Tropical forests play a critical role in the exchange of carbon between land and atmosphere, highlighting the urgency of understanding the effects of climate change on these ecosystems. The most optimistic predictions of climate models indicate that global mean temperatures will increase by up to 2 0C with some tropical regions experiencing extreme heat. Drought and heat-induced tree mortality will accelerate the release of carbon to the atmosphere and decrease evaporative cooling, creating a positive feedback that greatly exacerbates global warming. Thus, under a warmer and drier climate, tropical forests may become net sources, rather than sinks, of carbon. Earth system models have not reached a consensus on the magnitude and direction of climate change impacts on tropical forests. Also, these models are simplistic, calling into question the reliability of their predictions. Thus, there is an immediate need to improve the representation of tropical forests in earth system models to make robust predictions. The goal of our study is to quantify the responses of tropical forests to climate variability and improve the predictive capacity of terrestrial ecosystem models.
Results/Conclusions
We have collected species-specific physiological and functional trait data from 144 tree species in a Puerto Rican rainforest to parameterize the Ecosystem Demography model (ED2). The large amount of data generated by this research will lead to better validation and lowering the uncertainty in future model predictions. To best represent the forest landscape in ED2, all the trees have been assigned to three plant functional types (PFTs): early, mid, and late successional species. Trait data for each PFT were synthesized in a Bayesian meta-analytical model and posterior distributions of traits were used to parameterize the ED2 model. Model predictions show that biomass production of late successional PFT (118.89 ton/ha) was consistently higher than early (71.33 ton/ha) and mid (13.21 ton/ha) PFTs. However, mid successional PFT had the highest contributions to NPP for the modeled period. Ensemble runs were conducted to propagate parameter uncertainty through the model. These runs were used to construct error estimates around model forecasts, to compare modeled and observed aboveground biomass, and to identify which processes and tree species need further study.