depths. Photographic sampling was repeated in summers of 1986 

 and 1987 and will be continued in the future (Table 1). In 

 summer 1986, we constructed 24 concrete bases ( 40 x 40 x 10 cm 

 approx. dimensions) each with two horizontal granite slabs and 

 one or two vertical slabs (20-30 cm tall) per base. One side of 

 each vertical slab was covered with a one cm mesh vexar cage ( 15 

 x 15 x 5 cm ) to exclude large predators (3 replicates per 

 depth). Three other vertical faces per depth were covered with 

 the same type of cage but with two sides removed to allow 

 predator access but to reduce water flow as in complete cages 

 (cage control). Six other vertical surfaces were left 

 unobstructed at each depth. These units were collected after one 

 year (September 1987) and were photographed on all surfaces using 

 a 35 mm camera and a macro lens ( Kodachrome 64 film). In 

 addition, all mobile fauna on each rock surface were collected 

 and identified to determine if small predators were present that 

 could affect the community development. 



Patches on Rock Walls 



A series of artificially scraped patches were created in 

 summer 1987 by SCUBA divers using a knife and wire brush to 

 remove all encrusting species in patches 10 x 10 cm on natural 

 rock walls at 30 m depth (method in Sebens 1986). Even heavy 

 brushing does not remove all crustose algal thalli or pieces of 

 invertebrate colonies wedged in small crevices. However, 

 community development in such patches mimics that in natural 

 patches and the early successional community that develops is 

 subject to the same influences as are natural patches, including 

 the effects of larger benthic predators that are unlikely to 

 climb onto settling plate racks. These experiments will be 

 monitored twice annually for as many years as possible into the 

 future. 



In 1985 we began photographing a series of 10 unmanipulated 

 0.25 m^ quadrats on rock walls at 30 m using a Nikonos camera 

 with 15 mm wide angle lens, mounted on an aluminum quadrapod 

 (Witman 1985). Natural mortality of sponges, sea anemones, and 

 other encrusting species has been observed in these quadrats. 

 Subsequent photographs will determine the nature of early 

 community development in these natural patches as well, for 

 comparison to results in artificial patches and on settling 

 panels. 



RESULTS 



The first year's community development on settling panels 

 was striking in several aspects. Panels at the shallowest 

 station (30 m) were almost completely covered with erect and 

 encrusting bryozoans (19 species, Table 2), as were the panels 

 from the next depth (50 m) which had 14 species (Figures 2-5). 

 Panels from the deeper two stations were far less heavily 

 encrusted and had smaller numbers of bryozoan species (13 at 65 

 m, 12 at 80 m). Encrusting bryozoans were most abundant at the 



50 



