762 
Fishery Bulletin 96(4), 1998 
Table 4 
Breakdown into the most important prey groups of the 
mean dissimilarity between stomach contents (percent 
weight) of bonefish, Albula vulpes, from the two principal 
clusters shown in Figure 2. Small bonefish ranged from 
260 to 439 mm FL and large bonefish ranged from 400 to 
699 mm FL. Prey groups are listed in order of decreasing 
contribution to the overall dissimilarity between the two 
bonefish length groups. is the mean contribution of the 
ith species to the dissimilarity between the two groups, 5 / 
SD( 5 ; ) is the ratio between the mean contribution of the 
ith species (5 ; ) and the standard deviation of the values for 
that species [SD(8 ; )], 5 : % is the contribution to the total 
dissimilarity scaled as a percentage, and Cum 8 ; % is the 
cumulative contribution to the total dissimilarity scaled as a 
percentage. Taxa that are likely to be reliable discriminators 
of the two length groups are indicated by ** in the 5 /SD(8 ) 
column. Taxa proportionally more important in the diet of 
large bonefish than small bonefish are shown in bold type. 
Species 
8, 
S/SD(8.) 
8, % 
Cum 8^ % 
Penaeidae 
8.13 
1.94** 
10.47 
10.47 
Xanthidae 
7.97 
2.48** 
10.26 
20.73 
Alpheidae 
5.29 
1.84** 
6.81 
27.54 
O. beta 
5.28 
1.84** 
6.80 
34.34 
Portunidae 
(unidentified) 
4.62 
1.26 
5.95 
40.29 
Brachyura 7 
3.68 
1.37 
4.73 
45.02 
Majidae 
3.25 
1.05 
4.19 
49.21 
Portunus spp. 
2.80 
0.45 
3.60 
52.81 
Stomatopoda 
2.51 
1.16 
3.23 
56.03 
Callinectes spp. 
2.50 
0.97 
3.23 
59.26 
1 Excluding xanthids, portunids, and majids. 
stomach contents of Bahamian bonefish collected 
over different bottom types, and this variation prob- 
ably also occurs in the Florida Keys. There was also 
no evidence that bonefish do not feed in deeper wa- 
ters than those traditionally fished by anglers; prey 
availability and bonefish feeding may be quite dif- 
ferent at greater depths than in the shallow waters 
we sampled. 
Most (77%) of the fish in our sample were longer 
than 500 mm (Fig. 1); consequently, the diet of large 
bonefish is better described by our data than that of 
small bonefish. The inadequacy of our description of 
the diet of small bonefish is reflected in the low lev- 
els of similarity among 20-mm length intervals of 
bonefish smaller than 480 mm (Fig. 2). Many of the 
length intervals smaller than 500 mm contained few 
fish and resulted in greater variation in diet among 
length intervals and probably caused the lower lev- 
els of similarity among 20-mm length groups of small 
bonefish than among large fish. 
The changes in diet as length of bonefish increased 
in general reflect the expansion of the diet to include 
Table 5 
Breakdown into the most important prey groups of the 
mean dissimilarity between stomach contents (percent 
weight) of bonefish, Albula vulpes (480-699 mm FL), 
caught on the ocean side of the Florida Keys (n=50) and in 
Florida Bay (n= 50). Prey groups are listed in order of de- 
creasing contribution to the overall dissimilarity between 
the two study areas. Taxa proportionally more important 
in the diet of ocean-side bonefish than Florida Bay bone- 
fish are shown in bold type. The low values of 8 ( /SD(8 ; ) 
suggest that the data were variable and that no taxa were 
reliable discriminators of either area. Symbols are ex- 
plained in the legend of Table 4. 
Species 
5, 
8/SD(8,) 
8 ; % 
Cum 8 ; % 
Xanthidae 
13.47 
1.15 
17.17 
17.17 
Alpheidae 
8.56 
1.06 
10.91 
28.08 
O. beta 
8.04 
0.82 
10.24 
38.32 
Penaeidae 
7.49 
0.82 
9.55 
47.87 
Callinectes spp. 
4.53 
0.46 
5.78 
53.65 
larger prey such as O. beta and crabs of the genus 
Callinectes. In some cases, for example xanthid and 
portunid crabs, prey size was positively correlated 
with predator length. In other cases, for example O. 
beta, prey were not eaten at all by small bonefish, 
and there was no correlation between prey size and 
predator length among large bonefish. Small but 
abundant crustaceans such as Thor spp., H. zoster- 
icola, and P. americanus were not important in the 
diet of the bonefish we examined. These small crusta- 
ceans may be eaten by smaller bonefish (<228 mm), 
but they are apparently outside of the size range of 
prey typically consumed by the size of bonefish we 
considered. 
Bruger (1974) examined the stomachs of 129 bone- 
fish ranging from 221 to 679 mm FL (reported as 
210 to 656 mm SL) collected from the waters off the 
Keys and reported the frequency of occurrence of 
crustaceans in the diet. Of these 129 stomachs, 19 
were empty. We recalculated his frequency of occur- 
rences on the basis of only the number of stomachs 
with prey (n = 110) to compare with our frequency 
data: the recalculated results were 85% crustaceans, 
33% mollusks, and 17% teleosts. These results are 
in general agreement with our findings of 89% crus- 
taceans, 51% mollusks, and 45% teleosts, except that 
our samples contained more teleosts than Bruger’s. 
Among crustacean prey, Bruger’s results resemble 
ours; penaeid shrimp, alpheid shrimp, portunid 
crabs, and xanthid crabs were the most frequently 
occurring crustacean prey. Although Bruger did not 
quantify by species the fishes found in stomachs, he 
did not include O. beta in his list of teleosts eaten by 
