114 
Fishery Bulletin 111(2) 
Table 1 
Summary of mean total length (TL, mm), sample size ( n ), and mean carbon (5 13 C) and nitro- 
gen (5 15 N) stable isotope values (%e) and their standard deviations (SD) by sample location 
for Spotted Seatrout (Cynoscion nebulosus) collected at 10 sites in coastal Mississippi and 
the western Mississippi Sound in 2007-09. Data for “reference” samples were obtained from 
laboratory-reared fish used to calculate intrinsic variability. Map code refers to site labels in 
Figure 1. The map code na (not available) indicates that the reference fish are not show in 
Figure 1. 
Location 
Map code 
TL |mm] 
(SD) 
n 
Mean 5 13 C 
(SD) 
Mean § 15 N 
(SD) 
Bayou Caddy 
BC 
323(61) 
30 
-20.8(0.9) 
14.3(0.9) 
Dupont 
DP 
289(54) 
9 
-21.9(1.8) 
13.7(1.2) 
Bayou Portage 
BP 
279(45) 
31 
-22.6(1.3) 
13.5(0.8) 
Oyster reefs 
OR 
357(51) 
24 
-21.7(1.1) 
14.3(0.6) 
Back Bay 
BB 
366(53) 
10 
-22.1(1.9) 
14.8(0.6) 
Fort Bayou 
FB 
290(54) 
4 
-21.0(1.3) 
14.7(1.0) 
Davis Bayou 
DB 
296(71) 
16 
-21.3(1.4) 
13.1(1.1) 
Graveline Bayou 
GB 
271(47) 
15 
-20.6(1.1) 
12.8(0.8) 
Marsh Lake 
ML 
439(40) 
2 
-20.6(1.4) 
14.4(0.6) 
Barrier Islands 
BI 
500(76) 
33 
-19.1(1.7) 
15.1(1.0) 
Reference 
na 
210(13) 
20 
-18.03(0.24) 
12.74(0.2) 
Finally, to assess the potential bias on sample vari- 
ability with pooling fish for analysis of S 13 C isotope 
values, a simple resampling numerical experiment was 
conducted on data from all site and date combinations 
that produced at least 30 fish for analysis. Subsamples 
that consisted of between 2 and 20 randomly selected 
fish were averaged to produce a “pooled” value for 5 13 C, 
and means of pooled data were calculated from each 
pooled subsample (n = 10 subsamples selected with re- 
placement) for each resampled size. These data were 
then plotted to assess bias in the result by compari- 
son of the resampled mean with the nonpooled sample 
mean for each set of data. The pattern in the absolute 
residuals was examined as a function of resample size 
to estimate minimum effective pooled sample size for 
each data set. 
Results 
From 10 sites over the period of 2007-09, 187 individu- 
al Spotted Seatrout were captured (Fig. 1, Table 1). All 
fish were captured between May and October and sam- 
pling coverage was relatively even across months with 
the exception of September, which yielded only 3 fish. 
Fish ranged in size from 212 to 622 mm TL, but mean 
TL by site was 250-350 mm TL with the exception of 
fish collected near the Barrier Islands. The mean 5 13 C 
for all fish across all sites was -21.4%o (standard devia- 
tion [SD] of 1.6%e.) The system-wide standard devia- 
tion was generally similar to the within-site standard 
deviation but close to the within-site maximum (Table 
1). The mean S 15 N across all sites was 13.9 %c (SD of 
l.l%e). As with 8 13 C, the overall standard deviation was 
within the range of within-site variability but toward 
the high end of the range. On the basis of the 8 15 N 
data, Spotted Seatrout collected from field sites feed 
at a mean trophic level of 3.4. The standard deviation 
of both 8 13 C and 8 15 N for reference fish raised in the 
laboratory was < 20 % of field-based estimates of within- 
sample variability (Table 1). 
The 8 13 C and 8 15 N isotope ratios of Spotted Seat- 
rout showed an inshore-to-offshore and summer-to- 
fall enrichment trend. Both S 13 C and 8 15 N values 
examined showed a statistically significant trend 
across sites ( 8 13 C ANOVA, F 9161 =2.61, P=0.008; 8 15 N 
F 9 jgj=3.44, P<0.0001) and sample months ( 8 13 C ANO- 
VA, jF 5 161 =7.35, P<0.0001; 8 15 N F 5 161 =9.29, P<0.0001). 
Post-hoc analysis indicated a significant difference be- 
tween fish collected at inland, marsh-associated sites, 
such as Graveline Bayou, Davis Bayou, Bayou Portage, 
and Dupont, and fish captured near the Barrier Is- 
lands (Fig. 2A). Isotope values overlapped considerably 
among more open-water sites in Mississippi Sound 
(Bayou Caddy and oyster reefs), Marsh Lake, and Back 
Bay at both marsh-associated sites and Barrier Islands 
sites, but the open-water sites were closer than the 
Barrier Islands sites in isotopic value to marsh-associ- 
ated sites. In terms of months, June, July, and August 
were statistically different from October, but this dif- 
ference was confounded with site differences because 
most fish captured in October were from around the 
Barrier Islands (Fig. 2B). There was also a small but 
significant positive linear trend in both 8 13 C and S 15 N 
with fish TL (Fig. 3). 
The lipid-corrected stable isotope values of Spotted 
Seatrout liver samples were similar to results from 
white muscle for both 8 13 C and 8 15 N but contained no- 
