Responses of a wetland ecosystem to the controlled introduction of invasive fish

Abstract

Summary The ecological consequences of species invasions can vary in time and space, complicating efforts to generalise invader impacts across ecosystems. This challenge is particularly relevant when using small‐scale experiments to derive predictions for freshwater ecosystems. In this study, our aims were to document the effects of a controlled fish introduction within an ecosystem‐scale experiment and to test possible factors driving variation in invasion outcomes. We measured community and ecosystem responses to the introduction of western mosquitofish (Gambusia affinis: Poeciliidae) using a ‘Before‐After‐Control‐Impact’ design within a large, experimentally divided natural wetland in California, USA. We then used a replicated outdoor mesocosm experiment to address how two factors that vary widely in natural wetlands – habitat complexity and alternative prey availability – mediate the effects of mosquitofish on native wetland amphibians. In the natural wetland, mosquitofish increased in population size by ˜90‐fold over the course of a single summer. Mosquitofish introduction was associated with a 50% decrease in macroinvertebrate density and a 90% decrease in zooplankton abundance relative to a fishless control treatment. We observed no effects of mosquitofish on the abundance or total biomass of two native pond‐breeding amphibians – the Pacific chorus frog (Pseudacris regilla: Hylidae) and California newt (Taricha torosa: Salamandridae) – likely because more preferable alternative prey were abundant, vegetation provided refuges from predation, and the mosquitofish introduction occurred after amphibian larval stages were most susceptible to predation. Surprisingly, mosquitofish were also associated with a 50% decrease in both relative phytoplankton fluorescence and total phosphorus, and a sharp increase in N:P ratios in the water column, possibly due to the assimilation of fish biomass acting as a nutrient sink. In contrast to our ecosystem experiment, mosquitofish consumed native amphibians and reduced their growth rates in outdoor mesocosms. The strength of predation within the smaller scale venue, however, varied with the availability of alternative prey (i.e. zooplankton), and the complexity of the habitat (i.e. presence of aquatic macrophytes). Our mesocosm results support the hypothesis that alternative prey and habitat complexity facilitated coexistence between invasive mosquitofish and native amphibian larvae in our ecosystem experiment. Our findings highlight the potential for invasive fish to drive rapid shifts in freshwater ecosystems while also emphasising the roles of environmental characteristics in mediating whether native and non‐native species will coexist.