598 
Fishery Bulletin 96(3), 1998 
Table 6 
Summary statistics and estimated parameters for the Jolly-Seber analysis of Kemp’s ridley turtle, Lepidochelys kempii , mark- 
recapture data (standard errors given in parentheses). Descriptions of the notation are as follows: n = total number of turtles 
captured and released each year; m = number of marked turtles captured each year; u = number of unmarked turtles captured 
each year; r = number of turtles released each year that were captured again later; 2 = number of turtles captured before a given 
year and captured again later; M = annual estimate of the number of marked turtles in the population; N = annual estimate of 
population size; and B = annual estimate of the number of individuals recruited to the population. 
Summary statistics 
Estimated parameters 
Year 
n 
u 
m 
r 
2 
M 
N 
B 
1986 
17 
17 
0 
3 
— 
— 
— 
— 
1987 
17 
16 
1 
3 
2 
10.91 
(5.49) 
98.05 
(32.11) 
139.87 
(42.90) 
1988 
35 
31 
4 
1 
1 
18.20 
(7.41) 
187.03 
(55.83) 
96.38 
(32.86) 
1989 
29 
28 
1 
3 
1 
7.30 
(4.70) 
159.79 
(50.22) 
107.25 
(34.87) 
1990 
32 
29 
3 
2 
1 
14.17 
(6.72) 
172.66 
(52.53) 
45.54 
(24.72) 
1991 
20 
18 
2 
6 
1 
11.32 
(6.05) 
109.35 
(35.92) 
201.60 
(58.21) 
1992 
49 
43 
6 
2 
1 
28.61 
(10.55) 
262.79 
(75.59) 
25.64 
(28.35) 
1993 
22 
19 
3 
16.34 
(9.53) 
119.81 
(39.34) 
ter shell fragments were identified in both of the fe- 
cal specimens. Furthermore, one of these specimens 
contained hooked mussels (Ischadium recurvus ) at- 
tached to an oyster shell fragment. 
Discussion 
The results of this study indicate the importance of 
seagrass beds and oyster reefs as developmental 
habitats for Kemp’s ridley, loggerhead, and green 
turtles. Furthermore, these species may be prefer- 
entially utilizing the two habitat types on the basis 
of their respective feeding strategies. The extensive 
seagrass flats along the west coast of Florida have 
been identified as foraging habitat for the herbivo- 
rous green turtle (True, 1887; Carr and Caldwell, 
1956; Caldwell and Carr, 1957). Netting effort at the 
seagrass shoals of Waccasassa Reefs resulted in cap- 
tures of mid- to late subadult green turtles, compa- 
rable to the size class of green turtles reported by 
Carr and Caldwell (1956). The Kemp’s ridley turtle 
is primarily cancivorous (Shaver, 1991; Burke et al., 
1994), and the distribution of this species can be cor- 
related to areas with abundant crab populations 
(Ogren, 1989). Intertidal oyster bars provide refuge 
for stone crabs (McRae, 1950; Bender, 1971; Wilber 
and Herrnkind, 1986), whereas the mud bottom ad- 
jacent to these bars is the preferred substrate of blue 
crabs (Evink, 1976; Wolfe, 1990). Subadult Kemp’s 
ridley turtles dominated the aggregation of marine 
turtles captured in the vicinity of Corrigan Reef and 
the food items for these turtles were typical of the 
macroinvertebrate fauna inhabiting nearshore oys- 
ter bars. Subadult and adult loggerhead turtles were 
also captured at Corrigan Reef; this species also feeds 
on benthic invertebrates, particularly molluscs 
(Dodd, 1988). The possibility of competition for food 
resources between loggerhead and Kemp’s ridley 
turtles is unknown and could be investigated by com- 
paring fecal specimens of both turtles captured in 
the same location. 
Tagging studies of Kemp’s ridley turtles have re- 
vealed reproductive migrations of females in the Gulf 
of Mexico (Pritchard and Marquez M., 1973), sea- 
sonal migrations of subadults along the Atlantic coast 
(Henwood and Ogren, 1987; Schmid, 1995), and an 
east-west movement of subadults in the northern 
Gulf (Ogren, 1989). However, there are no mark-re- 
capture data to indicate a seasonal migration of sub- 
adult turtles in the eastern Gulf of Mexico. As stated 
by Carr (1980) and observed in the present study, 
turtles apparently immigrate to the nearshore wa- 
ters of the Cedar Keys in April and emigrate to some 
