Taylor and Gervasi: Feeding habits and dietary overlap of age-0 flounder 
173 
a percentage of total variation) (Anderson et al., 2008). 
Moreover, by using Pearson correlations, vectors of the 
dominant prey taxa (%/i?/>1.8%) were superimposed 
onto the PCO biplots, which correspond to the mono¬ 
tonic relationships between the dietary importance of a 
prey and the PCO axes (Anderson et al., 2008). 
Diet diversity for winter and summer flounder was 
estimated by using the Levins index of niche breadth 
(B) (Levins, 1968), such that 
«p-l 
■-1 
(4) 
where B = the standardized index of niche breadth for 
winter or summer flounder; 
Pk = the proportional contribution of prey taxon k 
to the diet of a flounder species (based on 
7i?/k); and 
Tip = the total number of prey taxa identifled in 
the stomachs of a flounder species. 
Values of B range between 0 and 1, and a value of 0 in¬ 
dicates maximum dietary specialization and a value of 
1 indicates nondiscrimination of prey. Moreover, B>0.6 
denotes a high niche breadth, whereas values of 0.4- 
0.6 and <0.4 represent moderate and low diet diversity, 
respectively (Novakowski et al., 2008). In this study, 
B was calculated for winter and summer flounder ir¬ 
respective of their body sizes (the size-independent es¬ 
timate) and at specific size classes (the size-dependent 
estimate). For the latter, for each sampling site, B was 
estimated for winter and summer flounder at 5-mm TL 
and 10-mm TL size-class intervals, respectively. Non¬ 
linear (exponential) regression models were used to 
examine the effect of body size of winter and summer 
flounder on diet diversity. 
Interspecific dietary overlap 
The extent of dietary overlap between winter flounder 
and summer flounder was evaluated by using 2 ap¬ 
proaches. First, after the procedures described above, 
2-way PERMANOVA models, PCO biplots, and SIM¬ 
PER analyses were used to examine similarities in 
diet composition as a function of species type (winter 
or summer flounder) and size categorization (small, 
small-medium, medium-large, or large). Second, the 
Schoener index was used to assess interspecific dietary 
overlap (Schoener, 1970), such that 
a = l-0.5nLaiPhk-^j 
jk 
(5) 
where a estimates the degree of resource overlap be¬ 
tween flounder species; and 
Phk Pjk = the proportional contributions of prey 
taxon k (based on IRI^ to the diet of win¬ 
ter flounder {h) and summer flounder (/), 
respectively. 
The result is an a value that ranges between 0 and 
1, and a>0.6 denotes biologically significant overlap 
in the use of prey resources (Schoener, 1970), whereas 
values of 0.4-0.6 and <0.4 represent moderate and low 
dietary overlap, respectively (Hartman and Brandt, 
1995; Novakowski et al., 2008; Zahn Seegert et al., 
2014). In this study, the Schoener index was calculated 
for winter and summer flounder irrespective of their 
body sizes (size-independent estimate) and for winter 
and summer flounder of equivalent sizes (size-depen- 
dent estimates). For the latter, for each sampling site, 
a was evaluated for winter and summer flounder at 
5-mm-TL increments ranging from 20 to 90 mm TL (14 
total increments). A nonlinear (logarithmic) regression 
model was used to examine the effect of fish body size 
on the extent of dietary overlap. 
Results 
Intraspecific diet composition and diversity 
During 2009-2015, winter flounder and summer floun¬ 
der were collected from 186 and 132 seine hauls, re¬ 
spectively, in the Seekonk and Taunton Rivers (Table 
2). The stomachs of 1109 winter flounder and 749 sum¬ 
mer flounder were examined in total, of which 89.8% 
and 95.3% contained prey. In stomachs of winter floun¬ 
der, 33 unique prey taxa were identified (mean num¬ 
ber of prey taxa per stomach: 3.0 [standard error (SE)] 
0.04) (Table 2). The dominant prey of winter flounder, 
with respect to the %IRI, were amphipods, harpacti- 
coid and calanoid copepods, polychaetes, bivalves (e.g., 
clam siphons), insects (e.g., chironomid larvae), and 
isopods (Table 2). Collectively, these taxa accounted for 
98.8% of the relative diet of winter flounder 20-90 mm 
TL. Other prey that were of lesser dietary importance, 
yet relatively common in stomachs of winter flounder 
(%F>1%), included ostracods, nematodes, and mysid 
shrimp. Summer flounder 19-172 mm TL consumed 32 
different prey taxa and each stomach contained, on av¬ 
erage, 2.1 (SE 0.03) prey taxa (Table 2). Five prey taxa 
composed 98.9% of the diet of summer flounder (based 
on %IRI): mysid shrimp, sand shrimp, amphipods, 
fish, and copepods (Table 2). Other prey encountered 
at relatively high frequencies in stomachs of summer 
flounder (%F'>1%) were polychaetes, cumaceans, clam 
siphons, and ostracods. 
According to the niche breadth index, a low level of 
diet diversity was observed for both winter and sum¬ 
mer flounder when calculations were made irrespective 
of body size {B: 0.24 and 0.30; Table 2). Niche dietary 
breadth of winter and summer flounder, however, sig¬ 
nificantly expanded with increasing body lengths (expo¬ 
nential regression: winter flounder, F=11.05, coefficient 
of determination [r^]=0.104, P<0.005; summer flounder, 
F=6.61, r^=0.09, P<0.05) (Fig. 4, A and B). Specifically, 
winter flounder had a moderate and high degree of 
diet diversity when individuals exceeded ~75 mm TL 
(R>0.4), and this broader-based diet was most evident 
in the Taunton River (Fig. 4A). For summer flounder. 
