Realm: Terrestrial
Climate: Temperate/Tropical
Biome: Tropical and subtropical moist broadleaf forests
Central latitude: 32.709250
Central longitude: -17.029830
Duration: 2 years, from 2016 to 2021

257 records

3 distinct species

Across the time series Pipistrellus maderensis is the most frequently occurring species

Methods

In both study periods, in 2016 and 2021 (see Ferreira et al 2022 for the results associated to 2016), we sampled bats' echolocation calls on 74 grid cells with the size of 1 km2. The grid cells were selected and randomly stratified according to altitude and habitat type. In total, we conducted 217 sampling nights (74 grids x 3 sampling nights). However, here we only present the data from 58 sites affected by a large-scale fire in 2016. This dataset is focusing on the 33 burned sites and on 25 unburned control sites located close to the burned areas or in the same grids (total of 58 sites per year). During the sampling period of the island-wide bioacoustics survey we recorded bats from mid-August to late-September of 2016 and between July and late August of 2021. In addition, in 2016 we used an early prototype of the AudioMoth acoustic detector, that used the same audio front-end and microphone, while in 2021 we used the version 1.0.0. of the AudioMoth with the 1.6.0. firmware update. We attached detectors to poles or trees at ca. 1 m above the ground and programmed them to record at a 192 kHz sample rate from sunset until sunrise for 1 min. in every 5 min in 2016 and continuously throughout the night in 2021. Our unit of measure for bat activity was a bat-pass, defined as two or more pulses of a single species detected in a 5-second recording. Files containing bat calls were separated from those without any call. We selected recordings containing pulses with a frequency of maximum energy between 10 and 96 kHz, with a minimum pulse length of 2 ms and a maximum of 500 ms. When two or more sonotypes were identified in one 5-second file, a single bat-pass for each sonotype was counted. File tagging was conducted manually and, to optimise the process and manage the files, all 5-second files were clustered according to their similarity using a custom R script (see Ferreira et al 2022 for more details). Finally, due to the differences in effort between 2016 and 2021 data presented in this dataset was standardized by total minutes of recording per night (110 min in 2016 and 660 min in 2021). Hence, unit depicts by number of bat-passes per min.

Citation(s)

@Mastersthesis{BioTIME_cit410, Study_id = {668}, Citation_id = {410}, Address = {Bremen, Germany}, Author = {Nouioua, Rym}, School = {University of Bremen}, Title = {Effects of large-scale forest fires on insular bats}, Year = {2022}, Url = {https://www.researchgate.net/publication/363380695_Effects_of_large-scale_forest_fires_on_insular_bats} }<br/>@Article{BioTIME_cit411, Study_id = {668}, Citation_id = {411}, Author = {Ferreira, Diogo F. and Gibb, Rory and L{\'o}pez-Baucells, Adri{\`a} and Nunes, Nuno J. and Jones, Kate E. and Rocha, Ricardo}, Journal = {Journal for Nature Conservation}, Pages = {126177}, Title = {Species-specific responses to land-use change in island insectivorous bats}, Volume = {67}, Year = {2022}, Doi = {10.1016/j.jnc.2022.126177}, Language = {English} }