Accounting for spatial pattern when modeling organism-environment interactions

Timothy H. Keitt, Ottar N. Bjørnstad, Philip M. Dixon, Steve Citron-Pousty

Research output: Contribution to journalArticle

255 Citations (Scopus)

Abstract

Statistical models of environment-abundance relationships may be influenced by spatial autocorrelation in abundance, environmental variables, or both. Failure to account for spatial autocorrelation can lead to incorrect conclusions regarding both the absolute and relative importance of environmental variables as determinants of abundance. We consider several classes of statistical models that are appropriate for modeling environment-abundance relationships in the presence of spatial autocorrelation, and apply these to three case studies: 1) abundance of voles in relation to habitat characteristics; 2) a plant competition experiment; and 3) abundance of Orbatid mites along environmental gradients. We find that when spatial pattern is accounted for in the modeling process, conclusions about environmental control over abundance can change dramatically. We conclude with five lessons: 1) spatial models are easy to calculate with several of the most common statistical packages; 2) results from spatially-structured models may point to conclusions radically different from those suggested by a spatially independent model; 3) not all spatial autocorrelation in abundances results from spatial population dynamics; it may also result from abundance associations with environmental variables not included in the model; 4) the different spatial models do have different mechanistic interpretations in terms of ecological processes - thus ecological model selection should take primacy over statistical model selection; 5) the conclusions of the different spatial models are typically fairly similar - making any correction is more important than quibbling about which correction to make.

Original languageEnglish (US)
Pages (from-to)616-625
Number of pages10
JournalEcography
Volume25
Issue number5
DOIs
StatePublished - Oct 1 2002

Fingerprint

organisms
modeling
autocorrelation
statistical models
environmental factors
organism
plant competition
environmental gradient
mite
mites
population dynamics
case studies
habitats
habitat
experiment

All Science Journal Classification (ASJC) codes

  • Ecology, Evolution, Behavior and Systematics

Cite this

Keitt, Timothy H. ; Bjørnstad, Ottar N. ; Dixon, Philip M. ; Citron-Pousty, Steve. / Accounting for spatial pattern when modeling organism-environment interactions. In: Ecography. 2002 ; Vol. 25, No. 5. pp. 616-625.
@article{c5e79424e6b64968af7a4636f38c71b9,
title = "Accounting for spatial pattern when modeling organism-environment interactions",
abstract = "Statistical models of environment-abundance relationships may be influenced by spatial autocorrelation in abundance, environmental variables, or both. Failure to account for spatial autocorrelation can lead to incorrect conclusions regarding both the absolute and relative importance of environmental variables as determinants of abundance. We consider several classes of statistical models that are appropriate for modeling environment-abundance relationships in the presence of spatial autocorrelation, and apply these to three case studies: 1) abundance of voles in relation to habitat characteristics; 2) a plant competition experiment; and 3) abundance of Orbatid mites along environmental gradients. We find that when spatial pattern is accounted for in the modeling process, conclusions about environmental control over abundance can change dramatically. We conclude with five lessons: 1) spatial models are easy to calculate with several of the most common statistical packages; 2) results from spatially-structured models may point to conclusions radically different from those suggested by a spatially independent model; 3) not all spatial autocorrelation in abundances results from spatial population dynamics; it may also result from abundance associations with environmental variables not included in the model; 4) the different spatial models do have different mechanistic interpretations in terms of ecological processes - thus ecological model selection should take primacy over statistical model selection; 5) the conclusions of the different spatial models are typically fairly similar - making any correction is more important than quibbling about which correction to make.",
author = "Keitt, {Timothy H.} and Bj{\o}rnstad, {Ottar N.} and Dixon, {Philip M.} and Steve Citron-Pousty",
year = "2002",
month = "10",
day = "1",
doi = "10.1034/j.1600-0587.2002.250509.x",
language = "English (US)",
volume = "25",
pages = "616--625",
journal = "Ecography",
issn = "0906-7590",
publisher = "Wiley-Blackwell",
number = "5",

}

Accounting for spatial pattern when modeling organism-environment interactions. / Keitt, Timothy H.; Bjørnstad, Ottar N.; Dixon, Philip M.; Citron-Pousty, Steve.

In: Ecography, Vol. 25, No. 5, 01.10.2002, p. 616-625.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Accounting for spatial pattern when modeling organism-environment interactions

AU - Keitt, Timothy H.

AU - Bjørnstad, Ottar N.

AU - Dixon, Philip M.

AU - Citron-Pousty, Steve

PY - 2002/10/1

Y1 - 2002/10/1

N2 - Statistical models of environment-abundance relationships may be influenced by spatial autocorrelation in abundance, environmental variables, or both. Failure to account for spatial autocorrelation can lead to incorrect conclusions regarding both the absolute and relative importance of environmental variables as determinants of abundance. We consider several classes of statistical models that are appropriate for modeling environment-abundance relationships in the presence of spatial autocorrelation, and apply these to three case studies: 1) abundance of voles in relation to habitat characteristics; 2) a plant competition experiment; and 3) abundance of Orbatid mites along environmental gradients. We find that when spatial pattern is accounted for in the modeling process, conclusions about environmental control over abundance can change dramatically. We conclude with five lessons: 1) spatial models are easy to calculate with several of the most common statistical packages; 2) results from spatially-structured models may point to conclusions radically different from those suggested by a spatially independent model; 3) not all spatial autocorrelation in abundances results from spatial population dynamics; it may also result from abundance associations with environmental variables not included in the model; 4) the different spatial models do have different mechanistic interpretations in terms of ecological processes - thus ecological model selection should take primacy over statistical model selection; 5) the conclusions of the different spatial models are typically fairly similar - making any correction is more important than quibbling about which correction to make.

AB - Statistical models of environment-abundance relationships may be influenced by spatial autocorrelation in abundance, environmental variables, or both. Failure to account for spatial autocorrelation can lead to incorrect conclusions regarding both the absolute and relative importance of environmental variables as determinants of abundance. We consider several classes of statistical models that are appropriate for modeling environment-abundance relationships in the presence of spatial autocorrelation, and apply these to three case studies: 1) abundance of voles in relation to habitat characteristics; 2) a plant competition experiment; and 3) abundance of Orbatid mites along environmental gradients. We find that when spatial pattern is accounted for in the modeling process, conclusions about environmental control over abundance can change dramatically. We conclude with five lessons: 1) spatial models are easy to calculate with several of the most common statistical packages; 2) results from spatially-structured models may point to conclusions radically different from those suggested by a spatially independent model; 3) not all spatial autocorrelation in abundances results from spatial population dynamics; it may also result from abundance associations with environmental variables not included in the model; 4) the different spatial models do have different mechanistic interpretations in terms of ecological processes - thus ecological model selection should take primacy over statistical model selection; 5) the conclusions of the different spatial models are typically fairly similar - making any correction is more important than quibbling about which correction to make.

UR - http://www.scopus.com/inward/record.url?scp=0036779345&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0036779345&partnerID=8YFLogxK

U2 - 10.1034/j.1600-0587.2002.250509.x

DO - 10.1034/j.1600-0587.2002.250509.x

M3 - Article

AN - SCOPUS:0036779345

VL - 25

SP - 616

EP - 625

JO - Ecography

JF - Ecography

SN - 0906-7590

IS - 5

ER -