Testing the link between perceived and actual risk of predation: mosquito oviposition site selection and egg predation by native and introduced fish

Abstract

According to the threat-sensitivity hypothesis, prey avoidance behaviour should reflect the magnitude of predation risk. Since predation can strongly affect reproduction success, ovipositing females are expected to adaptively adjust their predator avoidance response, or local breeding patch selectivity, in accordance with the perceived level of threat posed for their progeny by specific predators. However, association between avoidance and predation can be disrupted when the prey and the predator lack spatiotemporal opportunities to co-evolve, such as in cases of non-native predator introductions. We examined the interactions between mosquitoes (from the genus Culex) and three species of sympatric predaceous freshwater fish, a native cyprinid Barbus paludinosus, a cichlid Pseudocrenilabrus philander and an introduced poeciliid, the western mosquitofish Gambusia affinis. In an outdoor mesocosm experiment, we quantified patterns of Culex oviposition site selection across fish species using free-roaming, caged and fish-free treatments. In a complementary laboratory experiment, we tested the effectiveness of each fish species as predators of mosquito eggs and larvae. Synthesis and applications. We found evidence for: (i) mosquito egg raft predation by free-roaming fish; (ii) fish-specific avoidance by ovipositing Culex; and (iii) a positive association between fish-specific oviposition avoidance and fish-specific efficiency as an egg predator. These results contribute towards a better understanding of predator–prey co-evolution, predator-borne cue recognition, and suggest local native fish, the southern mouthbrooder Pseudocrenilabrus philander, as an alternative to Gambusia for the biocontrol of Culex mosquitoes.