Attracted to the enemy: Aedes aegypti prefers oviposition sites with predator-killed conspecifics

Abstract

Oviposition habitat choices of species with aquatic larvae are expected to be influenced by both offspring risk of mortality due to predation, and offspring growth potential. Aquatic predators may indirectly influence growth potential for prey by reducing prey density and, for filter-feeding prey, by increasing bacterial food for prey via added organic matter (feces, partially eaten victims), creating the potential for interactive effects on oviposition choices. We tested the hypothesis that the mosquito Aedes aegypti preferentially oviposits in habitats with predatory Toxorhynchites larvae because of indirect effects of predation on chemical cues indicating bacterial abundance. We predicted that A. aegypti would avoid oviposition in sites with Toxorhynchites, but prefer to oviposit where bacterial food for larvae is abundant, and that predation by Toxorhynchites would increase bacterial abundances. Gravid A. aegypti were offered paired oviposition sites representing choices among: predator presence; the act of predation; conspecific density; dead conspecific larvae; and bacterial activity. A. aegypti preferentially oviposited in sites with Toxorhynchites theobaldi predation, and with killed conspecific larvae, but failed to detect preferences for other treatments. The antibiotic tetracycline eliminated the strongest oviposition preference. Both predation by Toxorhynchites and killed larvae increased bacterial abundances, suggesting that oviposition attraction is cued by bacteria. Our results show the potential for indirect effects, like trophic cascades, to influence oviposition choices and community composition in aquatic systems. Our results suggest that predators like Toxorhynchites may be doubly beneficial as biocontrol agents because of the attraction of ovipositing mosquitoes to bacterial by-products of Toxorhynchites feeding.

Figure 1

The six treatments presented to gravid females in this study: (A) Predator Presence, (B) Actual Predation, (C) Conspecific Density, (D) Mechanical Killing, (E) Mechanical Killing vs. Predation, (F) Bacterial Activity. The control in all treatments is represented by the container on the left, while the treatment is represented by the container on the right.

Figure 2

Oviposition response (OAI values) of Ae. aegypti gravid female to control and treatment containers within the six treatments. ***P<0.001, **P<0.01, P>0.05 = NS. Predators used in this experiment were Toxorhynchites theobaldi.

Figure 3

Production of new bacterial biomass based on measurement of protein synthesis (quantified by measuring incorporation of tritiated H-leucine (4,5-[H])) in oviposition containers with: 50 Ae. aegypti prey larvae alone (prey alone), 33 prey alive plus 17 crushed (prey crushed) and 50 prey larvae plus one predator (prey+predator). Different letters indicate significant difference among means. Predators used in this experiment were Toxorhynchites rutilus.