probable that, on some scales, hydrodynamic processes play an 

 important role in initial settlement and subsequent colonization 

 in the deep sea. 



For many years, studies of settlement onto hard substrates 

 were conducted almost exclusively in shallow water, with the 

 exceptions of experiments on wood-boring bivalves (Turner, 1973; 

 1977) and a series of deep-water biodeterioration studies 

 (DePalma, 1962; Muraoka, 1964; 1966a; 1966b; 1966c). The 

 scarcity of research on settlement and colonization of deep-sea 

 hard substrates is surprising since several studies of softbottom 

 communities have shown that dispersing colonists may have an 

 important influence on community structure in the deep sea 

 (Grassle, 1977; Desbruyeres, Bervas and Khripounoff, 1980; Levin 

 and Smith, 1984; Smith, Jumars and DeMaster, 1986). A recent 

 experiment conducted at 1300 m in the Catalina Basin near 

 Southern California showed that initial settlement of 

 foraminifers and metazoans onto hard substrates (manganese 

 nodules and ceramic models of nodules) was strongly influenced by 

 the elevation of the substrate off the seafloor. This pattern 

 may have been due to larval responses to the difference in flow 

 characteristics directly over the seafloor and 20 cm above it 

 (Mullineaux, 1987a; 1988). The study also suggested that 

 several metazoan species may be actively selecting 

 f erromanganese substrates over ceramic models. The indications 

 that deep-sea larvae respond to flow conditions during 

 settlement, and that initial colonization of deep-water habitats 

 may be quite different than shallow water environments, motivated 

 the choice of study site and the experimental design of the 

 present study. 



The major goal of this study is to investigate the response 

 of hard-substrate larvae to different flow regimes in a deepwater 

 environment. The experimental substrates were flat plates, 

 chosen for their hydrodynamic simplicity, as the flow regime over 

 them could be adequately described and quantified with laboratory 

 flume measurements. When a flat plate is placed in steady, 

 unidirectional flow, a boundary layer develops and thickens as a 

 function of distance from the leading edge. For a very thin flat 

 plate, the thickness of the boundary layer depends only on the 

 velocity of the oncoming flow, the distance from the leading edge 

 and the viscosity and density of seawater, and can be calculated 

 from empirical equations. At the leading edge of a thick plate, 

 however, the plate acts as a bluff body, and can cause the flow 

 to separate. This separation is due to an adverse pressure 

 gradient which develops over the plate, forming an eddy 

 downstream of the leading edge. In this case, the boundary-layer 

 thickness can be predicted from equations only downstream from 

 the eddy, where the boundary layer has reattached. 



Experiments were conducted at the deep-water site on Cross 

 Seamount to address several specific questions. The responses 



252 



