Predicting Tree Function and Mortality in a Changing Tropical Environment

Forests face uncertain futures as a changing climate yields increases in temperature, atmospheric carbon, and drought frequency across the globe. In water-limited environments, trade-offs emerge between hydraulic safety and photosynthetic productivity, as photosynthesis has an associated water cost. Physiological diversity within and between forests can dictate which trees survive, which trees die, and the mechanisms that govern these outcomes. This diversity and complexity is often poorly represented in earth systems models; there is a need to bridge the gap between large-scale approaches and empirical examinations of plant response to environmental stress.

Here, we've employed a model originally developed by John Sperry to evaluate a tree's proximity to hydraulic failure over time. The model assumes that trees will selectively open their stomata to photosynthesize only when the potential carbon gains outweigh the associated water losses, tracking proximity to hydraulic failure across a time series of real-world weather drivers for a given set of parameters derived directly from plant functional traits. These traits are partitioned across tissues so that points of hydraulic failure can be identified precisely. In our approach, we've parameterized the model based on field collections in a tropical census plot, evaluated the resulting forest dynamics against data and observed mortality, and extended the model into future years using weather data generated from climate models. This analysis predicts how trees die, considers which trees are most vulnerable to a changing climate, and may inform subsequent whole-earth approaches that leverage physiological data to forecast forest dynamics at a massive scale.