A new method for detecting species associations with spatially autocorrelated data

Stephen H. Roxburgh, Peter Chesson

Research output: Contribution to journalArticle

56 Citations (Scopus)

Abstract

Many organisms display patchiness in their distribution patterns over a wide range of spatial scales. Patchy distribution patterns can be caused by processes such as growth, migration, reproduction, and mortality, which result in neighboring areas being more likely to contain a species than distant areas, a phenomenon known as positive spatial autocorrelation. When species are patchily distributed, the within-species spatial randomness assumptions of the standard statistical tests for detecting species associations are seriously violated. Using these tests under such circumstances can lead to incorrect rejection of the null hypothesis. To address this problem we introduce a new test for detecting species associations-the random patterns test. This test takes into account spatial autocorrelation by including the characteristics of the spatial pattern of each species into the null model. A randomization procedure was used to generate the null distribution of the test statistic. The random patterns test is illustrated with data collected from an herbaceous understory community of a Eucalyptus forest near Canberra, Australia.

Original languageEnglish (US)
Pages (from-to)2180-2192
Number of pages13
JournalEcology
Volume79
Issue number6
StatePublished - 1998
Externally publishedYes

Fingerprint

testing
autocorrelation
methodology
patchiness
Eucalyptus
understory
method
test
statistical analysis
statistics
organisms
mortality
distribution
organism

Keywords

  • Pattern analysis
  • Randomization test
  • Semivariograms
  • Spatial autocorrelation
  • Species association tests

ASJC Scopus subject areas

  • Ecology

Cite this

A new method for detecting species associations with spatially autocorrelated data. / Roxburgh, Stephen H.; Chesson, Peter.

In: Ecology, Vol. 79, No. 6, 1998, p. 2180-2192.

Research output: Contribution to journalArticle

@article{e168ef71d17e4b80a5e3c5ae45ee9749,
title = "A new method for detecting species associations with spatially autocorrelated data",
abstract = "Many organisms display patchiness in their distribution patterns over a wide range of spatial scales. Patchy distribution patterns can be caused by processes such as growth, migration, reproduction, and mortality, which result in neighboring areas being more likely to contain a species than distant areas, a phenomenon known as positive spatial autocorrelation. When species are patchily distributed, the within-species spatial randomness assumptions of the standard statistical tests for detecting species associations are seriously violated. Using these tests under such circumstances can lead to incorrect rejection of the null hypothesis. To address this problem we introduce a new test for detecting species associations-the random patterns test. This test takes into account spatial autocorrelation by including the characteristics of the spatial pattern of each species into the null model. A randomization procedure was used to generate the null distribution of the test statistic. The random patterns test is illustrated with data collected from an herbaceous understory community of a Eucalyptus forest near Canberra, Australia.",
keywords = "Pattern analysis, Randomization test, Semivariograms, Spatial autocorrelation, Species association tests",
author = "Roxburgh, {Stephen H.} and Peter Chesson",
year = "1998",
language = "English (US)",
volume = "79",
pages = "2180--2192",
journal = "Ecology",
issn = "0012-9658",
publisher = "Ecological Society of America",
number = "6",

}

TY - JOUR

T1 - A new method for detecting species associations with spatially autocorrelated data

AU - Roxburgh, Stephen H.

AU - Chesson, Peter

PY - 1998

Y1 - 1998

N2 - Many organisms display patchiness in their distribution patterns over a wide range of spatial scales. Patchy distribution patterns can be caused by processes such as growth, migration, reproduction, and mortality, which result in neighboring areas being more likely to contain a species than distant areas, a phenomenon known as positive spatial autocorrelation. When species are patchily distributed, the within-species spatial randomness assumptions of the standard statistical tests for detecting species associations are seriously violated. Using these tests under such circumstances can lead to incorrect rejection of the null hypothesis. To address this problem we introduce a new test for detecting species associations-the random patterns test. This test takes into account spatial autocorrelation by including the characteristics of the spatial pattern of each species into the null model. A randomization procedure was used to generate the null distribution of the test statistic. The random patterns test is illustrated with data collected from an herbaceous understory community of a Eucalyptus forest near Canberra, Australia.

AB - Many organisms display patchiness in their distribution patterns over a wide range of spatial scales. Patchy distribution patterns can be caused by processes such as growth, migration, reproduction, and mortality, which result in neighboring areas being more likely to contain a species than distant areas, a phenomenon known as positive spatial autocorrelation. When species are patchily distributed, the within-species spatial randomness assumptions of the standard statistical tests for detecting species associations are seriously violated. Using these tests under such circumstances can lead to incorrect rejection of the null hypothesis. To address this problem we introduce a new test for detecting species associations-the random patterns test. This test takes into account spatial autocorrelation by including the characteristics of the spatial pattern of each species into the null model. A randomization procedure was used to generate the null distribution of the test statistic. The random patterns test is illustrated with data collected from an herbaceous understory community of a Eucalyptus forest near Canberra, Australia.

KW - Pattern analysis

KW - Randomization test

KW - Semivariograms

KW - Spatial autocorrelation

KW - Species association tests

UR - http://www.scopus.com/inward/record.url?scp=0031671399&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0031671399&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:0031671399

VL - 79

SP - 2180

EP - 2192

JO - Ecology

JF - Ecology

SN - 0012-9658

IS - 6

ER -