Prey persistence can be difficult where predators are generally abundant, particularly when prey cannot behaviorally select times and places of low risk. However, persistence can be enhanced by passive mechanisms to exploit spatial and temporal refugia, such as spatial inheritance of refugia and variance in timing of vulnerable stages. Here, we evaluate the potential for both mechanisms to explain gypsy moth (Lymantria dispar) persistence in the presence of an abundant predator of their pupae, the white-footed mouse (Peromyscus leucopus), in upstate New York forests. We measured small-scale mouse activity with extensive track-plate sampling in 2002-2005 and 2007-2009, and used the empirical relationship between mouse track activity and pupal predation rate to translate the map of activity into a map of predicted pupal survival. We then quantified the magnitude, spatial scale, and temporal persistence of heterogeneity in predicted survival, and combined those results with empirically fitted dispersal kernel models to estimate the spatial inheritability of refugia and the expected impact on gypsy moth population growth. We combined a simple model of seasonal population growth of mice with hypothesized relationships between mouse abundance and mouse-free space, to produce a model predicting overall pupal survival under scenarios with different variances in pupation timing.
Results/Conclusions
The magnitude, spatial autocorrelation, and between-year correlation of predicted pupal survival all we depended strongly on overall mouse abundance. Years with moderate-to-high mouse abundance had showed the strongest spatial structuring and temporal persistence of relatively safe areas, indicating the greatest potential for spatial inheritance of refugia when these predators were abundant. However, empirical estimates of dispersal kernels indicated that short-distance dispersal of larvae (which is necessary to inherit refugia the allowed maternal survival) may be less prevalent than originally though. As a result, we estimated that spatial inheritance of refugia could increase l modestly at best (median generally < 10% increase). Alternatively, our modeling efforts suggest that developmental “stagger” can substantially reduce the impact of mouse predation during relatively high mouse years, by ensuring that some larvae pupate before the landscape saturates with mice. During a high-mouse year (20 mice/ha in June), even a modest variance in pupation time increased l from <0.2 to ca 0.5, nearly a 3-fold increase in resulting density that could enable the highly fecund gypsy moth to rapidly achieve high densities during subsequent low or moderate mouse populations. Our findings highlight the importance of passive exploitation of temporal and spatial refugia in enhancing prey persistence.