584 



Fishery Bulletin 102(4) 



Cmax 



LSM= X Uft-ft-i), 



Cmin 



where Cmax = the length or weight class of the largest or 

 heaviest sexually immature animal; 

 Cmin = the length or weight class of the smallest 

 or lightest sexually mature animal; 

 L = the lower value of the length or weight 



class t; and 

 f t = fraction of mature animals in the length 

 or weight class t. 



The specimens were pooled into length and weight inter- 

 vals of 4 cm and 4 kg, respectively. 



The estimated variance of this method is also modi- 

 fied and is calculated as 



var(MS) = M> 2 £ [(/; (1 -/",)/ A^, -l]. 



(x=33.6 mm), respectively. The weight and length of the 

 right testes ranged from 0.17 to 9.98 g (.v=2.62 g) and 

 from 17.9 to 60.0 mm (.v=34.5 mm), respectively. The 

 relationship of testes weight and testes length resulted 

 in significant regression (P<0.0001) and correlation 

 <r 2 =0.91; F=823.9; P<0.0001; y=0.000012x 333 ). The male 

 with the heaviest relative testes weight was 141.6 cm in 

 length and 31.2 kg in weight, and its combined-testes 

 weight was 20.1 g, which is 0.064% of its total weight. 

 The mean of the relative testes weight from 23 mature 

 males was 0.036% of their total weight. 



The testes of the franciscana are characterized by a 

 high lateral symmetry. There was no statistically dif- 

 ference in weight (£=-0.09; P=0.93; n=ll) and length 

 (*=-0.4; P=0.69; « = 100) between testes of the same 

 animal. A strong correlation was found between left 

 and right testes length (6 = 0.95; F=1073.0; r 2 = 0.92; 

 P<0.0001; ra=100; y = 1.232 x 095 ) and between left 

 and right testes weight (6 = 0.99; F=7262.8; r 2 = 0.99; 

 P<0.0001; ra=71; y = 1.02.v 1 °i, where x and y represent 

 values of the left and right testis, respectively. 



where N : = the number of specimens in the length or 

 weight class f; and 

 w = the interval width, a constant equal to 4 in 

 these cases. 



For estimating age, length, and weight at sexual matu- 

 rity, pubertal animals were grouped together with imma- 

 ture animals. 



Results 



The weight and length of the left testes ranged from 

 0.23 to 10.42 g (.r=2.60 g) and from 15.7 to 59.7 mm 



Seminiferous tubule diameter 



A nonlinear regression demonstrated positive allometry 

 (6>0.333) of the seminiferous tubule diameter to the 

 combined testicular weight (6 = 0.39; 95% CI=0.35-0.44) 

 (Fig. 2), and a strong correlation between these two vari- 

 ables (F=343.6; r 2 = 0.86; P<0.0001; y=59.4.x° 39 ). 



The relationship between the seminiferous tubule di- 

 ameter and testes length is shown in Figure 3 and the 

 relationship between the seminiferous tubule diameter 

 and standard length is shown in Figure 4. In immature 

 males, there was almost no increase in the seminif- 

 erous tubule diameter with the increase of standard 

 length (0.26 jim/cm) and total weight (0.5 ^m/kg). In 

 mature males, however, seminiferous tu- 

 bule diameter was significantly correlat- 

 ed with standard length (6 = 1.06; F=4.4; 

 r 2 = 0.18; P=0.048; y = 1.4775.v-43.572) 

 and there was no correlation with total 

 weight (6 = 0.23; P=1.28; r 2 =0.07; P=0.27; 

 y = 1.6132.\-+108.54). 



The differences of the seminiferous tu- 

 bule mean diameters were statistically 

 significant between immature, pubertal, 

 and mature male franciscanas (ANOVA, 

 Fs=255.4; df=87; P<0.001). 



Combined-testes weight and length, 

 and sexual maturity 



There was almost no increment in mass 

 of the combined-testes weight in imma- 

 ture dolphins. An increment of only about 

 2.0 g in the combined-testes mass was 

 observed in animals of 70.0 to 125.0 cm in 

 length. For dolphins about 120.0-130.0 cm 

 in length, the combined-testes mass sud- 

 denly increased (Fig. 5), indicating the 



