Assessing the use of long-term lek survey data to evaluate the effect of landscape characteristics and wind facilities on sharp-tailed grouse lek dynamics in North Dakota and South Dakota

Abstract

The contribution of renewable energy to meet worldwide demand continues to grow. In the United States, wind energy is one of the fastest growing renewable energy sectors. Throughout the Great Plains of the United States, wind facilities often are placed in open landscapes of high-elevation grasslands, and those same habitats support sharp-tailed grouse ( Tympanuchus phasianellus ), a resident gamebird species. To assess the feasibility of using independently derived, long-term datasets gathered in North Dakota and South Dakota to determine whether wind facilities affected lek metrics, the U.S. Geological Survey obtained six datasets and identified 37 study sites, 9 of which contained wind turbines at varying densities. The association between explanatory variables that described geographic, landscape, and climatic attributes with two primary response metrics that described lekking activity within study sites—lek density (leks per square kilometer) and mean number of males per lek—was examined. The explanatory variables included number of turbines, geographic location, elevation, land-cover attributes available from satellite-derived land-cover data, soil moisture, precipitation, and temperature. Sampling units consisted of township-sized blocks, and lek information came from roadside surveys. Low sample sizes of constructed wind facilities available at the time of analysis did not lend itself to advanced statistical techniques, such as employing a rigorous design structure or assessing accuracy on landscape, geographic, or climatic variables. Given the quality of the data, the estimates obtained for lek density and mean number of males per lek should be considered approximations; however, these estimates have value in designing future studies, such as providing estimates for power analyses to determine sufficient sample size. No strong associations were found between the included explanatory variables and response variables (when these variables were measured as described in this report for township-sized blocks). The strongest association was that lek density and mean number of males per lek increased from South Dakota to North Dakota. Owing to the highly unbalanced distribution of turbine and nonturbine study sites across the study area, the analysis with wind turbines was inconclusive. The constraints under which the analysis can be used and the limitations of the independently derived datasets in attempted applications are discussed.