Figure 2

Principal coordinate analysis approximating the multivariate homogeneity of groups variance. We used the Bray–Curtis dissimilarity index. The graph shows how much the data points (smaller symbols), analyzed by year, species and sex, disperse from a group centroid (larger symbols) representing the mean value calculated across the multivariate space. The first axis (PCoA1) explains 18.23% of the total variance, while the second axis (PCoA2) explains 5.84% of the total variance. The organization of points in the multivariate space suggests a deep differentiation between sampling years.