88 
Fishery Bulletin 11 7(1-2) 
dependent surveys for sharks exist, most of them are 
spatially limited (Peterson et al., 2017). 
The Southeast Reef Fish Survey (SERFS) is a fish¬ 
ery-independent monitoring and research program that 
targets reef fish species in continental shelf waters off 
the southeastern United States and is funded by the 
National Marine Fisheries Service. Annual sampling of 
this collaborative program, for which the Beaufort Lab¬ 
oratory of the Southeast Fisheries Science Center is a 
contributor, is done by using video cameras attached to 
chevron traps to survey a region from St. Lucie Inlet, 
Florida, to Cape Hatteras, North Carolina (Bacheler 
et al., 2014). This fishery-independent trap-video sur¬ 
vey generates -1500 video recordings, or samples, per 
year and targets species from hard substrates (lime¬ 
stone or hard-bottom reefs) on the continental shelf 
and continental shelf break of the southeastern Unit¬ 
ed States. Because of time and personnel constraints, 
SERFS staff primarily evaluate videos for assessment 
of spatial distribution of reef fish species in the snap¬ 
per-grouper complex and use a video-sampling method 
that involves viewing a subset of frames (number of 
frames [/i]=41) from each video sample to calculate 
relative abundance (Bacheler et al., 2014; Schobernd 
et al., 2014). This video-sampling method may not be 
effective for species, such as sharks, that tend to be 
relatively rare or highly mobile (Schobernd et al., 2014; 
Bacheler and Shertzer, 2015). 
Video is increasingly used as a fish survey and re¬ 
search tool in marine ecosystem research (see the re¬ 
view by Murphy and Jenkins, 2010), including for the 
study of elasmobranchs (Meekan and Cappo, 2004; 
Bond et al., 2012; White et al., 2013; Barley et al., 
2017). The advantages of the use of remotely viewable 
underwater video (often recorded by cameras on bait¬ 
ed stations) over traditional, direct underwater visual 
census (UVC) by scuba divers include the potential for 
longer survey times, removal of diver effects (attraction 
or avoidance) on target species (Kulbicki, 1998), and 
greater depth range and spatial replication. Particu¬ 
larly for relatively rare, large, and mobile species, such 
as elasmobranchs, UVC may produce biased estimates 
of population density (MacNeil et al., 2008; Ward-Paige 
et al., 2010; McCauley et al., 2012; Osgood and Baum, 
2015). 
The work described in this paper was conducted as 
part of a study (Bacheler et al., 2017) that compared 
observations of reef fish abundance from 3 techniques: 
diver census, video analysis, and sampling with traps 
(sampling of videos and traps was conducted with 
SERFS methods). The objective of the work reported 
here was to assess the utility of an alternative, rapid 
method of video analysis for generating data specifi¬ 
cally for sharks. For this assessment, we used video 
recordings collected from sampling locations used by 
Bacheler et al. (2017). We determined species compo¬ 
sition and relative abundance of sharks encountered 
in the Atlantic Ocean off east-central Florida by using 
UVC conducted by divers and analysis of videos from 
baited remote underwater video stations (BRUVS), 
and we compared our results with those of Bacheler 
et al. (2017). Given that diver and video methods are 
known to sample different members of fish communi¬ 
ties (Colton and Swearer, 2010; Langlois et al., 2010; 
Watson et al., 2010; Barley et al., 2017), we also spe¬ 
cifically compared the composition and relative abun¬ 
dance of shark species observed with these sampling 
approaches. The SERFS was designed to sample rela¬ 
tively sedentary reef fish species at a regional scale 
with a large number of sites across the continental 
shelf of the southeastern United States. We hypothe¬ 
sized that this work would reveal that the use of video 
can be a non-consumptive alternative or supplement 
to traditional sampling methods that require extrac¬ 
tion (e.g., with longlines and gillnets) for providing es¬ 
timates of relative abundance for a variety of sharks in 
the Atlantic Ocean. 
Materials and methods 
Study area 
We conducted our study at a number of hard-bottom 
reef sites in temperate waters on the U.S. continental 
shelf off east-central Florida (Fig. 1). The continental 
shelf along the southeastern United States consists 
primarily of sand and mud substrates, but patches of 
hard, rocky reefs in temperate waters are scattered 
throughout the region and are important habitat for 
many reef fish species. Specific hard-bottom sites were 
chosen from sampling locations in the study area of 
the SERFS (Bacheler et al., 2014) by using 2 criteria: 
depths safe for diving (<32 m) and a history of traps 
catching reef fish. Hard-bottom habitats sampled in 
our work ranged from rocky ledges to patchily distrib¬ 
uted rock outcrops and pavement, sometimes covered 
in a veneer of sand. All sampling occurred at depths 
between 18 and 30 m on 2-7 July 2014 aboard the MW 
Spree, a 30.2-m aluminum-hulled vessel converted from 
use for oil field services to a live-aboard dive platform 
that was fitted with a pot hauler for this effort. 
Chevron traps 
We deployed a single chevron trap at each site sampled 
in this study. Chevron traps were shaped like an ar¬ 
rowhead and measured 1.7 m x 1.5 m x 0.6 m, with a 
volume of -0.91 m 3 (Bacheler et al., 2014). The mouth 
of each trap, shaped like an upside-down teardrop, 
measured -45 cm high and 18 cm wide, and the mesh 
size of each trap was 3.4 cm 2 . We baited each trap with 
-4 kg of menhaden (Brevoortia spp.). A soak time of 
90 min was targeted for each trap, but actual soak 
time varied from 51 to 99 min, with a mean of 80.2 
min (standard deviation [SD] 11.2). Each trap had a 
single 8-mm polypropylene line connecting it to 2 sur¬ 
face floats. Chevron traps were deployed with an aver¬ 
age distance of 813 m separating them to provide some 
measure of independence between traps. The minimum 
