Tropical dry forests (TDF) are complex, functionally diverse ecosystems that exist across a wide range of environmental gradients. TDFs are also threatened by extensive land-use change and often existing as recovering secondary forests. This combination of environmental stressors imposed by climate and disturbance results in a high potential for multiple resource constraints on forest growth, including water, nitrogen, and phosphorus limitation. Here, we ask: How do resource constraints impact TDF ecosystem structure, function, composition, and successional recovery?
We use a new version of the Ecosystem Demography (ED2) model that is parameterized for TDFs and incorporates multiple resource constraints (carbon, nitrogen, phosphorus, and water). We simulate eighteen 0.1-hectare inventory plots in Guanacaste, Costa Rica. These plots span a large nutrient gradient and vary greatly in composition. We first ran a 10-year baseline simulation that was initialized with the observed stand structure, soil properties, and meteorology. We then ran “swap” simulations where we swapped one component, such as swapping the soil nutrients of the nutrient rich and poor sites, among the sites while keeping the other components the same. We are currently expanding our modeling analyses to a broader climatic and geographic range of sites with additional TDFs sites in Yucatan, Mexico.
Results/Conclusions
Across our nutrient gradient of Costa Rican sites, we found plots with higher soil nitrogen and phosphorous showed greater biomass accumulation in the baseline simulations. Nitrogen in particular was important to forest recovery. The variation in soil nitrogen among sites profoundly affected biomass accumulation, differences in vegetation stoichiometry, and ecosystem functioning such as litter recycling. Although the coefficient of variation in soil phosphorous was three times the coefficient of variation of soil nitrogen, the impact of phosphorous on biomass accumulation and decomposition was smaller, but it shaped the demographic structure. Sufficient phosphorous resulted in greater basal area in the larger tree size classes. At least for our Costa Rican plots, our “swap” simulations indicated that the nutrient context was more important in explaining site differences in biomass accumulation than either meteorology or initial vegetation. Overall, our results demonstrate a strong link between soil nutrients and TDF structure and function and highlight the importance of including soil nutrient variation in ecosystem modeling. In addition, our results are especially relevant to recovering secondary TDFs.