Acoustic Indices For Biodiversity Assessment and Landscape Investigation
Acoustic Indices For Biodiversity Assessment and Landscape Investigation
Acoustic Indices For Biodiversity Assessment and Landscape Investigation
Jérôme Sueur1) , Almo Farina2) , Amandine Gasc1,3) , Nadia Pieretti2) , Sandrine Pavoine3,4)
1)
Muséum national d’Histoire naturelle, Département Systématique et Évolution, UMR 7205-CNRS ISYEB,
45 rue Buffon, Paris, France. sueur@mnhn.fr
2)
Department of Basic Sciences and Foundations, Urbino University, Urbino, Italy
3)
Muséum national d’Histoire naturelle, Département Ecologie et Gestion de la Biodiversité, UMR 7204
CNRS-UPMC CESCO, 55-61 rue Buffon, 75005 Paris, France
4)
Mathematical Ecology Research Group, Department of Zoology, University of Oxford, South Parks Road,
Oxford, OX1 3PS, UK
Summary
Bioacoustics is historically a discipline that essentially focuses on individual behaviour in relation to population
and species evolutionary levels but rarely in connection with higher levels of ecological complexity like com-
munity, landscape or ecosystem. However, some recent bioacoustic researches have operated a change of scale
by developing acoustic indices which aim is to characterize animal acoustic communities and soundscapes. We
here review these indices for the first time. The indices can be divided into two classes: the α or within-group
indices and the β or between-group indices. Up to 21 α acoustic indices were proposed in less than six years.
These indices estimate the amplitude, evenness, richness, heterogeneity of an acoustic community or soundscape.
Seven β diversity indices were suggested to compare amplitude envelopes or, more often, frequency spectral pro-
files. Both α and β indices reported congruent and expected results but they may still suffer some bias due, for
instance, to anthropic background noise or variations in the distances between vocalising animals and the sensors.
Research is still needed to improve the reliability of these new mathematical tools for biodiversity assessment and
monitoring. We recommend the contemporary use of some of these indices to obtain complementary informa-
tion. Eventually, we foresee that this new field of research which tries to build bridges between animal behaviour
and ecology should meet an important success in the next years for the assessment and monitoring of marine,
freshwater and terrestrial biodiversity from individual-based level to landscape dimension.
PACS no. 43.80.+p
773
ACTA ACUSTICA UNITED WITH ACUSTICA Sueur et al.: Biodiversity assessment
Vol. 100 (2014)
efficient acoustic tools. In particular, research in ecology within- and between-group acoustic diversity. Several α
has a tradition of several indices that describe with a sin- acoustic indices have been developed to try to assess
gle value the ecological complexity from community to the richness or complexity of an acoustic community or
landscape scale (e.g. [25]). The requirement for acoustic soundscape and some β acoustic indices have been pro-
tools leads to the concomitant development of acoustic- posed to evaluate a level of acoustic disparity between
based ecological indices that could be used for biodiver- acoustic communities. We will review these two families
sity assessment, investigations on community dynamics of indices successively.
and landscape patterns. We aim with this paper to review
and comment some of these recent indices to better under- 3. Within-group indices: α acoustic diver-
stand the relationship between environmental proxies and sity
the acoustic complexity of vocal animals. All indices have
been developed in terrestrial communities and landscapes Acoustic indices estimating within-group diversity can be
so far. We will therefore focus our review on terrestrial divided into three categories: (1) indices using the ampli-
environments even if several indices can be used without tude, or intensity i.e. sound energy, (2) indices that esti-
major conceptual and technical issues to marine and fresh- mate a level of complexity in terms of time, frequency
water environments [26]. and/or amplitude, and (3) indices that take into account the
three components (biophony, geophony, and anthrophony)
2. Biodiversity indices of a soundscape (Table I).
774
Sueur et al.: Biodiversity assessment ACTA ACUSTICA UNITED WITH ACUSTICA
Vol. 100 (2014)
α indices
Relative avian abundance – Area under spectrum in relation with [43]
an amplitude threshold [43]
Temporal Entropy Ht Envelope complexity [44]
Spectral Entropy Hf Spectrum complexity [44]
Acoustic Entropy Index H Envelope and spectrum complexity [44]
Ratio of biophony to anthrophony ρ Ratio of biophony to anthrophony [54]
Acoustic Complexity Index ACI Spectrogram complexity [50]
Biophony – Biophony level [45]
Biophony peak bioPeak Biophony level [56]
Acoustic Entropy Index AEI (= H) Envelope and spectrum complexity [45]
Shannon’s Index H Spectrum complexity [46]
Acoustic Richness AR Envelope complexity and intensity [48]
Median of amplitude envelope M Median of amplitude envelope [48]
Normalised Difference Soundscape Index NDSI Ratio of anthrophony to biophony [55]
Acoustic Diversity Index ADI (= H ) Spectrum complexity [47]
Sound pressure level parameters L Ratio of sound pressure relative to [35]
a reference value [35]
Number of peaks NP Spectrum complexity [19]
Mid-band activity – Fration of spectrum above [42]
an amplitude threshold [42]
Entropy of spectral maxima Hm Spectrum composition [42]
Entropy of spectral variance Hv Spectrum complexity [42]
Spectral diversity – Number of clusters [42]
Spectral persistence – Duration of repeated clusters [42]
β indices
Spectral Dissimilarity Df Spectrum dissimilarity [44]
Temporal Dissimilarity Dt Envelope dissimilarity [44]
Acoustic Dissimilarity Index D Envelope and spectrum dissimilarity [44]
Kolmogorov-Smirnov distance KS Spectrum dissimilarity [60]
Kullback-Leibler distance KL Spectrum dissimilarity [60]
Vectorial correlation coefficient RV Spectrum similarity [60]
Cumulative Dissimilarity Dcf Spectrum similarity [61]
the relative abundance and composition of bird commu- trum scaled by its integral with species being replaced
nities [43]. This index computes the area under the fre- by frequency bins. Similarly, the temporal entropy, Ht ,
quency spectrum above a specific dB threshold and within was computed on the amplitude envelope obtained with
a specific frequency range. This metric, which is a function the Hilbert transform of the signal, scaled by its integral
of both sound level and the number of frequency bands as well. These two indices were multiplied to obtain an
used by the bird community, facilitated the monitoring acoustic entropy named H ranging between 0 and 1, with
of species across Hawaiian bird submontane ecosystems low values indicating pure tones and high values sound
[43]. with numerous and even frequency bands. This index re-
One year later, one of the most used indices in biodi- turned expected results on recordings made in a Tanzanian
versity assessment, the Shannon Shannon evenness index, forest where animal acoustic activity was high and back-
was applied on sound emitted by animal communities by ground noise due to wind, rain and human activity very
computing two acoustic sub-indices Hf , and Ht [44] . The low leading to a high signal-to-noise ratio [44]. A slight
Shannon index derives from the computation of entropy. modification of the spectral entropy index Hf was later
For a set of S species, this index is calculated with the introduced by reducing the frequency resolution of the av-
equation erage spectra to 1 kHz or by applying Shannon diversity
(=Shannon evenness/ln number of frequency bins) (index
H= − pi ln pi / ln S, (1) H and AEI [45]; index H [46]; index ADI [47]). It later
appears that entropy, in particular its spectral component
where pi is the proportion of individuals found in the ith
Hf , could give counter-intuitive results when applied to
species.
recordings where background noise dominates over ani-
The spectral entropy, Hf , was therefore obtained by ap-
mal sound production as it often occurs in temperate habi-
plying Shannon evenness to the average frequency spec-
775
ACTA ACUSTICA UNITED WITH ACUSTICA Sueur et al.: Biodiversity assessment
Vol. 100 (2014)
776
Sueur et al.: Biodiversity assessment ACTA ACUSTICA UNITED WITH ACUSTICA
Vol. 100 (2014)
777
ACTA ACUSTICA UNITED WITH ACUSTICA Sueur et al.: Biodiversity assessment
Vol. 100 (2014)
Four other distance metrics were used to compare the sity indices used so far were mainly based on the Shannon
average spectrum of bird songs: (1) the Kolmogorov- evenness index but other classical indices, like the expo-
Smirnov distance that is the maximum distance between nential of the entropy or the Hill index [64, 65] could in-
two cumulated frequency spectra, (2) the symmetric Kull- spire quickly new α acoustic indices. However, it is highly
back-Leibler distance that computes the relative entropy probable that a single index will never cover all biodiver-
between two probability frequency spectra, (3) the simi- sity facets and be reliable in all contexts. Combinations of
larity RV vectorial correlation coefficient that measures the indices could lead to more efficient results as already ex-
correlation between two matrices [60], (4) the cumulative plored by [42]. We therefore recommend the use of several
frequency dissimiliarity Dcf [61]. The index Dcf works complementary α indices.
as Df but takes cumulative frequency spectra as inputs.
This index has the advantage to be sensitive not only to Compared to the important number of α indices, few β
the spectral overlap between two spectra but also to the indices have been conceived so far (Table I). Bioacous-
mean distance between the different frequency peaks of tics shows a great interest in sound comparison to identify
the two spectra. All indices, including Df , were proved automatically species or individuals [66, 67, 68, 69, 70].
to be highly correlated generating similar results. Df and However, the comparative methods used in these contexts
Dcf could be preferred due to their simplicity in terms of are adapted to closely related sounds, as vocalisations pro-
computation. duced by a single individual, but are in most cases irrele-
vant to compare sounds emerging from communities and
landscapes where strict time and frequency homologies
5. Comments and prospectus are difficult to define. The β acoustic indices in current
use are all based on simple distances between time en-
To find a single index that summarizes all biodiversity velopes or spectral profiles. These indices are very simple
facets is undoubtedly utopian. There will not be any sin- and might need a refinement. In particular, they may not be
gle value that will reliably estimate all levels of local or optimal as they are all based on a pointwise comparison.
regional diversity. The motivation to find such a unique This is particularly the case of the Df index that operates a
parameter derives probably from the request of managers, subtraction of homologous frequency bins. These indices,
politicians and policy makers who need a single and like Df , can return unexpected important differences for
easy to understand value to take conservation decisions two frequency spectra with similar shapes but only slightly
[62, 63]. The same phenomenon seems to happen with shifted in frequency. Other metrics that compare vectors of
acoustic indices: there is a current search for a single index proportions (here frequency spectra) can be envisaged to
that would give the most reliable and complete informa- replace the index Df used so far. Such metrics could be for
tion on the acoustic and diversity states of a population, a example the Orlóci chord distance [71] and the Morisita-
community or a landscape. This explains why, after a few Horn metric [72, 73]. Eventually, another method to com-
years only, several indices have been proposed in the same pare community or landscape acoustics could be to use
time but have been used very rarely together [45] (Table I). the symbolic aggregate approximation (SAX, [74]) [55].
The α acoustic indices achieved an important success SAX consists in converting a numerical series into a char-
probably because they aspire to give a single value, a kind acter string by transforming the data into a discrete string
of signature, to an acoustic community or a soundscape. of letters. The size of the string can be chosen as well as
These indices returned congruent results revealing, for in- the length of the alphabet. This results in a dimensional-
stance, changes in bird species richness in accordance with ity reduction and lower bounding. SAX is currently in use
aural identification [48] or complex patterns of the sound- for data mining, in particular for online search of similar
scape across different temporal scales [51]. However it is soundscapes (see http://lib.real.msu.edu/) but it could be
important to note that they may be affected by several fac- used to address ecological questions where spectral dis-
tors like transitory or permanent background noise, varia- similarities have to be computed.
tion in the distance of the animals to the sensor, the relative
intensity and calling repetition of the calling animals, time All the developments in relation to acoustic indices for
and/or frequency overlap between sounds arising from dif- biodiversity assessment and landscape ecology can be con-
ferent sources. These variations should be evaluated soon sidered as a new turn in bioacoustics with a change of
in different contexts, such as different habitats (vegetation scale from species to community and landscape. A ma-
structure and composition) and different sampling efforts. jor issue in ecology is to collect data over large areas and
In addition, a clear correlation between the α indices and long time periods with a high and regular repetition rate
the level of community diversity or soundscape complex- [75]. By investigating acoustic communities and sound-
ity has not been established yet. There is therefore still scapes, bioacoustics provides an efficient way to sample
a need for a confrontation between classical direct field- large ecological units. If the scaling up towards commu-
based data like individual and/or species aural counting by nities and landscapes sounds a promising avenue for bioa-
volunteer observers and acoustic inferences. The research coustics and ecology, this process should not discredit the
of α indices is currently in development: the improvement historical species-specific approach that provides accurate
of former indices and the emergence of new indices are information on populations and species dynamics. Forth-
expected in the next years. As an example, acoustic diver- coming efforts should consider all units of the ecological
778
Sueur et al.: Biodiversity assessment ACTA ACUSTICA UNITED WITH ACUSTICA
Vol. 100 (2014)
scale shedding new light on the acoustic behaviour and [15] P. J. Morin: Community ecology. Wiley-Blackwell, Ox-
acoustic ecology of animals. ford, 1999.
[16] A. A. Chek, J. P. Bogart, S. C. Loughheed: Mating signal
Acknowledgement partitioning in multi-species assemblages: a null model test
using frogs. Ecology Letters 6 (2003) 235–247.
We thank Stuart Gage and Eric Kasten for having checked [17] A. Amézquita, S. V. Flechas, A. P. Lima, A. H. Gasser, W.
the section dealing with landscape indices. Jerôme Sueur Hödl: Acoustic interference and recognition space within
and Sandrine Pavoine were supported by a Fondation pour a complex assemblage of dendrobatid frogs. Proceedings
la Recherche sur la Biodiversité (FRB) grant and JS by of the National Academy of Sciences 108 (2011) 17058–
the SABIOD MASTODONS Big Data CNRS MI project 17063.
2012-2017. [18] A. K. D. Schmidt, H. Römer, K. Riede: Spectral niche seg-
regation and community organization in a tropical cricket
assemblage. Behavioral Ecology (2012) doi: 10.1093/be-
References heco/ars187.
[19] A. Gasc, J. Sueur, S. Pavoine, R. Pellens, P. Grandcolas:
[1] C. E. Shannon, W. Weaver: The mathematical theory of
Biodiversity sampling using a global acoustic approach:
communication. Illinois University Press, 1949.
contrasting sites with microendemics in New Caledonia.
[2] M. D. Greenfield: Signalers and receivers. Mechanisms and PLoS ONE 8 (2013) e65311.
evolution of arthropod evolution. Oxford University Press,
[20] R. T. T. Forman, M. Godron: Landscape ecology. Wiley &
Oxford, 2002.
Sons, New York, 1986.
[3] P. K. McGregor: Animal communication networks. Cam- [21] Z. Naveh, A. S. Lieberman: Landscape ecology. Theory
bridge University Press, Cambridge, 2005. and application. Springer Verlag, Berlin, 1984.
[4] T. U. Grafe: Anuran choruses as communication networks. [22] A. Farina: Ecology, cognition and landscape. Linking nat-
– In: Animal communication networks. P. McGregor (ed.). ural and social systems. Springer, Dordrecht, NL, 2010.
Cambridge University Press, Cambridge, 2005, 277–299.
[23] M. A. Acevedo, L. J. Villanueva-Rivera: Using automated
[5] D. J. Mennill, M. Battiston, D. R. Wilson, J. R. Foote, S. M. digital recording systems as effective tools for the monitor-
Doucet: Field test of an affordable, portable, wireless mi- ing of birds and amphibians. Wildlife Society Bulletin 34
crophone array for spatial monitoring of animal ecology (2006) 211–214.
and behaviour. Methods in Ecology and Evolution 3 (2012)
704–712. [24] R. S. Sousa-Lima, T. F. Norris, J. N. Oswald, D. P.
Fernandes: A review and inventory of fixed autonomous
[6] A. R. Ridley, E. M. Wiley, A. M. Thompson: The ecological recorders for passive acoustic monitoring of marine mam-
benefits of interceptive eavesdropping. Functional Ecology mals. Aquatic Mammals 39 (2013) 23–53.
28 (2013) 197–205.
[25] W. J. Sutherland: Ecological census techniques. A hand-
[7] H. Slabbekoorn: Singing in the wild. – In: Nature’s music, book. Cambridge University Press, Cambridge, 1996.
the science of birdsong. P. Marler, H. Slabbekoorn (eds.).
Elsevier, 2004, 178–205. [26] S. Parks, J. L. Miksis-Olds, S. L. Denes: Assessing marine
ecosystem acoustic diversity across ocean basins. Ecologi-
[8] J. W. Bradbury, S. L. Vehrencamp: Principles of animal cal Informatics (2014) doi: 10.1016/j.ecoinf.2013.11.003.
communication. Sinauer, Sunderland, 1998.
[27] A. E. Magurran: Measuring biological diversity. Blackwell,
[9] C. Huetz, T. Aubin: Bioacoustics approaches to locate and Malden, USA, 2004.
identify animals in terrestrial environments. – In: Sensors
for ecology, towards integrated knowledge of ecosystems. [28] S. Pavoine, M. B. Bonsall: Measuring biodiversity to ex-
J. J. Le Galliard, J. M. Guarini, F. Gaill (eds.). CNRS, Paris, plain community assembly: a unified approach. Biological
2002, 83–96. Reviews 86 (2011) 792–812.
[10] F. Samaran, N. Gandilhon, R. Prieto Gonzalez, F. Pace, [29] A. E. Magurran, B. J. McGill: Biological diversity, frontiers
A. M. Kennedy, O. Adam: Passive hydro-acoustics for in measurement and assessment. Oxford University Press,
cetacean census and localisation. – In: Sensors for ecology, Oxford, 2011.
towards integrated knowledge of ecosystems. J. J. Le Gal- [30] R. H. Whittaker: Evolution and measurement of species di-
liard, J. M. Guarini, F. Gaill (eds.). CNRS, Paris, 2012, versity. TAXON 21 (1972) 213–251.
63–82. [31] J. Liu, J. Kang, T. Luo, H. Behm, T. Coppack: Spatiotempo-
[11] B. Krause: Anatomy of the soundscape: evolving perspec- ral variability of soundscapes in a multiple functional urban
tives. Journal of the Audio Engineering Society 56 (2008) area. Landscape and Urban Planning 115 (2013) 1–29.
73–80. [32] Y. Matsinos, A. Mazaris, K. Papadimitriou, A. Mniestris,
[12] B. Pijanowski, A. Farina, S. Gage, S. Dumyahn, B. G. Hatzigiannidis, D. Maioglou, J. Pantis: Spatio-temporal
Krause: What is soundscape ecology? An introduction and variability in human and natural sounds in a rural land-
overview of an emerging new science. Landscape Ecology scape. Landscape Ecology 23 (2008) 945–959.
26 (2011) 1213–1232. [33] A. Mazaris, A. Kallimanis, G. Chatzigianidis, K. Papadim-
[13] B. Pijanowski, L. J. Villanueva-Rivera, S. L. Dumyahn, A. itriou, J. Pantis: Spatiotemporal analysis of an acoustic
Farina, B. L. Krause, B. M. Napoletano, S. H. Gage, N. environment: interactions between landscape features and
Pieretti: Soundscape ecology: the science of sound in the sounds. Landscape Ecology 24 (2009) 817–831.
landscape. Bioscience 61 (2011) 203–216. [34] W. J. Davies, M. D. Adams, N. S. Bruce, R. Cain, A.
[14] J. Roughgarden, J. Diamond: Overview: the role of species Carlyle, P. Cusack, D. A. Hall, K. I. Hume, A. Irwin, P.
interactions in community ecology. – In: Community Ecol- Jennings, M. Marselle, C. J. Plack, J. Poxon: Perception
ogy. J. Diamond, T. J. Case (eds.). Harper & Row Publish- of soundscapes: an interdisciplinary approach. Applied
ers, New York, 1986, 333–343. Acoustics 74 (2013) 224–231.
779
ACTA ACUSTICA UNITED WITH ACUSTICA Sueur et al.: Biodiversity assessment
Vol. 100 (2014)
[35] M. Rychtáriková, G. Vermeir: Soundscape categorization [52] A. Farina, N. Pieretti: Sonic environment and vegetation
on the basis of objective acoustical parameters. Applied structure: a methodological approach for a soundscape
Acoustics 74 (2013) 240–247. analysis of a Mediterranean maqui. Ecological Informat-
[36] A. J. Torija, D. P. Ruiz, A. Ramos-Ridao: Required sta- ics (2014) doi: 10.1016/j.ecoinf.2013.10.008.
bilization time, short-term variability and impulsiveness [53] D. Bormpoudakis, J. Sueur, J. Pantis: Spatial heterogeneity
of the sound pressure level to characterize the temporal of ambient sound at the habitat type level: ecological im-
composition of urban soundscapes. Applied Acoustics 72 plications and applications. Landscape Ecology 28 (2013)
(2011) 89–99. 495–506.
[37] J. R. Barber, K. R. Crooks, K. M. Fristrup: The costs of [54] J. Qi, S. H. Gage, W. Joo, B. Napoletano, S. Biswas: Sound-
chronic noise exposure for terrestrial organisms. Trends in scape characteristics of an environment. A new ecological
Ecology and Evolution 25 (2009) 180–189. indicator of ecosystem health. – In: Wetland and water re-
[38] C. Francis, J. Paritsis, C. Ortega, A. Cruz: Landscape pat- source modeling and assessment. J. Wei (ed.). CRC Press,
terns of avian habitat use and nest success are affected by Boca Raton, 2008, 201–211.
chronic gas well compressor noise. Landscape Ecology 26 [55] E. P. Kasten, S. H. Gage, J. Fox, W. Joo: The remote en-
(2011) 1269–1280. vironmental assessment laboratory’s acoustic library: an
[39] J. A. González-Oreja, A. A. De La Fuente-Díaz-Ordaz, L. archive for studying soundscape ecology. Ecological In-
Hernández-Santín, C. Bonache-Regidor, D. Buzo-Franco: formatics 12 (2012) 50–67.
Can human disturbance promote nestedness? Songbirds [56] B. Krause, S. H. Gage, W. Joo: Measuring and interpreting
and noise in urban parks as a case study. Landscape and the temporal variability in the soundscape at four places
Urban Planning 104 (2012) 9–18. in Sequoia National Park. Landscape Ecology 26 (2011)
[40] A. Rodriguez, A. Gasc, S. Pavoine, P. Grandcolas, P. 1247–1256.
Gaucher, J. Sueur: Temporal and spatial dynamics of an-
[57] H. R. Karfunkel, B. Rohde, F. J. J. Leusen, R. J. Gdanitz,
imal sound within a neotropical forest. Ecological Infor-
G. Rihs: Continuous similarity measure between nonover-
matics (2014) doi: 10.1016/j.ecoinf.2013.12.006.
lapping X-ray powder diagrams of different crystal modi-
[41] S. H. Gage, A. C. Axel: Visualization of temporal fications. Journal of Computational Chemistry 14 (1993)
change in soundscape power of a Michigan lake habitat 1125–1135.
over 4-year period. Ecological Informatics (2014) doi:
10.1016/j.ecoinf.2013.11.004. [58] R. de Gelder, R. Wehrens, J. A. Hageman: A generalized
expression for the similarity of spectra: application to pow-
[42] M. Towsey, J. Wimmer, I. Williamson, P. Roe: The use der diffraction pattern classification. Journal of Computa-
of acoustic indices to determine avian species richness in tional Chemistry 22 (2001) 273–289.
audio-recordings of the environment. Ecological Informat-
ics (2014) doi: 10.1016/j.ecoinf.2013.11.007. [59] L. Bodis, A. Ross, E. Pretsch: A novel spectra similarity
measure. Chemometrics and Intelligent Laboratory Sys-
[43] N. T. Boelman, G. P. Asner, P. J. Hart, R. E. Martin: Multi- tems 85 (2007) 1–8.
trophic invasion resistance in Hawaii: bioacoustics, field
surveys, and airborne remote sensins. Ecological Appli- [60] A. Gasc, J. Sueur, F. Jiguet, V. Devictor, P. Grandcolas, C.
cations 17 (2007) 2137–2144. Burrow, M. Depraetere, S. Pavoine: Assessing biodiversity
with sound: Do acoustic diversity indices reflect phyloge-
[44] J. Sueur, S. Pavoine, O. Hamerlynck, S. Duvail: Rapid netic and functional diversities of bird communities? Eco-
acoustic survey for biodiversity appraisal. PLoS ONE 3 logical Indicators 25 (2013) 279–287.
(2008) e4065.
[61] L. Lellouch, S. Pavoine, F. Jiguet, H. Glotin, J. Sueur: Mon-
[45] W. Joo, S. H. Gage, E. P. Kasten: Analysis and interpreta- itoring temporal change of bird communities with dissimi-
tion of variability in soundscapes along an urban rural gra- larity acoustic indices. Methods in Ecology and Evolution
dient. Landscape and Urban Planning 103 (2011) 259–276. (2014) doi: 10.1111/2041–210X.12178.
[46] L. Villanueva-Rivera, B. Pijanowski, J. Doucette, B. Pekin:
[62] J. Loh, R. E. Green, T. Ricketts, J. Lamoreux, M. Jenk-
A primer of acoustic analysis for landscape ecologists.
ins, V. Kapos, J. Randers: The Living Planet Index: Using
Landscape Ecology 26 (2011) 1233–1246.
species population time series to track trends in biodiver-
[47] B. Pekin, J. Jung, L. Villanueva-Rivera, B. Pijanowski, J. sity. Philosophical Transactions of the Royal Society B:
Ahumada: Modeling acoustic diversity using soundscape Biological Sciences 360 (2005) 289–295.
recordings and LIDAR-derived metrics of vertical forest
structure in a neotropical rainforest. Landscape Ecology [63] A. E. Magurran, S. R. Baillie, S. T. Buckland, J. M. Dick,
27 (2012) 1513–1522. D. A. Elston, E. M. Scott, R. I. Smith, P. J. Somerfield,
A. D. Watt: Long-term datasets in biodiversity research and
[48] M. Depraetere, S. Pavoine, F. Jiguet, A. Gasc, S. Duvail, J. monitoring: assessing change in ecological communities
Sueur: Monitoring animal diversity using acoustic indices: through time. Trends in Ecology and Evolution 25 (2010)
implementation in a temperate woodland. Ecological Indi- 574–582.
cators 13 (2012) 46–54.
[64] M. O. Hill: Diversity and evenness: a unifying notation and
[49] C. Gini: Variabilità e mutabilità. – In: Memorie di Meto- its consequences. Ecology 54 (1973) 427–432.
dologica Statistica. E. Pizetti, T. Salvemini (eds.). Liberia
Eredi VirgilioVeschi, Roma, Italy, 1912. [65] L. Jost: Entropy and diversity. Oikos 113 (2006) 363–375.
[50] N. Pieretti, A. Farina, D. Morri: A new methodology to in- [66] T. S. Brandes, P. Naskrecki, H. K. Figueroa: Using image
fer the singing activity of an avian community: the Acoustic processing to detect and classify narrow-band cricket and
Complexity Index (ACI). Ecological Indicators 11 (2011) frog calls. Journal of the Acoustical Society of America
868–873. 120 (2006) 2950–2957.
[51] A. Farina, N. Pieretti, L. Piccioli: The soundscape method- [67] V. B. Deecke, V. M. Janik: Automated categorization of
ology for long-term bird monitoring: a Mediterranean Eu- bioacoustic signals: avoiding perceptual pitfalls. Journal
rope case-study. Ecological Informatics 6 (2011) 354–363. of the Acoustical Society of America 119 (2006) 645–653.
780
Sueur et al.: Biodiversity assessment ACTA ACUSTICA UNITED WITH ACUSTICA
Vol. 100 (2014)
[68] M. A. Acevedo, C. J. Corrada-Bravo, H. Corrada-Bravo, [72] M. Morisita: Measuring of interspecific association and
L. J. Villanueva-Rivera, T. M. Aide: Automated classifica- similarity between communities. Memoirs of the Faculty
tion of bird and amphibian calls using machine learning: a of Science, Kyushu University, Series E (Biology) 3 (1959)
comparison of methods. Ecological Informatics 4 (2009) 65–80.
206–214. [73] H. S. Horn: Measurement of “overlap" in comparative eco-
[69] R. Bardeli, D. Wolff, F. Kurth, M. Koch, K. H. Frommolt: logical studies. American Naturalist 100 (1966) 419–424.
Detecting bird sounds in a complex acoustic environment [74] J. Lin, E. Keogh, S. Lonardi, B. Chiu: A symbolic represen-
and application to bioacoustic monitoring. Pattern Recog- tation of time series with implications for streaming algo-
nition Letters 31 (2010) 1524–1534. rithms. Proceedings of the 8th ACM SIGMOD Workshop
on Research Issues in Data Mining and Knowledge Discov-
[70] M. Towsey, B. Planitz, A. Nantes, J. Wimmer, P. Roe: A ery, San Diego, California, USA, 2003.
toolbox for animal call recognition. Bioacoustics 21 (2012)
107–125. [75] J. J. D. Greenwood, R. A. Robinson: Principles of sam-
pling. – In: Ecological census techniques. W. J. Suther-
[71] L. Orloci: An agglomerative method for classification of land (ed.). Cambridge University Press, Cambridge, 2006,
plant communities. Journal of Ecology 55 (1967) 193–206. 11–86.
781