Top-down cascade from a bitrophic predator in an old-field community

We tested the hypothesis that a bitrophic (third and fourth level) arthropod predator can exert a cascading, top—down influence on other arthropods and plants in an early successional old field. First—stadium mantids, Tenodera sinensis, were added to replicated open—field plots in numbers corresponding to naturally occurring egg hatch density and allowed to remain for $\approx 2$ mo. Sticky—trap dispersal barriers around both control and mantid—addition plots allowed us to monitor emigration of arthropods continuously during the experiment. Biomass of herbivores, carnivores, and plants, and abundances of arthropod taxa within plots were determined at the beginning, middle, and end of the experiment. The impact of mantids on the community was a top—down trophic cascade, beginning at the fourth trophic level and evident at each of the lower three levels. Mantids induced marked behavioral responses in other predators, but inteference among predators did not prevent the trophic cascade. The most common predators, cursorial spiders, emigrated from mantid addition plots in significantly greater numbers than from controls. This behavioral response may have resulted from avoidance of predation or competition. Mantids decreased biomass of herbivorous arthropods through predation, and this decrease in turn increased biomass of plants. Therefore, these generalist predators were able to decrease herbivory enough to affect plant growth. This and other recent studies indicate that top—down effects can be important in structuring terrestrial communities. Ours is the first example of a top—down cascade by a generalist arthropod predator in a nonagricultural ecosystem and illustrates the importance of detecting behavioral responses in studies of trophic interactions.