320 



Abstract.— We used hydroacoustic 

 techniques to obtain biomass estimates 

 ofyellowtail rockfish, Sebastes flavidus, 

 in a lOO-km-'area off the west coast of 

 British Columbia, Canada. The purpose 

 of our study was to estimate sampHng 

 variance and explore the effect of diel 

 aggregation behavior on the precision 

 of biomass estimates. A set of eight 

 transects was sampled eight times: four 

 at night, four during daylight. Although 

 we observed a pronounced diel behav- 

 ioral pattern of diurnal aggregation and 

 nocturnal dispersion, we found no sig- 

 nificant differences between nocturnal 

 and diurnal estimates of mean biomass. 

 Diurnal estimates showed a tendency 

 towards higher variance, but the dif- 

 ferences were not significant in most 

 comparisons and were too small to in- 

 fluence survey design. The coefficient 

 of variation of the eight observations 

 for any individual transect ranged from 

 13 to V28'7r. The coefficient of variation 

 in biomass for the whole survey area, 

 based on repeating the set of eight 

 transects eight times, was IS.B*?? and 

 the estimate of mean biomass for the 

 survey area was 1152 t. The observed 

 diel behavioral patterns did not, in this 

 study, produce different estimates of 

 yellowtail rockfish biomass. Survey 

 time might therefore be optimized with- 

 out concern for this source of variance 

 for this species. 



Diel vertical migration by yellowtail rockfish, 

 Sebastes flavidus, and its impact on acoustic 

 biomass estimation 



Richard D. Stanley 

 Robert Kieser 

 Bruce M. Leaman 



Ken D. Cooke 



Biological Sciences Branch 



Fisheries and Oceans 



Pacific Biological Station 



Nanaimo, British Columbia, V9R 5K6 Canada 



E-mail address (for R D Stanley) stanieyrffidfo-mpo gc ca 



Manuscript accepted 2.') June 1998. 

 Fi.sh. Bull. 97.320-3.31 (1999). 



Yellowtail rockfish, Sebastes flavidus, 

 are an important component of the 

 British Columbia (B,C,) trawl fish- 

 ery. Annual catches of 4000-5000 t 

 represent about 207r of the total 

 rockfish {Sebastes spp.) landings 

 and over 57( of the total domestic 

 trawl landings from the B.C. coast 

 (Rutherford, 1996). The stock as- 

 sessments for these species have 

 typically relied upon population 

 dynamics models tuned with catch- 

 per-unit-of-effort indices (Stanley, 

 1993) or swept-area surveys (Lea- 

 man and Stanley, 1993). However, 

 problems with both these methods 

 have led us to investigate more di- 

 rect methods. 



Previous hydroacoustic work in- 

 dicated that some of the trawl- 

 caught rockfishes are aggregating 

 species and that they are usually 

 found associated with specific 

 bathymetric features (Wilkins, 

 1986; Leaman et al., 1990; Kieser 

 et al., 1992; Richards et al., 1991). 

 Submersible observations and 

 acoustic tagging of a small number 

 offish confirmed this association for 

 yellowtail rockfish (Pearcy et al., 

 1989), although the acoustic tagging 

 did not indicate consistent diel pat- 

 terns (Pearcy, 1992). This aggregat- 

 ing behavior off the bottom in pre- 

 dictable locations suggests that this 

 species may be a suitable candidate 



for hydroacoustic biomass estima- 

 tion. However, because individual 

 yellowtail rockfish stocks are thought 

 to occupy large areas (Stanley, 1993; 

 Stanley et al., 1994), such compre- 

 hensive hydroacoustic surveys 

 would be costly and time-consum- 

 ing. This accentuated the need for 

 a prior study of fish behavior and 

 sampling variance to assess the fea- 

 sibility of such an approach and to 

 optimize survey effort. We were also 

 concerned about the impact of diel 

 aggregating behavior on the preci- 

 sion and bias of biomass estimates 

 (Olsen, 1990; Appenzeller and Leg- 

 gett, 1992; Simmonds et al., 1992). 

 Within the same genus, Richards et 

 al. (1991) reported strong diurnal 

 aggregation by Pacific ocean perch 

 (S. alutus), and Wilkins ( 1986) has 

 reported strong nocturnal aggrega- 

 tion by widow rockfish (S. entomelas). 

 We investigated these issues with 

 an experimental survey in Novem- 

 ber 1991 in waters 10 km west of 

 Vancouver Island, British Colum- 

 bia, Canada (Fig. 1). The objective 

 was to estimate yellowtail rockfish 

 abundance within a small area and, 

 in doing so, to examine the variance 

 within and among transects. In ad- 

 dition, we wished to characterize 

 diel patterns of behavior and docu- 

 ment the effects of such behavior on 

 the precision of the biomass esti- 



