414 
Fishery Bulletin 115(3) 
12 
HBi All black sea bass 
I i Nuchal black sea bass 
_ 10 
o 
Sand Sand+rock Live bottom 
Habitat type 
Figure 3 
Average values of MeanCount, the mean number of fish counted in 
a sample of frames from a video, for black sea bass ( Centropristis 
striata) observed in the 3 classified habitat types (sand, sand+rock, 
live bottom) on videos collected from 14 June to 4 August 2011 at 6 
sampling sites in waters off the coast of Maryland. Average values 
are given for 2 categories: all black sea bass and nuchal black sea 
bass, (the latter fish were distinguishable from other individuals by 
a darker body coloration, a nuchal hump, and white fin stripes). The 
error bars indicate the 95% confidence intervals. 
the residuals and the standard error (SE) 
of the random effects around the popula- 
tion intercept were relatively small for each 
model (Table 2). However, intraclass cor- 
relation coefficients (ICC= [Intercept SE]/ 
[Intercept SE+Residual variance]; Zurr 
et al., 2009) were fairly high, indicating 
moderate to strong correlations between 
MeanCount observations from the same 
sampling sites. On average, untransformed 
values of MeanCount and their associated 
variances (all black sea bass, sand=0.51, 
sand+rock=19.21, live bottom=27.07; nuchal 
black sea bass, sand=0.05, sand+rock=0.25, 
live bottom=0.57) were greatest in live bot- 
tom habitats (Fig. 3). Log-transformed val- 
ues of MeanCount were significantly dif- 
ferent between habitat types for both cat- 
egories of black sea bass (Table 2; Fig. 3); 
bait method was not significant (P> 0.05). 
Results from pairwise Tukey’s honestly sig- 
nificant difference tests with a Bonferroni 
correction indicated that log-transformed 
values of MeanCount differed significantly 
between sand and sand+rock habitats (all 
black sea bass, P=0.0Q4; nuchal black sea 
bass, P=0.016) and between sand and live 
bottom habitats (all black sea bass, P=0.003; 
nuchal black sea bass, P=0.002) but not be- 
tween sand+rock and live bottom habitats 
(P>0.05). 
Results of Spearman’s rank correlation 
analysis indicated that values of Mean- 
Count for nuchal black sea bass were significant 
and positively correlated with those for non-nuchal 
black sea bass (p=0.829, P<0.001). Time of first ar- 
rival, ranging from 0.5 to 27.5 min, was latest in 
sand habitats and earliest in live bottom habitats for 
both categories of black sea bass. The range of TFA 
was 0.5-21.5 min for baited trap deployments, with 
a mean of 2.9 min (95% confidence interval [Cl]: 1.7- 
5.1), and 0.5-27.5 min for unbaited trap deployments, 
with a mean of 3.8 min (95% Cl: 0.8-6. 8). MeanCount 
was significantly and negatively correlated with TFA 
for all black sea bass (p= -0.397, P=0.011) but not 
for nuchal black sea bass (P>0.05). Black sea bass 
also arrived earliest in live bottom habitats and lat- 
est in sand habitats when the trap was baited. Mean 
TFA was 3.4 min (95% Cl: -0.3-7. 1) in sand, 2.7 min 
(95% Cl: -0.2-5. 6) in sand+rock, and 0.5 min (95% Cl: 
0. 0-0.0) in live bottom and 8.6 min (95% Cl: 2.2-15.0) 
in sand, 0.6 min (95% Cl: 0.4-0. 8) in sand+rock, and 
0.5 min (95% Cl: 0. 0-0.0) in live bottom for baited 
and unbaited trap deployments, respectively. TFA was 
significantly and negatively correlated with Mean- 
Count for the category of all black sea bass during 
baited trap deployments (p=-0.738, P<0.001) but not 
for unbaited trap deployments or for nuchal black sea 
bass for either baited or unbaited trap deployments 
(P>0.05). 
It was clear from processing videos that general be- 
haviors observed around the camera system depended 
on the type of habitat in the camera view regardless of 
whether bait was present in the trap or not. On sand 
habitats, fish swam past quickly or entered the view 
slowly by moving short distances of 1 m or so before stop- 
ping and resting on the bottom; some fish would lie on 
the bottom without moving for up to 10 min or more. In- 
frequently, antagonistic behaviors were observed when 
large nuchal males chased smaller fish out of the camera 
view. Fish also aggregated when nuchal and non-nuchal 
fish would lie next to each other in groups of 2 or more. 
Other behaviors included nipping at the sediment and 
‘back rubbing’ when fish turned over and rubbed their 
dorsal surface or head on the sand. On structured (e.g., 
rocks, boulders) and live bottom habitats, fish were gen- 
erally present when the camera system landed on the 
bottom. Occasionally black sea bass approached the 
camera system, however they spent the majority of the 
time swimming around and above structures or rest- 
ing on the bottom next to or under rocks and in holes 
or crevices of outcroppings. Approximately 20-30% of 
the behaviors displayed by nuchal black sea bass were 
antagonistic and territorial. For example, in one case, a 
large nuchal male continuously returned to and swam 
around the same rocky outcropping after repeatedly 
chasing other nuchal and non-nuchal fish away. 
