To test whether colonizing macrofauna specialize on different types of small-scale patches of food and disturbance in the deep sea, sediment tray and artificial depression colonization experiments were conducted on the deep-sea floor at 900-m depth, south of St. Croix, U.S. Virgin Islands. Trays and depressions were unenriched (Unenriched Controls) or enriched with either Thalassiosira sp. or Sargassum sp. Concurrent deployment of different types of enrichment and disturbance made it possible to evaluate whether macrofauna specialize on different patches, and thus avoid species interactions that might lead to competitive exclusion. Depressions create a hydrodynamic regime that traps passive particles, allowing tests of the relative importance of active selection of different patch types versus passive deposition for abundant colonizers. After 23 d, total densities and densities of the four abundant colonizers (Capitella spp., Nereimyra punctata, Cumella sp. and Nebalia sp.) were extremely high in enriched trays, despite relatively low ambient densities. Densities in Unenriched Control Trays were very low, and did not attain ambient densities. After 24 d, total densities in all depression treatments were considerably lower than in enriched tray treatments, and only Sargassum Depression densities exceeded those in the ambient environment. Lower densities of organisms in depression treatments compared with trays and differences in densities among depression treatments suggest that the dominant colonizers were highly active and selective, and were not passively entrained in depressions. Faunal analysis indicated that trays and depressions were very different, and Sargassum Depression fauna was very different from other depression types. A strong difference was not observed between fauna in ambient sediments and Thalassiosira sp. or Unenriched Control Depressions, perhaps because Thalassiosira sp. was dropped in depressions on the sediment surface and may have been more readily available to consumers and more rapidly consumed than in trays. Thalassiosira Trays were colonized by a lower diversity fauna than Sargassum Trays, and Unenriched Control Trays were colonized by very low densities of a fauna that was comparable in diversity to the ambient community. Diversity in Sargassum Depressions was higher than in enriched trays but lower than in other artificial depressions and the ambient fauna. Thalassiosira Depressions and Unenriched Control Depressions were comparable in diversity to ambient fauna and natural depressions, which were highly diverse. These experiments suggest that fauna may respond quickly and selectively to artificial food patches and disturbance, and this fauna is different from that observed in the ambient sediment. Thus, a patch mosaic may be part of the reason for the high species diversity that is observed in deep-sea ecosystems. The different, highly diverse, fauna observed in natural depressions compared with flat ambient sediment suggests that natural analogs of these experiments have unique faunas that may contribute to the species richness of deep-sea habitats.