Population structure based on microsatellite markers
In this study, we divided the populations into three groups; 1) greenhouses (commercial) populations 2) nearby greenhouses and 3) far from the greenhouses populations and assessed genetic structure of populations and groups. PIC values were higher than 0.5 in all studied loci (0.88 ± 0.08). All microsatellite loci were highly polymorphic, with 4–27 alleles per locus and an average of 18.30 alleles (Table 2; Table S1). F-statistic and gene flow (Nm ) values for 20 microsatellite loci in all populations are given Table 2.FIT, FIS andFST values of the total population consisting of 14 subpopulations were determined as 0.244, 0.203 and 0.053, respectively in B. terrestris . These values indicated medium inbreeding and low genetic differentiation among B. terrestrispopulations. Gene flow (Nm ) for the total population was found to be high (mean 4.9). This suggests that 5 out of every 100 individuals have migrated between populations. Allelic richness (AR ) values ranged from 8.95 to 10.30 among populations (Table 3). The lowest observed heterozygous (Ho ) value (0.63) was detected in the second group, while the highestHo value (0.75) was observed in the first group (commercial populations). The mean Ho and He were 0.68 and 0.86 respectively for all populations. The mean inbreeding coefficients were 0.18, 0.17 and 0.22 in group 1, group 2 and group 3, respectively (Table 3).
Pairwise FST values indicating the genetic differentiation between the sub-populations were given in Table 4. The lowest pairwise FST values were observed between TP5 and TP6 populations (0.014), while the highest pairwise FST values were determined between TP2 and TP5 (0.045). It was found that there was no significant difference between the populations in accordance with the pairwise FSTvalues (p> 0.05). A large part of molecular variance (AMOVA, Table 5) was found within individuals (65%) rather than between individuals (32%) or between all populations (3%).
Based on the Factorial Correspondence Analysis (FCA), populations were divided into four main clusters. The first cluster consisted of TP1, TP2 and TP3 populations, the second cluster consisted of TP4, TP5 and TP6 populations, the third cluster consisted of TP7, KM, DM and GB populations, and the last cluster consisted of AK, BB and TM populations. The FS population was located between the third and fourth clusters (Fig. 1). The STRUCTURE results were consistent with the FCA. STRUCTURE HARVESTER suggested that K = 4 was the most plausible model (Fig. 2). Based on the model populations were divided into four clusters (K = 4). Cluster one consisted of TP1, TP2 and TP3 populations (in the purple colored), cluster two consisted of TP4, TP5 and TP6 populations (in the green colored), cluster three consisted of TP7, KM, DM, GB and FS populations (in the orange colored) and cluster four consisted of AK, TM and BB populations (in the pink colored). Commercial populations were separated from wild populations except TP7.