382 



Fishery Bulletin 101(2) 



(T. truncatus) (Hohn*). This hypothesis still needs to be 

 confirmed. However, all the S. guianensis individuals that 

 were still nursing, but which already had remains of solid 

 food in their stomachs («=5), had only an accessory layer 

 that is closer to the neonatal line — they did not have the 

 layer that we are assuming marks the end of weaning. 



According to Rosas (20001, there was no significant 

 difference in incidental catches between mature and im- 

 mature individuals of S. guianensis caught on the coast of 

 Parana, suggesting a similar vulnerability of young and 

 adult estuarine dolphins to fisheries. Because the animals 

 analyzed in our study were the same ones used by Rosas 

 (2000), this lack of significant difference between mature 

 and immature individuals can suggest a representative age 

 distribution of the individuals analyzed. 



Because the maximum estimated age in our study was 

 30 years, and because the dolphins here analyzed were 

 incidentally caught in fishing nets, it seems reasonable to 

 assume that the longevity of the estuarine dolphin may 

 be 30-35 years. This hypothesis is also corroborated by 

 the study carried out by Ramos (1997) with S. guianensis 

 on the coast of Rio de Janeiro State (southeastern Brazil). 

 Although the age of the oldest male observed in our study 

 was 29 years, the frequency of males older than 21 years 

 was less than 3%, which is extremely low when compared 

 with the frequency of 21.5% for females older than 21 years. 

 These results suggest a greater life expectancy for females, 

 which is also corroborated by a study carried out by Ramos 

 (1997) in Rio de Janeiro. 



Growth 



The use of Schnute's model is helpful in deciding which 

 growth model should be used. Even though the researcher 

 can usually decide which model is most appropriate by 

 looking at the data, subtle differences in data distribution 

 could cause one or another model to be more adequate. Use 

 of a generic model allows this choice without intervention 

 of the researcher and avoids any unconscious bias towards 

 or against any model. 



The discontinuity of growth in male S. guianensis in our 

 study could have been due to the small sample size or may 

 have been due to a second growth spurt, which has already 

 been observed in the total length oiStenella attenuata (Per- 

 rin et al., 1976), Lissodelphis borealis (Ferrero and Walker, 

 1993), and Phocoenoides dalli (Ferrero and Walker, 1999), 

 and in the weight of male Tursiops truncatus (Cockroft and 

 Ross, 1990). The k value obtained for male S.guianensis 

 up to five years was very high, meaning that asymptotic 

 length in this phase of life was reached quickly. The ces- 

 sation of growth exhibited by the model for males up to 

 5 years probably is not true in the biological sense but 

 could be an artifact created by the model and the small 

 sample size. Most probably there is a marked reduction in 

 growth with the start of sexual maturation and a greater 

 investment in the weight or reproductive apparatus (or 



' Hohn, A. A. 1999. Personal comimiii licMiiDirt l.ahoralory, 

 Sc)ulli(>ast Fishiories Scienct' ('enter. National Marine Fi.slieries 

 Ser\'ice, 101 I'river.s Island Road. Bcaulort, NC 28516-9722. 



both). The hypothesis of a greater investment in weight is 

 supported by the observed difference in the weight-length 

 coefficient between males and females. Additionally, sexual 

 investment of male estuarine dolphins is very high — testes 

 of adult males can reach up to 32 cm in length and weigh 

 up to 3.3% of the total body weight (Rosas and Monteiro- 

 Filho, 2002). 



After the secondary growth spurt in males, the final 

 asymptotic length did not differ very much from that in 

 females. Previous growth studies carried out by Borobia 

 (1989), Schmiegelow (1990), and Ramos et al. (2000) with 

 the estuarine dolphin did not mention the existence of a 

 second growth spurt in males, possibly because the authors 

 did not analyze the growth of males and females separately. 

 According to Ramos et al. (2000), male and female data 

 were combined because of the absence of sexual dimor- 

 phism in the body size of adults of this species. 



Borobia (1989) and Schmiegelow (1990), who also used 

 the von Bertalanfiy model, obtained different values for the 

 growth equation parameters (Table 1). The sample used by 

 Borobia ( 1989 ) did not have many individuals in ages 1 and 2, 

 and none in the age class. The absence of animals that "an- 

 chor" the beginning of the curve could result in low estimates 

 of ^ and ?Q. Additionally, Borobia ( 1989) examined individuals 

 from different locations along the distribution of the species 

 and thus did not take into consideration possible geographi- 

 cal variations. The results obtained by Schmiegelow (1990) 

 are similar to those of our study, probably because both of 

 them used animals from the same region. 



Ramos et al. (2000) analyzed the growth of S. guian- 

 ensis using the Gompertz growth model and obtained an 

 asymptotic length (191.7 cm) which was much greater 

 than that obtained in our study and in previous studies 

 (Borobia, 1989; Schmiegelow, 1990) (Table 1). This differ- 

 ence could be due to 1) the small number of individuals 

 older than 12 years (/!=3) in their sample; or 2) a difference 

 in asymptotic lengths between southeastern and southern 

 Brazil populations. Similar differences have been observed 

 between asymptotic lengths of Pontoporia hlainvillei from 

 Rio de Janeiro (southeastern Brazil) and Sao Paulo and 

 Parana (same area of the present study), where larger in- 

 dividuals were found in Rio de Janeiro (Ramos et al., 2000; 

 Rosas, 2000). Therefore, it is possible that environmental 

 variables could be responsible for larger sizes in the area 

 studied by Ramos et al. (2000), both for S. guianensis and 

 for f! hlainvillei. 



Although no significant difference was observed in the 

 asymptotic length between adult males and females, the 

 differentiated growth in time between the two sexes is 

 probably responsible for the difference observed in the 

 weight-length relationship. 



In most species, the length exponent (0) of the weight- 

 length relationship is usually close to 3 (Santos, 1978). The 

 estimated values of this exponent for the estuarine dolphin 

 (3.2 for males and 2.6 for females) suggest that the longi- 

 tudinal and transversal body growth in this species follows 

 a similar pattern. 



Our results suggest that it is important to study growth 

 by analyzing the sexes separately, because there may be dif- 

 ferential growth between the sexes before the adult age. 



