OOS 35-8 - Role of community composition in determining tropical dry forest ecosystem responses to altered water and nutrient regimes

Thursday, August 11, 2016: 4:00 PM
Grand Floridian Blrm G, Ft Lauderdale Convention Center
David M. Medvigy, Department of Geosciences, Princeton University, Princeton, NJ, Annette Trierweiler, Department of Geosciences, Xiangtao Xu, Department of Geosciences, Princeton University, NJ and Jennifer S. Powers, Ecology, Evolution and Behavior, University of Minnesota, St. Paul, MN
Background/Question/Methods

Tropical dry forests (TDFs) are characterized by a pronounced dry season.  Especially in the neotropical dry forests, climate change is expected to bring reduced precipitation and altered precipitation seasonality.  Furthermore, increased temperatures may lead to increased evapotranspiration.  At the same time, many neotropical soils are recovering from past land use and may be nutrient-limited. The ultimate impact of these evolving resource regimes on forest composition and functioning is not well understood. Models, which are our basic tools for projecting ecosystem responses to global change, can be used to address this knowledge gap. However, few models have been specifically parameterized for TDFs.  Here, we present a new version of the Ecosystem Demography 2 (ED2) model that includes a parameterization of TDFs based on field data from Costa Rica. In particular, we focus on the model’s framework for representing limitation by multiple resources (carbon, water, nitrogen, and phosphorus). We have carried out simulations designed to illustrate (i) model-data comparison corresponding to Costa Rican field sites; (ii) simulated responses to recently-established rainfall exclusion and nutrient fertilization experiments; (iii) simulated responses to an extreme drought of the type that may be expected under future climate.

Results/Conclusions

First, we find that the model is able to accurately simulate individual-level growth and mortality rates at Costa Rican field sites over the past 7 years.  The phenology of different plant functional groups is also realistically simulated.  Second, we find that simulations that included a 3-year period of nitrogen fertilization generally led to sustained increases in plant carbon storage.  However, the magnitude of this effect was sensitive to proportion of trees that were capable of symbiotic dinitrogen fixation.  Third, simulations of a 3-year rainfall exclusion experiment and a severe long-term drought were highly sensitive to the details of stand composition.  Our preliminary simulations indicate that if (i) a stand is reliant on nitrogen fixation to maintain growth and (ii) the drought is severe enough to induce mortality of nitrogen fixers, several decades may elapse before ecosystem carbon storage recovers to pre-drought levels.  Overall, our results show that tropical dry forests may be very sensitive to changes in climate and nutrient regimes, but that it is difficult to make generalizations about ecosystem responses without controlling for variations in community composition. Modeling studies that seek to predict ecosystem and community responses to environmental change should account for these variations in composition.