Effects of invasive fish on wetland communities: Linking field data and mechanistic experiments at multiple scales
Invasive species sometimes outnumber native species and can act as drivers of ecosystem change. Understanding and managing their effects will be facilitated by combining mechanistic experiments with field data. Here, we use experiments at the community and ecosystem level in conjunction with cross-sectional field data to examine how invasive fish alter wetland communities in California. We first conducted community level experiments in outdoor mesocosms to determine how fish restructure native communities of zooplankton, invertebrates and pond-breeding amphibians. We then assessed whether such results ‘scale-up’ to a real ecosystem by using a ‘before-after control-impact’ design involving the controlled introduction of fish to a natural wetland. Lastly, we analyze a field data set of 171 wetlands to determine if experimental results are consistent with broad patterns in natural ecosystems.
In mesocosms, nonnative fish strongly reduced the survival of native amphibians and altered invertebrate communities. Importantly, however, such effects were mediated by differences in habitat complexity and the availability of alternative prey. Consistent with the smaller scale studies, nonnative fish reproduced rapidly and altered the abundance and composition of zooplankton and macroinvertebrates in the ecosystem level experiment. At the broadest scale of field surveys, the presence of fish strongly reduced the probability of occupancy of three native amphibians, one of which is federally endangered. Taken together, these results support the growing evidence that nonnative fish in wetlands have potential to reshape community structure and negatively affect native species of conservation concern. Our experimental results also indicate that variation in environmental characteristics will mediate the strength of invasive fish effects, making it important for land managers to recognize that invasion outcomes will vary across discrete ecosystems.