614 



Fishery Bulletin 89(4), 199! 



o 



c 

 cu 



cr 

 cu 



150 



100 



Dec Jan P>b Mar Apr 



Predicted Date of Birth 



Figure 4 



Frequency distribution of birthdates (date of extrusion check) 

 for all shortbelly rockfish caught during the May- June survey 

 off central California, 1989. Note the existence of a small but 

 discrete second cohort spawned in March and April. 



smallest planktonic larvae, we believe this mark was 

 formed at, or shortly after, extrusion. The size distribu- 

 tion of the few planktonic larvae that lacked this 

 feature (henceforth called the extrusion check) was 

 similar to the size distribution of full-term, gestating 

 larvae (Fig. 1). Thus, shortbelly rockfish do not appear 

 to grow substantially from the time of extrusion to the 

 time the extrusion check is formed. Likewise, all 

 planktonic larvae greater than 5.5 mm in length had 

 an extrusion check and more than 50% had the extru- 

 sion check, at a size of 4.7 mm (Fig. 2). The average 

 size of the full-term gestating larvae we measured 

 was 4.7 mm, and we take this as an estimate of the 

 size-at-extrusion, which is independent of our growth 

 models. 



Validation 



We compared how the mean SL of fish in a cohort 

 changed over a 30-day period, with the growth rate 

 estimated from a regression of SL on nominal age 

 determined independently by examining otoliths. For 

 the 55 fish we aged from the extended May-June 

 cruise, a linear relationship existed between assumed 

 daily age and SL, with a slope of 1.82 days/mm (SE 

 0.0484) and an r 2 of 0.963 (Fig. 3). The distribution of 

 estimated birthdates for all juveniles caught during the 

 cruise (N 1262) contained a small secondary mode 

 (N 25) in March and April (Fig. 4). In our primary 

 analysis, we considered only those fish with birthdates 

 occurring in January or February (N 1237). 



