Estimating species richness of tropical bird communities from rapid assessment data

Sebastian K. Herzog, Michael Kessler, Thomas Cahill

Research output: Contribution to journalArticle

102 Citations (Scopus)

Abstract

Rapid assessment surveys of tropical bird communities are increasingly used to estimate species richness and to determine conservation priorities, but results of different studies are often not comparable due to the lack of standardization. On the basis of computer simulations and six years of field testing, we evaluated the recently proposed "20-species-list" survey method and statistical estimators for assessing species richness of tropical bird communities. This method generates a species-accumulation curve by subdividing consecutive observations of birds into lists of 20 species, thus relating cumulative species richness to the number of observations rather than time or space and thereby accounting for moderate differences in observer qualification and field conditions. Species accumulation curves from computer-simulated communities and two empirical data sets from Bolivia were analyzed with nine species richness estimators to evaluate estimator accuracy with respect to variations in species-list size, sample size, species-pool size, and community structure. For empirical and most simulated data sets, the MMMEAN estimator performed best, but it was more sensitive to differences in community structure than most other estimators. The CHAO 2 estimator, which was recommended by previous studies, performed reasonably well but was considerably more sensitive to sample size than MMMEAN. The bootstrap and first- and second-order jackknife estimators performed poorly. We recommend using MMMEAN or, when standard deviations of richness estimates are indispensable, CHAO 2 with 10-species lists for estimating species richness of tropical bird communities and propose a set of standard survey rules. Careful examination of estimator accumulation curves is required, however, and a technique based on the ratio between estimator and species accumulation curve is suggested to control for the confounding effects of sampling effort. Overall, the species-list method combined with statistical richness estimation is doubtlessly much more standardized and valuable than simple comparisons of one-dimensional locality lists and represents a promising tool for conservation assessment and the study of avian diversity patterns in the tropics.

Original languageEnglish (US)
Pages (from-to)749-769
Number of pages21
JournalAuk
Volume119
Issue number3
StatePublished - Jul 2002
Externally publishedYes

Fingerprint

species richness
bird
species diversity
birds
community structure
Bolivia
methodology
standardization
computer simulation
sampling
tropics
species pool
survey method
size structure
testing

ASJC Scopus subject areas

  • Animal Science and Zoology

Cite this

Estimating species richness of tropical bird communities from rapid assessment data. / Herzog, Sebastian K.; Kessler, Michael; Cahill, Thomas.

In: Auk, Vol. 119, No. 3, 07.2002, p. 749-769.

Research output: Contribution to journalArticle

Herzog, Sebastian K. ; Kessler, Michael ; Cahill, Thomas. / Estimating species richness of tropical bird communities from rapid assessment data. In: Auk. 2002 ; Vol. 119, No. 3. pp. 749-769.
@article{dc802c17962a45c8b9492c775d3218f3,
title = "Estimating species richness of tropical bird communities from rapid assessment data",
abstract = "Rapid assessment surveys of tropical bird communities are increasingly used to estimate species richness and to determine conservation priorities, but results of different studies are often not comparable due to the lack of standardization. On the basis of computer simulations and six years of field testing, we evaluated the recently proposed {"}20-species-list{"} survey method and statistical estimators for assessing species richness of tropical bird communities. This method generates a species-accumulation curve by subdividing consecutive observations of birds into lists of 20 species, thus relating cumulative species richness to the number of observations rather than time or space and thereby accounting for moderate differences in observer qualification and field conditions. Species accumulation curves from computer-simulated communities and two empirical data sets from Bolivia were analyzed with nine species richness estimators to evaluate estimator accuracy with respect to variations in species-list size, sample size, species-pool size, and community structure. For empirical and most simulated data sets, the MMMEAN estimator performed best, but it was more sensitive to differences in community structure than most other estimators. The CHAO 2 estimator, which was recommended by previous studies, performed reasonably well but was considerably more sensitive to sample size than MMMEAN. The bootstrap and first- and second-order jackknife estimators performed poorly. We recommend using MMMEAN or, when standard deviations of richness estimates are indispensable, CHAO 2 with 10-species lists for estimating species richness of tropical bird communities and propose a set of standard survey rules. Careful examination of estimator accumulation curves is required, however, and a technique based on the ratio between estimator and species accumulation curve is suggested to control for the confounding effects of sampling effort. Overall, the species-list method combined with statistical richness estimation is doubtlessly much more standardized and valuable than simple comparisons of one-dimensional locality lists and represents a promising tool for conservation assessment and the study of avian diversity patterns in the tropics.",
author = "Herzog, {Sebastian K.} and Michael Kessler and Thomas Cahill",
year = "2002",
month = "7",
language = "English (US)",
volume = "119",
pages = "749--769",
journal = "Auk",
issn = "0004-8038",
publisher = "Ornithological Societies of North America",
number = "3",

}

TY - JOUR

T1 - Estimating species richness of tropical bird communities from rapid assessment data

AU - Herzog, Sebastian K.

AU - Kessler, Michael

AU - Cahill, Thomas

PY - 2002/7

Y1 - 2002/7

N2 - Rapid assessment surveys of tropical bird communities are increasingly used to estimate species richness and to determine conservation priorities, but results of different studies are often not comparable due to the lack of standardization. On the basis of computer simulations and six years of field testing, we evaluated the recently proposed "20-species-list" survey method and statistical estimators for assessing species richness of tropical bird communities. This method generates a species-accumulation curve by subdividing consecutive observations of birds into lists of 20 species, thus relating cumulative species richness to the number of observations rather than time or space and thereby accounting for moderate differences in observer qualification and field conditions. Species accumulation curves from computer-simulated communities and two empirical data sets from Bolivia were analyzed with nine species richness estimators to evaluate estimator accuracy with respect to variations in species-list size, sample size, species-pool size, and community structure. For empirical and most simulated data sets, the MMMEAN estimator performed best, but it was more sensitive to differences in community structure than most other estimators. The CHAO 2 estimator, which was recommended by previous studies, performed reasonably well but was considerably more sensitive to sample size than MMMEAN. The bootstrap and first- and second-order jackknife estimators performed poorly. We recommend using MMMEAN or, when standard deviations of richness estimates are indispensable, CHAO 2 with 10-species lists for estimating species richness of tropical bird communities and propose a set of standard survey rules. Careful examination of estimator accumulation curves is required, however, and a technique based on the ratio between estimator and species accumulation curve is suggested to control for the confounding effects of sampling effort. Overall, the species-list method combined with statistical richness estimation is doubtlessly much more standardized and valuable than simple comparisons of one-dimensional locality lists and represents a promising tool for conservation assessment and the study of avian diversity patterns in the tropics.

AB - Rapid assessment surveys of tropical bird communities are increasingly used to estimate species richness and to determine conservation priorities, but results of different studies are often not comparable due to the lack of standardization. On the basis of computer simulations and six years of field testing, we evaluated the recently proposed "20-species-list" survey method and statistical estimators for assessing species richness of tropical bird communities. This method generates a species-accumulation curve by subdividing consecutive observations of birds into lists of 20 species, thus relating cumulative species richness to the number of observations rather than time or space and thereby accounting for moderate differences in observer qualification and field conditions. Species accumulation curves from computer-simulated communities and two empirical data sets from Bolivia were analyzed with nine species richness estimators to evaluate estimator accuracy with respect to variations in species-list size, sample size, species-pool size, and community structure. For empirical and most simulated data sets, the MMMEAN estimator performed best, but it was more sensitive to differences in community structure than most other estimators. The CHAO 2 estimator, which was recommended by previous studies, performed reasonably well but was considerably more sensitive to sample size than MMMEAN. The bootstrap and first- and second-order jackknife estimators performed poorly. We recommend using MMMEAN or, when standard deviations of richness estimates are indispensable, CHAO 2 with 10-species lists for estimating species richness of tropical bird communities and propose a set of standard survey rules. Careful examination of estimator accumulation curves is required, however, and a technique based on the ratio between estimator and species accumulation curve is suggested to control for the confounding effects of sampling effort. Overall, the species-list method combined with statistical richness estimation is doubtlessly much more standardized and valuable than simple comparisons of one-dimensional locality lists and represents a promising tool for conservation assessment and the study of avian diversity patterns in the tropics.

UR - http://www.scopus.com/inward/record.url?scp=0036656048&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0036656048&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0036656048

VL - 119

SP - 749

EP - 769

JO - Auk

JF - Auk

SN - 0004-8038

IS - 3

ER -