Assessing the Impacts of Agricultural Land Use on Wood Frog (Lithobates sylvaticus) Presence and Health in Central Saskatchewan

Globally, amphibian populations are declining in response to many factors including habitat loss and degradation, environmental contamination, invasive species, emerging diseases, climate change, and overexploitation. Amphibians are particularly susceptible to habitat loss and contaminants because of their diverse habitat requirements and unique life histories and ecologies. The Canadian Prairie Pothole Region (PPR) is home to several amphibian species, but they are threatened by large-scale conversion of habitat to agriculture. One of the more common amphibians in this region is the wood frog (Lithobates sylvaticus), a wide-ranging species that occupies a variety of ecosystems, from forests to prairies to tundra. This makes it an ideal model species to compare results in ecological and toxicological studies. Given anecdotal reports of their abundance in the PPR and simultaneous exposure to a number of anthropogenic stressors, I investigated the effects of environmental variables across multiple scales on wood frog presence and tadpole and metamorph health in central Saskatchewan. I visited wetlands at five sites near Saskatoon, SK along a gradient of agricultural intensity with two grassland sites (Allan and St. Denis) and three cropland sites (Burr, Colonsay, and Humboldt). I collected data on water quality including nutrients and pesticides, wetland habitat, and surrounding land use and used environmental DNA (eDNA) to detect the presence of ranavirus and wood frogs. To assess the effects of these variables on both wood frog presence and health (condition, mass, and neutrophil to lymphocyte (N:L) ratios), I used boosted regression trees, a relatively novel but growing modelling technique in the ecological sciences. Wood frogs were present in both grassland and cropland sites. eDNA was more successful at detecting wood frogs in wetlands compared to traditional survey methods – visual encounter surveys and dipnetting. However, for both wood frogs and ranavirus, detection varied seasonally with greater success in the summer than in the spring. Several environmental variables influenced wood frog presence, the most influential being those associated with wetland productivity, vegetation buffer width, and proportion of the surrounding landscape that is comprised of other waterbodies. Wood frog presence was positively associated with higher dissolved phosphorus (≥ 0.4 mg/L), a range of dissolved nitrogen (0.1 to 0.2 mg/L), lower chlorophyll a (≤ 15 µg/L), wider vegetation buffers (≥ 10 m), and more water on the landscape (≥ 0.25). Wood frog detection was also positively influenced by lower total dissolved solid values (<1000 mg/L TDS) and negatively influenced by very low catch-per-unit-effort values (< 0.01 CPUE). In contrast pesticides and ranavirus were poor predictors of wood frog presence, suggesting either the inability to avoid these stressors or resilience towards them. These results are consistent with previous studies regarding the importance of vegetation buffers and land use and cover, but highlight the effects of environmental factors at multiple scales on wood frog presence. Tadpoles completed their larval development in both grassland and cropland sites. Body condition and N:L ratios were affected only by Gosner stage (GS); both were stable or slightly declined until metamorphic climax (GS 41-42), after which they declined greatly. There were, however, effects of environmental variables on tadpole and metamorph body mass. Besides Gosner stage, influential variables included total dissolved solids, proportion of pesticides detected, ammonia, and wetland surface area. Total dissolved solids and pesticide detection had marked negative effects on body mass at and above 600-700 mg/L TDS and 0.01 proportion of pesticides detected. Wetlands in the PPR are naturally saline, but the ionic composition is unique in that it is primarily sulfate ions and little research has investigated the effects of sulfates on tadpoles. Pesticide concentrations were lower than most lethal doses reported in the literature, but in the field setting where these tadpoles are simultaneously exposed to multiple stressors, it appears to have an impact on body mass. These results again emphasize the importance of multiple, interacting stressors on tadpole health as reduced mass at metamorphosis can have negative implications for survival and fecundity as an adult. I also observed unique neutrophils that warrant further research in wood frog hematology, especially with tadpoles at metamorphic climax. It is clear that wood frogs can survive in these agricultural landscapes, but in order to maintain populations we need to monitor habitat characteristics at the water quality, wetland, and landscape scales. Agricultural activity can alter wood frog habitat at all of these scales, and all have implications for wood frog occupancy. Contaminant exposure may also affect life stages of the wood frog differently. Adult presence was not greatly influenced by pesticides, but tadpole and metamorph size was reduced which may have individual- and potentially population-level impacts. The results of these studies contribute new information to our understanding of wood frog ecology in a unique part of its North American range