524 



Fishery Bulletin 100(3) 



5. 8000 - 



W 4000 - 



10000 20000 30000 40000 



Stock biomass (t) 



20 



 I I ' I ' r I 



1970 1975 1980 1985 1990 1995 2000 

 Year 



— •— Observed 



O Normal 

 -▼- LognormalJ 

 Gamma 



015 - 



00 



-0 25 00 25 50 75 1 GO 1 25 

 Intrinsic population growth rate {rj 



Figure 2 



(A) Surplus production estimates P^, and 

 biomass estimates B^, for Georges Bank 

 yellowtail flounder during 1973-97, The 

 biomass estimates and parabolic curve 

 fitted to production estimates are from 

 Cadrin.'' (B) Time series of annual intrin- 

 sic rate of growth (r^, I parameter value 

 estimates from Equation 6. (C) Probability 

 distributions for r^, values used in simula- 

 tions for Georges Bank yellowtail flounder. 



age 55 yr, assumed M=0.055/yr), relatively unproductive 

 stock (r^,=0. 027-0. 037). Cowcod rockfish are much less pro- 

 ductive than Georges Bank yellowtail flounder because 

 of their long lives and slower growth and because adult 

 habitat is limited to steep rocky areas in relatively deep 

 water (90-500 m, Butler et al.^). Production process errors 

 show a high level of autocorrelation (p=0.83-0.94. Table 

 3). Means, variances and autocorrelations for r^, were not 

 very sensitive to assumptions about 7<^ (Table 3). There was 



1000 2000 3000 4000 

 Stock biomass (t) 



1950 1960 1970 1980 1990 2000 



Year 



-•— Observed 

 O Normal 

 ▼- Lognormal 

 — Gamma 



-0 02 00 02 04 06 08 10 

 Intrinsic population growth rate {r^) 



Figure 3 



(A) Surplus production estimates P^ and 

 biomass estimates B^, from Butler et al.'' 

 for cowcod rockfish during 1951-97 in the 

 Southern California Bight. The parabolic 

 curve fitted to production estimates (P^,) 

 shows trends only and is not recommended 

 for management purposes. (B) Time series 

 of annual intrinsic rate of growth (r^,) 

 parameter value estimates from Equation 

 6. (C) Probability distributions for r^, values 

 used in simulations for cowcod rockfish. 



no clear relationship between surplus production and bio- 

 mass, but Pj, was lowest at the highest and lowest B^, lev- 

 els and autocorrelation in production process errors was 

 obvious (Fig. 3A). The distribution of r^, values (Fig. 3B) 

 was skewed to the right and there were no negative val- 

 ues. Gamma distributions fitted by maximum likelihood 

 and the method of moments (Appendix 1) were similar in 

 shape (Fig. 3). 



