Experimental Aquatic Food Webs: Interactions between Two Predators and Two Prey

Publisher Website
Google Scholar
Abstract
We performed four replicates of all 16 combinations of the presence and absence of two predators (larvae of the dragonfly Anax junius and adult salamanders Notophthalmus viridescens) and two anuran prey (Bufo americanus and Rana palustris larvae) in an array of artificial temporary ponds. The two species of anurans were introduced at densities high enough to cause density-dependent reductions in survival and body size at metamorphosis and to increase larval period. The two species were in competition when raised together. Notophthalmus reduced the density of Bufo and caused the survivors to metamorphose early and at a small size. Anax caused an even greater reduction in survival but not as strong an acceleration of metamoprhosis. Newts also reduced the density of Rana, but survivors benefited by growing rapidly. The effects of Anax were even stronger; the only Rana tadpoles that were able to metamorphose in the 2 mo of the experiment were from ponds in which Anax reduced densities enough to permit rapid growth of the surviving tadpoles. In systems with one predator and two prey, the anurans shared the risk of predation and were mutualists rather than competitors. In systems with two predators and one prey, the risks of predation were additive. In systems with all four species, risks were again spread over both anurans, making them mutalists. There was no evidence for interactions between the predators within the limits of the experiment''s design. One- and two-species systems predicted low and nearly equal survival of the two prey at the end of the experiment (4% for Bufo and 7% for Rana). The observed survival in the four-species systems was 35% for Bufo and 59% for Rana. Linear additive models of species interactions may give useful qualitative predictions but may be systematically biased in their quantitative predictions.