FISHERY BULLETIN: VOL. 74, NO. 2 



a single-school sample of seven containing no 

 "fused" individuals is aboutO.Ol (= 0.5''). If schools 

 consisting only of juveniles were common, many 

 more all-juvenile samples would have been en- 

 countered. Conversely, juveniles (two-tone, speck- 

 led, and/or mottled) occurred in all but 3 of the 324 

 samples. It must be concluded that juveniles prob- 

 ably do not school separately in the eastern 

 Pacific. Another possibility, albeit unlikely, is that 

 all-juvenile schools exist but are not captured by 

 tuna seiners. 



COMPARISON WITH 

 THE JAPANESE POPULATION 



Many of the estimates of life history parameters 

 presented here differ from those published by 

 Kasuya et al. (1974) for the relatively unexploited 

 population of S. attenuata in Japanese waters (Ta- 

 ble 12). The differences could be caused by 1 ) differ- 

 ential procedures or analytical methods, 2) in- 

 trinsic racial differences between the populations, 

 or 3) differential population status, e.g., exploited 

 versus unexploited. The comparisons below of 

 similarly calculated average estimates, of course, 

 rest on the assumption that the overall samples in 

 both cases were not biased with respect to age or 

 sex. The major sampling differences between the 

 two studies is that the Japanese sample consisted 

 mostly of large samples from a few schools. 



whereas our sample consisted mainly of aggre- 

 gated, small samples from many schools. Both 

 studies assume no sampling bias. Comparison of 

 large, single-month samples in the present study 

 with large, single-school samples in the Japanese 

 study (e.g., the October 1972 sample in Figures 15 

 and 16 with sample number 2 in Figure 2 of 

 Kasuya et al. 1974) indicate very similar length- 

 frequency distributions and support the idea that 

 the aggregated samples are probably not biased, 

 or, if biased, are biased in the same way. This 

 inference is, of course, based on the assumption 

 that the underlying population structures are 

 about the same in the two populations. 



The estimate of Kasuya et al. ( 1974) of length at 

 birth was based on only 5 full-term fetuses and 

 newborn calves versus 86 in the present study. 

 Our estimate can, therefore, be considered more 

 reliable, although the possibility does exist that 

 length at birth is greater in the Japanese popula- 

 tion. The difference between the estimated 

 lengths at 1 yr for the two populations is about the 

 same as the difference between the estimates of 

 length at birth. Estimated length at attainment of 

 sexual maturity and maximum length (for males) 

 are also greater for the Japanese samples. The 

 estimate of length at maturity of males is greater 

 in spite of the fact that Kasuya et al. used a lower 

 testis-weight criterion than we did (68 versus 100 

 g). The average lengths of both adult males and 



Table 12. — Comparison of estimates of average life history parameters of Stenella attenuata by Kasuya et al. (1974) and 



in present paper. 



266 



