South Mediterranean forests are known for their spatial and temporal variability in fauna and flora species abundance. Using acoustic surveys and generalized linear models, we investigated the nocturnal activity of bats in five forest sites in northeast Algeria from March to November. A total of 12 species were detected: Rhinolophus blasii, R. hipposideros, Tadarida teniotis, Miniopterus schreibersii, Eptesicus isabellinus, Hypsugo savii, Pipistrellus kuhlii, P. pipistrellus, Plecotus gaisleri, Myotis cappaccinii, M. emarginatus, and M. punicus. Activity levels varied significantly among months and sites. Overall bat activity and P. kuhlii activity increased from spring to summer, peaked in August, and slightly decreased in early autumn. Activity levels also differed markedly among forest sites, with low activity levels in the urban forest site, and E. isabellinus was mostly active in only two forest sites. These results suggest that both environmental conditions and location-specific factors shape the activity patterns.

Agricultural intensification is a key driver of biodiversity loss in Mediterranean agroecosystems, yet its effects on protected bat assemblages remain poorly documented in northern Africa. All six bat taxa recorded are legally protected in Algeria. Several of the recorded species are also covered by international conservation instruments for their European populations, while the probable North African endemic Plecotus cf. gaisleri is of particular biogeographic interest. We conducted a preliminary assessment of bat activity, diversity, and community composition at two representative agricultural sites—a low-input citrus orchard and an intensively managed vineyard—in northern Algeria. Acoustic monitoring was conducted from March to November 2023. A total of 349 acoustic contacts were recorded at the citrus orchard site and 74 at the vineyard site, corresponding to a 4.6-fold difference in observed activity. Negative binomial modelling indicated a substantially lower fitted contact rate at the vineyard than at the orchard (IRR = 0.215, 95% CI: 0.130–0.355). Monthly taxon richness and Shannon diversity were also higher at the orchard. The vineyard assemblage was strongly dominated by Pipistrellus kuhlii, whereas a more taxonomically diverse assemblage was detected at the orchard. NMDS suggested site-associated structuring of bat assemblages, and a paired PERMANOVA detected a significant difference between the two sampled sites (R² = 0.281, pseudo-F = 6.26, p = 0.004). No statistically significant difference in multivariate dispersion was detected (PERMDISP: F = 0.808, p = 0.100), although the small sample size limits the power of this diagnostic test. Although based on a single orchard–vineyard pair, these results suggest that the low-input orchard examined here may provide favourable conditions for a protected bat assemblage. Replicated multi-site studies are needed before generalizing this pattern to low-input perennial agroecosystems across North African farmland.