Table 4.— Sample regressions of ratio of round to dressed weight on length. 











n' 







Mean ratio at 



Sample 



Year 



Month 



No. 



e C 



(b' -b) 



S(b' -b) 



mean length 



Gutted 

















1 



1942 



Apr. 



46 



-1.140 



0.301* 



0.122 



1.16:1 



2 



1953 



May 



29 



-0.419 



0.151* 



0.061 



1.16 



1 



3 



1953 



June 



22 



0.151 



0.010 



0.072 



1.21 



1 



4 



1953 



June 



20 



0.147 



0.006 



0.080 



1.19 



1 



5 



1953 



Dec. 



34 



0.098 



0.009 



0.092 



1.14 



1 



6 



1954 



Jan. 



25 



0.052 



0.017 



0.203 



1.13 



1 



7 



1954 



Jan. 



22 



-1.075 



0.291 



0.176 



1.12 



1 



8 



1954 



Feb. 



23 



-0.066 



0.059 



0.180 



1.18 



1 



9 



1954 



June 



39 



-0.314 



0.122 



0.070 



1.14 



1 



Total 



0.129 



0.004 



0.020 



1.16:1 



Gutted and 



gilled 















1 



1942 



Apr. 



21 



-0.621 



0.192* 



0.060 



1.17:1 



2 



1954 



Apr. 



46 



-1.171 



0.333 



0.208 



1.22 



1 

 1 



Total 



-0.595 



0.187* 



0.061 



1.20 



*Significantly greater than zero (/>^=05) 



the desired conversion factor. Three of the 11 

 samples were found to have slope values (b'- b) 

 significantly greater than zero, and all samples 

 had positive slopes. The slightly positive slopes, 

 when extrapolated to zero length, gave negative 

 or very low intercept values, which means a 

 ratio of round to dressed less than or near 

 unity, even though the total regression coeffi- 

 cient was not significantly greater than zero. 

 Therefore, because landed fish range only from 

 40 to 80 cm, it is appropriate to use the mean 

 ratios of round to gutted weight at the mean 

 length of the samples (Table 4). No seasonal 

 trends were evident. Thus, the overall ratio of 

 1.16:1 appears to be the best available estimate 

 for converting gutted to round weights. The 

 overall ratio estimated for converting gutted 

 and gilled weight to round weight was 1.20:1. 

 In order to use the length-weight equations 

 to estimate round weights, the following ad- 

 justments should be made: 



log e Y = log e C + 0.1442 + b log e X 

 for gutted, and 



log e V = loge C + 0.1857 + b log e X 

 for gutted and gilled 



Loge C is the intercept and b the coefficient of 

 the regression of dressed weight on length. 



CONCLUSIONS 



Several conclusions 

 these analyses: 



were evident from 



1 . Subsample differences were significant. 



2. Large differences existed among sam- 

 ples (trips) within strata. 



3 . The sorting of fish into scrod and large 

 categories produced significantly offset 

 regression lines. 



4. Year-to-year changes were not sig- 

 nificant. 



5. Samples within Georges Bank and Nova 

 Scotian regions were homogeneous. 



6. Differences were found between the 

 Georges Bank and the Nova Scotian 

 region. 



