Effects of environmental heterogeneity on predictions of tree species' abundance in response to climate warming

Most predictions of tree species' distributions at broad spatial scales are based only on one possible characterization of environmental heterogeneity. Evaluating the effects of multiple heterogeneities on predictions may help in quantifying prediction uncertaintie. In this study we investigated the effects of three levels of environmental heterogeneity on landscape-scale predictions. In addition, we analyzed how seed dispersal and interspecies competition contributes to prediction uncertainty. We used a coupled ecosystem and landscape modeling approach to predict tree species' abundance at the landscape scale. We designed multiple-species and single-species scenarios, each with the three levels of environmental heterogeneity. Our results showed the importance of considering environmental heterogeneity when predictioning tree species' abundance. For early-successional species landscape-scale predictions differed significantly among heterogeneity levels. For late-successional species, prediction uncertainties based on different heterogeneity levels were comparatively low. Seed dispersal may be a source of variation in predictions, whereas interspecies competition may reduce such variation.