Forecasting suitable areas for wind turbine occurrence to proactively improve wildlife conservation

Abstract

The United States is rapidly expanding production of renewable energy to meet increased energy demands and reduce greenhouse gas emissions. Wind energy is at the forefront of this transition. A central challenge is understanding the nexus between wind energy development and its capacity for negative effects on wildlife causing population declines and habitat loss. Collaboration among conservationists and developers, early in the planning process, is crucial for minimizing wind-wildlife conflicts. Such collaborations require data showing where wind and wildlife impacts occur. To meet this challenge and inform decision-making, we provide natural resource agencies and stakeholders information regarding where future wind turbines may occur, and the potential affects on natural resource management, including the conservation of priority species and their habitats. We developed a machine learning model predicting suitability of wind turbine occurrence (hereafter, wind turbine suitability) across an eight-state region in the United States, representing some of the richest areas of wind potential. Our model incorporates predictor variables related to infrastructure, land ownership, meteorology, and topography. We additionally created a constraint layer indicating areas where wind would likely not be developed because of zoning, protected lands, and restricted federal agency proximity guidelines. We demonstrate how the predictive wind turbine suitability model informs conservation planning by incorporating animal movement models, relative abundance models coupled with spatial conservation planning software, and population density models for three exemplar, high priority species often affected by wind energy: whooping cranes (Grus americana), golden eagles (Aquila chrysaetos), and lesser prairie-chickens (Tympanuchus pallidicinctus). By merging the wind turbine and biological models, we identified conservation priority areas (i.e., places sharing high suitability for wind turbines and species use), and places where wind expansion could minimally affect these species. We use our “species-wind turbine occurrence relationships” to demonstrate applications, illustrating how forecasting areas of wind turbine suitability promotes wildlife conservation. These relationships inform wind energy siting to reduce negative ecological impacts while promoting environmental and economic viability.