362 
Fishery Bulletin 96(2), 1 998 
100 
90 
■1 
80 
||||||| 90-120 cm 
70 
||» M <o.i yr 
>, 60 
O 
c 
a) 
g- 50 
Q) 
H Neonates 
111 ^li 
LL 
40 
30 
■i iii 
20 
- 
!■ In 
10 
- 
0 
Jan 
Feb Mar Apr May Jun Jul Aug Sep Oct Nov Dec 
Month 
Fsgure 4 
Calving seasonality illustrated as the monthly distribution of strandings of neonates, specimens <0.1 
years old, and specimens 90-120 cm in length. 
Table 3 
Estimated length at birth (cm) for bottlenose dolphins, 
Tiirsiops truncatus, stranded along the coast of Texas. The 
two methods of defining a newborn animal (“neonates” and 
“specimens <0.1 yr”) are described in the text. The 
Gompertz values were the predicted lengths at age 0 (L 0 ) 
from the Gompertz curve fitted to age-at-length data for 
males and females. The samples for neonates and speci- 
mens <0.1 yr are independent, n = number of dolphins in 
each sample. 
Sex 
Males Females unknown Total 
Neonates 
mean 109.5 111.2 
SD 9.3 13.6 
n 12 5 
range 94-124 91-128 
Specimens < 0.1 yr 
mean 109.7 107.4 
SD 7.7 9.7 
n 13 5 
range 93-121 97-122 
Gompertz 
mean 128.2 115.1 
SD 3.6 4.9 
107.3 109.5 
5.0 9.3 
4 21 
102-112 91-128 
113.7 109.7 
3.5 7.5 
3 21 
110-117 93-122 
n 
range 
Bottlenose dolphins from the coast of Texas are not 
significantly different in adult size from those in 
Sarasota Bay, Florida (Read et al., 1993), the only 
other western Atlantic or Gulf of Mexico area for 
which growth curves have been fitted to length-at- 
age data. Contrary to the analysis of Read et al. 
(1993), however, we found that the Gompertz curve 
did not adequately describe growth across all age 
classes. The curve did not respond readily enough to 
simulate rapid growth during the first couple of 
years. The fit was improved by excluding specimens 
less than 1 year of age, the phase of growth where the 
rate and absolute incremental increase are greatest. 
Although this approach provided a data set comparable 
in range of ages to that used in Read et al. (1993), the 
Gompertz model proved a better predictor for the 
Sarasota Bay, Florida, sample where the predicted 
length of young animals was larger. The large and un- 
biased sample of live animals that represented most of 
the individuals in Sarasota Bay probably accounts for 
these differences. The fit might be further improved if 
the sample was large enough to allow for a two-stage 
Gompertz fit (e.g. see Perrin et al., 1976) that would 
accommodate the growth spurt occurring at matu- 
ration (Cheal and Gales, 1992; Read et al., 1993). 
Estimates of length at birth are not available for 
the Sarasota Bay animals. Along the mid-Atlantic 
