Direct and indirect effects of multiple enemies on water strider mating dynamics

Previous studies have shown that green sunfish, Lepomis cyanellus, have strong effects on the activity, habitat use, social interactions and mating dynamics of a stream-dwelling water strider, Aquarius remigis (family Gerridae, hence, gerrids). In nature, however, stream pools often contain not just sunfish and water striders, but also smaller fish such as minnows. Here, we used factorial experiments in seminatural streams to document the direct and indirect effects of sunfish and fathead minnows, Pimephales promelas, on water strider survival and behavior. Sunfish, minnows and gerrids all consume surface prey (here, crickets); thus these three species are potential food competitors. Sunfish eat minnows. Accordingly, the presence of sunfish caused minnows to increase their schooling behavior and shift their activity from the surface toward the bottom substrate. The presence of sunfish was also associated with an increase in the number of missing gerrids, whereas minnows caused relatively little gerrid disappearance. Most interestingly, the presence of minnows decreased the effect of sunfish on gerrid disappearance rates; that is, minnows apparently had an indirect positive effect on water strider survival. We suggest that this indirect positive effect reflects the fact that minnows are alternative prey for sunfish. The effects of sunfish and minnows on gerrid mortality explained the influence of these fish on gerrid behavior. Sunfish caused decreases in male gerrid activity, female availability, mating activity, mating frequency and mating duration. Larger males had a mating advantage over smaller males only in pools with sunfish and no minnows. Sunfish also caused a borderline significant decrease in the large female mating advantage. These results were all observed in previous studies and can be viewed as adaptive responses to predation risk. These patterns were not consistent with the expected effects of sunfish as food competitors with water striders. In contrast, minnows had relatively little influence on water strider behavior and the few significant effects were the opposite of those of sunfish. Minnows caused increases in female activity and in mating duration, a decrease in the large male mating advantage and an increase in the large female mating advantage. These patterns fit the view that minnows caused an increase in gerrid hunger, i.e., that minnows acted as food competitors with gerrids. Finally, planned contrasts against controls showed that, in the presence of both sunfish and minnows, water striders showed no significant behavioral responses to fish (i.e., gerrid behavior in pools with sunfish and minnows did not significantly differ from behavior in fishless pools). The most likely mechanism explaining this pattern is a dilution of sunfish predation risk due to the presence of minnows serving as alternative prey for sunfish.