of deep-water larvae to boundary-layer flow conditions were 

 tested by measuring gradients in settlement downstream from the 

 leading edges of plates, and by comparing leading-edge settlement 

 on plates predicted to be with and without eddies. Behavioral 

 responses of larvae to substrate composition were also tested 

 with plates composed of different materials. These plates were 

 designed to determine whether deep-sea larvae actively selected 

 naturally occurring substrates, and to compare the magnitude of 

 larval responses to substrate composition with responses to flow. 

 The plates used for the flow and composition tests were elevated 

 above the seafloor to isolate them from flow disturbances caused 

 by small topographic irregularities. Additional plates were set 

 directly on the seafloor in order to collect demersal larvae that 

 were potentially excluded from the elevated plates. 



STUDY SITE 



Cross Seamount is a flat-topped, submerged peak located 60 

 km to the southeast (18°0' N, 158°17' W) of the Hawaiian Islands. 

 The summit rises to within 375 m of the surface and the base is at 

 a depth of greater than 4500 m (topography illustrated in 

 Malahoff, 1985). The site chosen for this study was near the 

 center of the summit at a depth of 410 m. All experiments and 

 current measurements were conducted on an oblong, rippled sand 

 flat, approximately 85-m long and 35-m wide. Ripple orientation 

 was generally north-south. Observers in the Pisces V submersible 

 noted that ripples were asymmetrically shaped and estimated that 

 ripple height was less than 2 cm and ripple spacing averaged 5 cm. 

 Small boulders and outcrops of f erromanganese crust surrounded the 

 site, but the nearby surface relief was no greater than 20-30 cm 

 (with the exception of one large basalt block, 1-m tall and 3-m 

 long ) . 



A sand flat, rather than a site covered with f erromanganese 

 crust, was selected as the study site for its hydrodynamic 

 simplicity. A flat site was needed to ensure that the 

 experimental substrates rested horizontally on the bottom, with 

 the collecting surfaces oriented parallel with the flow. This 

 particular site was expansive enough to distance the substrates 

 from nearby surface topography which could shed eddies, or 

 otherwise complicate the flow regime. Experiments and 

 measurements were conducted near the center of the sand flat and 

 well away from the large basalt block. Ripples and other small 

 surface relief do influence boundary- layer flows (Schlicting, 

 1936), but at this site they were relatively small and regularly 

 spaced. A few centimeters above the seafloor, the influence of 

 the ripples is probably reflected in the flow as bed roughness, 

 rather than as individual flow disruptions. 



253 



