Predicting the spatial distribution of wintering golden eagles to inform full annual cycle conservation in western North America.

Wildlife conservation strategies focused on one season or population segment may fail to adequately protect populations, especially when a species' habitat preferences vary among seasons, age-classes, geographic regions, or other factors. Conservation of golden eagles (Aquila chrysaetos) is an example of such a complex scenario, in which the distribution, habitat use, and migratory strategies of this species of conservation concern vary by age-class, reproductive status, region, and season. Nonetheless, research aimed at mapping priority use areas to inform management of golden eagles in western North America has typically focused on territory-holding adults during the breeding period, largely to the exclusion of other seasons and life-history groups. To support population-wide conservation planning across the full annual cycle for golden eagles, we developed a distribution model for individuals in a season not typically evaluated-winter-and in an area of the interior western U.S. that is a high priority for conservation of the species. We used a large GPS-telemetry dataset and library of environmental variables to develop a machine-learning model to predict spatial variation in the relative intensity of use by golden eagles during winter in Wyoming, USA, and surrounding ecoregions. Based on a rigorous series of evaluations including cross-validation, withheld and independent data, our winter-season model accurately predicted spatial variation in intensity of use by multiple age- and life-history groups of eagles not associated with nesting territories (i.e., all age classes of long-distance migrants, and resident non-adults and adult "floaters", and movements of adult territory holders and their offspring outside their breeding territories). Important predictors in the model were wind and uplift (40.2% contribution), vegetation and landcover (27.9%), topography (14%), climate and weather (9.4%), and ecoregion (8.7%). Predicted areas of high-use winter habitat had relatively low spatial overlap with nesting habitat, suggesting a conservation strategy targeting high-use areas for one season would capture as much as half and as little as one quarter of high-use areas for the other season. The majority of predicted high-use habitat (top 10% quantile) occurred on private lands (55%); lands managed by states and the Bureau of Land Management (BLM) had a lower amount (33%), but higher concentration of high-use habitat than expected for their area (1.5-1.6x). These results will enable those involved in conservation and management of golden eagles in our study region to incorporate spatial prioritization of wintering habitat into their existing regulatory processes, land-use planning tasks, and conservation actions.

Modeling spatial variation in density of golden eagle nest sites in the western United States.

In order to contribute to conservation planning efforts for golden eagles (Aquila chrysaetos) in the western U.S., we developed nest site models using >6,500 nest site locations throughout a >3,483,000 km2 area of the western U.S. We developed models for twelve discrete modeling regions, and estimated relative density of nest sites for each region. Cross-validation showed that, in general, models accurately estimated relative nest site densities within regions and sub-regions. Areas estimated to have the highest densities of breeding golden eagles had from 132-2,660 times greater densities compared to the lowest density areas. Observed nest site densities were very similar to those reported from published studies. Large extents of each modeling region consisted of low predicted nest site density, while a small percentage of each modeling region contained disproportionately high nest site density. For example, we estimated that areas with relative nest density values <0.3 represented from 62.8-97.8% ([Formula: see text] = 82.5%) of each modeling area, and those areas contained from 14.7-30.0% ([Formula: see text] = 22.1%) of the nest sites. In contrast, areas with relative nest density values >0.5 represented from 1.0-12.8% ([Formula: see text] = 6.3%) of modeling areas, and those areas contained from 47.7-66.9% ([Formula: see text] = 57.3%) of the nest sites. Our findings have direct application to: 1) large-scale conservation planning efforts, 2) risk analyses for land-use proposals such as recreational trails or wind power development, and 3) identifying mitigation areas to offset the impacts of human disturbance.

Detection of Influenza A viruses at migratory bird stopover sites in Michigan, USA.

Introduction: Influenza A viruses have the potential to cause devastating illness in humans and domestic poultry. Wild birds are the natural reservoirs of Influenza A viruses and migratory birds are implicated in their global dissemination. High concentrations of this virus are excreted in the faeces of infected birds and faecal contamination of shared aquatic habitats can lead to indirect transmission among birds via the faecal-oral route. The role of migratory birds in the spread of avian influenza has led to large-scale surveillance efforts of circulating avian influenza viruses through direct sampling of live and dead wild birds. Environmental monitoring of bird habitats using molecular detection methods may provide additional information on the persistence of influenza virus at migratory stopover sites distributed across large spatial scales. Materials and methods: In the current study, faecal and water samples were collected at migratory stopover sites and evaluated for Influenza A by real-time quantitative reverse transcriptase PCR. Results and Discussion: This study found that Influenza A was detected at 53% of the evaluated stopover sites, and 7% and 4.8% of the faecal and water samples, respectively, tested positive for Influenza A virus. Conclusion: Environmental monitoring detected Influenza A at stopover sites used by migratory birds.