Abstract
Human development is increasing pressure on North America's mainly intact boreal forest. We outline the need for a comprehensive synthesis of existing data and for effective scientific tools to support conservation of this biome and of the birds that depend on it. To illustrate how broad collaborations can address these needs, we introduce and report on the Boreal Avian Modelling Project. This is a new partnership involving universities, government, private, and nongovernment groups that was created to develop spatially explicit, predictive models of boreal bird habitat associations across Canada. This initiative is designed to improve our understanding of the influence of environmental factors and human activities on boreal bird species, leading to spatially explicit predictive models of the distribution of avian populations. The intended applications of these models are land use planning and avian conservation across the nearctic boreal forest. In this essay, we present a description of the extensive collection of point count survey data assembled by the Project, and the library of spatial covariates used for modeling. We show how it is possible to account for a number of nuisance variables related to differences in survey protocol among source data sets and make some preliminary suggestions as to how future surveys could be standardized. We present a distance-sampling approach used to convert standardized point count data to density estimates, which we illustrate by providing habitat-specific densities and total population estimates for one species in a part of western Canada. We also illustrate the use of Classification and Regression Trees to develop species niche models from the standardized data. We conclude with a discussion of the need for a monitoring program for boreal birds in Canada, the role of predictive statistical models in developing such a program, and how monitoring could be related to boreal bird conservation through adaptive management.