Density dependence and stochastic variation in overabundant, unmanaged white-tailed deer herds of eastern national parks

In many parts of the eastern U.S., white-tailed deer populations have existed for decades at densities high enough to cause public concern due to decreased forest regeneration, more frequent deer-vehicle collisions, and increased spread of human and animal diseases. The National Park Service (NPS) has initiated management actions at several sites focused on decreasing deer density without a broader understanding of population dynamics that contribute to population size and fluctuation over time. Using 7-year to 13-year sets of deer monitoring data from 16 NPS sites, I estimated parameters of the theta-logistic model (TLM), a widely-used population model which integrates the two key factors of population dynamics: density dependence and environmental stochasticity. I found that: (1) the density dependent parameter of the TLM (theta) can be estimated from time-series if the ratio of process variance to sampling variance is more than 1.6 and if parameter bias is simulated and corrected; (2) deer populations existing at a lower equilibrium density fluctuated more over time and conversely larger populations fluctuated less; and (3) overall population variation is positively related to environmental stochasticity.