| 1. |
- Carpenter, Stephen R., et al.
(författare)
-
Accelerate Synthesis in Ecology and Environmental Sciences
- 2009
-
Ingår i: BioScience. - : Oxford University Press (OUP). - 0006-3568 .- 1525-3244. ; 59:8, s. 699-701
-
Tidskriftsartikel (refereegranskat)abstract
- Ecology is a leading discipline in the synthesis of diverse knowledge. Ecologists have had considerable experience in bringing together diverse, multinational data sets, disciplines, and cultural perspectives to address a wide range of issues in basic and applied science. Now is the time to build on this foundation and invest in ecological synthesis through new national or international programs. While synthesis takes place through many mechanisms, including individual efforts, working groups, and research networks, centers are extraordinarily effective institutional settings for advancing synthesis projects.
|
|
| 2. |
- Ripa, Jörgen, et al.
(författare)
-
Interaction assessments in correlated and autocorrelated environments
- 2007
-
Ingår i: The impact of environmental variability on ecological systems. The Peter Yodzis Fundamental Ecology Series Vol. 2. - 9781402058509 ; 2, s. 111-131
-
Bokkapitel (övrigt vetenskapligt/konstnärligt)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.
|
|
