Latour et al. : The trophic dynamics of Paralichthys dentatus in Chesapeake Bay 
49 
76 ° 30 ' 76 ° 00 ' 75 ° 30 ' 
Figure 1 
Average catch of summer flounder ( Paralichthys dentatus) in the mainstem (i.e., nontribu- 
tary waters) of Chesapeake Bay by sampling month (Mar, May, Jul, Sep, Nov) from 2002 
through 2006. Horizontal histograms represent raw catch data by 0.1 latitudinal degrees 
corresponding to the map scale. 
of the data (Garrison and Link, 2000). Size, 
month, and year were coded by using ordinal 
variables. Observations (sampling sites) con- 
taining fewer than three summer flounder and 
explanatory variable blocks (size, month, year 
categories) containing fewer than three obser- 
vations were excluded to eliminate variance 
issues related to small sample size. 
The CCA was used to determine the amount 
of variability in the summer flounder diet ex- 
plained by the canonical axes, which are linear 
combinations of the three explanatory variables 
correlated to weighted averages of prey within 
blocks (ter Braak, 1986; Garrison and Link, 
2000). The significance of the ontogenetic and 
temporal factors was determined by using per- 
mutation tests (ter Braak, 1986). A prey species- 
explanatory factor biplot was constructed to 
examine the correlations between the factors 
and the canonical axes and to explore the di- 
etary trends associated with these variables. 
Detailed diet descriptions were then generated 
for each of the significant factors identified by 
the CCA. The CCA was performed with the 
program CANOCO, vers. 4.5 (Microcomputer 
Power, Ithaca, NY). 
Total length (mm) 
Figure 2 
Size frequency of summer flounder (Paralichthys dentatus ) 
sampled in the mainstem (i.e., nontributary waters) of Chesa- 
peake Bay from 2002 through 2006. 
