Invertebrate Predation and Lotic Prey Communities: Evaluation of In Situ Enclosure/Exclosure Experiments

The influence of the stonefly Kogotus nonus on prey communities was assessed by varying the predator density within small enclosures in a first—order stream in southwestern Alberta. Experimental containers with natural densities of prey, standardized substrate and detritus, and zero, one, two, or three Kogotus were buried in a riffle for 10 during three periods: July 1981, June 1982, and July 1982. Kogotus depressed the densities of Thienemaniella and some other Orthocladiinae, but had no effect on the densities of Corynoneura (Orthocladiinae), Stempellinella (Tanytarsini), and Baetis tricaudatus (Ephemeroptera). Significant reduction of the densities of Thienemaniella and other Orthocladiinae occurred only when these prey items were present at high densities. The extent to which prey densities were depressed was found to be dependent on predator density within the enclosures: increased predator densities resulted in lower prey densities. However, when predator density in the containers exceeded that commonly observed in the riffle, predator efficiency dropped, probably because of interference between predators, the existence of prey refuges, or changes in predator search effort. An experimental design in which containers that excluded predators were compared with controls accessible to predators was used to determine if results would be similar to results from the enclosure experiments. Using this design, observed results could not clearly be attributed to the presence or absence of the predator, and were more likely due to differences in abiotic conditions between the experimental and control containers. It was concluded that observation of predator effects in field manipulative experiments may be strongly dependent on the selection of an experimental design in which predator effects are not confounded with container effects.