capacity offspring of the first experiment was so low, that
The lines also differed clearly with respect to further efforts in this direction were stopped.
their flight capacity. The number of wasps from Also M. euphorbiae and not A.solani is the
the German line that left the bottle was preferred aphid for a mass rearing system.
significantly lower at 15 and 18°C. (MWU-test) It can be speculated that a further exploration
(Fig. 1) In the numbers that actually flew across of A. abdominalis coming from other parts of its
the water the same pattern was visible (Fig. 2). distribution area can be interesting.
The flight-pattern could be well observed in
this set-up. When A. abdominalis are taking off, THE SUITABILITY OF A. solani HONEYDEW
they jump up a little, and subsequently lose AS A FOOD SOURCE FOR Aphidius spp.
altitude very quickly before regaining height
again. Only when they did not hit the water at the Introduction
lowest point, could they reach the other side of
the barrier.
The great majority of parasitoids and many
arthropod predators depend on sugar sources to
cover their energetic needs. There is strong
% of A. abdominalis that left
100
theoretical as well as empirical evidence that the
75
availability of suitable sugar sources can be a key
50 factor determining the population dynamics of
predator-prey and parasitoid-host systems
25
(KRIVAN & SIROT 1997; WÄCKERS 2003).
0 Besides (extra) floral nectar, honeydew is the
15 18 21 24 27 most prevalent source of exogenous sugars in
French German Temperature in °C nature. Due to the fact that agricultural
Fig. 1. Percentage adult A. abdominalis that left the ecosystems often lack flowering plants,
the bottle at different temperatures. honeydew is likely of particular importance in
agriculture. A recent study by Wäckers and
Steppuhn (unpublished) demonstrated that 80%
60
of the larval parasitoid Cotesia glomerata
collected in a cabbage field contained honeydew
% of A.abdominalis that flew
50
40
specific sugars, indicating a high incidence of
honeydew feeding by this parasitoid. Parasitoids
30
of honeydew-producing insects are believed to be
20 even more intimately linked to this food source.
10 Despite this intimate link, honeydew can vary
considerably with respect to its nutritional
0
15 17 21 24 27
quality. Whilst certain types of honeydew can be
French German Temperature in °C equally suitable as nectar or sugar solutions,
others are clearly inferior or even toxic
Fig. 2. Percentage adult A. abdominalis that flew (WÄCKERS 2000). To test the suitability of A.
across the waterbarrier. solani honeydew from various A. solani-plant
23
combinations for aphid parasitoids, we compared controls, separate A. colemani cohorts were
the longevity of honeydew-fed A. colemani, to subjected to water only and a 1μl droplet of a 1M
parasitoids fed with a sucrose solution. Though A. sucrose solution respectively. The petridishes
colemani is not a natural parasite of A.solani it were kept at T=20ºC, RH=95-100%, 16L8D. The
served as a model for aphid parasitoids in general high humidity prevented the honeydew and
and Aphidius spp. in particular. sucrose solution from becoming too viscuous.
Survival of wasps was scored daily. Every
Materials and method second day the food droplets were renewed. For
each treatment we tested 50-60 individuals.
Aulacorthum solani honeydew was collected Parasitoids that were found dead in the honeydew
from aphid colonies feeding on the following droplets were discarded.
plant species: Brassica nigra, Capsicum We used HPLC (High Performance Liquid
frutescens, Gossypium herbaceum, and Vicia Chromatography) to analyse the sugar
sativa. The plants had been grown in 1l. potting composition of the honeydew types used in the
soil in greenhouses at the Dutch Institute for longevity experiment
Ecology (NIOO-CTE) in Heteren. Growing
conditions were T=20ºC, RH=50-80%, 16L8D. Results
The homopteran-plant combinations were
kept in fine-mesh screen cages to prevent Longevity
contamination. To collect the honeydew, a glass
plate was placed underneath the plant. After 24 Parasitoids provided with water only (control)
hours the honeydew was collected using a glass lived 4.0 days on average. Access to the sucrose
micropipette and subsequently stored in a freezer solution increased parasitoid livespan more than
at -15ºC. twofold (8.4 days). The honeydew, on the other
Newly emerged parasitoids were placed in a hand, had only a marginal effect on parasitoid
petridish and provided with a piece of water- longevity (Fig 3). Only honeydew from A. solani
soaked cotton wool as well as a 1μl droplet of one on V. sativa raised parasitoid longevity relative to
of the honeydew types placed on the lid. As the water control (Mann-Whitney U-test).
12
c
10
8
b a a a a
6
0
herbaceum
Gossypium
frutescens
Brassica
Capsicum
Vicia
sucrose
nigra
water
sativa
Honeydew from Aulacorthum solani feeding on Controls

Fig. 3. Longevity of A. colemani when provided with various sugar sources, or water only (control). Different
letters indicate significant differences among treatments (MWU-test).
Honeydew Sugar Composition the aphid’s host plant (Table 1). Most notably, the
The HPLC sugar analysis showed that the aphid synthesized sugar erlose was absent in
composition of A. solani honeydew depends on honeydew from G. herbaceum, while occuring in
24
substantial amounts in honeydew collected from however, the honeydew was dominated by
the other three plants. Maltose was only found in sucrose and its hexose components glucose and
the honeydew collected from Vicia faba. Overall, fructose.
Table 1
Sugar composition of honeydew collected from various Aulacorthum solani-plant combinations. Numbers
represent percentage of total sugar content (weight/weight).
Sucrose Glucose Fructose Unidentified Erlose Trehalose Maltose Sorbitol Mannitol
Capsicum frutescens 46.9 12.0 22.5 1.4 15.6 1.1 0.0 0.0 0.4
Gossypium herbaceum 35.5 19.4 39.5 2.0 0.0 0.0 0.0 1.3 2.3
Brassica nigra 35.0 15.2 27.7 3.3 16.7 0.8 0.0 0.0 1.4
Vicia sativa 26.8 19.7 30.5 2.1 11.5 0.0 7.7 0.8 0.9
Discussion the other honeydew oligosaccharides studied,

erlose was moderately suitable as a food source,
Based on the fact that honeydew is primarily a while trehalose and raffinose were unsuitable
sugar solution, it is often assumed that it makes (WÄCKERS 2001). Similar results were obtained
for a suitable insect food source. However, the in sugar longevity studies for A. colemani
data presented here, as well as previous reported (Wäckers, unpublished). However, given the
studies show that there can be a considerable relatively low concentration of these sugars in the
variation in honeydew suitability. It is yet unclear honeydew tested here, it is unlikely that the sugar
which honeydew components are responsible for composition is responsible for the poor survival
this variation, but two possibilities have been of A. colemani in this study.
proposed: (i) secondary plant compounds and (ii) Irrespective of the underlying mechanism, the
sap-feeder synthesized sugars. It has been poor quality of A. solani honeydew may have
reported that secondary plant compounds appear considerable ecological and applied implications.
in honeydew (e.g. MALCOLM 1990). However, The poor performance of A. ervi in controlling A.
little is known about the effects of these solani in the glasshouse may be partly due to the
compounds on honeydew-feeding insects. While limited survival of Aphidius spp. on A. solani
secondary plant compounds in the honeydew honeydew.
might have been responsible for the poor
suitability of A. solani honeydew, it would be THE USE OF A. abdominalis ON BANKER
remarkable that the effect on the parasitoid would PLANTS
be similar in the four plants tested, as they
represent a wide variation in plant secondary
chemistry. In order to obtain large numbers of A.
The few studies addressing the effect of sap- abdominalis in the greenhouse at the right time, a
feeder synthesized sugars indicate that these banker plant system was developed. Winter wheat
specific sugars might have a negative effect on (cv. “Vivant”) was sown in hanging baskets (
insect longevity (WÄCKERS 2000). ZOEBELEIN 20 cm.) and infected with the cereal aphid
(1956) reported that melezitose feeding reduced Sitobion avenae. Small numbers of parasites (20-
the life span of the parasitoid Microplectron 30 /banker) were introduced on the bankers. It
uscipennis in comparison to food-deprived took two generations before a large number of
individuals, while sucrose, glucose or fructose mummies (2-3000/banker) could be found. Most
increased M. uscipennis longevity by a factor of bankers were started in February (20/ha.) and
more than 2. In studies with C. glomerata, peaked in parasite production at the beginning of
melezitose prolonged the parasitoid’s life span April when first infections with A. solani could be
relative to control individuals provided with water expected.
only, but reduced longevity by 44-47% in By putting yellow sticky traps in the
comparison to sucrose, fructose and glucose. Of greenhouse it was possible to monitor the
25
production and dispersal of adult A. abdominalis growth and subsequently aphid control. We
over time. Peak production of the bankers observed large differences in aphid growth
occurred in week 15 (around 8-10 weeks after between different greenhouses and we concluded
their introduction); bankers were removed at that this could only be explained by differences in
week 20, because all aphids were parasitised and plant physiology. As an example in figure 4 the
all mummies were emerged. Until week 29 A. number of newly formed pepper fruits per week is
abdominalis was trapped on yellow sticky traps. plotted for the same greenhouse in two
The impact of A. abdominalis on the population consecutive years. In 2001 fruit set occurred
of A. solani was hard to quantify. Marked continuously during the year. In 2002 however,
colonies of A. solani often had disappeared when oscillations with a period of 4 weeks occurred.
inspected again after one week. This can be partly Especially in peppers fluctuations in fruit set is a
explained by host feeding of the parasites. Black “natural” phenomenon, though plant breeders are
mummies however were found not only on selecting for varieties with a more even fruit set.
leaves, but also at the base of the stem, on ropes, Also abiotic factors (like day/night temperature
plastic of the rock wool pot, and even on the and feeding regime) can influence fruit set.
ground. Finally growers know that if fruit set is unequal at
In 2002 the banker plant system was used on the beginning of the season it is almost
approximately 100 ha. One grower (10 ha.) did impossible to go back to a more even fruit set
not need to use any chemicals for aphid control, later on. As a consequence it is observed that the
though several infections with A. solani were aphid population growth is reflected in the fruit
observed. Most other growers were able to reduce set pattern. In the vegetative phase the aphid
the number of Pirimicarb applications from population will grow much faster than in the
around 7 to 2- 3. fruiting phase. Small infections with A. solani
will explode very fast in this period. Until now
practical consequences for the release of
THE INFLUENCE OF TIMING OF FRUIT SET beneficial insects have not been considered.
ON APHID CONTROL However if the biological control is already
poorly established and the crop is in a vegetative
An aspect which is easily overlooked is the phase, an advice to use a chemical correction will
influence of the physiology of the plant on aphid be given sooner.
Setting of new fruit per week per m²

9
8
7
6
5
4
3
2
1
0
1 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37
2001 2002
2
Fig. 4. Setting of new peppers per week/m for 2 consecutive years.
26
OVERALL CONCLUSIONS KLAPWIJK, J.N. 1999. Biological control of tomato

pests in the Netherlands. IOBC Bulletin 22(1): 125-
128
The biological control of the foxglove aphid KRIVAN, V. & E. SIROT 1997. Searching for food or
A.solani in peppers has been improved by several hosts: the influence of parasitoids behavior on host-
measurements. By comparing two lines of A. parasitoid dynamics. Theoretical Population
abdominalis, a line was chosen which performs Biology 51: 201-209.
better at lower temperatures. This is important MALCOLM, S.B. 1990. Chemical defences in chewing
and sucking insect herbivores: plant-derived
because there is a tendency to lower the
cardenolides in the monarch butterfly and oleander
temperatures in greenhouses during winter and aphid. Chemoecology 1: 12-21.
spring. Because the rate of increase of A. MULDER, S., H. HOOGERBRUGGE, K. ALTENA & K.
abdominalis is rather low, a banker plant system BOLCKMANS 1999. Biological pest control in
was developed to rear high numbers of parasites cucumbers in the Netherlands. IOBC Bulletin
before A. solani can be expected in the 22(1): 177-180
greenhouse. With an average of 20 bankers per SCHELT, J.VAN 1994. The selection and utilisation of
hectare and 3000 mummies per banker, around 6 parasitoids for aphid control in glasshouses.
parasites/m2 were reared in the greenhouse. Proceedings of the Section Experimental and
Numbers were sometimes that high that the Applied Entomology of the Netherlands
Entomological Society 5: 151-157
number of predated aphids was even higher than
SCHELT, J.VAN 1999. Biological control of sweet
the number of mummies found. pepper pests in the Netherlands. IOBC Bulletin
The hypothesis that the honeydew of the 22(1): 217-220
aphids was toxic for Aphidius spp. could not be SCHWORER, U. & W. VÖLKL 2001. Foraging behavior of
confirmed. However the result that this honeydew Aphidius ervi (Haliday) (Hymenoptera:
had the same nutritional value as water was Braconidae: Aphidiinae) at different spatial scales:
remarkable. Resource utilization and suboptimal weather
Future research will focus on further conditions. Biological Control 21(2): 111-119
improvement of the banker plant system (optimal WÄCKERS, F.L. 2000. Do oligosaccharides reduce the
suitability of honeydew for predators and
timing, quality of the banker plant), the screening
parasitoids? A further facet to the function of
of new parasites and predators. The influence of insect-synthesized honeydew sugars. Oikos 90:
the plant physiology on the aphids and 197-201.
subsequently on biological control is challenging WÄCKERS, F.L. 2001. A comparison of nectar- and
but can only be solved in cooperation with other honeydew sugars with respect to their utilization by
partners. the hymenopteran parasitoid Cotesia glomerata.
Journal of Insect Physiology 47: 1077-1084.
WÄCKERS, F.L. 2003. The effect of food supplements
REFERENCES on parasitoid-host dynamics. Pp. 226-231 in R.
VAN DRIESCHE (Ed). International symposium on
biological control of arthropods. USDA Forest
HÖLLER, C. & H. HAARDT 1993. Low field Service, publication FHTET-03-05
performance of an aphid parasitoid, Aphelinus ZOEBELEIN, G. 1956. Der Honigtau als Nahrung der
abdominalis, efficient in the laboratory (Hym., Insekten. Zeitschrift für angewandte Entomologie
Aphelinidae). Entomophaga 38: 115-124 38: 369-416.
27
28
COMPARISON OF THE REPRODUCTIVE INVESTMENT IN

COCCIDOPHAGOUS AND APHIDOPHAGOUS LADYBIRDS (COLEOPTERA:
COCCINELLIDAE).
A. MAGRO, J-L. HEMPTINNE, A. NAVARRE & A.F.G. DIXON
MAGRO, A., J-L. HEMPTINNE, A. NAVARRE & A.F.G. DIXON 2003. Comparison
of the reproductive investment in Coccidophagous and Aphidophagous ladybirds
(Coleoptera: Coccinellidae). Pp. 29-31 in A.O. SOARES, M.A. VENTURA, V.
GARCIA & J.-L. HEMPTINNE (Eds) 2003. Proceedings of the 8th International
Symposium on Ecology of Aphidophaga: Biology, Ecology and Behaviour of
Aphidophagous Insects. Arquipélago. Life and Marine Sciences. Supplement 5: x
+ 112 pp.
The prey of coccidophagous ladybird beetles has a slower rate of development and is less
mobile than that of aphidophagous ladybirds. These differences are paralleled by a suite of
characters suggesting that coccidophagous species live at a slower pace than aphidophagous
species. Data in the literature tend to indicate that coccidophagous ladybirds live longer and
have a lower fecundity than aphidophagous species. Thus the expectation is that
coccidophagous species allocate proportionally less resources to their gonads than
aphidophagous species. The reproductive investment in C. montrouzieri and A. bipunctata
support this prediction. The fat in gonads represents 27.4 % of the total body fat in C.
montrouzieri and 37.1 % in A. bipunctata.
Alexandra Magro (e-mail: alexandra.magro@educagri.fr), A. Navarre & J-L. Hemptinne,

Ecole Nationale de Formation Agronomique, Jeune Equipe 000271JE1 “Laboratoire
d’Agro-écologie”, BP 87, FR-31326 Castanet-Tolosan, France & A.F.G. Dixon, School of
Biological Sciences, University of East Anglia, Norwich NR4 7TJ UK.
INTRODUCTION egg weight multiplied by the ovariole number and

related this to adult weight for 8 aphidophagous
Worldwide there have been 155 attempts to ladybirds. There is a very strong positive
control aphids by introducing ladybirds, and only correlation between reproductive investment and
one was supposedly substantially successful, adult weight. Gonads make up a fixed proportion
whereas of the 613 such attempts to control of the body weight, and there is a trade-off in the
coccids resulted in 53 complete or substantial number of eggs a species can lay per day and egg
successes (DIXON 2000). Coccidophagous size. MAGRO (1997) performed the same kind of
ladybirds are therefore very successful biocontrol study on 3 coccidophagous ladybirds, where the
agents compared to aphidophagous species. reproductive biomass is again a fixed proportion
DIXON (2000) suggested, therefore, that the study of the total body mass. However, the reproductive
of ladybird successes and failures in biological biomass in coccidophagous ladybirds makes up a
control programmes might enable us to smaller percentage of the total body mass than in
identifying the specific attributes of successful aphidophagous species. That is, coccidophagous
biological control agents. ladybirds invest proportionally less in
Reproductive potential is generally regarded reproduction than aphidophagous ladybirds.
as an important attribute of natural enemies (e.g., DIXON (2000) states that there is a slow-fast
DEBACH 1973; MESSENGER et al. 1976; VAN continuum in the life history characteristics of
DRIESCHE & BELLOWS Jr. 1996). STEWART et al. ladybirds. When the reproductive investment of
(1991) measured reproductive investment as the coccinellids is considered along with other
29
parameters such as rate of development, relative replaced and the extraction continued for another
growth rate, the rate of metabolism, speed of 9 hours. The tissues were then removed and the
movement, reproductive rate and rate of ageing, solvent evaporated off overnight, after which the
coccidophagous species appear to have a slow tissues were weighed.
pace of life and aphidophagous species a fast pace Fat contents were estimated as indicated by
of life. These different paces of life reflect those equations 1 and 2:
of their prey.
As fat is the chief form in which energy is Ovfat Ovdw Ovdws (1)
stored (WIGGLESWORTH 1972), we decided to
verify the above conclusions by comparing the Tfat Tdw Tdws (2)
allocation of fat to gonads and soma in an
aphidophagous and a coccidophagous ladybird.
This paper reports the preliminary results of this where Ovfat is the fat content of gonads, Ovdw is
study. the dry weight of the gonads and Ovdws is the dry
weight of the gonads after fat extraction. Tfat is
the total fat content of the body, Tdw is the total
MATERIAL & METHODS dry weight of the body and Tdws is the total dry
weight of the body after fat extraction.
Biological material Mean values for both species were compared
using a t test.
Two similar-sized ladybird species were studied.
Cryptolaemus montrouzieri Mulsant is a well- RESULTS & DISCUSSION
known Australian coccid predator, which has
been introduced all around the world to control Figure 1 presents the results of this study in terms
mealybugs. Adults weigh on average 11.1 mg of the mean percentage as well as the standard
(MAGRO 1997). Adalia bipunctata (L.) is a deviation of fat invested in the gonads of the two
European aphidophagous ladybird. Adult weight ladybirds.
is on average 15.8 mg (STEWART et al. 1991).
Both species came from our laboratory stock
cultures: C. montrouzieri was reared on potato
sprouts infested with Planococcus citri Risso and
A. bipunctata on Acyrtosiphon pisum Harris
reared from bean plants.
Measure of reproductive investment

Fig. 1. Mean reproductive investment in an
The reproductive investment was measured in aphidophagous ladybird - Adalia bipunctata - and a
terms of mg of fat. coccidophagous ladybird - Cryptolaemus montrouzieri.
The analyses were made on 15 females of C. Results presented in terms of fat in gonads as a
montrouzieri and 14 of A. bipunctata. Females percentage of total body fat. Standard deviation values
are between parentheses.
were between 15 and 30 days old and were laying
eggs regularly.
Each female was weighed (Sartorius Figure 1 indicates that A. bipunctata invests
Supermicro SC2 balance), dissected and its body more of its energy reserves in reproduction than
cut in three parts: head and thorax, ovaries and C. montrouzieri: the percentage of fat in the
abdomen without ovaries. Each part was dried at gonads of the aphidophagous species is greater
35°C until it reached a constant weight. The dried than in the coccidophagous species
tissues was put in Petroleum spirit (boiling range (aphidophagous: 37.1%; coccidophagous: 27.4%;
40–60 °C) for 9 hours and then the solvent was P<0.05). This supports the idea that
30
aphidophagous species invest proportionally more REFERENCES

of their fat reserves in reproduction than DIXON, A.F.G. 2000. Insect predator-prey dynamics.
coccidophagous species of ladybirds. Ladybird beetles and biological control.
Cambridge: Cambridge University Press. 257 pp.
CONCLUSIONS DOUTT, R.L. & P. DEBACH 1973. Some Biological
Control Concepts and Questions. Pp.118-142 in P.
DEBACH (Ed.) Biological Control of Insect Pests
The present work strengthens the conclusions of and Weeds. London: Chapman and Hall Ltd. 844
STEWART et al (1991) and MAGRO (1997) as the pp.
aphidophagous ladybird allocated proportionally KREBS, C.J. 1994. Ecology. The experimental analysis
more of its energy reserves to reproduction than of distribution and abundance (4th ed.). New York:
the coccidophagous ladybird. Harper Collins College Publishers. 801 pp.
This conclusion is important from two points MAGRO, A. 1997. Os Coccinelídeos dos Citrinos:
of view: estudo comparativo do seu interesse em luta
The evolutionary point of view: Organism are biológica. Unpublished doctoral dissertation,
energy limited (KREBS 1994). Limited resources University of Évora, Portugal. 177 pp.
MESSENGER, P.S., F. WILSON & M.J. Whitten 1976.
should be allocated in a way that maximises
Variation, Fitness and Adaptability of Natural
fitness. That is, reproduction is maximised within Enemies. Pp. 209-231 in C.B. Huffaker & P.S.
the constraints of other energy requirements. Messenger (Eds). Theory and Practice of
Therefore it is relevant to ask: Why do Biological Control. New York: Academic Press,
coccidophagous species invest proportionally Inc. 788 pp.
more of their fat reserves in soma than STEWART L.A., J.-L. Hemptinne & A.F.G. Dixon 1991.
aphidophagous species? DIXON (2000) argues Reproductive tactics of ladybird beetles:
that coccids in nature are generally much less relationships between egg size, ovariole number
abundant than the aphid prey of aphidophagous and developmental time. Functional Ecology 5:
380-385.
ladybirds and therefore it would appear to be
VAN DRIESCHE, R.G. & T.S. BELLOWS Jr. 1996.
advantageous for coccidophagous ladybirds to Biological Control. New York: Chapman and Hall.
allocate more of their energy reserves to 539 pp.
searching for prey. WIGGLESWORTH, V.B. 1972. The principles of Insect
In terms of biological control: A high Physiology (7th ed.). London: Chapman and Hall.
reproductive investment, although important, 827 pp.
does not seem to be an essential characteristic of
a successful biological control agent. Accepted, 31 May 2003.
31
32
BIOLOGICAL CONTROL OF THE ROSY APPLE APHID, Dysaphis plantaginea

(PASSERINI) (HOMOPTERA: APHIDIDAE): LEARNING FROM THE
ECOLOGY OF LADYBIRD BEETLES.
J.-L. HEMPTINNE, A.F.G. DIXON & E. WYSS
HEMPTINNE, J.-L., A.F.G. DIXON & E. WYSS 2003. Biological control of the rosy
apple aphid, Dysaphis plantaginea (Passerini) (Homoptera: Aphididae): learning
from the ecology of ladybird beetles. Pp. 33-41 in A.O. SOARES, M.A. VENTURA,
V. GARCIA & J.-L. HEMPTINNE (Eds) 2003. Proceedings of the 8th International
+ 112 pp.
D. plantaginea is one of the major pests of apple. Technical advisers use empirically
derived action thresholds, which depending on the geographic area, vary from 1 to 10
aphids per 100 fruit clusters. Not surprisingly the majority of orchards are sprayed every
year. As a consequence, clones of this aphid are now resistant or tolerant to insecticides.
There is therefore a need to develop other strategies for controlling this pest. Conservation
and enhancement of natural enemies in apple orchards is one of the possible strategies but
this technique does not produce consistent results. Releases of larvae of aphidophagous
predators are promising but they are still expensive. In the near future, significant
improvements in the biological control of D. plantaginea require of a threshold of
economic damage, a better understanding of the ecology of this aphid and of its natural
enemies, and the utilization of more resistant varieties of apple.
Jean-Louis Hemptinne (e-mail: jean-louis.hemptinne@educagri.fr), École Nationale de

Formation Agronomique, BP 87, FR-31326 Castanet-Tolosan, France; A.F.G. Dixon,
School of Biological Sciences, University of East Anglia, Norwich, NR4 7TJ, UK & E.
Wyss, Research Institute of Organic Agriculture, Ackerstrasse, Postfach, CH-5070 Frick,
Switzerland.
INTRODUCTION pest status of this aphid has increased over the
last decade. Spring infestations have become
Dysaphis plantaginea (Passerini) is one of the more severe and last longer. In addition, D.
major pests of apple in Western Europe and North plantaginea has become more tolerant and in
America (BONNEMAISON 1959; BASSINO 1982; some cases resistant to the commonly used
PASQUALINI & BRIOLINI 1982; WHALON & insecticides (DELORME et al. 1998). Reflecting
CROFT 1984; CRUZ DE BOELPAEPE et al. 1987; these trends, the average number of insecticide
BARONIO et al. 1988; GENDRIER et al. 1989; treatments per growing season in Swiss orchards
TOURNEUR et al. 1992; BLOMMERS 1994; SAVINI increased from 1.2 in 1990 to 1.8 in 1995 (HOEHN
2000). It causes the leaves to curl and distorts et al. 1996; GRAF et al. 1999). In South-western
current year shoots. Heavy infestations, France, in 1998 the majority of farmers sprayed 3
particularly just after flowering, reduce the to 4 times against D. plantaginea (DEDIEU 1998).
growth of the fruit and yield by 20 to 30 % (GRAF Although D. plantaginea severely affects
et al. 1999). However, the aphid is rarely apple production (DE BERARDINIS et al. 1994) the
abundant several years running, but tends to reach threshold of economic damage, surprisingly, has
outbreak levels approximately every 5 to 6 years. not been determined (WHALON & CROFT 1984).
Farmers and technical advisers do not seem to be Technical advisers use empirically derived action
aware of these fluctuations and spray or thresholds, which depending on the geographic
recommend spraying of orchards every year. The area, vary from 1 to 10 aphids per 100 fruit
33
clusters (WHALON & CROFT 1984). In Europe the more natural communities (ELTON 1958). As the
action threshold is 1 aphid (BLOMMERS 1994). later harbour more species than the former, the
Thus not surprisingly, the majority of orchards stability of these communities was seen as a
are sprayed every year in early spring. This consequence of their greater biological diversity.
frequent use of insecticides probably accounts for The structure of food webs was often invoked as
the existence of resistant clones of D. an explanation. In complex communities, there
plantaginea. This threshold hampers any change are not only more species of herbivore but also of
in the control of apple aphids. It indicates that D. carnivore. That is, there are more pathways to
plantaginea is extremely harmful and as a channel the energy from one trophic level to the
consequence technical advisers and farmers are next. If one or few carnivore species are
reluctant to reconsider the methods used to temporarily absent, there are still enough
control this aphid. The widespread and frequent consumers to exploit the herbivore productivity.
use of insecticides in apple orchards is contrary to As a consequence, populations show slight
the political desire for the development of fluctuations in abundance around an equilibrium
sustainable agriculture in Europe. Therefore, (KREBS 1994). This ecological concept probably
there is an urgent need to develop other strategies inspired the use of biological control based on
for controlling D. plantaginea. natural enemy conservation and enhancement by
The guilds of natural enemies associated with making fields more hospitable to natural enemies
apple aphids have been extensively recorded by cultivating hedgerows, cover crops or weed
(HODEK 1973; TOURNEUR et al. 1992; HODEK & strips.
HONċK 1996). They generally consist of a This idea is not new and was already practised
minimum of 50 species of insects belonging to in the first decade of the XXth Century in
several families. Coccinellidae is one of the three attempts at controlling aphids (DIXON 2000).
most important families contributing 4 to 6 Later it received some theoretical support from
species to these guilds. This diversity of ROOT (1973) who proposed two hypotheses to
beneficial insects encourages the view that account for the fact that herbivorous insects are
biological control is feasible. Thus the less abundant in complex agroecosystems than in
exploitation of the entire guild as a biological simple ones (Fig. 1).
control agent is an attractive strategy. The
objective is to keep the beneficial insects in the a)
orchards and increase their abundance. To Simple Complex
encourage them overwintering sites are provided Diversity of prey/hosts species Low High
Fluctuations of natural enemies Large Small
and/or additional sources of food such as cover Availability of pollen and nectar Low High
crops, strips of flowers or hedgerows are sown or Functional response of natural enemies Low High
planted. Alternatively the guilds can be used as a
catalogue of potential biological control agents.
Those selected can be reared in factories and b
released in the orchards. This paper considers
these two approaches to the biological control of
apple aphids with special emphasis on ladybird
beetles.
CONSERVATION AND ENHANCEMENT OF

NATURAL ENEMIES IN APPLE ORCHARDS
b)
Fig. 1. Summary of Root’s two hypotheses: a) the four
Since at least the beginning of the XXth Century main characteristics of the enemies hypothesis and b)
circumstantial evidence was thought to favour the the relationship between the number of species and the
concept that simple communities on cultivated number of individuals per species in a community
land were more prone to insect outbreaks than according to the resource concentration hypothesis.
34
Firstly, the enemies hypothesis postulates that aphidophagous predators in determining aphid
predators and parasitoids are more efficient in abundance.
diverse than simple communities of plants.
RUSSELL (1989) reviewed the literature for
60
evidence in favour of this hypothesis. Secondly, strip area
the resource-concentration hypothesis stipulates control area
% of infested trees
that specialist herbivores more easily find, stay in, 40
and reproduce in monocultures of their host
plants. Modern apple orchards typically are
20
simplified communities: the soil below the trees
is kept bare and grass between the rows of trees is
mown regularly. Thus it is not surprising that 0
attempts have been made to control aphids by 1 2 3 4
increasing plant diversity in European and North Infestation classes
American orchards (BROWN & WELKER 1992; Fig. 2. Percentage of infested trees in the strip sown
WYSS 1995; BROWN & SCHMITT 1996; BROWN et area and the control area in an experimental organic
al. 1997; KIENZLE et al. 1997; SOLOMON et al. orchard in May 1993 (After WYSS 1995). Infestation
1999; VOGT & WEIGEL 1999). WYSS’ seminal classes: 1, 1 to 3, 2, 4 to 10, 3, 11 to 50 and 4, 51 to 200
study (1995) lasted for 3 years during which D. aphid colonies.
plantaginea infestation and aphidophagous
predators were monitored in two parts of an
organic orchard. In the first year aphid infestation Diversity and Stability
and predator abundance were identical in the two
zones. In the second year a mixture of species of It is surprising that the enhancement of plant
indigenous dicotyledons was sown in six one- diversity in or near fields in attempts to control
meter wide strips located in one of the two zones. pests attracted such wide support, especially as it
These plants flowered successively from early was not founded on sound scientific principles or
spring to late autumn. Some of them also host experiments (PIMM 1984). As early as 1968,
aphids when they are rare on apple. Therefore, HAIRSTON et al. failed to increase the stability of
pollen, nectar and aphids were available to the experimental communities of bacteria and ciliates
aphidophagous predators throughout the year. by increasing the number of species in his
Later in the second and in the third year, microcosms. Mathematical models of food webs
aphidophagous predators appeared to be more showed that complex networks of consumers
abundant on the trees in the zone with the strip were not more stable than simple ones (MAY
planting and there were fewer trees with large D. 1973). Finally, ROOT (1973) and RUSSELL (1989)
plantaginea colonies in this zone than in the based their hypotheses on circumstantial
control area (Fig. 2). Unfortunately, when VOGT evidence. Currently, the link between diversity
& WEIGEL (1999) repeated WYSS’ experiment in and stability in communities is still a highly
a much smaller orchard, they recorded more D. controversial issue (KREBS 1994; DIAZ & CABIDO
plantaginea on the trees in the zone with the strip 2001). However, properties of communities such
planting than in the control zone. as resistance and/or resilience to perturbations are
These cases studies show that a greater thought to depend on plant functional diversity
abundance and/or diversity of natural enemies rather than number of species. Plant functional
achieved by manipulating plant diversity does not types are sets of species showing similar
automatically translate into aphid control responses to the environment and similar effects
(ANDOW 1986, 1988; VAN EMDEN 1990; VAN on ecosystem functioning (DIAZ & CABIDO
DRIESCHE & BELOW 1996; OBRYCKI & KRING 2001). It should be noted, however, that the
1998; DIXON 2000; LANDIS et al. 2000). This impact of plant functional diversity has been
forces a reconsideration of the link between studied in relatively few cases (DIAZ & CABIDO
diversity and stability and the role of 2001). Recently, field trials in Sweden and
35
Switzerland yielded a positive correlation of aphid colonies and lay a few eggs during the
between plant functional richness and reproductive window (KINDLMANN & DIXON
composition, and the number of aphids, and a 1993). This is more fully discussed by DIXON &
negative one with the number of parasitoids HEMPTINNE (2003).
(KORICHEVA et al. 2000). Although an interesting
result the mechanism linking plant functional
RELEASE OF APHIDOPHAGOUS
diversity and the abundance of these insects is
PREDATORS TO CONTROL APPLE APHIDS
unknown. Finally, in the absence of a well
founded theoretical understanding it is not
surprising that the results of studies on the There have been few releases of natural enemies
conservation and enhancement of natural enemies in apple orchards (BOUCHARD et al. 1988;
in apple orchards are contradictory. HAGLEY 1989; GRASSWITZ & BURTS 1995; WYSS
et al. 1999a, b).
In Europe, Wyss experimentally assessed the
The role of aphidophagous predators ability of predators to keep the numbers of D.
plantaginea below the action threshold. As
The two spot ladybird beetle Adalia bipunctata is producers routinely spray against this aphid early
one of the most abundant predators of D. in spring, because of the low value of the action
plantaginea in European apple orchards. Its threshold, the study was aimed at determining the
reproductive behaviour provides an explanation effectiveness of predators to reduce the numbers
of why it is unable to regulate aphid abundance of fundatrices of the aphid. According to the
(HEMPTINNE et al. 1992; DOUMBIA et al. 1998). literature and field observations, A. bipunctata,
Aphids are smaller and grow much faster than Episyrphus balteatus (De Geer) and Aphidoletes
two spot ladybirds (DIXON 1998, 2000). In the aphidimyza (Rondani) are the most abundant
field, the developmental time, from egg to adult, enemies of D. plantaginea. In Northern Europe,
of A. bipunctata is slightly shorter than the climatic conditions are often harsh when aphids
duration of an aphid colony. If a female ladybird hatch from overwintering eggs. Therefore,
is to maximize its fitness it has to carefully select preliminary trials were made in 1995 to evaluate
its oviposition sites. If its larvae hatch in a very the searching ability of these predators in the
young colony, the probability of finding and field. Apple seedlings kept in 1 m3 cages were
catching prey is extremely low so they are likely infested with fundatrices of D. plantaginea. Eggs
to die of starvation. An old colony is not better or larvae of the three predators were then
because the number of prey is more likely to introduced in the cages. Larvae of A. bipunctata
become scarce before the larvae can complete were the most resistant to frost and efficient at
their development and they then have to compete finding and killing the fundatrices (WYSS et al.
for a dwindling resource and most if not all of 1999a). The effectiveness of this ladybird was
them will die, mainly as a result of cannibalism. further studied on 3-year-old apple trees and on
Between these two extremes, there is a narrow apple branches in a commercial orchard.
reproductive window, oviposition during which On 3 year old apple trees, each infested with 5
results in maximum larval survival. However, fundatrices, were placed ladybird eggs or larvae
laying too many eggs in an aphid colony or in to give four predator-prey ratios: 0:5, 1:5,1:1 and
colonies where there are already ladybird larvae 5:1. The treatments with eggs were unsuccessful
is likely to result in poor survival. In both cases because all the batches of eggs were either
the many predators hasten the decline in aphid destroyed by rain or frost. Larvae at the two
abundance and increases competition for food. In highest predator-prey ratios prevented the
addition, the youngest larvae will be the first to be increase in aphid abundance (Fig. 3; WYSS et al.
eaten by older larvae (AGARWALA & DIXON 1999b). These results were confirmed using
1992). Natural selection is likely to have favoured naturally infested branches of apple trees (Wyss,
ladybird females that are able to assess the quality unpublished results).
36
significantly affects the number of fundatrices

40
next spring (KEHRLI & WYSS 2001). So there is a
potential for biological control in autumn, which
Nr of aphids
needs to be further explored. At first sight,

20 autumn is probably more favourable for
biological control than spring: producers are less
busy and the risk of bad weather interfering with
natural enemies is less.
0
Thirdly, the protection of apple has to be
control 1:5 1:1 5:1
thought of in terms of IPM. More attention has to
Treatments be paid to planting more aphid resistant varieties
Fig. 3. The average number of aphids per tree on of apple. They continue to bean effective mean of
control trees and on trees on which larvae of A. reducing the population growth rate of D.
bipunctata were released at one of the three predator- plantaginea and so maximizing the impact of
prey ratios: 1:5, 1:1 and 5:1 (After WYSS et al. 1999b). natural enemies. Florina, Delorina, FloRub,
Goldrush, Red Devil are resistant to D.
plantaginea (WÜRTH et al. 1999, 2002). In
INTEGRATED PEST MANAGEMENT addition, Ariwa, Renora, Rewena, Rubinola,
Saturn are less susceptible to this aphid. All these
At a time when sustainable agriculture appears varieties are hardly planted in Europe and the US.
more and more frequently at the top of political The major problem is to introduce them onto the
agendas, it is important to convince farmers that market and to convince consumers to buy them.
the biological control of D. plantaginea is as A lower productivity or a poor ability to sustain
reliable and efficient as the current methods of conservation might create additional problems for
control. To reach this objective, research has to be some of these apples. However, one should learn
developed in three directions. how to use this genetic potential in commercial
Firstly, a real economic threshold of damage orchards.
is needed. In conjunction with forecasts of aphid The cost of ladybird larvae is high so their use
abundance, it will enable technical advisers and for aphid control is expensive. One way of
farmers to make more effective decisions- and is reducing the cost is to target the fundatrices
the corner stone of IPM in apple orchards. because fewer larvae are required to achieve an
Secondly, we have to learn more about the efficient predator-prey ratio. Time series analyses
weak points in the D. plantaginea life cycle and of suction trap catches show a cyclical pattern in
how to use this knowledge to control the the variation in abundance of D. plantaginea in
abundance of this aphid. For example, field France. Years with the lowest catches of aphids
observations (BONNEMAISON 1959) and time are separated by 6 to 4 years of higher catches
series analysis of the suction trap catches (HEMPTINNE et al. in press; Fig. 4). If one could
(HEMPTINNE et al. in press) indicate that mortality predict aphid outbreak years the cost of releasing
in autumn is important and influences the number natural enemies could be spread over more than
of aphids next spring. Would it be possible to one year. Currently analysis of suction trap
increase autumn mortality? According to two catches has revealed that the abundance of D.
recent studies the answer to this question might plantaginea is regulated by density-dependence
be positive. WYSS et al. (1995) showed that with weather acting as a disturbing factor
spider webs caught many migrants. The more (HEMPTINNE et al. in press). We do not know
spider webs the more aphids caught and the fewer what regulates its abundance and cannot therefore
fundatrices next year. Release of predators seems devise a reliable forecasting system. However, the
also to be efficient in autumn. The impact of results provide working hypotheses for field and
larvae of A. bipunctata, released before mid- laboratory experiments, the objective of which is
October, on sexuparae and gynoparae to identify the regulating mechanism.
37
3,5
efficient in autumn when D. plantaginea returns
to the apple. If it is decided to release predators,
log10(tot number of aphids
3,0 then it should be larvae rather than adults because

2,5
they tend to stay in or in the vicinity of patches of
prey where they are released. A farmer that
2,0 introduces ladybird larvae in his orchard behaves
1,5 like a gravid ladybird female, carefully selecting
prey colonies where its larvae will later develop.
1,0
However, this method of biological control is
78 83 88 93 98
Year expensive because the price per predator larva is
high and large numbers are required to treat an
Fig. 4. The logarithm of the numbers of Dysaphis
orchard.
plantaginea caught every year from 1978 to 1999 in the
suction trap at Rennes (Brittany, France). Like Janus biological control by natural
enemies has two faces. Schematically and in the
evolutionary framework, natural enemies tend to
aggregate where prey or hosts are abundant, that
CONCLUSION
is in the more profitable patches. As their
abundance gradually increases there they compete
Conservation and enhancement of natural for the resource and they interfere with each
enemies is an attractive strategy for controlling other. Good patches gradually lose their value and
aphids, which has been tested in apple orchards. natural enemies begin to leave and search for
However, the results are not consistent. The better patches. Two opposing forces are at work:
problem is twofold. Firstly, it rests on the idea attraction to resources and repulsion (BEGON et
that diversity creates stability. Undoubtedly al. 1996). The first force is emphasized in
accumulating evidence indicate that diverse biological control strategies, while the second is
systems have interesting properties that are neglected. Intercrops, cover crops or hedgerows
already visible when organic fields are compared provide more resources and support more
with conventionally managed fields (MÄDER et consumers. But how do consumers distribute
al. 2002). Unfortunately we still do not know how themselves between patches of resource and what
these properties are generated and it is therefore is their impact on resource abundance?
difficult to engineer a system to achieve desirable To answer such a question, one needs to focus
properties (HINDMARCH & PIENKOWSKI 2000). on the ecology of the protagonists, which in this
Secondly, stability does not necessarily mean case are aphids and ladybird beetles. Aphids are
regulation of a population in the sense of keeping particularly adapted to exploit transient and
its abundance under a threshold of economic ephemeral resources, the sap flow rich in
damage. The main weakness of conserving and nitrogen, which only occurs in spring and in
enhancing natural enemies in order to control autumn (DIXON 1998). Their life history is
pests was pointed out by RUSSELL in 1989: the summarized by a short motto: “going fast”.
almost complete lack of field studies “on the Ladybird beetles evolved the ability of exploiting
behavior of individual arthropod enemies”. If we such prey. Their reproductive behaviour, as
do not correct this, techniques of biological described for A. bipunctata, is an expression of
control conceived from theory built on their adaptation. It clearly shows the two forces
speculation will continue to yield inconsistent mentioned above in action. Ladybird females are
results. attracted to aphid colonies but they also avoid
From a purely technical point of view, it is those colonies that are marked by the tracks left
possible to control D. plantaginea using natural by conspecific larvae. As a result, female A.
enemies. Experiments have revealed that larvae bipunctata tend to distribute their eggs in many
of A. bipunctata can locate and kill fundatrices of colonies of prey and show a weak numerical
this aphid early in spring even when the weather response to aphid abundance (HEMPTINNE et al.
is cold and humid. They are also active and 1992). Field work aimed at understanding how
38
ladybird beetles behave in strip managed orchards BROWN, M.W., E. NIEMCZYK, T. BAICU, K. BALAZS, V.
and at assessing their impact on more diverse JAROSIK, G. JENSER, F. KOCOUREK, R. OLSZAK, A.
food sources is needed. SERBOIU, & T. VAN DER ZWET 1997. Enhanced
biological control in apple orchards using ground
covers and selective insecticides: an international
REFERENCES study. Zahradnictvi - Horticultural Sciences 24:
35-37.
AGARWALA, B.K. & A.F.G. Dixon 1992. Laboratory CRUZ DE BOELPAEPE, M.O., N.M. FILIPE & V.C.
study of cannibalism and interspecific predation in AFONSO 1987. Dynamique des populations
ladybird. Ecological Entomology 17: 303-309. aphidiennes en verger de pommiers. Leurs
ANDOW, D.A. 1986. Plant diversification and insect fluctuations saisonnières dans la province de Beira-
population control in agroecosystems. Pp. 277-386 Baixa (Portugal) en fonction des facteurs de
in D. PIMENTEL (Ed.) Some aspects of integrated l’environnement. La Défense des Végétaux 246: 5-
pest management. Ithaca: Cornell University Press. 16.
ANDOW, D.A. 1988. Management of weeds for insect DE BERARDINIS, E., P. BARONIO & J. BAUMGÄRTNER
manipulation in agroecosystems. Pp. 265-301 in M. 1994. The effect of aphid (Dysaphis plantaginea
A. ALTIERI & M. LIEBMAN (Eds). Weed Pass., Hom., Aphididae) feeding on apple fruit
management in agroecosystems: ecological growth. Ecological Modelling 72: 115-127.
approaches. Boca Raton, CRC Press. DEDIEU, J.-L. 1998. Les pucerons cendrés en 1998 :
BARONIO, P., G. BRIOLINI, A. BUTTURINI & G. FACCIOLI année atypique ou nouvelle situation. La Letttre du
1988. Dinamica di popolazioni dell’afide grigio del Cercle Arbo 11: 1.
melo (Dysaphis plantaginea Pass.). 1. Ricerche DELORME, R., D. Auge, P. Touton, C. Vergnet & F.
etologiche su ospite secondario. Bolletin Istituto Villatte 1998. La résistance des pucerons aux
Entomologico « G. Grandi” Università di Bologna. insecticides. Enquête 1997. Pp. 375-384 in 1er
43: 9-16. Colloque transnational sur les luttes biologiques,
BASSINO, J.P. 1982. Les pucerons des arbres fruitiers. intégrée et raisonnée. Bilan et perspectives au
Travaux récents et perspectives. Pp. 237-242 in Les regard d’expériences régionales et européennes.
Pucerons des cultures. Journée d'Etudes et Lille. 576 pp.
d'Information. ACTA: Le Carroussel. DIAZ, S. & M. CABIDO 2001. Vive la difference: plant
BEGON, M., J.L. HARPER, C.R. Townsend 1996. functional diversity matters to ecosystem processes.
Ecology. Individuals, populations and communities TRENDS in Ecology & Evolution, 16: 646-655.
(3rd edition). Oxford: Blackwell Science. 1068 pp. DIXON, A.F.G. 1998. Aphid ecology. An optimization
BLOMMERS, L.H.M. 1994. Integrated pest management approach. London: Chapman & Hall. 300 pp.
in European apple orchards. Annual Review of DIXON, A.F.G. 2000. Insect predator-prey dynamics.
Entomology 39: 213-241. Ladybird beetles & biological control. Cambridge:
BONNEMAISON, L., 1959. Le puceron cendré du Cambridge University Press. 257 pp.
pommier (Dysaphis plantaginea Pass.). DIXON, A.F.G. & J.-L. HEMPTINNE 2003. Ladybirds and
Morphologie et biologie - Méthodes de lutte. the biological control of aphid populations. Pp. 1-
Annales des Epiphyties 10: 257-320. 10 in A.O. SOARES, M.A. VENTURA, V. GARCIA &
BOUCHARD, D., S.B. Hill & J.G. Pilon 1988. Control of J.-L. HEMPTINNE (Eds) 2003. Proceedings of the
green apple aphid populations in an orchard 8th International Symposium on Ecology of
achieved by releasing adults of Aphidoletes Aphidophaga: Biology, Ecology and Behaviour of
aphidimyza (Rondani) (Diptera: Cecidomyiidae). Aphidophagous Insects. Arquipélago. Life and
Pp. 257-260 in E. NIEMCZYK & A.F.G. DIXON Marine Sciences. Supplement 5: x + 112 pp.
(Eds). Ecology and Effectiveness of Aphidophaga. DOUMBIA, M., J.-L. HEMPTINNE & A.F.G. DIXON 1998.
The Hague: SPB Academic Publishing. 341 pp. Assessment of patch quality by ladybird: role of
BROWN, M.W. & W.V. WELKER 1992. Development of larval tracks. Oecologia 113: 197-202.
the phytophagous arthropod community on apple as ELTON, C.S. 1958. The ecology of invasions by Animals
affected by orchard management. Environmental and Plants. London: Methuen.
Entomology 21: 485-492. GENDRIER, J.P., Y. MONNET, C. MAILLET & H.
BROWN, M.W. & J.J. SCHMITT 1996. Impact of ground AUDEMARD 1989. Les arthropodes ravageurs :
cover plants on beneficial arthropods in an apple situation et lutte. L’Arboriculture fruitière 420: 33-
orchard in West Virginia, USA. IOBC/WPRS 40.
Bulletin 19: 332-333. GRAF, B., H. HOEHN & H.U. HOEPLI 1999. Optimizing
39
the control of rosy apple aphid Dysaphis Biological Agriculture and Horticulture 15: 233-
plantaginea (Pass.) (homoptera: Aphididae). in F. 240.
POLESNY & B. GRAF (Eds). IOBC workshop on KINDLMANN, P. & A.F.G. DIXON 1993. Optimal
Arthropod Pest Problems in Pome fruits. Bulletin foraging in ladybird beetles (Coleoptera:
IOBC/WPRS. 22(5): 71-76 Coccinellidae) and its consequences for their use in
GRASSWITZ, T. & E.C. BURTS 1995. Effect of native biological control. European Journal of
natural enemies and augmentative releases of Entomology 90: 443-450.
Chrysoperla rufilabris Burmeister and Aphidoletes KORICHEVA, J., C.P.H. MULDER, B. SCHMID, J. JOSHI &
aphidimyza (Rondani) on the population dynamics K. HUSS-DANELL 2000. Numerical responses of
of the green apple aphid, Aphis pomi DeGeer. different trophic groups of invertebrates to
International Journal of Pest Management 41: 176- manipulations of plant diversity in grassland.
183. Oecologia 125: 271-282.
HAGLEY, E.A.C. 1989. Release of Chrysoperla carnea KREBS, C.J. 1994. Ecology. The experimental analysis
Stephens (Neuroptera: Chrysopidae) for control of of distribution and abundance (4th edition). Menlo
the green apple aphid, Aphis pomi DeGeer Park: HarperCollins College Publishers. 801 pp.
(Homoptera: Aphididae). The Canadian LANDIS, D.A., S.D. WRATTEN & G.M. GURR 2000.
Entomologist 121: 309-314. Habitat management to conserve natural enemies of
HAIRSTON, N.G., J.D. ALLEN, D. COLWELL, D.J. arthropod pests in agriculture. Annual Review of
FUTUYMA, J. HOWELL, M.D. JUBIN, J. MATHIAS & Entomology 45: 175-201.
J.H. VANDERMEER 1968. The relationship between MÄDER, P., A. FLIEßBACH, D. DUBOIS, L. GUNST, F.
species diversity and stability: An experimental PADRUOT & U. NIGGLI 2002. Soil fertility and
approach with protozoa and bacteria. Ecology 49: biodiversity in organic farming. Science 296: 1694-
1091-1101. 1697.
HEMPTINNE, J-L., A.F.G. DIXON & J. COFFIN 1992. MAY, R.M. 1973. Stability and complexity in model
Attack strategy of ladybird beetles (Coccinellidae): ecosystems. Princeton: Princeton University Press.
factors shaping their numerical response. 265 pp.
Oecologia 90: 238-245. OBRYCKI, J.J. & T.J. KRING 1998. Predaceous
HEMPTINNE, J-L., A. MAGRO, A. MAUREAU, M. HULLÉ coccinellidae in biological control. Annual Review
& A.F.G. DIXON in press. Development of a of Entomology 43: 295-321.
forecasting system for the integrated pest PASQUALINI, E. & G. BRIOLINI 1982. Stratégies de lutte
management of Dysaphis plantaginea. Proceeding contre les pucerons du pommier. Pp. 251-252 in
of the symposium Aphids in the New millennium. Les Pucerons des cultures. Journée d'Etudes et
Rennes: INRA Editions. d'Information. ACTA: Le Carroussel.
HINDMARCH, C., & M. PIENKOWSKI 2000. Land PIMM, S.L. 1984. The complexity and stability of
management : the hidden costs. The British ecosystems. Nature 307: 321-326.
Ecological Society, Ecological Issues Series. ROOT, R.B. 1973. Organization of a plant-arthropod
Oxford: Blackwell Publishing. 60 pp. association in simple and diverse habitats: The
HODEK, I. 1973. Biology of coccinellidae. The Hague: fauna of collards (Brassica oleracea). Ecological
Dr. W. Junk Editor and Praha: Academia. 260 pp. Monographs 43: 95-124.
HODEK, I. & A. HONċK 1996. Ecology of Coccinellidae. RUSSEL, E.P. 1989. Enemies hypothesis: A review of
Dordrecht: Kluwer Academic Publishers. 464 pp. the effect of vegetational diversity on predatory
HOEHN, H., H.U. HOEPLI & B. GRAF 1996. Mehlige insect and parasitoids. Environmental Entomology
Apfelblattlaus: Befallsentwicklung und 18: 590-599.
Bekaempfung. Schweizerisch Zeitschrift Obst- und SAVINI, J.R. 2000. Problématique puceron cendré.
Weinbau 132: 60-62. Réussir Fruits & Légumes 182: 52-54.
KEHRLI, P. & E. WYSS 2001. Effects of augmentative SOLOMON, M., J. FITZGERALD & R. JOLLY 1999.
releases of the coccinellid, Adalia bipunctata, and Artificial refuges and flowering plants to enhance
of insecticide treatments in autumn on the spring predator populations in orchards. IOBC/WPRS
population of aphids of the genus Dysaphis in apple Bulletin 22: 31-38.
orchards. Entomologia Experimantalis at Applicata TOURNEUR, J.C., D. BOUCHARD & J.-G. PILON 1992. Le
99: 245-252. complexe des enemies naturels des pucerons en
KIENZLE, J., C.P.W. ZEBITZ & S. BRASS 1997. Floral pommeraie au Québec. Pp. 179-193 in C. VINCENT.
and faunal species diversity and abundance of & D. CODERRE (Eds). La lutte biologique. Québec:
aphid predators in ecological apple orchards. G. Morin éditeur. 671 pp.
40
VAN DRIESCHE, R.G., & T.S. BELLOWS JR. 1996. Einsatzes von Pflanzenbehandlungsmitteln. Institut
Biological control. New York: Chapman & Hall. Transfrontalier d’Application et de Développement
539 pp. Agronomique. 23 pp.
VAN EMDEN, H.F. 1990. Plant diversity and natural WYSS, E. 1995. The effects of weed strips on aphids
enemy efficiency in agroecosystems. Pp 63-80 in and aphidophagous predators in an apple orchard.
M. MACKAUER, L.E. EHLER & J. ROLAND (Eds). Entomologia Experimentalis et Applicata 75: 43-
Critical issues in Biological Control. Andover: 49.
Intercept. 330 pp. WYSS, E., U. NIGGLI & W. NENTWIG 1995. The impact
VOGT, H. & A. WEIGEL 1999. Is it possible to enhance of spiders on aphid populations in a strip-managed
biological control of aphids in an apple orchard apple orchard. Journal of Applied Entomology 119:
with flowering strips? IOBC/WPRS Bulletin 22: 39- 473-478.
46. WYSS, E., M. VILLIGER J.-L. HEMPTINNE & H. MÜLLER-
WHALON, M.E., & B.A. CROFT 1984. Apple IPM SCHÄRER 1999a. Effects of augmentative releases
implementation in North America. Annual Review of eggs and larvae of the ladybird beetle, Adalia
of Entomology 29: 435-470. bipunctata, on the abundance of the rosy apple
WÜRTH, M., H. BENTZ, J. GUIOT, F. WEIBEL & M. aphid, Dysaphis plantaginea, in organic apple
LITTERST 1999. Abschlussbericht zum ITADA- orchards. Entomologia Experimentalis et Applicata
Projekt A3.3: Obstbau. Institut Transfrontalier 90: 167-173.
d’Application et de Développement Agronomique. WYSS, E., M. VILLIGER & H. MÜLLER-SCHÄRER 1999b.
42 pp. The potential of three native insect predators to
WÜRTH, M., J. GUIOT, H. BENTZ, R. KEIM & F. WEIBEL control the rosy apple aphid, Dysaphis plantaginea.
2002. Abschlussbericht des Arbeitsprogramm BioControl 44: 171-182.
ITADA IIbis Projekt 2-1-4: Prüfung
krankheitsresistenter neuer Apfelsorten für Tafel–
und Industrieobst mit dem Ziel der Reduzierung des Accepted 31 May 2003.
41
42
FITNESS OF FIVE PHENOTYPES OF Harmonia axyridis PALLAS

(COLEOPTERA: COCCINELLIDAE).
L. SERPA, H. SCHANDERL, C. BRITO & A. O. SOARES
SERPA, L., H. SCHANDERL, C. BRITO & A.O. SOARES 2003. Fitness of five
phenotypes of Harmonia axyridis Pallas (Coleoptera: Coccinellidae). Pp. 43-49 in
A.O. SOARES, M.A. VENTURA, V. GARCIA & J.-L. HEMPTINNE (Eds) 2003.
Proceedings of the 8th International Symposium on Ecology of Aphidophaga:
Biology, Ecology and Behaviour of Aphidophagous Insects. Arquipélago. Life
and Marine Sciences. Supplement 5: x + 112 pp.
H. axyridis is a highly polymorphic ladybird beetle with phenotypes that vary in their
elytral patterns. In natural populations the relative frequency of phenotypes is not constant,
but is subject to a strong geographic and temporal variation. The selective forces that are
the basis for those variation are diverse and difficult to determine. In this work we suggest
that fitness of phenotypes could contribute to these variations. That is, a specific genotype
could also confer a specific reproductive capacity. We compare the reprodutive capacity of
adults of five phenotypes [succinea (h) or s0, succinea-3 (h3) or s9, conspicua -1 (hC1) or
c1, aulica-1 (hA1) or au and nigra [? (h?)] or ni]. Our results show that under the same
biotic and abiotic conditions, there are differences in biological parameters of the
phenotypes that affect their fitness. The rare phenotype nigra, and the relatively uncommon
aulica have lower reproductive capacity than the most frequent phenotypes succinea and
conspicua.
Luís Serpa, H. Schanderl, C. Brito & A.O. Soares (e-mail: onofre@notes.uac.pt),

Department of Biology, University of the Azores, Rua da Mãe de Deus, Apart. 1422 – PT-
9501-801 Ponta Delgada, Azores, Portugal.
INTRODUCTION remains fairly constant. The relative frequency of

phenotypes is genetically determined as the result
Fitness is defined as a measure of the of mating between individuals of the same
reproductive success of a genotype relative to populations. The phenotypes will persist as long
another in the same environment. It measures the as different genotypes are favorably selected in
capacity of a genotype to transmit its genes to the different parts of the ecosystem or at different
next generation (BEGON et al. 1990). The direct times (BUSKIRK et al. 1997; RICKLEFS 1990).
evaluation of fitness has rarely been performed This phenomenon is called genetic polymorphism
(HOFFMAN 1994). Generally, researchers measure (MAJERUS 1994; FUTUYMA 1998). The genetic
fitness indirectly by using biological parameters polymorphism determines the existence of
such as viability (survival, fecundity and fertility). geographical races, subspecies, seasonal variation
If genotypes differ in their viability then the in the relative frequency of phenotypes, and the
relative frequency of genotypes will differ in the occurrence of rare phenotypes, which otherwise
next generation. For this reason, one simple would be eliminated from populations (MAJERUS
mesure of fitness is considered to be the number 1994).
of offspring produced by a specific genotype Harmonia axyridis Pallas, like other ladybird
relative to another (FUTUYMA 1998). beetle species, shows a highly polymorphic
Various natural populations of insects are elytral pattern. Some of its phenotypes were
made up of individuals of different phenotypes, previously considered different species and/or
more or less similar, whose relative frequency different genera (KOMAI 1956). The phenotypes
43
are classified in two main groups; (i) the succinea changes in the relative frequency of elytral
group, or light forms, characterized by the phenotypes in a population at Suwa, Japan. This
presence of elytras with color varying between author suggested that these change was due to
shiny red and brown, and by the presence of a natural selection, probably in relation to the
variable number of spots between 0 and 19, and milder winters in Suwa in more recent years.
(ii) the dark or melanic group, characterized by Those observations suggest that
the presence of red spots on a dark background. polymorphism in H. axyridis can be adaptive and
The melanic group includes the phenotypes that the relative frequency of phenotypes in
conspicua, spectabilis and aulica (TAN 1946, different populations could be related to their
1949; KOMAI 1956). differing fitnesses in different environments.
Different populations of H. axyridis show Recently, we showed that the nigra and aulica
considerable geographic variation in the relative phenotypes of H. axyridis differ in their fitness
frequency of phenotypes (so called micro- (SOARES et al. 2001). Nevertheless remained to
geographic variation). In China and Manchuria, be tested the possibility that the most common
the phenotypes succinea, conspicua and phenotypes have highest fitnesses. We
spectabilis are very frequent (TAN 1946, 1949; hypothesized that (i) a particular genotype that
KOMAI 1956). In Japan, conspicua and spectabilis confers specific coloration pattern could also
phenotypes are the most frequent while in Korea confer other specific attributes, as stated by
the succinea phenotype predominates LAMANA & MILLER (1995) and (ii) the most
(DOBZHANSKY 1933; KOMAI 1956; OSAWA & abundant phenotypes in Asian populations should
NISHIDA 1992). On the other hand the phenotype show a higher fitness than the less abundant and
aulica is at low frequency or is even absent in rare phenotypes. We therefore predicted that the
certain regions (DOBZHANSKY 1933). Very rare rare phenotype nigra [? (h?)], should show a
phenotypes such as corvine may also occur in lower performance than the less frequent, aulica-
natural populations (MADER 1932 in TAN 1946). 1 (hA1), or the most abundant succinea (h),
Some attempts have been made to relate succinea-3 (h3) and conspicua-1 (hC1).
elytral phenotypes in H. axyridis with
geographical and seasonal factors (KRYLTSOV MATERIAL AND METHODS
1956; ABBAS & NAKAMURA 1985; ABBAS et al.
1988; OSAWA & NISHIDA 1992; HODEK & Biological material
HONċK 1996). However, the precise selective
forces associated with the phenomenon are still The phenotypes of H. axyridis individuals came
very difficult to determine (MUGGLETON 1978). from mass rearing at 25 ± 1 °C, 75 ± 5 % RH and
Macro-geographical variation of Asian photoperiod of 16L:8D. Coccinellids were fed on
populations of H. axyridis is apparently related to a diet of Aphis fabae Scopoli and eggs of
climatic factors, whereas micro-geographical Ephestia kuehniella Zeller. Field collected
variation, that is, differences in the relative individuals of H. axyridis were added regularly to
frequency of elytral phenotypes between host minimize inbreeding.
plants in the same habitat, appear mostly related We used two phenotypes classified in the
to food availability (KOMAI & HOSINO 1951; succinea group, succinea (h) or s0 and succinea-3
KOMAI 1956). Seasonal variation in the (h3) or s9 and three classified in the melanic
proportion of dark and light forms in Japanese group, conspicua-1 (hC1) or c1, aulica-1 (hA1) or
populations has also been observed (TAN 1949; au and nigra [? (h?)] or ni. These forms occurred
OSAWA & NISHIDA 1992). OSAWA & NISHIDA naturally in our mass rearing.
(1992) demonstrated a significant increase in the Biology of adults: longevity, body-weight and
relative frequency of light phenotypes during reproductive parameters
spring and early summer. This result, supported
by laboratory experiments, suggest the occurrence In order to evaluate the biology of adults of
of assortative mating between different different phenotypes we sexed and paired 200
phenotypes. KOMAI (1956) reported long-term individuals of five phenotypes, thereby creating
44
20 couples for each phenotype. Each couple was the contribution of the latter to the statistical
isolated in a 60 ml Petri dish Mino-Gaillard (Ø: 5 separation of phenotypes based on their biological
cm, height: 3 cm). A daily surplus of apterous A. characteristics.
fabae females was provided. Egg clusters were
removed from Petri dishes every day. All
individuals were monitored until they death. The RESULTS
experiment was performed at 25 ± 1 °C, 75 ± 5 %
RH and a photoperiod of 16L:8D, under
fluorescent lamps (Sanyo FL 40 SS W/37). Biological attributes of the elytral phenotypes,
Different biological features were evaluated: including longevity and reproductive capacity of
longevity, percentage of fertile couples, number adults, are summarized in Table 1. The longevity
of egg clusters, number of days of oviposition, of nigra females and males was significantly less
total and daily fecundity, daily fertility (including than the other phenotypes. Longevity did not
sibling cannibalism), percentage of hatching, differ between aulica and succinea (s0 e s9)
sterile and embrionate eggs (3 days after the last males, which lived longer on average than
hatched eggs), sex ratio and body-weight. Body- conspicua males. No significant differences were
weight was determined for the parental generation found among longevities of conspicua, succinea
(P), in the first day after emergence and in the (s0) and aulica females. The lowest percentage of
first day of sexual maturation, and also for a pool fertile couples was observed in the nigra
including both parental and progeny of the same phenotype. The nigra and succinea (s9) couples
elytra, in the first day after emergence. Adults showed a significant lower fecundity than the
were weighed to an accuracy of 10-4 mg on a other phenotypes (Table 1). The nigra couples
Mettler AM 50 analytical balance. Longevity, showed the lowest value for daily fertility
fecundity, fertility (including sibling cannibalism) (including sibling cannibalism). That is, females
and percentage of hatching were compared. of this phenotype produced fewest larvae per day,
Proportions were arcsin¥% transformed (ZAR followed by females of aulica phenotypes. No
1984). One-factor ANOVA was used to compare significant differences in daily fertility were
longevity, fecundity, fertility (including sibling observed between succinea s0 and succinea s9.
cannibalism), percentage of hatching and body The highest fertility values were observed in
weight. All statistical tests were done using JMP conspicua couples (Table 1). Nigra and aulica
(SALL et al. 2001). The Ȥ2 test was used to females showed the lowest hatching percentage of
compare the sex ratio in each phenotype. eggs. The highest hatching percentage was
Raw multivariate data were arranged in a observed in conspicua couples. Aulica, nigra and
square matrix, assigning individuals to rows and succinea s9 showed no significant differences in
biological features (variables) to columns, and terms of percentage of embrionate eggs, nor were
standardised (ROHLF 1994; FIELD et al. 1982). there differences between conspicua and succinea
Individuals were compared using euclidean s0. We observed the highest percentage of sterile
distances arranged in a symmetrical matrix of eggs for the nigra couples, followed by the aulica
association (ROHLF 1994; LEGENDRE & and succinea s9 couples. No significant
LEGENDRE 1984; CLIFFORD et al. 1975), followed differences were observed between succinea s0
by a cluster analysis, using Unweighted Pair- and conspicua (Table 1). The proportion of
Group Arithmetic Average (UPGMA). Principal females obtained from conspicua (Ȥ2=0.093, df=1,
component analysis (PCA) was used, in order to p=0.76) and aulica (Ȥ2=39.68, df=1, p<0.0001)
reduce the dimensionality of the data, to a few couples was significantly higher a 50:50 sex ratio.
artificial variables (principal components), However, no significantly differences from a
linearly uncorrelated and arranged in order of 50:50 sex ratio were obtained in nigra (Ȥ2=0.093,
importance, in terms of the variance accounted df=1, p=0.76), succinea (s0) (Ȥ2=0.477, df=1,
for. The final plot of both phenotypes and p<0.489) and succinea s9 (Ȥ2=0.253, df=1,
biological parameters permitted us to visualize p<0.614) (Table 1).
45
Table 1
Comparations of longevity, body-weight and reproductive attributes of five phenotypes of H. axyridis [conspicua
(c1), aulica (au), nigra (ni) and succinea (s9 and s0)]. Means and standard errors followed by different letters
indicate post-hoc significantly differences. Last column indicate F values, degrees of freedom and p value of
ANOVA.
Phenotypes
Biological parameters conspicua (c1) aulica (au) nigra (ni) succinea (s9) succinea (s0)
Longevity (days)
Males (LM) 66.6±6.3c 76.3±5.4b 43.6±3.4a 75.3±7.4b 79.7±6.8b F=5.89, df=4, 94, P=0.0003
Females (LF) 62.2±6.5b 60.3±4.2b 30.7±3.1a 52.2±5.5c 64.8±4.5b F=8.04, df=4, 94, P<0.0001
Percentage of fertile couples (FrC) 95.0 85.0 55.0 85.0 90.0
Number of batches (Bt) 33.5±4.1b 31.1±2.7b 11.4±1.4a 24.0±2.4c 34.9±3.4b F=10.8, df=4, 94, P<0.0001
Number of day of oviposition (Ov) 26.8±3.1b 25.4±2.0b 10.7±1.3a 18.7±1.8c 27.7±2.6b F=10.3, df=4, 94, P<0.0001
Total fecundity (TFc) 885.3±96.6b 808.3±85.1b 311.8±47.7a 579.5±72.9c 815.8±90.8b F=8.58, df=4, 94, P<0.0001
Daily fecundity (DFc) 34.4±2.0a 31.1±1.8a 27.3±2.1a 31.0±2.3a 28.0±1.8a F=2.06, df=4, 94, P=0.0925
Daily fertility (DFr) 12.8±2.1d 4.0±1.0b 0.56±0.3a 7.1±2.0c 8.1±1.3c F=9.35, df=4, 94, P<0.0001
Percentage of hatching (Ha) 47.2±2.7c 22.8±2.7b 2.3±0.5a 26.3±2.9b 39.4±2.7c F=72.7, df=4, 74, P<0.0001
Percentage of embrionates (Em) 7.6±0.7b 5.8±0.7a 5.6±0.7a 5.6±0.9a 7.4±0.7b F=6.02, df=4, 74, P<0.0001
Percentage of sterile (St) 45.1±2.9a 71.2±3.4b 94.1±0.9c 68.2±3.3b 53.2±9.9a F=39.9, df=4, 74, P<0.0001
Sex ratio (females/males) (Sr) 1.11 1.62 1.09 0.96 1.03
Body-weight after emergence (mg)
Males (W1M) 25.8±0.7a 27.4±0.6a 26.2±0.6a 25.2±0.6a 26.3±0.3a F=2.07, df=4, 99, P=0.091
Females (W1F) 32.1±1.2a 32.3±0.6a 31.3±0.6a 34.5±0.7a 32.2±0.4a F=2.77, df=4, 99, P=0.310
Body-weight in the first day of sexual maturation
Males (W2M) 31.7±0.8a 33.7±0.7a 31.8±0.6a 31.3±0.8a 32.3±0.3a F=1.97, df=4, 99, P=0.105
Females (W2F) 44.3±1.6a 44.2±1.4a 44.6±0.9a 43.5±1.1a 43.9±0.6a F=0.21, df=4, 99, P=0.935
Body-weight after emergence (P+F)
Males (W3M) 26.2±0.4a 33.2±0.4b 25.8±0.4a 26.9±0.3a 26.6±0.3a F=62.65, df=4, 458, P<0.0001
Females (W3F) 32.8±0.5a 38.2±0.5b 31.5±0.6a 32.3±0.4a 33.7±0.4a F=31.11, df=4, 468, P<0.0001
Among the three estimates of body weight, correspond to parameters related to longevity,
significant differences are evident between males fecundity and fertility, which could be associated
and females. Only the pooled estimate, included to a general fitness dimension. On the other hand,
weights of both parents and progeny of the same the main contributions to axis 2 correspond to sex
elytral genotype, showed a significant difference ratio of the offspring and body weight, with
among phenotypes: both females and males were special importance of both sexes for W3, and
heavier for aulica than for other phenotypes. males W1 and W2. While the first axis clearly
The UPGMA dendogram using of euclidean separates the phenotype nigra from the other
distances, based on standardized biological data, phenotypes, axis 2, discriminates phenotype
cluster C1, S0 and S9, apart from both aulica and aulica, and to lesser extent, phenotype s9, from
nigra phenotypes (Fig. 1). the core group made by c1 and s0 (Fig. 2).
Fig. 1. UPGMA cluster of elytral phenotypes, based

on euclidean distances using standardized biological
parameters. Symbols are those presented in the text
(Material and methods).
This grouping is also supported by the results Fig. 2. PCA ordination was based on biological data of
of principal component analysis (PCA) that H. axyridis. PC1 and PC2 refer to the first principal
reduced the multidimensional data set to two components that explain, respectively, 52.5 % and 29.2
principal axes, which explained 81.7% of the % of the variance. Elytral morphotypes are underlined.
variance. The main contributions for the first axis Symbols correspond with those in Table 1.
46
DISCUSSION 1978) and Olla v-nigrum (Mulsant) (KREITER &

IPERTI 1984). Despite the fact that experimental
temperature was fixed at 25 °C, nigra phenotype
Stability of genetic polymorphism in a population was found to have a relatively short adult
is a consequence of the balance of diverse lifespan. This could be a disadvantage in
selective forces (SHEPPARD 1975). Most of the biological control programs, from a practical
published studies focus on one selective factor point of view, since the decrease in time spent in
thought to be responsible for the success of a foraging activity and ovipositing may reduce its
particular phenotype relative to another. Direct or fitness.
indirect fitness evaluation of phenotypes, under For polymorphic species, such as H. axyridis,
diverse ambient conditions, has rarely been done. the degree of melanization can also be an
Biological control programs generally focus important factor affecting longevity. Melanic
on the selection of the best parasitoid or predator forms can reach a higher body temperature than
species to be introduced based on criteria related non-melanic forms (DE JONG et al. 1996), and
to their capacity to have substantial impact on a consequently they can reach higher metabolic rate
pest population (WAAGE & MILLS 1992). and biological activity (DIGBY 1955; LUSIS 1961;
However, selection of the most fit phenotypes of DIXON 1972; BENHAM et al. 1974; MUGGLETON
a selected species is rarely attempted (SOARES et et al. 1975; WILLMER & UNWIN 1981;
al. 2001). Selection of the best phenotype will BRAKEFIELD & WILLMER 1985; STEWART &
depend on the biological control strategy used. DIXON 1989; DE JONG et al. 1996). Our
Certain biological parameters, such as experiment was performed using low radiation
reproductive capacity, will favor population fluorescent lamps. We therefore doubt that
establishment and predator-prey stabilization and melanism decreased longevity of nigra through
may be most desirable in classical biological elevated body temperature. Instead, we suggest
control programs, whereas short-term efficacy that longevity may be directly determined
may be favored in inundative biological control genetically, and the short lifespan of nigra may
programs. Our study showed that, under our be adaptive when the environment is
experimental conditions, phenotypes of H. characterized by low temperature.
axyridis apparently differed in fitness. The most Indeed macro-geographic variation in Asiatic
abundant phenotypes in nature, succinea and populations of H. axyridis is associated with
conspicua, were also the phenotypes with higher climatic factors, such as temperature, as the result
longevity, fertility, fecundity and percentage of of physiological adaptations. In cold regions the
hatching, in a clear contrast with the rare/less dark phenotypes are, generally, more frequent
abundant phenotypes of nigra and aulica, witch than the light forms (KOMAI 1956).
confirm our hypotheses that is the most abundant Every species of coccinellid has genetically
phenotypes in Asian populations should show a fixed rates of fecundity and egg sizes that are also
higher fitness than the less abundant and rare affected by food availability and environmental
phenotypes. Our results suggest, also, that conditions. If food supply is limited, coccinellids
conspicua and succinea (s0 and s9) individuals decreases their fecundity but will maintain egg
should be used in classical biological control size (HODEK & HONċK 1996). More voracious
programs because they have a greater longevity and heavier coccinelid females have more
and higher reproductive capacity. These resources to invest in egg production (FERRAN et
characteristics favor population adaptation and al. 1984; HODEK 1993; STEWART et al. 1991a, b;
stabilization. SOARES et al. 2001). Observed differences in
Temperature determines longevity in body weight of aulica adults versus adults of
ecothermic species. In coccinellids longevity other phenotypes may reflect the particular
generally decreases with the increase of dietary condition of our experiment. It remains to
temperature. This effect was reported for C. be investigated whether the same differences will
novemnotata Herbst (MCMULLEN 1967), occur in mass rearing programs. Nevertheless,
Coleomegilla maculata lengi (WRIGHT & LAING given (1) the absence of significant differences in
47
female body weight among the parental DOBZHANSKY, TH.G. 1933. Geographical variations in
generations of the different phenotypes and (2) lady-beetles. American Naturalist 709: 97-126.
the significant differences between fecundity and FERRAN, A., M.O. CRUZ DE BOELPAEPE, H. SCHANDERL
fertility, among the various phenotypes, our & M.-M. LARROQUE 1984. Les aptitudes trophiques
et reproductrices des femelles de Semiadalia
results suggest that reproductive capacity of undecimnotata (Col.: Coccinellidae). Entomophaga
phenotypes could be correlated with 29: 157-170.
environmental conditions and prey quality. Thus FIELD, J.C., K.R. CLARKE & R.M. WARWICK 1982. A
additional experiments using other temperatures practical strategy for analyzing multispecies
and prey should be performed. distribution patterns. Marine Ecology-Progress
Series 8: 37-52.
REFERENCES FUTUYMA, D.J. 1998. Evolutionary Biology (3rd ed.).
Sinauer Associates, Inc., Massachusetts. 763 pp.
ABBAS, I. & K. NAKAMURA 1985. Variation of elytral HODEK, I. & A. HONċK 1996. Ecology of Coccinellidae.
spot patterns in a field population of lady beetle Kluwer Academic Publishers, Netherlands. 260 pp.
Epilachna aff. sparsa (Coleoptera, Coccinellidae) HOFFMAN, A.A. 1994. Genetic analysis of territoriality
feeding on bitter cucumber in Sumatra. Scientific of Drosophila melanogaster. Pp. 188-205 in: C.
Report Kanazawa University 30: 27-32. BOAKE (Ed) Quantitative genetic studies of
ABBAS, I., K. NAKAMURA, H. KATAKURA & H. SASAJI behavioral evolution. Chicago University Press,
1988. Geographical variation of elytral spot Chicago. 390 pp.
patterns in the ladybird Epilachna HODEK, I. 1993. Habitat and food specificity in
vigintiduopunctata (Coleoptera, Coccinellidae) in aphidophagous predators. Biocontrol Science and
the province of Sumatrea Barat, Indonesia. Technology 3: 91-100.
Research of Population Ecology 30: 43-56. KOMAI, T. 1956. Genetics of ladybeetles. Advances in
BEGON, M., J.L. HARPER & C.R. TOWNSEND 1990. Genetics 8: 155-189.
Ecology: Individuals, population and communities KOMAI, T. & Y. HOSINO 1951. Contributions to the
(2nd ed.). Blackwell Scientific Publications, Evolutionary Genetics of the Lady-beetle,
London. 945 pp. Harmonia. II. Microgeographic Variations.
BENHAM, B.R., D. LONSDALE & J. MUGGLETON 1974. Genetics 36: 382-390.
Is polymorphism in two-spot ladydird an example KREITER S. & G. IPERTI 1984. Étude des potentialités
of non-industrial melanism? Nature London 249: biologiques et écologiques dún prédateur
179-180. aphidiphage Olla-v-nigrum Muls. (Coleoptera;
BRAKEFIELD, P.M. & P.G. WILLMER 1985. The basis of Coccinellidae) en vue de son introduction en
thermal-melanism in the ladybird Adalia France. 109e Congrès National des Sociétés.
bipunctata: differences in reflectance and thermal Savantes, fasc. II: 275-282.
properties between the morphs. Heredity 54: 9-14. KRYLTSOV, A.I. 1956. Geographical variability of lady-
BUSKIRK, J.V., S.A. MCCOLLUM & E.E. WERNER 1997. birds (Coleoptera, Coccinellidae) in north
Natural selection for environmentally induced Kirghisia. Entomological Obozr 35: 771-781.
phenotypes in tadpoles. Evolution 51: 1983-1992. LAMANA, M.L. & J.F. Miller 1995. Temperature-
CLIFFORD, H.T. & W. STEPHENSON 1975. An dependent development in polymorphic lady bettle,
Introdution to Numerical Classification. Academic Calvia quatuordecimguttata (Coleoptera,
Press. New Press. 229 pp. Coccinellidae). Annals of the Entomological
DE JONG, P.W., S.W.S. GUSSEKLOO & P.M. Society of America 88: 785-790.
BRAKEFIELD 1996. Differences in thermal balance, LEGENDRE, L. & P. LEGENDRE 1984. Écologie
body temperature and activity between numérique 2: La structure des donneés ecologiques
non-melanic and melanic two-spot ladybird beetles (2nd ed.). Masson Presse Université du Québec. 254
(Adalia bipunctata) under controlled conditions. pp.
Journal Experimental Biologie 199: 2655-2666. LUSIS, J.J. 1961. On the biological meaning of colour
DIGBY, P.S.B. 1955. Factors affecting the temperature polymorphism of ladybeetle Adalia bipunctata L.
excess of insects in sunshine. Journal Experimental Latvijas Entomologs 4: 3-29.
Biologie 32: 279-298. MAJERUS, M.E.N. 1994. Ladybirds. The New
DIXON, A.F.G. 1972. Control and significance of the Naturalist Library, Harper Collins, London. 367 pp.
seasonal development of colour forms in the MCMULLEN, R.D. 1967. The effects of photoperiod,
sycamore aphid, Drepanosiphum platanoides temperature, and food supply on rate of
(Schr.). Journal of Animal Ecology 41: 689-697. development and diapause in Coccinella
48
novemnotata. The Canadian Entomologist 99: 578- STEWART, L.A. & A.F.G. DIXON 1989. Why big
586. species of ladybird beetles are not melanics.
MUGGLETON, J. 1978. Selection against the melanic Functional Ecology 3: 165-177.
morphs of Adalia bipunctata L. (two-spot STEWART, L.A., A.F.G. Dixon, Z. RĤžiþka & I. Iperti
ladybird): a review and some new data. Heredity 1991a. Clutch and egg size in ladybird beetles.
40: 269-280. Entomophaga 36: 329-333.
MUGGLETON, J., D. Lonsdale & B.R. Benham 1975. STEWART, L.A., J.-L. HEMPTINNE & A.F.G. DIXON
Melanism in Adalia bipunctata L. (Col., 1991b. Reproductive tactics of ladybird beetles:
Coccinellidae) and its relationship to atmospheric relationships between egg size, ovariole number
pollution. Journal Applied Ecologie 12: 451-464. and developmental time. Functional Ecology 5:
OSAWA N. & T. NISHIDA 1992. Seasonal variation in 380-385.
elytral colour polymorphism in Harmonia axyridis TAN, C.C. 1946. Mosaic Dominance in the inheritance
(the ladydird beetle): the role of non-random of colour patterns in the Lady-birds Beetle,
mating. Heredity 69: 297-307. Harmonia axyridis. Genetics 31:195-210.
RICKLEFS, R.E. 1990. Ecology. (3rd ed.). W.H. TAN, C.C. 1949. Seasonal variations of color patterns
Freeman and Company, New York. 896 pp. in Harmonia axyridis. Proceedings of the 8th
ROHLF, F.J. 1994. NTSYS-pc Numerical Taxonomy International Congress of Genetics: 669-670.
System & Multivariate Analysis System Exeter WAAGE, J.K. & N.J. MILLS 1992. Biological control.
Software. New York. Pp. 421-430 in M.J. CRAWLEY (Ed) Natural
SALL, J., A. LEHMAN & L. CREIGTHON 2001. JMP start enemies. Blackwell Scientific Publications, Oxford.
statistics: a guide to statistical and data analysis 592 pp.
using JMP and JMP IN software. Duxbury Press, WILLMER, P.G. & D.M. UNWIN 1981. Field analysis of
Toronto. 656 pp. insect heat budget: reflectance, size and heating
SHEPPARD, P.M. 1975. Natural Selection and Heredity rates. Oecologia 50: 250-255.
(5th ed.). Hutchinson, London. 239 pp. WRIGHT, E.J. & J.E. Laing 1978. The effects of
SOARES, A.O., D. CODERRE & H. SCHANDERL 2001. temperature on development, adult longevity and
Fitness of two phenotypes of Harmonia axyridis fecundity of Coleomegilla maculata lengi and its
(Coleoptera: Coccinellidae). European Journal of parasite Perilitus coccinellae. Proceedings of
Entomology 98: 287-293. Entomological Society of Ontario 109: 33-48.
SPSS Prodution Facility 2001. SPSS, version 11th ed. ZAR, J.H. 1984. Biostatistical analysis (2nd ed.).
LEAD Technologies Inc., Chicago, IL. Prentice-Hall, New Jersey. 718 pp.
49
50
INFLUENCE OF PREY QUALITY ON THE REPRODUCTIVE CAPACITY OF

TWO PHENOTYPES OF Harmonia axyridis PALLAS (COLEOPTERA:
COCCINELLIDAE).
A.O. SOARES, D. CODERRE & H. SCHANDERL
SOARES, A.O., D. CODERRE & H. SCHANDERL 2003. Influence of prey quality

on the reproductive capacity of two phenotypes of Harmonia axyridis Pallas
(Coleoptera: Coccinellidae). Pp. 51-54 in A.O. SOARES, M.A. VENTURA, V.
GARCIA & J.-L. HEMPTINNE (Eds) 2003. Proceedings of the 8th International
+ 112 pp.
H. axyridis is a highly polymorphic ladybird beetle with a wide geographical distribution in

the palearctic region. Genetic polymorphism seems to be the strategy adopted in order to
face different habitats at different times. In this paper we suggest that a genotype that
confers a specific elytral pattern may also confer a particular reproductive capacity when
fed on single diets of essential preys such as A. fabae or M. persicae. We discuss KOMAI &
HOSINO (1951) hypothesis, which predicts that the differences in the relative frequency of
elytral pattern phenotypes in relation to the host plants in a same habitat could be related
with specific composition of aphid populations. Our results revealed that A. fabae
significantly increased reproductive capacity of nigra phenotype without affecting
reproductive capacity of aulica phenotype.
António O. Soares (e-mail: onofre@notes.uac.pt), H. Schanderl, Department of Biology,

University of the Azores, Rua da Mãe de Deus, Apart. 1422, PT-9501-801 Ponta Delgada,
Azores, Portugal & D. Coderre, Département des Sciences Biologiques, Université du
Québec à Montréal, C.P. 8888 Succ. Centre-ville, Montréal, H3C 3P8, Canada.
INTRODUCTION H. axyridis is a euriphagous predator that

prefers aphids (HUKUSIMA & KAMEI 1970;
HUKUSIMA & OHWAKI 1972; OSAWA 1992;
Asian populations of Harmonia axyridis Pallas IABLOKOFF-KHNZORIAN 1982; SCHANDERL et al.
show a highly level of genetically polymorphic, 1988). It also feeds on psillids (FYE 1981),
with more than one genotype occurring in a coccids (MCLURE 1987; HODEK & HONċK 1996),
population. To identify precisely selective forces spider mites (LUCAS et al. 1997) and lepidopteran
that act to maintain different phenotypes of H. eggs (SCHANDERL et al. 1988).
axyridis in the same population is a very difficult A particular phenotype of H. axyridis may
task. Genetic polymorphism should persist only confer different specific morphological,
when different genotypes are selectively favoured physiological and behavioural attributes (TAN
in varying parts of the environment or at different 1946, 1949; KOMAI 1956; FUTUYMA 1998). If so,
times. Thus a single phenotype cannot be equally then the differences in nutritive demands of
fit in all environmental conditions. This fact phenotypes, or in nutritive value of different prey,
challenges our understanding of adaptation. One could affect relative fitnesses of different
of the important issues raised concerns how phenotypes, when fed different prey. In Japan,
environmental heterogeneity facilitates the KOMAI & HOSINO (1951) reported differences in
coexistence of more than one genotype within a the relative frequency of elytral pattern
population (RICKLEFS 1991). phenotypes between host plants in the same
51
habitat (so called micro-geographic variation). Measure of reproductive capacity

They suggested that such differences could be
correlated somehow with differences in the
composition of the aphid populations, and the To evaluate reproductive capacity, we sexed and
paired 40 individuals of each phenotype. Each of
feeding preferences of the phenotypes. However,
the hypothesis was not addressed further until this the twenty pairs was isolated in a 60 ml Petri dish
study. (Ø: 5 cm, height: 3 cm). One of two diets was
provided to 10 couples of each phenotype. The A.
We have shown recently that the nigra and
aulica phenotypes of H. axyridis differ in their fabae and M. persicae were supplied daily in
fitness fed on the same prey (SOARES et al. 2001, excess. Abiotic conditions were 20 ± 1°C, 75 ±
5% RH and a photoperiod of 16L:8D, using
2003), but we do not know if the differences
would remain consistent on other diets. Despite fluorescent lamps (Philips ref.: TDL 23W/54 and
the well-documented polyphagy in H. axyridis, TDL 18W/54). Egg clusters were removed from
Petri dishes twice daily. Fecundity, fertility and
we hypothesise that aulica and nigra phenotypes
will display different reproductive capacity fed on percentage hatch were compared over the first
Aphis fabae Scopoli or Myzus persicae (Sulzer). fifteen days of a females’ reproductive life.
Percentages were arcsin¥% transformed (ZAR
In this study, we evaluate the reproductive
capacity of aulica and nigra fed diets of either A. 1984). One-factor ANOVA was used to compare
fabae or M. persicae. fecundity, fertility and percentage hatch of the
eggs of the phenotypes fed on the different diets.
All statistical tests were done using JMP (SALL
MATERIAL & METHODS et al. 2001).
Biological material
RESULTS
H. axyridis individuals of the aulica and nigra
phenotypes came from mass cultures reared at 22
± 1°C, 75 ± 5% RH and a photoperiod of 16L:8D, Comparative reproductive capacity of aulica and
using fluorescent lamps (Philips ref.: TDL nigra phenotypes fed A. fabae or M. persicae
23W/54 and TDL 18W/54). Coccinellids were Independent of the prey offered, fecundity,
fed a mixed diet of the aphids, A. fabae and M. fertility and percentage hatch of the eggs of
persicae, and eggs of Ephestia kuehniella Zeller. aulica females of H. axyridis were significantly
Prey were provided on bean plants (Vicia fabae higher than for nigra females (Table 1).
L. Major) placed in the rearing cages. Individuals Influence of prey quality on the reproductive
of M. persicae and A. fabae strongly differed in capacity of aulica and nigra phenotypes
size. The average weight of a wingless female of There was no significant difference in the
M. persicae was 0.48 mg and of A. fabae 1.09 fecundity, fertility and percentage hatch of the
mg. Field collected ladybirds were added to the eggs in the aulica phenotype of H. axyridis fed
mass culture at regular intervals in order to either M. persicae or A. fabae (Table 2).
minimize inbreeding. The aulica and nigra However, the reproductive capacity of nigra
phenotypes occurred naturally in our cultures. females was significantly higher than that of
The different phenotypes were reared separately. aulica females when fed on A. fabae (Table 2).
Table 1
Fecundity (average number of eggs / female / day ± s.e.), fertility (average number of embrionate eggs / female /
day ± s.e.) and hatching (percentage of hatching ± s.e.) of aulica and nigra phenotypes of H. axyridis fed on A.
fabae and M. persicae. Different letters indicate significant differences at p < 0.05
A. fabae M. persicae
aulica nigra aulica nigra
Fecundity 22.8±1.6a* 17.2±1.1b F=7.88, df=1, 298, P0.005 20.9±1.4a 12.6±0.9b F=23.57, df=1, 297, P0.0001
Fertility 20.6±1.5a 6.2±0.7b F=76.75, df=1, 298, P0.0001 19.3±1.4a 2.6±0.6b F=124.7 df=1, 298, P0.0001
Percentage of hatching 63.8±2.7a 25.1±2.8b F=100.8, df=1, 229, P0.0001 63.2±2.7a 13.7±2.8b F=209.4, df=1, 227, P0.0001
52
Table 2
Fecundity (average number of eggs / female / day ± s.e.), fertility (average number of embrionate eggs / female /
day ± s.e.) and hatching (percentage of hatching ± s.e.) of aulica and nigra phenotypes of H. axyridis fed on A.
fabae and M. persicae. Different letters indicate significant differences at p < 0.05
aulica nigra
A. fabae M. persicae A. fabae M. persicae
Fecundity 22.8±1.6a* 20.9±1.4a F=0.74, df=1, 298, P=0.39 17.2±1.1a 12.6±0.9b F=9.78, df=1, 297, P=0.001
Fertility 20.6±1.5a 19.3±1.4a F=0.42, df=1, 298, P=0.515 6.2±0.7a 2.6±0.6b F=13.2 df=1, 297, P=0.0003
Percentage of hatching 63.8±2.7a 63.2±2.7a F=0.171, df=1, 298, P=0.679 25.1±2.8a 13.7±2.8b F=19.1, df=1, 234, P0.0001
DISCUSSION REFERENCES
Generalist predators attack a variety of prey, BLACKMAN, R.L. 1967. The effects of different prey on
which differ in energy content and cost of capture Adalia bipunctata L. and Coccinella 7-punctata L.
and ingestion (ROGER et al. 2000). Food Annual Applied Biology 59: 207-219.
availability (KAWAUCHI 1981; FERRAN et al. CRAWLEY, M.J. & J.R. KREBS 1992. Foraging theory.
Pp. 90-114 in M.J. CRAWLEY (Ed.). Natural
1984; STEWART et al. 1991; DIXON & GUO 1993; Enemies. Blackwell Scientific Publications,
DIXON 2000) and prey quality (BLACKMAN 1967; Oxford. 592 pp.
NIIJIMA & TAKAHASHI 1980; NIIJIMA et al. 1986) DIXON, A.F.G. 2000. Insect Predator-Prey Dynamics:
affect the fecundity of ladybird beetles, which is Ladybirds Beetles & Biological Control.
an important determinant of their fitness Cambridge University Press, Cambridge. 257 pp.
(HEMPTINNE et al. 1993; EVANS et al. 1999). DIXON, A.F.G. & Y. GUO 1993. Eggs and cluster size in
Significant differences in the reproductive ladybird beetles (Coleoptera: Coccinellidae): the
capacity of nigra and aulica females, and a direct and indirect effects of aphid abundance.
significant increase in the reproductive capacity European Journal of Entomology 90: 457-463.
EVANS, E.W., A.T. STEVENSON & D.R. RICHARDS 1999.
of nigra females when fed on A. fabae versus M.
Essential versus foods of insect predators: benefits
persicae suggest the quality of A. fabae and M. of a mixed diet. Oecologie 121: 107-112.
persicae as prey differs for the two phenotypes. FERRAN, A., M.O. CRUZ DE BOELPAEPE, H. SCHANDERL
As predicted by optimal foraging theory, & M.M. LARROQUE 1984. Les aptitudes trophiques
predators utilize the different prey types available et reproductrices des femelles de Semiadalia
so as to maximize their rate of gain of energy undecimnotata (Col.: Coccinellidae). Entomophaga
(STEPHENS & KREBS 1986; CRAWLEY & KREBS 29: 157-170.
1992) or maximize the ingestion of nutrients FYE, R.E. 1981. Rearing and release of coccinellids for
(WALDBAUER et al. 1984). Such results support potential control of pear Psylla. Agricultural
the hypothesis of KOMAI & HOSINO (1951), that Research Service (Western Region) 20: 1-9.
FUTUYMA, D.J. 1998. Evolutionary Biology (3rd ed.).
micro-geographic variation in the relative
Sinauer Associates Inc., Massachusetts. 763 pp.
frequency of elytral patterns within H. axyridis HEMPTINNE, J.-L., A.F.G. DIXON, J.L. DOUCET & J.E.
populations could be correlated somehow with PETERSEN 1993. Optimal foraging by hoverflies
differences in the composition of aphid (Diptera: Syrphidae) and ladybirds (Coleoptera:
populations. The key may lie in intra-specific Coccinellidae): Mechanisms. European Journal of
differences in the food preferences of phenotypes Entomology 90: 141-150.
in order to increase their fitness. Requirements for HODEK, I. & A. HONċK 1996. Ecology of Coccinellidae.
high-energy resources for metabolism and Dordrecht: Kluwer Academic Publishers. 260 pp.
reproduction could determine prey choice by HOUCK, M.A. 1991. Time and resourse partitioning in
females of ladybird beetles (HOUCK 1991). As Stethorus punctum (Coleoptera: Coccinellidae).
Environmental Entomology 20: 479-497.
demonstrated in the laboratory by our
HUKUSIMA, S. & M. KAMEI 1970. Effects of various
experimental results, different phenotypes of H. species of aphids as food on development,
axyridis may thrive on different preys species in fecundity and longevity of Harmonia axyridis
nature. This may provide some selective basis for Pallas (Coleoptera: Coccinellidae). Research
maintenance of genetic polymorphism in this Bulletin. Faculty Agriculture. Gifu University 29:
species. 53-66
53
HUKUSIMA, S. & T. OHWAKI 1972. Further notes on ROGER, C., D. CODERRE & G. BOIVIN 2000. Differential
feeding biology of Harmonia axyridis Pallas prey utilization by generalist predator Coleomegilla
(Coleoptera: Coccinellidae). Research Bulletin. maculata lengi according to prey size and species.
Faculty Agriculture. Gifu University 33: 75-82. Entomologia Experimentalis et Applicata 94: 3-13.
IABLOKOFF-KHNZORIAN, S.M. 1982. Les coccinelles; SALL, J., A. LEHMAN & L. CREIGHTON 2001. JMP start
Coléoptères-Coccinellidae. Société Nouvelle des statistics: a guide to statistical and data analysis
Éditions Boubée, Paris. 569 pp. using JMP and JMP IN software. Duxbury Press,
KAWAUCHI, S. 1981. The number of oviposition, Toronto. 656 pp.
hatchability and the term of oviposition of SCHANDERL, H., A. FERRAN & M.M. LARROQUE 1985.
Propylea japonica Thunberg (Coleoptera, Les besoins trophiques et thermiques des larves de
Coccinellidae) under different food condition. la coccinelle Harmonia axyridis Pallas. Agronomie
Kontyu 49: 183-191. 5: 417-421.
KOMAI, T. 1956. Genetics of ladybeetles. Advances in SCHANDERL, H., A. FERRAN & V. GARCIA 1988. L’
Genetics 8: 155-189. élevage de deux coccinelles Harmonia axyridis et
KOMAI, T. & Y. HOSINO 1951. Contributions to the Semiadalia undecomnotata à l’ aide d’ oeufs d’
Evolutionary Genetics of the Lady-beetle, Anagasta kuehniella tués aux rayons ultraviolets.
Harmonia. II. Microgeographic Variations. Entomologia Experimentalis et Applicata 49: 417-
Genetics 36: 382-390. 421.
LUCAS, E., D. CODERRE & C. VINCENT 1997. Voracity SOARES, A.O., D. CODERRE & H. SCHANDERL 2001.
and feeding preferences of two aphidophagous Fitness of two phenotypes of Harmonia axyridis
coccinellids on Aphis citricola and Tetranychus (Coleoptera: Coccinellidae). European Journal of
urticae. Entomologia Experimentalis et Applicata Entomology 98: 287-293.
85: 151-159. SOARES, A.O., D. CODERRE & H. SCHANDERL 2003.
MCLURE, M.S. 1987. Potential of the asian predator, Effect of temperature and intraspecific allometry on
Harmonia axyridis Pallas (Coleoptera: predation activity of two phenotypes of Harmonia
Coccinellidae), to control Matsucoccus resinosae axyridis (Coleoptera: Coccinellidae).
Bean & Godwin (Homoptera: Margarodidae) in the Environmental Entomology 32: 939-944.
United States. Environmental Entomology 16: 224- STEPHENS, D.W. & J.R. KREBS 1986. Foraging theory.
230. Princeton University Press, Princeton. 237 pp.
NIIJIMA, K. & H. TAKAHASSHI 1980. Nutrional studies STEWART, L.A., A.F.G. Dixon, Z. RĤžiþka & I. Iperti
of an aphidophagous coccinellid, Harmonia 1991. Clutch and egg size in ladybird beetles.
axyridis (IV). Effects of chemically defined diet Entomophaga 36: 329-333.
and some fractions of drone honeybee on the TAN, C.C. 1946. Mosaic Dominance in the inheritance
reprodution. Bulletin Faculty Agriculture. of color patterns in the Lady-birds Beetle,
Tamagawa University 20: 47-55. Harmonia axyridis. Genetics 31:195-210.
NIIJIMA, K., M. MATSUKA & I. Okada 1986. Artificial TAN, C.C. 1949. Seasonal Variations of color patterns
diets for an aphidophagous coccinellid, Harmonia in Harmonia axyridis. Proceedings of the 8th
axyridis, and its nutrition (Minireview). Pp: 37-50 International Congress of Genetics: 669-670.
in I. HODEK (Ed). Ecology of Aphidophaga. WALDBAUER, G.P., R.W. COHEN & S. FRIEDMAN 1984.
Academia, Prague. 562 pp. Self-selection of an optimal nutrient mix from
OSAWA, N. 1992. A life table of the ladybird beetle defined diets by larvae of the corn earworm,
Harmonia axyridis Pallas (Coleoptera: Heliothis zae (Boddie). Physiological Zoology 57:
Coccinellidae) in relation to the aphid abundance. 590-597.
Japan Journal of Entomology 60: 575-579.
RICKLEFS, R.E. 1990. Ecology. (3rd ed.). W.H.
Freeman and Company. New York. 896 pp. Accepted 31 May 2003.
54
DEVELOPMENTAL ISOMORPHY IN LADYBIRDS (COLEOPTERA:

COCCINELLIDAE)
V. JAROŠÍK, J. POLECHOVÁ, A.F.G. DIXON & A. HONċK
JAROŠÍK, V., J. POLECHOVÁ, A.F.G. DIXON & A. HONċK 2003. Developmental

isomorphy in ladybirds (Coleoptera: Coccinellidae). Pp. 55-64 in A.O. SOARES,
M.A. VENTURA, V. GARCIA & J.-L. HEMPTINNE (Eds) 2003. Proceedings of the
8th International Symposium on Ecology of Aphidophaga: Biology, Ecology and
Behaviour of Aphidophagous Insects. Arquipélago. Life and Marine Sciences.
Supplement 5: x + 112 pp.
Analysis of the time spent in the egg, larval and pupal stages by insects reveal a distinct
pattern. Although, well fed individuals kept at high temperatures complete their
development much faster than poorly fed individuals kept at low temperatures,
nevertheless, they all spend the same proportion of the total time required for
development in each developmental stage. Data will be presented that indicate that
ladybirds conform to this pattern. All stages of development appear to have the same lower
developmental threshold. If this is true than it will greatly facilitate practical studies on the
development of ladybirds and improve our understanding of how selection has shaped their
life history strategies and those of insects in general.
VojtČch Jarošík (e-mail jarosik@mbox.cesnet.cz), Jitka Polechová, Department of Zoology,

Charles University, Viniþná 7, CZ-128 44 Prague 2, Czech Republic; Anthony F.G. Dixon,
University of East Anglia, School of Biological Sciences, Norwich NR4 7TJ, UK & Alois
HonČk, Research Institute of Crop Production, Drnovská 507, CZ-161 06 Prague 6 –
RuzynČ, Czech Republic.
INTRODUCTION between the rate of development and temperature

enables us to calculate two virtual constants: the
That the developmental rate of insects increases sum of effective temperatures, SET, i.e. the
with temperature is described by many models. amount of heat needed to complete a
To the best of our knowledge, the most developmental stage, and the lower
appropriate model is that of SHARPE & DE developmental threshold, LDT, i.e. the
MICHELLE (1977), which predicts a linear temperature below which development ceases
relationship between the developmental rate and (e.g., HODEK HONċK 1996, pag. 65-71). The
temperature in the middle of the temperature thermal constants SET and LDT can be used for
range. The linear part of the relationship the timing of life-history events, the
corresponds to temperatures that are ecologically determination of pre-adult thermal requirements
relevant for insect development, and gives the when forecasting and monitoring agricultural and
most reliable approximation of developmental forestry pests, assessing natural enemies suitable
rate (J. Polechová personnal communication). for biological control, comparing populations,
Here we show that the linear approximation is the and determining the effect of phylogeny,
most appropriate for ladybirds (Coleoptera: geography, body size, food and other constraints
Coccinellidae), and introduce the notion of on the duration of development. Since LUDWIG
developmental isomorphy. (1928) introduced this method it has been widely
When data on duration of development are used, and there are several reviews of the results,
converted into its reciprocal, the developmental including those for ladybirds (HONċK
rate, the linear approximation of the relationship KOCOUREK 1990; HONċK 1996; HODEK
55
HONċK 1996; KIRITANI 1997; DIXON 2000). for predicting thermal requirements, the accuracy
If the proportion of total developmental time of linear and exponential approximations of
spent in a particular developmental stage does not developmental rates were compared. Because the
change with temperature, than the LDT is the developmental rates at the very low and high
same for all developmental stages of a species, temperatures for a species, where mortality
and it shows developmental isomorphy. sharply increases, are of little practical
Developmental isomorphy is known for 7 species importance, only data for those temperatures
of mites and 342 species from 11 insect orders where the mortality was lower than 10% of the
(JAROŠÍK et al. 2002). Here developmental maximum survivorship were included in the
isomorphy is demonstrated in non-dormant analysis. Then, the residuals of the linear and
ladybirds. We discuss how developmental exponential approximations were compared using
isomorphy can greatly facilitate practical studies a paired T-test. The data used for this was that for
on the development of ladybirds. Hyperapsis notata (DREYER et al. 1997),
Harmonia axyridis (LAMANA MILLER 1998),
MATERIAL & METHODS Scymnus levaillanti and Cycloneda sanguinea
(ISIKBER 1999). These data sets are exceptional as
The linear model they include the mortality experienced by the
species at each temperature.
The linear approximation of the relationship
between the developmental rate, DR (i.e.
proportion of development occurring per unit Testing of developmental isomorphy
time) and temperature, t, can be described as DR
= a + b.t, where a is the intercept with the y-axis,
and b the slope of the linear function. From this The ratios of the times spent in each
equation, the lower developmental threshold, developmental stage at different constant
LDT, i.e. the temperature when development temperatures (ºC) were recalculated from the data
ceases (DR = 0, t = LDT) can be estimated as on duration of non-dormant development. In most
LDT = -a/b. Graphically, LDT is the value at cases, it was calculated as a ratio of time spent in
which the relationship intercepts the temperature a particular stage divided by the total pre-
axis. Using the relationship between DR and t, the imaginal development, i.e.,
sum of effective temperatures, SET, i.e. number (egg)/(egg+larva+pupa). However, data on a
of day degrees above the LDT necessary for the particular stage and an uncompleted total
completion of a particular developmental stage, development, e.g., (larva)/(larva+ pupa) were
can also be estimated. At the moment of also analysed. The calculations used data for three
completion of a development stage, DR = 1 and t or more temperatures. The data was obtained
= SET. Then, shifting y-axis so that a = 0, SET = from the studies listed in Table 1. All the 66
1/b. populations of 48 species and subspecies were
analysed to avoid bias in favour of the hypothesis
Reliability of linear vs. exponential models of being tested. All the data for each particular stage
developmental rates evaluated fell within the range of the linear
relationship between the rate of development and
To assess the suitability of a simple linear model temperature.
56
Table 1
The species and source of the data used for determining developmental isomorphy.
Species Temp. (0C)
Reference
#1 Range
Adalia bipunctata (L.) OBRYCKI TAUBER (1981) 4 18.3-26.7
Adalia bipunctata (L.) HONċK KOCOUREK (1988) 4 15-24
Adalia bipunctata (L.) GURNEY HUSSEY (1970) 3 16-24
Adalia flavomaculata DeGeer MICHELS BATEMAN (1986) 3 25-29
Brumus suturalis F. DE FLUITER (1939) 5 23.1-32.2
Calvia quattuordecimguttata (L.) LAMANA MILLER (1995) 3 14-26
Calvia quattuordecimguttata (L.) SEMYANOV (1980) 4 15-30
Cheilomenes sulphurea (Olivier) OKROUHLÁ et al. (1983) 3 20-28
Chilocorus stigma (Say) MUMA (1955) 3 16.7-26.7
Hyperaspis notata (Mulsant) CORREJO et al. (1991) 3 22-30
Coccinella novemnotata Herbst MCMULLEN (1967) 3 15.6-26.7
Coccinella quinquepunctata L. HONċK KOCOUREK (1988) 4 15-24
Coccinella septempunctata (L.) HONċK KOCOUREK (1988) 4 15-24
Coccinella septempunctata (L.) HODEK (1958) 3 15-25
Coccinella septempunctata (L.) BUTLER (1982) 4 17-25
Coccinella septempunctata (L.) OBRYCKI TAUBER (1981) 4 18.3-26.7
Coccinella septempunctata (L.) XIA et al. (1999) 3 15-25
Coccinella septempunctata (L.) TRITISCH (1997) 3 17-25
Coccinella septempunctata brucki Mulsant KAWAUCHI (1983) 4 15-25
Coccinella septempunctata brucki Mulsant KAWAUCHI (1979) 3 20-30
Coccinella septempunctata brucki Mulsant SAKURAI et al. (1991) 3 22-30
Coccinella transversalis F. VEERAVEL BASKARAN (1996) 3 18-30
Coccinella transversoguttata Brown OBRYCKI TAUBER (1981) 4 18.3-26.7
Coccinella trifasciata L. MILLER LAMANA (1995) 5 18-34
Coccinella undecimpunctata L. ERAKY NASSER (1993) 4 14-26
Coelophora quadrivittata Fauvel CHAZEAU (1981) 3 20-30
Coleomegilla maculata (DeGeer) OBRYCKI TAUBER (1978) 4 18.3-26.7
Coleomegilla maculata (DeGeer) GURNEY HUSSEY (1970) 3 16-24
Coleomegilla maculata lengi Timberlake WRIGHT LAING (1978) 4 19-25
Curinus coeruleus Mulsant DIRAVIAM VIRAKTAMATH (1991) 4 22.8-25.8
Cycloneda sanguinea (L.) ISIKBER (1999) 4 20-27.5
Cycloneda sanguinea (L.) GURNEY HUSSEY (1970) 3 16-24
Delphastus catalinae (Horn) HEMACHANDRA (1994) 3 20-26
Eriopis connexa (Germar) MILLER PAUSTIAN (1992) 4 14-26
Harmonia axyridis (Pallas) KAWAUCHI (1979) 3 20-30
Harmonia axyridis (Pallas) LAMANA MILLER (1998) 5 14-30
Hippodamia convergens Guerin BUTLER DICKERSON (1972) 4 20-28.9
Hippodamia convergens Guerin OBRYCKI TAUBER (1982) 5 15.6-26.7
Hippodamia parenthesis (Say) ORR OBRYCKI (1990) 4 14-26
Hippodamia quinquesignata (Kirby) KADDOU (1960) 3 15.6-30
Hippodamia sinuata Mulsant MICHELS BEHLE (1991) 4 15-30
Hippodamia variegata Goetz MICHELS BATEMAN (1986) 3 25-29
Hyperaspis notata Mulsant DREYER ET AL. (1997) 5 18-32
Lemnia biplagiata (Swartz) SEMYANOV BEREZNAYA (1988) 3 20-30
Lioadalia flavomaculata (DeGeer) BROWN (1972) 6 13-27
Menochilus sexmaculatus (F.) KAWAUCHI (1979) 3 20-30
Menochilus sexmaculatus (F.) VEERAVEL BASKARAN (1996) 3 18-30
1
Number of temperatures
57
Table 1 (continued)
The species and source of the data used for determining developmental isomorphy.
Temp. (0C)
Species Reference
#1 Range
Olla v-nigrum (Mulsant) KREITER (1985) 5 15-30
Pharoscymnus flexibilis (Mulsant) SHARMA et al. (1990) 3 24-32
Pharoscymnus numidicus (Mulsant) KEHAT (1967) 3 24-31
Propylea japonica (Thunberg) KAWAUCHI (1979) 3 20-30
Propylea japonica (Thunberg) KAWAUCHI (1983) 4 15-25
Propylea quatuordecimpunctata (L.) HONċK KOCOUREK (1988) 3 15-24
Propylea quatuordecimpunctata (L.) BAUMGAERTNER et al. (1987) 3 15.3-25.3
Scymnus apiciflavus Motschulsky DE FLUITER (1939) 6 19.1-32.2
Scymnus frontalis (F.) NARANJO et al. (1990) 3 15-26.2
Scymnus hoffmani Weise ZHAO WANG (1987) 4 18-30
Scymnus hoffmani Weise KAWAUCHI (1983) 4 15-25
Scymnus interruptus (Goeze) TAWFIK et al. (1973) 3 15.5-27.9
Scymnus levaillanti Mulsant ISIKBER (1999) 5 17.5-27.5
Scymnus roepkei de Fluiter DE FLUITER (1939) 6 19.1-32.2
Semiadalia undecimnotata (Schneider) HONċK KOCOUREK (1988) 4 15-24
Stethorus bifidus Kapur PETERSON (1993) 5 12.5-27.5
Stethorus punctillum Weise BERKER (1958) 3 19-35.6
Stethorus japonicus H.Kamiya TANAKA (1966) 13 17-29
Subcoccinella vigintiquatuorpunctata (L.) ALI (1971) 3 18-28
1
Number of temperatures
Angular transformed proportion of total size), which is independent of sample size, and
developmental time spent in a particular stage the null hypothesis that the overall effect size
was plotted against temperature, and the existence indicates a zero slope was tested. The assumption
of developmental isomorphy inferred from a zero that the individual analyses share a common
change in proportion. Temperature was first population effect size was tested by the
regressed with a different intercept and a different homogeneity statistic Q (SHADISH HADDOCK
slope for each stage (using average proportion for 1994). Details of the statistical procedure are
replicated data due to origin or photoperiod), and described in JAROŠÍK et al. (2002).
the significance was then evaluated by
simultaneous deletion test. Individual studies on
RESULTS
populations of the same species were analysed
separately because the results varied due to
differences in experimental design. The Linear vs. exponential models of developmental
calculations were performed using general linear rate
modelling in GLIM v. 4 (FRANCIS et al. 1994).
To reach a general conclusion, all the data was For the four ladybird species for which there is
then tested using meta-analysis, a statistical sufficient data on developmental rate and
synthesis of the results of separate, independent mortality, the linear model gave a better fit (R2 =
experiments (HEDGES OLKIN 1985; 0.991 r 0.0102) than the exponential (R2 = 0.973
GUREWITCH HEDGES 1993; COOPER r 0.0194) (t = 2.62; df = 3; P < 0.05) (Table 2).
HEDGES 1994). The outcome of each analysis was That is, the simple linear model is the best for
represented by a quantitative index (the effect practical purposes.
58
Table 2 in a particular developmental stage with change

Comparison of explained variance (R2) obtained by in temperature. It is illustrated for Propylea
fitting exponential (exp) and linear (lin) japonica in Fig. 1. Therefore, within the
approximations to the relationships between temperature ranges with a linear relationship
developmental rate and temperature in 4 species of between RD and t, for each species there is (1) a
ladybirds.
proportional SET for completion of each
Species Stage R2 lin R2 exp
developmental stage at each temperature and (2) a
Cycloneda sanguinea Pupa 0.990 0.960
common LDT for all developmental stages.
Harmonia axyridis Pupa 0.999 0.976
Hyperaspis notata Egg - Adult 0.998 0.999
Scymnus levaillanti Pupa 0.977 0.957 DISCUSSION
Average 0.991 0.973
Standard error 0.0102 0.0194 Linear vs. non-linear models of developmental
rate
There are three categories of models of the

Developmental rate (day-1)
relationship between temperature and duration of

Egg development in insects and other ectotherms
28% (HONċK 1999): (i) Non-linear. The objective of
Larva this kind of data fitting is the description of the
50% developmental rate over a wide range of
Pupa temperatures (e.g. STINNER et al. 1974;
22%
HAGSTRUM MILLIKEN 1991). This type of
LD T model gives a good fit to the data, but the
9.2 15 18 20 25 Tem perature ( 0 C ) parameters have little biological meaning. (ii)
Non-linear incorporating physiological and
Fig. 1. Effect of temperature (t) on rate of
development (RD) within the linear range of the biochemical constants. These not only describe
relationship for Propylea japonica. Data from but also attempt to explain the relationship in
Kawauchi (1983). The population is isomorphic and terms of physiological mechanisms (e.g. LOGAN
spent 28% of total development in egg, 50% in larva, et al. 1976; SHARPE DEMICHELLE 1977;
and 22% in pupa at temperatures 15, 18, 20 and 25 0C. SCHOOLFIELD et al. 1981; WAGNER et al. 1984,
All developmental stages have a common lower 1991). (iii) A linear approximation.
developmental threshold (LDT = 9.2) and for Models of type (ii) are often theoretically
assessment of the number of day-degrees above the correct, but not tractable for most ecologists. In
LDT necessary for completion of a particular
development stage the sum of effective temperatures
addition, only relatively recent models of this
can be determined at any temperature within the linear type (e.g. LACTIN et al. 1995) enable one to
range. calculate lower developmental threshold. On the
other hand, the type (iii) linear approximation
Developmental isomorphy within the range of temperatures ecologically
relevant to where an insect lives, is a reliable
Combining statistically the results of the 68 model, and can be easily used to calculate two
studies in a meta-analysis indicated the overall virtual constants: the lower developmental
prevalence of developmental isomorphy in threshold, LDT (the temperature below which
ladybirds (the overall weighted average effect development ceases), and the sum of effective
size: 7.10E-04 with 95% confidence interval from temperatures, SET (the amount of heat needed for
–2.94E-03 to 4.36E-03; homogeneity statistic Q = completing a developmental stage). This greatly
0.64, df = 65, p = 1). That is, all the populations simplifies modelling. Therefore, we argue against
and species showed developmental isomorphy. non-linear models for practical purposes, and
Developmental isomorphy in ladybirds advocate the simple linear model. We believe that
implies no change in the proportion of time spent the simple linear model of the relationship
59
between developmental rate and temperature Even if LDTs are calculated from data
gives the best approximation of LDT and SET in collected over a range of ecologically relevant
insects. temperatures, and the regression of development
rate on temperature is linear, the accuracy of the
The existence of a common LDT for all the estimates is affected by errors in the estimates of
developmental stages of a species the developmental rate (CAMPBELL et al. 1974).
The low precision of LDTs is obvious from their
When the data for ladybirds were plotted against standard errors (CAMPBELL et al. 1974), which
temperature, the developmental isomorphy are typically between 1-3 0C (J. JANÁýEK A.
hypothesis was supported by a zero change in the HONċK, unpublished data). Crucial from a
proportion of the total developmental time spent statistical point of view is any bias in the
in a particular stage of a development. measurements made at extreme temperatures.
Developmental isomorphy in the overall pattern Important determinants of the slopes of the linear
of the data thus indicates that all the regressions, from which the LDTs are inferred,
developmental stages of each ladybird species are the extreme values (see CRAWLEY 1993, p.
have a common LDT. If so, there should be little 78-82). Therefore, a relatively small bias in the
variation in the LDT between stages and instars developmental rates measured at extreme
within a species, and within populations of temperatures will cause a large shift in the LDT.
individual species. This is not supported by the Poor estimates of developmental rate are most
literature on ladybird development (HONċK likely at high temperatures because the precision
KOCOUREK 1990; HONċK 1996; HODEK with which the duration of development is
HONċK 1996). Therefore, if developmental measured is poor and the error large
isomorphy is a common feature of ladybirds, then (development rate is the reciprocal of duration of
a significant proportion of the variation in LDTs development).
within species is illusory and possibly a
consequence of how it is estimated from CONCLUSIONS
experimental data.
What are the sources of error in estimating The existence of rate isomorphy in ladybirds has
LDT? First, the values of developmental rate important practical implications for the timing of
obtained at extreme (high or low) temperatures life-history events. The experimental procedure
may violate developmental isomorphy (JAROŠÍK for determining the thermal development
et al. 2002). At low temperatures there may be constants, LDT and SET, can be simplified. The
differential mortality. The individuals with the lower developmental threshold can be determined
fastest development complete their development based on data for one stage, preferably the pupa,
but the rest are more likely to succumb to adverse which is little affected by factors other than
conditions, because their development is temperature, and has a duration usually longer
prolonged. Second, imprecise measurement of than that of the egg stage. SET may also be
developmental time, particularly at high calculated from the duration of development at
temperatures. As developmental rate increases one temperature (Fig. 1). Thus more effort can be
with temperature, the number of observations per invested in greater precision in determining the
stage should also increase. To measure the rate of length of development.
development with the same precision at low and
high temperatures, the time interval must be ACKNOWLEDGEMENTS
proportional to the length of the development
stage at each temperature. This is not the case in The work was supported the Ministry of
most studies (SHAFFER 1983; VAN RIJN et al. Education Youth and Sport of the Czech Republic
1995). A constant monitoring is the most (grant no. J13/98113100004), and the Grant
probable source of bias in data collected at high Agency of the Czech Republic (grant no.
temperatures. 522/01/0864).
60
REFERENCES Mulsant (Coleoptera: Coccinellidae). Journal of

Biological Control 5: 14-17.
ALI, M. 1971. The effect of temperature and DIXON, A.F.G. 2000. Insect predator-prey dynamics:
photoperiod on the behaviour and development of ladybird beetles and biological control. Cambridge
Subcoccinella vigintiquatuorpunctata L. (Col., University Press, Cambridge.
Coccinellidae). Acta Phytopathologica Academiae DIXON, A.F.G., J.-L. HEMPTINE P. KINDLMANN 1997.
Scientiarum Hungaricae 6: 185-189. Effectiveness of ladybirds as biological control
BAUMGAERTNER, J., M. BIERI V. DELUCCHI 1987. agents: patterns and processes. Entomophaga 42:
Growth and development of immature life stages of 71-83.
Propylaea 14-punctata L. and Coccinella 7- DREYER, B.S., P. NEUENSCHWANDER, B. BOUYJOU, J.
punctata L. (Col.: Coccinellidae) simulated by the BAUMGÄRTNER S. DORN 1997. The influence of
metabolic pool model. Entomophaga 32: 415-423. temperature on the life table of Hyperaspis notata.
BERKER, J. 1958. Die natürlichen Feinde der Entomologia Experimentalis et Applicata 84: 85-
Tetranychiden. Zeitschrift für angewandte 92.
Entomologie 43: 115-172. ERAKY, S.A. M.A.K. NASSER 1993. Effect of
BROWN, H.D. 1972. On the biology of Lioadalia constant temperatures on the development and
flavomaculata (Deg.) (Col., Coccinellidae), a predation prey efficiency of the ladybird beetle,
predator of the wheat aphid (Schizaphis graminum Coccinella undecimpunctata L. (Coleoptera:
(Rond.)) in South Africa. Bulletin of Entomological Coccinellidae). Assiut Journal of Agricultural
Research 61: 673-679. Sciences 24: 223-230.
BUTLER, G.D. 1982. Development time of Coccinella FRANCIS, B., M. GREEN C. PAYNE (Eds). 1994. The
septempunctata in relation to constant temperatures GLIM system. Release 4 manual. Clarendon Press,
(Col.: Coccinellidae). Entomophaga 27: 349-353. Oxford.
BUTLER, G.D. U.A. DICKERSON 1972. Life cycle of FRAZER, B.D. R.R. MCGREGOR 1992. Temperature-
the convergent lady beetle in relation to dependent survival and hatching rate of eggs of
temperature. Journal of Economic Entomology 65: seven species of Coccinellidae. Canadian
1508-1509. Entomologist 124: 305-312.
CAMPBELL, A., B.D. FRAZER, N. GILBERT, A.P. GILBERT, N. D.A. RAWORTH 1996. Insects and
GUTIERREZ M. MACKAUER 1974. Temperature temperature - a general theory. Canadian
requirements of some aphids and their parasites. Entomologist 128: 1-13.
Journal of Applied Ecology 11: 431-438. GUREVITCH, J. L.V. HEDGES 1993. Meta-analysis:
CORREJO, N.S., A.C. BELLOTTI & R. GONZÁLEZ 1991. Combining the results of independent experiments.
Evaluacion de algunos factores determinantes de la Pp. 378-398 in S.M. SCHEINER J. GUREVITCH
efficiencia de Cleothora notata (Col.: (Eds). Design and analysis of ecological
Coccinellidae) como depredator del piojo harinoso experiments. Chapman ҏ Hall, New York..
de la Yuca Phenococcus herreni (Hom.: GURNEY, B. N.W. HUSSEY 1970. Evaluation of some
Pseudococcidae). Revista Columbiana de coccinellid species for the biological control of
Entomologia 17: 21-27. aphids in protected cropping. Annals of Applied
CHAZEAU, J. 1981. Données sur la biologie de Biology 65: 451-458.
Coelophora quadrivittata (Col.: Coccinellidae), HAGSTRUM, D.W. G.A. MILLIKEN 1991. Modelling
prédateur de Coccus viridis (Hem: Coccidae) en differences in insect developmental times between
Nouvelle-Calédonie. Entomophaga 26: 301-312. constant and fluctuating temperatures. Annals of
COOPER, H. L.V. HEDGES (Eds). 1994. The handbook the Entomological Society of America 84: 369-379.
of research synthesis. Russell Sage Foundation, HAMALAINEN, M. M. MARKKULA 1977. Cool storage
New York. of Coccinella septempunctata and Adalia
CRAWLEY, M.J. 1993. GLIM for ecologists. Blackwell, bipunctata (Col., Coccinellidae) eggs for use in the
London. biological control in greenhouses. Annales
DE FLUITER, H.J. 1939. Beitrag zur Kenntnis der Agricultural Fennicae 16: 132-136.
Biologie und Ökologie einiger Coccinelliden von HEDGES, L.V. I. OLKIN 1985. Statistical methods for
Java. 7th International Congress of Entomology, meta-analysis. Academic Press, New York.
Berlin, 1939: 1106-1119. HEMACHANDRA, K.S. 1994. Development, reproduction
DIRAVIAM, J. C.A. VIRAKTAMATH 1991. Biology of and feeding behaviour of Delphastus catalinae, a
the introduced ladybird beetle, Curinus coeruleus coccinellid predator of glasshouse whitefly,
61
Trialeurodes vaporariorum. M.Sc. Thesis, the adult form of Cocinella septempunctata brucki,
University of London. Propylea japonica and Scymnus (Pullus) hoffmani
HODEK, I. 1958. Influence of temperature, rel. humidity (Coleoptera, Coccinelldiae). Kurume University
and photoperiodicity on the speed of development Journal 32: 45-51.
of Cocinnella septempunctata L. ýasopis KEHAT, M. 1967. Studies on the biology and ecology of
ýeskoslovenské Spoleþnosti Entomologické 55: Pharoscymnus numidicus [Coccinellidae] an
121-141. important predator of the date palm scale
HODEK, I. A. HONċK 1996. Ecology of Parlatoria blanchardi. Annales de la Societé
Coccinellidae. Kluwer, Dordrecht. Entomologique de France (N.S.) 3: 1053-1065.
HONċK, A. 1996. The relationship between thermal KIRITANI, K. 1997. The low development threshold
constants for insect development: a verification. temperature and the thermal constant in insects,
Acta Societatis Zoologicae Bohemoslovaca 60: mites and nematodes in Japan. The Miscellaneous
115-152. Publications of the National Institute of Agro-
HONċK, A. 1997. Incidence of protogynous and Environmental Sciences 21: 1-72.
protandrous development in the pre-imaginal stage KREITER, S. 1985. Etude bioecologique dÓlla v-
of insect development: an overview. Acta Societatis nigrum et essai de quantification de léfficacite
Zoologicae Bohemoslovaca 61: 113-128. predatrice dÁdalia bipunctata (L.) contre les
HONċK, A. 1999. Constraints on thermal requirements aphides en verger de pechers (Coleoptera:
for insect development. Entomological Science 2: Coccinellidae). Ph.D. Thesis, Universite de Droit,
615-621. dÁix, Marseille.
HONċK, A., V. JAROŠÍK, Z. MARTINKOVÁ I. Novák LACTIN, D.J., N.J. HOLLIDAY, D.L. JOHNSON R.
2002. Food induced variation of thermal constants CRAIGEN 1995. Improved rate model of
of development and growth of Autographa gamma temperature-dependent development by arthropods.
(Lepidoptera: Noctuidae) larvae. European Journal Environmental Entomology 24: 68-75.
of Entomology 99: 241-252. LAMANA, M.L. S.C. MILLER 1995. Temperature-
HONċK, A. F. KOCOUREK 1988. Thermal dependent development in a polymorphic lady
requirements for development of aphidophagous beetle, Calvia quatuordecimguttata (Coleoptera:
Coccinellidae (Coleoptera), Chrysopidae, Coccinellidae). Annals of Entomological Society of
Hemerobiidae (Neuroptera), and Syrphidae America 88: 785-790.
(Diptera): some general trends. Oecologia 76: 455- LAMANA, M.L. S.C. MILLER 1998. Temperature-
460. dependent development in an Oregon population of
HONċK, A. F. KOCOUREK 1990. Temperature and Harmonia axyridis (Coleoptera: Coccinellidae).
development time in insects: a general relationship Environmental Entomology 27: 1001-1005.
between thermal constants. Zoologische LAMB, R.J. 1998. Insects and temperature - some
Jahrbücher Abteilung für Systematik und Ökologie comments on a general theory. Canadian
der Tiere 117: 401-439. Entomologist 130: 111-114.
ISIKBER, A.A. 1999. Evaluation of two predatory LOGAN, J.A., D.J. WOLLKIND, S.C. HOYT L.K.
coccinellid beetles, Scymnus levaillanthi and TANIGOSHI 1976. An analytic model for description
Cycloneda sanguiea (Coleoptera: Coccinellidae) of temperature dependent rate phenomena in
for the biocontrol of cotton aphid, Aphis gossypii arthropods. Environmental Entomology 5: 1133-
(Homoptera: Aphididae). Ph.D. Thesis, University 1140.
of London. LUDWIG, D. 1928. The effects of temperature on the
JAROŠÍK V., A. HONċK A.F.G. DIXON 2002. development of an insect (Poppilia japonica
Developmental rate isomorphy in insects and mites. Newman). Physiological Zoology 1: 358-389.
American Naturalist 160: 497-510. MCMULLEN, R.D. 1967. The effects of photoperiod,
KADDOU, I.K. 1960. The fleeding behavior of temperature, and food supply on rate of
Hippodamia quinquesignata (Kirby) larvae. development and diapause in Coccinella
University of California Publications in novemottata. Canadian Entomologist 99: 578-586.
Entomology 16: 181-232. MICHELS, G.J. A.C. BATEMAN 1986. Larval biology
KAWAUCHI, S. 1979. Effect of temperatures on the of two imported predators of the greenbug,
aphidophagous coccinellids. Kurume University Hippodamia variegata Goetz and Adalia
Journal 28: 47-51. flavomaculata DeGeer, under constant
KAWAUCHI, S. 1983. The threshold temperature and temperatures. Southwestern Entomologist 11: 23-
thermal constant for development from the egg to 30.
62
MICHELS, G.J. R.E. BEHLE 1991. Effects of two prey its development. Polish Journal of Ecology 35:
species on the development of Hippodamia sinuata 755-765.
(Coleoptera: Coccinellidae) larvae at constant ORR, C.J. J.J. OBRYCKI 1990. Thermal and dietary
temperatures. The Journal of Economic requirements for development of Hippodamia
Entomology 84: 1480-1484. parenthesis (Coleoptera: Coccinellidae).
MILLER, J.C. 1992. Temperature-dependent Environmental Entomology 19: 1523-1527.
development of the convergent lady beetle PETERSON, P.G. 1993. The potential ability of Stethorus
(Coleoptera: Coccinellidae). Environmental bifidus (Kapur) to regulate populations of
Entomology 21: 197-201. Tetranychus lintearius (Dufour). M.Sc. Thesis,
MILLER, J.C. M.L. LAMANA 1995. Assessment of Massey University, New Zealand.
temperature-dependent development in the general ROSENTHAL, R. 1994. Parametric measures of effect
population and among isofemale lines of size. Pp. 231-244 in H. COOPER L.V. HEDGES
Coccinella trifasciata (Col.: Coccinellidae). (Eds). The handbook of research synthesis. Russell
Entomophaga 40: 183-192. Sage Foundation, New York.
MILLER, J.C. J.W. PAUSTIAN 1992. Temperature- SAKURAI, H., N. YOSHIDA, C. KOBAYASHI S. TAKEDA
dependent development of Eriopis connexa 1991. Effects of temperature and day length on
(Coleoptera: Coccinellidae). Environmental oviposition and growth of lady beetle, Coccinella
Entomology 21: 1138-1142. septempunctata bruckii. Research Bulletin of
MUMA, M.H. 1955. Some ecological studies on the Faculty of Agriculture Gifu University 56: 45-50.
twice-stabbed lady beetle Chilocorus stigma (Say). SEMYANOV, V.P. 1980. Biology of Calvia
Annals of the Entomological Society of America 48: quatuordecimguttata L. (Coleoptera:
493-498. Coccinellidae). Revue dÉntomologie de lÚRSS
NARANJO, S.E., T.L. GIBSON D.D. WALGENBACH 59: 757-763.
1990. Development, survival, and reproduction of SEMYANOV, V.P. E.B. BEREZNAJA 1988. Biology and
Scymnus frontalis (Coleoptera: Coccinellidae), on prospects of using Vietnam´s lady beetle Lemnia
imported predator of Russian wheat aphid, at four biplagiata (Swartz) for control of aphids in
fluctuating temperatures. Annals of the greenhouses. Pp. 267-269 in E. Niemczyk
Entomological Society of America 83: 527-531. A.F.G. Dixon (Eds). Ecology and effectiveness of
NYLIN S., C. WIKLUND, P.O. WICKMAN E. GARCIA- aphidophaga. Academic Publishers, The Hague.
BARROS 1993. Absence of trade-offs between SHADISH, V.R. C.K. HADDOCK 1994. Combining
sexual size dimorphism and early male emergence estimates of effect size. Pp. 261-281 in H. COOPER
in a butterfly. Ecology 74: 1414-1427. L.V. HEDGES (Eds). The handbook of research
OBRYCKI, J.J. M.J. TAUBER 1978. Thermal synthesis. Russell Sage Foundation, New York.
requirements for development of Coleomegilla SHARMA, D.C., U.S. RAWAT & A.D. PAWAR 1990.
maculata (Coleoptera: Coccinellidae) and its Effect of temperature and humidity on the
parasite Perilitus coccinellae (Hymenoptera: development, longevity and predatory potential of
Braconidae). Canadian Entomologist 110: 407-412. Pharoscymnus flexibilis Muls. on San Jossé scale.
Journal of Biological Control 4: 11-14.
OBRYCKI, J.J. M.J. TAUBER 1981. Phenology of three
SHAFFER, P.L. 1983. Prediction of variation in
coccinellid species: thermal requirements for
development period of insects and mites reared at
development. Annals of the Entomological Society
constant temperatures. Environmental Entomology
of America 74: 31-36.
12: 1012-1019.
OBRYCKI, J.J. M.J. TAUBER 1982. Thermal SHARPE, P.J.H. & W. DE MICHELE 1977. Reaction
requirements for development of Hippodamia kinetics of poikilotherm development. Journal of
convergens (Coleoptera: Coccinelldiae). Annals of Theoretical Biology 64: 649-670.
the Entomological Society of America 75: 678-683. SHARPE, P.J.H., G.L. CURRY, D.W. DEMICHELE C.L.
OKROUHLÁ, M., S. CHAKRABARTI I. HODEK 1983. COLE 1977. Distribution model of organism
Developmental rate and feeding capacity in development times. Journal of Theoretical Biology
Cheilomenes sulphurea (Coleoptera: 66: 21-38.
Coccinellidae). VČstník ýeskoslovenské Spoleþnosti SCHOOLFIELD, R.M., P.J.H. SHARPE & C.E. MAGNUSON
Zoologické 47: 105-117. 1981. Non-linear regression of biological
OLSZAK, R.W. 1987. The occurrence of Adalia temperature-dependent rate models based on
bipunctata (L.) (Coleoptera, Coccinellidae) in absolute reaction-rate theory. Journal of
apple orchards and the effect of different factors on Theoretical Biology 88: 719-731.
63
SLANSKY, F. 1993. Nutritional ecology: the Annals of Applied Biology 125: 521-532.
fundamental quest for nutrients. Pp. 29-91 in N.E. VAN RIJN, C.J., C. MOLLEMA G. STEENHUIS-BROERS
STAMP T.M. CASEY (Eds). Caterpillars. 1995. Comparative life-history studies of
Chapman and Hall, New York. Frankliniella occidentalis and Thrips tabaci
SLANSKY, F. J.M. SCRIBER 1985. Food consumption (Thysanoptera: Thripidae) on cucumber. Bulletin of
and utilization. Pp. 87-163 in G.A. Kerkuth L.I. Entomological Research 85: 285-297.
Gilbert (Eds). Comprehensive Insect Physiology, VEERAVEL, R., P. BASKARAN 1996. Temperature-
Biochemistry and Pharmacology, Vol. 4. dependent development, adult longevity, fecundity
Pergamon, Oxford. and feeding potential of two coccinellid predators
under laboratory conditions. Entomon 21: 13-18.
STINNER, R.E., A.P. GUTIERREZ G.D. BUTLER 1974.
An algorithm for temperature-dependent growth WAGNER, T.L., R.L. OLSON J.L. WILLERS 1991.
rate simulation. Canadian Entomologist 106: 519- Modeling arthropod development time. Journal of
524. Agricultural Entomology 8: 251-270.
WAGNER, T.L., H.I. WU, P.J.H. SHARPE, R.M.
TANAKA, M. 1966. Fundamental studies on the
SCHOOLFIELD R.N. COULSON 1984. Modeling
utilization of natural enemies in the citrus grove in
insect development rates: a literature review and
Japan I. The bionomics of natural enemies of the
application of a biophysical model. Annals of the
most serious pests II. Stethorus japonicus H.
Entomological Society of America 77: 208-225.
Kamiya (Coccinellidae), a predator of the citrus red
mites, Panonychus citri McG. Bulletin WRIGHT, E.J. J.E. LAING 1978. The effects of
Horticultural Research Japan Station Series D 4: temperature on development, adult longevity and
22-42. fecundity of Coleomegilla maculata lengi and its
parasite, Perilitus coccinellae. Proceedings of the
TAWFIK, M.F.S., S. ABDUL-NASR B.M. SAAD 1973. Entomological Society of Ontario 109: 33-47.
The biology of Scymnus interruptus Goeze
XIA, J.Y., W. VAN DER WERF R. RABBINGE 1999.
(Coleoptera: Coccinellidae). Bulletin de la Societe
Temperature and prey density on bionomics of
Entomologique dÉgypte 57: 9-26.
Coccinella septempunctata (Coleoptera:
TRITISCH, H. 1997. Der Marienkäfer Coccinella Coccinellidae) feeding on Aphis gossypii
septempunctata L. im Komplex Winterweizen- (Homoptera: Aphididae) on cotton. Environmental
Getriedeblattläusen Antagonist. Agrarökologie, Entomology 28: 307-314.
Vol. 24. Verlag Agrarökologie Berne, Hannover. ZHAO, D.X. Z.W. WANG 1987. Influence of
TRUDGILL, D.L. 1995. Why do tropical poikilothermic temperature on the development of the coccinellid
organisms tend to have higher threshold beetle, Scymnus hoffmani Weise. Acta
temperature for development than temperate ones. Entomologica Sinica 30: 47-53.
Functional Ecology 9: 136-137.
TRUDGILL D.L. J.N. PERRY 1994. Thermal time and
ecological strategies - a unifying hypothesis. Accepted 31 May 2003.
64
PAST AND FUTURE MODELS OF PREDATOR-PREY POPULATION

DYNAMICS: THE IMPORTANCE OF GENERATION TIME RATIO
P. KINDLMANN & A.F.G. DIXON

KINDLMANN, P. & A.F.G. DIXON 2003. Past and future models of predator-prey
population dynamics: the importance of generation time ratio. Pp. 65-71 in A.O.
SOARES, M.A. VENTURA, V. GARCIA & J.-L. HEMPTINNE (Eds) 2003. Proceedings
of the 8th International Symposium on Ecology of Aphidophaga: Biology,
Ecology and Behaviour of Aphidophagous Insects. Arquipélago. Life and Marine
Sciences. Supplement 5: x + 112 pp.
Mathematical models of predator-prey population dynamics are widely used for predicting
the effect of predators as biological control agents, but the assumptions of the models are
more relevant to parasite-host systems. Predator-prey systems, at least in insects,
substantially differ from what is assumed by these models. The main differences are: (i)
Juveniles and adults have to be considered as two different entities, as the former stay
within a patch and do not reproduce, while the latter move between patches of prey where
they reproduce. (ii) Because of their high mobility, food availability is likely to be less
restrictive for adults than juveniles, which are confined to one patch. Therefore, a
functional response to prey abundance may not be important for adults. (iii) Egg and larval
cannibalism are common in insect predators. Therefore, the quality of patches of prey for
their larvae determines the reproductive strategy of adult predators more than the
availability of food for the adults. Here we develop a new model, based on the above
considerations, which is suitable for modelling these interactions. We show that selection
should favour mechanisms that enable predators to avoid reproducing in patches with
insufficient prey and those already occupied by predators.
Pavel Kindlman (e-mail: pavel@entu.cas.cz), Faculty of Biol. Sci., Univ. South Bohemia
and Inst. Landscape Ecology CAS, Branisovska 31, CZ-37 005 Ceske Budejovice, Czech
Republic & A.F.G. Dixon, School of Biological Sciences, University of East Anglia,
Norwich, NR4 7TJ, UK.
INTRODUCTION – THE PAST are treated similarly and a stable age distribution
assumed. Similarly, all predators and prey are
assumed to interact with each other and migration
Since the sixties, mathematical models have been between patches ignored in the hope that the
widely used to describe the population dynamics resulting population dynamics will be similar
of predator-prey systems. The main emphasis was
everywhere.
put on calculating stability of their equilibrium
points (BEDDINGTON et al. 1976, 1978; HASSELL
CLASSICAL MODELS VS. EVOLUTION OF
1978; GODFRAY & HASSELL 1987; MURDOCH
LIFE HISTORIES
1994). These models have a lot in common.
Stability and oscillatory behaviour of these In insect predator-prey systems, however, the
models follow from the negative feedback above assumptions are not satisfied. While
between prey density and predator fecundity, juvenile predators (larvae) are confined to the
which in turn is a function of predator voracity. patch of prey1 where they were born, as they can
Thus this feedback, which is dependent mainly on
prey density (functional response), is the crucial 1
Patch in this sense means the space that a larva of a predator
factor determining the population dynamics of
can reach by walking, usually one or only a few adjacent
predator-prey systems (HASSELL 1978). The plants, or even only a part of an individual plant as in the case
individual developmental stages of the predator of plants like trees.
65
only walk, adult predators can fly, which enables illustrated by a simple mental exercise: Assume a
them to move between patches and therefore to univoltine population (genotype) P of predators,
find patches where food is abundant. Thus food each female of which produces 200 eggs and that
availability is much less limiting for adult insect mortality during development from egg to adult is
predators than for their larvae. In addition, the 99%. Fitness of a female in this population is the
fitness of a female predator is determined by the number of her offspring that survive to
number of her offspring that survive to reproductive age and equals 200u0.01 = 2.
reproductive age. Probability that an egg will Assume a mutant, genotype A, selected for better
survive and develop into a reproductive survival, the fecundity of which is the same as
individual strongly depends on the quality that of P, i.e., 200 eggs, but has a marginally
(measured as food availability and risk of being lower mortality: 98%. Its fitness is 200u0.02 = 4.
killed during the period of larval development) of Genotype P would be out-competed by genotype
the patch where the egg is laid. Therefore, there A. For P to have the same fitness as A, its
are two life history traits that determine fitness of fecundity would have to be 400. Thus, in terms of
a female predator: number of eggs laid and fitness, a decline in mortality from 99% to 98% is
quality of the patches where these eggs are laid. equivalent to increasing fecundity from 200 to
Which of these is likely to be the stronger 400 eggs. Assuming the same scenario but a
selection force? lower mortality, e.g., 50% for P and 49% for A,
Before dealing with this question, it is would yield a different result: fitness of P is then
necessary to accept that the fecundity of insect 200u0.5 = 100, and of A 200u0.51 = 102. For P
predators is enormous and can be a thousand eggs in this case to have the same fitness its fecundity
per female under laboratory conditions (DIXON would have to be 204, i.e., a very small increase.
2000). Their realised fecundity in nature is most This simple exercise illustrates a general
likely lower because of the harsher conditions phenomenon: in organisms suffering high
prevailing there, but assuming that several mortality, very small improvements in survival
hundreds of eggs are laid on average per female can be very important in terms of fitness. In order
does not seem to be unrealistic. Although the to achieve the same result fecundity would have
numbers of insect predators may fluctuate from to be increased considerably. Such a change is
year to year, their numbers over the long term are likely to be more severely constrained than a
stable, i.e., they neither approach infinity nor small change in oviposition behaviour resulting in
zero. Thus an average individual produces one improved offspring survival. Assumption of
offspring that survives to reproductive age. univoltinism is convenient, as it simplifies the
Assuming the sex ratio is close to 1:1, then a calculation of fitness, but is not crucial to the
female predator produces on average two outcome.
offspring that survive to reproductive age over the The finding that selection for survival could
long-term (several decades). Combining the be much more important for insect predators than
above, i.e., assuming that an average mother selection for fecundity also changes our view of
produces several hundreds of eggs, but only two the nature of the models needed for insect
survive to reproductive age then the mortality predator-prey systems. Selection for voracity is
during development from egg to adult must be still important, as a large voracity implies large
enormous: reaching values close to 99%. This adult size. However, in adults, voracity is
theoretical prediction is strongly supported by positively correlated with fecundity (HEMPTINNE
empirical data: in nature mortality often reduces et al. 1992; DIXON 2000), but probably not
juvenile survival dramatically, as only about 1% survival – adults can fly and therefore are less
of the eggs laid in a patch may survive (OSAWA likely to die from starvation, compared to their
1989; AGARWALA & DIXON 1993; HIRONORI & much less mobile larvae. An adult can, however,
KASUHIRO 1997; KINDLMANN et al. 2000). strongly influence the survival of its offspring by
The relative strength of selection acting on an carefully selecting oviposition sites. The best sites
increase in fecundity and better survival can be are those, which contain enough food for their
66
larvae and in which the risk of their being killed use cues that enable them to avoid being eaten by
is low. other conspecific predators will be strongly
Existence of patches of prey is time limited in selected for. This prediction is empirically
many species. Aphids are one example: their supported by the discovery of a pheromone that
colonies last only for a few months (DIXON deters adults of many different groups of
1997). By definition, the same is true for many predatory insects from ovipositing in patches of
other insect pests: they often severely damage prey contaminated with conspecific larval tracks
their host plant. If this is true, then it does not (HEMPTINNE et al. 1992; DOUMBIA et al. 1998).
make much sense to stay on a dead plant and it is Avoiding oviposition in patches that are being
advantageous for them to migrate somewhere attacked by larvae of predators could be
else, which means that the existence of patches is important for another reason: presence of larvae
time limited. In addition, the developmental time is an indicator of the age of a patch of prey.
of many insect predators is comparable with the Predators need some time to find a patch and the
duration of the patches of their prey (DIXON existence of larvae in a patch means that the patch
2000; KINDLMANN & DIXON 1993, 1999, 2001). must have existed for at least E time units ago,
For example, developmental time of where E is the developmental time of the egg. In
aphidophagous ladybirds spans more than 1 contrast the probability that a patch is not found
month (about 6 weeks, depending mainly on by a predator clearly declines with the age of the
temperature), which is comparable to the duration patch. Therefore, a patch that has not yet been
of an aphid colony. In such circumstances it is found by a predator is not likely to exist for long.
selectively advantageous to oviposit only during a Existence of predatory larvae in a patch of prey
short period, “egg window”, early in the existence may indicate there is insufficient time for
of a patch of prey, as late laid eggs are unable to offspring, to complete their development because
complete their development before prey becomes the patch is in a relatively advanced stage of
scarce. Therefore genotypes that are able to use development.
cues that enable them to estimate the age of a Thus the reason why adults migrate between
patch of prey, and lay eggs only in patches in the prey patches so frequently is that it enables them
early stages of development will be strongly to optimise the distribution of their eggs
selected for. (numerical response), rather than optimise
Another trait that increases the likelihood of foraging in the terms of maximising food
survival of a juvenile predator is its ability to consumption per unit time (functional response).
avoid being killed. This can be partially achieved This numerical response, however, differs from
via a large voracity as by eating a lot larvae that normally used as it reflects the ages of
become strong and better able to defend patches of a prey, rather than the amount of food
themselves, but is most likely more determined they contain. It is surprising that this was
by adults avoiding ovipositing in “dangerous” completely ignored in the models of predator-
patches. Which patches are “dangerous”? Most prey population dynamics.
insect predators are cannibalistic (FOX 1975;
OSAWA 1989, 1992a, 1992b, 1993; DIXON 2000).
GENERATION TIME RATIO
This is a strong selective force: on meeting
another predator that is edible, belongs to a
different genotype and is weaker than the Ovipositing during the short egg window early in
potential cannibal, it is advantageous to eat it, as the existence of a patch of prey is advantageous,
it is a source of energy and a potential competitor only if developmental time of the predator is
for food. Thus non-sibling cannibalism is comparable with the average duration of patches
undoubtedly advantageous. Even sibling of prey. If the generation time of the organism
cannibalism may have a selective advantage, if consuming the prey is short, then this advantage
prey becomes scarce (OSAWA 1989, 1992b). If disappears, as there is no penalty for reproducing
cannibalism is an important cause of mortality in late. This is the case, e.g., of insect parasites, the
insect predators, then genotypes that are able to developmental time of which is comparable with
67
that of their host and which are therefore able to THE PREDATOR – PREY MODEL
complete several generations during the existence
of a patch of prey. Adults of each generation Biological Assumptions
redistribute themselves: they can either oviposit
in the patch where they were born, if there are not
many competitors and a lot of prey. However, Insect herbivores have frequently been observed
they can also fly away and reproduce in another to first increase and then decline in abundance,
patch of prey. Therefore, short lived predators even in the absence of natural enemies (DIXON
(like parasites) have several opportunities during 1997, 2000). As explained above, such declines
the existence of a patch of prey to “adjust their are not likely to be due to regulation by predators.
strategy”: by ovipositing or flying away they can It was shown that prey individuals respond
respond to changes in prey abundance and negatively to either lowering the carrying
therefore deplete the patch more effectively – a capacity of their host plant (deterioration of food
possibility not available to predators that quality during the season) or to their cumulative
complete only one generation in a patch. density, or to both by emigrating from patches
The situation is similar to an aircraft trying to and colonizing empty patches (KINDLMANN &
follow the terrain: a slow flying aircraft (short- DIXON 1996, 1997). Here we assume that the
lived predator) can adjust its trajectory to the regulatory term for prey, when alone, is its
changes in the terrain much better than a fast cumulative density, h, instead of some function of
flying one (long-lived predator). Thus in long- its instantaneous density. In contrast to the
lived predators selection favours those individuals logistic or exponential growth models, this
that are able to select patches for oviposition that function allows prey to decline in abundance with
contain sufficient prey and which will remain increasing time even in the absence of natural
abundant for long enough to sustain its larvae enemies. Analysis of other alternatives can be
(which results in egg windows etc.). These found in KINDLMANN et al. (2003).
predators are unlikely to reduce prey abundance We assume that predators only oviposit during
to the same extent as short-lived predators. In the egg window early in the development of
other words, if the ratio of the developmental patches of prey. We follow the dynamics within a
time of the predator to that of its prey patch of prey from the instant when the egg
(“generation time ratio”, GTR – KINDLMANN & window closes. Thus the initial density of a
DIXON 1999) is large, then from an evolutionary predator in a patch is defined by the number of
perspective the predator has to "project" far into eggs laid there by adults that developed in other
the future. If the existence of a patch of prey is patches of prey and oviposited there during the
limited in time, then it is advantageous for “egg window”. Changes over time in the number
predators to lay eggs early in the existence of a of predators within a patch are assumed to be due
patch, as future prey availability is uncertain. This to larval cannibalism and not reproduction.
uncertainty also makes cannibalism We assume the predator is cannibalistic but
advantageous. Because of the risk of cannibalism, has a preference, p, for eating prey, as opposed to
predators tend to lay fewer eggs in a patch, but conspecifics. If they prefer prey, then p > 1, but p
continue to oviposit until cues indicate that it is may also be smaller than one, as for example
highly likely that their eggs will be eaten by when the larvae of a predator prefer to eat
conspecific larvae. This is when the “egg conspecific eggs, which cannot defend
window” closes and ovipositing predators themselves. If p = 1, the predator shows no
abandon a prey patch. Cannibalism thus acts to preference for either prey or conspecifics (the
regulate the numbers of predators per patch “meet and eat” hypothesis).
(MILLS 1982). Between-season dynamics are determined by
68
iterating the within-season dynamics. In this 10000

illustrative example we assume that the predator a
Prey numbers, x
is univoltine, its prey achieves only one peak in 1000
y=0
abundance during a season and that both predator y = 40
and prey redistribute themselves uniformly 100
between seasons, so that the initial numbers of
both predator and prey are the same for all 10
patches. The numbers of prey next spring is
calculated by multiplying its autumn numbers by 1
winter mortality, and that of the predator by 0 10 20 30 40 50
multiplying its autumn numbers by winter Time, t
mortality and its fecundity. Between seasons both 50
b
Predator numbers, y
prey and predators redistribute themselves among
the many patches that make up the population. 40
30
Within-Season Dynamics 20
10
Within-season, the dynamics of insect predator -
prey system can be described by (KINDLMANN & 0
DIXON 1993, 2002): 0 10 20 30 40 50
Time, t
dh
ax , h(0) = 0 (1a) Fig. 1. Trends in time in prey (a) and predator (b)
dt abundance predicted by the model when a = .000005, r
= .3, v = 1, b = 0, p = 1, x0 = 100, y0 = 0 and y0 = 40. In
dx vpxy (a) prey density in the absence of predators and the
( r h) x , x(0) = x0 (1b)
dt b px y presence of 40 predators (see inset) is also presented.
Equation (1a) describes changes in cumulative

dy vy 2 ,
y(0) = y0 (1c) density of prey, (1b) describes changes in prey
dt b px y density, and (1c) describes the decrease in
predator density due to cannibalism. A typical
where: trend in numbers in a patch predicted by model
(1) is shown in Fig. 1. There is no further predator
h(t) - cumulative density of the prey at time t reproduction in a patch; therefore, predator
x(t) - density of prey at time t numbers monotonously decline. As a
a - scaling constant relating prey consequence, if prey abundance (x) increases at
cumulative density to its own dynamics the beginning (i.e., if y0 is sufficiently small, so
r - maximum potential growth rate of the dx 0 ), then as time proceeds the
prey that lim
t o0 dt
y(t) - density of predator at time t dynamics of the prey is less and less influenced
v - predator voracity by the declining numbers of the predator. Because
b - parameter of the functional response of of the way the diet of the predator is defined (the
the predator terms containing v in [1b] and [1c]), the decline in
p - predator's preference for prey predator numbers is more pronounced when there
T - time when predator matures; coincides are few prey individuals relative to predator
with the duration of a patch of prey, yielding individuals. That is, when the ratio x/y is small at
initial values x(T) and y(T) for the next season. the beginning and when prey numbers have
69
passed their peak and become small again due to Biologically:

the negative effect of cumulative density.
Initial Self-regulation
Predators have almost no influence on the prey number of prey
Effect of predator
numbers on prey
dynamics in this system (KINDLMANN & DIXON of prey abundance this dynamics this year
1993). Not surprisingly, the number of predators next year year
= +
that survive is positively influenced by the initial
number of prey and negatively influenced by the Initial Allometric Self-regulation
initial number of predators (KINDLMANN et al. number of function of No. of predator
predators of prey this abundance this
unpub.). next year year year
The predicted trends in abundance (Fig. 1) = u
closely match those observed in nature in aphids In this system when predators are absent, the
(DIXON et al. 1996; KINDLMANN & DIXON 1996, number of prey next year increases when the
1997; DIXON & KINDLMANN 1998; KINDLMANN number of prey this year is low, but is strongly
et al. 2003) and ladybird beetles (OSAWA 1993; regulated by itself (Xt+1 = a1 – b1Xt). Influence of
HIRONORI & KATSUHIRO 1997; YASUDA & the predator on prey dynamics is expressed by the
OHNUMA 1999; KINDLMANN et al. 2000; YASUDA term (c1 - d1Xt).Yt. The number of predators next
et al. 2004). year is positively, but less than linearly,
influenced by the number of prey this year (see
Between-Season Dynamics the term a 2 X tb 2 ). The shape of the between-
season relation in the number of predators is
The above equations (1) translate the initial prey parabolic, indicating that predators do best at
and predator numbers (x[0] and y[0]) into their intermediate densities - when there are few
final numbers (x[T] and y[T]) at the end of the predators, few of them survive, and because of
season. Between-season dynamics of this system cannibalism, few survive even when they are
can then be obtained iteratively: If we denote the initially numerous.
total number of prey and the total number of
predators in a population consisting of n patches CONCLUSIONS
at the beginning of the year t, by xt and yt,
respectively, and if xt = n. x(0) and yt = n. y(0), To conclude, in situations where the
then xt+1 = n.dx.x(T) and yt+1 = n. dy.f. y(T), where developmental time of the predator is long
dx and dy are the probabilities of prey and relative to its prey, predators are unlikely to be
predators, respectively, surviving from the end of effective classical biological control agents. This
one season to the beginning of the next and f is is because the predator abundance is strongly
predator fecundity. It is not possible to derive the regulated by cannibalism (KINDLMANN & DIXON
exact relations between xt and yt and xt+1 and yt+1 2001). This is well illustrated by the
from model (1), but approximate relations were aphidophagous ladybird - aphid system.
obtained by KINDLMANN & DIXON (2002), who
linearised the dependence by making Xt = ln(xt) REFERENCES
and Yt = ln(yt). This results in the following
system of difference equations: AGARWALA, B.K. & A.F.G. Dixon 1992. Laboratory
study of cannibalism and interspecific predation in
X t 1 a1 b1 X t (c1 d1 X t ).Yt ladybirds. Ecoogical Entomology 17: 303-309.
BEDDINGTON, J.R., C.A. FREE & J.H. LAWTON 1976.
(2) Concepts of stability and resilience in predator-prey
Yt 1 a 2 X tb .(c2Yt d 2Yt 2 )
2
models. Jornal of Animal Ecology 45: 791-816.
70
BEDDINGTON, J.R., C.A. FREE & J.H. LAWTON 1978. KINDLMANN, P. & A.F.G. DIXON 1999. Generation
Characteristics of successful natural enemies in time ratios - determinants of prey abundance in
models of biological control of insect pests. Nature insect predator - prey interactions. Biological
273: 513-519. Control 16: 133-138.
DIXON, A.F.G. 1997. Aphid ecology: an optimization KINDLMANN, P. & A.F.G. DIXON 2001. When and why
approach. London, UK: Chapman and Hall. 312 top-down regulation fails in arthropod predator-
pp. prey systems. Basic and Applied Ecology 2: 333-
DIXON, A.F.G. 2000. Insect predator-prey dynamics 340.
ladybird beetles and biological control. KINDLMANN, P. & A.F.G DIXON 2003. Insect predator-
Cambridge, UK: Cambridge University Press. 257 prey dynamics and the biological control of aphids
pp. by ladybirds. pp. 118-124 in R. VAN DRIESCHE
DIXON A.F.G. & P. KINDLMANN 1998. Population (Ed). International symposium on biological
dynamics of aphids. Pp. 207-230 in J.P. DEMPSTER control of arthropods. USDA Forest Service,
& I.F.G. MACLEAN (Eds). Insect populations. publication FHTET-03-05.
Dordrecht, The Netherlands: Kluwer Academic KINDLMANN, P., A.F.G. DIXON & V. JAROSIK in press.
Publishers. 512 pp. Population dynamics. in H. VAN EMDEN & R.
DIXON, A.F.G., P. KINDLMANN & R. SEQUEIRA 1996. HARRINGTON (Eds). Aphids as insect pests.
Population regulation in aphids. Pp. 77-88 in R.B. KINDLMANN, P., H. YASUDA S. SATO & K. SHINYA
FLOYD, A.W. SHEPPARD & P.J. DEBARRO (Eds). 2000. Key life stages of two predatory ladybird
Frontiers of population ecology. Melbourne, species. European Journal of Entomology 97: 495-
Australia: CSIRO Publishing. 639 pp. 499.
DOUMBIA, M., J.L. HEMPTINNE & A.F.G. DIXON 1998. MILLS, N.J. 1982. Voracity, cannibalism and
Assessment of patch quality by ladybirds: role of coccinellid predation. Annals of Applied Biology
larval tracks. Oecologia 113: 197-202. 101: 14-148.
FOX, L.R. 1975. Cannibalism in natural populations. MURDOCH, W.W. 1994. Population regulation in theory
Annual Review of Ecology and Systematics 6: 87- and practice. Ecology 75: 271-287.
106. OSAWA, N. 1989. Sibling and non-sibling cannibalism
GODFRAY, H.C.J. & M.P. HASSELL 1987. Natural by larvae of a lady beetle Harmonia axyridis Pallas
enemies may be a cause of discrete generations in (Coleoptera: Coccinellidae) in the field. Research
tropical insects. Nature 327: 144-147. in Population Ecology 31: 153-160.
HASSELL, M.P. 1978. The dynamics of arthropod OSAWA, N. 1992a. Sibling cannibalism in the lady
predator-prey systems. Princeton, NJ: Princeton beetle Harmonia axyridis: fitness consequences for
University Press. 237 pp. mother and offspring. Research in Population
HEMPTINNE, J.L., A.F.G. DIXON & J. COFFIN 1992. Ecology 34: 45-55.
Attack strategy of ladybird beetles (Coccinellidae): OSAWA, N. 1992b. Effect of pupation site on pupa;
factors shaping their numerical response. cannibalism and parasitism in the ladybird beetle
Oecologia 90: 238-245. Harmonia axyridis Pallas (Coleoptera:
HIRONORI, Y. & S. KATSUHIRO 1997. Cannibalism and Coccinellidae). Japanese Journal of Entomology
interspecific predation in two predatory ladybird 60: 131-135.
beetles in relation to prey abundance in the field. OSAWA, N. 1993. Population field studies of the
Entomophaga 42: 153-163. aphidophagous ladybird beetle Harmonia axyridis
KINDLMANN, P. & A.F.G. DIXON 1993. Optimal (Coleoptera: Coccinellidae): life tables and key
foraging in ladybird beetles (Coleoptera: factor analysis. Research in Population Ecology
Coccinellidae) and its consequences for their use in 35: 335-348.
biological control. European Journal of YASUDA, H. & N. OHNUMA 1999. Effect of
Entomology 90: 443-450. cannibalism and predation on the larval
KINDLMANN, P. & A.F.G. DIXON 1996. Population performance of two ladybirds. Entomologia
dynamics of tree-dwelling aphids: from individuals Experimentalis et Applicata 93: 63-67.
to populations. Ecological Modelling 89: 23-30. YASUDA H., P. KINDLMANN, K. SHINYA & S. Sato in
KINDLMANN, P. & A.F.G. DIXON 1997. Patterns in the press. Intra- and interspecific interactions on
population dynamics of the Turkey-oak aphid. Pp. survival of two predatory ladybirds in relation to
219-225 in J.M. NIETO NAFRIA & A.F.G. DIXON prey abundance: competition or predation. Applied
(Eds). Aphids in natural and managed ecosystems. Entomology and Zoology.
Leon, Spain: Universidad de Leon. 688 pp.
71
72
OVIPOSITION BEHAVIOUR OF THE TWO-SPOT LADYBIRD BEETLE Adalia

bipunctata (L.) (COLEOPTERA: COCCINELLIDAE) ON PLANTS WITH
CONSPECIFIC LARVAL TRACKS.
B. FRÉCHETTE, C. ALAUZET & J.-L. HEMPTINNE
FRÉCHETTE, B., C. ALAUZET & J.-L. HEMPTINNE 2003. Oviposition behaviour of

the two-spot ladybird beetle Adalia bipunctata (L.) (Coleoptera: Coccinellidae) on
plants with conspecific larval tracks. Pp. 73-77 in A.O. SOARES, M.A. VENTURA,
V. GARCIA & J.-L. HEMPTINNE (Eds) 2003. Proceedings of the 8th International
+ 112 pp.
There are many laboratory studies on the oviposition-deterring effect of conspecific larval
tracks on the two-spot ladybird beetle, Adalia bipunctata (L.). However, this phenomenon
has not been studied in the field. In near natural field conditions A. bipunctata foraging on
aphid infested plants, with or without conspecific larval tracks, were reluctant to lay eggs
on and quickly left the plants with conspecific larval tracks compared to the control plants.
The results of this study indicate that A. bipunctata reacts to conspecific larval tracks on
plants in near natural conditions, as they do in the laboratory.
Bruno Fréchette (e-mail : frechette_bruno@yahoo.ca), Claude Alauzet, Université Paul

Sabatier. Laboratoire d'écologie terrestre. UMR 5552. 118 route de Narbonne (4 R 3). FR-
31062 Toulouse cedex 04. France & Jean-Louis Hemptinne, Ecole Nationale de
Formation Agronomique. Laboratoire d'Agro-écologie 000271JE1. B.P. 87. FR-31326
Castanet-Tolosan cedex. France.
INTRODUCTION observations led KINDLMANN & DIXON (1993,

1999) to propose that there is in the development
Aphidophagous ladybird beetles seeking of aphid patches a time, "egg window", when it is
oviposition sites have to move between patches of most advantageous for ladybirds to lay their eggs.
aphids, evaluate patch quality and decide whether This stresses the importance of the age of an
to feed and/or lay eggs. Patch quality is important aphid colony in determining its suitability for
since it determines the probability of survival of oviposition. Since ladybird larvae and aphid
their progeny (KINDLMANN & DIXON 1993, colonies have similar development times, the
1999). theory stipulates that ladybird should lay a limited
Ladybirds should avoid ovipositing in patches number of eggs early in the development of aphid
of aphids that are already exploited by colonies. If eggs are laid later the aphid colony
conspecific larvae. Laying eggs in such patches might disperse and disappear before the ladybird
will greatly increase the risk of cannibalism larvae complete their development. In the absence
(MILLS 1982), which is an important mortality of food the risk of cannibalism greatly increases
factor in natural population of several species of (AGARWALA & DIXON 1992).
ladybird beetles (OSAWA 1989, 1993; HIRONORI Females of the two-spot ladybird beetle,
& KATSUHIRO 1997). Adalia bipunctata (L.), do not use the age
Many field observations indicate that ladybird structure of aphid colonies to asses their quality
eggs are often laid near aphid colonies that are in as oviposition sites (HEMPTINNE et al. 2000).
an early phase of development (BANKS 1955; However, in Petri dishes they refrain from
HEMPTINNE et al. 1992). These and other ovipositing when confined with conspecific
73
larvae or adults (HEMPTINNE et al. 1992), or on any eggs laid were counted and removed daily.
substrates contaminated with conspecific larval Ladybirds selected for the experiment were
tracks (DOUMBIA et al. 1998). between 10-20 days old and had laid at least one
The oviposition-deterring effect of larvae on egg batch in the last 5 consecutive days.
the reproductive behaviour of aphidophagous was
first described for Adalia bipunctata L. by
Bean plants
HEMPTINNE et al. (1992) and that it was their
larval tracks for Chrysopa oculata Say by
RģŽIýKA (1994). Subsequently the same Broad-bean plants, Vicia faba L., used in the
phenomenon was recorded for many species of experiment were about 15 cm high and had 6
chrysopids (RģŽIýKA 1996, 1997a, 1997b, 1998), leaves. Experimental plants were those on which
ladybirds (RģŽIýKA 1997b, 2001; YASUDA et al. 5 fourth instar larvae had walked for 45 minutes
2000; HEMPTINNE et al. 2001) and the 16 hours before the beginning of the experiment.
cecidomyiid fly Aphidoletes aphidimyza Both control and experimental plants were then
(Rondani) (RģŽIýKA & HAVELKA 1998). infested with about 100 aphids and left for 15
The experiments on the reproductive hours for the aphids to settle.
behaviour of the two-spot ladybird beetle A.
bipunctata were done in the laboratory. The goal Experiments
of this study was to determine the oviposition
response of A. bipunctata in near natural Experiments were performed outdoors in the
conditions on broad bean plants contaminated shade between 10:00 am and 1:00 pm in June and
with larval tracks. July 2001 and 2002. One female was released at
YASUDA et al. (2000) counted the number of the base of a bean plant and (1) whether or not it
eggs laid by Harmonia axyridis Pallas on plants laid eggs, (2) the time spent Walking, Resting,
previously walked on by larvae of H. axyridis and Eating and Ovipositing and (3) its residence time
Coccinella septempunctata L. This study was were recorded. Whether the beetles walked or
done in cages and the behaviour of the females flew off the plants was also recorded. Mean
was not observed. HEMPTINNE et al. (1992) temperature at the beginning of the experiment
observed that A. bipunctata confined in Petri was 22.3 r 0.3°C. There were 40 replicates for
dishes with conspecific larvae were more active each treatment.
than control beetles. Therefore, we also recorded
the behaviour of the females.
Statistical analysis
MATERIALS & METHODS Differences in proportions were compared using
Pearson tests. Percentages were arcsin
Ladybirds transformed before analysis and then compared
using a t-test or a Wilcoxon test if the distribution
The eggs of the two-spot ladybird, A. bipunctata, of the results was not normal according to the
were obtained from a stock culture. Larvae were Shapiro-Wilk test. All analysis were made using
reared at 20 r 1°C, LD 16:8, and fed an excess of Jmp In£ (SALL et al. 2001). For the residence
pea aphids, Acyrthosiphon pisum Harris. Adults times, the values for experimental and control
were isolated within 24 hours of their emergence beans were sorted into three categories: short (0-
from pupa. When their integuments had hardened, 60 min), medium (60-120 min) or long residence
their sex was determined and they were kept in times (more than 120 min). The distributions of
male female pairs. Every day, these pairs of frequencies were compared using a Log
adults were put in clean Petri dishes with a piece Likelihood Ratio test computed by hand
of corrugated paper and an excess of pea aphids; following ZAR (1996).
74
RESULTS percentage of the time allocated to Oviposition

(Wilcoxon: F²=7.9929; df=1; p=0.0047), there
Oviposition were no significant differences in that allocated to
Walking (t-test: F=0.9087; df=1; p=0.3447),
A significantly greater proportion of females laid Resting (t-test: F=0.0060; df=1, p=0.9387) or
eggs on control than on experimental plants (Fig. Eating (t-test: F=0.0991; df=1; p=0.7541) (Fig.
1; Pearson: F²=11.114; df=1; p=0.0009). Only 3). In both treatments, a similar percentage of
one female (2.5%) laid eggs on an experimental beetles (12.5%) left the plants without eating,
plant. while 87.5% ate at least one aphid. Of the females
that laid eggs and completed oviposition before
Ov ipos ition the experiment ended, 80% (8 out of 10) ate at
No Ov ipos ition least 1 aphid before leaving the plant.
Ex perimental 97,5%
60
%
50
40
Control 30% 70% 30
20
0 50 100 10 *
%
0
Fig. 1. The percentage of A. bipunctata females that Resting Walking Eating Oviposition
laid eggs on control and experimental plants. Control Experimental
Fig. 3. The percentage of time spent in different

Residence time activities by A. bipunctata females on control and
experimental plants. Significant difference (p<0.05) is
Ladybirds tended to stay for a shorter period of indicated by an asterisk (*).
time (0-60 min) on experimental plants than on
control plants (Figure 2; Log Likelihood Ratio: Way of leaving the plants
G=8.48; df=2; p<0.05).
Of the ladybirds that left plants before the end of
25 3 hours (n=24 for control plants, n=31 for
20 experimental plants) most left by flying (70.8%
Nb Individuals
for control plants, 77.4% for experimental plants),

15
and the rest by walking (29.2% for control plants,
10 22.6% for experimental plants). The differences
5 in percentages between treatments were not
0 significant (Pearson: F²=0.309; df=1; p=0,5782).
Short Medium Long However, temperature had a significant effect
Experimental Control on the tendency to fly (Pearson: F²=8.744; df=1;
Fig. 2. The number of A. bipunctata females that had p=0.0031). For both treatments combined, the
short, medium, and long residence times on control and percentage flying was significantly higher at t
experimental plants. 21°C (83.7%, n=43) than at < 21°C (41.7%,
n=12).
Behaviour
DISCUSSION
Only the behaviour of beetles that stayed at least
30min on the plants (n=30 for control, n=26 for This study provides strong evidence that the
experimental) was analysed. Other than in the effects of the pheromone in larval tracks on the
75
oviposition behaviour of A. bipunctata observed increase in abundance of its aphid prey on lime
in the laboratory are also relevant to natural trees in the field (WRATTEN 1973). When cereal
conditions. A. bipunctata females leavesearlier aphid density is augmented in wheat fields there
and are reluctant to lay eggs on plants on which is an augmentation in the adult numbers of both
conspecific larvae have previously walked, even Hippodamia convergens Guérin-Méneville and C.
if food is abundant. septempuctata, but egg density is very low and
However, the agitated behaviour observed by about the same in both augmented and control
HEMPTINNE et al. (1992) was not observed in this patches of prey (ELLIOTT & KIECKHEFER 2000).
study. Females spend the same percentage of time The effect of larval tracks on the distribution
walking, resting and eating on both control and of aphidophagous ladybirds could affect their
experimental plants. Moreover, even though potential as biological control agents. By only
females tended to leave experimental plants laying a few eggs early in the development of an
earlier than control plants, the same percentage aphid colony (KINDLMANN & DIXON 1993),
(87.5%) ate at least one aphid before leaving in ladybirds are unable to marked aggregative
both treatments. This indicates that females may response to patches of prey, a characteristic
refrain from ovipositing on aphid infested bean feature of effective biological control agent
plants with larval tracks, but will stay on these (BEDDINGTON et al. 1978).
plants and feed when hungry. That is, the only
difference in behaviour was the acceptance or
ACKNOWLEDGEMENTS
rejection of the plant as an oviposition site.
Even though the results clearly demonstrate
that larval tracks deter oviposition, the low The first author was supported by a graduate
percentage of females that laid eggs on control scholarship from Fond québécois de la recherche
plants (30%) is worrying. The reason may be that sur la nature et les technologies. We are grateful
the ladybirds used in the experiment were not all to Dr. A.F.G. Dixon and Dr. P. Kindlmann for
equally 'ready' to oviposit within the 3 hour comments, discussion and language improvement
observation period. It is possible that female of this manuscript.
foraging behaviour depends on some intrinsic
conditions, such as their egg load (MIKENBERG et
REFERENCES
al. 1992). A female that is not ready to oviposit
may respond differently to environmental cues
than one that is about to lay eggs. In addition, a AGARWALA, B.K. & A.F.G. DIXON 1992. Laboratory
female that takes a long time to find a 'suitable' study of cannibalism and interspecific predation in
site for oviposition may accept a site that she ladybirds. Ecological Entomology 17: 303-309.
BANKS, C.J. 1955. An ecological study of
would have normally rejected as in HEMPTINNE et
Coccinellidae (Col.) associated with Aphis fabae
al. (1992). There is need for a better Scop. on Vicia faba. Bulletin of Entomological
understanding of the physiological ecology of Research 46: 561-587.
ladybirds. BEDDINGTON, J.R., C.A. FREE & J.H. LAWTON 1978.
A better knowledge of a ladybird's oviposition Characteristics of successful natural enemies in
cycle could lead an improvement in experimental models of biological control of insect pests. Nature
design and clearer results. Working only with 273: 513-519.
beetles ready to, but not desperate to oviposit may DOUMBIA, M., J.-L. HEMPTINNE & A.F.G. DIXON 1998.
result in a greater percentage ovipositing. Assessment of patch quality by ladybirds: role of
larval tracks. Oecologia 113: 197-202.
The response of adult ladybirds to larval
ELLIOTT, N.C. & R.W. KIECKHEFER 2000. Response by
tracks could help to explain how ladybirds coccinellids to spatial variation in cereal aphid
distribute themselves in the environment. It may density. Population Ecology 42: 81-90.
account for why even in patches of high aphid HEMPTINNE, J.-L., A.F.G. DIXON & J. COFFIN 1992.
density there are often few ladybird eggs. For Attack strategy of ladybird beetles (Coccinellidae):
example, there is not a density dependent increase factors shaping their numerical response.
in oviposition by A. bipunctata in response to an Oecologia 90: 238-245.
76
HEMPTINNE, J.-L., M. Doumbia & A.F.G. DIXON 2000. RģŽIýKA, Z. 1996. Oviposition-deterring pheromone in
Assessment of patch quality by ladybirds: role of chrysopid: intra- and interspecific effects.
aphid and plant phenology. Journal of Insect European Journal of Entomology 93: 161-166.
Behavior 13: 353-359. RģŽIýKA, Z. 1997a. Persistence of the oviposition-
HEMPTINNE, J.-L., G. LOGNAY, M. DOUMBIA & A.F.G. deterring pheromone in Chrysopa oculata (Neur.:
DIXON 2001. Chemical nature and persistence of Chrysopidae). Entomophaga 42: 109-114.
the oviposition deterring pheromone in the tracks RģŽIýKA, Z. 1997b. Recognition of oviposition
of the larvae of the two spot ladybird, Adalia deterring allomones by aphidophagous predators
bipunctata (Coleoptera: Coccinellidae). (Neuroptera: Chrysopidae, Coleoptera:
Chemoecology 11: 43-47. Coccinellidae). European Journal of Entomology
HIRONORI, Y. & S. KATSUHIRO 1997. Cannibalism and 94: 431-434.
interspecific predation in two predatory ladybirds RģŽIýKA, Z. 1998. Further evidence of oviposition-
in relation to prey abundance in the field. deterring allomone in chrysopids (Neuroptera:
Entomophaga 42: 153-163. Chrysopidae). European Journal of Entomology
KINDLMANN, P. & A.F.G. DIXON 1993. Optimal 95: 35-39.
foraging in ladybird beetles (Coleoptera: RģŽIýKA, Z. 2001. Oviposition responses of
Coccinellidae) and its consequences for their use in aphidophagous coccinellids to tracks of coccinellid
biological control. European Journal of (Coccinellidae) and chrysopid (Chrysopidae)
Entomology 90: 443-450. larvae. European Journal of Entomology 98: 183-
KINDLMANN, P. & A.F.G DIXON 1999. Strategies of 188.
aphidophagous predators: lesson for modelling RģŽIýKA, Z. & J. HAVELKA 1998. Effects of
insect predator-prey dynamics. Journal of Applied oviposition-deterring pheromone and allomones on
Entomology 123: 397-399. Aphidoletes aphidimyza (Diptera : Cecidomyiidae).
MIKENBERG, O.P.J.M., M. TATAR & J.A. ROSENHEIM European Journal of Entomology 95: 211-216.
1992. Egg load as a major source of variability in SALL, J., A. LEHMAN & L. CREIGHTON 2001. JMP£
insect foraging and oviposition behavior. Oikos 65: start statistics (2nd ed.). Pacific Grove (CA):
134-142. Duxbury. 656 pp.
MILLS, N.J. 1982. Voracity, cannibalism and WRATTEN, S.D. 1973. The effectiveness of the
coccinellid predation. Annals of Applied Biology coccinellid beetle, Adalia bipunctata (L.), as a
101: 144-148. predator of the lime aphid, Eucallipterus tiliae L.
OSAWA, N. 1989. Sibling and non-sibling cannibalism Journal of Animal Ecology 42: 785-802.
by larvae of a lady beetle Harmonia axyridis Pallas YASUDA, H., T. TAKAGI & K. KOGI 2000. Effect of
(Coleoptera: Coccinellidae) in the field. Researches conspecific and heterospecific larval tracks on the
on Population Ecology 31: 153-160. oviposition behaviour of the predatory ladybird,
OSAWA, N. 1993. Population field studies of the Harmonia axyridis (Coleoptera: Coccinellidae).
aphidophagous ladybird beetle Harmonia axyridis European Journal of Entomology 97: 551-553.
(Coleoptera: Coccinellidae): life tables and key ZAR, J.H. 1996. Biostatistical analysis. (3rd ed.). Upper
factor analysis. Researches on Population Ecology Saddle River (NJ): Prentice-Hall International Inc.
35: 335-348. 718 pp.
RģŽIýKA, Z. 1994. Oviposition-deterring pheromone in
Chrysopa oculata (Neuroptera: Chrysopidae).
European Journal of Entomology 91: 361-370. Accepted 31 May 2003.
77
78
ANT TRAIL: A HIGHWAY FOR Coccinella magnifica REDTENBACHER

(COLEOPTERA: COCCINELLIDAE)
J.-F. GODEAU, J.-L. HEMPTINNE & J.-C. VERHAEGHE
GODEAU, J.-F., J.-L. HEMPTINNE & J.-C. VERHAEGHE 2003. Ant trail: a highway
for Coccinella magnifica Redtenbacher (Coleoptera: Coccinellidae). Pp. 79-83 in
A.O. SOARES, M.A. VENTURA, V. GARCIA & J.-L. HEMPTINNE (Eds) 2003.
Proceedings of the 8th International Symposium on Ecology of Aphidophaga:
Biology, Ecology and Behaviour of Aphidophagous Insects. Arquipélago. Life
and Marine Sciences. Supplement 5: x + 112 pp.
Red Wood Ant species chemically mark .trails connecting their nest to aphid colonies. The
trail pheromones are produced in the ant's hindgut. The large trails are maintained thanks to
visual clue and because new pheromone marks are regularly laid down. As a consequence,
Red Wood Ant's workers create a network of trails equivalent to hundreds of meters around
every nest mound. Coccinella magnifica is an aphidophagous ladybird species living only
in the vicinity of Red Wood ant's nests. It preys upon aphid colonies regardless of ant's
presence. We show here that, in the laboratory, C. magnifica can follow a recruitment trail
laid down by Formica polyctena under artificial conditions with its hindgut's pheromone.
This is confirmed by field observations of ladybirds walking along natural Red Wood Ant's
trails. As C. magnifica appears to be a specialist, this behaviour is adaptive to efficiently
locate prey.
Jean-François Godeau (e-mail: jean-francois.godeau@umh.ac.be), Laboratoire De

Zoologie (Université de Mons-Hainaut) Avenue du Champs de Mars, 6, BE-7000 Mons,
Belgium; Jean-Louis Hemptinne, Ecole nationale de Formation agronomique) Dept. ARGE
BP 87 FR-31326 Castanet-Tolosan, France & Jean-Claude Verhaeghe, Université de
Mons-Hainaut Avenue du Champs de Mars, 6, BE-7000 Mons, Belgium.
INTRODUCTION important role in red wood ants orientation.

Ants are largely known as invertebrate
In most temperate forests, lives a well-known ant predators but they also gather honeydew and
species: the red wood ant (Formica rufa sensu occasionally prey upon aphids dwelling in trees
lato) which is a taxonomical complex including and shrubs (e.g. WAY 1963; SKINNER 1980).
some ecological species separated by their habitat What are the relationships between ants,
preferences (YARROW 1955; SEIFERT 1996). It aphids and aphid predators?
lives in large colonies whose nests are built as a Since more than half a century, it is largely
hillock of dead plant fragments. This mound can admitted that ants increase attended aphid
reach 1.5 to 2 meters high and shelter hundreds of populations by cleaning the honeydew droplets
thousands ant individuals (HÖLLDOBLER & glued on the colony and by protecting them
WILSON 1990). The other striking aspect of that against intruders, especially ladybirds, lacewings,
ant are their wide foraging trails, linking the nest bugs and syrphids (e.g. BANKS 1962; WAY 1963;
to food resources scattered in the surrounding HÖLLDOBLER & WILSON 1990).
(SKINNER 1980). Trails are marked by workers Ladybirds found in red wood ant’s habitat are
which use chemical pheromones probably mainly woodland-specialised species whose
produced in the hindgut and in other glands adults and larvae feed on tree aphids. In
(GABBA & PAVAN 1970; PARRY & MORGAN comparison with open areas, aphid resources are
1979; ATTYGALE & MORGAN 1984). Moreover, more dispersed vertically in the forest vegetation
visual cues have been demonstrated to play an stages. All these hotspots of attended aphids are
79
scattered in a three-dimension maze of branches, MATERIAL AND METHODS

which create a more complex situation for food
location than a two-dimensions system in Tests with a naturally-laid trail
herbaceous vegetation. Some ladybirds species
encountered in woodland are: Myzia In the first experiment, we used a naturally-laid
oblongoguttata L. in deciduous ones, or Myrrha ant trail which is obtained as follows. A nest is
octodecimguttata L. in pine forests, or Coccinella connected by a bridge to an area with a sugar
magnifica Redtenbacher in mixed forests, which solution. When the recruitment of workers was
is morphologically very similar to Coccinella important, a cardboard sheet, on which a grid was
septempunctata L. but scarcer because of its traced with a pencil, was place on the bridge
localised distribution (MAJERUS 1994). Some of (shaded squares on Fig. 1). Only a row of squares
these species are occasionally observed on ants was accessible to the ants, the others were
trails, being brought back to the nest. Field covered with two smaller cardboard sheets.
observations revealed that M. oblongoguttata and Workers walked for two hours on this row when
Anatis ocellata L. are sometimes observed as travelling back and forth between the nest and the
prey, carried by F. rufa L. on foraging trails sugary solution. Then, the cardboard sheet was
(SLOGGETT 1999). In that cases, ladybirds have uncovered and deposited in an arena where the
been killed by ants or found already dead. C. tests were performed with C. magnifica adults.
magnifica is the only species usually observed
alive on trails, walking among ant workers
(SLOGGETT 1999, pers. obs.). This species is
known to live around F. rufa’s colonies
(DONISTHORPE 1920a, b; MAJERUS 1989), and it
is suspected to benefit from ant-attended colonies,
by preying upon it, in spite of the ant aggressive
behaviour.
C. magnifica has evolved as a specialised
predator, adapted to avoid ants aggressiveness.
Though, we ignore whether this apparent
immunity is due to the use of appeasement
chemicals produced by the ladybird or only to
behavioural adaptations. Since it is clearly Fig. 1. Cardboard sheet on which a row of squares
established that adults and larvae of C. magnifica (shaded) have been marked by trail pheromones laid by
eat attended aphids, there is no information foraging ants. An sample of a path walked by a
attesting that it also follows ants trails to enhance ladybird is figured.
its food location efficiency. If it was the case, this
adaptation, combined with an apparent immunity Marked and non-marked squares on which the
toward ants, could represent a strong selective ladybird walked were counted during two
advantage. minutes. The comparison between ladybird’s path
We want to test the fact that C. magnifica is and the rows of marked square was obtained by
found along F. rufa’s trails not as a prey but as an using a similarity index SI (1) (VERHAEGHE
intruder. 1982), indicating if there is a preference for
We report here the results of two experiments marked squares:
using a Formica polyctena Förester colony kept
in laboratory and two species of ladybirds, C. Nc 2
S .I . (1)
magnifica and C. septempunctata . Na u Nb
80
where Nc = number of marked squares crossed by All these observations were filmed and
the ladybird, Na = total number of marked analysed afterward. For each sequence, the
squares and Nb = total number of squares crossed detection rate was measured and the “following
by the ladybird. score” was noted. The detection rate is the
Occupation time of each square was also number of times an animal crossed the trail,
measured as an estimation of walk speed. stopped and eventually followed it. The
Each adult was firstly tested with a marked “following score” is the number of 10° arcs
cardboard and then presented with an unmarked consecutively followed by an insect. This score
cardboard. There were 9 replicates with the ant was log-transformed. Both the detection rates and
trails and 8 with a clean cardboard. the “following scores” are presented by their
mean. The statistical difference between detection
rates is obtained by comparing the proportion of
Tests with an artificial circular trail
defectives (detection of trail or not) during all
tests. The result is a binomial distribution on
In the second experiment, we aimed at confirming which we calculated a value for z (NIST/
our first results and obtaining more precise SEMATECH 2003), representing the significance
information about the nature of the observed ant- level for the difference between the two
trail. We made an artificial circular trail, by proportions (test vs. control). The differences
dissecting ants, extracting the gland which between following scores were tested with a
produces the trail pheromones. Ant workers were Median test.
collected on the foraging trails from the
laboratory nest and dissected one by one under
water, with ophthalmologic dissection tools. RESULTS
When one hindgut was isolated, it was transferred
to 75 ml water and crushed. The results of five Tests with a naturally-laid trail
dissections were gathered and deposited on a
strong filter paper with a Stadler® metallic pen. As the trail is deposited by ants, we assume that
This method, described by Pasteels and the right pheromones are present on the cardboard
VERHAEGHE (1974), has already been used in sheet. In comparison with the control,
numerous trail-following experiments with ants C. magnifica walked more often (S.I.Test= 0.4219;
and other invertebrates (CAMMAERTS et al. 1990; S.I.Control= 0.0417; Mann-Whithney U = 109.0;
LENOIR et al. 1991; QUINET & PASTEELS 1995). P<0.01, Fig. 2) and slower (Median occupation
The 12-cm diameter circular trail on a filter time: marked squares: 2.313 sec; covered squares:
paper was put in a circular arena of 25-cm in 1.259 sec; Mann-Whithney U = 126.0, P< 0.001
diameter and left for 10 minutes before the Fig. 3) on trail-marked squares. These results are
beginning of the tests. Two identical sequences of consistent with an olfactory detection and
tests have been performed with the same trail: (1) orientation of C. magnifica thanks to F. polyctena
four ants observed during two minutes; (2) four trail pheromones.
ladybirds (C. magnifica or C. septempunctata)
during four minutes; (3) four ladybirds of the
Tests with an artificial circular trail
second species (C. septempunctata or C.
magnifica) during four minutes; (4) four ants
Over the course of the experiments the quality of
observed again during two minutes to check out
whether the trail is still active. The first sequence the trails deteriorated so that the detection rate of
took place 10 min after the trail has been laid ant F5 was not different in the presence or
down and the second 70 min later. absence of the pheromone (Table 1). C. magnifica
The artificial trail was compared to a control show a strong tendency to respond positively to
trail obtained with pure water. The sequence of the pheromone and not to the control. On the
tests was replicated 5 times with the trail and 4 contrary, the detection rate of C. septempunctata
times with the control. is not modified by ant trails.
81
Median of the following index (+ quartiles)

DISCUSSION
0,6
In the laboratory, C. magnifica, is able to follow a
SI Test
0,5 natural foraging trail delineated by pheromone
SI Control
spots.
Similarity index
0,4
Artificial trails made up of hindgut extracts
0,3
are less efficient than natural trails. They,
however, trigger detection and trail-following by
0,2 ants and C. magnifica.
Our preliminary results might be confirmed
0,1
by testing an extract of 10-hindguts instead of 5,
0,0 expecting a response twice bigger with ants and
C. magnifica.
Fig. 2. Following index performed by C. magnifica on Knowledge about pheromones produced by
a marked sheet (in grey) and on control (in white). Dufour, poison and mandibular glands are quiet
scant in the literature, so that we do not know
Median of the occupation time of squares (+ quartiles) their respective role in the formation of F. rufa
3,5
foraging trails. Additional tests must be
Marked squares performed to check whether one or more of these
3,0
Hidden squares glands could be used to enhance trail-following
2,5 efficiency.
Time (sec)
2,0 The fact that C. magnifica is able to follow

1,5 ants trails is adaptive because these ladybirds are
1,0
able to discover and exploit large aphid colonies
attended by ants.
0,5
This study leads to another question: does the
0,0 trail pheromone intervenes in mate encountering,
selection of oviposition sites and egg laying?
Fig. 3. Occupation time of squares, used as an
estimation of walk speed on marked (in grey) and non-
marked (in white) squares. REFERENCES
ATTYGALE, A. & E.D. MORGAN 1984. Chemicals from

Table 1. the glands of ants. Chemical Society Reviews 13:
The detection rates of F. polyctena (F2, F3, F4 and F5), 245-278.
C.magnifica (CM_1 and CM_2) and C. septempunctata BANKS, C.J. 1962. Effects of the ant Lasius niger (L.)
(C7_1 and C7_2) measured in two sequential on insects preying on small population of Aphis
experiments fabae Scop. on bean plants. Annals of Applied
Detection rate (%) Biology 50: 669-679.
Sequence
Test (N=) Control (N=) z= CAMMAERTS, R., C. DETRAIN & M.-C. CAMMAERTS
F2 60 (225) 45.99 (137) 4.28 (***) 1990. Host trail following by the myrmecophilous
CM_1 45.78 (166) 36.14 (166) 2.58 (*) beetle Edaphopaussus favieri (Fairmaire)
Nr. 1
C7_1 56.25 (64) 51.19 (84) 0.81 (N.S.) (Carabidae Paussinae). Insectes Sociaux 37 (3):
F3 62.11 (190) 28.26 (92) 2.74 (**) 200-211.
F4 57.22 (180) 48.62 (181) 2.31 (*) DONISTHORPE, H.St.J.K. 1920a. The Myrmecophilous
CM_2 41.38 (145) 27.52 (109) 3.74 (***)
Nr. 2 Lady-Bird, Coccinella distincta, Fald., its Life
C7_2 35.58 (104) 37.93 (58) -0.61 (N.S.)
History and Association with Ants. The
F5 50.48 (198) 51.48 (154) -0.61 (N.S.)
Entomologist's Record XXXI (12): 214-222.
DONISTHORPE, H.St.J.K. 1920b. The Myrmecophilous
For all the insects tested, there was no Lady-Bird, Coccinella distincta, Fald., its Life
difference between the “following scores” History and Association with Ants. The
measured with and without pheromone. Entomologist's Record XXXII (1): 1-3 + II
82
GABBA, A & M. PAVAN 1970. Researches on trail and Acta Zoologica Fennica 133: 1-106
alarm substances in ants. Pp. 161-194 in JOHNSTON, SEIFERT, B. 1996. Ameisen: beobachten, bestimmen.
J. W., D. G. MOULTON, AND A. TURK. (Eds). Naturbuch. Verlag. 352 pp.
Advances in chemoreception, Vol. 1. SKINNER, G.J. 1980. The feeding habits of the Wood-
Communication by chemical signals. Ant, Formica rufa (Hymenoptera: Formicidae) in
HÖLLDOBLER, B. & E.O. Wilson 1990. The Ants. Limestone woodland in north-west England.
Springer Verlag, Berlin and Heidelberg. 732 pp. Journal of Animal Ecology 49: 417-433.
MAJERUS, M.E.N. 1989. Coccinella magnifica SLOGGETT, J.J., R.A. WOOD & M.E.N. MAJERUS 1998.
(redtenbacher): a myrmecophilous ladybird. British Adaptations of Coccinella magnifica Redtenbacher,
Journal of Entomological Natural History 2: 97- a Myrmecophilous Coccinellid, to Aggression by
106. Wood Ants (Formica rufa Group). I. Adult
MAJERUS, M.E.N. & P.W.E. KEARNS 1994. Ladybirds. Behavorial Adaptation, Its Ecological Context and
Naturalist's Handbook 10: 339pp. Evolution. Journal of Insect Behavior 11(6): 889-
NIST/SEMATECH e-Handbook of Statistical Methods, 904.
http://www.itl.nist.gov/div898/handbook/, 2003 SLOGGETT, J.J., A. MANICA, M.J. Day & M.E.N.
PARRY, K. & E.D. MORGAN 1979. Pheromones of ants: MAJERUS 1999. Predation of ladybirds (Coleoptera:
a review. Physiological Entomology 4: 161-189. Coccinellidae) by wood ants, Formica rufa L.
PASTEELS, J.M. & J.-C. VERHAEGHE 1974. Dosage (Hymenoptera: Formicidae). Entomologist's
biologique de la phéromone de piste chez les Gazette 50: 217-221.
fourrageuses et les reines de Myrmica rubra. VERHAEGHE, J.-C. 1982. Food recruitment in
Insectes Sociaux 21: 167-180. Tertramorium impurum (Hymenoptera:
QUINET, Y. & J.M. PASTEELS 1995. Trail following and Formicidae). Insectes Sociaux 29 (1): 67-85.
stowaway behaviour of the myrmecophilous WAY, M.J. 1963. Mutualism between ants and honey-
staphylinid beetle, Homoeusa acuminata, during producing homoptera. Annual Review of
foraging trips of its host Lasius fuliginosus Entomology 8: 307-344.
(Hymenoptera: Formicidae). Insectes Sociaux 42: YARROW, I.H.H. 1955. The british ants allied to
31-44. Formica rufa L. (Hym. Formicidae). Transaction
ROSENGREN, R. 1971. Route fidelity, visual memory of the Society for British Entomology 12 (1): 1-48.
and recruitment behaviour in foraging Wood ants
of the genus Formica (Hymenoptera, Formicidae). Accepted 31 May 2003.
83
84
EFFECTS OF CONSPECIFIC AND HETEROSPECIFIC LARVAL TRACKS ON

MOBILITY AND SEARCHING PATTERNS OF Cycloneda limbifer SAY
(COLEOPTERA: COCCINELLIDAE) FEMALES
Z. RģŽIýKA & R. ZEMEK
RģŽIýKA, Z. & R. ZEMEK 2003. Effects of conspecific and heterospecific larval

tracks on mobility and searching patterns of Cycloneda limbifer Say (Coleoptera:
Coccinellidae) females. Pp. 85-93 in A.O. SOARES, M.A. VENTURA, V. GARCIA &
J.-L. HEMPTINNE (Eds) 2003. Proceedings of the 8th International Symposium on
EthoVision, a computerised video tracking system was used to monitor the behaviour of females of an
aphidophagous coccinellid Cycloneda limbifer Say. Time spent on, and distance walked within 30
minutes, were recorded on simultaneously provided clean substrates and substrates with fresh tracks of
conspecific or heterospecific first instar larvae. Females spent longer and walked a greater distance on
substrates with fresh tracks of conspecific larvae than on clean substrates. In contrast, females of C.
limbifer spent less time and walked a shorter distance on substrates with fresh tracks of larvae of the
coccinellid Ceratomegilla (Semiadalia) undecimnotata (Schneider) than on clean substrates. During
the middle period (10-20 minutes) of the tests, the speed of movement of C. limbifer females was
significantly lower on substrates with conspecific larval tracks, and significantly higher on substrates
with tracks of C. undecimnotata larvae than on clean substrates. Previous results show a strong intra-
and interspecific oviposition-deterring effect of fresh larval tracks of these coccinellid species on C.
limbifer. Results presented paradoxically indicate that the tracks of conspecific larvae significantly
increase the time C. limbifer spent on a substrate, while those of C. undecimnotata larvae decrease it.
Zdenik RĤžiþka (e-mail: ruzicka@entu.cas.cz) & R. Zemek, Institute of Entomology,

Academy of Sciences, Branišovská 31, CZ-370 05 ýeské BudČjovice, Czech Republic.
INTRODUCTION coccinellids C. limbifer and C. undecimnotata.

The intensity of the oviposition-deterring effects
Mechanisms regulating complex behaviour in of fresh conspecific tracks and tracks of C.
insects are often found after laborious and time undecimnotata larvae on C. limbifer were similar,
consuming experiments. The incidental discovery but the density of faecal spots on substrates with
of the oviposition-deterring effect of the tracks of tracks of C. undecimnotata larvae was
an aphidophagous insect (RģŽIýKA 1994) is an significantly lower than on clean substrates, that
exception. It came after a lot of effort had been on substrates with fresh conspecific tracks was
devoted to investigating spacing in different slightly higher, though not significantly so, than
predators (SOLOMON 1949, 1964; KUCHLEIN on clean substrates (RģŽIýKA 2001). This
1966; MARKS 1977; HEMPTINNE & DIXON 1991; indicates that fresh conspecific tracks only deter
FERRAN & DIXON 1993). C. limbifer females from ovipositing, but
Tracks of conspecific larvae deter females of heterospecific tracks deter them from ovipositing
many aphidophagous coccinellids from and searching.
ovipositing (RģŽIýKA 1997b; DOUMBIA et al. In contrast, females of the coccinellid
1998; YASUDA et al. 2000; RģŽIýKA 2001). Harmonia axyridis Pallas laid fewer eggs and
Interspecific oviposition-deterring effects of were observed less frequently on plants with
larval tracks in coccinellids (RģŽIýKA 2001) are tracks of conspecific larvae than with tracks of
less frequently recorded than in chrysopids Coccinella septempunctata larvae (YASUDA et al.
(RģŽIýKA 1998). Intra- and interspecific 2000). The authors suggested that the less time
oviposition-deterring effects of the tracks of first spent on plants contaminated with conspecific
instar larvae were especially strong between the than on those with heterospecific tracks may
85
account for why fewer eggs were laid on plants females of the aphidophagous ladybird C.
with conspecific tracks. limbifer was studied in experiments similar in
Advanced technologies offer reliable ways of design to those used previously (RģŽIýKA 1997b,
monitoring searching behaviour in insects. The 2001). The behaviour of a female was monitored
new modes of digital data collection provide on a clean substrate and an adjacent substrate
highly accurate information on insect mobility, with larval tracks within a circular arena (Fig. 1).
which is less dependent on the skill and or time The arena was the bottom of glass Petri dish, 18.5
available to researchers. In this study, cm in diameter. The rim of the dish was painted
computerised video monitoring of coccinellid with Fluon (polytetrafluorethylene), which
females enabled a more detailed analysis of the prevented the coccinellid from leaving the arena.
effects of oviposition-deterring larval tracks. The test substrates consisted of strips (40 x 200
In order to find out whether repellent effects mm) of transparent plastic sheet transversally
are associated with the oviposition-deterring folded every 10 mm, which resulted in a total
effects of larval tracks, the searching behaviour of length of 130 mm. The transparency of the
single C. limbifer females was compared on material enabled females to be monitored on both
simultaneously provided clean substrates and the upper and lower surfaces of each strip. The
substrates with either conspecific or C. strips were placed parallel and 2.5 cm apart, each
undecimnotata larval tracks. The mobility of inside a 53x145mm rectangle substrate zone,
single females was recorded by EthoVision, a within the monitored circular arena. They were
computer-aided automatic video tracking system, fixed to the bottom of the Petri dish with narrow
which enables continuous monitoring of small strips of clear adhesive tape. The substrate zones
objects within selected areas of an arena. covered 57% of the area of the circular arena.
In order to ensure recording of females
MATERIALS AND METHODS present also on the margin of substrates, the
substrate zones around the folded strips were
Insects slightly larger then the substrates. The positions
of a female on the test substrates and outside
Experiments were done using Cycloneda limbifer substrate zones were recorded for 30 minutes
Casey (origin Cuba 1996) and Ceratomegilla with a computerised video tracking system. The
undecimnotata (Schneider) [=Semiadalia monitoring started three minutes after the female
undecimnotata (Schneider)] (origin North was carefully placed in the centre of the arena. In
Bohemia, Czech Republic). Females used in order to avoid effects of bias the substrates with
experiments were 10-25 days old. The laboratory larval tracks were placed alternately in the left
culture of C. limbifer was reared on Aphis fabae and right rectangles. In addition, blank tests with
Scopoli, and that of C. undecimnotata on two clean substrates were carried out to reveal
Acyrthosiphon pisum Harris. Cultures of both whether females had a preference for one of the
aphids were maintained on horse bean, Vicia faba two rectangles, which would indicate a bias due
L. to, e.g. temperature, light intensity or
geomagnetism. The behaviour of each female was
Experimental design monitored first in a blank test and then, after 5-10
minutes, in a choice test. This was replicated 14
The effect of larval tracks on the behaviour of times. No female laid eggs during the tests.
86
Fig. 1. Diagram of the arena (A) with two substrates and the monitoring apparatus (B). Fluon
(polytetrafluorethylene) painted on inner wall of Petri dish.
Contamination with larval tracks the lamps. The light intensity at floor level in the
arena was 800 Lux. Temperature in the room
A pair of plastic strips was exposed to 40 unfed during the experiments was kept at 26ґ 1 C. A
first instar conspecific or heterospecific larvae for Petri dish with a beetle in it was placed on a white
4 hours in a Petri dish, 18.5 cm in diameter. The sheet of paper. White was used to ensured a high
inner rim of the dish was painted with Fluon to contrast between the background and the beetle,
keep larvae in the dish. Strips were used for necessary for detection by the computer. The
experiments within 6 hours of exposure to larvae. experimental arena was scanned by a colour CCD
camera equipped with a zoom lens and fixed to
the ceiling above the centre of the arena. The
Video tracking composite video signal from the camera was fed
into a computerised video tracking system placed
The experiments were carried out in a dark room outside the room. The recording system consisted
(2.6H2.7H2.7 m) illuminated from above by a of a video monitor (Sony), a computer (486DX2,
series of eight 150 cm long fluorescent tubes 66 Mhz) with a frame grabber (Targa Plus,
(Narva LS 65 W-1 coolwhite). Diffused lighting TrueVision) and EthoVision software (Noldus
was achieved by placing a thin white cloth under Information Technology, 1997). The location of a
87
beetle in the arena was determined automatically walked on the two substrates nearest to the
by the software using a grey scaling method of average value, in each blank test is illustrated in
object detection. The co-ordinates of the centre of Fig. 2. The distances walked during the three
the animal's body were calculated using a spatial subsequent periods of 10 minutes were similar
resolution of 254H238 pixels. Tracking was done (P=1), [P=0.5416]; (P=0.1040), [P=0.6698] and
10 times per second which was the highest (P=0.7148), [P=0.5416] (Fig. 3 and 4). The
possible sample rate (BELL 1991) taking into average speed of females on clean substrates in
consideration the speed of the processor and the left and right positions when associated with
storage capacity of the computer. Nevertheless, conspecific larval tracks was 6.7 mm/sec and 6.0
this gave an accurate representation of the track. mm/sec. The average speed of females on clean
substrates in the left and right positions when
associated with tracks of heterospecific larvae
Data analysis
was 5.6 mm/sec and 5.1 mm/sec. In both blank
tests, the speed of walking on clean substrates, in
The digitised paths of individual beetles were the right and in the left position, did not differ
used to calculate distances walked and time spent either in the whole test (P=0.3258), [P=0.5416],
on each of the two test substrates and on the or in three consequent periods (P=0.3590),
surrounding glass. For details of the algorithms [P=0.8311]; (P=0.9700), [P=0.5771], and
used in the data analysis see Noldus Information (P=0.1940), [P=0.1040] (Fig. 5).
Technology (1997). Female behaviour on both During the blank tests, females of C. limbifer
substrates was compared in three subsequent spent (21 %) and [18 %] of the total time on the
periods: 0-10, 10-20 and 20-30 minutes. glass bottom of the Petri dish outside substrate
Differences in the time spent on the two zones. The average distances (SE) walked on
substrates were tested using non parametric the glass were (52792 cm) and [39288 cm],
Wilcoxon signed paired sample test (SIEGEL & i.e. (40 %) and [34 %] of the total distance
CASTELLAN 1988). The same test was used to walked in the monitored area. The average speed
compare distance walked and speed of movement of females on the glass away from the substrates
on the substrates. was (13.8 mm/sec) and [12.6 mm/sec].
RESULTS Choice experiments with fresh tracks of

conspecific larvae
Blank tests
The total residential time and the distance walked
The effect of bias on individual females of C. on substrates with fresh tracks of unfed
limbifer walking on two clean substrates was not conspecific first instars were significantly longer
significant. The results for females on the two than on clean substrates (P=0.0134 and
clean substrates, in the first choice test with P=0.0203). The average speed of females was
conspecific (values in parenthesis) and the 10.9 mm/sec on the clean substrate and 8.9
second, with heterospecific larval tracks (values mm/sec on the substrate with tracks. The average
in brackets), revealed no significant bias in the speed on both substrates in the test did not differ
time spent on substrates in the left and right significantly (P=0.1531).
positions (P=0.5830), [P=0.8077]. The values for Females of C. limbifer spent significantly
the periods 0-10, 10-20 and 20-30 minutes were longer on substrates with tracks than on clean
(P=0.3258), [P=0.3910]; (P=0.9515), [P=0.3575] substrates only during the first and the second
and (P=1), [P=0.1937]. Total distances walked by period of 10 minutes (P=0.0353 and P=0.0203).
emales in the blank tests on the right and left In the last period, the difference in favour of the
substrates were similar (P=0.7148), [P=0.7148]. substrate with tracks was not significant
Records of the tracks of the most representative (P=0.5830) (Fig. 3). Also distances walked on
female, i.e. the one with the ratio of the distances substrates with larval tracks were significantly
88
longer than on clean substrates in the first and in the middle period, i.e. between 10-20 minutes
second period (P=0.0245 and P=0.0419), but not (P=0.0250), but not in the first and the last
significantly so in the last period (P=1). Walking periods (P=0.6770), and (P=0.6360), (Fig. 5). A
speed of females on substrates with tracks was record of the tracks of the most representative
significantly lower than on clean substrates only female is illustrated in Fig. 2.
Fig. 2. Tracks of average Cycloneda limbifer females in blank tests (A and B) and in choice tests with conspecific
(C) and Ceratomegilla undecimnotata (D) larval tracks.
Blank test Blank test

Site Site
D is ta n c e w a lk e d ( c m )
left right left right

T im e o f s t a y ( s e c )
400 400
ns ns ns ns ns ns
300 300
200 200
100 100
0 0
0-10 10-20 20-30 0-10 10-20 20-30
Period of test (minutes) Period of test (minutes)
Choice test Choice test
Site Site
D is t a n c e w a lk e d ( c m )
clean tracks clean tracks

400 400
* * ns * * ns
300 300
200 200
100 100
0 0
0-10 10-20 20-30 0-10 10-20 20-30
Fig. 3. Effects of conspecific larval tracks on the behaviour of females of Cycloneda limbifer. Results (mean
SE) both for the behaviour on two clean substrates in blank tests and on one clean and one contaminated substrate
in subsequent choice tests. Wilcoxon paired sample test (two-tailed P value), * = P<0.05, ns = not significantly
different (P0.05).
89
During the tests, females of C. limbifer spent [P=0.0107].

24 percent of the time on the glass outside C. limbifer spent significantly less time on
substrate zones. The average distance walked by a substrates with fresh tracks of unfed first instars
female on the glass was 863158 cm (43% of the of C. undecimnotata than on clean substrates in
total distance walked in monitored area). The the second period (10-20 minutes) of the test
average speed of females on glass outside [P=0.0040]. Differences in the first and the last
substrate zones was 19.8 mm/sec. period of 10 minutes were not significant
[P=0.1189 and P=0.2166] (Fig. 4). The distances
walked on clean substrates were significantly
Choice experiments with tracks of C. longer than on substrates with tracks of
undecimnotata larvae heterospecific larvae during the second period
(10-20 minutes) of the test [P=0.0295], but not in
The total residential time and the distance walked the first and the last period [P=0.0906 and
on substrates with fresh tracks of unfed P=0.1726] (Fig. 4). Walking speed of females on
heterospecific first instars were significantly substrates with tracks was also significantly
shorter than on clean substrates [P=0.0023 and higher than on clean substrates only in the middle
P=0.0134]. The average speed of females during period of the test [P=0.0005], but not in the first
the choice test was 7.5 mm/sec on the clean (0-10 minutes) and the last (20-30 minutes)
substrate and 9.2 mm/sec on the substrate with periods [P=0.4631 and P=0.583] (Fig. 5). A
tracks. The speed on substrates with tracks was record of the track of the most representative
significantly higher than on clean substrates female is illustrated in Fig. 2.

Site Site
left right left right

T im e o f s ta y ( s e c )
400 400
ns ns ns ns ns ns
300 300
200 200
100 100
0 0
0-10 10-20 20-30 0-10 10-20 20-30
Choice test Choice test
Site Site
clean tracks clean tracks

400 400
ns * ns ns ** ns
300 300
200 200
100 100
0 0
0-10 10-20 20-30 0-10 10-20 20-30
Fig. 4. Effects of larval tracks of Ceratomegilla undecimnotata on the mobility of Cycloneda limbifer. Results are
for the behaviour (mean SE) on two clean substrates in blank tests and on one clean and one contaminated
substrate in subsequent choice tests. Wilcoxon paired sample test (two-tailed P value), ** = P<0.01, * = P<0.05, ns
= not significantly different (P0.05).
90
Site Site
15 left right left right
15
12 12
m m /s e c
m m /s e c
9 9
6 6
ns ns ns ns ns ns
3 3
0-10 10-20 20-30 0-10 10-20 20-30
Choice test with conspecific tracks Choice test with heterospecific tracks
Site Site
15 clean tracks 15 clean tracks
12 12
m m /s e c
m m /s e c
9 9
6 6
ns * ns ns *** ns
3 3
0-10 10-20 20-30 0-10 10-20 20-30
Fig. 5. The speed of movement (mean SE) of Cycloneda limbifer females on two clean substrates in blank tests
and on a clean substrate and substrate with tracks of conspecific or Ceratomegilla undecimnotata larvae in
subsequent choice tests. Wilcoxon paired sample test (two-tailed P value), *** = P<0.001, * = P<0.05, ns = not
significantly different (P0.05).
During the tests, females of C. limbifer spent The current results confirm a former
26 percent of the time on the glass outside assumption that the higher densities of faecal
substrate zones. The average distance (SE) spots left by females on clean substrates than on
walked by a female on the glass was 937150 those with larval tracks of C. undecimnotata
cm, (48 % of the total distance walked in indicate a repellent effect of contaminated
monitored area). The average speed of females on substrates (RģŽIýKA 2001). Residential time and
the glass was 19.1 mm/sec. total distance walked by C. limbifer females on
clean substrates were significantly longer than on
substrates with heterospecific tracks. This effect
DISCUSSION was strongest during the middle period (10-20
minutes) of the test. Also, the speed of females
Adult coccinellids usually do not stay on plants was higher on contaminated than on clean
very long. If not laying eggs, they walk or fly substrates at this time.
away after several minutes, exceptionally after The analysis revealed that the fresh tracks of
hours. Therefore, the presence of larval tracks is conspecific larvae affect the searching behaviour
most likely to change the searching behaviour of of C. limbifer females. In the first 20 minutes of
females soon after their arrival on a plant. the test, residential time and the distance walked
Automatic monitoring of females of C. limbifer were significantly higher on substrates with
provided considerably more information on the conspecific tracks than on clean substrates. Both
effects on their behaviour of substrates with larval parameters were also significantly lower on clean
tracks than did faecal spot densities left by substrates over the whole 30 minutes of this test.
females in choice tests designed to study these This effect of larval tracks was not evident in the
effects over 20 hours (RģŽIýKA 2001). previous study (RģŽIýKA 2001). In addition, the
91
speed of females was higher on contaminated This study showed that the fresh tracks of C.
than on clean substrates in the middle period. undecimnotata larvae and those of conspecific
This effect of fresh conspecific tracks was larvae affect the searching of C. limbifer females
surprising, because the oviposition-deterring differently. This was unexpected because the
effect of conspecific larval tracks is at least as oviposition-deterring effects of fresh conspecific
strong as the oviposition- deterring effect of and heterospecific larval tracks were almost
heterospecific tracks (RģŽIýKA 2001). This is the identical (RģŽIýKA 2001).
first report that larval tracks can increase the The difference in the effect of fresh tracks of
time of stay and decrease the speed of search of conspecific and heterospecific first instar larvae
conspecific females in an insect predator. Fresh on the searching behaviour of females is here
tracks of conspecific first instar larvae may reported in aphid predators for the first time. The
indicate, at least in some coccinellid species, that different behaviour of females on substrates with
prey might be present, even though the tracks conspecific tracks can have an adaptive
indicate it is an unsuitable site for oviposition. significance. Fresh larval tracks may stimulate
In the absence of aphids, the effect of females to search sites more thoroughly, because
conspecific tracks on C. limbifer declined after 20 the food is likely to be present. The prolongation
minutes. Results confirm that the strongest effects of the search on sites with tracks may also give a
of larval tracks on female mobility can be better assessment of the ratio of prey to
expected shortly after their arrival on a plant. conspecific competitors. This may enable females
BÄNSCH (1966) observed that adult coccinellids to more effectively asses site quality.
search model plants without aphids for 23
minutes.
Larvae of the pyralid Ephestia kuehniella ACKNOWLEDGEMENTS
Zeller secrete an oviposition-deterring pheromone
from salivary glands and contaminate their food This research was supported by the grant of the
with the secretion (CORBET 1971). While low Grant Agency of the Czech Republic, No.
numbers of larvae attract conspecific females to 206/00/0809, from the Entomology Institute
lay eggs, high numbers deter them (CORBET project Z5007907 (Acad. Sci. CR) and the grant
1973). Females of another species, Plodia project S5007102 (Grant Agency Acad. Sci. CR).
interpunctella (Hübner), lay significantly more We thank M. ýervenská and R. Guttwirthová for
eggs in sites contaminated with 1 or 5 larvae than their assistance with experiments and cultures of
on clean sites, but lay significantly fewer eggs on insects.
sites contaminated with 10 larvae (PHILLIPS &
STRAND 1994). A similar response is not
REFERENCES
observed in aphidophagous predators. The
oviposition-deterring effect of larval tracks
steadily increases with increase in density of BÄNSCH, R. 1966. On prey seeking behaviour of
tracks (RģŽIýKA 1997a; DOUMBIA et al. 1998) as aphidophagous insects. Pp. 123-128 in: I. HODEK
well as size of larvae (RģŽIýKA 1997b). (Ed.). Ecology of Aphidophagous Insects. Prague:
Academia.
Chemical analysis of the larval tracks of the
BELL, W.J. 1991. Searching Behaviour: The
coccinellid Adalia bipunctata (L.) revealed a
Behavioural Ecology of Finding Resources.
wide spectrum of semiochemicals, mainly London: Chapman & Hall.
hydrocarbons (HEMPTINNE et al. 2001), however,
CORBET, S.A. 1971. Mandibular gland secretion of
effects of individual compounds or sets of larvae of the flour moth, Anagasta kuehniella,
compounds on conspecific females are unknown. contains an epideictic pheromone and elicits
The strong intraspecific oviposition-deterring oviposition movements in a hymenopteran parasite.
effect of conspecific larval tracks (RģŽIýKA Nature 232: 481-484.
2001) and their effects on the foraging of C. CORBET, S.A. 1973. Oviposition pheromone in larval
limbifer may either be stimulated by the same or mandibular glands of Ephestia kuehniella. Nature
different substances present in the tracks. 243: 537-538.
92
DOUMBIA, M., J.-L. HEMPTINNE & A.F.G. DIXON 1998. RģŽIýKA, Z. 1994. Oviposition-deterring pheromone in
Assessment of patch quality by ladybirds: role of Chrysopa oculata (Neuroptera: Chrysopidae).
larval tracks. Oecologia 113: 197-202. European Journal of Entomology 91: 361-370.
FERRAN, A. & A.F.G. Dixon 1993. Foraging behaviour RģŽIýKA, Z. 1997a. Persistence of the oviposition-
in ladybird larvae (Coleoptera: Coccinellidae). deterring pheromone in Chrysopa oculata (Neur.:
European Journal of Entomology 90: 383-402. Chrysopidae). Entomophaga 42: 109-114.
HEMPTINNE, J.-L. & A.F.G. DIXON 1991. Why RģŽIýKA, Z. 1997b. Recognition of oviposition-
ladybirds have generally been so ineffective in deterring allomones by aphidophagous predators
biological control? Pp. 149-157 in L. POLGÁR, R.J. (Neuroptera: Chrysopidae, Coleoptera:
CHAMBERS, A.F.G. DIXON & I. HODEK (Eds). Coccinellidae). European Journal of Entomology
Behaviour and Impact of Aphidophaga. The 94: 431-434.
Hague: SPB Academic Publishing. RģŽIýKA, Z. 1998. Further evidence of oviposition-
deterring allomone in chrysopids (Neuroptera:
HEMPTINNE, J.-L., G. LOGNAY, M. DOUMBIA & A.F.G.
Chrysopidae). European Journal of Entomology
DIXON 2001. Chemical nature and persistence of
95: 35-39.
the oviposition deterring pheromone in the tracks
RģŽIýKA, Z. 2001. Oviposition responses of
of the larvae of the two spot ladybird, Adalia
aphidophagous coccinellids to tracks of coccinellid
bipunctata (Coleoptera: Coccinellidae).
(Coccinellidae) and chrysopid (Chrysopidae)
Chemoecology 11: 43-47.
larvae. European Journal of Entomology 98: 183-
KUCHLEIN, J.H. 1966. Some aspects of the 188.
prey-predator relation. Pp. 237-242 in I. HODEK SIEGEL, S. & N.J. CASTELLAN JR. 1988. Nonparametric
(Ed). Ecology of Aphidophagous Insects. Prague: Statistics for the Behavioral Sciences. New York:
Academia. McGraw-Hill Book Company.
MARKS, R.J. 1977. Laboratory studies of plant SOLOMON, M.E. 1949. The natural control of animal
searching behaviour by Coccinella septempunctata populations. Journal of Animal Ecology 18: 1-35.
L. larvae. Bulletin of Entomological Research 67: SOLOMON, M.E. 1964. Analysis and Processes Involved
235-241. in the Natural Control of Insects. Pp. 1-54 in J.B.
Noldus Information Technology 1997. EthoVision: CRAGG (Ed). Advances in Ecological Research 2.
Video Tracking, Motion Analysis & Behavior London and New York Academic Press.
Recognition System. Reference Manual, Version YASUDA, H., T. Takagi & K. Kogi 2000. Effects of
1.90., Wageningen. conspecific and heterospecific larval tracks on the
PHILLIPS, T.W. & M.R. STRAND 1994. Larval secretion oviposition behaviour of the predatory ladybird
and food odors affect orientation in female Plodia Harmonia axyridis (Coleoptera, Coccinellidae).
interpunctella. Entomologia Experimentalis et European Journal of Entomology 97: 551-553.
Applicata 71: 185-192.
93
94
PREDATORS AND PARASITOIDS ON DIFFERENT CEREAL APHID SPECIES

UNDER CAGED AND NO CAGED CONDITIONS IN HUNGARY
Z. BASKY
BASKY, Z. 2003. Predators and parasitoids on different cereal aphid species under
caged and no caged conditions in Hungary. Pp. 95-101 in A.O. SOARES, M.A.
VENTURA, V. GARCIA & J.-L. HEMPTINNE (Eds) 2003. Proceedings of the 8th
International Symposium on Ecology of Aphidophaga: Biology, Ecology and
Behaviour of Aphidophagous Insects. Arquipélago. Life and Marine Sciences.
Supplement 5: x + 112 pp.
Wheat plants, randomly chosen, were infected with Diuraphis noxia, Sitobium avenae,
Rhopalosiphum padi and Metopolophium dirhodum, repectively, regardless of the presence
or absence of other aphids. Half of the artificially infected plants were caged whereas the
other half was left no caged. One month after artificial aphid infection plants were sampled
weekly, and insects were counted.
R. padi was the most abundant species, followed by S. avenae, M. dirhodum and D. noxia
on caged tillers. Significant relationships were found between C. septempunctata and R.
padi densities and between Aphidius ervi and S. avenae densities. Activity of parasitoids
was hampered by hyperparasitoids. Chalcididae hyperparasitoids were more than twice as
abundant as parasitoids.
D. noxia was ca. ten to twenty times more frequent on no caged tillers, compared to other
cereal aphid species. Populations of indigenous aphid species: R. padi, S. avenae and M.
dirhodum were under the economic threshold on the no caged tillers. The results suggest
that the indigenous aphid species were more influenced by local natural enemies than
Diuraphis noxia predators and parasitoids presented in this study were not able to keep D.
noxia populations under economic injury level.
Zsuzsa Basky (e-mail: h10433bas@ella.hu), Plant Protection Institute of the Hungarian

Academy of Sciences P.O.B. 102, HU- 1525 Budapest, Hungary.
INTRODUCTION cultivars (e.g., NIASSY et al. 1987; OGECHA et al.

1992; PUTERKA et al. 1992; MILLER et al. 1994;
Cereals are one of the main staple crops world- BASKY et al. 2001). Aphid resistant cultivars may
wide (FAO 2001) with the largest area in the negatively impact the third trophic level (PRICE
temperate zone. Several species of pests attack 1986). Therefore the enhancement of predator
cereals (AFONINA et al. 2001). Aphids and parasitoid activity in a growing system that
(Homoptera Aphididae are one of the most allows natural enemies to maintain aphid
important pests of cereals, causing both direct populations below the economic injury level
damage by feeding and indirect damage by remains a desirable goal (MARASAS et al. 1997).
transmitting plant viruses (QUIROZ 1992). To This study reports the result of a cage
sustain cereal production and profitability, experiment where susceptible wheat plants
attempts to decrease aphid damage by using infected naturally by aphids and natural enemies,
resistant cultivars and enhancing naturally were artificially infected with indigenous cereal
occurring biological control have been made aphids and with the recently occurred Diuraphis
(STECHMANN 1986). However, the introduction of noxia (Kurdjumov) (BASKY & EASTOP 1991).
resistant cultivars resulted in developing new The aim of this study was to monitor aphid,
aphid biotypes which are able to damage resistant predator and parasitoid density under caged and
95
no caged conditions between tillering and placed into Berlese funnel for 5 days to extract
ripening. insects. Afterwards, plants were transferred to
emergence canisters for one week to collect
emerging parasitoids. The numbers of Aphidius
MATERIAL AND METHODS
and Aphelinus - type mummies were counted
from both the extracted samples and dry plant
The trial was carried out in a suburb of Budapest material. The numbers of different aphid species
(18º53’E 47º35’N, 342 m a.s.l.) on a 2 ha wheat were counted. The numbers of each species of
field. adult parasitoid and hyperparasitoid from Berlese
funnels and the emergence canisters were
Artificial infection counted, species were identified.
Winter wheat variety “MV 17” was sown at a Data analysis

seed rate of 220 kg/ha on 13 November. When
the plants reached the growth stage GS 30
Analysis of covariance using the sampling date,
(TOTTMAN & BROAD 1987) beginning of stem
isolation level (cage vs. no cage) and artificial
elongation on 11 May (2001), randomly chosen
aphid infection as categorical variables and
groups of 10-12 tillers were artificially infected
number of predators as continuous predictor were
with aphids.
used to identify the effect of these variables on
Plants were infected regardless of the
the number of different aphid species.
presence or absence of other aphids and natural
The effect of parasitoids was calculated as a
enemies.
regression between the number of the most
Five aptera or L4 nymph individuals of the
frequent parasitoid (Aphidius ervi Haliday) and
indigenous cereal aphid species: Sitobion avenae
the numbers of Diuraphis noxia, S. avenae, R.
(Fabr.), Rhopalosiphum padi (L.), or
padi and M. dirhodum on caged and no caged
Metopolophium dirhodum (Walker) and the
tillers.
newly occurred Diuraphis noxia (Kurdjumov),
The similarity between cage – no cage were
respectively were used to infect wheat plants.
calculated by the Renkonen - index (RENKONEN
Half of the artificially infected plants were caged
1938)
the other half was left no caged. Twenty-five
Statistical significance of goodness of fit of
pairs of caged and no caged group of tillers were
numbers of different species collected from caged
infected with different aphid species. Each aphid
and no caged tillers was tested by Yates corrected
species was released on other twenty-five pairs of
Chi-square test.
caged and no caged group of tillers. The distance
Analyses were made using the Statistica
between caged and no caged pairs was 1.5 m and
program package (STATISTICA 1997)
3 m between pairs. Five pairs of caged and no
caged groups of tillers were infected in each of
five replicate blocks at each aphid species RESULTS
yielding a total of 200 plants.
Aphids
Sampling
Rhopalosiphum padi reached the highest
Sampling started one month after the artificial numbers among the aphid species, followed by S.
aphid infection of the plants. Five pairs of caged avenae, M. dirhodum and D. noxia on the caged
and no caged tillers infected with different aphid tillers (Fig.1). R. padi was the dominant naturally
species, selected at random, were sampled occurring aphid species; it was present in 92 % of
destructively at weekly intervals for five weeks the cages, although only 25 % of the cages were
starting on 11 June. The plants were individually artificially infected with this species. S. avenae
96
was present in 52 % of the cages followed by M. ten times higher compared to S. avenae and R.
dirhodum and D. noxia 41 and 37 %, padi and twenty times higher that that of M.
respectively. R. padi colonised winter wheat more dirhodum (Fig. 2). D. noxia was present on 39 %
frequently compared to other aphid species. of the no caged tillers. S. avenae infected 49 % of
However, on the no caged tillers D. noxia was no caged tillers, followed by R. padi and M.
the prevailing aphid species, its abundance was dirhodum 40 and 29 %, respectively.
D. noxia S. avenae
360
550
240
Mean number of aphids on caged tillers
350
120
150
0 -50
2 July
9 July
11 June
25 June
2 July
9 July
18 June
11 June
18 June
25 June
R. padi M. dirhodum
1400 350
1000 250
600 150
200 50
-200 -50
2 July
9 July
2 July
9 July
11 June
18 June
25 June
11 June
18 June
25 June
Sampling date
Fig. 1. Seasonal abundance of cereal aphids in the cages.
D. noxia S. avenae
14
120
10
80
Mean number of aphids on nocaged tillers
6
40 2
0 -2
11 June
18 June
25 June
11 June
18 June
25 June
2 July
9 July
2 July
9 July
R. padi M. dirhodum
6
8,5
4
6,5
4,5 2
2,5 0
11 June
18 June
25 June
11 June
18 June
25 June
2 July
9 July
2 July
9 July
Sampling date
Fig. 2. Seasonal abundance of cereal aphid species on no caged tillers.
97
Predators and parasitoids mummies, always found on Diuraphis noxia

hosts, were 5 times more frequent on caged than
The survey yielded 10 species of aphidophagous no caged tillers.
insects associated with cereal aphids. The Seventy eight percent of the emerged adult
ladybird C. septempunctata (Coccinellidae) was parasitoids were Aphidius ervi.
the most abundant predator, 67 % of the predators Apart from A. ervi, Aphidius uzbekistanus
belonged to C. septempunctata. Hoverflies Luzhetzki, Diaretiella rapae (M’Intosh),
(Syrphidae) and lacewings (Chrysopidae) Ephedrus plagiator (Nees), Praon volucre
comprised only 33 % of the predators. –96% of (Haliday) and Aphelinus spp. parasitoid species
the collected predators were in the cages (Fig 3). were present. The majority (94%) of the
Aphidius – type mummies infecting mainly S. hyperparasitoids belonged to the Chalcididae
avenae were 5.5 times more abundant on the family. The number of hyperparasitoids was more
caged than on no caged tillers. Aphelinus - type than twice as much at that of parasitoids (Fig 4).
caged plants no caged tillers

2,4
2,0 0,10
Mean number of Coccinellids
1,6
0,06
1,2
0,8
0,02
0,4
0,0 -0,02 2 July
9 July
2 July
9 July
11 June
18 June
25 June
11 June
18 June
25 June
Sampling date
Fig. 3. Seasonal abundance of Coccinellids on no caged tillers.
parasitoids/cage parasitoids/plant
Mean number of parasitoids
6 0.6
4 0.4
2 0.2
0 0.0
18 June
25 June
18 June
25 June
2 July
9 July
2 July
9 July
11 June
11 June
hyperparasitoids/cage hyperparasitoids/plant
Mean no of hyperparasitoids
22 1.6
14
0.8
6
-2 0.0
18 June
25 June
18 June
25 June
2 July
9 July
2 July
9 July
11 June
11 June
Sampling date
Fig. 4. Seasonal abundance of parasitoids and hyperparasitoids on caged and no caged tillers.
98
Relationships between aphids and natural R. padi, M. dirhodum and D. noxia or between
enemies the numbers of parasitoids and hyperparasitoids.
Analysis of covariance revealed significant Cage versus no cage

relationship between sampling date, isolation
level, artificial aphid infection, number of C.
The low Renconen index value (0.3527) indicated
septempunctata, and number of R. padi S. avenae,
that the community structure of aphids and
M. dirhodum and D. noxia (adjusted R2=0.17, =,
natural enemies was different on caged and no
0.13, 0.28 and 0.18, R2 respectively. F=2.07
caged tillers. Yates corrected Chi-sqaured test
1.75, 2.98 and 2.12, respectively df= 40, 158,
revealed significant differences between the
P<0.00) for the whole model. However, within
frequencies of different aphid species on caged
the whole model the only significant relationship
and no caged tillers. Yates corrected Chi2 value
occurred between number of C. septempunctata
was the highest for D. noxia (Chi2=9257,
and that of R. padi at P<0.10 (P=0.06, F=3.37).
P=0.000). The high Chi2 value reflects that D.
The significant relationship between number of
noxia was the least abundant species in the cages,
C. septempunctata and R. padi is in agreement
but it was the most abundant species on the no
with the observations that relatively few R. padi
caged tillers. R. padi was the most abundant in
remained in cages where C. septempunctata
the cages (Chi2=1256 P=0.000), it was less
larvae were present.
abundant than D. noxia and S. avenae on the no
However, the number of C. septempunctata
caged tillers. S. avenae (Chi2=881 P=0.000) was
did not affect significantly those of S. avenae, M.
the second most abundant aphid species in the
dirhodum and D. noxia. It is in agreement with
cages and it was the second most abundant on the
the observations that D. noxia survived in large
no caged tillers. M. dirhodum (Chi2=634
numbers in cages containing C. septempunctata.
P=0.000) was the third most abundant cereal
Sampling date and isolation level significantly
aphid in the cages and it was the least abundant
affected aphid numbers. Artificial aphid infection
on the no caged tillers
significantly affected number of D. noxia and R.
There was a significant difference between
padi, but not of S. avenae and M. dirhodum.
abundance C. septempunctata and that of
In the cages, the aphid/parasitoid ratio was
hyperparasitoids on no caged and caged tillers
highest for R. padi (132 aphids/parasitoid,
(Chi2 was 5.25 and 7.49, respectively P=0.000).
followed by S. avenae (86 aphids/parasitoid, M.
However, the abundance of hover flies, lacewings
dirhodum (45 aphids/parasitoid), and D. noxia (39
and parasitoids did not differ significantly
aphids/parasitoid).
between caged and no caged tillers.
While on the no caged tillers the
aphid/parasitoid ratio was the highest for D. noxia
(150 aphids/parasitoid, and was much lower for DISCUSSION
R. padi, S. avenae and M. dirhodum (19, 15 and 7
aphids/parasitoid, respectively. The parasitoid R. padi, S. avenae, M. dirhodum and D. noxia
/hyperparasitoid ratio was higher on no caged populations reached very high numbers on
tillers than in the cages (3.0 vs. 2.6 , parasitoid artificially- infected caged tillers. The high aphid
/hyperparasitoid respectively). abundance in the cages was probably due to
Simple linear regression analyses revealed successful artificial aphid infection. On the other
significant relationship between Aphidius ervi and hand, cages blocked aphid emigration. Cages
Sitobion avenae (Adjusted R2=0.83 F=37.42 prevented alate aphids to leave the host plants.
P<0.00) on no caged tillers. The relationship was Therefore several generations developed on these
not significant between Aphidius ervi and the plants resulting in very high aphid populations.
other aphid species regardless of the isolation Cages blocked natural enemy emmigration as
level. There was no significant regression either well, resulting in much higher natural enemy
between the numbers of predators and numbers of abundance in cages than on no caged tillers.
99
Although the majority of predators and hyperparasitisation often occurs in the Sitobion
parasitoids were in the cages, the high number of avenae - Aphidius ervi host - parasitoid
aphids in cages indicated that aphid reproduction relationship (ABO KAF 1991).
rate was higher than the consumption rate of The community structure of aphids and
predators or the reproduction rate of parasitoids. natural enemies is significantly different on caged
Abundance of indigenous aphid species was and no caged tillers (Renkonen index). The
under economic injury level in spite of artificial relative abundance of aphids was significantly
aphid infection on no caged tillers, indicating that higher on caged than on no caged tillers. The high
naturally occurring natural enemies were able to aphid abundance in the cages was partly the result
control indigenous aphid species in spite of the of the blocked aphid emigration. Alata aphids left
artificial aphid introduction. However, D. noxia no caged wheat when they developed. While in
surpassed the economic injury level. It was 10-20 the cages they continued feeding and producing
times more abundant than indigenous aphid progenies as long as the wheat was suitable for
species. D. noxia feeds inside the rolled leaves aphid feeding. Therefore the number of aphids on
(AALBERSBERG 1988) predators cannot easily caged tillers was much higher than on no caged
reach D. noxia individuals. Colonies of other ones.
cereal aphids are exposed to predators on the The relative abundance of hover flies,
extended leaf surfaces (REED et al. 1991). The lacewings and parasitoids did not differ
lower proportion of indigenous cereal aphids on significantly on caged and no caged tillers.
no caged tillers indicated that C. septempunctata, Based on these results, it can be concluded
syrphid and chrysopid predators limited the that the natural enemy complex played a
numbers of R. padi, S. avenae and M. dirhodum substantial role in suppressing populations of
more efficiently than those of D. noxia. The indigenous aphid species under no caged
significant relationship between number of C. conditions. The number of naturally occurring
septempunctata and number of R. padi supported and artificially introduced indigenous aphid
the assumption that C. septempunctata efficiently species was below the economic threshold during
suppressed R. padi. the survey under no caged conditions. However,
The significant relationship between A. ervi D. noxia originated from artificial infection built
and S. avenae indicated that this aphid was up high populations on no caged tillers. It
efficiently suppressed by A. ervi on no caged indicates that parasitoids and predators present in
tillers. A. ervi was the most frequent parasitoid this study were not able to maintain the D. noxia
present in this study. A. ervi more often infection below the economic injury level.
parasitized S. avenae than R. padi, M. dirhodum
and D. noxia. This is due to host specialisation of
ACKNOWLEDGEMENT
this parasitoid. Aphidius ervi is common on
Sitobion avenae (STARY 1973; ABO KAF 1991).
However, hardly any parasitoids attacked R. padi I thank to Dr. Petr Starý for identifying the
and M. dirhodum. Aphelinus spp. always parasitoids and hyperparasitoids, Dr. Balázs Kiss,
attacking D. noxia was the second most abundant Ferenc Kádár for suggestions in statistical
parasitoid. However, it was not abundant enough analysis and Dr. Gábor Lövei for helpful
to efficiently decrease the number of D. noxia comments on the earlier version of this
below the economic injury level. manuscript.
Higher parasitoid/hyperparasitoid rate
occurred on no caged tillers than on caged ones. REFERENCES
Although hyperparasitoids were closed in the
cages together with aphids and parasitoids; ABO KAF, N. 1991. Parasitic Hymenoptera associated
therefore they could easily reproduce on with cereal aphids in fields of wheat and barley in
parasitized mummies. The lower hyperparasitoid the region of Lublin, Poland. Pp. 17-21 in L.
rate in the cages indicated that hyperparasitoids POLGÁR, R.J. CHAMBERS, A.F.G. DIXON & I. HODEK
continued to arrive as the season progressed. High (Eds). Behaviour and impact of Aphidophaga. SPB
100
Academic Publishing bv, The Hague, The PUTERKA, G.J., J.D. BURD & R.L. BURTON 1992.
Netherlands. Biotipic variation in a worldwide collection of
AFONINA, V.M., V.B. TSHERNYSHEV, I.I. SOBOLEVA- Russian wheat aphid (Homoptera, Aphididae).
DOKUCHAEVA, A.V. TIMOKHOV, O.V. TIMOKHOVA Journal of Economic Entomology 85: 1497-1506.
& R.R. SEIFULINA 2001. Arthropod complex of PRICE, P.W. 1986. Ecological aspects of host plant
winter wheat crops and its seasonal dynamics. resistance and biological control: Interactions
Integrated Control in Cereal Crops. IOBC wprs among three trophic levels. Pp. 11-30 in D.J.
Bulletin 24: 153-163. BOETHEL & R.D. EIKENBARY (Eds). Interactions of
AALBERSBERG, Y.K. 1988. Natural enemies and their Plant Resistance and Parasitoids and Predators of
impact on Diuraphis noxia (Mordvilko) Insects. Ellis Horwood, Ltd., Chichester.
(Hemiptera: Aphididae) populations. Bulletin of REED, D.K., J.A. WEBSTER, B.G. JONES & J.D. BURD
Entomological Research 78: 111-120. 1991. Tritrophic Relationships of Russiann Wheat
BASKY, Z. & V.F. EASTOP 1991. Diuraphis noxia in Aphid (Homoptera: Aphididae), a Hymenopterous
Hungary. Newsletter barley Yellow Dwarf 4: P. 34. Parasitoid (Diaretiella rapae McIntosh), and
BASKY, Z., R.L. HOPPER, J. JORDAAN & T. SAAYMAN Resistant and Susceptible Small Grains. Biological
2001. Biotypic differnces in Russian wheat aphid Control 1: 35-41.
(Diuraphis noxia) between South African and RENKONEN, O. 1938. Statistisch-ökologische
Hungarian agro-ecosystems. Agriculture, Untersuchungen über die terrestrische Käferwelt
Ecosystems and Environment 83: 121-128. der finnischen Bruchmoore. Annales Zoologici
FAO 2001. Production Yearbook (1999). (Vol. 53), Societatis Zoologicae - Botanicae Fennicae
FAO, Rome, Italy. Vanamo 6: 1-231.
MARASAS, C., P. ANANDAJAYASEKERAM, V.L. QUIROS, C., R.M. LISTER, R.H. SHUKLE, J.E. ARAYA &
TOLMAY, D. MARTELLA, J.L. PURCHASE & G.J. J.E. FOSTER 1992. Selection of symptom variants
PRINSLOO 1997. Socio-economic impact of the from the NY-MAV strain of barley yellow dwarf
Russian wheat aphid control research program. virus and their effects on the feeding behaviour of
Southern African Centre for Cooperation in the vector Sitobion avenae (Homoptera:
Agricuture and Natural Resources & Training; Aphididae). Environmental Entomology 21: 376-
P/Bag 00108, Gaborone, Botswana. 381.
MILLER, H., D.R. PORTER, J.D. BURD, C.W. STARÝ, P. 1973. A review of the Aphidius – species
MORNHINWEG & R.L. BURTON 1994. Physiological (Hymenoptera, Aphidiidae) of Europe.
effects of Russian wheat aphid (Homoptera: Annotationes Zoologicae et Botanicae 84: 85 pp.
Aphididae) on resistant and susceptible barley. STATISTCA 1997. Statsoft. Tulsa, Oklahoma, USA
Journal of Economic Entomology 87: 493-499. STECHMANN, D.H. 1986. Cereal aphids – Aphidophaga
NIASSY, A., J.D. RYAN & D.C. PETERS 1987. Variation associations in hedges and fields: Can habitat
in feeding behaviour, fecundity and damage of interaction contribute to integrated pest
biotypes B and E of Schizaphis graminum management? Pp. 273-278 in I. HODEK (Ed).
(Homoptera: Aphididae) on three wheat genotypes. Ecology of Aphidophaga. Academia Prague & Dr.
Environmental Entomology 16: 1163-1168. W. Junk, Dordrecht.
OGECHA, J., J.A. WEBSTER & D.C. PETERS 1992. TOTTMAN, D.R. & A. BROAD 1987. Decimal code for
Feeding behaviour and development of biotypes E, the growth stage of cereals. Annales of Applied
G, H of Schizaphis graminum (Homoptera: Biology 110: 683-687.
Aphididae) on ‘Wintermalt’ and ‘Post’ barley.
Journal of Economic Entomology 85: 1522-1526. Accepted 31 May 2003.
101
102
INTRAGUILD PREDATION AMONG THE APHIDOPHAGOUS LADYBIRD

BEETLES Harmonia axyridis PALLAS AND Coccinella undecimpunctata L.
(COLEOPTERA: COCCINELLIDAE): CHARACTERIZATION OF THE
DIRECTION AND SYMMETRY
S. FÉLIX & A.O. SOARES
FÉLIX, S. & A.O. SOARES 2003. Intraguild predation among the aphidophagous
ladybird beetles Harmonia axyridis Pallas and Coccinella undecimpunctata L.
(Coleoptera: Coccinellidae): characterization of the direction and symmetry. Pp.
103-106 in A.O. SOARES, M.A. VENTURA, V. GARCIA & J.-L. HEMPTINNE (Eds)
2003. Proceedings of the 8th International Symposium on Ecology of
Aphidophaga: Biology, Ecology and Behaviour of Aphidophagous Insects.
Direction and symmetry of intraguild predation (IGP) between all developmental stages of
the aphidophagous Harmonia axyridis Pallas and Coccinella undecimpunctata L. were
characterized.
Our results revealed that H. axyridis predation level was significantly higher than C.
undecimpunctata and eggs were the more vulnerable developmental stage. Significantly
asymmetric IGP on eggs occurred after the second and fourth larval stages of H. axyridis
and C. undecimpunctata, respectively. Asymmetric IGP on pupas exclusively occurred in
the presence of the fourth larval stage of H. axyridis.
Sandra Félix (e-mail: felixs@mail.telepac.pt) and A. O. Soares, Department of Biology,

Azores, Portugal.
INTRODUCTION HOLT 1992; LUCAS et al. 1998; YASUDA et al.

2001). IG prey populations may suffer local
Intraguild predation is the killing and extinctions due to IGP, which represents an
consumption of a species that uses similar important mortality factor (LUCAS et al. 1998).
resources being potential competitors. It is a The two most important factors that determine
common and often important interaction. In a the symmetry and the direction of IGP are the
diverse array of communities, IGP occurs among size and the degree of feeding specificity
members of the same guild (POLIS & HOLT exhibited by the protagonists. IGP occurs mainly
1992). The aggressor is the intraguild predator with generalist predators attacking prey of smaller
(IG predator), the victim the intraguild prey (IG size, including conspecifics (POLIS et al. 1989;
prey), and the common resource is the extraguild POLIS & HOLT 1992; LUCAS et al. 1998). The size
prey (LUCAS et al. 1998). IGP is considered to be of the prey attacked generally increases with the
an extreme form of competition or a type of size / age of the IG predator (SABELIS 1992), with
classical predation that may affect the smaller individuals being more vulnerable to a
distribution, abundance and evolution of the larger number of predators (WERNER & GILLIAM
animal species. IGP not only provides an 1984; LUCAS et al. 1998). In ladybird beetles,
additional food resource to the IG predators, but it large species usually take advantage of small
may reduce inter- or intraspecific competition and ladybird species in terms of food consumption,
predation risk for the extraguild prey when and this could result in asymmetrical interactions
mutual IGP occurs (POLIS et al. 1989; POLIS & between the two species (OBRYCHI et al. 1998). A
103
specialized predator should be less adapted to and adults) of H. axyridis and C.

attack a nonpreferred prey, a disadvantage when undecimpunctata.
confronted with a generalist predator (LUCAS et As there are no possible interaction
al. 1998). respectively between eggs and pupae of the two
Coccinella undecimpunctata L. is a native species, there remain 45 experimental
aphidophagous predator of the Azorean combinations. Larvae and adults used in the
ecosystems, which can be found mostly by the experiments were 24h old. Prior to the beginning
sea (salt lands) and also in plants and flowers of the tests, second to fourth instars larvae and
from where they feed on pollen (RAIMUNDO & adults were starved for 24h then, they were
ALVES 1986). Harmonia axyridis Pallas is a weighed on a 10-4 mg Mettler AM 50 analytical
paleartic species native from Asia (TAN 1946, balance. First instars larvae were not starved they
1949; KOMAI 1956; IABLOKOFF-KHNZORIAN were also weighed before the experiments. One-
1982). An euryphagous predator which prefers weighed individuals of one species of ladybirds
aphids (HUKUSIMA & KAMEI 1970; HUKUSIMA & was placed with a weighed individual of the other
OHWAKI 1972; IABLOKOFF-KHNZORIAN 1982; species in a 2L transparent plastic box that
OSAWA 1992), psillids (FYE 1981; DREA & contains also a potted broad bean without aphid.
GORDON 1990), coccids (MCLURE 1987; HODEK Twenty-four hours later, the box was checked to
& HONċK 1988), the immature stages and eggs of determine which species survived if any. There
lepidopteran (SCHANDERL et al. 1988; DREA & were 15 replicates for each combination.
GORDON 1990), and spider mites (DREA & The natural mortality of each developmental
GORDON 1990; CLOUTIER & CLOUTIER 1991; instar of the two species was used as a control. A
LUCAS et al. 1997). It was found to be the best single individual of each instar was kept for 24h
biological agent tested in laboratory against many in a 2L transparent plastic box with a broad bean
phytophagous species. The possibility to rear this without aphid. There were 3 replicates for each
ladybird beetle successfully on eggs of Ephestia instar.
kuehniella Zeller (SCHANDERL et al. 1988) favors All trials were performed at 20 ± 1 ºC, 75 ±
its mass production to sustain inundative releases. 5% of RH and a photoperiod of 16L:8D, under
However, studies on the possible negative fluorescent lamps (Philips ref.: TDL 23W/54 and
impacts of its presence on native ladybird beetles TDL 18W/54).
and on other non-target arthropods are rarely A symmetry index was adapted from LUCAS
performed (LUCAS et al. 2002). et al. (1998) and expresses the number of
The aims of this study were to characterize the replicates in which a given predator was eaten
(i) direction and (ii) the symmetry of intraguild over the total numbers of replicates in which there
predation (IGP) between the aphidophagous H. was IGP for a particular combination of predator.
axyridis and C. undecimpunctata.
Statistical analysis
MATERIAL & METHODS
The symmetry indices for each combination were
H. axyridis individuals were mass reared at 22 ± 1 compared to a theoretical index of 50%
°C, with 75 ± 5 % RH and a photoperiod of corresponding to a symmetric interaction, using a
16L:8D, using fluorescent lamps (Philips ref.: 2
TDL 23W/54 and TDL 18W/54). Coccinellids Chi-square test (Ȥ , P < 0.05) (SPSS PRODUCTION
were fed a mixture of Aphis fabae Scopoli and FACILITY 2001).
Myzus persicae (Sulzer), and eggs of Ephestia
kuehniella Zeller. C. undecimpunctata adults RESULTS
were collected in S. Maria Island and reared
under de same biotic and abiotic conditions. Among all significantly asymmetric
Predation level, direction and symmetry of combinations, H. axyridis was IG predator in 16
IGP were characterized and compared between all times and C. undecimpunctata only in 3
developmental stages (eggs, 4 larval stages, pupae combinations. IGP were not significantly
104
asymmetric in 19 combinations and symmetric in eggs. IGP between adults did not occurred
one of them. Absence of IGP was observed 6 significantly asymmetric IGP, on pupae,
times. Among all combinations with H. axyridis exclusively occurred in the presence of the fourth
adults except on pupas and adult, the IGP was larval stage of H. axyridis. Larvae of both species
significantly asymmetric. On the other hand, in attacked heterospecific. In general, predation rate
all combinations of C. undecimpunctata adults, of immature stages of H. axyridis is higher (Fig.
significantly asymmetric IGP only occurred on 1).
Fig. 1. Representation of the intraguild predation (IGP), between various developmental stages of H. axyridis and
C. undecimpunctata. Legend: Close arrow - significant asymmetry IGP; Dashed arrow - not significantly
asymmetric IGP; Dashed line- symmetric IGP; Lack of arrows and lines- absence of IGP; C. undecimpunctata: in
central position (F2, df=1, P < 0.05).
DISCUSSION immobility do not confers any disadvantaged in

this biotic relation because they were almost
We characterized, under controlled conditions, invulnerable to IGP, the biggest body size, can
the direction and symmetry of IGP between H. contribute to its lower vulnerability.
axyridis and C. undecimpunctata. Among all Concerning the combinations between the
combinations tested, 19 of them were found to be larval stages, the differences size of IG predator
significantly asymmetrical, with a 16 and IG prey could determine the direction and
combinations having H. axyridis as the IG symmetry.
predator and only in 3 of them C. In all combinations with adults versus either
undecimpunctata was the IG predator. Those larval stages or eggs, adults took advantage,
results are according with our previously mainly H. axyridis. However IGP between adults
prediction in which H. axyridis would be the did not occurs.
aggressor (IG predator) and C. undecimpunctata
the victim (IG prey) in most of the combinations
ACKNOWLEDGEMENTS
among their developmental stages.
Eggs were the most vulnerable developmental
stage. We think that ease capture, in consequence We thank Helena Figueiredo and Roberto
of immobility, is the main reason for their Resendes for technical assistance and Professor
vulnerability. It seems, however, that pupae’s Luís Silva for his help in statistical analysis.
105
REFERENCES Oecologia 117: 279-285.

OSAWA, N. 1992. A life table of the ladybird beetle
Harmonia axyridis Pallas (Coleoptera:
CLOUTIER, C. & C. CLOUTIER 1991. Les solutions
Coccinellidae) in relation to the aphid abundance.
biologiques de la lutte pour la répression des
Japanese Journal of Entomology 60: 575-579.
insectes et acariens ravageurs des cultures. Pp. 19-
88 in C. VINCENT & D. CODERRE (Eds). La Lutte MCLURE, M.S. 1987. Potencial of the asian predator,
Harmonia axyridis Pallas (Coleoptera:
Biologique. Gaent Morin Éditeur, Québec. 671 pp.
Coccinellidae), to control Matsucoccus resinosae
DREA, J.J. & R.D. GORDON 1990. Coccinellidae. Pp.
Bean & Godwin (Homoptera: Margarodidae) in the
19-40 in D. ROSEN (Ed). Armored Scale Insects-
United States. Environmental Entomology 16: 224-
Their Biology, Natural Enemies and Control Vol.
230.
4B. New York; Amsterdam: Elsevier. 688 pp.
POLIS, G.A. & R.D. HOLT 1992. Intraguild predation:
FYE, R.E. 1981. Rearing and release of coccinellids for
The dynamics of complex trophic interactions.
potential control of pear Psylla. Agricultural
Trends in Ecology and Evolution 7: 151-154.
Research Service (Western Region) 20: 1-9.
RAIMUNDO, A.A.C. & L.L.G ALVES 1986. Revisão dos
HODEK, I. & A. HONċK 1988. Sampling, rearing and
coccinelídeos de Portugal. Publicações da
handling of aphid predators. Pp. 311-321 in A.K.
Universidade de Évora, Évora. 103 pp.
MINSK & P. HARREWIJN (Eds). Aphids-their
SABELIS, M.W. 1992. Predatory arthropods. Pp. 225-
Biology, Natural Enemies and Control. vol. 2B.
264 in M.J. CRAWLEY (Ed). Natural Enemies.
Elsevier Science Publishers B. V., Amsterdam. 364
pp. Blackwell Scientific Publications, Oxford. 592 pp.
IABLOKOFF-KHNZORIAN, S.M. 1982. Les coccinelles; SCHANDERL, H., A. FERRAN & V. GARCIA 1988. L’
élevage de deux coccinelles Harmonia axyridis et
Coléoptères-Coccinellidae. Société Nouvelle des
Semiadalia undecomnotata à l’ aide d’ oeufs d’
Éditions Boubée. Paris. France. 568 pp.
Anagasta kuehniella tués aux rayons ultraviolets.
KOMAI, T. (1956). Genetics of ladybeetles. Advances in
Entomologia Experimentalis et Applicata 49: 417-
Genetics 8: 155-189.
421.
LUCAS, E., D. Coderre & C. VINCENT 1997. Voracity
TAN, C.C. 1946. Mosaic Dominance in the inheritance
and feeding preferences of two aphidophagous
of color patterns in the Ladybird Beetle, Harmonia
coccinellids on Aphis citricola and Tetranychus
axyridis. Genetics 31: 195-210.
urticae. Entomologia Experimentalis et Applicata
85: 151-159. TAN, C.C. 1949. Seasonal Variations of color patterns
in Harmonia axyridis. Proceedings of the 8th
LUCAS, E., D. CODERRE & J. BRODEUR 1998. Intraguild
International Congress Genenetic: 669-670.
predation among aphids predators: characterization
WERNER, E.E. & J.F. GILLIAM 1984. The ontogenetic
and influence of extraguild prey density. Ecology
79: 1084- 1092. niche and species interactions in size- structured
populations. Annual Review of Ecology and
LUCAS, E., I. GAGNE & D. CODERRE 2002. Impact of
Systematics 15: 393-425.
the Harmonia axyridis on adults of Coccinella
YASUDA, H., T. KIKUCHI, P. KINDLMANN & S. SATO
septempunctata and Coleomegilla maculata
(Coleoptera: Coccinellidae). European Journal of 2001. Relations between attack and escape rates,
Entomology 99: 457-463. cannibalism, and intraguild predation in larvae of
two predatory ladybirds. Journal of Insect Behavior
OBRYKI, J.J., K.L. GILES & A.M. ORMORD 1998.
14: 373-384.
Interactions between an introduced and indigenous
coccinellid species at different prey densities. Accepted 31 May 2003
106
APPROACH TO THE KNOWLEDGE OF COCCINELLIDADE (COLEOPTERA)

SPECIES DIVERSITY OF MADEIRA AND PORTO SANTO ISLANDS
A.O. SOARES, R.B. ELIAS & A. RAIMUNDO
SOARES, A.O., R.B. ELIAS & A. RAIMUNDO 2003. Approach to the knowledge of
Coccinellidade (Coleoptera) species diversity of Madeira and Porto Santo islands.
Pp. 107-112 in A.O. SOARES, M.A. VENTURA, V. GARCIA & J.-L. HEMPTINNE
(Eds) 2003. Proceedings of the 8th International Symposium on Ecology of
Aphidophaga: Biology, Ecology and Behaviour of Aphidophagous Insects.
We present a list of coccinellids collected in Madeira archipelago and notice the presence
of two new species; of Scymnini; Scymnus (Scymnus) abietis (Paykull) and Nephus hiekei
Fürsch. We analyse the distribution and abundance at the community level, using the
following measures: species richness index, abundance and relative abundance, similarity
index, diversity, and evenness. 13 species from 3 families were recorded. Hippodamia
variegata was the most abundant one with 29.7 % of the 141 specimens collected. Two
main groups of relative abundance were identified. The first group included the 6 most
common H. variegata, S. interruptus, C. arcuatus, C. septempunctata, S. levaillanti and R.
litura and the second one included S. abietis, L. lophante and R. chrysomeloides. Diversity
index values are relatively high. The 6 most common species comprised more than 90 % of
the individuals without a very different relative abundance.
Many factors other than extinction and immigration could explain the variation of the
number of species on islands. The highest number of species collected in Madeira island,
compared to Porto Santo, could be related with its higher elevation, soil and substrate types,
plant species richness, number of habitat, habitats diversity, structure and heterogeneity of
the former.
António O Soares (e-mail: onofre@notes.uac.pt), R.B. Elias, Department of Biology,

Portugal & A. Raimundo, Department of Biology, University of Évora, PT-7000 Évora,
Portugal.
INTRODUCTION factors not included in the model, other than

extinction and immigration, could explain the
According to the MACARTHUR and WILSON’S variation on the number of species on islands
(1967) equilibrium theory of island biogeography, such as elevation, number of soil types, substrate
the number of species on islands balances types, plant species richness, number of habitats,
regional process governing immigration against habitat diversity, structure and heterogeneity
local process governing extinction. The (BORGES 1992; BORGES & BROWN 1999).
immigration rate of new species to the island will Species diversity is an expression of
be dependent on the potential mainland colonists community structure. High species diversity
and the extinctions rate increases with the number indicates a highly complex community, for a
of species already present on the island. The greater variety of species allows for a larger array
extinction rate through competition between of species interactions (ZAR 1984). Information
species on islands increases more rapidly as on community characteristics, mainly diversity,
species diversity also increases. According to the has remained a central issue in ecology and
same model, small and more isolated islands are measures of diversity are frequently seen as
thought to support fewer species. However, many indicators of the stability and maturity of the
107
community and the well being of ecological trees. The natural vegetation of Porto Santo has
systems. Diversity can be divided into two also been totally destroyed. It was probably
components: the variety or species richness and thermophilous laurel forest (KARSHOLT 2000).
the relative abundance of species or heterogeneity Term community used in this work, has its
(MAGURRAN 1991). Diversity measures can be a broadest sense, i.e., a set of coccinellid
useful tool in environmental monitoring and individuals present in a given habitat at a
conservation management (MAGURRAN 1991), particular time (HODEK & HONċK 1996). Thus we
used to evaluate how far the impact of will consider each island as a whole despite the
immigration, extinctions and environmental differences in habitats and the mosaic of
factors, namely human activity, can affect its environments that each habitat can have. The
components. Thus all information concerning aims of this work were (i) to provide a
previous history of the community, such as the contribution to the knowledge of the biological
permanence of community in time is very control agents of the Coccinellidae family, in
important. Madeira Archipelago (ii) to characterise species
The main characteristics of the Madeiran diversity of the Coccinellidae community, and
archipelago are as follows. Geography and (iii) to test the species-area hypothesis, which
geology: located in the North Atlantic Ocean predicts that assemblages will be more species
between 32º20’-33º10’ N latitude and 16º20’- rich on larger islands.
17º20’ W longitude. Is 978 km distant from
Lisbon and 630 km from the west coast of
Morocco, to the Azores it roughly 800 km and to MATERIAL & METHODS
Canary Islands 380 Km. It is formed by two main
islands (Madeira - 728 km2 and Porto Santo - 57
km2) and three small islands located at 24 km Twenty-three sampling sites located in Madeira
southeast of Madeira (Chão, Deserta Grande e and Porto Santo was selected. In which one 30
Bugio). Pico Ruivo (1862 m) and Pico do Facho samples were taken [MADEIRA: Funchal (1),
(478 m) are Madeira’s and Porto Santo’s highest Camacha (1), Santo da Serra (2) Caniçal (1),
elevations. The archipelago has a volcanic origin. Câmara de Lobos (1), Quinta Grande (1), Anjos
It was formed 60-70 million years ago but Porto (1), Laranjeiras (1), Porto Moniz (1), Seixal (1),
Santo has been subaerial for only 12-13 million Chão da Ribeira (1), Santana (2), Faial (1), Curral
years whereas Madeira has been aerial for 2-3 das Freiras (1), Fajã da Rocha do Barco (1);
million years. Climate: Mediterranean type, rather PORTO SANTO: Pico do Castelo (1), Fonte da
homogenous, but varies depending on the Areia (1), Aeroporto (1), Portela (2) and Pedrogal
elevation and exposure. The annual precipitation (1)] (Fig. 1). Depending on the type of vegetation,
varies much depending on the altitude (Funchal, different methods were used such as direct
645 mm; Encumeada 2675 mm and Porto Santo observation, beating and sweeping methods,
338 mm). Monthly temperatures range from 6 to collecting the ladybeetles with a suction tube
20 ºC, depending on the region. Porto Santo is aspirator. We never took more than an hour in
semi-arid. Vegetation: Madeira presents an each plot. The collected specimens were
evergreen laurel forest, called laurisilva, which preserved in 70% alcohol and afterwards mounted
covers about 10 % of the island. The vegetation and classified. The specimens collected are
on most of the southern part of the island, as well deposited at the Ecology Section of the Biology
as along portions of the northern coast and into Department of the Azores University (ref. CC-
the valleys presents many introduced plants and UA-SC-MAD).
108
ª ni º
pi « »100 (2)
¬ n1 n2 n3... ni ¼
Similarity index (Cs): This is a simple measure

of the extent to which two habitats have species
(or individuals) in common. Applied as defined,
purely in terms of species number, this coefficient
gives equal weight to all species and hence tends
to place too much significance on the rare
species, whose capture will depend heavily on
chance (SOUTHWOOD & HENDERSON 2000). In
this case we use this index to evaluate to which
extent two species have common habitats. It was
calculated from the Czekanowski or Sørensen
equation:
2j
Fig. 1. Madeira and Porto Santo Islands with reference Cs (3)
ab
to the 23 sites; 1-Funchal, 2-Camacha, 3 and 4-Santo
da Serra, 5-Caniçal, 6-Câmara de Lobos, 7-Quinta
Grande, 8-Anjos, 9-Laranjeiras, 10-Porto Moniz, 11- where j is the number of habitats common to the
Seixal, 12 Chão da Ribeira, 13 and 16-Santana, 14- two species, and a and b are the total numbers of
Faial, 15-Curral das Freiras, 17-Fajã da Rocha do habitats, where the species are present,
Barco, 18-Pico do Castelo, 19-Fonte da Areia, 20- respectively.
Aeroporto, 21 and 23-Portela and 22-Pedrogal.
Diversity index (H’): Species diversity
To characterise the community structure of (sometimes called species heterogeneity), a
the Coccinellidae fauna, the following measures characteristic unique to the community level of
of diversity were used: biological organisation, is an expression of
Species richness index (D): Species richness community structure. Diversity can be measured
provides an extremely useful measure of recording the number of species and describing
diversity. A number of simple indices have been their relative abundance, (MAGURRAN 1991). We
derived using some combination of S (number of used the Shannon and Wiener diversity index. It
species recorded) and N (total number of assumes that individuals are randomly sampled
individuals summed over all species) such as the from an “indefinitely large” population (PIELOU
Margalef’s index (MAGURRAN 1991): 1975), and is calculated from the equation:
DMg
( S 1)
(1)
H' ¦ pi log pi (4)
ln N
where pi is the proportion of individuals found in
Abundance (ni) and relative abundance (pi): ith species estimated as
Relative proportions of different species in the
community can be measured. The abundance and ni
relative abundance express the total number of pi (5)
N
individuals collected and the distribution of
individuals among species, respectively. If n1, n2, Evenness (J): This diversity index takes into
n3, … ni, are the abundance of species 1, 2, 3 …, account both species richness and evenness of the
i, the relative abundance expressed in percentage, individuals’ distribution among the species.
becomes (MAGURRAN 1991): Evenness (also referred as homogeneity) may be
109
expressed by considering how close a set of For the first time it was recorded the presence of
observed species abundances is to those from an Scymnus (Scymnus) abietis (Paykull) and Nephus
aggregation of species having maximum possible hiekei Fürsch in Madeira archipelago. In Porto
diversity for a given N and S (ZAR 1984). It is Santo Island we collected Coccinella
calculated as follows: septempunctata L. and Hippodamia variegata
(Goeze) (Table 1). Among all the Coccinellidae
H' species already known from Madeira archipelago
J (6)
H ' max (33 species) (JANSSON 1940; LUNBLAD 1958;
BIELAWSKY 1963; MITTER 1984; FÜRSCH 1987;
H’máx is the maximum possible diversity for a RAIMUNDO & LVES 1986; ERBER &
collection of N individuals in a total of S species, HINTERSEHER 1988, 1990; ERBER 1990; ERBER &
when the N individuals are distributed as evenly AGUIAR 1996) we r ecollected 33% (11 species),
as possible among the S species. It is calculated from which some of them were little abundant.
as follows: The short period of time selected to carry out the
samples, could explain the reason of the low
H ' max log S (7) number of species recollected.
Many factors other than extinction and
J is constrained between 0 and 1 with 1 immigration could explain the variation on
representing a situation in which all species are species numbers on islands. For instance,
equally abundant. elevation, number of soil types, substrate types,
plant species richness, number of habitats, habitat
RESULTS & DISCUSSION diversity, structure and heterogeneity (BORGES
1992; BORGES & BROWN 1999). Thus, the higher
Species richness and species richness index (D): variability of environmental conditions in
We collected 141 specimens distributed by 13 Madeira island could explain the higher number
species of 9 genus, 4 tribes and 3 subfamilies. of species collected.
Table 1
List of coccinellid species collected on Madeira (Mad) and Porto Santo (Ps) Islands. Legend: + Present on island;
* New species to the archipelago
Subfamilies Tribes Genus Species Mad Ps Note
Scymninae Scymnini Stethorus S. wollastoni Kapur +
Clitostethus
C. arcuatus (Rossi) +
Scymnus S. interruptus (Goeze) +
S. levaillanti Muls. +
S. abietis (Paykull) + *
Nephus N. flavopictus Woll. +
N. hiekei Fürsch + *
Coccidulinae Coccidulini Lindorus L. lophante (Blaisds.) +
Rhizobius R. litura F. +
R. chrysomeloides (Herbst.) +
Coccinellinae Coccinellini Adalia A. decempunctata (L.) +
Coccinella C. septempunctata (L.) +
Hippodamiini Hippodamia H. variegata (Goeze) + +
The values of species richness based on the abundance are given. Almost 75% of all
number of species, were 2.627 and 2.519 to the individuals collected (ni =89) belong to the
archipelago and Madeira Island alone, following species: H. variegata (ni =42, pi
respectively (Table 2). =29.7%), Scymnus interruptus (Goeze) (ni =29, pi
Abundance (ni) and relative abundance (pi): =20.6%) and Clitostethus arcuatus (Rossi) (ni
In figure 2, values of abundance and relative =18, pi =12.8%). Those values rise up to 81%
110
(ni=101) and 87% (ni=112) when we include C. variegata/ S. levaillanti and S. interruptus/ S.
septempunctata and Scymnus levaillanti Mulsant, levaillanti. The similarity index values in the
respectively. The rare species collected, with a following combinations C. arcuatus - S.
uniform distribution of the individuals, were S. interruptus and C. arcuatus - S. levaillanti were,
abietis, Lindorus lophante (Blaisds.) and 0.46 and 0.40 respectively (Table 3). Higher
Rhizobius chrysomeloides (Herbst.), with only values of similarity suggest the possibility of
one female in each (0.7%). population’s niche apportionment and interaction
such as competition and intraguild predation.
Table 2
Data characterising the Coccinellid community from
Madeira; Sr- species richness; DMg- species richness H. variegata 29,7
42
index, H’- diversity index, H’máx.-the maximum S. interruptus 20,6

29
possible diversity and J-evenness. Legend: Mad- C. arcuatus 12,8

18
Madeira island;Ps-Porto Santo island C. septem punctata 9,9

14
Indices Sr DMg H’ H’máx J’ S. levaillanti 8,6

12
Values R. litura 5
Species
11
13 2.627 0.875 1.114 0.785
(Mad+Ps) N. hiekei 2,8
4
Values N. flavopictus 2,8
12 2.519 0.764 1.079 0.708 4
(Mad) A. decem punctata 1,4
2
S. wollastoni 1,4
2
Similarity index (Cs): H. variegata (s= 1, 4, 7, 8, R. chrysomeloides 0,7
1
Relative Abundance (%)
Abundance (N)
9, 10, 14, 15(2x), 17) and S. interruptus (Goeze) L. lophante 0,7
1
(s= 4, 8, 9, 11, 12, 13, 14, 15, 16 e 17) were found S. abietis 0,7
1
in 10 samples. S. levaillanti were found in 7 0 10 20 30 40 50

samples (s= 7, 9, 10, 13, 24, 16, 17), C. arcuatus
Abundance
and R. litura in 3 samples (s= 9, 15, 16 and s= 2,
3, 12, respectively). The highest values of Fig. 2. Absolute and relative number of specimen and
similarity were obtained in the following species among species (Relative Abundance-%) found in
combinations: H. variegata/ S. interruptus, H. Madeira and Porto Santo islands.
Table 3
Similarity index values (Czekanowski or Sørensen) between the most abundant species.
H. variegata S. interruptus S. levaillanti R. litura C. arcuatus
H. variegata ___ 0.60 0.59 0 0.31
S. interruptus ___ 0.59 0.15 0.46
S. levaillanti ___ 0 0.40
R. litura ___ 0
C. arcuatus ___
Diversity index (H): Relatively high values of of the archipelago (0.875) can be explained by the
diversity were obtained in archipelago (Madeira absence of C. septempunctata in Madeira island.
and P. Santo) and in Madeira island (Table 2). Evenness (J): The maximum evenness is
Despite a high relative abundance of H. variegata obtained when values of diversity index and
in the community, we found that half of the maximum possible diversity are the same (Table
species had more than 90% of the individuals. 2). Despite not having an identical distribution of
The relative abundance of the 6 commonest individuals among species, we found that the
species was high, with abundance ranging from 5 community presented two groups in which the
to 29.7%. Diversity index and maximum possible relative abundance wasn’t very different. The
diversity, were slightly different (Table 2). The first group includes the 6 common species and the
lowest value of the diversity index obtained in second one, the other species in which we
Madeira island (0.764) when compared to the one recorded a uniform distribution of the individuals.
111
In our opinion this fact contributes to the Funchal 42: 147-181.

relatively high values of evenness. ERBER, D. & W. HINTERSEHER 1988. Contribution to
The amount of time spent in sampling and the the knowledge of the Madeira beetles. Boletim
temporal and spatial variations in the abundance Museu Municipal do Funchal 40: 139-214.
ERBER, D. & W. HINTERSEHER 1990. Additional notes
of Coccinellidae, related with the life cycle of
to the Knowledge of the Madeira beetles. Boletim
each species are, in our opinion, the main Museu Municipal do Funchal 42: 141-146.
restriction to the utilization of these results in the ERBER, D. & A.M.F. AGUIAR 1996. New rewmarkable
characterisation of the distribution and abundance species of the coleopterous fauna of Madeira.
of the entire coccinellid community. Exhaustive Boletim Museu Municipal do Funchal 48: 41-62.
sampling programmes must be carried out in FÜRSCH, H. 1987. Die Scymninae der Kanaren, Azoren
order to evaluate seasonal and annual community und Madeiras. Acta Coleopterologica 3: 1-14.
structure. The approach that we are proposing to HODEK, I. & A. HONċK 1996. Ecology of
know species diversity in Madeira and Porto Coccinellidae. Dordrecht: Kluwer Academic
Santo islands, can be a useful tool in Publishers. 260 pp.
JANSSON, A. 1940. Die Arthropodenfauna von Madeira
environmental monitoring and conservation
nach den Ergebnissen von Prof. Dr. O. Lundblad
management, and can be used to evaluate how far Juli-August. XXIX. Coleoptera: Sämtliche
the impact of immigration, extinction and Familien unter Ausschluss der familiee der
environmental factors namely human activity, Carabidae, Dytiscidae, Hydrophilidae und der
may affect its components. Gattung Cryptophagus Herbst aus der Familie
Cryptophagidae. Arkiv för Zoologi 30: 1-64.
KARSHOLT, O. 2000. Contribution to the Lepidoptera
ACKNOWLEDGEMENTS fauna of the Madeiran Islands. Beiträge zur
Entomologie: 397-405.
We would like to thank Professor Helmut Fürsch, LUNDBLAD, O. 1958. Die Arthropodenfauna von
from Passau University, for identifying N. hiekei Madeira nach den Ergebnissen der Reise von Prof.
Fürsch for confirming S. abietis (Paykull) Dr. O. Lunbblad Juli-Augustus 1935. XXXV. Die
Käferfauna der insel Madeira. Arkiv för Zoologi 11:
classification. Thanks are also due to António
461-524.
Franquinho Aguiar from Laboratório Agrícola da MACARTHUR, R.H. & E.O. WILSON 1967. The theory
Madeira for sending the list of Coccinellid of of island biogeography. Princeton: Princeton
Madeira and the bibliographic references and to University Press. 203 pp.
Anunciação Ventura for her useful comments. MAGURRAN, A.E. 1991. Ecological diversity and its
measurement. Cambridge: Chapman and Hall. 179
pp.
REFERENCES MITTER, H. 1984. Beitrag zur Kenntnis der Käferfauna
der Insel Madeira. Bocagiana 80: 1-7.
BORGES, P.A.V. 1992. Biogeography of the Azorean PIELOU, E.C. 1975. Ecological Diversity. New York:
Coleoptera. Boletim Museu Municipal do Funchal Wiley & Sons Inc. 165 pp.
44: 5-76. RAIMUNDO, A.A.C. & M.L.L.G. ALVES 1986. Revisão
BORGES, P.A.V. & V.K. BROWN 1999. Effect of island dos Coccinelídeos de Portugal. Évora, Publicações
geological age on the arthropod species richness of da Universidade de Évora. 103 pp.
Azorean pastures. Biolological Journal of the SOUTHWOOD, T.R.E. & P.A HENDERSON 200.
Linnean Society 66: 373-410. Ecological methods (3rd ed.). London: Chapman
BIELAWSKY, R. 1963. Coccinellidae (Coleoptera) von and Hall. 575 pp.
Madeira. Communications in Biology 25: 72-102. ZAR, J.H. 1984. Biostatistical analysis (2nd ed.). New
EERBER, D. 1990. New and little known Coleoptera Jersey: Prentice-Hall. 718 pp.
from Madeira. Results of excursions to Madeira in
the years 1986-1990. Boletim Museu Municipal do Accepted 31 May 2003.
112
ISBN 972-8612-15-X
ISSN 0873-4704

Biology, Ecology and Behaviour of Aphidophagous Insects - PDF Room

Uploaded by

Copyright:

Available Formats

Biology, Ecology and Behaviour of Aphidophagous Insects - PDF Room

Uploaded by

Document Information

Original Description:

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

Biology, Ecology and Behaviour of Aphidophagous Insects - PDF Room

Uploaded by

Copyright:

Available Formats

Ecology of Aphidophaga:

Biology, Ecology and Behaviour of

Ponta Delgada 1-6 September 2002

António O. Soares, Maria A. Ventura,

Arquipélago - Life and Marine Sciences

GUEST EDITORS FOR SUPPLEMENT 5

INFORMATION ON THE INTERNET

SCIENTIFIC COMMITTEE FOR THIS SUPPLEMENT

For purpose of citation, this supplement should be cited as follows:

DIXON, A.F.G. & J.-L. HEMPTINNE

BUENO, V.H.P., A.B. CARNEVALE & M.V.SAMPAIO

SCHELT, J.- V. & F. WÄCKERS.

MAGRO, A., J-L. HEMPTINNE, A. NAVARRE & A.F.G.DIXON

HEMPTINNE, J.-L., A.F.G. DIXON & E. WYSS

SERPA, L., H. SCHANDERL, C. BRITO & A. O. SOARES

SOARES, A.O., D. CODERRE & H. SCHANDERL

JAROŠÍK, V., J. POLECHOVÁ, A.F.G. DIXON & A. HONċK

KINDLMANN, P. & A.F.G. DIXON.

FRÉCHETTE, B., C. ALAUZET & J.-L. HEMPTINNE

GODEAU, J.-F., J.-L. HEMPTINNE & J.-C. VERHAEGHE

RģŽIýKA, Z. & R. ZEMEK

FÉLIX, S. & A. O. SOARES

SOARES, A.O., R. B. ELIAS & A. RAIMUNDO

Basky, Zsuzsa Duelli, Peter

Bianchi, Félix Evans Edward W.

Hindayana, Dadan British-Colombia V5A 156, Canada. E-mail:

Sato, Satoru Yamagata JP-997-8555 Japan. E-mail:

1 - António Martins 14 - Peter Duelli 27 - Tessa Grasswitz

LADYBIRDS AND THE BIOLOGICAL CONTROL OF APHID POPULATIONS

A.F.G. DIXON & J.-L. HEMPTINNE

Anthony F.G. Dixon (e-mail: a.f.dixon@uea.ac.uk), School of Biological Sciences,

THEORY (PRICE et al. 1980; Fig 1B). Predators are

Fig. 7. The relationship between egg cannibalism and

ADDITIVE EFFECTS OF PEA APHID NATURAL ENEMIES DESPITE

Wolfgang W Weisser (e-mail: wolfgang.weisser@uni-jena.de), Institute of Ecology,

INTRODUCTION presence of generalist predators can reduce the

week, therefore, the control was better the more

HOST PREFERENCE OF Lysiphlebus testaceipes (CRESSON) (HYMENOPTERA:

V.H.P. BUENO, A.B. CARNEVALE & M.V. SAMPAIO

INTRODUCTION explain the final parasitization rates of the host

MATERIAL & METHODS species were placed in a Petri dish, and 15 L.

Accepted 31 May 2003.

THE BIOLOGICAL CONTROL OF Aulacorthum solani (KALTENBACH)

J.V. SCHELT & F. WÄCKERS

INTRODUCTION years. Because A. solani is a very polyphagous

No differences could be seen between the Discussion

Honeydew from Aulacorthum solani feeding on Controls

Discussion the other honeydew oligosaccharides studied,

Setting of new fruit per week per m²

OVERALL CONCLUSIONS KLAPWIJK, J.N. 1999. Biological control of tomato

Accepted 31 May 2003.

COMPARISON OF THE REPRODUCTIVE INVESTMENT IN

A. MAGRO, J-L. HEMPTINNE, A. NAVARRE & A.F.G. DIXON

Alexandra Magro (e-mail: alexandra.magro@educagri.fr), A. Navarre & J-L. Hemptinne,

INTRODUCTION egg weight multiplied by the ovariole number and

Measure of reproductive investment

aphidophagous species invest proportionally more REFERENCES

BIOLOGICAL CONTROL OF THE ROSY APPLE APHID, Dysaphis plantaginea

J.-L. HEMPTINNE, A.F.G. DIXON & E. WYSS