BOWERING: AGE, GROWTH, AND MATURITY OF GREENLAND HALIBUT 



the mean sizes at age, however, giving equal weight to 

 each age may overestimate if given the lower num- 

 bers and questionable age readings in the very 

 large fish. 



Back-Calculated Growth Curves 



The direct proportionality method (Lea 1919) for 

 back-calculating size at age from otoliths appears 

 valid for Greenland halibut (Fig. 4). The linear least 

 squares regression for total fish length on total oto- 

 lith length based on 123 observations yielded a cor- 

 relation coefficient (r) of 0.98 and a i-value for r of 

 48.59 which is highly significant. 



Back-calculated growth curves up to age 5 of the 

 1972 year class show very close agreement (Fig. 5) 

 between males and females, particularly for the more 

 southerly areas where the curves essentially coincide 

 (Fig. 5). A covariance analysis of males versus fe- 

 males, however, indicated no significant difference in 

 the males and females for both the regression coef- 



10 12 14 16 18 20 22 24 26 28 30 32 34 36 38 40 42 44 46 48 50 



TOTAL OTOLITH LENGTH ( micrometer units) 



FIGURE 4. — Linear least squares regression of fish length against to- 

 tal otolith length of Greenland halibut. 



ficients (slopes) or adjusted means (y- intercepts) 

 throughout the range under consideration (P= 0.44). 

 The sexes, therefore, were combined. A covariance 

 analysis on the slopes of the fitted lines for the sexes 

 combined was performed and indicated that, with the 

 exception of the Baffin Bank data, the regression 

 coefficients (slopes) were not significantly different 

 (P >0.05). This would indicate that the growth rates 

 in these areas were similar. Paired comparisons using 

 £-tests to test for differences between the adjusted 

 means indicated that both the Gulf of St. Lawrence 

 data and the Nain Bank data were significantly dif- 

 ferent from those of all other areas. The Saglek Bank 

 data differed from those of the Northeast Newfound- 

 land Shelf but did not differ from either the Ham- 

 ilton Bank data or the northern Grand Bank data. 

 There were no significant differences amongst the 

 adjusted means of the Hamilton Bank, Northeast 

 Newfoundland Shelf, and the northern Grand Bank 

 data. The combined back-calculated growth curves 

 (Fig. 6) show the Baffin Bank growth rate to be the 

 slowest with a considerably smaller size at age. The 

 largest size at age is found in the Gulf of St. Lawrence 

 with all other areas in the midrange. 



Sexual Maturity 



The sexual maturity curves (Fig. 7) indicate a clear 

 shift to the right in the curves going from northern 

 Labrador to southern Labrador with the Northeast 

 Newfoundland Shelf and southern Labrador curves 

 approximately the same. The curve for the Baffin 

 Bank area falls near the middle (Fig. 7), close to that 

 of the Nain Bank area. The Gulf of St. Lawrence 

 curve, however, is well to the left and appears nearly 

 isolated from all others. 



The results of the probit transformation analysis of 

 sexual maturity data, by area are shown in Table 1. 

 All chi-square tests indicate acceptance of the fitted 

 lines to the observed data at the 5% significance 

 level. The length at the 50% maturity level, or M 50 , 



Table 1. — Results of probit analyses of sexual maturity data for the Greenland 

 halibut, by area, with results of x 2 tests for acceptability of the fitted lines to the 

 observed data. 



605 



