Exploring the risk of savanna expansion in Tropical South America under climate change

Abstract

Tropical South American landscapes are covered by massive forests and savannas as dominant biomes. Forests are usually found in the wetter regions and savannas in the drier, yet both biomes can co-exist where rainfall conditions are moderate. In the coming decades, climate is projected to become drier in the tropical part of the continent, implying that savannas may expand their range over forested areas. This study will address the risk of savanna expansion, as well as the mechanisms and the regions where such transitions are most likely to happen. It will contribute to a larger thematic research project financed by the program MICROSOFT-FAPESP, entitled "Towards an understanding of tipping points within Tropical South American biomes", which follows a multi-scale framework and is inspired by the theory of dynamical complex systems with alternative attractors. In collaboration with researchers that have a well known expertise in the topic we will add two objectives to the large thematic project. First we ask the question: which mechanisms may drive savanna expansion? We will combine information from remote sensing and field assessments to understand whether savanna expansion can be driven by (a) plant-soil-fire interactions at the local scale, and (b) the relative capacities of forest and savanna trees to disperse at the regional scale. Both approaches will be developed in two regions with contrasting climatic conditions. Second, we will assess how forest resilience varies spatially using two methods: (a) comparing forest recovery after disturbances along a gradient of climatic conditions, and (b) testing if the current distribution of forest and savanna tree species can be explained by the mechanisms identified in the first objective. For this we will use data on tree species abundance from > 1500 field plots across tropical South America. We will also investigate the hypothesis that forests with more species typical of savannas in the tree community may have been savannas in a recent past, implying that if climate becomes drier, these forests may be the first to collapse. Our combination of approaches should reveal how and where savannas are more likely to expand as climatic conditions change, allowing South American societies to manage such risk and its consequences. (AU)