Crabtree et a!.: Feeding habits of Albu !a vulpes 
761 
Tab!e 3 
Food items found in stomachs of bonefish, Albula vulpes , caught in the waters of the Florida Keys ( m =384 ) by bonefish length 
interval. W = percent weight, F = percent frequency of occurrence, N = percent numerical abundance. 
<440 mm FL fn=611) >439 mm FL (n=323) 
laxon ana 
prey item 
W 
F 
N 
W 
F 
N 
Annelida 
Polychaeta 
3.06 
19.67 
9.54 
1.34 
43.34 
40.08 
Mollusca 
Unidentified Mollusca 
2.29 
19.67 
1.82 
1.56 
14.24 
0.67 
Gastropoda 
4.99 
21.31 
13.18 
2.53 
32.82 
5.07 
Bivalvia 
4.64 
22.95 
3.51 
2.39 
25.70 
2.81 
Crustacea 
Stomatopoda 
2.26 
3.28 
2.52 
2.03 
12.38 
0.79 
Decapoda 
Penaeidae 
40.45 
36.07 
11.50 
6.68 
26.01 
4.37 
Alpheidae 
0.35 
8.20 
1.12 
9.36 
51.70 
17.52 
Hippolytidae 
1.87 
11.48 
7.99 
0.33 
26.01 
6.76 
Majidae 
3.28 
8.20 
0.70 
3.92 
8.98 
0.89 
Portunidae 
15.64 
8.20 
3.37 
10.91 
19.50 
1.83 
Xanthidae 
7.43 
13.11 
1.68 
31.49 
56.97 
8.94 
Chordata 
Teleostei 
3.90 
19.67 
3.09 
21.10 
49.54 
5.07 
Batrachoididae 
Opsanus beta 
0.00 
0.00 
0.00 
17.83 
34.67 
3.43 
fish were significantly different from those of bone- 
fish collected in January-March ( ANOSIM, /?=0.284, 
P=0.018), we have little confidence in this test be- 
cause of the small sample size, and these results are 
not included in Table 7. 
Bonefish fed selectively on some prey groups but 
did not select others. The suite of epibenthic crusta- 
ceans and fishes found in bonefish stomachs was sig- 
nificantly different from that collected with throw 
traps both on the ocean side of the Florida Keys 
(ANOSIM, i?=0.261, P<0.001) and in Florida Bay 
(ANOSIM, f?=0.419, P<0.001). Bonefish on the ocean 
side of the Florida Keys fed selectively on alpheid 
shrimp, xanthid crabs, P. duorarum , and O. beta , 
whereas they did not select the small but abundant 
crustaceans Thor spp. and Periclimenes americanus 
(Table 8). Similarly, Florida Bay bonefish fed selec- 
tively on xanthid crabs, alpheid shrimp, O. beta, P. 
duorarum, and Callinectes spp. but did not select the 
abundant but small crustaceans Thor spp., Hippolyte 
zostericola, and P. americanus, as well as the abun- 
dant goby Gobiosoma robustum (Table 9). 
Discussion 
A variety of factors could have biased our descrip- 
tion of the diet of bonefish. Some prey, particularly 
soft-bodied prey, may have been digested more rap- 
idly than others with bony or chitinous skeletons. 
Consequently, we may have underestimated the im- 
portance of soft-bodied organisms such as polychae- 
tes. Furthermore, bonefish have massive pharyngeal 
tooth plates capable of crushing shells and other hard 
structures. If bonefish are able to expel the crushed 
shells of mollusks and swallow only the soft-bodied 
organism, then we could have underestimated the 
importance of mollusks. This might explain why 
mollusks were relatively unimportant in our samples. 
Our sample consisted principally of bonefish caught 
with hook-and-line gear; therefore most of the fish 
we examined were probably actively foraging or they 
would not have consumed the bait presented by an- 
glers. We do not believe that the number of fish with 
empty guts in our sample reflects the number of fish 
in the area that were not feeding, therefore we did 
not attempt to evaluate temporal feeding patterns. 
We cannot eliminate the possibility that some bone- 
fish regurgitated prey during capture trauma; if some 
prey taxa were more likely to be regurgitated than 
others, this could have biased our results. Most of 
the bonefish in our sample came from relatively shal- 
low (<2 m) grass, sand, or hard-bottom flats, but be- 
cause the fish were caught by anglers, we did not 
have corresponding data on bottom type for each fish. 
Colton and Alevizon (1983) found differences in the 
