FISHERY BULLETIN: VOL. 84, NO. 4 



females during the spawning season and can be used 

 to identify females approaching the end of their 

 spawning season as well as those in postspawning 

 condition (Hunter and Macewicz 1985b). 



METHODS 



Skipjack tuna were captured either by pole and 

 line or were catches associated with moored fish at- 

 traction devices or free floating natural flotsam. 

 Two sets of collections of skipjack tuna were ana- 

 lyzed: a group of 12 females taken near Noumea, 

 New Caledonia on 23 February 1984; and a group 

 of 87 females taken in 8 different collections at 

 various locations in the South Pacific from 20 Octo- 

 ber to 30 November 1984 (Table 1). Our samples 

 were opportunistically taken and spanned a great 

 latitudinal range (0°-23°S). At present the peak 

 spawning months of skipjack tuna are poorly defined 

 over this range of latitudes. Spawning is known to 

 occur throughout the year in some areas (Nishikawa 

 et al. 1985), but regional differences may exist in 

 the peak months of spawning, and the spawning 

 season also varies with skipjack size (Naganuma 

 1979). Naganuma concluded from analysis of gono- 

 somatic indices (GSI) that peak spawning period for 

 small skipjack tuna (40-60 cm) in the South Pacific 

 is October to December. Argue et al. (1983) ex- 

 amined 11,000 adult skipjack tuna for cannibalism 

 of juveniles (15-70 mm) and for GSI over the same 

 latitude range as this study, but covering 80° of 

 longitude (140°W-140°E). They found that canni- 

 balism and female GSI was highest between Octo- 

 ber and March in this broad area. More data are 

 needed to identify the regional variation about this 

 general pattern. 



The 8 collections of gonads (collections 2-9, Table 



Table 1 .—Characteristics of 9 collections of female skipjack tuna 

 taken in the South Pacific in 1984. 



1) were treated statistically as 8 "clusters" of ran- 

 dom samples of unequal size. The mean proportion 

 of postovulatory follicles <24 h old was calculated 

 as the total number of females with such follicles 

 divided by the total number of mature females. 

 Cochran (1977) pointed out that estimation of vari- 

 ance by the simple binomial probability formula can 

 produce serious errors. The variance was calculated 

 by the appropriate formula recommended by Coch- 

 ran (1977). 



Three female skipjack tuna were spawned in cap- 

 tivity (23°-24°C; June 1985) at the Kewalo Research 

 Facility of the National Marine Fisheries Service 

 using the stress spawning technique of Kaya et al. 

 (1982). One fish (48 cm fork length [FL]) was 

 sacrificed at the time of spawning, another (43.8 cm 

 FL) 12 h later and the third (44 cm FL) 24 h after 

 spawning. The ovaries of these females were used 

 to establish histological criteria for the aging of 

 the postovulatory follicles of the sea-caught 

 females. 



Ovaries were preserved in 10% Formalin 3 and 

 embedded in Paraplast. Histological sections were 

 cut at 5-6 /^m and stained with Harris hematox- 

 ylin followed by eosin-phloxine-B counter stain 

 (H&E). 



Histological Classification 



To estimate reproductive condition of skipjack 

 tuna, we used two histological classification systems: 

 one for estimating spawning frequency and the 

 other for assessing the likelihood that a female will 

 continue to spawn (atretic state of the ovary). Each 

 ovary was classified histologically according to both 

 systems. These classification systems were devel- 

 oped for northern anchovy, Engraulis mordax, by 

 Hunter and Goldberg (1980) and Hunter and 

 Macewicz (1980, 1985a, b) and are used here with 

 a few modifications appropriate to skipjack tuna 

 ovarian structure and their rates of postovulatory 

 follicle resorption. The descriptions of postovulatory 

 follicles of different ages are from the three captive 

 Hawaiian skipjack tuna. As these fish resorbed their 

 postovulatory follicles much more rapidly than did 

 the northern anchovy, we used stages of shorter 

 duration. The atretic classification system remains 

 unchanged, except for a few minor details of histo- 

 logical structure based on our observations of sea- 

 caught fish. We believe that the reproductive inter- 

 pretations we associate with the atretic classes are 



PL = pole and line; PS = purse seine catch of skipjack tuna attracted 

 to either a fish attraction device moored in waters of 350-450 m deep or natural 

 flotsam. 



Latitude and longitude given in degrees and minutes when available. 



3 Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



896 



