Ziig et al Age and growth of Hawaiian Chclonia 



123 



40 60 



Straight carapace length (cm) 



80 



100 



Figure 6 



Difference in estimated age (yr) between the correction-factor method and 

 the spline-integration method within the comparable range of carapace 

 lengths (SCL). For positive values (black region!, the CF method gives older 

 ages and for negative values (striped region), younger ages. 



methoci and 4 to 10 years old by the SI method. These age 

 estimates for Hawaiian greens in the last years of their 

 pelagic developmental stage are similar to those reported 

 for C. mydoK populations in the southern Great Barrier 

 Reef (SGBR) (5-6 yr; Chaloupka et al., in press) and for 

 the Atlantic coast of central Florida (3-6 yr; Zug and 

 Glor, 1999). The smallest C. mydas turtle in the Florida 

 sample was 28 cm SCL (several others were less than 35 

 cm), whereas the smallest SGBR turtle was 38.5 cm CCL 

 (Limpus and Chaloupka, 1997) and the smallest Hawai- 

 ian specimen was 34.8 cm SCL, indicating an earlier shift 

 from pelagic to benthic life in Florida greens. 



The growth of the juvenile turtles predicted by both CF 

 and SI models is consistent with the growth observed in 

 the Molokai mark-recapture sample, but predictions of the 

 CF model depart from the tag-recapture results in older 

 turtles (Fig. 6). Mean growth rates for smaller (30-60 cm) 

 turtles estimated from our transformed humerus incre- 

 ment data (Table 2) were about half as high as growth 

 rates reported for turtles of the same size in most Atlan- 

 tic and Caribbean locales (based on tagging and skeleto- 

 chronology: Tables 2 and 3 in Zug and Glor, 1999). Our 

 estimates were similar to the observed growth rates of 

 tagged turtles in Kiholo Bay, Hawai'i (Balazs et al., 2000) 

 and nearly twice as high as rates observed in some other 

 Pacific samples (Galapagos. Heron Island; Tables 2 and 3 

 in Zug and Glor, 1999). Subsequent studies of Australian 

 populations (Limpus and Chaloupka, 1997; Chaloupka et. 

 al., in press) have shown a mid-juvenile growth rate more 

 similar to our estimates; however, growth rate is associated 

 with a gi-owth surge in the Australian turtles over a narrow 

 mid-juvenile length range (50-60 cm SCL). Such a spurt 



in growth was not evident in our Sl-based growth cui"ve, 

 and growth rates were fairly constant in the 40-80 cm 

 size classes (Fig. 8). 



Our age and growth estimates pertain to the Hawaiian 

 population as a whole, because the sampled turtles orig- 

 inated from locations throughout the archipelago. Some 

 variation in age and growth between island foraging 

 groups is likely, given the extensive latitudinal range of 

 the habitats and the associated variation in physical and 

 biological parameters affecting growth (Balazs, 1982). A 

 geographic analysis of age and growth will be the subject 

 of a future study. Future study will also investigate the 



