Water and bed-sediment samples were collected by the U.S. Geological Survey (USGS) in 2009 and 2010 from 11 sites within California and 18 sites total in Colorado, Georgia, Idaho, Louisiana, Maine, and Oregon, and were analyzed for a suite of pesticides by the USGS. Water samples and bed-sediment samples were collected from perennial or seasonal ponds located in amphibian habitats in conjunction with research conducted by the USGS Amphibian Research and Monitoring Initiative and the USGS Toxic Substances Hydrology Program. Sites selected for this study in three of the states (California, Colorado, and Orgeon) have no direct pesticide application and are considered undeveloped and remote. Sites selected in Georgia, Idaho, Louisiana, and Maine were in close proximity to either agricultural or suburban areas. Water and sediment samples were collected once in 2009 during amphibian breeding seasons. In 2010, water samples were collected twice. The first sampling event coincided with the beginning of the frog breeding season for the species of interest, and the second event occurred 10–12 weeks later when pesticides were being applied to the surrounding areas. Additionally, water was collected during each sampling event to measure dissolved organic carbon, nutrients, and the fungus, Batrachochytrium dendrobatidis, which has been linked to amphibian declines worldwide. Bed-sediment samples were collected once during the beginning of the frog breeding season, when the amphibians are thought to be most at risk to pesticides. Results of this study are reported for the following two geographic scales: (1) for a national scale, by using data from the 29 sites that were sampled from seven states, and (2) for California, by using data from the 11 sampled sites in that state.<br />
<br />
Water samples were analyzed for 96 pesticides by using gas chromatography/mass spectrometry. A total of 24 pesticides were detected in one or more of the 54 water samples, including 7 fungicides, 10 herbicides, 4 insecticides, 1 synergist, and 2 pesticide degradates. On a national scale, aminomethylphosphonic acid (AMPA), the primary degradate of the herbicide glyphosate, which is the active ingredient in Roundup®, was the most frequently detected pesticide in water (16 of 54 samples) followed by glyphosate (8 of 54 samples). The maximum number of pesticides observed at a single site was nine compounds in a water sample from a site in Louisiana. The maximum concentration of a pesticide or degradate observed in water was 2,880 nanograms per liter of clomazone (a herbicide) at a site in Louisiana. In California, a total of eight pesticides were detected among all of the low and high elevation sites; AMPA was the most frequently detected pesticide, but glyphosate was detected at the highest concentrations (1.1 micrograms per liter).<br />
<br />
Bed-sediment samples were analyzed for 94 pesticides by using accelerated solvent extraction, gel permeation chromatography for sulfur removal, and carbon/alumina stacked solid-phase extraction cartridges to remove interfering sediment matrices. In bed sediment, 22 pesticides were detected in one or more of the samples, including 9 fungicides, 3 pyrethroid insecticides, p,p’-dichlorodiphenyltrichloroethane (p,p’-DDT) and its major degradates, as well as several herbicides. Pyraclostrobin, a strobilurin fungicide, and bifenthrin, a pyrethroid insecticide, were detected most frequently. Maximum pesticide concentrations ranged from less than their respective method detection limits to 1,380 micrograms per kilogram (tebuconazole in California). The number of pesticides detected in samples from each site ranged from zero to six compounds. The sites with the greatest number of pesticides were in Maine and Oregon with six pesticides detected in one sample from each state, followed by Georgia with four pesticides in one sample. For California, a total of 10 pesticides were detected among all sites, and 4 pesticides were detected at both low and high elevation sites; tebuconazole and pyraclostrobin were the two most frequently detected pesticides in California. For the other six selected states, the most frequently detected pesticides in bed sediment were pyraclostrobin (detected in 17 of 42 samples), bifenthrin (detected in 14 of 42 samples), and tebuconazole (detected in 10 of 42 samples).<br />
<br />
The fungus, Batrachochytrium dendrobatidis (Bd), was detected in water samples in sites from four of the seven states during 2009 and 2010, and the number of zoospore equivalents per liter of water in samples where Bd was detected ranged from 1.6 to 343. Bd was not detected in water samples from sites in Georgia, Louisiana, and Oregon.

Disturbances are often expected to magnify effects of disease, but these effects may depend on the ecology, behavior, and life history of both hosts and pathogens. In many ecosystems, wildfire is the dominant natural disturbance and thus could directly or indirectly affect dynamics of many diseases. To determine how probability of infection by the aquatic fungus Batrachochytrium dendrobatidis (Bd) varies relative to habitat use by individuals, wildfire, and host characteristics, we sampled 404 boreal toads (Anaxyrus boreas boreas) across Glacier National Park, Montana (USA). Bd causes chytridiomycosis, an emerging infectious disease linked with widespread amphibian declines, including the boreal toad. Probability of infection was similar for females and the combined group of males and juveniles. However, only 9% of terrestrial toads were infected compared to >30% of aquatic toads, and toads captured in recently burned areas were half as likely to be infected as toads in unburned areas. We suspect these large differences in infection reflect habitat choices by individuals that affect pathogen exposure and persistence, especially in burned forests where warm, arid conditions could limit Bd growth. Our results show that natural disturbances such as wildfire and the resulting diverse habitats can influence infection across large landscapes, potentially maintaining local refuges and host behaviors that facilitate evolution of disease resistance.

1) Epidemiological studies are crucial for understanding the distribution and dynamics of emerging infectious diseases. To accurately assess infection states in wild populations, researchers need to account for observational uncertainty. We focus on two sources of uncertainty when estimating epidemiological parameters: non-detection of infection in sampled individuals and sampling error when quantifying infection intensity for infected individuals. <br />
2) We developed new analytical methods to simultaneously estimate prevalence and the distribution of infection intensities based on repeated sampling of individuals in the wild. The methods are an extension of those used for occupancy estimation and address both sources of observation error. At the same time, we account for heterogeneity in detection probability that results from individual variation in infection intensity. <br />
3) We use two estimation approaches to account for detection. The first is to use the complete likelihood in a hierarchical Bayesian model, fit using Markov chain Monte Carlo sampling. The second is to estimate the detection relationship using a mark-recapture abundance estimator and uses those results to calculate weighted estimates for prevalence and mean infection intensities.<br />
4) We use data from a field survey of Batrachochytrium dendrobatidis (Bd) in Illinois amphibians to test these methods. We show that detection probability using quantitative PCR is strongly related to infection intensity, measured in zoospore equivalents. Sites in the study varied greatly in estimated prevalence and to a lesser extent in mean infection intensities of infected individuals. We did not find evidence of a relationship of snout-vent-length to infection intensity or prevalence. Naïve estimates of prevalence that do not account for detection were smaller than estimates for either of our methods, which yielded similar prevalence values for most sites. <br />
5) Uncertainty when assessing disease state is a characteristic of most diagnostic tests. The estimators presented here account for this uncertainty and thus, can improve accuracy when assessing the relationship of ecological factors to prevalence and infection intensity.