Fish cultivation varies from large scale stocking of free-ranging or caged fish in large reservoirs, to small scale introduction in artificial ponds in rural properties embedded in variable spatial networks. The resulting widespread presence of generalist fish predators in freshwater habitats may have important consequences for freshwater metacommunity structure. One of the premises of metacommunity theory is that the importance of niche-based processes is dependent on the spatial context of habitat patches. Here we experimentally tested whether and how spatial isolation influences the impact on insect community assembly of introduced generalized fish predators commonly used in aquaculture. We constructed twenty-four 1,200-L artificial ponds to conduct a fully factorial field experiment crossing three levels of spatial isolation (30, 120 and 480 m from a source wetland), with absence or presence of fish. The experiment was conducted in a savanna landscape in southeast Brazil. Predators were Redbreast Tilapias (Coptodon rendalli), a widely introduced fish and voracious generalized omnivore, stocked at a density of one individual per mesocosms. Each predator-by-distance treatment was replicated 4 times. The experiment lasted three months and mesocosms were spontaneously colonized by aquatic and semiaquatic insects dispersing in the terrestrial landscape from the source wetland.
Results/Conclusions
The abundance of predatory insects such as dragonflies and backswimmers was reduced by the presence of fish. This might happen because fish are visually-oriented predators preferentially preying upon large prey, which are usually insect predators. In fishless ponds, spatial isolation had a negative effect on the abundance of insect predators, and a strong positive effect on the abundance of midges. Insect predators have smaller population sizes and therefore a low colonization rate in isolated ponds. Midges, on the other hand, are known to reach isolated habitats easily, probably because of their greater population sizes and the possibility of passive dispersal. This can lead to an increase in their abundance through trophic cascades. The impacts of fish presence on community structure were modulated by spatial isolation with the (probably direct) negative effects of fish on predatory insects being stronger in more isolated ponds than in highly connected ones. However, most of the indirect positive effects of fish on non-predatory insects were stronger in less isolated habitats. A strong effect of fish in more isolated ponds is predicted by metacommunity theory. However, our results also show that this pattern might not be true for indirect effects that involve trophic interactions.