Abstract. Monitoring species in National Parks facilitates inference regarding effects of climate change on population dynamics because parks are relatively unaffected by other forms of anthropogenic disturbance. Even at early points in a monitoring program, identifying climate covariates of population density can suggest vulnerabilities to future change. Monitoring landbird populations in parks during the breeding season brings the added benefit of allowing a comparative approach to inference across a large suite of species with diverse requirements. For example, comparing resident and migratory species that vary in exposure to non-park habitats can reveal the relative importance of park effects, such as those related to local climate. We monitored landbirds using breeding-season point-count data collected during 2005-2014 in three wilderness areas of the Pacific Northwest (Mount Rainier, North Cascades, and Olympic National Parks). For 39 species, we estimated recent trends in population density while accounting for individual detection probability using Bayesian hierarchical N-mixture models. Our analyses integrated several recent developments in N-mixture modeling, incorporating interval and distance sampling to estimate distinct components of detection probability while also accommodating count intervals of varying duration, annual variation in the length and number of point-count transects, spatial autocorrelation, random effects, and covariates of detection and density. As covariates of density, we considered metrics of precipitation and temperature hypothesized to affect breeding success. We also considered effects of park and elevational stratum on trend. Regardless of model structure, we estimated stable or increasing densities during 2005-2014 for most populations. Mean trends across species were positive for migrants in every park and for residents in one park. A recent snowfall deficit in this region might have contributed to the positive trend, because population density varied inversely with precipitation-as-snow for both migrants and residents. Densities varied directly but much more weakly with mean spring temperature. Our approach exemplifies an analytical framework for estimating trends from point-count data, and for assessing the role of climatic and other spatiotemporal variables in driving those trends. Understanding population trends and the factors that drive them is critical for adaptive management and resource stewardship in the context of climate change.

Dispersal of whitebark pine (Pinus albicaulis Engelm.), a keystone species of many highelevation ecosystems in western North America, depends on Clark's nutcracker (Nucifraga columbiana Wilson), a seed-caching bird with an affinity for whitebark seeds. To the extent that this dependence is mutual, declines in whitebark seed production could cause declines in nutcracker abundance. Whitebark pine is in decline across much of its range due to interacting stressors, including the non-native pathogen white pine blister rust (Cronartium ribicola J. C. Fisch.). We used avian point-count data and tree surveys from four national park units to investigate whether trends in whitebark pine can explain trends in Clark's nutcracker. Spatial trends were modeled using recent data from two parks, while temporal trends were modeled using longer time-series of nutcracker and whitebark data from two additional parks. To assess the potential dependence of nutcrackers on whitebark, we linked a model of nutcracker density (accounting for detection probability) with a model of whitebark trends, using a Bayesian framework to translate uncertainty in whitebark metrics to uncertainty in nutcracker density. In Mount Rainier National Park, temporal models showed dramatic declines in nutcracker density concurrent with significant increases in whitebark crown mortality and trees infected with white pine blister rust.

No abstract

No abstract