COS 34-4 - Non-linear impacts of succession on survival and growth of naturally-recruited tree seedlings during tropical forest restoration

Tuesday, August 9, 2016: 2:30 PM
Palm A, Ft Lauderdale Convention Center
T. Trevor Caughlin, School of Forest Resources & Conservation, University of Florida, Gainesville, FL and Marinés De la Peña, Facultad de Ciencias, Universidad Nacional Autonoma de Mexico, Mexico, Mexico
Background/Question/Methods

Non-linear and unpredictable changes in tree stem density, biomass and species composition are the norm during tropical forest succession. Tree demography drives these changes and could explain why divergent successional trajectories emerge during tropical forest restoration. The seedling-to-sapling transition is a key environmental filter for tree populations in old-growth forests, but remains poorly understood during tropical reforestation. We used hierarchical Bayesian models to quantify the demography of >1000 naturally-recruited individuals of 50 tree species during the first seven years of secondary succession in tropical pastures. Our study design includes experimental manipulation of community structure in two treatments, planted with either animal or wind-dispersed tree species, as well as an unplanted control. We hypothesized that changing environmental conditions in plots during the first seven years of succession would initially result in low growth and survival rates, as trees compete with pasture grass, followed by a period of high growth and survival, as establishing trees outcompete the grass, and ending with lower growth and survival due to increased tree competition. We also hypothesized that time-dependent vital rates would vary between the treatments, with higher initial growth and survival rates in planted treatments.

Results/Conclusions

We found strong evidence for time-dependent changes in growth, with a non-monotonic (hump-shaped) relationship between time and growth rate. This relationship varied between treatments, as peak growth rate occurred later in time for the control treatment, with a community-level maximum growth rate at 3.19 years (95% CI: 2.98-3.45), relative to the planted treatments with community-level maximum growth rates at 2.81 (95% CI: 2.55-3.02) and 2.80 (95% CI: 1.76-3.14) years for the animal and wind treatments, respectively. In contrast, tree survival declined over time, with highest survival rates during the first 3 years of succession. For both growth and survival, there were important differences between treatments. In the wind-dispersed treatment, relative performance of species changed dramatically over time, while in the animal-dispersed and control treatments, tree species with high initial growth and survival tended to remain top-performers throughout the seven year study period. For example, the most abundant species (Bursera simaruba) showed a sharp decline in survival from >80% to <20% over the seven year period in the wind-dispersed treatment, but not in the animal or control treatments. Overall, our results demonstrate that successional age can have major consequences for the demography of naturally-recruited seedlings during restoration.