Environmental Surrogates for Predicting and Conserving Adaptive Genetic Variability in Tree Species

Abstract: Adaptive genetic variability within species is an essential component of biodiversity but has been largely ignored in studies aimed at assessing and predicting biodiversity of the forest environment. We used factorial regression and structuring models to test easily measured surrogates, such as ecological attributes, as predictors of adaptive genetic variation between populations of a tree species ( Eucalyptus delegatensis ). Adaptive variability was defined in terms of variation in average growth performance of 68 populations and of population-by-environment interaction across seven different environments. The best surrogates of genetic variability were measures of solar radiation and temperature range, each predicting more than 50% of the genetic variability within the species. Rock and understory types, when used either alone or in combination with other covariates, also were very efficient in discriminating between populations in groups showing similar adaptation. Significant relationships between particular surrogates and growth patterns of variation were attributed to effects of natural selection that had occurred in the population source locations. We recommend the development of studies focusing on the population level of biodiversity to improve the conservation of forest ecosystems in Australia.