Mercury (Hg) contamination from industrial sources is pervasive throughout North America and is recognized by the US Environmental Protection Agency as a health hazard for wildlife and humans. Avian species are commonly used as bioindicators of Hg because they are sensitive to contaminants in the environment and are relatively easy to sample. However, it is important to select the appropriate avian species to use as a bioindicator, which should be directly related to the project objectives. In this study, we tested the utility of two tidal marsh sparrows, Seaside (Ammodramus maritimus) and Saltmarsh (Ammodramus caudacutus) sparrows, as bioindicator species of the extent of Hg contamination in tidal marshes along the Delaware Bay. To determine the possibility of using one or both of these species, we estimated sparrow blood Hg burden in five Delaware watersheds. We found no difference in Hg concentrations between species (F (1,133) < 0.01, P=0.99), but Saltmarsh Sparrows had limited sample size from each site and were, therefore, not appropriate for a Delaware Bay-wide Hg indicator. Seaside Sparrows, however, were abundant and relatively easy to sample in the five watersheds. Seaside Sparrow blood Hg levels ranged from 0.15 to 2.12 ppm, differed among drainages, and were greatest in two drainages distant from the Delaware Bay shoreline (F (4,95) =2.51, P=0.05). Based on a power analysis for Seaside Sparrow blood Hg, we estimated that 16 samples would be necessary to detect differences among sites. Based on these data, we propose that Seaside Sparrows may be used as a tidal marsh Hg bioindicator species given their habitat specificity, relative abundance, widespread distribution in marsh habitats, ease of sampling, and limited variation in blood Hg estimates within a sampling area. In Delaware Bay, Saltmarsh Sparrows may be too rare (making them difficult to sample) to be a viable tidal marsh Hg bioindicator.

Nutria (Myocastor coypus) were introduced to the eastern shore of Chesapeake Bay, USA in the 1940s. They reached peak densities in the late 1990s, causing massive wetland loss. Beginning in 2002, a systematic plan to eradicate nutria from the 1.7M ha Delmarva Peninsula was implemented. Since that time the nutria population has been effectively reduced, and no nutria have been detected since May 2015. A lack of detection does not equate with complete absence. We address the following three questions. (1) What is the expected probability of nutria eradication from the Delmarva Peninsula as of the end of 2020? (2) If the probability of eradication is below the management target of 0.95, how much more surveillance is required? (3) How sensitive is the estimated probability of eradication to varying levels of public surveillance and modelled population growth rates? These questions were addressed by employing a stochastic spatially-explicit surveillance model that uses data in which no nutria were detected to quantify the probability of complete absence (PoA) over the entire Delmarva Peninsula. We applied an analytical framework that decomposes the spatial risk of survivors and data into management zones, and took advantage of low-cost public reporting of nutria sightings. Active surveillance by the eradication program included detector dog and tracker surveys, shoreline surveys, detection with ground and water platforms (with hair snares), and camera traps. Results showed that the PoA increased with time and surveillance from a beginning PoA in May 2015 of  0.01 to a mean of 0.75 at the end of 2020. This indicates that the PoA on the Delmarva was well below the target threshold of 0.95 for declaring eradication success. However, given continued surveillance without detection, a PoA of 0.95 would be achieved by June 2022. This analysis provides an objective mechanism to align the expectations of policy makers, managers and the public on when eradication of nutria from the entire Delmarva Peninsula should be declared successful.

No abstract