760 
Fishery Bulletin 96(4), 1 998 
significantly different from zero, the difference was 
small and thus the differences between the stomach 
contents of Florida Bay and ocean-side bonefish were 
probably slight. Xanthid crabs, alpheid shrimp, O. 
beta, penaeid shrimp, and Callinectes spp. together 
made up over 50% of the dissimilarity between the 
two areas (Table 5). Although these taxa contributed 
to the overall level of dissimilarity, the ratios (8 / 
SD(8 ; )) between the mean contribution (5 ; ) to the 
overall level of dissimilarity and the standard devia- 
tion of the 8 values across all stomachs were low for 
each prey taxa. Thus, these taxa did not consistently 
contribute to the level of dissimilarity, and are prob- 
ably not reliable discriminating prey taxa character- 
istic of either area. In both areas, the same prey taxa 
dominated the diet (Table 6). 
A seasonal effect on feeding was 
found for ocean-side bonefish 
Bray-Curtis similarity 
0. 20. 40. 60. 80. 
FL 
GP 
n 
290 
2 
2 
310 
3 
6 
270 
1 
1 
330 
4 
5 
350 
5 
11 
370 
6 
10 
390 
7 
10 
430 
9 
5 
410 
8 
10 
450 
10 
7 
470 
11 
10 
490 
12 
10 
510 
13 
12 
530 
14 
17 
550 
15 
22 
670 
21 
24 
590 
17 
48 
610 
18 
41 
570 
16 
33 
650 
20 
41 
630 
19 
43 
690 
22 
15 
710 
23 
2 
100 . 
Figure 2 
Dendrogram and multidimensional scaling (MDS) ordination showing similari- 
ties in diet among bonefish, Albula vulpes, grouped into 20-mm length intervals. 
Stress for the MDS plot = 0.13. FL = median fork length of the 20-mm length 
intervals, GP = group number used on the MDS ordination, and n = the number 
of bonefish in each 20-mm length interval. The vertical bars to the right of the 
cluster show the two principal length groupings referred to in the text and the 
480-699 mm length grouping used for all statistical comparisons. The circled 
groups in the MDS ordination correspond to the principal groupings shown on 
the cluster dendrogram and referred to in the text. 
(ANOSIM, 72=0.069; P=0.002) but 
not for bonefish collected from 
Florida Bay (ANOSIM, 72=0.066; 
P=0.080). For both tests, the R val- 
ues were close enough to zero to 
suggest that any seasonal differ- 
ences in diet were slight. On the 
ocean side of the Keys, pairwise 
tests between bonefish caught dur- 
ing January-March and those 
caught during all other seasons 
were significant at P<0.05; no 
other pairwise seasonal compari- 
sons were significant. Xanthid 
crabs, alpheid shrimp, brachyuran 
crabs (excluding xanthids, por- 
tunids, and majids), O. beta, 
penaeid shrimp, and stomatopods 
accounted for most of the dissimi- 
larity between stomach contents of 
bonefish collected during Janu- 
ary-March and those of bonefish 
collected during other seasons 
(Table 7). The ratios (8-/SD(8.)) be- 
tween the mean contribution (8 ; ) 
to the overall level of dissimilarity 
and the standard deviation of the 
values across all stomachs were 
low for each prey taxa. No taxa 
consistently contributed to the 
level of dissimilarity, and there 
were no reliable discriminating 
prey taxa characteristic of any par- 
ticular season. Variable sample 
sizes between seasons in both ar- 
eas reduced the power of our sea- 
sonal comparisons; most ocean- 
side bonefish were caught during 
January-May, and most Florida 
Bay bonefish were caught during 
June-December. Only six stom- 
achs were examined from ocean- 
side bonefish captured during 
July-September. Although the 
stomach contents of these six bone- 
