I 10 



Fishery Bulletin 94( 



1996 



cal stress, as shown for captive skipjack tuna (Hunter 

 etal., 1986). 



The yellowfin tuna postovulatory follicle is simi- 

 lar to that described for most other marine fishes. 

 The rate of postovulatory follicle degeneration in 

 yellowfin tuna, however, is more rapid than that for 

 fish inhabiting cooler waters. The degeneration pro- 

 cess of the postovulatory follicle, along with the oo- 

 cyte developmental rate, is apparently correlated 

 with temperature. Clarke (1987) showed that the 

 Hawaiian anchovy (Encrasicholina purpurea) 

 spawns at intervals of about 2 d, as compared with 

 about 7 d for northern anchovy (Engraulis mordax) 

 (Hunter and Goldberg, 1980) and about 6 d for the 

 Peruvian anchovy (Engraulis ringens) (Alheit et al., 

 1984), both cooler-water species. The postovulatory 

 follicle can be detected in the northern anchovy for 

 just under 48 hours (Hunter and Goldberg, 1980), 

 whereas for yellowfin and Hawaiian anchovy that 

 period is no more than 24 hours. 



This paper provides a method for direct estima- 

 tion of the frequency of spawning of males by histo- 

 logical examination of testicular tissues. Structural 

 characteristics of the vas deferens, specifically the 

 amount of sperm present, the shape of the duct, and 

 the staining of the epithelium are useful for detect- 

 ing whether a male yellowfin tuna has recently 

 spawned. Specific cellular characteristics have not 

 been utilized in the separation of the testis into 

 prespawning and postspawning groups. Although 

 positive identification of cell types along with nuclear 

 and cytoplasmic features may be of additional value 

 in detection of reproductive activity, this would have 

 required an ultrastructural investigation of cell 

 types, which was beyond the scope of the present 

 investigation. The main limitation to the technique 

 is that spawning cannot be detected more than 12 h 

 after it occurs. Because yellowfin tuna are normally 

 captured during daylight hours in the purse-seine 

 fishery in the eastern Pacific, spawning-frequency 

 estimates from males could be determined only for 

 fish caught during the relatively short time between 

 about 0600 and 1200 h. Spawning-frequency estimates 

 for yellowfin tuna from throughout the eastern Pacific 

 are thus better estimated from the presence of 

 postovulatory follicles that are detectable in ovaries 

 from fish captured throughout the entire day, with the 

 assumption that the males spawn at similar times. 



Batch fecundity 



There is only a short period from the late afternoon 

 until about 2200 h (previous to spawning) when ova- 

 ries with migratory-nucleus or hydrated oocytes are 

 found in yellowfin tuna. McPherson ( 1991) reported 



hydrated oocytes from yellowfin tuna collected be- 

 tween 1200 h and 2000 h in Australian waters, and 

 Hunter et al. (1986) reported the presence of migra- 

 tory-nucleus oocytes from skipjack tuna collected at 

 1955 h in the South Pacific. The use of migratory- 

 nucleus or hydrated oocytes for batch-fecundity de- 

 terminations is crucial because only oocytes in these 

 stages can be distinguished from the less-developed 

 subsequent batch of oocytes. 



The estimate of mean relative fecundity for yel- 

 lowfin tuna obtained from this study (68.0 oocytes 

 per gram of body weight) is considerably less than 

 that of Joseph (1963) ( 106 oocytes per gram of body 

 weight). Analysis of covariance applied to the log- 

 transformed batch fecundity and length data from 

 the present study and the truncated data set from 

 Joseph (1963) indicated a significant difference in 

 the test for equality of slopes (F ()5(1 65) =10.74, 

 P<0.005). Furthermore, the adjusted mean batch fe- 

 cundity of 1,461,465 oocytes for the data from the 

 present study is less than the adjusted mean batch 

 fecundity of 2,454,049 oocytes from the data set of 

 Joseph ( 1963). The differences in these estimates are 

 probably due to the methods used, rather than to 

 spatial or temporal variation. The biases associated 

 with the method used by Joseph created an overesti- 

 mate of the number of oocytes in the spawning batch. 

 The batch fecundities for numerous species of fish 

 have been shown to vary geographically and tempo- 

 rally. This may be a function of inter- and intra-popu- 

 lation variation influenced by both genetic and envi- 

 ronmental components (Wootton, 1979). 



Cost of spawning 



The best estimates of the average daily cost of spawn- 

 ing in yellowfin tuna are 0.97% and 0.28% of the body 

 weight per day in females and males, respectively. 

 Although the two methods employed for estimating 

 the costs of spawning for females are both appropri- 

 ate and produce close estimates, the method based 

 on oocyte weight and relative fecundity is more 

 precise. 



An estimate of total daily energy costs for repro- 

 ductively active yellowfin tuna can be obtained from 

 a bioenergetics approach. Olson and Boggs (1986) 

 estimated a mean daily ration of 5.29^ body weight 

 from a bioenergetics model incorporating energy ex- 

 penditures for swimming, standard metabolism, 

 growth, excretion, egestion, and food assimilation. 

 The estimated mean energy expenditure for growth 

 was about 0.41% of body weight per day. If the esti- 

 mated daily costs of spawning for females and males 

 of 0.97% and 0.28^ of the body weight per day are 

 added to the overall bioenergetics estimate, daily 



