PS 72-3 - Poisson regression for count data: Application to bird and bat mortality at the wind farm La Venta II, Oaxaca, Mexico

Friday, August 12, 2011
Exhibit Hall 3, Austin Convention Center
Rafael Villegas-Patraca, Clara Perez and Marissa Mora, Environmental and sustentability, Instituto de EcologĂ­a AC, Xalapa, Mexico
Background/Question/Methods

Wind power production is a fast growing source of renewable energy worldwide. Wind energy development can have ecological impacts, such as the increase of bird and bat fatalities through collision with the structures built at wind energy installations (e.g. turbines, rotors, towers). The wind farm La Venta II is located on one of the most important bird migration routes; therefore, the effects of the wind energy facility on bird populations and diversity are expected during the migration season. Birds fly at lower altitudes due to unfavorable weather conditions, especially at night, which is likely to increase the risk of collision. The number of bat fatalities has increased in the wind farm, but the causes of collisions are unknown. Identifying the factors associated with bird and bat turbine-related mortality is crucial to understand the effects of wind farms on their populations. In this study, counts of bird and bat collisions were obtained from field carcass searches and analyzed using a generalized linear model (GLM). Poisson regression belongs to the GLM family and was therefore used to model non-normally distributed count data and to identify the factors most related with the counts of bird and bat collisions.

Results/Conclusions

The results indicated the vegetation and the wind are the factors most related with bird and bat mortality in the wind farm La Venta II. The results also allowed identifying the turbines and the section of the wind farm with high number of collisions. This information will assist decision support in terms of management and mitigation practices at this wind energy facility located in a semi-humid, tropical environment.

Copyright © . All rights reserved.
Banner photo by Flickr user greg westfall.