SPAWNING OF YELLOWFIN TUNA 



57 



This decline is to be expected if there are succes- 

 sive spawnings by each fish. 



Table 4. — Proportionate numbers of ova more than 0.S8 

 mm. in diameter for all fish with ovaries in stage F, 

 June-September 1950 



Month 



June 



July 



August 



September, 



Total 

 number 

 of ova 



2,100 

 2,100 

 1,500 

 1,200 



Number of ova 



over 0.38 mm. 



in diameter 



349 

 385 

 313 

 283 



Probability of 



ova over 0.38 



mm. in 



diameter ' 



0.166 

 0.183 

 0.209 

 0.236 



> The ratio of pooled ova in the two groups for each month. 



In summary, the presence of several groups of 

 ova in the maturing ovaries as indicated by the 

 ova-diameter frequency polygons, the significant 

 correlation between successive groups of maturing 

 ova, the presence of remnants of mature ova from 

 a previous spawning in the lumina of the ovaries 

 at the same time that successive groups are matur- 

 ing, and the decrease in the proportionate numbers 

 of ova composing the intermediate group as the 

 spawning period progressed appear to be ample 

 evidence that individual yellowfin tuna spawn 

 more than once during the breeding season in 

 Hawaiian waters. However, from this study, it 

 is not possible to give any estimate of the number 

 of times an individual fish may spawn during the 

 season or the period of time that elapses between 

 spawnings. 



Among ovary samples collected during Novem- 

 ber and early December 1950, after spawning had 

 terminated, were several with scattered remnants 

 of mature ova in the lumina, indicating that the 

 fish had spawned the previous summer or fall. Ex- 

 cept for these remnants, which consisted of shells, 

 some with the oil globules still discernible, only 

 immature ova were present. Ovaries collected 

 during January, February, and March, 1950, con- 

 tained only immature ova, and no remnants were 

 evident. Thus, it appears that ova in the inter- 

 mediate stage at the close of the spawning period 

 are absorbed and are not carried over to the fol- 

 lowing year. 



RELATION OF OVARY SIZE TO FISH SIZE 

 AS A MEASURE OF MATURITY 



Having defined the stages of maturity according 

 to the position of the mode of the most mature 



group of ova in the ova-diameter frequency dis- 

 tributions, the question then arises : Is there some 

 simpler technique by which these stages might be 

 approximated without the laborious measurements 

 of ova diameters? If a relation exists between 

 ovary weight at particular stages of maturity and 

 fish weight (as a measure of fish size), it would be 

 possible to compare the changes in the growth of 

 the ovaries for fish of comparable sizes. 



One means of estimating this relation is to deter- 

 mine the percentage that the ovary composes of the 

 total weight of the fish. This method was used 

 by Hoek (1895) as a measure of the degree of ripe- 

 ness in Rhine salmon. Masterman (1913a and 

 1913b) later applied the method for estimating 

 sexual maturity in European salmon and smelt 

 (Osmerus eperlanus). Others investigating the 

 spawning of the European salmon have followed 

 Masterman. This method also has been used to a 

 certain extent by Olsen and Merriman (1946) in 

 their study of the spawning of the ocean pout 

 {Macrozoarces americanus) in the North Atlantic. 



While the relation of the ovary weight to the 

 body weight gave a fair approximation of relative 

 maturity in the yellowfin tuna, the ratio expres- 

 sing this relation was not directly proportional 

 over the size range of fish for a particular stage 

 of maturity. When ovary weight was plotted 

 against body weight, 3 great variation among the 

 ovary weights was found for fish of a particular 

 size. These observations appeared to fall into two 

 generally distinct groups, one consisting of fish 

 with ovaries in the immature stages (A, B, and C) 

 and the other consisting of fish with ovaries in 

 the maturing and ripe stages (D through J) ; 

 although some of the values were intermediate 

 between the two groups. 



The relation between body weight and ovary 

 weight for the immature and intermediate stages 

 of maturity (A, B, and C) was found to be linear 

 over the ranges of fish size and ovary size. Fur- 

 thermore, when weights of ovaries in stages B 

 and C were plotted against fish weights, the result- 

 ing regression was nearly identical with the linear 

 regression obtained for fish with ovaries in stage 

 A. Covariance analysis indicated that the two 

 regression coefficients were sufficiently similar 

 (7^=2.30; / > >0.05) to warrant the assumption 

 that they arose from a single homogeneous group 



* Both flsli and ovaries were weighed while fresh. 



