Human-induced climate change is increasingly recognized as a fundamental driver of biological processes and patterns, and a threat to the persistence of many species. Recent climate change has already caused shifts in the geographic ranges of myriad species and future climate change is expected to result in even greater redistributions of taxa. As a result, predicting the impact of climate change on future patterns of biodiversity has become a fundamental aspect of conservation planning. Here we use Audubon Christmas Bird Count and North American Breeding Bird Survey data in combination with detailed climate data and projections to estimate the current and future ranges of birds throughout the United States and Canada. Our results address three topics of general interest for broad-scale bird conservation: (1) the impact of climate change on bird diversity in the United States and Canada, (2) identification of areas that are expected to remain important to birds; namely “refugia” that are forecasted to remain climatically suitable from 2000–2080 for individual species and communities, and (3) in-depth analyses of potential climate change impacts on priority species. Before generating predicted responses of birds to future climate change, we assessed the predictive performance of three modeling algorithms (i.e., Maxent, Generalized Additive Models, and Boosted Regression Trees) when confronted with independent observations of birds made in historical time periods and climate spaces. Boosted regression trees performed as well as, or better than, the other algorithms for 512 of 543 species (94%). We then used our Boosted Regression Tree models to forecast species distributions to future time periods based on climate estimates described by the Intergovernmental Panel on Climate Change (IPCC).When assuming that species can—and will— track their climatic niches perfectly through time and across geographic space, we show that winter species richness is expected to increase over much of the continent using each of two distinct modeling approaches. In the summer, richness is expected to decline over much of the conterminous United States and increase in more northern latitudes though the predictive power of summer species distribution models was lower than for winter models. To bracket our optimistic assumptions that birds can—and will—track their climatic niches through space and time we also develop a complementary approach in which we identify in situ refugia. Refugia are areas that we expect to remain climatically suitable for species and communities into the future. When areas also remain suitable across emissions scenarios, they can be considered “no regrets” areas that are likely to remain suitable regardless of the climate conditions that come to pass. We show that the highest numbers of overlapping refugia persist in areas of high current species richness during both winter and summer seasons. When examining how the integrity of existing communities may erode, however, we detected marked variation in community erosion across space, and especially, time. Going forward in time, bird communities in the western United States and southern Canada are much less certain to remain intact than communities in the Midwest and parts of the Great Plains, particularly during the summer. Relatively little of the variation in our refugia predictions could be attributed to emissions scenarios, though the potential benefits of mitigation become clear by 2080 with the low (B2) emissions scenario fostering the persistence of approximately 13.0-13.9% more species than the high (A2) emissions scenario. The present document, Developing a Management Model of the Effects of Future Climate Change on Species: A Tool for the Landscape Conservation Cooperatives, serves as a general technical report describing our mapping methodology and basic summary results. It accompanies Version 1.0 of our Geographic Information Systems (GIS) library containing over 100,000 spatially explicit predictions of the past, present, and future distributions of North America’s birds. We anticipate further expanding our efforts to address several challenges, including focused approaches to characterizing the relative influence of climate and land cover on species distributions, broadening the study area to incorporate species in Central and South America, and more detailed analyses of the spatial and temporal scales at which climate influences species distributions.
