Validation of landscape suitability indices for Black Terns (Chlidonias Niger) in the U.S. Great Lakes region

Remotely sensed land cover data enable characterization of avian habitat over large spatial scales, which can be applied to regional conservation planning. Black Terns (Chlidonias Niger) have been declining in the U.S. Great Lakes region for decades. Although causes of the decline are not well understood, habitat loss and degradation are viewed as major barriers to recovery. The Upper Mississippi River and Great Lakes Region Joint Venture, a partnership for bird conservation in the Upper Midwest, developed a landscape suitability index for the wetland-breeding Black Tern based on remotely sensed land cover to choose habitat to enhance or restore. We tested the validity of this index and compared it to an alternative index we constructed from results of field surveys. Over 2 breeding seasons, we surveyed for presence of Black Tern breeding colonies in 158 wetlands across the Great Lakes region and compared our findings with predictions of the existing index. We used our first season of surveys, numerous remotely sensed landscape variables, and random forest classification to build our alternative index. Key predictors of Black Tern colony presence in the alternative index were wetland area, wetland type (emergent vegetation, open water, or combination), and area of wetlands available for foraging within 2 km. The 2 indices performed similarly well, correctly identifying categories of sites most likely to host Black Terns, but even high-suitability sites had <20% predicted occupancy probability. The best use of these and similar indices is in prioritizing restoration projects at a regional scale by identifying sites with potential for use.