Density dependence impacts our understanding of population resilience

Current metrics of demographic resilience (e.g., resistance, recovery) summarize how populations respond to the frequent, varied disturbances that ecological systems experience. Much of the application of these metrics has focused on the potential response of populations represented by time-invariant, density-independent structured population models to hypothetical disturbances. Here, we show that density dependence has profound and complex impacts on our understanding of resilience. We examine resilience measures in a flexible structured model with five vital rate parameters (juvenile survival, adult survival, juvenile progression, adult retrogression, and adult reproductive output) with density dependence operating on one vital rate at a time. Depending on which vital rate was subject to density effects, existing measures of demographic resilience (compensation, resistance, and recovery time) either increased or decreased with population density. Moreover, the density-independent model under-predicted the recovery time of the corresponding density-dependent model, with a greater offset for species with longer generation times and higher iteroparity. Our findings demonstrate the importance of underlying non-linear processes when examining demographic resilience, particularly if we hope to predict how natural populations will respond to real disturbances.