Bats in a tropical wind farm: species composition and importance of the spatial attributes of vegetation cover on bat fatalities

Abstract

Although many tropical countries have functional wind farms, most of the information on their impact on bat populations has come from temperate zones. Our study is based on a 5-year study (2009–2013) of bat captures using mist nets, acoustic recordings, and carcass searches at a wind farm in tropical southern Mexico. We investigated the composition of bat species, estimated the number of bat fatalities per turbine, and assessed the effect of the spatial attributes of vegetation cover near turbines on numbers of bat carcasses recovered by trophic guild. We recorded 29 bat species at the wind farm. The family Phyllostomidae was best represented in terms of number of species and individuals captured. Glossophaga soricina and G. morenoi exhibited the highest relative abundance, whereas Balantiopteryx plicata had the highest acoustic activity. We found 203 carcasses, including 73 Pteronotus davyi individuals (35.9%); other frequent species were Mormoops megalophylla, Molossus sinaloae, and Lasiurus intermedius. The total number of carcasses found within a year ranged from 17 to 83 (2012 and 2009, respectively), with the corrected estimates ranging from 410 to 1,980, or 4.18–20.20 fatalities/turbine. The number of carcasses recorded was positively correlated with secondary vegetation surrounding turbines but negatively correlated with agricultural fields. The spatial attributes of vegetation surrounding turbines influenced numbers of bat carcasses differentially depending on the bats’ trophic guild and habitat use. Contrary to findings from United States and Canadian wind farms, most of the carcasses observed in our study were resident species. Notably, the most commonly captured and acoustically active species were not the most commonly found in carcass searches. To obtain more accurate information about the most vulnerable species and how to reduce the impact on bat mortality, we advise the use of alternative monitoring methods in pre-construction studies. Aunque muchos países tropicales han promovido el establecimiento de parques eólicos en sus territorios, mucha de la información referente al impacto de esta actividad sobre las poblaciones de murciélagos proviene de zonas templadas. Usando datos de 5 años (2009–2013) de muestreos con redes de niebla, grabaciones acústicas, y búsquedas de cadáveres en un parque eólico al sur del trópico mexicano, investigamos la composición de especies de murciélagos, estimamos el número de cadáveres/turbina y evaluamos el efecto de los atributos espaciales de la vegetación que rodea a las turbinas sobre el número de cadáveres de murciélagos por gremio trófico. Registramos 29 especies de murciélagos en el parque, la familia Phyllostomidae fue la mejor representada en cuanto al número de especies e individuos capturados. Glossophaga soricina y G. morenoi fueron las especies con mayor abundancia relativa, mientras que Balantiopteryx plicata fue la especie con mayor actividad acústica. Se registraron 203 cadáveres de murciélagos en el parque eólico, el 35.9% pertenecían a Pteronotus davyi, otras especies frecuentes fueron Mormoops megalophylla, Molossus sinaloae y Lasiurus intermedius. El número total de cadáveres encontrado va de 17 a 83 (2012 y 2009 respectivamente), mientras que la estimación corregida va de 492–1,980, o 4.18–20.20 cadáveres/turbina. La cantidad de cadáveres estimada para los 5 años combinados es de 4,782. El área ocupada por vegetación secundaria alrededor de los aerogeneradores se relacionó significativa y positivamente con el número de cadáveres registrados. Nuestros resultados también indican que la ubicación de los aerogeneradores afecta de forma diferencial a las especies de murciélagos según sus gremios tróficos y uso de hábitat. Contrario a lo encontrado en parques eólicos de Estados Unidos y Canadá, los cadáveres de murciélagos registrados en el parque eólico tropical son de especies residentes, incluyendo aquellas que forman grandes colonias. Ya que algunas de las especies frecuentemente registradas en las redes y acústicamente, no fueron las que se encontraron frecuentemente en los cadáveres, sugerimos diversificar los métodos de monitoreo en estudios de pre-construcción para identificar las especies más vulnerables y tomar acciones que reduzcan el impacto sobre sus poblaciones.

Effects of a wind farm installation on the understory bat community of a highly biodiverse tropical region in Mexico

“…Carollia subrufa , Desmodus rotundus , Dermanura phaeotis and Lasiurus blossevillii were not captured during the construction or operational stages, while Molossidae species only appeared in mist nets during these stages. Bat surveys in the region ( López et al, 2009 ; Bolívar-Cimé et al, 2016 ; Muñoz et al, 2016 ), including the present study, suggest that Carollia subrufa and Dermanura phaeotis are more sensitive to forest cover here than in other regions, because they were only recorded in primary vegetation or in the least perturbed stage. Contrarily, Artibeus and Sturnira species ranked in the first places of species captured in all stages an even increased their abundance during the two operational stages.…”

“…In spite of the disturbance caused by wind farm in the study area, bat species richness is relatively high in the understory (22 species). In a wind farm near our study area, Bolívar-Cimé et al (2016) recorded 29 species of the entire bat community sampled with nets and acoustic monitoring, but of these only 14 species were recorded using mist nets. Moreover, our data reveal that, for bats flying in the understory, after the loss of diversity during construction and the first operational stage (immediate effect), bat diversity and composition by the second operational stage (delayed effect) became similar to the initial stage.…”

“…Recent studies in Brazil ( Barros, Gastal de Magalhaes & Rui, 2015 ) and Puerto Rico ( Rodríguez-Durán & Feliciano-Robles, 2015 ) describe mortality caused by collision with turbines after their construction. In Mexico studies have focused on assessing bat species richness near wind farms ( Briones-Salas, Peralta-Pérez & García-Luis, 2013 ; Muñoz et al, 2016 ), documenting the main scavengers that remove bat carcasses ( Villegas-Patraca et al, 2012 ), and analyzing the effect of the spatial attributes of vegetation cover near turbines on species composition and bat fatalities ( Bolívar-Cimé et al, 2016 ).…”

Influence of landscape and time of year on bat-wind turbines collision risks

“…Santos et al (2013) found that mortality probabilities decreased with the distance to eucalyptus forests. Other studies showed that open habitats such as prairies, pastures and croplands triggered fewer collisions than closed habitats such as forests (Bolívar-Cimé et al, 2016;Piorkowski and O'Connell, 2010;Thompson et al, 2017). Piorkowski and O'Connell (2010) found more fatalities in ravine topography than in low topography relief and Santos et al (2013) found that distance to the slope was negatively correlated with mortality risks.…”

Resource Availability May Not Be a Useful Predictor of Migratory Bat Fatalities or Activity at Wind Turbines

“…A second objective of the study was to compare available resources with acoustic bat activity across the site, as some studies have shown that activity levels are useful for predicting wildlife fatalities and vice versa [22][23][24]. However, emerging studies at wind energy facilities indicate that bat activity monitoring using acoustic detectors in particular is not as effective at predicting bat fatalities as originally anticipated [9,25,26]. Thus, there is a real need to further investigate this uncertainty and determine whether there is a positive correlation between resource availability, bat acoustic activity, and bat fatalities at operational wind farms.…”