GAP model validation for a subtropical dry forest and predicting poor transition of abandoned fields into secondary forests in Puerto Rico

Background/Question/Methods

Because of human pressures, the need to understand and predict the long-term dynamics and development of subtropical dry forests is urgent. Simulation modeling is designed to predict how a forest will respond to various disturbances and can lead to better management and conservation practices of threatened and biologically diverse subtropical dry forests. Through modifications to the ZELIG simulation model, including the development of species- and site-specific parameters and internal modifications, the capability to predict forest change within the 4500-ha Guanica State Forest in Puerto Rico can now be accomplished.  Published data-sets and additional data from the Puerto Rican Forest Inventory Analysis were used to parameterize the new gap model, ZELIG-TROP. We used data from the 1.44-ha permanent plot located inside the Guanica Forest to test the model. The first objective of this study is to accurately re-create the observed forest succession for a Puerto Rican subtropical dry forest using ZELIG-TROP. The second objective consists in testing the realism of ZELIG-TROP to predict the successional pattern of secondary forests across a gradient of abandoned fields that currently are being reclaimed as forests.

Results/Conclusions

For the first objective, the gap model ZELIG-TROP is successfully used for the first time in a subtropical dry forest ecosystem. The simulated results, which started from bare ground, have a strong resemblance to the observed forest structure measured over the past 30 years. Simulated total basal area, species composition, total stem density, and biomass all closely resembled the observed Puerto Rican forest. The total above ground biomass (Mg ha^-1) was the least likely to be accurately predicted. For the second objective, abandoned fields that are on degraded lands do not fully recover and reach a mature forest status during the simulated time period. Abandoned fields that are not on degraded lands do reach a mature forest status, but there is a lag time. The forest recovery trends matched predictions published in other studies; attributes involving resources acquisition (increase in height, canopy coverage, density) are the fastest to recover, but attributes used for structural development (biomass, basal area) are relatively slow in recovery. Recovery of these latter degraded systems may take longer time periods, as simulated in this report. Biomass and basal area, two attributes that increase during later successional stages, are significantly lower than a mature forest throughout 200 years of simulation, suggesting that the definition of resiliency in subtropical dry forests needs to be partially redefined.