4 



30 



H 

 I 



C3 



LLI 



< 



> 

 O 



Oct 79 (88, 0.08) 

 Dec 79 (8, 0.85) 

 Jan 80 (14. 0.51 

 Feb 80 (40, 0.52 

 Mar 80 (54, 0.91 

 Apr 80 (92. 0.77) 

 May 80 (99. 0.40 

 Jun 80 (108, 0.47 

 Jul 80 (91, 0.62 

 Aug 80 (103, 0.36) 

 Sep 80 (74. 0.29) 

 Dec 80 (12, 0.78) 

 Jan 81 (38, 0.84 

 Feb 81 (22, 0.87 

 Apr 81 (43, 0.61) 



FISHERY BULLETIN: VOL. 82, NO. 2 



Sep 80 



Jun 80 



Mar 80 

 Dec 80 



Jan & Feb 81 



Jan 80 



120 150 



TOTAL LENGTH (mm) 



180 



Rgure 2. — Monthly regressions of ovary weight on total length for Larimus fasciatus. Length of each line shows the observed 

 size range. Sample sizes and r 2 values are tabulated for each period. Regressions presented were significant at a = 0.05. 



were calculated by using initial x- intercepts to 

 readjust the initial x- variable (time) scale, so that 

 each final growth curve passed through the origin 

 (Table 4). Final calculated hatching dates are 

 mean values because regressions predict aver- 

 ages. Rate parameters — e.g., regression coeffi- 

 cients, growth/30 d, and von Bertalanffy K 

 values — fitted to observed size data are the same 



within rounding error regardless of the hatching 

 dates used, because curves are fitted to the same 

 time dimension between the initial and last collec- 

 tion of a cohort. 



Total annual mortality rates (1 — S) were calcu- 

 lated on a time-specific and cohort-specific basis 

 from the expression S = N t /N where S = rate of 

 survival, and N t and N are the numbers offish at 



342 



