Summary: | Specialization: Conservation Biology Degree: Doctor of Philosophy Abstract: Northwestern North America is predicted to experience some of the world’s greatest human-caused climate change. Understanding the impacts of associated changes will be imperative to the conservation and management of northern-breeding birds. In particular, long distance migrants and aerial insectivores, such as the Olive-sided Flycatcher (Contopus cooperi, OSFL) and the Western Wood-Pewee (C. sordidulus, WEWP), may be impacted disproportionately, in interaction of climate change with additional ex situ population stressors, such as habitat loss and changes to insect populations. Climate-mediated effects on distribution and abundance of these species, both of which have experienced dramatic population declines over the past half century, might be mediated independently of habitat loss by identification of areas of potential climate macrorefugia. To describe contributions of habitat elements on distribution and abundance of these species I generated boosted regression tree models, using data from point counts conducted between 1992 and 2014 at 1049 unique locations in the boreal and hemiboreal zones of northwestern North America. Bootstrap runs of models randomly selected abundances and absences from each location, stratified by number of observations, then built stagewise models from a suite of climate, landcover, topographical and disturbance covariates. I included offsets for unequal detectability of birds by observers in model development. Covariates describing vegetation and landcover were most important in describing abundance, followed by those for climate and topography. Influences of many covariates were non-linear and specific to species, but overall they described habitat that was forested, mid-elevation, topographically complex, and moderate in temperature, precipitation and length of summer season. Relative habitat associations revealed that open forest types, tundra, wet areas, riparian habitat, and old burns were positively ...
|