Building better wetlands for amphibians: Investigating the roles of engineered wetland features and mosquitofish (Gambusia affinis) on amphibian abundance and reproductive success

Compensatory wetland mitigation efforts have been broadly criticized for failing to replace lost ecological functions, even when net gains in wetland area are achieved. Of particular concern is the suitability of constructed wetlands as habitat for vulnerable taxa. Amphibians are vital components of wetland ecosystems, yet one third of all species are threatened. Habitat loss and alteration are primary reasons for these declines so it is imperative that constructed wetlands significantly contribute to amphibian conservation. Nevertheless, few studies have sought to understand how altering engineered wetland features influences amphibian abundance and reproductive success. Results from such studies are necessary to guide wetland planners and managers in their efforts to construct and restore wetlands that aid amphibian conservation. The aim of my dissertation research was to investigate features that can be altered by wetland planners and managers to enhance wetland suitability for amphibians. I ask two basic wetland planning questions that are essential to providing appropriate habitat: 1) how do aquatic habitat features influence reproductive success and species richness; and, 2) how do features of the core terrestrial habitat influence individual species abundances. The first question addresses how the wetland is designed, whereas the second question addresses where the wetland is placed in the landscape. First, I surveyed amphibian populations in 49 existing constructed wetlands throughout northern Missouri to investigate influences of both design and placement features on amphibian abundance. Design features typical of open water ponds best explained abundances of cricket frogs (Acris crepitans ), bullfrogs (Lithobates catesbeianus), and green frogs (L. clamitans); all commonly captured species. At the placement level, models that included nearby aquatic habitat ranked highest for common species. Salamanders and most hylid frogs were rarely captured and responded positively to aquatic vegetation, but negatively to fish and anthropogenic disturbance-related features in the terrestrial habitat. Next, I conducted field studies at 18 experimental constructed wetlands to test design feature effects on amphibian metamorph production and species richness. I examined the effects of within-wetland slope, vegetation, and introduced mosquitofish (Gambusia affinis). Mosquitofish significantly reduced metamorph production in both survey years. Vegetation cover had significant effects on production the second year. Regression models revealed that total metamorph production was greatest at shallow-sloped, fish-free wetlands during the first year, but shallow-sloped wetlands with high vegetation amounts were best the second year. Species richness was negatively associated with fish and positively associated with vegetation in both years. Finally, I focused on the role of predators, particularly mosquitofish, in shaping amphibian and invertebrate communities in constructed wetlands. I also investigated the role of predators in increasing the severity of sub-lethal tail injuries to tadpoles, and the role of vegetation in attenuating these injuries. My results indicate that boreal chorus frogs (Pseudacris maculata) are intolerant to large populations of aquatic predators, including mosquitofish. Gray treefrogs (Hyla versicolor/chrysoscelis complex) were also sensitive to mosquitofish based on larval abundance assessments. Additionally, mosquitofish increased the severity of larval ranid tail injuries, and they reduced both invertebrate abundance and richness. Fish removal increased chorus frog reproduction, bolstered invertebrate populations, and reduced the severity of tadpole tail injuries. Vegetation was important for increasing invertebrate taxa richness, but did not provide larval ranids adequate refuge from mosquitofish based on my tail injury analyses. The results of my research indicate that constructed wetlands can be effective for amphibian conservation if appropriate habitats for target species are provided at both the design and placement levels. Fish-free, heavily vegetated, shallow-sloped wetlands, placed in landscapes with low anthropogenic disturbance, appear to provide the best habitat for the most uncommon amphibian species captured in my surveys. Mosquitofish should not be introduced into wetlands because of their potential to negatively impact native amphibian and invertebrate communities. Wetlands that attract breeding native salamanders, whose larvae also feed on mosquitoes, are a better alternative and likely more effective for amphibian conservation. Furthermore, designing some wetlands to dry occasionally will reduce populations of fish and predatory aquatic invertebrates that prevent colonization and lower the reproductive success of some amphibian species