Species–area relationships in honeycomb worm bioconstructions

Abstract The increase of species richness with the available or sampled area is a fundamental principle in ecology, but explanatory models of this relationship have never been tested in worm reefs representing priority systems as biodiversity hotspots and habitats of conservation concern. Alternative models based on the ‘random placement hypothesis’ versus the occurrence of processes operating at different scales were experimentally tested in patches of the tubeworm Sabellaria alveolata . Macrofaunal assemblage structure and Hill numbers giving increasing weight to the most abundant species were compared between cores collected from the central portion of patches of three sizes (‘small’, ‘medium’, ‘large’) and between cores collected from the centre versus the edge of medium and large patches. Faunal richness per sampled unit was independent of patch size, consistently with passive sampling theory. Instead, diversity estimated by Hill number 1 D (exponential of Shannon entropy) progressively decreased across large, small and medium patches, while small and large patches showed higher Hill number 2 D (inverse of Simpson diversity) than medium patches, suggesting the occurrence of size‐specific underlying processes. Our findings can be explained with effects of processes that would vary with the perimeter‐to‐area ratio of S. alveolata patches of varying size, such as wave disturbance, predation pressure and colonization patterns. This challenges the traditional view of species–area relationships, especially when focusing on small rather than biogeographic scale and multiple diversity measures, with potential implications for conservation and ecological theory. It suggests, in particular, that a mosaic of sabellariid patch sizes may be equally or even more effective than preserving or establishing large and continuous habitats for maintaining associated biodiversity and ecosystem function. It also emphasizes the importance of not assuming paradigmatically the wide validity of species–area models and incorporating within‐habitat factors when attempting to apply general ecological theories to conservation and management.