How Does Tropical Forest Respond to Drought? Using Remote Sensing to Test Predictions of Functional Ecology and Resource Theory

Abstract

The Amazon is one of the largest terrestrial contributors to global carbon cycling. Photosynthesis of the Amazon rainforest, however, despite considerable in situ and space-based observations, acts as one important source of uncertainty for the carbon cycle. Slight changes in Amazon forest photosynthesis would have a substantial impact on ecosystem dynamics, biosphere-atmosphere exchange, and global carbon cycling. However, the response of Amazon forests to climate variability and long-term change - including increasing droughts - is highly uncertain, with some models predicting catastrophic forest collapse while others predict resilience. This highlights the importance of the understanding of tropical forest photosynthesis dynamics under climate change. Climate drivers alone, though important, are evidently insufficient to predict the complexity of drought responses across heterogeneous landscapes. Particularly missing is an understanding of the comprehensive biogeographic drivers of differences in photosynthetic dynamics across tropical regions from wet to dry forests and across multiple forest ecotopes defined by hydraulic environments, soil fertility and texture, and functional forest traits. Thus, this dissertation asks if a “functional biogeography” of forest behaviors can address the question: why do some forests show green-up, responding positively to water stress, suggesting a sign of genuine ecosystem resilience, whereas other forests show brown-down, responding decreases in their photosynthetic function to water stress, suggesting potential vulnerability? To determine the extent of contrasting forest responses to water stresses at different timescales, from dry seasons to interannual droughts, across different environments of Amazonia, and to develop a more mechanistic understanding of how those responses emerge along gradients of forest ecotopes and climate, I addressed this basic question in three key contexts in each of three chapters: (1) At inter-annual timescales, in the context of the large interannual Amazon droughts of 2005, 2010, and 2015; (2) At seasonal timescales (during annual dry seasons), for the global tropics as well as Amazonia; and, finally, (3) In the context of using past drought patterns (as developed in Chapter 1 for 2005, 2010, and 2015) to predict future drought response (in particular the response during the drought of 2023). I used remote sensing explorations of water-stress-induced vegetation greening/browning patterns and how they correlate with ecotopes and climate across landscapes. I further used ground-based studies of tree drought resilience or plot-scale tree demography to validate remote sensing inferences at multiple levels. The result is a biogeography that uses the insights of resource theory and functional ecology, as tested by remote sensing indices, to reveal underlying ecological mechanisms. The functional biogeography can predict tropical forest resilience and vulnerability, in response to droughts and dry seasons. This new functional biogeography of forest responses to water stress provides a framework for conservation decisions and improved predictions of heterogeneous forest responses to future climate changes, warning that Amazonia’s most productive forests are also at greatest risk, and that longer/more frequent droughts are undermining multiple ecohydrological strategies and capacities for Amazon forest resilience. This new biogeography lays the foundation for improving ecosystem and global models of vegetation feedbacks to climate.