314 



Fishery Bulletin 92[2), 1994 



tively correlated with depth (Westrheim, 1970; 

 Boehlert, 1980; Wilkins, 1980; Richards, 1986). 



Nearshore rockfish are usually found in close as- 

 sociation with the substrate or vertical relief (e.g. 

 kelp beds) and their density may be dependent on 

 the type of habitat available (e.g. boulder fields, 

 shelter holes, etc.) (Patten, 1973; Moulton, 1977; 

 Richards, 1986, 1987; Pearcy et al., 1989; Stein et 

 al., 1992). Depth and density distributions of near- 

 shore rockfish in British Columbia have been as- 

 sessed primarily by hook-and-line surveys (e.g. 

 Richards et al., 1988) and observations from 

 submersibles are lacking. To date, only one study 

 has assessed the distribution of nearshore rockfish 

 in British Columbia using direct observations from 

 a submersible (Richards and Cass, 1985; Richards, 

 1986). In their study, the depth and habitat distri- 

 butions of rockfish were surveyed at depths of 21- 

 140 m in coastal waters of the northeastern Strait 

 of Georgia. In addition to observations on depth and 

 type of habitat frequented by various rockfish spe- 

 cies, however, submersibles also provide a unique 

 opportunity to observe the behavioral activities of 

 the fishes and their associations with conspecifics 

 and heterospecifics. To date, there has been a lack 

 of submersible studies that attempt to quantita- 

 tively assess the in situ activities of rockfish. 



In the present study, we examined the distribu- 

 tion of rockfish observed from a submersible de- 

 ployed in Saanich Inlet, a coastal fjord in the south- 

 ern Strait of Georgia, British Columbia (Fig. 1). 

 Fjords in British Columbia differ from open coastal 

 areas, such as those surveyed by Richards (1986), 

 in that they typically rely on estuarine-type circu- 

 lation for mixing but have submerged sills which 

 restrict mixing of waters below the sill depth 

 (Thomson, 1981). For fjords of Vancouver Island, 

 B.C., limited circulation results in low dissolved 

 oxygen levels in relatively deep water (Pickard, 

 1963). In Saanich Inlet, restricted mixing at depth 

 results in oxygen-deficient waters (<2.0 mg-Lr 1 ) be- 

 low 100 m throughout most of the year and inter- 

 mittent, seasonal (usually during January-August) 

 anoxia with production of hydrogen sulfide 

 (Herlinveaux, 1962; Liu, 1989), which is toxic to 

 aerobic organisms (Martin et al., 1981). Various 

 studies of invertebrates in Saanich Inlet have shown 

 that the concentration of dissolved oxygen in the 

 water limits their depth distribution (e.g. Burd, 

 1983; Mackie and Mills, 1983; Jamieson and Pikitch, 

 1988; Liu, 1989). Field studies have also demon- 

 strated that the vertical or horizontal distribution 

 of fishes is positively correlated with oxygen concen- 

 tration (reviewed in Kramer, 1987). We therefore 

 speculated that depth distributions for rockfish spe- 



cies in Saanich Inlet would be relatively shallow 

 when compared to their reported maximum depths 

 in open coastal areas. 



The present study is the first to examine in situ 

 species composition and density of rockfish in a 

 coastal fiord in British Columbia. In addition, in situ 

 behavioral activities and species associations of 

 nearshore rockfish at depths greater than 30 m in 

 coastal waters of British Columbia have been de- 

 scribed for the first time. 



Methods 



The Pisces PV submersible (Department of Fisher- 

 ies and Oceans, Canada) was used to survey rock- 

 fish populations in Saanich Inlet on 9-10 December 

 1986. A comprehensive description of the Pisces PV 

 submersible was given by Mackie and Mills (1983). 

 The inlet has a steep, rocky slope bottom inter- 

 spersed with sand-shell valleys and is 7.2 km at its 



Figure 1 



Location of transect sites ( • ) in Saanich Inlet, 

 Vancouver Island, British Columbia. Inset: Location 

 of Saanich Inlet in relation to the mainland of B.C. 



