544 
Fishery Bulletin 95(3), 1997 
Tab!e 2 {continued) 
Collection site 
Sex 
Fork length 
(cm) 
Month of 
collection 
Range of 5 13 C 
(%c) for each 
spine 
Range of 8 15 N 
(%c) for each 
spine 
Panama City, Florida 
female 
84 
6 
1.5 
0.6 
Panama City, Florida 
female 
81 
6 
3.0 
0.6 
Panama City, Florida 
female 
80 
6 
2.2 
0.3 
Panama City, Florida 
female 
90 
6 
2.0 
0.4 
Fort Pierce-Palm Beach, FL 
female 
102 
5 
2.2 
0.5 
Fort Pierce-Palm Beach, FL 
male 
100 
5 
3.5 
0.3 
Fort Pierce-Palm Beach, FL 
male 
87 
5 
3.5 
0.7 
Fort Pierce-Palm Beach, FL 
male 
71 
3 
1.3 
nd 
Fort Pierce-Palm Beach, FL 
male 
75 
3 
2.1 
0.4 
Fort Pierce-Palm Beach, FL 
male 
70 
3 
1.7 
0.2 
SX [*»] = Sample /R «an<tar<, “« * 1° 3 . <« 
where X = the heavier isotope (either 13 C or 15 N); and 
R = the ratio (either 13 C: 12 C or 15 N: 14 N). 
The working standard for carbon was tank C0 9 which 
was identified as 8 13 C pDB = -1.85%c, and the standard 
for nitrogen was N„ from air, which is 0%e by defini- 
tion (see Eq. 1). 
Because spines were too large to measure whole, 
they were divided into segments that were analyzed 
separately, and the weighted average was calculated 
as 
72 = 1 
±W„ < 2 > 
71=1 
where W n = weight of the segment in milligrams; 
and 
b n = isotopic value for the segment. 
Multivariate analysis of covariance (MANCOVA) was 
used to determine which independent variables had 
significant effects in the general linear models (GLM) 
(Eqs. 3 and 4) (SAS, 1990). 
S 15 N or <5 13 C = collection site + season + sex; 
[length was used as a covariate.] (3) 
8 15 N or <5 13 C = region + season + sex ; 
[length was used as a covariate.] (4) 
If an independent variable did not have a significant 
effect on the model, the variable was eliminated and 
the GLM was conducted again. Least squared means 
(LSmeans) and a pairwise comparison, with a 95% 
confidence interval (P=0.05), were performed to de- 
termine significant differences in nitrogen and car- 
bon isotope data between sample collection sites and 
regions. As king mackerel increased in fork length, 
an increase in 15 N was observed (Fig. 2B); therefore, 
analysis of covariance was used to control for differ- 
ences in fish size. Fork length was used as the 
covariate. 
Results 
No significant correlation between weight of the fin 
spine sample segment and 5 15 N (r 2 =0.03) or 8 13 C 
(r 2 =0.06) was detected for any of the samples, sug- 
gesting that the mineral phase had been completely 
removed, and 100% collagen carbon and nitrogen as 
C0 9 and N 2 had been recovered, respectively. 
Isotopic variations within individual spines were 
examined to try to determine the life history of indi- 
viduals. Spines were delineated into three portions 
(tip, mid, and base) (Fig. 3). The base of the spine is 
believed to contain more recently acquired material. 
Isotopic trends in carbon, along the length of the 
spine, were observed for many sites. Isotopic values 
for carbon became lighter as the fish aged (from tip 
to base); however, few trends existed for nitrogen. In 
general, the isotopic difference within the spine was 
generally less for nitrogen compared with carbon. 
Nitrogen and carbon isotopic differences were ob- 
served between various sites and regions (Table 1; 
Fig. 4). In the pairwise comparison, more significant 
differences were found between individual sites for 
nitrogen isotope ratios than for carbon isotopic ra- 
tios (Fig. 4). Nitrogen isotopic data displayed a geo- 
graphical pattern (Fig. 5). In general, king mackerel 
from Mississippi, Louisiana, and Texas were 15 N- 
enriched in contrast with those from the Mexican 
