In the future, many of world’s tropical forests will be recovering from human disturbance and, where prevalent, subject to increasing impacts from severe tropical cyclones caused by climate change. Long-term monitoring, experimentation, and modeling are necessary to understand the compound effects of human and natural disturbance on tropical forests. Stabilizing influences that have the potential to explain community dynamics in the context of repeated disturbance include life history trade-offs in colonization and competitive ability (competition-colonization hypothesis) or in the ability of species to survive at low resource conditions and exploit the temporary resource-rich conditions often generated in the wake of disturbance (successional niche hypothesis).

We developed Hierarchical Bayesian models to investigate impacts of repeated hurricane disturbance on reproduction, growth, and survival of 10 tree species at a 16-ha mapped plot in Puerto Rico. The species represent a range of life histories, successional stages, and responses to hurricane damage. Unlike previous approaches, our model accommodates temporal variation in process error and observations from multiple sources. We parameterized the model using growth and mortality data from four censuses of the plot taken at 5-year intervals (1990-2005) together with seed production data collected over a 13 year period (1992 -2005). We used this model to investigate the life history tradeoffs that drive community dynamics and allow tree species coexistence in this forest under a regime of repeated hurricane disturbance.

Results/Conclusions

Species’ susceptibilities to hurricane damage as reflected by changes in diameter growth and fecundity immediately following a storm were weak, highly variable, and generally unpredictable by traditional life history groupings. However, responses to the high light conditions generated in the wake of storms followed the expected patterns, with pioneer and secondary forest species exhibiting greater gains in growth and fecundity relative to shade-tolerant species. We found moderately strong relationships among species traits that are likely to influence responses to hurricanes including a tradeoff between mean species fecundity in high light versus survival in low light, and a positive relationship between colonization ability and fecundity in high light environments. In general, tradeoffs between species’ susceptibilities to hurricane damage and their response after a hurricane were weak. However, species that suffered greater reductions in diameter growth as a result of storm damage appeared to compensate by their ability to grow rapidly after hurricanes. These results suggest that some weak tradeoffs in competitive and colonization ability, in addition to successional niche processes are operating in this forest.