13 



a 

 z 

 s 

 o 



DRESSED WEIGHT (Kg) 



FIGURE 2. — Functional (GM) regression of round weight on dressed weight for 685 western Atlantic 

 bluefintuna 1974-77. (Number of fish indicated, star signifies number >9.) 



Table 2.— Regression statistics for logio round weight (Yi - log,o straight fork length (X). round weight 

 (Y) - dressed weight (XA and straight fork length (Y) - curved fork length (Xj of western Atlantic bluefin 

 tuna. Weights in kilograms and lengths in centimeters. 



Ix' 



\~v2 



2.xy 



Syx' 



Log,o round weight - log,o straight fork length 



3,578 2 19254 187739 222 745 1.898 17 648 054 00356051 0,00399809 



Round weight dressed weight 



685 256.993 325 158 832.635 1.398.650 1.009.090 257 261 0175776 



Straight fork length - curved fork length 

 606 271.477 259.444 120.979 112.262 103,959 37.9615 0.0177140 



mates of weight from length differed most from 

 estimates which I calculated using equations of 

 Rodriguez-Roda (1964, 1971). The largest varia- 

 tion (12%) was found for a prespawning fish 

 measuring 48 cm. 



No size range was reported by Mather et al. 

 (1974) for estimating length from weight. How- 

 ever, estimated length corresponding to the ex- 

 tremes and middle of the size range in weight I 

 studied agree closely to values I calculated using 

 their regression equation for Newfoundland, with 

 the greatest difference being only 39r for a 5 kg 

 fish. A greater difference (13%) was noted when 

 comparing estimates from my functional (GM) re- 



gression with estimates obtained using their re- 

 gression equation for the Bahamas for a 5 kg fish. 

 This large difference may have resulted from their 

 not including fish in this size range when calcu- 

 lating their equation because differences at the 

 middle and upper end of my size range were small, 

 4% or less. There appears to be a typographical 

 error in the equation these authors gave for 

 bluefin tuna from Libya, so no comparison was 

 made. 



My functional (GM) regression estimates of 

 round weight from dressed weight agree well with 

 the estimates I obtained using the regression 

 equation of Mather et al. (1974). The largest dif- 



998 



