52 
Fishery Bulletin 1 14(1) 
Table 3 ( continued) 
AS/NHB 
SS 
Taxa 
84.42 h, 63-126 m 
16.48 h, 40-126 m 
Labridae 
Tautoga onitis, tautog 
0.008 
Tautogolabrus adspersus, cunner 
0.729 
0.254 
Labridae (unidentified) 
0.002 
Caproidae 
Antigonia capros , deepbody boarfish 
4.776 
21.628 
Paralichthyidae 
Paralichthys dentatus, summer flounder 
0.002 
0.085 
Paralichthys oblongus, fourspot flounder 
0.004 
0.085 
Bothidae (unidentified), lefteye flounders 
0.254 
Pleuronectidae 
Glyptocephalus cynoglossus, witch flounder 
0.002 
Hippoglossoides platessoides, American plaice 
0.002 
Cynoglossidae 
Symphurus stigmosus, blotchfin tonguefish 
0.085 
Molidae 
Mola mola, ocean sunfish 
0.002 
Unidentified fishes 
0.096 
0.339 
Chain dogfish were less abundant on the natural 
hard bottom (site NHB) than on the shipwreck (AS) 
or SS habitats. They occurred in massive numbers at 
the shipwreck sites, where they were often so densely 
packed that they lay on top of each other in layers that 
were several individuals thick (Fig. 2B). Individuals 
and aggregations of individuals were observed on all 
areas of the shipwreck sites, including on and within 
the trawl nets that covered sections of the shipwrecks. 
Although many chain dogfish were observed lying on 
soft bottom, they did so generally within tens of meters 
of the shipwrecks. The aggregations of chain dogfish 
probably reflect activity related to spawning because 
thousands of their egg cases were attached to the ship- 
wreck structures and the nets that covered them (Fig. 
2B). 
Because shelf communities are subjected to sea- 
sonal environmental variability and may exhibit sea- 
sonal distribution patterns, multivariate analysis was 
used to examine seasonal differences (fall 2012 versus 
spring 2013) in fish distributions at the 4 study sites 
that were sampled during both seasons; 17 video sam- 
ples (8 for fall, 9 for spring) and 29 species were ex- 
amined in this analysis. Season did not have a signifi- 
cant impact on fish assemblages (ANOSIM, coefficient 
of multiple correlation [f?] = -0.024, P= 0.55). Likewise, 
there were no differences in assemblage structure over 
the limited depth range examined (8 sites, depths of 
68-126 m, dive NF-21 at site SS excluded, f?=0.026, 
P=0.400). The greatest distances between sites were 
no more than 50 km, and fish assemblages (excluding 
observations from dive NF-21) at the 8 sites were not 
significantly different (P=0.130, P=0.090) in regard to 
distance from one another or distance from Norfolk 
Canyon. Therefore, all data were combined for analysis 
of habitat influence on fish assemblages. 
Multivariate analysis of 26 video samples (exclud- 
ing those from the shallow dive NF-21) and 41 taxa, 
indicated a significant difference (P=0.499, P=0.001) in 
fish assemblage structure between the soft bottom (SS) 
and hard bottom (AS/NHB) habitat types (Fig. 3). The 
video samples from the AS/NHB habitat type were 60% 
dissimilar from the 2 sample groups affiliated with the 
SS habitat type; the video sample associated with nat- 
ural hard-bottom habitat (at site NHB) grouped with 
the shipwreck hard-bottom (AS) samples (dive number 
28, Fig. 3). The fishes most influencing the group of 
samples from the AS/NHB habitat type (on the basis 
of SIMPER analysis) were the chain dogfish, members 
of Anthiinae, the yellowfin bass, the deepbody boarfish, 
the conger eel (Conger oceanicus ), Scoi'paena sp., the 
red barbier, the cunner ( Tautogolabrus adspersus ), and 
the blueline tilefish ( Caulolatilus microps). Fishes most 
influencing the groups of samples from the SS habitat 
type were the chain dogfish, Scorpaena sp., the deep- 
body boarfish, and the black sea bass. 
Data indicates at least some difference in fish com- 
munities along isobaths. Within the hard bottom group, 
video samples (n=9) from the 4 deeper shipwreck habi- 
tats (AS) north of Norfolk Canyon (depths of 91-126 m; 
Fig. 1) grouped closely together (Fig. 3), although data 
were collected in 2 different years and seasons. The 3 
shipwreck hard-bottom (AS) video samples from the 
shallower middle shelf (depths of 68-85 m; Fig. 1) were 
offset together in the overall group of samples from the 
AS/NHB habitat type (dive numbers 22, 29, 30; Fig. 3). 
The 3 samples from sandy bottom on the middle shelf 
(SS habitat type, dive numbers 22, 29, 30; Fig. 3) also 
