Burton: Age, growth, and mortality of Lutjanus griseus 
259 
cn 
c 
0 ) 
To 
,o 
Figure 3 
Theoretical growth curves for gray snapper by (A) area and < B ) sex, areas pooled. 
Back-calculated sizes at age of 
gray snapper by area are shown in 
Tables 3 and 4. Tbe mean total back- 
calculated lengths for gray snapper 
from north Florida, back-calculated 
to the last annulus, were larger than 
those for fish from south Florida for 
all ages except age 2. Given the ob- 
vious differences in size at age be- 
tween the two areas, it seemed in- 
appropriate to pool the data for the 
purposes of analyzing growth. 
The linear regressions of the mea- 
surements to the first and second an- 
nuli on age for headboat and commer- 
cial specimens from north Florida 
were significantly different from zero 
(77=519, P=0.0001, r 2 =0.04; and 77 = 
506, P=0.0012, r 2 =0.02). Although 
this result indicates the presence of 
Lee’s phenomenon for samples from 
these fisheries (Potts et al., 1998), 
the model explains very little of the 
variation, so that Lee’s phenomenon 
was weak and likely masked by other 
environmental variables. The lineal- 
regressions of the measurements to 
the first and second annuli on age for 
headboat and commercial specimens 
from south Florida were not signif- 
icantly different from zero (77=720, 
P= 0.23, r 2 =0.002 and 77=720, P=0.81, 
7- 2 =0.0001), indicating that size-se- 
lective mortality was not detected. 
The von Bertalanffy ( 1938) growth 
equations, fitted to back-calculated 
lengths at age for the last annulus 
(Fig. 3A), were 
North Florida: 
L t = 717 (1 _ e -° 17 o + 0.0251), anc j 
South Florida: 
L t = 625 (1 -e-°- 13, ' + 133) ). 
Parameters and associated statistics 
for these equations as well as for 
equations fitted to the subset of data 
by sex, are listed in Table 5. Esti- 
mates of growth parameters were 
not significantly different between 
sexes [Hotelling's T 2 test, F { 0 . 025 , 3 , ~>= 
1.0 (Bernard, 1981)] (Fig. 3B). How- 
ever, estimates of growth parameters were significantly 
different between areas (P l0025 3r =84.1). 
Mortality 
Estimates of the instantaneous rate of natural mortality 
(M) varied considerably by method but were similar across 
areas for a given method. Estimates of M assume a con- 
stant natural mortality and so apply to all ages of fish. 
Hoenig’s (1983) longevity-mortality relation returned the 
lowest estimates, 0.18 and 0.29 (north and south Florida), 
and his equation adjusted for sample size returned similar 
values between areas, 0.33 and 0.35. Pauly’s (1980) equa- 
tion, with growth parameters and mean seawater temper- 
