NOTE Fernandez and Hohn: Age, growth, and calving season of Tursiops truncatus 
359 
males. Trends in the residuals, however, showed that 
the fit was poor for a number of age classes (Fig. 3). 
To obtain a better predictor, the model was fitted to 
a data set that excluded all specimens less than 1 
year of age, analogous to the procedure in Read et 
al. (1993) where the Gompertz model was found to 
describe well the growth of bottlenose dolphins from 
Sarasota Bay, Florida. With this subset of data, as- 
ymptotic length was slightly higher for males (268.0 
cm) and for females (246.7 cm) but neither were sig- 
nificantly different from the predicted values from 
the total sample (approximate Gtest, P> 0.05, Table 
2). Trends in the residuals showed that predicted 
length at age was a better fit for ages 1-9 years and 
was slightly overestimated at asymptote. 
In an attempt to find a fit with evenly dispersed 
residuals, the Gompertz model was modified to fix 
length at birth at 109.5 cm, as calculated below by 
independent means, rather than by solving for the 
asymptotic value, A: 
S(t)= 109.5^exp(6(l- 6exp(-£)))j. (2) 
Predicted asymptotic length was 261.7 cm for males 
and 244.4 cm for females. Predicted length from birth 
through age 2 years was very similar to that esti- 
mated from the total sample with the Equation 1 ver- 
sion of the Gompertz model. Length at age was still 
underestimated and this derivation of the model did 
not improve the fit. 
Although none of the iterations produced a fit that 
represented all age classes well, the best overall pre- 
dicted values occurred with Equation 1) when speci- 
mens <1 year of age were excluded (Fig. 3) and, for 
comparative purposes, predicted lengths were taken 
from this model (268 cm for males and 246.7 cm for 
females). Males were significantly larger than fe- 
males at asymptotic length (approximate Gtest, 
P<0.05). Asymptotic lengths of males and females 
were not significantly different (approximate Gtest, 
P>0.05) between dolphins from Texas and Sarasota 
Bay, Florida (Table 2). 
Length at birth and calving seasonality 
There was no difference in length at birth between 
male and female neonates or between males and fe- 
males <0.1 year old U-test, P>0.05). Therefore, data 
were combined to produce a single estimate with each 
method (Table 3). Mean length at birth from com- 
bined data was not significantly different between 
these two methods (Gtest, P>0.05), therefore the to- 
tal data set was combined (males, females, and sex the Gompertz curve was greater than length at birth 
unknown, n=42) to give an estimated length at birth estimated from the more direct measures. In addi- 
of 109.4 cm (SD =8. 5). Length at age 0 estimated from tion, because the best-fit version of the Gompertz 
Figure 1 
Midlongitudinal stained thin sections from the tooth of a 29- 
year-old, 261-cm female bottlenose dolphin, Tursiops trun- 
catus , stranded in Texas. Growth layer group (GLG) bound- 
aries were considered to be the relatively thin dark layers; 
most of these are easily visible. Boundaries between the first 
few layers are marked with a small black line on the upper 
left side of the section. The last few, very fine GLG’s near the 
pulp cavity are difficult to see at this low magnification. 
