Population genetic structure
Individual samples of M. myotis suitable for genetic analyses were 120 for the microsatellite analysis and 112 samples for the mitochondrial analysis, as detailed in Table 1. Data were obtained from 25-30 female bats sampled in May from four out of five colonies, while no more than nine samples were collected from colony three due to its small size and scarce accessibility. In order to secure the robustness of genetic analyses, we then run an additional set of tests using a reduced dataset including only nine individuals for each colony. These analyses showed consistent results with the main tests (Supplementary Methods).
Among the 15 microsatellites markers published for M. myotis(Castella & Ruedi, 2000), 12 were selected to carry out this study as they i) could be successfully amplified in most samples, ii) showed variability within the populations, iii) had no evidence for null allele and iv) respected the Hardy Weinberg equilibrium (Table S1). Genotyping of all individuals allowed for the detection of 109 distinct alleles, with an overall mean of 9.5 alleles across loci, ranging from 6.83 to 10.58 within each of the investigated roost (Table 2). Genetic analyses based on allele frequency for the 12 selected microsatellites supported low genetic differentiation between different roosts (FST = 0.030, SE 0.006). Indeed, pairwise FST values between colonies ranged between 0.009 and 0.036 (mean 0.0193) and showed significant although little differentiation only for the small colony three (p-value = 0.001) (Table 3). This same colony also showed the lowest mean allelic richness across loci (Table 2).
Regarding genetic analyses on the mitochondrial DNA, we aligned sequences of 413 bp, among which 20 sites were variable among the South Tyrolean dataset, seven within the HVI and 13 within the HVII regions, defining nine distinctive haplotypes included in two main clades (Figure 2, Table S4). Mean nucleotide distance was 0.12% within groups and 4.5% between groups. The most common haplotype was Hc_01, found in 66% of all individuals and in all colonies, with frequency ranging between 41% and 75% in different locations. The overall frequency of the other haplotypes varied between 0.9% and 17%, with five haplotypes found in a single colony, one found in two colonies and two found in four colonies (Table S4). In each colony, we found three to seven distinct haplotypes, with colony two showing the highest variability (n=7) and the highest number of private alleles (n=3) (Figure 3, Table 2, TableS4). Indeed, mean gene and nucleotide diversity confirmed to be higher within this colony (Table 2). In turn, colony two showed the highest mean nucleotide distance from other colonies based on sequences from HVI+HVII (Table 3). Similarly to what was observed for microsatellites, our data support a very limited geographical structuring in the area, with FST values ranging between -0.013 and 0.084 (mean 0.02), and showing significant although low differentiation only between colonies 1 and 2 (FST: 0.084; p value: 0.019) (Table 3). The Mantel test detected a no sign for isolation by distance based on either mitochondrial (R = 0.287, p=0.052), or nuclear markers (R= 0.011, p=0.43) (Figure S3).