Mechanistic, physiologically based models can explain much of the variation in forest productivity with climate and soils, but little of the variation in mortality fluxes and biomass stocks. As lead scientist of the Forest Carbon Research Initiative for the Center for Tropical Forest Science-Forest Global Earth Observatory (CTFS-ForestGEO), I strive to implement standardized measurements of forest carbon stocks and fluxes across this network of forest plots, assist network collaborators with associated analyses and manuscripts, and contribute to related syntheses. My lab is also quantifying landscape-scale variation in forest canopy structure and dynamics in Panama using camera-carrying unmanned aerial vehicles and photogrammetry.

Lianas (woody climbing plants) and vines (nonwoody climbing plants) vary widely in abundance within and among tropical forests, with important implications for forest carbon budgets. Lianas and vines reduce tree growth, increase tree mortality rates, and alter forest structure and composition. Past studies have established that climbing plants are more abundant in drier forests, younger forests, and on edges – but scientists currently lack a mechanistic understanding for what determines their abundance in any forest, much less its variation across forests. We are pursuing a variety of empirical and theoretical studies to identify the factors that control the abundances of these structural parasites.

Individual tropical forests commonly contain >100 woody plant species in an area the size of a football field, and can contain >1,000 species in a square kilometer. Ecologists have long debated what mechanisms prevent one or a few species from outcompeting and excluding the others. Tradeoffs in performance among spatially varying habitats and temporally varying climate conditions clearly play a role, as do interactions with specialized natural enemies that tend to disproportionately plague and disadvantage any given species in areas where it is locally more common. We work to evaluate the strength of different mechanism and quantify their roles through a combination of theory and empirical analyses.

Co-occurring plant species differ tremendously in their life history strategies and in related functional traits. A classic example is seed mass, which commonly varies more than 1-million-fold among tree species within any given tropical forest. The persistence of this variation clearly indicates that one or more underlying tradeoffs prevents any one strategy from being everywhere superior. But what exactly are these underlying tradeoffs? Using mutually informed theoretical development and empirical analyses, we seek to identify the underlying tradeoffs and capture them mechanistically in sufficient detail to quantitatively explain observed patterns of variation.

Earth system models (ESMs) simulate the earth’s climate system and its interaction with vegetation, and are used to predict the influences of policy scenarios on future atmospheric composition and climates. Among other things, these models simulate the feedbacks between tropical forests and the climate system. However, these models are currently unable to correctly capture observed spatial and temporal variation in tropical forest carbon budgets, and diverge wildly in their predictions of future responses. We are collaborating with two ESM development teams to improve the representation of tropical forests in these models by building more accurate representations of key processes and improving parameter estimates. Visit the pages NGEE-Tropics and GFLD LM3 for more details.

The amount of leaf area deployed and the age distribution of leaves varies seasonally and interannually in tropical forests in relation to climate, and is critical to determining temporal variation in ecosystem carbon, water, and energy fluxes. We hypothesize that interspecific variation in phenology is a key determinant of which species thrive in drier vs. wetter forests, in sunnier vs. cloudier years, and in rich vs. poor soils. We are quantifying the phenology of thousands of trees using camera-carrying unmanned aerial vehicles; analyzing how individual tree and liana phenology relates to temporal climate variation, spatial environmental variation, and species traits; and evaluating the implications of phenology for ecosystem fluxes by integrating these data with models.