β-diversity, coined as species turnover among communities, has been used for many years as a successful way to investigate biodiversity patterns and species assembly rules. However, most of previously studies have only focused on species composition without considering differences in species biology and/or abundances. We aim going one step further by measuring the functional turnover among communities, through the use of the Rao’s quadratic entropy index. Indeed, this index allows partitioning functional diversity into α-(within-site) and β-(among-site) components while including species relative abundances and functional distances among species. We applied this functional diversity partitioning on a large dataset of tropical estuarine fish communities from the Terminos lagoon region (south of Gulf of Mexico, Mexico). This region is marked by a strong environmental variability, high species richness, and is of primary interest from an ecological point of view. The 37 stations of a 150km-long transect were sampled monthly during two years. Two functions (food acquisition and locomotion) were studied through the measure of 15 functional traits on more than 1000 individuals belonging to 70 species. For each function of interest, functional β-diversity was investigated both through space (stations) and time (months). At the same scales, taxonomic β-diversity was estimated through the use of the Lande’s index.

Results/Conclusions

Species turnover was high both on spatial and temporal scales (around 0.8 for the taxonomic turn-over according to Lande’s index) whereas functional turnover was low on both scales and for the two functions (around 0.3). Moreover, when compared to a null model, which assumes a random assignation of traits values to observed abundances, functional turnover was lower than expected by chance. Besides, species clustering into functional groups for each function revealed that a couple of groups strongly dominated fish assemblages. Consequently, when species belonging to these two dominant functional groups were removed, functional β-diversity values were no longer significantly different than expected by chance. These results clearly indicate that despite a high species turnover, some core functional groups stabilize the functional structure of local communities, thus maintaining a low functional β-diversity through space and time in a highly variable environment. We believe that this new framework will be helpful in community ecology to go further than exploring only turnover in species composition. For instance, functional β-diversity used in parallel with taxonomic and phylogenetic β-diversities will help to analyze macroecological patterns and may contribute to disentangling the effects of dispersal, niche filtering and competitive interactions in communities.