BLACKBURN AND SERVENTY: DISTRIBUTION AND LIFE HISTORY OF SKIPJACK TUNA 



method (e.g., speed, lure type, lure size) between 

 that period and later. Robins (1975) took similar 

 fish in the period April-June off Western Aus- 

 tralia, mostly by purse seining but occasionally by 

 trolling. 



Weight as well as length was measured for 607 

 east coast skipjack tuna. The following significant 

 linear regression was found between the common 

 logarithms of the variables: 



log W = -6.0762 + 3.5202 (logL) 



where W is weight (grams) and L is fork length 

 (millimeters). The coefficient of determination, 

 i? -, is 0.856. Standard errors of the first and second 

 constants in the equation are 0.1595 and 0.0586, 

 respectively. The equation is equivalent to: 



W = 0.000000839 L 



3.5202 



The range of L in the data used is 410-645 mm. 

 The heaviest fish weighed 5.67 kg. 



The 95% confidence limits of the regression co- 

 efficient are 3.4048 and 3.6356, calculated from 

 the standard error. Other published regressions of 

 skipjack tuna weight on length for large samples 

 ( >200) indicate regression coefficients from 3.2164 

 to 3.67. Those samples were taken in the eastern, 

 central, and northwestern Pacific (Nakamura and 

 Uchiyama 1966, and references there) and off 

 North Carolina (Batts 1972). Standard errors of 

 the coefficients were not published in most cases. 

 The standard error can be calculated from data of 

 Hennemuth (1959), for a regression coefficient of 

 weight on length for 1,280 skipjack tuna from the 

 eastern Pacific (combined areas). The coefficient 

 was 3.336, with 95% confidence limits 3.296 and 

 3.376. Thus the east Australian and eastern 

 Pacific regressions are significantly different at 

 the 5% level of probability. The meaning of this 

 difference is not clear. The two groups of skipjack 

 tuna probably belong to different populations 

 (Fujino 1972; Sharp 1978). However, Hennemuth 

 (1959) found regression coefficients from 3.144 to 

 3.555 in different areas of the eastern Pacific north 

 of the Equator, a region considered to contain only 

 one skipjack tuna population (Fujino 1972; Sharp 

 1978), and some of those coefficients were signifi- 

 cantly different. 



SEXUAL CONDITION 



The gonads of 418 east coast skipjack tuna were 



weighed. Ovary weights ranged from 4 to 30 g. 

 Most ovaries were white to pink. Discrete small 

 ova were visible to the naked eye in some ovaries, 

 but large yolked ova were not observed. Ovaries of 

 a reddish flaccid appearance, which might have 

 been spent, were seen occasionally from April to 

 August in New South Wales and Tasmanian 

 waters. Testes were small, weighing mostly 1-2 g 

 with a maximum of 13 g. Milt could sometimes be 

 expressed from them by pressing. Similar obser- 

 vations were made on a small number of skipjack 

 tuna from the west and northwest coasts of Aus- 

 tralia, except that no gonads were weighed. A 

 specimen taken off Fremantle in July was possi- 

 bly spent. No gonad data are available from South 

 Australia. 



Orange (1961) compared ovaries of skipjack and 

 yellowfin tunas by means of a "gonad index" equal 

 to 



3 8 



(gonad weight) / (fish length ) 10 



with gonad weight in grams and fish length (fork 

 length) in millimeters. This is a ratio between 

 gonad weight and an estimate offish weight. The 

 estimate is not accurate for skipjack tuna, since 

 weight increases with fork length to some power 

 slightly higher than 3 in that species, as noted 

 earlier. However Orange also compared gonad in- 

 dices with the appearance of the ovaries and ova, 

 and found that only indices over 30 indicated ap- 

 proaching sexual maturity in skipjack tuna. 

 Naganuma (1979) made similar comparisons 

 which indicated that spawning skipjack tuna have 

 gonad indices of 80 or higher, measured on 

 Orange's scale. Thus calibrated, the index has 

 some utility, and it has been employed by other 

 skipjack tuna investigators. None of the gonad 

 indices in our east Australian material reached 30 

 (Table 2); only 2, out of 224, were slightly over 20. 

 Thus no females appeared to be mature on the 

 basis of gonad index, confirming the observations 

 on the gonads themselves. Yet virtually all these 

 skipjack tuna were at or over the size at which first 

 sexual maturity has been found in other Pacific 

 waters, i.e., about 45 cm (Kearney et al. 1972; 

 Blackburn and Williams 1975; Naganuma 

 1979). 



It is clear from these observations that skipjack 

 tuna do not spawn to any significant extent in east 

 Australian coastal waters, and there is no evi- 

 dence that they spawn in any Australian coastal 

 waters; nevertheless, they do spawn in the 



91 



