COS 66-2 - Demographic lability instead of buffering in a long-lived tropical tree

Abstract: Environmental fluctuations and anthropogenic disturbance are fundamental drivers of population dynamics. Understanding species response to environmental variation is critical for developing conservation strategies. The demographic buffering hypothesis suggests that natural selection will lead to low variability of highly influential demographic processes. Demographic buffering is expected due to the nonlinearity in dynamics in response to perturbations of demographic processes. Because some organisms are more likely to shrink instead of dying thereby limiting population size variation, shrinkage is expected to be a driver of demographic buffering. Previous tests of the demographic buffering hypothesis yield mixed results, but long-lived species are expected to buffer against environmental variability. In addition, our understanding of the role of chronic anthropogenic disturbance in influencing demographic buffering or lability is limited. The goals of this study are to test the demographic buffering hypothesis and examined the effects of disturbance, shrinkage, and nonlinearity in species response over the short- and long-terms. We used four years of demographic data of a long-lived tropical tree species in two contrasting climatic environments to develop matrix projection models. We used these matrices to estimate lower-level vital rates transient and asymptotic elasticities and bias corrected coefficient of variations for each vital rate. We tested for significant negative correlation between the square of the (transient or asymptotic) elasticities and coefficient of variations for each demographic process to detect demographic buffering. We tested the effect of shrinkage, seedling and adult survival rate and nonlinear population dynamics response on the strength of demographic buffering. We found no support for demographic buffering hypothesis but rather a positive relationship between the square of vital rates coefficient of variation and elasticities indicating demographic lability. This trend was consistent over the short and long terms and irrespective of climatic conditions. The strength of demographic lability increased with harsh environmental conditions and chronic anthropogenic disturbance. Contrary to expectation, shrinkage, seedling, and adult survival have no effects on the strength of demographic lability. However, we found a significant nonlinear positive relationship between the slope of the transfer function, a proxy for the degree of non-linearity between viral rates and population growth rate, and the strength of demographic lability both for short and long terms. Our study demonstrates that contrary to prediction that longevity buffers against demographic variation, demographic lability can be selected for in a long-lived species with a limited role of shrinkage but a stronger role of nonlinear population response to perturbation.