Experimental test of predator complementarity in ecomorphological traits on prey survivorship in tropical floodplain fish assemblages

Background/Question/Methods

Predation may strongly affect the structure and function of ecosystems, and most species are consumed by multiple predators. The functional response of prey subject to multiple predators may be shaped by the degree of overlap between predators in terms of hunting modes and habitat domains. Thus, ecomorphological traits that relate to predator foraging tactics and habitat use may be useful to predict functional responses of prey subjected to multiple predators, independently of predator identity, environment and biogeographic history. Specifically, prey mortality is expected to increase with increasing complementarity of predator traits (i.e. reduced multivariate overlap). Our objective was to test effects of ecomorphological distance between predators on mortality of prey subjected to multiple predators in experimental tropical floodplain food webs. We used principal coordinate analysis to categorize the distances between predator pairs based on ecomorphological traits (close, intermediate and far multivariate distances). The catfish Sorubim lima was included in all predator treatment pairs and ‘close’, ‘intermediate’ and ‘far’ predator combinations included Pseudoplatystoma corruscans, Hoplias aff. malabaricus, and Astronotus crassipinnis,respectively. Experiments were conducted in lagoon mesocosms (3 x 4 m; 25 cm depth) with 25% cover of floating macrophytes and the same prey assemblage for all predator treatments.

Results/Conclusions

Contrary to expectations, ecomorphological distance between predators in multiple predator assemblages had no effect on total prey mortality. Total prey mortality was significantly affected by predators in two of the three predator species pairs, but not for the intermediate-distance predator pairs. Examination of prey-specific responses to predator assemblages identified no effect of predators on Aphyocharax spp. survivorship, but significant mortality (all p<0.05) for Moenkhausia forestii. Although predators exhibited significant consumption of M. forestii, multivariate distances between predators did not affect mortality rate.  Rather than complementarity between predators that should enhance predation, our predator pairs (especially the intermediate-distance pair) may exhibit interference, leading to risk reduction for prey.  Key components of foraging behavior or habitat use may not be captured by our ecomorphological traits or may be affected by our experimental venue. Alternatively, identity-specific behaviors that depend on the composition of species present may be present, but were not testable with our experimental design. Further research is needed to better parameterize trait-based predictions of multiple predator effects.