UCHIYAMA and STRUHSAKER: AGE AND GROWTH OF SKIPJACK AND YELLOWFIN TUNAS 



Table l. — Experimental data on marked daily growth increments in skipjack and yellowfin tunas. 



and Kramer 1979), temperature (Taubert and 

 Coble 1977; Methot and Kramer 1979), and age of 

 fish (Pannella 1971; Brothers et al. 1976) have been 

 demonstrated to affect the formation of daily 

 growth rings on the sagitta. In our experiments 

 with nehu, skipjack tuna, and yellowfin tuna, the 

 amount of daily ration appeared to have an influ- 

 ence on otolith growth increments. Fishes fed once 

 daily did not have clear otoliths with countable 

 growth increments. Only when the fishes were fed 

 to satiation throughout the day were countable 

 growth increments formed. In the experiment on 

 the effect of winter conditions on the formation of 

 growth increments by Taubert and Coble (1977), 

 the green sunfish, Lepomis cyanellus, lowered 

 their activity level and fed less when the tempera- 

 ture fell below 10° C. Wild and Foreman (1980) also 

 suggested that the difference in their results be- 

 tween yellowfin and skipjack tunas may have been 

 due to feeding, citing the differences in the occur- 

 rence of full stomachs in the yellowfin and skipjack 

 tunas from the Revillagigedo Islands area 

 examined by Alverson (1963). Thus, it is evident 

 that there is need for further research on the vari- 

 ables affecting day-to-growth increment relation 

 of tuna otoliths. 



RESULTS AND DISCUSSION 



Skipjack Tuna 



A plot of fork length versus age for 51 central 

 Pacific tuna indicated three linear stanzas of 

 growth (Figure 2). A least squares procedure was 

 used to compute the lines of best fit for each stanza 

 (Figure 2, Table 2). Von Bertalanffy growth pa- 

 rameters were also calculated (Figure 2). As there 

 are no means available to statistically compare 

 the von Bertalanffy growth curve, which has three 

 parameters, with the three linear growth stanzas, 

 which have a total of six parameters, the com- 

 parison was performed by examining the distri- 

 bution of residuals (Figure 3). The residuals for 

 the linear growth stanzas were distributed ran- 



TABLE 2. — Length-age regression parameters for the three 

 linear growth stanzas of skipjack tuna: N = number of data 

 and RSS = residual sum of squares. 



domly about the x-axis, signifying a good fit (run's 

 test: Z = 0.39678, P ^ 0.6892). On the other 

 hand, the residuals for the von Bertalanffy curve 

 oscillated about zero and were largest at break- 

 points between linear stanzas and at the midpoint 

 of the stanzas, thus suggesting a deviation from 

 randomness (run's test: Z = -3.32287, P<0.001). 

 Therefore, the linear stanzas appeared to be the 

 preferable growth curve. 



It was noted earlier that the series of three 

 linear growth lines appeared to provide a better fit 

 to our data than the von Bertalanffy growth curve. 

 However, since many of the earlier growth studies 

 use the von Bertalanffy growth curve, it is of 

 interest to compare the von Bertalanffy growth 

 curve of this study with those of other growth 

 studies on the central Pacific skipjack tuna stock 

 (Figure 4). Growth rate estimates by Rothschild 

 (1967) using corrected data of 35 long-term tag 

 returns, by Joseph and Calkins (1969) using 

 Rothschild's uncorrected data, and by Skillman^ 

 using 356 tag returns obtained during a 2-yr 

 period, were all less than the rate obtained in this 

 study. Brock (1954) analyzed the modal progres- 

 sion in length-frequency distributions obtained 

 over a 5-yr period, and derived a growth curve 

 similar to ours. Both curves show similar growth 



^Skillman, R.A. Estimates of von Bertalanffy growth 

 parameters for skipjack tuna, Katsuwonus pelamis, from 

 capture-recapture experiments in the Hawaiian Islands. Man- 

 user, in prep. Southwest Fisheries Center Honolulu Labora- 

 tory, National Marine Fisheries Service, NOAA, Honolulu, 

 Hawaii. 



155 



