Impact of future climate change
The expected climatic conditions by the mid-21stcentury are likely to result in smaller climatically suitable areas for bats in South Asia (average reduction >8%), but with large portions of the current areas expected to remain suitable (average retention >59%. Figure 2). However, these average impacts hide considerable variation among species resulting partly from different climatic variables being important across species and models (Supplementary Material 1). In all scenarios there were potential winners and losers, with some species predicted to have no climatically suitable areas in the future while for others the area could double, and retention varied from nearly 0 to 99.8% (Figure 2). These ‘losing’ and ‘winning’ species were more common in some taxonomic groups, with species in the family Miniopteridae being consistently among the highest losers, and species in Rhinopomatidae and Pteropodidae predicted to be among the most affected in the CanESM5 model scenarios and the HadGEM3 model scenarios, respectively. On the other hand, Molossidae and Emballonuridae showed the highest gain in suitable areas under the CanESM5 model and the HadGEM3 scenarios, respectively. Nevertheless, in nearly all scenarios even winning families generally had species projected to lose suitable area (Supplementary Material 2). In the CanESM5 scenarios Molossidae showed the highest projected retention of suitable areas in the future for any family (73.5% in Can2-45 and 71.2% in Can5-85), while Hipposideridae was projected to retain the highest amount of suitable area in the HadGEM3 scenarios (80.7% in Had2-45 and 76.8% in Had5-85; Figure 2; Supplementary Material 2).
Overall patterns were largely consistent among the four future climate change scenarios explored, with smaller impacts predicted under ‘middle of the road’ SSP2-RCP4.5 socioeconomic scenarios than in the pessimistic ‘fossil-fuelled development’ SSP5-RCP8.5 scenarios (Figure 2). Many species show consistent losses, like Kerivoula malpasi , known from five localities, with predicted losses of 66% to 84% in climatically suitable area across scenarios; while others showed consistent gains, including Myotis hasseltii that was projected to increase the area of suitable climate by more than 60% andSaccolaimus saccolaimus which is projected to gain 106% of its current suitable area (Figure 2; Supplementary Material 2). For some individual species projected impacts were highly dependent on the scenario - for example, Myotis csorbai , known from seven localities, was projected to lose 100% of climatically suitable areas in the future in CanESM5 scenarios, but had losses around 47% in the HadGEM3 scenarios. Pteropus medius , with 439 occurrences, andCynopterus sphinx , with 312 occurrences showed moderate projected retention of suitable areas (25.9 – 57.1% in P. medius , 30.8 – 44.8% in C. sphinx ; Supplementary Material 2), with varied degrees of projected losses in all scenarios (18 – 61.7% in P. medius , 37.9 – 60.7% in C. sphinx ; Supplementary Material 2).
Across all climate scenarios and species, and including all projected spatial changes, future climatically suitable areas were on average 216 km from current climatically suitable areas. However, distances also varied between scenarios and species. For example, Murina pluvialis’ climatic suitable area was projected to shift by an average of 4.5 km in the future, the smallest distance in the CanESM5 scenarios; however, in the HadGEM3 scenarios the average distance between current and future areas was 128 km (Supplementary Material 2). For some species, the disparity in projected suitable areas between socioeconomic scenarios led to extremely large differences in distance between current and future. For example, Rhinolophus subbadius , in the SSP2-RCP4.5 scenarios for both climate models, had a projected distance of 28 km from current to future; however, in SSP5-RCP8.5, the distance increased to 2202 km. Most future suitable areas were located northward from current suitable areas with a trend for more north-eastern shifts under SSP2-RCP4.5 socioeconomic scenarios and more north-western shifts under the more pessimistic SSP5-RCP8.5 scenarios (Figure 3; Supplementary Material 2).