LAUSR

The Caatinga of northeastern Brazil is considered one of the most species-rich dry forests in the World. The area may harbor nearly 100 species of bats and correctly accessing such richness is a scientific challenge. Bat inventories have been usually based solely on mist netting, which is biased in drylands, where nearly half of the species are open-space insectivores flying high above nets, or species able to better detect and avoid nets. Here, we used mist netting and acoustic inventories in three protected areas in the Caatinga, so species richness could be compared based on both techniques. After an effort of 1,010 net.hours and 1,941 minutes of recordings, we recorded 29 species with nets, 27 on recordings, and five with both techniques. Acoustic sampling resulted in more species in 44% of nights sampled, and two nights had species recorded solely based on their echolocation calls. Species recording differed between techniques, and the driest the site, the lowest the captures with nets, while records based on echolocation remained constant. Based on species estimators, netting achieved 51% of the expected richness, acoustics achieved 87%, and both techniques achieved 70%. Using acoustic sampling we recorded ca. of 42% of the insectivorous species expected for the entire Caatinga and 20% solely on netting. Moreover, in our study bioacoustics indicated that some species were more abundant and widespread than expected. Therefore, studies using solely netting or bioacoustics will produce different results and, in Neotropical dry forests like Caatinga, for a more complete scenario on the local bat species richness a combination of both techniques is essential.