260 
Fishery Bulletin 106(3) 
that many species had similar neighbors across headboat mutton snapper, and yellowtail snapper (see Appendix 
and commercial data sets. For example, in both data for scientific names). This repeatability of results pro- 
sets, lane snapper was near blue runner, gray snapper, vides evidence of species assemblages. 
The £-medoid cluster analyses identified 
£ = 14 clusters as most compatible with the 
headboat data and £ = 7 clusters as most com- 
patible with the commercial data (Fig. 2). 
These optimal numbers of clusters were not 
cleanly defined because peaks in average 
silhouette widths lacked distinction (Fig. 2). 
In general, assemblages were similar across 
data sets, at least for species that were pres- 
ent in both data sets (Table 1). 
Hierarchical cluster analyses provided as- 
sociations among species that were consistent 
with the assemblages of £-medoid analyses. 
In hierarchical analysis of the headboat data 
(Fig. 3), three assemblages had the strongest 
similarities among member species, labeled 
here as the deepwater assemblage (blueline 
tilefish, snowy grouper, speckled hind, and 
yellowedge grouper), southern assemblage 
(blue runner, gray snapper, lane snapper, 
mutton snapper, and yellowtail snapper), 
and northern assemblage (bank sea bass, 
black sea bass, knobbed porgy, gag, gray 
triggerfish, greater amberjack, red porgy, 
red snapper, scamp, tomtate, vermilion snap- 
per, white grunt, and whitebone porgy). In 
hierarchical analysis of the commercial data, 
the same three assemblages were identified 
with few differences in constituent species 
(Fig. 4). In both data sets, these assemblages 
had the strongest coherence among member 
species, as measured by each cluster’s av- 
erage silhouette width (Table 1). Thus, the 
deepwater, southern, and northern assem- 
blages were examined further for synchrony 
in indices of abundance. 
indices of abundance 
Although data through 2005 were considered, 
indices of the deepwater assemblage were 
derived through 1993, because 1994 began 
regulations that would have invalidated 
catch per effort as an index of abundance 
(i.e., one speckled hind per vessel per trip). 
Deepwater species that met the criterion of 
at least 20% positive trips were speckled 
hind, snowy grouper, and blueline tilefish. 
All southern species met the 20% positive 
trip criterion, however small sample sizes 
of these species north of Cape Canaveral, 
Florida, necessitated combining geographic 
areas into two regions: southern Florida and 
all other areas. Northern species that met the 
20% criterion were white grunt, gag, tomtate, 
black sea bass, vermilion snapper, and gray 
triggerfish. 
Cotwck 
YtlSnp 
BstGnt 
FrnGnt Schmst 
o o 
Prkfsh 
GrySnp LanSnp BluRun 
° °o MtnSnp 
WhtGnt ... q Dl . n 
VrmSnp ° oJltP 9V BI £ Gp A roT rf 
Margat ° 0 OceTrf 
ScyPgy 
BlkMrg 
oTomtat 
RedPgy 
GryTrf 
o 
Gag 
RedGpr 
WtbPgy o 
BndRud 
o 
oBankSB 
RedSnp 
5 Scamp 
oSIkSnp 
Gr vsby“ °SndTlf 
qjKnbPgy 
AlnvJck o RckHnd 
_Hogfsh 
RedHnd 
oGrAjck QenTrf 
o BfnSnp 
Coney 
DogSnp 
NssGpr 
YlfGpr 
Sc up 
AtlSpf 
SpkHnd 
o 
SnwGpr 
o 
WrsGpr 
LgsPgy 
YlmGpr 
CbrSnp 
YdgGpr 
B 
o YtlSnp 
CrvJck GrySnp 
° o 
o LanSnp 
BluRun MtnSnp 
FrnGnt 
o 
BstGnt 
WtbPgy 
KnbPgy 
GrAjck 
BlkGpr 
o 
RedGpr 
o 
Gag 
o 
o GryTrf 
VrmSnp 
Hogfsh 
Margat 
JitPgy 
SnwGpr 
o 
° BluTIf 
RedSnp BIckSB 
o ° 
o RedPgy 
o Scamp 
LsAjck 
SlkSnp 
YdgGpr 
Dimension one 
Figure 1 
Nonmetric multidimensional scaling of species from the (A) head- 
boat and (B) commercial sectors. Distances between points are 
approximately proportional to the dissimilarities between species. 
Abbreviations are explained in the Appendix. 
