50 
Fishery Bulletin 106(1 ) 
Results 
General diet description 
From 2002 through 2006, summer flounder were 
collected at 877 sampling locations, and at 688 of 
these locations at least one summer flounder had 
prey in its stomach. Overall, prey were encountered 
in 1780 (57.8%) of the 3079 stomachs collected. The 
total observed diet was composed of 123 prey types, 
70 of which were identifiable to the species level (24 
fishes and 46 invertebrates). In an effort to pres- 
ent summer flounder diet composition in the most 
efficient manner, prey types contributing relatively 
little to the overall diet were combined at higher 
taxonomic levels. 
Mysid shrimp ( Neomysis spp.) and bay an- 
chovy ( Anchoa mitchilli) were the main prey of 
the summer flounder, accounting for approxi- 
mately 42% combined (24.1% and 17.9%, respec- 
tively, Fig. 4) of the diet by weight, and mantis 
shrimp (Squilla empusa — 11.2%) and weakfish 
( Cynoscion regalis — 11.1%) were of secondary 
and nearly equal importance. Of the remaining 
prey types, spot ( Leiostomus xanthurus), Atlantic 
croaker ( Micropogonias undulatus ), and spotted 
hake ( Urophycis regia ) were the most important 
fishes, and sand shrimp (Crangon septemspinosa) 
was the main invertebrate prey. Each of these 
species represented between 2% and 7% of the 
diet. All other identifiable prey types each con- 
tributed <2% to the diet. 
Unidentifiable prey items (i.e., unidentifiable 
fish and unidentifiable material) were prevalent, 
likely because of the shearing action of the teeth 
of these predators, and composed 6.0% of the diet 
by weight. Although many of the unidentifiable 
items were encountered in stomachs along with 
identifiable prey and were likely the same spe- 
cies as the latter, they were, however, classified 
as unidentifiable so as to provide a conservative 
diet description. 
Ontogenetic and temporal changes in diet 
The CCA indicated that summer flounder dietary 
changes by fish size, month, and year were sta- 
tistically significant. Taken together, the afore- 
mentioned factors explained 6.0% (P=0.001) 
of the variability in diet; the first and second 
canonical axes accounted for 51.2% and 34.5% of 
the explainable variation, respectively. Fish size 
(r=-0.459; P=0.001) more closely corresponded 
to the first canonical axis than the second and, 
of the three variables examined, accounted for 
the greatest portion of the variation that was 
explicable. Month (r=-0.481; P=0.001) and year 
(r=-0.094; P=0.001) were more closely correlated 
to the second axis (Fig. 5). 
The amount of fish in the diet of summer floun- 
der increased with increasing size (Fig. 6A). My- 
sid shrimp, sand shrimp, and mantis shrimp 
accounted for approximately 79% of the diet of 
the summer flounder <225 mm TL. Bay anchovy 
(9.5%) and weakfish (2.3%) were the main fish 
prey of these individuals. The diet of summer 
flounder ranging from 225 to 374 mm TL was al- 
so dominated by mysid shrimp. The contribution 
1.8 -i 
1.6 
1.4 - 
1.2 
1.0 -| 
0.8 
0.6 
0.4 
0.2 
0.0 
• • • • 
“I 1 1 1 - 
“I 1 1 1 
0 
575-599 - 
600-624 - 
625-649 - 
B 
y9 
<24 - y 
,40 ' 
4 6 8 10 12 14 16 18 
Number of clusters 
Trim observations 
Average distance between clusters 
Figure 3 
(A) Scree plot depicting average distance between clusters 
versus the number of clusters which was used to identify the 
number of clusters into which 25-mm size-classes of summer 
flounder ( Paralichthys dentatus ) should be grouped (four size 
groups were selected since the curve leveled out at five or more 
clusters), (B) cluster diagram representing the relationships 
among the diet compositions of 25-mm size-classes of summer 
flounder. Trim observations represent the 25-mm size-classes 
omitted from the analysis because of low probability density, and 
average distance represents the coefficient used as a measure 
of dissimilarity among size-classes. 
