PERRIN ET AL.: GROWTH AND REPRODUCTION OF SPOTTED PORPOISE 



220 

 200 



180 



160 



100-/ 



60 



1 2 3 4 5 6 7 8 9 10 II 12 13 14 15 16 17 18 

 POSTNATAL DENTINAL LAYERS (number) 



Figure 12. — Fit of the double Laird growth model (see text) to 

 2-cm mean values of body length on number of postnatal 

 dentinal layers for males and females of Stenella attenuata 

 from the offshore eastern Pacific. For samples greater than 

 30, ± standard errors indicated as vertical line. 



For females with 6 or more layers, the growth 

 equation is 



' 0.0657 



L = 159 exp 



0.3707 



[l - exp (- 0.3707U - 5.588))] 



In this case, both the growth rate and the rate of 

 decay of growth are sharply lower than for 

 juveniles. 



The fit for males is not as good (Figures 1 1, 12) as 

 it is for females, probably due to greater variabil- 

 ity and to inadequate sample sizes for the two 

 oldest strata (the tooth-reading effort was concen- 

 trated on females because of their importance in 

 population dynamics). Another possible explana- 

 tion for the relatively poorer fit for males is that 

 growth (real, or as inferred from tooth layers) in 

 adult males is more complex than in adult 

 females, and a model more complex than the Laird 

 model is called for. Inferred asymptotic length is 

 206 cm, achieved at predicted age of 26 layers. 

 Average length of adult males (defined as those 

 having testes weighing 200 g or more) is 200.7 cm, 

 based on a sample of 253 (Perrin 1975). The largest 

 male of 1,083 measured was 226 cm long. The 

 growth equation for males with 6 or more layers is 



L = 159.5 exp 



0.0524 



, 0.2032 

 [l - exp (- 0.2032(^ - 5.588))] 



The secondary growth rate (a, 0.0524) is very 

 slightly smaller than for females, but the rate of 

 decay {a, 0.2032) is sharply smaller, reflecting the 

 attainment of greater size in males. The equations 

 rearranged and reduced for estimating age (in 

 terms of layers) from length are 



^(M and F <160 cm) = -1.394 In (7.531 



- 1.48 In L) 

 = 5.588 - 2.698 In (29.606 



- 5.64 In L) 

 = 5.588 -4.921 In (20.669 



-3.878 In L). 



^(F ^160 cm) 

 ?(M^160cm) 



Note: These equations should not be used to esti- 

 mate age from length except for grouped samples 

 of smaller animals (about 180 cm or less), for 

 which growth rate is still large compared to indi- 

 vidual variation in length. 



The juvenile growth curve based on tooth layers 

 can be calibrated for the first year by comparison 

 with the growth curve derived from analysis of 

 modal progression (above) and by deduction from 

 what is known about juvenile growth of other 

 odontocetes (the fetal-postnatal growth argument 

 above). Estimated average length at 8 mo based on 

 analysis of modal progression is 125.5 cm. The 

 predicted number of layers at that length (Figure 

 12) is 1.53. If the average growth rate during the 

 first year is assumed to be the same as the average 

 during the first 8 mo, the predicted number of 

 layers at 1 yr (1.53 • 12 - 8) is 2.3. This extrapola- 

 tion, however, is a slight overestimate, because 

 while growth during the first year in delphinids is 

 approximately linear, there is some decay of rate. 

 The predicted number of tooth layers (using Fig- 

 ure 12) at 138 cm, the above-predicted length at 1 

 yr based on camparison with other odontocetes, is 

 2.0. It seems safe to assume that about 2 layers are 

 laid down during the first year of life. 



Calibration of the remainder of the tooth-layer 

 curve is more difficult. Kasuya et al. (1974) 

 examined the innermost layer in teeth of S. at- 

 tenuata and related type and thickness of layer to 

 season of capture. They concluded that one layer 

 (one transparent plus one opaque subunit) repre- 

 sents 1 yr of growth. We found no correlation 

 between thickness of the innermost layer and 

 season of capture. Almost all of the samples for 

 which teeth were sectioned, however, were col- 

 lected in the first few months of the year. Lacking 

 such direct calibration, several alternative pos- 

 sibilities can be examined. The results, however. 



241 



