Search

Interactions among predators that prey on each other and are potential competitors for shared prey (intraguild [IG] predators) are widespread in terrestrial ecosystems and have the potential to strongly influence the dynamics of terrestrial food webs. Ants and spiders are abundant and ubiquitous terrestrial IG predators, yet the strength and consequences of interactions between them are largely unknown. In the leaf-litter food web of a deciduous forest in Kentucky (USA), we tested the direct and indirect effects of ants on spiders and a category of shared prey (Collembola) by experimentally subsidizing ants in open plots in two field experiments. In the first experiment, ant activity was increased, and the density of ants in the litter was doubled, by placing carbohydrate and protein baits in the center of each plot. Gnaphosa spiders were almost twice as abundant and Schizocosa spiders were half as abundant in baited plots relative to controls. There were more tomocerid Collembola in baited plots, suggesting possible indirect effects on Collembola caused by ant-spider interactions. The second experiment, in which screening of two mesh sizes selectively excluded large and small worker ants from a sugar bait, revealed that the large ants, primarily Camponotus, could alone induce similar effects on spiders. Gnaphosa biomass density was almost twice as high in the plots where large ants were more active, whereas Schizocosa biomass density was reduced by half in these plots. Although tomocerid densities did not differ between treatments, tomocerid numbers were negatively correlated with the activity of Formica, another large ant species. Path analysis failed to support the hypothesis that the ant Camponotus indirectly affected tomocerid Collembola through effects on densities of spiders. However, path analysis also revealed other indirect effects of Camponotus affecting tomocerids. These results illustrate the complexity of interactions between and within two major IG predator groups with disparate predatory behaviors, complexities that will have consequences for functioning of the forest-floor food web. [ABSTRACT FROM AUTHOR]

1. A field experiment was conducted to determine how short-term changes in moisture can alter activity-densities of spiders and springtails.2. In a Kentucky forest 10 unfenced 4-m2 plots were divided into two rainfall treatments. A clear roof over five plots excluded rainfall to simulate severe drought conditions (droughttreatment). Water was sprayed on the five uncovered plots at a rate equal to two times the long-term mean in order to establish thehigh-rainfalltreatment. Activity-densities of Collembola and spiders were measured using pitfall traps designed to sample the top, middle, and bottom layers of leaf litter. The experiment ran from 20 July to 23 September 2001.3. Overall (i.e. litter layers pooled) activity-density (mean number trapped each sampling date) of Collembola was≈ 60% lower in drought plots than in plots receiving increased precipitation. Surprisingly, overall spider activity-density was≈ 1.6 times greater in the drought plots.4. Differences in rainfall affected the spatial stratification of Collembola and spiders in strikingly different ways. Activity-densities of neither group differed between drought and high-rainfall treatments in the bottom litter layer. Collembola activity-density was three times greater in the top and middle litter layers in high-rainfall plots than in drought plots. In contrast, spider activity-density did not differ between treatments in the top layer, but activity-density was decreased by 50% in the middle layer of high-rainfall plots compared with drought plots.5. Three Collembola families (Sminthuridae, Tomoceridae, and Entomobryidae) accounted for most of the Collembola pattern. The spider response was due to altered activity-density of one family of wandering spider, the Gnaphosidae. [ABSTRACT FROM AUTHOR]