Abstract
Imagine going to a garden store and buying every kind of vegetable seed packet for sale. Sow them all in a patch of bare ground, and then try to predict which particular species will be harvestable at the end of the season. Even the best ecological science is stymied by this simple challenge. An open problem is predicting or controlling the dynamics of communities, whether they are vegetable gardens or tropical forests. The goal of this work is to develop biodiversity theory that can link across scales from individuals to communities to ecosystems. A more predictive ecological science could 1) delineate the processes and mechanisms that operate at or across different spatial scales and 2) generate falsifiable predictions for biodiversity dynamics in specific systems over time. To begin to address these aims, I will test two key ideas: first, that organisms physiology needs to be built in to models of population growth; and second, that the matching between physiology and environment is not necessarily optimal, because of disequilibrium caused by lagged responses of organisms to rapid change. Over the next five years I will expand and test these ideas in a set of long-term study sites spanning the lowland tropics of Malaysia to the high alpine of Colorado.