Woodbury and Ralston: Growth rates and birthdate distributions of Sebastes spp. off central California 



531 



distributions of all five species was evident in 1987, be- 

 ing particularly marked in shortbelly, chilipepper, and 

 widow rockfish. What is most striking in these data, 

 however, is the broad coherence among the species in 

 the yearly shifts that occurred in the distributions (Fig. 

 10). For example, 1986 was a year in which successful 

 parturition occurred late in the season for all species 

 examined. Conversely, 1985 and 1988 were years 

 marked by earlier birthdate distributions. 



Principal component analysis demonstrates the great 

 degree to which these data are coherent in time 

 (1985-88). Results show that 92.9% of the variation in 

 mean back-calculated calendar birthdate standard 

 scores was accounted for by the first principal compo- 

 nent (Xi 4.64). Given the tight coupling that occurred 

 between bocaccio and shortbelly rockfish in 1984-85, 

 these values may well have been higher if data had been 

 available for all species in all years (1983-88). First 

 component scores for each year were ranked in de- 

 scending order as follows: 1986 fo 2.93)>1987 (fc 

 0.13)>1985 (£j -1.16)>1988 (ft -1.89). This se- 

 quence mirrors the interannual shifting of distributions 

 evident in Figures 9 and 10. 



Discussion 



Growth 



Growth rates of field-collected larval and pelagic juve- 

 nile Sebastes spp. have been determined by others 

 using otolith microstructure (reviewed in Kendall and 

 Lenarz 1987, Moser and Boehlert 1991). It is note- 

 worthy that in all cases a linear growth model fitted 

 the data best, similar to results reported here (Figs. 

 2, 4-7). Data for S. melanostomus (N 4) indicated that 

 at 15-30 mm SL the growth rate was 0.24 mm/day 

 (Moser and Ahlstrom 1978). Research on S. diploproa 

 (Boehlert 1981) showed a growth rate of 0.19 mm/day 

 for fish 10-40 mm SL. Both studies were conducted off 

 southern California. Penney and Evans (1985) esti- 

 mated that redfish (composite of S. marinus, S. men- 

 tella, and S. fasciatus) from the northwest Atlantic 

 grew at the rate of 0.11-0.16 mm/day when in the 

 range 8-25 mm TL. 



These growth rates are somewhat lower than those 

 presented in Table 2. Our data show growth rates rang- 

 ing 0.52-0.64 mm/day for shortbelly rockfish, 0.56-0.97 

 mm/day for bocaccio, 0.40-0.56 for chilipepper, and 

 0.30-0.61 mm/day and 0.19-0.46 mm/day for widow and 

 yellowtail rockfishes, respectively. This is likely due to 

 the span of ages and/or sizes caught, since our fish were 

 larger and older than in these other studies. It is con- 

 ceivable, however, that the growth rates of pelagic 

 juvenile Sebastes spp. along the central California coast 



140 



120- 

 100 

 80- 

 60 

 40 

 20 

 



-20 



1982 1983 1984 



1985 1986 



Year 



1987 1988 1989 



Figure 10 



Annual trends (1983-88) in the means of back-calculated 

 birthdate distributions for shortbelly, bocaccio, chilipepper, 

 widow, and yellowtail rockfishes. 



are enhanced in response to superior feeding conditions 

 and increased ration (Brett 1979), a consequence of 

 elevated production associated with coastal upwelling 

 (Smith 1968). Dietary studies of these fishes show that 

 they feed primarily on various life stages of euphausiids 

 and calanoid copepods (Reilly et al. In prep.). 



Interannual variation in growth performance has 

 been implicated as a major factor affecting year-class 

 strength (Miller et al. 1988). Houde (1987) showed with 

 simple calculations that, in some situations, a doubling 

 of the instantaneous growth rate during the early life 

 history can increase the number of recruits by nearly 

 100-fold. Moderate and significant interannual varia- 

 tions in growth rates/performance were evident for the 

 species in this study. Likewise, widow rockfish year- 

 class strength has varied 10-fold during the 1974-89 

 period (Hightower and Lenarz 1989). These results are, 

 therefore, consistent with the hypothesis that fluctua- 

 tions in rockfish recruitment are due, at least in part, 

 to interannual variation in growth rate. 



Birthdate distributions 



Back-calculation of birthdate distributions using daily 

 increment data has been reported before (e.g., Bolz and 

 Lough 1983, Brothers et al. 1983, Methot 1983, Jones 

 1985, Penney and Evans 1985, Thorrold 1989). One 

 difficulty with this approach is that estimates of the 

 numbers born by date should be adjusted for differ- 

 ences in mortality (Penney and Evans 1985), i.e., to 

 have appeared in our samples, older juveniles must 

 have survived for a longer period of time than younger 

 ones. However, without an estimate of early larval and 

 juvenile mortality rates we were unable to adjust for 



