covered most space on the inner surfaces of plates at all depths, 

 whereas T. lilliacea covered more space on outer surfaces. 

 Coverage by Idmidronea could be limited either by the 

 availability of larvae in the water column, the rate of water 

 flow, or by events that occur post-settlement, such as 

 interactions with the other species present. The inverse 

 relationship between Idmidronea and all encrusting species 

 combined indicates either that substantial cover of Idmidronea 

 inhibits successful recruitment by the other species or that 

 significant cover of encrusting species inhibits Idmidronea . 

 These hypotheses cannot be separated at present, however, during 

 much of the first year space must have been available to both 

 groups. Idmidronea canopy appears to develop and expand rapidly 

 and could thus have prevented further recruitment or lateral 

 growth of encrusting species. Encrusting species preventing 

 Idmidronea recruitment by precedence would depend on a high 

 percent cover. Even at the 30 m depth, these species reached 40 

 percent cover only after one year and would thus have been at 

 much lower cover during most of the time period when the panels 

 were invaded by Idmidronea larvae. 



Considering the two erect species together, or Idmidronea 

 alone, there is a greater percent cover on inner plate surfaces 

 at all depths. This could result from a more favorable position 

 for settlement, for example in eddies created behind plates. Or 

 it could result from more favorable growth conditions during the 

 year, again because of the particular flow environment on the 

 back surfaces of plates. In 1987 we placed an Interocean S-4 

 recording current meter 0.4 m off the bottom near the DRS units 

 at each depth. 30 hour records under similar surface wave 

 conditions indicate that the 30 m site received more than twice 

 the water movement of the 65 or 80 m sites and that the 50 m site 

 was intermediate in flow speed. Much of the flow at the shallow 

 sites was wave-generated surge, although there were also periods 

 of strong unidirectional current passing over the top of the 

 pinnacle and lasting for 10 minutes or more. The deeper two 

 sites had little surge but had unidirectional tidally-generated 

 currents that reversed direction as the tide changed. Flow data 

 from September 1987, taken at 0.5 second intervals for one minute 

 of every five minutes, showed a range where most values were 5-25 

 cm.s -1 with peak flows over 50 cm.s" 1 at 30 m, a range of 2-15 

 cm.s -1 with peaks to 40 cm.s -1 at 50 m, a range of 0-12 cm.s -1 

 with peaks to 25 cm.s -1 at 65 m and at a range of 0-18 cm.s -1 

 with peaks to 27 cm.s -1 80 m (Fig. 6). Means for the period 

 were approximately 15, 8, 6, and 6 cm.s -1 for 30, 50, 65 and 80 m 

 respectively. Water flow alone could, therefore, result in 

 decreased advection of larvae to settling plates at the deeper 

 sites. This does not rule out, however, the possibility that 

 larvae are stratified in the water column and that certain 

 species are thus more likely to settle at any one depth. The 

 depth-specific patterns of many of the bryozoan species indicate 

 that this is very likely. For example, Idmidronea was far more 

 abundant at 65 than at 80 m although flow environments appear 

 similar. 



61 



