Search for a command to run...
We conducted field and laboratory experiments to examine the relative contributions of mortality from natural enemies and cannibalism to population regulation of the wolf spider Schizocosa ocreata. This cursorial spider is a common generalist predator in the leaf litter of forests of the eastern United States. In our study site in Maryland, densities of recently dispersed young S. ocreata can reach 60—90 individuals/m 2 , but decline to <15 spiders/m 2 by early autumn. We investigated the cause of this rapid decline in spiderling density. We first performed a predator—exclusion field experiment to determine if natural enemies contributed significantly to spider mortality. Replicated enclosures with lip barriers to prevent emigration were stocked with newly dispersed S. ocreata spiderlings at either 0.75 X or 3 X mean field density, or had no S. ocreata added (0X). Before spiderlings were introduced, we reduced potential enemies in half of the plots by sifting the litter and removing other spiders, centipedes and predaceous beetles, and by installing 3—mm mesh netting to exclude birds, wasps and other predators of S. ocreata. Over a period of 2.5 mo we measured effects of natural enemies and spider density on prey (Collembola), spider growth, and spider survival. Spider mortality was high (≈ 80%) and strongly density dependent (DD). Survival in the 3$\times$ field—density plots was 50% lower than in the 0.75 X plots. Increasing spiderling density did not significantly reduce Collembola density and resulted only in a minor DD reduction in spiderling growth (≈ 10%). Although overall spider mortality was high, excluding natural enemies did not significantly improve spider survival. The absence of emigration and the high spider mortality rates in the predator—reduction treatment strongly implicate cannibalism as the major DD mortality factor. We then conducted two laboratory experiments in order to obtain direct measurements of the rates of cannibalism. Rates of cannibalism were measured in 0.3—m 2 arenas stocked with S. ocreata at densities equivalent to those used in the field experiment. In Experiment 1, habitat complexity and prey availability were manipulated in arenas stocked with S. ocreata at 3 X mean field densities. In Experiment 2, we manipulated spider density (0.75 X and 3 X) and prey density to evaluate the DD component of cannibalism. The laboratory experiments confirmed the importance of cannibalism as a DD mortality factor. DD mortality from cannibalism was influenced by both prey abundance and habitat complexity. The presence of prey significantly reduced rates of cannibalism, and there was a significant interaction between habitat complexity and prey availability. When prey were present, rates of cannibalism were higher in the complex habitat than the simple habitat. In the absence of prey, increasing habitat complexity had no effect on rates of cannibalism. Rates of cannibalism in the treatment that most closely mimicked field conditions–complex leaf—litter with prey at natural densities–did not differ significantly from the rate of mortality of S. ocreata in the field experiment. Considered together, the results of our field and laboratory experiments lead to the conclusion that cannibalism can act as a strong DD mortality factor for young instars of S. ocreata, with the potential to regulate densities of this abundant cursorial spider.