526 



Fishery Bulletin 104(4) 



for differences among sites with a one-way ANOVA. 

 Depth and size distributions were compared among sites 

 with a Kolmogorov-Smirnov (K-S) test. To estimate 

 population sizes at each site, densities were applied 

 to depth-area polygons derived from multibeam sonar 

 bathymetry. 



To examine the fine-scale spatial distribution of white 

 abalone on the banks we determined and plotted linear 

 distances between individuals along transects and re- 

 corded the number of groups (individuals within 2 m of 

 another abalone) versus individuals (no other abalone 

 within 2 m). Additionally, the distances between all 

 individuals at each site were calculated, and for each 

 abalone the distance to the nearest subsequent abalone 

 sighting was determined. This was not a traditional 

 nearest neighbor analysis, because transects were not 

 designed for such an analysis. To test the degree of 

 dispersion of the population on a broader scale, the cu- 

 mulative number of white abalone observed within 250- 

 m^ segments of transect between 40 and 50 m depth 

 (the depth range with the highest abalone densities) at 

 Tanner Bank in 2002 were fitted to a negative binomial 

 distribution (Elliott, 1977). The Tanner Bank 2002 sur- 

 veys were appropriate for this analysis because of the 

 high number of abalone sightings there. 



Results 



Habitat 



The amount of rocky habitat between 30 and 60 m depth 

 at Tanner Bank, Cortes Bank, and the west side of San 

 Clemente Island was 1359, 1139 and 889 ha, respec- 

 tively (Fig. 1). Side-scan sonar data indicated that the 

 majority of the sea floor shallower than 60 m at all sites 

 is hard substrate (Fig. 1). High-resolution bathymetry 

 from multibeam sonar revealed a complex structure on 

 the offshore banks. Tanner Bank and Cortez Bank are 



uplifted anticlines that have been eroded by wave action 

 from the northwest during past periods of lower sea level 

 (Vedder et al., 1986). Erosion of more recent geologic 

 units has exposed erosion-resistant marine sedimentary 

 and volcanic rocks that through differential erosion have 

 developed layering and fracture features favorable to 

 abalone. San Clemente Island is on the uplifted southern 

 side of the San Clemente fault (Vedder et al., 1986), and 

 is composed of erosion-resistant volcanic rocks of the 

 same age as those at Tanner and Cortez Banks. 



Microhabitat analysis 



At Tanner Bank in 2002, white abalone were generally 

 found in association with two species of brown algae 

 (Laminaria farlowii and Agarum fimbriatum [88% of 

 white abalone sightings included one or both species]), 

 a foliose red alga (Rhodymenia spp. [45% of sightings]), 

 genticulate coralline algae (Calliarthron spp.) and Coral- 

 Una spp. [79% of sightings]), and encrusting coralline 

 algae (Lithothamnion spp. [100% of sightings]). The 

 brown alga Dictyotaceae spp. was present in 9% of white 

 abalone sightings (mostly in deeper waters), and Eise- 

 nia arhorea in 6% of sightings (mostly in shallower 

 waters). 



A PCA examination of these data revealed a strong 

 positive relationship between the presence of white aba- 

 lone and several factors; the strongest loadings occurred 

 with depth, microhabitat type (Mhab 4 (deformed with 

 sand)), the magnitude of seafloor slope (slope 1, flat), 

 moderate seafloor complexity (Com 2 (low)), and the 

 relative abundance (common or abundant) of brown 

 algae A (Agarum fimbriatum), L {Laminaria spp.), and 

 Br (brown algae spp. in general); (Fig. 2, Table 1). In 

 contrast, a strong negative relationship existed between 

 zero white abalone density and habitat characteris- 

 tics; the strongest negative relationships occurred with 

 the frequent presence of various algal species (same 

 as above). There were strong positive relationships 



