Challenges in Estimating the Effectiveness of Low Wind Speed Curtailment to Reduce Take of Bats in Hawaii

Low wind speed curtailment (LWSC) has been demonstrated as an effective operational measure to reduce fatalities of migratory tree roosting bat species, including the hoary bat, on the U.S. mainland and Canada at wind farms.... [ view full abstract ]

Low wind speed curtailment (LWSC) has been demonstrated as an effective operational measure to reduce fatalities of migratory tree roosting bat species, including the hoary bat, on the U.S. mainland and Canada at wind farms. On this basis, state and federal wildlife agencies recommend LWSC be implemented at wind farms in Hawaii and incorporated in project Habitat Conservation Plans as a minimization measure to reduce fatalities of the endangered Hawaiian hoary bat. Accurate analysis of post-construction monitoring data requires an understanding of the effectiveness of operational changes that alter fatality risk; however, because bat fatalities in Hawaii are rare events, evaluating the potential benefits of this measure is challenging. Data from the only two commercial wind farms in Hawaii with both pre- and post-LWSC data are used to investigate evidence for the potential effectiveness of LWSC in Hawaii and to explore challenges to this goal in a data-poor environment.

We evaluated post-construction mortality monitoring data at the Kaheawa I and Auwahi wind Projects on Maui for 2-3 years prior to the implementation of LWSC and 1–2 year post-implementation of LWSC to examine evidence for the effectiveness of curtailment. The implementation approach at the two facilities differed, with a seasonal implementation (February 15 – December 15) of 5 – 5.5 meters/second curtailment at the Kaheawa I Wind Project and year-round curtailment at 5 meters/second at the Auwahi Wind Project. We used Dalthorp and Huso’s Evidence of Absence analysis tool to estimate the annual rate of take and its 95 percent confidence interval for the LWSC and non-LWSC periods. Within project results show broad overlap of the confidence intervals among years. Annual pre-LWSC estimates at Kaheawa I were 1.76 (0.00 – 9.03), 7.8 (1.23 – 21.1), and 11.44 (3.37 – 24.7) bats/year. The post-LWSC estimate was 2.53 (0.00 – 12.9) bats/year. At the Auwahi Project, pre-LWSC estimates were 5.83 (0.41 – 18.5) and 9.01 (2.65 – 19.40) bats/year. The Auwahi Wind Project post-LWSC estimates were 3.79 (0.27 – 11.90) and 19.5 (7.98 – 36.5) bats/year. The apparent spike in fatalities in the second year of LWSC at the Auwahi Wind Project suggests that other factors could mask benefits of LWSC and that additional information and greater sample sizes are necessary to understand the potential benefit of LWSC. We address how the occurrence of anomalous fatality events, temporal changes in fatality risk, differences between the projects (e.g., bat population, wind intensity, weather patterns), and the inter-annual variability of data may affect approaches to the analysis of curtailment effectiveness.