Background/Question/Methods   Species migration in response to climatic change—both modern and historical—is most often modeled by taking modern correlations between a species' occurrences and the environment at those locations. While this is a good first approximation of species ranges, this approach ignores many factors known to influence species distributions, particularly biotic effects on range limits and migration, the role of population outliers and fine-scale environmental variation in the landscape, and basic questions of whether the fundamental or realized niche is being used to predict species ranges. Here we use data from a long term study of tree plots and associated experiments along an Andes to Amazon elevation gradient, as well as niche modeling, to ask (1) what is the role of microclimatic variability on species distribution along environmental gradients? (2) what is the role of biotic interactions in migration in this system, and when would it be likely to act?, and (3) what kind of uncertainty does our assumptions about ecological niches introduce into range models and species extinctions?

Results/Conclusions   We find that microclimatic environmental variability in solar radiation, disturbance, and cloud immersion causes large variability in species distribution, even within the same basic temperature and rainfall regimes. Biotic interactions due to changes in herbivore communities are important along the gradient, and demographic modeling shows that they will have large and differential effects on plant effective fecundity and migration rate. Finally, the basic question of fundamental versus realized niche has critical importance in predicting species responses to climate change, and associated extinctions.