Testing consistency of modelled predictions of the impact of climate change on bats

Changing with the times: Seasonal environmental gradients unveil dynamic bat assemblages and vulnerability

“…Our study identified the upland bat M. mystacinus as the most vulnerable species in the region due to its permanent association with high‐mountain forested habitats. This species has already been acknowledged as one of the European species most threatened by climate change (McGowan et al, 2021 ; Rebelo et al, 2010 ). Moreover, M. bechsteinii (forest/edge bats) is known to be a sedentary species (Napal et al, 2013 ) and very restricted to the mid‐altitude forest, making it the second most vulnerable bat species in the region.…”

Mapping Potential Regions of Human Interaction with Acuminate Horseshoe Bats (Rhinolophus acuminatus) in Thailand

“…To map the current and future distributions of the target species, we considered environmental variables that are thought to significantly affect bats [43]. Because R. acuminatus inhabit lowland dipterocarp forests, primary and secondary tropical forests, urban areas, caves, tree hollows, and buildings [15,16], we incorporated elevation, slope, land cover (19 categories), forest canopy cover height, and cave presence as predictor variables [21,44]. Elevation and slope data were obtained from the Shuttle Radar Topography Mission 90-m Digital Elevation Database v4.1.…”

“…Representative Concentration Pathway 4.5 (RCP4.5) was utilized to predict the climate for 2050 (representing 2041-2060) and 2070 (representing 2061-2080) using the HadGEM2-ES model, which was created by the Met Office Hadley Centre for Coupled Model Intercomparison Project Phase 5 (CMIP5) centennial simulations. HadGEM2-ES was selected for its detailed representation of climatic variation, which makes it suitable for mapping species distributions [44]. We excluded four data layers (BIO8: mean temperature in the wettest quarter; BIO9: mean temperature in the driest quarter; BIO18: precipitation in the warmest quarter; and BIO19: precipitation in the coldest quarter) that demonstrated inconsistent climatic variations among neighboring pixels because they combined information on temperature and precipitation [52,53].…”

“…Test points were then randomly selected from the species' presence locations using a "random test percentage" option, with the entered value of "25" instructing the program to allocate 25% of the occurrence records for testing [66]. Model performance was evaluated using 10 replicates with cross-validation, 500 iterations, and 10,000 background data added from samples for predictive accuracy testing [44]. Model accuracy was determined using the area under the curve (AUC) of the receiver operating characteristic (ROC) plot.…”

Foraging Time and Temperature Affected Birth Timing of Rhinolophus ferrumequinum and Predicted Year-To-Year Changes for 25 Years in a Population in West Wales, U.K.

“…Pipistrellus kuhlii has shifted its range northwards and it is likely that it will lose the southern portion of Europe (Sachanowicz et al, 2006;Smeraldo et al, 2021). In Ireland Nyctalus leisleri, Pipistrellus pipistrellus and P. pygmaeus are predicted to alter their distribution (Roche et al, 2020) with contraction of the range for Myotis mystacinus (McGowan et al, 2021) and in the Carpathian Mountain region a population of Rhinolophus euryale is predicted to adapt its distribution (Uhrin et al, 2021). Also, in North America Lasiurus seminolus has expanded (True et al, 2021).…”