β-diversity is currently receiving increasing attention, after being neglected for a long time, especially in marine environments. Recent works introduced the distinction within β-diversity between turnover and variation. The former relates to directional changes in β-diversity along any gradient, the latter to non-directional changes, or, in other words, to the heterogeneity of assemblages within any spatial, temporal, or environmental extent. However, the quantification of assemblage heterogeneity in assessing patterns of β-diversity is still largely unexplored. Here, we investigate the potential of classical and multivariate measures of β-diversity in highlighting patterns of assemblage heterogeneity examining eight cases of study from Mediterranean Sea, involving different marine organisms and a variety of environmental settings. Multivariate analyses were employed to assess differences in assemblage structure imputable to the investigated source of variability. ANOVAs on a set of diversity indices were also performed to test for effects on patterns of α-diversity. Differences in assemblage heterogeneity were tested using both classical and distance-based multivariate dispersion measures of β-diversity as variation. Mean values of classical β-diversity metrics were analyzed using ANOVA, whereas, for distance-based multivariate dispersion, permutational tests based on a set of resemblance measures were carried out. In all study cases, analyses of β-diversity as variation showed significant effects of the investigated source of variability in modifying patterns of assemblage heterogeneity, even when the multivariate structure of assemblages and/or α-diversity were not affected. The assessment of β-diversity as variation can potentially unveil patterns of change in assemblages that could remain unnoticed analyzing other components of diversity, providing complementary information crucial to the understanding of the effects of natural and anthropogenic disturbances on natural assemblages.
