102 
Fishery Bulletin 108(1) 
Table 1 
Results of pairwise permutational multivariate analysis of variance (PERMANOVA) of juvenile gray snapper ( Lutjanus 
griseus) otolith stable isotope signatures among sampling sites, within sampled region, for Florida Bay (June 2002 and Septem- 
ber 2003), and the Ten Thousand Islands (September 2002). 
Region and month 
Sites compared 
Test statistic 
(pseudo-F) 
P-value 
Number of 
permutations 
Florida Bay 
June 2002 
Black Betsy Key, Old Dan Bank 
3.69 
0.002 
997 
Black Betsy Key, Sprigger Bank 
4.07 
0.001 
993 
Black Betsy Key, Schooner Bank 
0.38 
0.864 
991 
Old Dan Bank, Sprigger Bank 
2.57 
0.008 
990 
Old Dan Bank, Schooner Bank 
2.97 
0.001 
991 
Sprigger Bank, Schooner Bank 
3.07 
0.002 
993 
Sept 2003 
Old Dan Bank, Sprigger Bank 
4.04 
0.001 
995 
Old Dan Bank, Schooner Bank 
1.79 
0.083 
994 
Sprigger Bank, Schooner Bank 
1.86 
0.073 
991 
Ten Thousand Is. (west coast) 
Sept 2002 
Blackwater Bay, Faka Bay 
4.34 
0.003 
637 
Blackwater Bay, Palm Bay 
1.29 
0.252 
35 
Blackwater Bay, Pumpkin Bay 
1.70 
0.132 
441 
Faka Bay, Palm Bay 
3.03 
0.015 
275 
Faka Bay, Pumpkin Bay 
3.74 
0.004 
986 
Palm Bay, Pumpkin Bay 
0.56 
0.625 
165 
populations (Martin and Wuenschel, 2006). Lara et 
al. (2008), using trace elemental analysis, found an 
overall cross validation success rate of 82% for clas- 
sifying randomly selected samples to a juvenile nurs- 
ery habitat signature in one of five regions: Biscayne 
Bay, Florida Bay, lower Florida Keys, Ten Thousand 
Islands, and Dry Tortugas. However, that study had a 
lower percentage for successfully classifying unknown 
samples to two regions, Dry Tortugas (50%) and Ten 
Thousand Islands (52%). In the current study, isotopes 
allowed us to distinguish between sites that the 
elemental ratios in the Lara et al. study were 
unable to separate and provided a clearer suc- 
cessful classification for all regions. The Dry 
Tortugas and Ten Thousand Islands regions 
showed the highest percentage for successful 
classification, 90% and 100%, respectively, com- 
pared to other regions analyzed. These two re- 
gions are geographically the farthest separated, 
Dry Tortugas has the most oceanic influence, 
and Ten Thousand Islands has the most fresh- 
water influence from Taylor Slough and Shark 
River Slough through the Florida Everglades. 
There was also some interannual variability in 
classification success in the Lower Keys. 
Temporal variation of stable isotopic ratios of 
carbon and oxygen in the otoliths of gray snap- 
per has not been previously studied in south- 
ern Florida. This study indicates that temporal 
variation in carbon and oxygen isotopic ratios 
was expressed in otolith samples collected from 
years 2001 through 2005. Swearer et al. (2003) 
tested temporal variation of three species (top- 
smelt [ Atherinops affinis], arrow goby [Cleve- 
landia ios], and California halibut [Paralichthys 
californicus ]), collected in spring and autumn 
of 1996 from Carpinteria Marsh in southern 
1 -| 
0.8 - 
0.6 - 
0.4 - 
0.2 - 
0 
- 0.2 4 
-0.4 
-6 
x 
x 
X 
X X 
o 4 
30 
O’ 0> 
o 
-4 
S 13 C 
0 2001 
A 2002 
■$> 2004 
X 2005 
Figure 3 
Scatter plot depicting variation in juvenile gray snapper (Lut- 
janus griseus) stable isotope otolith signatures (<5 18 0 and <5 13 C) 
from the summer of 2001, 2002, 2004, and 2005 from Sprigger 
Bank, Florida Bay. 
