Surveys of amphibians, abnormalities, pathogens, triazines, breeding-site characteristics, and reptiles in five areas managed by the National Park Service and the U.S. Fish and Wildlife Service in the Upper Midwest, 2002-2007

Abstract/Summary

Our goals were to assess the statuses of amphibian populations relative to key environmental factors in Voyageurs National Park (VOYA), the St. Croix National Scenic Riverway (SACN), the Upper Mississippi River National Wildlife and Fish Refuge (UMR), and the Neal Smith National Wildlife Refuge (NS)and to inventory amphibian and reptile species in Voyageurs National Park (VOYA), the St. Croix National Scenic Riverway (SACN), and the Mississippi National River and Recreation Area (MISS).

We 1) conducted visual-encounter and call surveys for amphibians at potential breeding sites in VOYA, the SACN, the UMR, and the NS, and targeted searches for amphibians in the MISS, 2) conducted visual-encounter surveys for reptiles, especially turtles, in the vicinity of amphibian breeding sites in VOYA and the SACN, and to a lesser extent, the MISS, 3) analyzed amphibian survey data across management units and years using information-theoretic models to estimate occupancy and the probability of detection in relation to site-specific and sampling-specific characteristics, 4) sampled water from amphibian breeding sites in VOYA, the SACN, the UMR, and the NS and analyzed them for the presence of triazine herbicides, 5) measured gross physical deformities among amphibian metamorphs at breeding sites in VOYA, the SACN, the UMR, and the NS, and 6) sampled amphibian populations in VOYA, the SACN, and the UMR for the presence of pathogens, parasites, and symptoms of disease.

A summary of our principal results is as follows:

• We observed most of the amphibian and reptile species we expected to find in each management unit. We found 84%, 75%, 72%, 71%, and 31% of the amphibian species potentially present at SACN, NS, UMR, VOYA, and MISS (where we targeted fewer species and devoted less resources). Similarly, we observed 43% of the possible snakes and 67% of the possible lizards in the SACN and 75% of the possible snakes in VOYA. No lizards were likely in VOYA, and we did not observe any.

• Estimates of occupancy and probabilities of detection varied widely for amphibian species sampled during the day across VOYA, the SACN, the UMR, and the NS. Occupancy estimates ranged from https://0.29 (Pseudacris triseriata/maculata ) to https://0.79 (R. clamitans) with standard errors for occupancy from https://0.05 (R. clamitans) to 0.9 (Bufo americanus ). Ranges for the probabilities of detection were from https://0-0.34 (Bufo americanus ) to 0.18-1.0 (R. clamitans) with standard error ranges for detection probabilities from https://0-0.13 (R. clamitans) to 0-157.8 (Pseudacris crucifer ). Sampling-specific covariates that were influential in the best-fitted occupancy models were time, observer, and date. Similarly, site-specific covariates were management unit and habitat type.

• Of the 523 water samples we collected from amphibian breeding sites across management units and analyzed for the presence of triazines, 30.8% were below the detection limit of https://0.050 ppb, mostly in the SACN and VOYA. Fifty-five percent of the values above the detection limit also were above 0.1 ppb, a threshold shown to have caused deformed gonads in frogs in the laboratory in some studies; most of these were samples from the NS and UMR. The relative triazine concentrations across management units followed their proximity to extensive areas of row crops, where triazines are applied in large quantities. Triazine concentrations at breeding sites in the floodplain of the Mississippi River in the UMR fluctuated in association with the hydrograph of the Mississippi River.

• We observed gross physical deformities in less than 5% of the metamorphs sampled in VOYA, the SACN, the UMR, and the NS. The percentage of deformed metamorphs across all species for VOYA, the UMR, and the SACN from 2002-2005 ranged from 0.5% to 4.5%, and means for each of these management units across years were 1.5, 2.4, and 1.4%, respectively.

• We observed the potentially deadly chytrid fungus, Batrachochytrium dendrobatidis, on R. clamitans and R. septentrionalis from practically all sites where we sampled them in VOYA (88%) and the SACN (100%) during 2006, but at fewer sites on R. pipiens in the UMR (25%). The pattern was similar for these species in 2007, except we also sampled R. clamitans in the UMR. We observed B. dendrobatidis on R. clamitans and R. septentrionalis at most sites where we sampled them in VOYA (71%) and the SACN (67%), and the majority of sites where we sampled R. clamitans in the UMR (80%). In contrast, we observed B. dendrobatidis on R. pipiens at only 13% of the sites where we sampled them in VOYA, the SACN, and the UMR.

• Most individual, apparently healthy amphibians we submitted for health screening were afflicted with parasites of one sort or another, ranging from flukes to trematodes. Ribeiroia ondatrae, a parasitic trematode we observed, is considered a primary inducer of gross physical abnormalities in amphibians.