Tropical tree mortality
Tree mortality, particularly in tropical forests, is poorly understood. One reason is that few trees die, and therefore the sample size is always small: there is always variability in the number of dead trees per year in a given place, but one cannot attribute that easily to any factor or just chance. Another reason is that when trees die they tend to disappear (faster in the tropics than in other systems), so the information about the causes of mortality is rapidly lost. Finally, all communities represent a biased sample of the trees that existed in that same place: the trees that remain tend to be those that tend to survive more than the average tree in that community. This (the survival bias) causes a lot of problems when it comes to do analyses and make inference about the functioning of the system. On the other hand, mortality is a critical componen of large-scale predictive models of forest dynamics, or terrestrial ecosystems functioning. It has been identified as a priority in the development of Global Vegetation Dynamic Models, and my main line of research now is obtaining useful data and results that could potentially be integrated into the FATES model, a GDVM being developed by NGEE-Tropics.
Commonness and rarity in tropical forests
Commonness and rarity are two sides of the same coin, and both aspects have occupied a central role in my research since my PhD thesis. Commonness and rarity can be measured along different dimensions simultaneously to classify species into different classes of rarity or to measure multi-dimensional commonness in a quantitative way.
An important part of my research is dedicated to the quantification and quantitative testing of the oligarchy hypothesis, which states that tropical forest communities are dominated by a limited and definable set of species that are locally abundant and spatially widespread (these are called "oligarchic species"). There are great challenges and opportunities in the development of a theory of two-dimensional commonness (= local abundance + frequency or spatial distribution) of broad application.
An important part of my research is dedicated to the quantification and quantitative testing of the oligarchy hypothesis, which states that tropical forest communities are dominated by a limited and definable set of species that are locally abundant and spatially widespread (these are called "oligarchic species"). There are great challenges and opportunities in the development of a theory of two-dimensional commonness (= local abundance + frequency or spatial distribution) of broad application.
Beta diversity in tropical forests and community assembly
The composition of any forest changes from one point to another. The understanding of the intensity and causes of such change, and its spatial configuration, are central issues of my research. I have invested a lot of time examining many variation partitioning methods, a widely employed but extremely tricky tool, especially when applied to multivariate methods. We have obtained some results indicating that spatially structured environment are more likely to reflect scale-dependent processes than spatially un-structured gradients. Besides, I am currently involved in a pair of methodological works that we are developing thanks to the insight gained in these methodologies. Finally, we have detailed soil data of one hundred plots of tropical forests, with which we are working to better understand how soil-mediated community assembly works across scales.
