The dynamics of soft-bottom disturbance mosaics may be strongly influenced by life stages of colonists, disturbance size, and patch isolation. We assessed the effects of postlarval immigration, patch size, and vertical isolation on colonization following small-scale disturbance in a mudflat in Barnstable Harbor, Massachusetts. Defaunated sediment plugs of two sizes (50 and 1,750 cm2 in plan area) and two levels of isolation (flush with the seafloor and elevated 5 cm) were implanted in the flat and sampled after 4-41 d. Postlarval immigration proved a major colonization mode for both treatment sizes. Colonization rates and successional patterns varied markedly between patch sizes, however. Fauna1 abundance and species number increased more rapidly, and species proportions differed, in smaller treatments primarily because the contribution of postlarval immigration varied inversely with patch size. Colonization in elevated plugs bore little resemblance to that in flush treatments, with macrofauna accumulating in raised plugs at markedly lower rates. We conclude that postlarval immigration may be a major mode of colonization at our site and perhaps in soft bottoms generally, following small-scale disturbance, that patch size must be considered in models of benthic colonization and succession, and that interpreting results from colonization studies with raised substrata may be problematic.Soft-bottom macrofaunal communities often are mosaics of small patches recovering from recent disturbance by rays, horseshoe crabs, mound builders, and other agents (e.g. Thistle 198 1;Levin 1984;Botton 1984). To elucidate and model the functioning of such mosaic systems (e.g. Paine and Levin 198 l), one must understand the mechanisms and rates of arrival of new organisms into disturbed patches.Colonists may arrive into a disturbed patch by one of several pathways or "modes": settlement of pelagic larvae from the water column; reproduction within the AcknowledgmentsWe thank H.