9701053 Chazdon This proposal addresses the critical need to create predictive models of seedling response to changing patterns of environmental heterogeneity. The PIs propose to develop mechanistic models of whole-plant carbon gain that will be coupled with studies of plant performance in relation to light heterogeneity in tropical forest environments. The PIs will examine levels of plant response to natural spatial and temporal variation of light to integrate leaf physiology with whole-plant architecture, leaf display, and biomass allocation. The approach involves two components. First, known genotypes of three shade-tolerant tree species will be planted along a continuous gradient of light availability from the center of a gap, through gap edges, and into the forest understory. An architectural model will be used to scale from leaf-level measures of photosynthesis to whole-plant estimates of carbon gain. From these data the PIs will develop and parameterize a maximum likelihood model of whole-plant carbon gain and growth as a function of continuous variation in light availability. This model will then be used to predict growth of transplanted seedlings of the same family lines in secondary forest and native tree plantations, two habitat types differing in patterns of light availability. Since changing land use patterns can dramatically affect both spatial and temporal patterning of light, this information will help to understand and predict how plants may respond to current and future land use changes. Predicting the consequences of heterogeneous light environments for seedling populations in tropical forests requires an explicit characterization of the pattern of heterogeneity as well as patterns of plant response to this heterogeneity.
