Mechanisms of recovery and resilience of different components of mosaics of habitats on shallow rocky reefs.

Understanding the mechanisms producing and maintaining discontinuities between patches in mosaics of habitats is necessary to predict changes in patterns of abundance and distribution of species. On temperate rocky reefs, physical and biological disturbance can result in a mosaic of patches of encrusting coralline, turf- or canopy-forming algae. We experimentally investigated the effects of disturbance within and at the boundary between these patches, in order to identify mechanisms accounting for re-colonisation of space and to assess whether the response of boundaries can alter the relative size of contrasting habitats. Also, we tested whether the resilience of the different types of assemblages depends on species richness (i.e. number of taxa present) of habitats or, alternatively, on other properties of systems like differences in life-history traits of dominating species. Although the nature of the mechanisms generating differences among habitats changed among different stages of the colonisation, local processes (within patches) prevailed over larger scale processes (among patches) in determining early patterns of colonisation of space in mosaics. By the end of the experiment, assemblages in clearings at boundaries had recovered to the undisturbed reference condition, in contrast to clearings within barren patches or algal turfs. Boundaries represent, therefore, relatively more stable components of the mosaic, with greater resilience than adjacent habitats. Although sea urchins are often indicated as the main force regulating the proportion of contrasting habitats on shallow rocky reefs, determining the nature of variation in the interaction between species dominating each type of habitat is crucial to understand dynamics of mosaics. Finally, our study provides evidence that resilience could not be entirely controlled by initial species diversity, suggesting that different proportions of dominant taxa could influence the stability of natural systems.