Conservation of habitat connectivity has been advocated as an effective approach to biodiversity conservation, especially in view of climate change, prompting initiatives to assess connectivity conservation opportunities at varying spatial scales. Little information is available, however, about how scale influences the comparative performance of different approaches to connectivity assessment. The Washington Wildlife Habitat Connectivity Working Group (WHCWG) has completed wildlife habitat connectivity assessments at two nested scales using two methods of analysis, allowing us to compare results across scales and analysis approaches within the same landscape. We conducted connectivity assessments at the scale of Washington State (447,000 km2), including a buffer of 100 to 200 km into adjoining areas, and at the scale of the Columbia Plateau ecoregion within Washington, including a 25 km buffer into adjoining ecoregions (132,200 km2). At each scale we conducted connectivity analyses using both a focal species approach and an approach based on landscape integrity or “naturalness.” We developed new automation tools to facilitate using both analysis approaches. Efforts to validate our connectivity models are ongoing. Project reports and modeling products are available on-line at waconnected.org.
Results/Conclusions
At the statewide scale we found that both analysis approaches identified broadly consistent habitat connectivity patterns. Focal species analysis revealed connectivity networks that were more amenable to ecological interpretation, while landscape integrity analysis provided seamless results across the entire analysis area. Initial comparisons of these approaches indicated the degree of overlap between focal species connectivity networks and landscape integrity networks varied among habitat-based guilds of focal species. Ninety-seven percent of the connectivity network identified for montane-associated focal species was captured within the network identified by the landscape integrity approach, while 76% of the connectivity network for shrub-steppe associates was captured within the landscape integrity network. At the ecoregional scale, both approaches again yielded broadly concordant patterns that were also consistent with expectations based on the statewide analysis. However, results of the two analysis approaches were more concordant at the larger scale. We found that results diverged at the ecoregional scale in areas where natural features (e.g., cliffs) were important contributors to landscape resistance to wildlife movement, near ecotones, and where agricultural lands and native cover types were interspersed. Examining patterns of concordance and complementarity in results provided insights about relative functionality of modeled linkages. These comparisons also suggested both analysis approaches have practical advantages and disadvantages depending on scale, assessment objectives, intended applications, and available resources.