How populations are adapted across habitats remains one of the central questions in ecology. The mode of adaptation is a product of the interaction between genetic and environmental influences. Often our inferences about phenotypes in wild populations can be confounded by the covariance between genetic and environmental effects. Cogradient variation occurs when the genetic and environmental effects on the phenotype accentuate one-another. Alternatively, countergradient variation is defined by the opposition of genetic and environmental effects on a phenotype. Countergradient variation can be particularly cryptic in that phenotypes in different populations in the wild may appear to be similar despite large environmental differences, but when these populations are held in a shared environment their phenotypes diverge due to genetic differences based on local adaptation. While much attention has been given to these phenomena, there still remain large questions as to whether there are predictable patterns across scale, taxa, or gradient types, and even the effect size to which co- and countergradient variation occur. Here we present the initial results of a meta-analysis of more than 300 studies, in which we find over 100 examples of co- and countergradient variation.
Results/Conclusions
The overall effect size for countergradient variation is 1.57 or an approximate 1.6 SD difference in response between populations expressing negative genetic and environmental covariance. Put another way, genetic compensation in locally adapted populations with negative covariance is 1.6 standard deviations different in shared environments. We also consider how taxa, the type of gradient, trait, and trait type might influence countergradient variation, finding significant effects in every category except trait type. When comparing the wild phenotypes between populations, we find that populations appear to be perfectly compensating. In other words, populations generally appear to be expressing similar values in the wild despite genetic differences. Finally, we consider the slopes of plastic responses across genotypes, finding that slopes are parallel and not expressing a Genetic x Environmental interaction. This meta-analysis represents the first quantitative and systematic investigation into the covariance of genetic and environmental effects on phenotypes across environmental gradients. As such, it has wide implications with regard to the distribution of local adaptation and trait values across environments, how those features might react to environmental disruption such as climate change, and even how phenotypic plasticity interacts with genetic evolution to promote adaptation.