146 



Fishery Bulletin 93(1), 1995 



The presence of subadult green turtles 

 in the Cape Canaveral area appeared 

 highly seasonal (Table 6); 73% (ra=30) 

 were captured from November to Janu- 

 ary. As with other turtle species, this 

 pattern resulted from a high number of 

 captures in 1987 and a high monthly 

 variation during other years. 



Four tagged green turtles were recap- 

 tured during this study. Three turtles 

 were originally tagged by contract per- 

 sonnel in the Cape Canaveral area. The 

 tag codes for the fourth green turtle 

 matched a set of NMFS tags that had 

 been distributed to Fort Lauderdale, 

 Florida. Of the three green turtles 

 tagged at Cape Canaveral, one was ini- 

 tially tagged in January 1987 and then 

 recaptured in April, approximately 68 

 km to the north. Another green turtle 

 that was initially tagged in the Cape 

 Canaveral area in January 1990 was re- 

 captured in this area the following Oc- 

 tober. A third green turtle was captured 

 off Port Canaveral in September 1990 

 and found stranded approximately 41 

 km to the north the following month. 



There was a strong correlation (r=0.9590; n=39) 

 between carapace width and carapace length for 



55 60 65 70 



Carapace Length (cm) 



Figure 5 



Length frequencies of Kemp's ridley turtles, Lepidochelys kempi, collected 

 along the Atlantic coast of Florida from 1986 to 1991. 



green turtles. Regression of carapace width on cara- 

 pace length resulted in the equation 



CW = 4.1763 + 0.6847 (SSCL). 



Table 3 



Estimated values of asymptotic length (a ) and intrinsic growth rate ik ) from 

 nonlinear regression of von Bertalanffy growth interval equation for logger- 

 head turtles (one asymptotic standard error in parentheses). 



Data treatment n a k 



All recaptures 



Capture/recapture by contract vessel 



All recaptures >90 days 



All recaptures >180 days 



All recaptures >365 days 



51 96.08 cm 0.0586 



(7.07) (0.0149) 



Residual mean square error = 9.1054 



17 112.52 cm 0.0365 



(24.74) (0.0204) 



Residual mean square error = 0.4088 



33 96.09 cm 0.0588 



(8.72) (0.0185) 



Residual mean square error = 13.8969 



24 96.40 cm 0.0569 



(7.97) (0.0162) 



Residual mean square error = 10.8231 



19 96.10 cm 0.0573 



(8.82) (0.0183) 



Residual mean square error = 13.5947 



Residual analysis of the length- 

 weight relationship indicated the 

 need for curvilinear terms in the re- 

 gression model. A strong correlation 

 (r=0.9587; n=37) was calculated for 

 the log-transformed weight to length 

 relationship, and regression of these 

 variables resulted in the equation 



log WT = 8.8784 + 2.9815 

 (log SSCL). 



There were no growth data available 

 for green turtles owing to the rela- 

 tively low number of recaptures and 

 the lack of data for the recoveries. 



Discussion 



The data for this project were col- 

 lected incidentally through the com- 

 mercial shrimp fishery of east-central 

 Florida. Bias in trawling effort oc- 



