Oviposition Site Selection in Response to Risk of Predation: Evidence from Aquatic Habitats and Consequences for Population Dynamics and Community Structure

Abstract

The ability of females to select oviposition sites based on risk of predation to their progeny should be of high selective value. Such behavior should also have large consequences for population dynamics and community structure. Given its potential importance, surprisingly few studies have tested for such a behavioral trait Predator detection, and subsequently oviposition site selection in response to risk of predation, is likely to evolve under the following combination of conditions: (i) immature stages are prone to high mortality from predation; (ii) females often have a number of patches from which they can oviposit; (iii) among-patch distributions of predators are random or contagious; (iv) among-patch distributions of predators are largely fixed from the time a prey female has oviposited until her progeny can leave the patch. Such conditions often occur for species with complex life cycles like mosquitoes and anurans that are aquatic during premetamorphic stages and terrestrial during postmetamorphic stages. I review the evidence for the existence and prevalence of such a behavioral response in aquatic habitats, detailing in particular, the system of the mosquito, Culiseta longiareolata, and its predators in temporary pools. The existing evidence suggests that this oviposition response is widespread in aquatic species, particularly those using temporary pools. Over the regional scale (encompassing all potential oviposition sites in an area), oviposition site selection in response to risk of predation should reduce the importance of consumption by predators but may increase the importance of competition. Hence, where predation at the local scale is typically thought to reduce competition, oviposition habitat selection in response to risk of predation should cause the opposite over a regional scale — i.e., increased competition. This behavior should also result in incorrectly estimating pest control by biological control agents based on classical control-treatment (predator) comparisons of pest densities.