Food web structure and the strength of transient indirect effects

The relative importance of direct and indirect effects in ecological communities remains unresolved. Indirect effects may diminish as they propagate through highly reticulate food webs. We tested this hypothesis by assembling replicate food webs of different complexity in laboratory microcosms, and comparing the transmission of indirect effects through these webs. By providing the top predator (Didinium) with either one (Paramecium) or two (Paramecium and Colpidium) species of protists as prey, we created linear or reticulate food webs where we could examine the transient response of predators to an indirect effect. Addition of Chlamydomonas, a small alga consumed by Paramecium, but not by Colpidium?, perturbed the system and generated an indirect effect on Didinium. We expected the proportional response of Didinium to Chlamydomonas addition would be smaller in the reticulate web containing alternative, unperturbed prey (Colpidium). We measured predator response as predator yield, the maximum number of predators produced prior to overexploitation of prey and subsequent predator decline. The ratio of yield in perturbed bottles to yield in unperturbed bottles measures the proportional response of Didinium to Chlamydomonas addition. We expected this ratio to be smaller with Colpidium present. Contrary to expectations, alternative prey enhanced rather than diminished predator response to the perturbation. This resulted from competition between the prey species, a factor ignored in some simple verbal arguments. Food web complexity may have unanticipated consequences for the strength of indirect effects.