U.S. imperiled species are most vulnerable to habitat loss on private lands

Abstract

To stem the ongoing loss of biodiversity, there is an urgent need to distinguish effective and ineffective approaches to protecting species and their habitats. Conservation laws may be strong on paper but ineffective in practice, or vary in effectiveness across different contexts, such as different land ownership and management settings. Using Google Earth Engine and 30 years of Landsat satellite images, we quantify annual habitat change for 24 species on the U.S. Endangered Species List and IUCN Red List across different categories of land ownership, such as federal, state, and private. We show that imperiled species lost very little habitat on federal lands (3.6%), while losses on non-protected private lands (8.1%) were twice as great. Patterns of loss suggest that listing species under the Endangered Species Act was one mechanism limiting habitat loss, and that the law was most effective on federal lands. These results emphasize the importance of federal lands for protecting habitat for imperiled species, but also highlight the need to improve habitat protection on private lands for long-term conservation. SIGNIFICANCE STATEMENTWe provide the first large-scale, long-term analysis of imperiled species habitat loss under the U.S. Endangered Species Act. Scientists have historically lacked the tools to evaluate patterns of habitat loss at large spatial and temporal scales, and instead focused on small-scale case studies that are insufficient for informing national policy debates. Our results substantiate the importance of federal protections in the United States, indicating that the ESA has successfully reduced habitat loss for imperiled species. They also identify gaps in protection that compromise wildlife conservation. Given the ongoing biodiversity crises, this information is urgently required to objectively assess the effectiveness of national conservation laws and formulate data-driven biodiversity conservation policies and approaches.