Appendix A. Experiment to examine potential artifacts in main ratio-dependent experiment.
In our functional response experiment we used different experimental durations and mussel sizes for C. maenas and H. sanguineus in order to standardize prey depletion. We conducted a supplementary experiment to verify that results obtained in the functional response experiment were not an artifact of these experimental differences between predator species. In this experiment, we examined predation by C. maenas and H. sanguineus on mussel prey over a constant experimental duration (24 h) and on the same sized mussels (1417 mm shell length) for each predator species. Our purpose was to capture the salient features of the functional response experiment (i.e., the relative importance of predator interference for C. maenas and H. sanguineus), but was not to determine the functional response, estimate parameters, or to compare between prey dependence and ratio dependence. In addition, we wanted to examine predation at a larger range of prey densities, consistent with the larger range of densities at our field site, to determine whether the effect of predator density differed at higher prey densities where competition for prey may not be as great. We therefore examined predation at four prey densities (10, 20, 100, and 200 prey per enclosure, using the same enclosures as described in the main text) and with two and four predators. Each of these eight predator-prey combinations was replicated four times for each species during separate trials.
We compared the effects of prey and predator density for each species separately using two-way ANOVAs on log transformed (to achieve homoscedasticty) per capita predation with prey density (four levels) and predator density (two levels) as fixed factors.
Results of this experiment indicated that C. maenas predation increased at higher prey densities (ANOVA, df = 3,24, F = 45.44, P < 0.0001, Fig. A1.A), and that increasing predator density caused a decrease in C. maenas per capita predation (ANOVA, df = 1,24, F = 22.78, P < 0.0001, Fig. A1.A) that was consistent across prey densities (predator density × prey density interaction df = 3,24, F = 0.38, P = 0.77). In contrast, while increasing prey density also had a positive impact on H. sanguineus predation (ANOVA, df = 3,24, F = 11.66, P < 0.0001, Fig. A1.B), increasing predator density had no influence on H. sanguineus predation (ANOVA, df = 1,24, F = 1.78, P < 0.20, Fig. A1.B), and this was again consistent across prey densities (predator density-prey density interaction df = 3,24, F = 0.60, P = 0.62).
These results are consistent with results in our main experiment where predation by both species increased asymptotically with prey density, and where predator density had a greater influence on predation by C. maenas than by H. sanguineus. Based on these results, we conclude that results of our functional response experiment given in the main text were not influenced by differences in experimental procedures with C. maenas and H. sanguineus, nor were they influenced by the range of experimental prey densities that we used.
FIG. A1. Per capita mussel consumption rates of C. maenas (part A) and H. sanguineus (part B) when experiments were conducted using the same size mussel prey for both predator species and the same experimental duration. Values are means ± SE (n = 4).