Interaction assessments in correlated and autocorrelated environments
(2007) 2. p.111131 Abstract
 Natural food webs are embedded in a variable environment, which causes population densities to fluctuate, despite a potential stable equilibrium. Population interactions as well as the characteristics of the environmental fluctuations determine the resulting population dynamics. Populations sensitive to the same kind of environmental disturbances will show correlated responses in their respective growth rates. Such 'environmental correlation' between species can have profound effects on the populations' dynamics, e.g. generating a positive correlation between the abundances of two competitors, which makes a direct correlation a highly inappropriate measure of population interactions. However, multivariate time series analysis will still... (More)
 Natural food webs are embedded in a variable environment, which causes population densities to fluctuate, despite a potential stable equilibrium. Population interactions as well as the characteristics of the environmental fluctuations determine the resulting population dynamics. Populations sensitive to the same kind of environmental disturbances will show correlated responses in their respective growth rates. Such 'environmental correlation' between species can have profound effects on the populations' dynamics, e.g. generating a positive correlation between the abundances of two competitors, which makes a direct correlation a highly inappropriate measure of population interactions. However, multivariate time series analysis will still identify and quantify population interactions correctly. The picture is more complicated if the environmental fluctuations are correlated over time – environmental autocorrelation causes biases in interaction assessments and possibly falsely identified delayed interactions. We present approximate expressions for the estimation bias, which show that the bias is the weakest when food web dynamics are close to unstable. In the absence of close to unstable dynamics the only way avoid this estimation error is to incorporate the most important environmental drivers as covariates in the time series analysis. (Less)
Please use this url to cite or link to this publication:
http://lup.lub.lu.se/record/629236
 author
 Ripa, Jörgen ^{LU} and Ives, Anthony R.
 organization
 publishing date
 2007
 type
 Chapter in Book/Report/Conference proceeding
 publication status
 published
 subject
 keywords
 food web dynamics, environmental stochasticity, multivariate time series analysis, interaction assessment, autocorrelation, correlation, community dynamics
 host publication
 The impact of environmental variability on ecological systems. The Peter Yodzis Fundamental Ecology Series Vol. 2
 editor
 Vasseur, David A. ; McCann, Kevin S. ; and
 volume
 2
 pages
 111  131
 publisher
 Springer
 external identifiers

 wos:000248254100006
 ISBN
 9781402058509
 language
 English
 LU publication?
 yes
 id
 b3844d4de429464c8485e7bc0ad042dc (old id 629236)
 date added to LUP
 20160404 10:54:01
 date last changed
 20181121 21:01:26
@inbook{b3844d4de429464c8485e7bc0ad042dc, abstract = {Natural food webs are embedded in a variable environment, which causes population densities to fluctuate, despite a potential stable equilibrium. Population interactions as well as the characteristics of the environmental fluctuations determine the resulting population dynamics. Populations sensitive to the same kind of environmental disturbances will show correlated responses in their respective growth rates. Such 'environmental correlation' between species can have profound effects on the populations' dynamics, e.g. generating a positive correlation between the abundances of two competitors, which makes a direct correlation a highly inappropriate measure of population interactions. However, multivariate time series analysis will still identify and quantify population interactions correctly. The picture is more complicated if the environmental fluctuations are correlated over time – environmental autocorrelation causes biases in interaction assessments and possibly falsely identified delayed interactions. We present approximate expressions for the estimation bias, which show that the bias is the weakest when food web dynamics are close to unstable. In the absence of close to unstable dynamics the only way avoid this estimation error is to incorporate the most important environmental drivers as covariates in the time series analysis.}, author = {Ripa, Jörgen and Ives, Anthony R.}, booktitle = {The impact of environmental variability on ecological systems. The Peter Yodzis Fundamental Ecology Series Vol. 2}, editor = {Vasseur, David A. and McCann, Kevin S.}, isbn = {9781402058509}, language = {eng}, pages = {111131}, publisher = {Springer}, title = {Interaction assessments in correlated and autocorrelated environments}, volume = {2}, year = {2007}, }