58 
Fishery Bulletin 109(1 ) 
Figure 1 
(A) side view and (B) front view of stereo-video baited 
camera system (BotCam). Components include (1) 
ultralow-light video camera, (2) controller-power supply- 
video capture device, (3) bait container, (4) stereo-video 
synchronization device, (5) bait release system, (6) acous- 
tic release, (7) syntactic foam flotation, (8) pressure and 
temperature sensor, (9) aluminum frame. Not shown 
below the acoustic release is the anchor (concrete blocks). 
or three concrete blocks that served as the sacrificial 
anchor. Concrete was used because it is environmentally 
benign, inexpensive, and readily available. BotCam 
was set to float 3 m above the seafloor, thus allowing 
deployments along steep, rocky slopes without risk- 
ing entanglement of the instrument on the bottom. It 
was recovered when it floated to the surface after the 
acoustic release was triggered to separate the sacrificial 
anchor from the buoyant instrument frame. The instru- 
ment can also be tethered to a surface buoy to allow 
recovery by a line haul. 
The complete system, as used during the pilot study, 
cost approximately $40,000; however, the systems be- 
ing used presently with very similar capabilities are 
about $25,000 per unit. The largest single expense is 
the pair of ultra-lowlight cameras. In addition, charter 
time for an appropriate survey vessel in Hawaii runs 
about $1000 per day. 
Study design 
During its development, BotCam was tested in approxi- 
mately 50 deployments around Hawai i, Wake Atoll, 
Guam, and the Commonwealth of the Northern Mariana 
Islands at depths down to 400 m. It was determined 
that 300 m was the maximum reliable deployment depth 
under ambient light conditions that would allow accurate 
species identification and sizing. Further, it was deter- 
mined that by using a 30- to 60-minute recording time, 
a single BotCam unit could be deployed, recovered, and 
ready for redeployment in 90 minutes (Merritt, 2005). 
Ten- to 60-minute deployments are also consistent with 
other shallow baited camera studies (Ellis and DeMar- 
tini, 1995; Willis et ah, 2003). 
Given these constraints and a limited number of 
available charter vessel days, a study site was selected 
relatively close to Honolulu, home port for the charter 
vessel and the Pacific Islands Fisheries Science Center. 
The site was centered on bottomfish habitat located 
along the west side of Penguin Bank, between the Ha- 
waiian Islands of Oahu and Molokai. Penguin Bank has 
historically been a productive bottomfish area and its 
proximity to the highly populated island of Oahu has 
resulted in high fishing pressure on both the east and 
west sides of the bank (Haight et ah, 1993b). 
Previous studies with submersibles and anecdotal 
evidence from bottomfish fishermen have indicated 
that the deep 7 bottomfish species generally prefer 
high-slope, hard-bottom habitats (Kelley et ah, 2006; 
Parke, 2007), which are present at Penguin Bank. 
Twenty-meter resolution bathymetry and backscatter 
data derived from multibeam sonar were available 
for the entire study area and were incorporated into 
a geographic information system in order to derive 
intersections of depth, slope, and substratum hard- 
ness (i.e., backscatter). The upper and lower depth 
boundaries for BotCam deployments were 100 and 300 
m, respectively, set by the biological and logistical con- 
straints given above, with a resulting sampling area of 
24.9 km 2 . Within this depth range, four habitat types 
