120 
Fishery Bulletin 109(1) 
long-chain monounsaturated fatty acids were higher in 
concentration in the winter, and the omega-3 fatty acids 
were lower in winter. However, only concentrations of 
20:5/2-3, not the biologically important 22:6n-3, were 
observed to be lower in winter, indicating mobilization 
of 20:5n-3 or a shift in diet during winter. 
The differences observed in total lipid content and 
fatty acid composition could also be a result of the dif- 
ferential allocation of resources to specific tissues with- 
in the body; however, this aspect of variability was not 
a goal of our research because Atlantic herring preda- 
tors consume their prey whole. A small subset of large 
individuals (n=31, at least 21 cm in fork length) was 
examined for differences in total lipid content by tissue 
type, and muscle tissue was found to have significantly 
more total lipid content than gonad tissue (PcO.OOOl). 
However, because few of these individuals were sexu- 
ally mature, these data indicate only that muscle tissue 
may be an important lipid store for fish as they begin 
to mature sexually. 
Iverson et al. (2002) also identified ontogeny as the 
main factor responsible for the variation in the lipid 
content and fatty acid composition of Pacific herring 
( Clupea pallasi), and similar fatty acids (omega-3, -6, 
and long-chain monounsaturates) were identified as 
important in distinguishing between the fatty acid sig- 
natures of fish by age. Huynh et al. (2007) also found 
ontogenetic variation in the fatty acid signatures of Pa- 
cific herring fillets. As we found with the data from this 
study, Jensen et al. (2007) identified ontogeny, year, and 
season as important factors contributing to differences 
in the fatty acid composition of Baltic herring collected 
from 2001 through 2003. Budge et al. (2002) determined 
that the fatty acid signatures of Atlantic herring were 
significantly different from all other species found on 
the Scotian Shelf, Georges Bank, and the southern Gulf 
of St. Lawrence, with the exception of capelin ( Mai - 
lotus villosus) and northern sand lance ( Ammodytes 
dubius). The separation of Atlantic herring from other 
fish species in the Budge et al. study may be due to the 
high concentration of 22:1/2-11 found in herring (17.27% 
±5.68%) compared to other fish (haddock: 1.65% ±1.58%, 
mackerel: 6.0t% ±3.44%, pollock: 2.68% ±1.45%). We al- 
so found high concentrations of 22:ln-ll in the Atlantic 
herring examined; in fact, 22 : 1 / 1-11 was the fatty acid in 
highest concentration in all fish (Tables 3 and 4). Com- 
bining previous work (Iverson et al., 2002; Jensen et 
al., 2007) on other ecosystems with the data presented 
here, it seems clear that there is significant variability 
in the nutritional quality of herring on ontogenetic, an- 
nual, and seasonal scales worldwide. In order to obtain 
a complete picture of prey quality, the total lipid content 
and fatty acid composition of the whole body, as well 
as body tissues (muscle, gonads), should be compared 
to identify possible differences due to sexual maturity 
stage. Further, these studies serve to emphasize that 
the characterization of prey from a physiological and 
biochemical perspective (quality) should be conducted 
synoptically with measures of biomass (quantity) to best 
determine the prey field of interest. 
The variation in Atlantic herring nutritional quality 
observed in this study could have large impacts on the 
health of herring predators. During years or seasons of 
low lipid content, predators relying on herring would 
either have to spend more time foraging to meet energy 
demands or cope with less energy intake from the same 
amount of prey. The latter could result in a decline in 
body condition, health, or reproductive success if min- 
imum caloric requirements are not met (e.g., Atkinson 
and Ramsay, 1995; Alonso-Alvarez and Telia, 2001). i 
For example, Diamond and Devlin (2003) demonstrated 
a decline in breeding success of Arctic and common 
terns (Sterna paradisaea, S. hirundo) in the Bay of 
Fundy from 1995 through 2000 and directly linked it 
to a decline in lipid content of their main prey item, 
Atlantic herring. Although we do not know at which 
point, or whether, the lipid content of Atlantic herring 
falls below a level that would make them unprofitable 
for predators, the lowest mean percent lipid value ob- 
served in this study (6%) is still considered high for fish 
in general (compared to that for Atlantic cod [Gadus 
morhua ]: 2.1%; haddock [Melanogrammus aeglefinus]\ 
1.4%; Atlantic mackerel [Scomber scombrus ]: 3.4%; At- 
lantic pollock [Pollachius pollachius ]: 3.0%; all values 
from Budge et al., 2002), many of which are important 
prey species despite their relatively low lipid content 
(Gannon et al., 1998; Pauly et al. 1998). Although it 
is likely that the lowest percentages observed in this 
study are still above this minimum level, the variation 
present in the nutritional quality of Atlantic herring 
has implications for the quality of food that predators 
are receiving. Predictability in the quantity and qual- 
ity of available resources is an important element of 
food web dynamics, and some predators may depend 
heavily on the availability of a consistent type and 
level of energy intake. Such reliance can have serious 
implications for predators accustomed to high-energy 
prey; these predators may not be able to adjust to a 
high volume of low-quality food in place of a lower but 
more consistent amount of high-quality food, as has 
been shown in Steller sea lions ( Eumetopias jubatus) 
by Rosen and Trites (2004). 
The results of this study indicate significant ontoge- 
netic, annual, and seasonal variation in the lipid con- 
tent and composition of Atlantic herring from the Bay 
of Fundy. Herring are a critical prey species in the Bay 
of Fundy and the variation in lipid content and composi- 
tion of these fish affects the nutritional quality of the 
prey that many upper predators in the Bay of Fundy 
are receiving. Because herring are a vital link between 
the upper and lower levels of the food chain, they are 
central to understanding the effects of variability in 
one species on the entire food web. Such variability has 
been shown to affect trophic dynamics in complex food 
webs across the globe (e.g., Duffy and Paul, 1992; Lang- 
vatn and Hanley, 1993; Toft and Wise, 1999; Rosen 
and Trites, 2004) and provides insight into the ecology 
and distribution of species in these environments. In 
the Bay of Fundy, predators of Atlantic herring are 
receiving nutritionally different “packages” depending 
