118 
Fishery Bulletin 109(1 ) 
than comparisons including fish from the other three 
years (2005 vs. 2008: dissimilarity=18.15%; 2005 
vs. 2007: 17.59%; 2005 vs. 2006: 16.36%; 2006 vs. 
2007: 14.06%; 2006 vs. 2008: 16.10%; 2007 vs. 2008: 
16.04%). As with the data on fatty acid signatures 
by fish age, the SIMPER analysis revealed that the 
separation of fish by year was also based on omega-3, 
-6, and long-chain polyunsaturated fatty acids (Table 
4). However, in contrast to the age data, 16:ln-ll 
contributed to over 10% of the differences found 
in the fatty acid signatures of fish from 2005 in 
contrast to fish collected during the other years of 
the study. The reason for this strong difference in 
percent composition is unclear because this fatty 
acid has not been previously identified as biologic- 
ally important. 
Herring from 2006 and 2007 were analyzed for dif- 
ferences in fatty acid composition by season. The AN- 
OSIM indicated significant differences in the fatty 
acid composition of fish by season (global r=0.254, 
P<0.01, Fig. 40. The SIMPER analysis showed that 
omega-3 and -6 (specifically 20:5n-3) and long-chain 
monounsaturated fatty acids (specifically 20:ln-ll 
and 22:1 n- id were the most important fatty acids 
contributing to the differences in fatty acid signa- 
tures of fish by season (Table 4). 
Discussion 
Age was the strongest determinant of the total lipid 
content and fatty acid composition of Atlantic herring. 
Younger fish had less total lipid content, but this was 
paired with higher concentrations of certain classes of 
fatty acids, such as omega-3 and -6 fatty acids which 
are important for vertebrate growth and development 
early in life (Sargent et ah, 1999; Szlinder-Richert et 
ah, 2010.). This was in contrast to older fish, which 
had higher total lipid content paired with higher 
concentrations of long-chain monounsaturated fatty 
acids which are important for energy storage (Hadley, 
1985). Although the r 2 value of the regression of 
fork length on total lipid content is relatively low, 
we believe this is indicative of the natural variabil- 
ity in the nutritional quality of Atlantic herring in 
this ecosystem. Although we do not know the direct 
cause of the ontogenetic variation in our sample, 
it is likely that small and large fish have different 
diets, especially considering the change in filtering 
ability as fish grow (Gibson, 1988). These differences 
in fatty acid composition could be manifested in one 
of two ways; either fish of different sizes are feeding 
on different prey items, or on different proportions of 
the same prey items (e.g., Iverson et ah, 1997; 2002). 
This study revealed significant annual variation 
in the lipid content and composition of herring. For 
example, 2005 fish had 37% less lipid than fish from 
2006 through 2008, and their fatty acid signatures 
were also significantly different from fish collected in 
the other three years. Also, although not statistically 
