at nSSR loci
The distribution of genetic diversity in M. fragrans (N = 393 individuals) was characterized at the population level (N POP = 5) by estimating the following indices using SPAGeDi v.1.5 : the number of alleles (N A), observed and expected heterozygosity (H O andH E respectively), rarefied allelic richness (AR ), and inbreeding coefficient (F IS).
We then calculated the F ST among populations and among sub-populations within each population using the hierfstatR package (Goudet, 2005) in R (R Core Team, 2013). Moreover, we characterized genetic structure using a Bayesian clustering approach as implemented in STRUCTURE v.x.x software with the following parameters: admixture model, sampling location as prior, independent allele frequencies. The number of clusters was set from K = 1 to 10, with 10 iterations for each K. We ran the analysis with 20,000 burn-in periods and 100,000 MCMC replicates. The value of K that best explained the data was estimated using STRUCTURE HARVESTER . We then summarized STRUCTURE results using CLUMPP and projected the barplot using CLUMPAK .
Genetic diversity and population structure of M. fragrans