268 



Fishery Bulletin 97(2), 1999 



n-ixi} 



'ii^^mm 



Figure 2 



Shortraker or rougheye rockfish tilted with the current while lying on the seafloor 

 in the eastern Gulf of Alaska in May 1992. Observed from a submersible. 



Spatial distribution 



Of the 115 above-bottom rockfish. 106 were grouped 

 with at least one other rockfish as they descended 

 together to the seafloor. They contacted the seafloor 

 within 5 m of at least one other rockfish in the group. 

 Because these fish maintained a <5-m spacing from 

 each other, all rockfish within 5 m of other rockfish 

 on the seafloor were considered grouped. Twelve sites 

 had 411 grouped fish, and 14 sites had 235 solitary 

 fish (Table 2). Most groups were small; 82 of the 113 

 groups contained only 2 or 3 rockfish, whereas only 

 two groups contained more than 12 rockfish. Twenty- 

 six pairs of rockfish consisted of a small and medium- 

 size fish separated by less than 0.5 m. 



Substrate associations 



Twenty-two combinations of primary and secondary 

 substrates were encountered at the 15 sites (Table 

 3). Substrates changed frequently within each site, 

 averaging 6.9 substrates/site. Rockfish were associ- 

 ated with 20 of the substrates, but no consistent pat- 

 tern of association was observed within a site. F'or 

 example, the 8 greatest densities (>16 rockfish/1000 

 m-) included 7 different substrates, the 23 lowest 

 densities (no rockfish) included 13 different sub- 

 strates, and 3 substrates had both more than 16 rock- 

 fish/1000 m^ and no rockfish. 



For all sites combined, the most abundant sub- 

 strate was cobble with sand (17.3%), whereas 7 of 

 the 22 substrates made up 1.0% or less each of the 

 total substrate (Table 4). The greatest densities of 

 rockfish were usually associated with soft substrates. 

 Sand with mud had the greatest average density (9.1 

 rockfish/lOOO m-), and 7 of the 10 greatest densities 

 were associated with primary substrates of sand or 

 mud. Nine of the 12 lowest densities of rockfish were 

 associated with primary substrates of cobble, rock, 

 or pebble. 



Boulder and slope associations 



Average boulder indices ranged from 0.0 to 2.1 (Table 

 3 ) and were not highly correlated (r=0.30 ) with rock- 

 fish densities. For all sites combined, high-abundance 

 boulder habitat contained greater densities of rock- 

 fish than did low-abundance boulder habitat (Fig. 3). 

 Habitats where boulders were absent (index=l) or 

 rare (index=2) were encountered 67% ofthe time and 

 had 56% of the rockfish, whereas habitats where 

 boulders were more abundant (index=3, 4, or 5) were 

 encountered 33% ofthe time and had 44% ofthe rock- 

 fish; 52 rockfish were found lying against boulders 

 ( F'ig. 4). The only site without boulders (site 2) was 

 the only site not containing rockfish. 



Average slope indices ranged from 0.9 to 3.7 and 

 were not highly correlated (r=0.56) with rockfish 



