HUNTER ET AL.: SPAWNING FREQUENCY OF SKIPJACK TUNA 



might reactivate their ovary sometime later in the 

 year if their physiological condition favored repro- 

 duction. Evidence for northern anchovy indicates 

 that the transitions from spawning to postspawn- 

 ing states and vice versa can occur rapidly. In the 

 laboratory at 16 °C, northern anchovy can resorb all 

 advanced oocytes within a few weeks (Hunter and 

 Macewicz 1985b) and can produce an active ovary 

 in 30 d (Hunter and Leong 1981). Owing to the 

 higher water temperatures and high metabolism of 

 skipjack tuna they are probably capable of even 

 faster reproductive cycling. 



Histological examination of females taken late in 

 the day (1955 h, collection 9) provided additional 

 evidence for daily spawning. Eight of 10 females 

 with postovulatory follicles in this collection also had 

 oocytes in the migratory nucleus stage. This stage 

 is the precursor to hydration. Thus, fish which had 

 spawned <24 h before were beginning to hydrate 

 a new batch of eggs which presumably would be 

 spawned in <12 h. The migratory nucleus stage was 

 observed only in this collection probably because it 

 was the only one taken in the evening, whereas all 

 others were taken in the morning (0645-0755). The 

 rarity of females with hydrated oocytes in our col- 

 lections and the age of the postovulatory follicles 

 imply that spawning usually took place at night. 

 Spawning in daylight hours has been observed by 

 fishermen and scientists, however (Iverson et al. 

 1970; Matsumoto et al. 1984). 



A single female taken during the morning (collec- 

 tion 8) had small (0.70 mm) early stage hydrated 

 oocytes (hydrated oocytes in which the yolk globules 

 had not fully fused). This female, the only one with 

 hydrated oocytes in our collections, also had new 

 postovulatory follicles despite the fact that the 

 hydrated oocytes were not fully advanced. This 

 female may have been induced to hydrate and spawn 

 by the stress of capture or may be simply an excep- 

 tion to the rule. To capture significant numbers of 

 females with hydrated oocytes would probably re- 

 quire sampling after 2100 h. It is important to cap- 

 ture eventually some females in the hydrated stage 

 because it is the best way to confirm that all oocytes 

 in the most advanced modal group, the group of 

 oocytes considered to be the next spawning batch 

 (Hunter and Goldberg 1980), are in fact spawned. 

 Counts of hydrated eggs are also the easiest and 

 most accurate method of estimating batch fecundity 

 (Hunter et al. 1985). 



The "stress" spawning technique of Kaya et al. 

 (1982) was used to produce the spawned skipjack 

 tuna for the aging of postovulatory follicles. In this 

 technique females captured at sea and placed in a 



tank spawn spontaneously, usually about 8 h after 

 capture presumably because of the stress of capture 

 and handling. Spawning typically takes place at 

 about 2400 h, which, by our estimate, appears to be 

 close to the usual time of spawning. It now seems 

 likely that many of these fish are naturally express- 

 ing their daily spawning activity. On the other hand, 

 eggs less than the normal size range, 0.8-1.17 mm 

 (Matsumoto et al. 1984), are occasionally spawned, 

 indicating that stress may induce premature hydra- 

 tion in some individuals. That the skipjack tuna do 

 not continue to spawn in the tanks is due probably 

 to the stress of captivity. Our examination of a cap- 

 tive skipjack 24 h after spawning indicated that 

 nearly all remaining oocytes containing yolk were 

 in the early stages of alpha atresia (Fig. 2e). Similar- 

 ly, female northern anchovy nearly always resorb 

 their advanced oocytes a few days after capture 

 although they will subsequently mature and spawn 

 (Leong 1971; Hunter and Macewicz 1985b). 



If female skipjack tuna spawn at the frequency 

 we observed (85% of the females per day), the cost 

 of reproduction and annual fecundity will be high 

 because skipjack tuna appear to have a long spawn- 

 ing season. The relative batch fecundity of skipjack 

 (number of eggs per spawning per body weight) is 

 about 100 eggs per gram (Matsumoto et al. 1984; 

 Goldberg and Au 1985). Skipjack tuna eggs are 

 about the same size as those of Scomber japonicus 

 which average in weight 0.04 mg (unpubl. data, Na- 

 tional Marine Fisheries Service, Southwest Fish- 

 eries Center). We estimate the cost of a single 

 spawning (excluding the metabolic cost of egg 

 maturation and reproductive behavior) to be about 

 2% of the body weight per spawning (Scomber egg 

 dry weight x relative batch fecundity x conversion 

 to wet weight; 4 x 10 ~ 5 x 100 x 5 = 0.02). If a 

 female spawned every 1.18 d over 3 mo (90 d), it 

 would produce about 7,600 eggs per gram body 

 weight at an average daily cost of 1.7% of the body 

 weight per day; a 4 kg skipjack tuna would spawn 

 about 30 million eggs over this period. 



If the collections used in this study were an un- 

 biased sample of the South Pacific skipjack tuna 

 population, then little doubt exists that spawning 

 occurs almost daily when they have active ovaries. 

 This preliminary study provides the tools necessary 

 for a population-wide assessment of reproduction. 

 We established the time-specific, histological criteria 

 for assessment of spawning rate, and the method 

 was applied to a small sample. A great deal more 

 remains to be done for a proper assessment of 

 reproduction in skipjack tuna. Specifically, many 

 more samples at different times of day, using a 



901 



