612 



Fishery Bulletin 89(4), 1991 



for older larvae, and relatively slow again 

 for juveniles. This suggests that it is pos- 

 sible to improve upon the linear model 

 and provide a more accurate description 

 of growth over the entire range of ages. 

 Since Pannella (1971) first discovered 

 daily growth increments in otoliths, many 

 researchers have used these microstruc- 

 tural features to study fish growth (see 

 reviews in Campana and Neilson 1985, 

 Jones 1986). By measuring the widths of 

 daily increments within the otolith, and 

 defining the relationship between stan- 

 dard length and otolith radius, one can 

 back-calculate the somatic length of fish at any given 

 age. This approach provides a powerful tool for esti- 

 mating growth rates (e.g., Thorrold and Williams 

 1989). In this study, we used otolith microstructure 

 to back-calculate the growth of larval and juvenile 

 shortbelly rockfish Sebastes jordani. To do this we 



(1) determined when daily increments begin to form, 



(2) validated the daily periodicity of the increments, 



(3) developed a growth model for back-calculation, and 



(4) assessed the effects of size-selective mortality on 

 back-calculated estimates of growth. 



Methods 



Field collections 



Larval and juvenile shortbelly rockfish were collected 

 on four cruises between 1 February and 13 June 1989 

 (Table 1). All sampling was conducted along the cen- 

 tral California coast from Cypress Pt. (Monterey Co.) 

 to Pt. Reyes (Marin Co.). Larval samples were collected 

 with oblique tows to maximum depths of 50-200 m, and 

 were towed at depth for 30 seconds, at an approximate 

 ship speed of one knot. Larval samples were collected 

 with aim plankton net (0.505 mm mesh) or, for the 

 18 February cruise, a 2x2 m Isaacs-Kidd trawl (2mm 

 mesh). For further information on larval hauls, see 

 Appendix A. Larvae were sorted and placed in 80% 

 ethanol. Juvenile samples were collected with a 26 x 

 26m midwater trawl (0.945cm mesh codend liner), 

 towed horizontally at depth for 15 minutes at a speed 

 of approximately one knot. Sampling for the juvenile 

 survey was at fixed stations spread throughout the 

 study area, with target depths of 10, 30, and 100m, 

 but with most samples taken at the standard depth of 

 30 m. Juvenile rockfish were removed from hauls and 

 immediately frozen. For further details on procedures 

 for the juvenile survey, see Wyllie Echeverria et al. 

 (1990). 



To determine when daily otolith increments begin to 

 form, gestating larvae (rockfish are viviparous live- 

 bearers) were collected and their otoliths examined. 

 Adult female rockfish were collected during the May- 

 June survey and from commercial fishermen at Fort 

 Bragg, California in May. Preextrusion larvae were re- 

 moved and placed in 80% alcohol for later examination. 



Laboratory procedures 



The stage of development of the gestating larvae was 

 determined by morphology and pigmentation (Moser 

 et al. 1977). Larvae from four females (two from the 

 May- June cruise and two collected from commercial 

 fishermen in May) were found to be in an advanced 

 stage of development and were used for further 

 analysis. 



The standard length (SL) of all larvae and juveniles 

 was measured to the nearest 0.1 mm. Ages were deter- 

 mined for juveniles spanning the entire size range col- 

 lected and for all planktonic larvae. Sagittae were 

 removed from each fish and affixed whole to slides with 

 a drop of clear fingernail polish. Intact otoliths from 

 larvae (<20mm) had discernable growth increments 

 with no further preparation. For fish with SL >20mm, 

 otoliths were sanded in the sagittal plane with 400-grit 

 sandpaper until the nucleus became visible. Concen- 

 trated HC1 was used to etch the otoliths until growth 

 increments were easily discernable. 



All otoliths were viewed at 600-1250 x magnifica- 

 tion with a compound microscope equipped with a video 

 camera and monitor. Only fish with a dark check mark 

 that clearly encircled the primordium were used in our 

 analysis of growth (see also Penney and Evans 1985). 

 Fish with this mark and no additional increments were 

 given a nominal age of zero. Increments were counted 

 from this mark to the most distal point along the 

 postrostral growth axis. We used a digitizer to record 

 the exact position of each increment; otolith radius (OR) 

 was measured to the nearest 0.1 ^m. 



