Modification of landscapes for agricultural production can result in many alterations to aquatic habitats, including increased sedimentation and increased inputs of agrochemicals via agricultural runoff. These alterations degrade habitats used by many amphibians for breeding and larval development. For example, agricultural pesticides can be detected over 90% of the time in streams nationwide and can persist at low concentrations throughout the year. Although many ecotoxicological studies have shown that contaminants commonly present in agricultural runoff have the potential to cause mortality, immunosuppression, or reproductive abnormalities in amphibians, we have a very poor understanding of how exposure to agricultural runoff may affect amphibian population persistence when exposures occur in realistic contexts.

I investigated amphibian responses to runoff at several levels—in laboratory studies, in field studies, and in landscape level surveys. In order to establish whether agricultural herbicides common in runoff have the potential to reduce larval amphibian performance at environmentally realistic levels, I chronically exposed larvae of three species (American toads [Bufo americanus], western chorus frogs [Pseudacris triseriata ], and gray treefrogs [Hyla versicolor]) to end-use herbicide formulations of atrazine, S-metolachlor, and glyphosate in the laboratory. Western chorus frog tadpoles exposed to the glyphosate formulation Roundup WeatherMax® at 572 ppb a.e. (the EPA drinking water standard) showed 80% mortality. In addition, development of American toads was slowed by exposure to Roundup WeatherMax® or Original Max® at 572 ppb a.e.

I then exposed tadpoles to water from agriculturally influenced and reference condition streams in northern Missouri—both in the laboratory and using in situ field enclosures. Under laboratory conditions, exposure to any stream water treatment increased growth and development rates relative to control water. In field enclosures, however, streams with greater agricultural influence offered less consistent larval habitat quality, based on patterns of larval performance. In particular, 70% of tadpoles from one agricultural site died in a pulse of post-exposure mortality. Measured concentrations of herbicides were not tightly correlated with agricultural land use, however.

Lastly, I documented amphibian use of stream habitats in agricultural areas as sites for breeding call activity and reproduction. Based on acoustic call surveys at 41 sites, eight of nine anuran species present in the north-central Missouri assemblage call from riparian corridors. Evidence of reproductive success was detected for five of these species in in-stream surveys at a subset of 26 sites. In occupancy and capture rate models, watershed or local (within 300 m) patterns of land use were generally poor predictors of both calling activity and reproductive success. Several physical habitat characteristics of stream sites, however, were closely related to capture rates of Rana catesbeiana and Rana blairi tadpoles and metamorphs.

Overall, these studies indicate that amphibians use riparian habitats to a great degree in agricultural landscapes. Resulting exposures to agricultural contaminants have the potential to result in reduced survival or fitness, although physical habitat quality at individual sites may be more important for encouraging population persistence.