Why are there so many species in deep-sea sediments?

High species diversity in samples of macrobenthos of deep-sea sediments is now well established. But a consensus on the processes regulating this unexpectedly species-rich coexistence in metazoan species at the deep-sea bed is still elusive. This review takes a broad approach by examining differences between marine and terrestrial biodiversity in the context of the following: scale of sampling effort on which our knowledge is based; the species concept as applied in the past to taxonomic studies on deep-sea benthic organisms; scaling differences and size related patterns in community structure and habitat complexity and differences in potential for co-evolution. Latitudinal and bathymetric patterns are summarised in relation to habitat variability and distributional range and in relation to J.S. Gray's (1994) claim that benthic diversity may be equally high on the continental shelf. Evidence is first reviewed for neo-Darwinian competitive co-existence based on niche specialisation and habitat partitioning. Although evidence for any dietary specialisation is sparse, biologically generated heterogeneity in the sediment, such as from mounds, burrows, tubes and 'mud balls', will persist longer and contribute more to niche diversification than in shallow water owing to slower obliteration by sedimentation and water turbulence. Second, the role of periodic small-scale disturbance in creating a shifting dynamic of invasion and species succession in patches of new, open habitat, is reviewed in relation to data from sea bed observations and experiments. Such disturbances may arise from effects ranging from organic enrichment as falls of carcases or dead plants, seaweed or wood, to patchy phytodetrital deposition from the surface. Environmental patchiness of both kinds may be important in determining small scale, predominantly non-competitive co-existence, particularly in a habitat lacking large-scale isolating barriers and open to chance, low-intensity recruitment from propagules of a wide taxonomic and functional spectrum of biota. Disturbance may also be expressed as diffuse, large-scale effects caused by sea bed currents. At their most extreme, 'benthic storms' erode and re-deposit sediment and create a regime of intermittent disturbance that reduces species richness; but more moderate Flow may enhance recruitment and remould the small-scale sediment landscape to create new habitat. The wide range in habitat partitioning and disturbance processes that may be involved in controlling species richness make realistic modelling studies challenging. Careful description of the full spatio-temporal range in environmental hetergeneity at the deep-sea bed is badly needed. Further manipulative experiments, and perhaps pilot 'industrial'-scale studies involving laying bulky inert or organically enriched waste on the bed, will certainly also be useful in improving our knowledge of processes controlling deep-sea macrobenthic species richness.