Tim Keitt (University of Texas, Austin)
Abstract. Spatial heterogeneity is a ubiquitous feature of ecological systems that interacts strongly with movement and dispersal, typically across many scales. I present several studies related to modeling the influence of spatial heterogeneity on ecological communities and populations. Ecologists today are placing more emphasis on functional traits of species in determining population- and community- level phenomena. In one case study, I present data from tropical forest plots that link species traits to environmental gradients. Unlike many previous studies, the modeling involves individual growth and survival rather than simple correlations based on presence and absence. We develop a Bayesian model for trait-based “filtering” of local tree assemblages based on traits, and extend this discussion to species coexistence. I present another study that abstractly models the influence of habitat fragmentation and environmental gradients on community patterns. This work revisits the old SLOSS debate adding the component of within-reserve gradients. Using a single unified model of spatial dependence, we contrast relative scales of fragmentation, underlying environmental gradients and dispersal distances. The results indicate that the interaction of these scales is critical for understanding ecological outcomes of habitat loss and fragmentation. Time permitting, I will sketch some results related to analyzing range-wide habitat connectivity of Jaguar populations.