Kathleen M. Trauth

Influences of Design and Landscape Placement Parameters on Amphibian Abundance in Constructed Wetlands

As natural wetlands disappear, constructed wetlands may play vital roles in amphibian conservation. However, previous investigations have concluded that artificial wetlands do not adequately replace lost wildlife habitat. Nevertheless, constructed wetlands serve as breeding habitat for amphibians where extensive natural wetland loss has occurred. To investigate the roles of engineered wetland features on amphibian abundance, we surveyed 49 constructed wetlands throughout northern Missouri. Cricket frogs (Acris crepitans), bullfrogs (Lithobates catesbeianus), and leopard frogs (Lithobates blairi/sphenocephalus complex) each occurred in over 80% of surveyed wetlands. Salamanders and hylid frogs were rarely encountered. We used an information theoretic approach to examine relationships between individual species and habitat features associated with wetland designs and placements. We found that models incorporating design features of open water ponds best explained abundances of most commonly encountered species. At the placement level, models that included nearby aquatic habitat ranked highest for common species. Salamanders and most hylid frogs responded positively to aquatic vegetative cover but negatively to fish abundance and anthropogenic disturbance-related features in the landscape. Our results indicate that to be effective amphibian conservation tools, constructed wetlands should be fish-free, heavily vegetated, include shallows, and placed within areas of low anthropogenic disturbance.

Testing wetland features to increase amphibian reproductive success and species richness for mitigation and restoration

Aquatic habitat features can directly influence the abundance, species richness, and quality of juvenile amphibians recruited into adult populations. We examined the influences of within-wetland slope, vegetation, and stocked mosquito fish (Gambusia affinis) on amphibian metamorph production and species richness during the first two years post-construction at 18 experimental wetlands in northeast Missouri (U.S.A.) grasslands. We used an information theoretic approach (AICc) to rank regression models representing total amphibian metamorph production, individual amphibian species metamorph production, and larval amphibian species richness. Total amphibian metamorph production was greatest at shallow-sloped, fish-free wetlands during the first year, but shallow-sloped wetlands with high vegetation cover were best the second year. Species richness was negatively associated with fish and positively associated with vegetation in both survey years. Leopard frog (Rana blairi/sphenocephala complex) metamorph quality, based on average metamorph size, was influenced by slope and the number of cohorts in the wetland. However, the tested variables had little influence on the size of American toads (Bufo americanus) or boreal chorus frogs (Pseudacris maculata). Our results indicate that wetlands designed to act as functional reproductive habitat for amphibians should incorporate shallows, high amounts of planted or naturally established vegetation cover, and should be fish-free.

A Vector Approach for Modeling Landscape Corridors and Habitat Connectivity

Landscape connectivity is an important consideration in understanding and reasoning about ecological systems. Two features within a landscape can be viewed as connected whenever a path exists between them. In many applications, the relevance of a potential path is assessed relative to the cost or resistance it presents to traversal. Typically, the least-cost paths between landscape features are used to approximate the potential for connectivity. However, traversal of a landscape between two locations may not necessarily conform to a least-cost path. Moreover, recent research has begun to cast some doubt on the how different types of landscape features may influence movement. Thus, it is important to consider the geographic bounds to movement more broadly. Continuous (i.e., raster) and discrete (i.e., vector) representations of connectivity are commonly used to model the spatial relationships among landscape features. While existing approaches can shed meaningful insights on system topology and connectivity, they are still limited in their ability to represent certain types of movement and are heavily influenced by scale of the areal units and how cost of landscape traversal is derived. In order to better address these issues, this paper proposes a new vector-based approach for delineating the geographic extent of corridors and assessing connectivity among landscape features. The developed approach is applied to evaluate habitat connectivity for salamanders to highlight the benefits of this modeling approach.