Analyzing vegetation-environment associations from regional to local scales: A case study in eastern Amazonia (Brazil)

This dissertation focuses on analyzing at regional and local scales the association between vegetation and physical environmental factors of eastern Amazonia. At a regional scale and within the conceptual setting of direct gradient analyses I developed comprehensive spatially explicit analyses of the relations between geographical ranges of major plant formations and the underlying template of their physical environment. The implementation of spatially explicit analyses is a departure from previous studies that have ignored the spatial component of the vegetation-environment association, so this work is a pioneering approach to the study of vegetation-environment relation(s) across biomes. Water balance factors derived as a function of climatic and geophysical variables were found to be the most important determinants of vegetation patterns. The variance in vegetation that could be explained by physical factors---while controlling for space---varied across vegetation types. More than half of the explainable variance in the vegetation can be accounted for by the spatial structure of the environment.; The results of analyzing the structure of plant communities at local scales indicate that each of the structural parameters measured is significantly and distinctively associated with a specific set of environmental variables. The use of functional vegetation types to characterize community complexity based on plant growth patterns provides evidence of the existence of optimal carbon allocation strategies, which explains the recognized correlation between plant physiognomy, community architecture and the physical characteristics of environments.; Spatially explicit empirical models were successfully implemented to map the geographical distribution of vegetation and in particular the transitions between major vegetation types. This study provides a comprehensive, realistic and interpretable set of biophysical maps that are important not only in terms of scientific value, but also for socioeconomic and environmental applicability. The approach taken here allows mapping the expected response patterns of vegetation across geographic regions, and its potential uses include the implementation of management and development strategies, the prediction of potential changes in land cover, and the testing and monitoring of broad scale models of vegetation responses to global change.