168 
Fishery Bulletin 107(2) 
column (J&W Scientific, Folsom, CA) was used for FAME 
separation. The GC oven temperature program used 
was as follows: 100°C for 2 minutes, ramp at 3°C/min. 
to 210°C, hold for 20 min, ramp l°C/min. to 220°C, hold 
for 10 min. 
Compound specific stable isotope analysis (CSIA) 
The FAMEs were analyzed for their stable carbon iso- 
tope compositions using a Hewlett Packard 5890 Series 
II gas chromatograph interfaced through a combustion 
furnace with a VG Isoprime IRMS (Fisons/VG/Micro- 
mass, Manchester, UK). The GC was equipped with 
the same column that was used for the GC-MS analysis 
and helium was the carrier gas. The GC oven tempera- 
ture program was identical to that used for the GC-MS 
FAME identification. Time elution was used to identify 
peaks. The C0 2 combustion products of the fatty acids 
eluting from the column were introduced into the mass 
spectrometer after passing through a water trap. 
All FAME 6 13 C values were corrected for the addi- 
tion of the methyl group to the original fatty acid. The 
derivatization of the fatty acids to their methyl esters 
results in a predictable and reproducible isotope effect 
(Ballentine et al., 1996; Uhle et al., 1997). Adding a 
methyl group to the fatty acid alters its isotope signa- 
ture. However, if the isotopic ratio of the methanol (in 
this case <5 13 C=-4 6%c, measured by injecting the metha- 
nol into the mass spectrometer through the GC) and 
Statistical analysis 
Kruskal-Wallis nonparametric procedures 
were used to test for differences in isotopic 
values among anadromous fish and the dif- 
ferent guilds (predators, carnivores, general- 
ists, and planktivores, (a=0.05)). The Dunn 
procedure was used to examine differences 
between groups (Rosner, 1990). Statview SE 
+ Graphics (Abacus Concepts, Inc., Cary, 
NC), JMP In (SAS, Cary, NC) and Microsoft 
Excel version 5.0 (Microsoft, Inc., Redmond, 
WA) were used for statistical tests. The 
Dunn procedure reduces the risk of type-1 
error inherent in multiple comparison tech- 
niques. It does so by increasing the Z-score 
needed to reject the null hypothesis as the 
number of individual groups being compared 
increases. In the present study, a Z-score of 
±3.02 (0.9975 confidence) was needed for a 
difference to be significant. 
Results 
The first objective of this study was to estab- 
lish that the spawning anadromous fish 
retained the marine isotope signal more than 
40 km upstream from saline waters. This 
was the case for all three isotopes examined. 
Table t 
Fish species examined by guild (including an anadromous group) from 
the Rappahannock River to assess the role of marine fish as nutrient 
vectors. Guild assignments are based on diet as reported in Jenkins 
and Burkhead (1993). 
Guild 
Species name 
Common name 
Predator 
Ictalurus furcatus 
blue catfish 
Lepisosteus osseus 
longnose gar 
Carnivore 
Micropterus salmoides 
largemouth bass 
Lepomis gibbosus 
pumpkinseed 
Hybognathus regius 
eastern silvery minnow 
Notemigonus crysoleucas 
golden shiner 
Lepomis macrochirus 
bluegill 
Perea flavescens 
yellow perch 
Generalist 
Anguilla rostrata 
American eel 
Ameiurus catus 
white catfish 
Ameiurus nebulosus 
brown bullhead 
Ictalurus punctatus 
channel catfish 
Planktivore 
Menidia beryllina 
inland silverside 
Dorosoma cepedianum 
gizzard shad 
Erimyzon oblongus 
creek chubsucker 
Anadromous 
Alosa aestivalis 
blueback herring 
Alosa pseudoharengus 
alewife 
Alosa sapidissima 
American shad 
Morone saxatilis 
striped bass 
Morone americana 
white perch 
the fatty acid methyl ester are known, then the isotopic 
signature of the original fatty acid can be determined 
using a mass balance Equation 2. 
^ 13 C FAME ~ /FA— <5 13 C FA + /"Methanol Methanol *2) 
where # 3 C FAME , 5 13 C FA , 
and 5 13 C Methanol = the carbon isotope signa- 
tures of the FAME, the 
underivatized fatty acid, 
and the methanol, respec- 
tively; and 
f FA and f Methanol = the fraction® of carbon in the 
FAME due to the underiva- 
tized fatty acid and metha- 
nol, respectively (Ballentine 
et al., 1996; Uhle et al., 
1997). 
Each sample was injected four to eight times (depending 
on the reproducibility of the analysis). Only d 13 C values 
that were within 1.5%e of each other were considered 
to reflect the 6 13 C of the FAME (MacAvoy et al., 2002). 
Therefore, the fi L3 C reported for each FAME identified is 
represented by an average value and a standard devia- 
tion. Every sixth sample injected was an internal, labo- 
ratory standard (naphthalene-d, 6 13 C-25.7 %c) to insure 
consistent performance of the GC, oxidation furnace, 
and mass spectrometer. 
