Somerton et al.: Quantifying the behavior of fish in response to a moving camera vehicle 
353 
after passage of the towed vehicle. Consequently, we 
felt that the passage of ~20 min after the last distur- 
bance would approximate undisturbed conditions. Un- 
fortunately, schools of vermilion snapper were not ob- 
served before the first passage of the camera vehicle to 
confirm this notion. 
The design of our experiment (i.e., placing fixed 
cameras on the bottom to observe fish behavior in re- 
sponse to moving camera systems) provided a unique 
perspective on how these systems actually sample the 
fish community. During the one pass of the camera ve- 
hicle, for example, no vermilion snapper were counted 
by the moving camera vehicle, but the alternate view 
provided by a benthic camera (Fig. 6) showed that ver- 
milion snapper were indeed within the tow path of the 
camera vehicle but had vacated it immediately before 
they would have been imaged by the forward directed 
cameras on the vehicle. 
Besides confirming that fish avoidance had occurred 
during the pass of the camera vehicle, the use of the 
fixed cameras, especially the use of stereo and optical 
target tracking, allowed quantification of the timing 
and strength of the avoidance behaviors, which in turn, 
provided an indication of the possible stimuli that may 
have elicited these behaviors. For example, vessel noise 
has been implicated repeatedly as a stimulus sufficient 
to trigger fish avoidance behavior (De Robertis and 
Handegard, 2013), but, in this particular pass of the 
camera vehicle, vessel noise did not trigger any overt 
avoidance behavior. The sight or sound of the tow cable 
likely triggered the initial overt avoidance behaviors 
but was insufficient to drive the fish out of the path of 
the camera vehicle. However, the sight or sound of the 
camera vehicle seemed to increase the avoidance be- 
havior sufficiently to cause the fish to vacate the count 
path before they could be seen and counted by the cam- 
era vehicle. These metrics of the timing and strength 
of the fish response could be used to help redesign the 
camera vehicle so that it could be stealthier and pro- 
duce stimuli below the thresholds required to trigger 
avoidance behaviors. 
The purpose for this article is to demonstrate the 
utility of fixed benthic cameras and stereo photography 
to measure the positions of a moving camera system, 
the target fish species, and the bottom topography, as 
well as how these measurements can be used to bet- 
ter understand how moving camera vehicles sample 
fish communities. The example we provide was based 
on one fortuitous pass of the camera vehicle in view 
of a single benthic camera while a school of vermil- 
ion snapper were present; consequently, our observa- 
tions could be quite different had the sampling been at 
greater depths or in more turbid water (Abrahams and 
Kattenfeld, 1997) and certainly would have been differ- 
ent for other target species. However, when the use of 
stereo photography from fixed benthic cameras is ap- 
propriate, it can allow quantification of fish behaviors 
and help define the effect of this behavior on density 
estimates derived from images obtained with moving 
camera vehicles. 
Acknowledgments 
Financial support for this research was provided by the 
National Marine Fisheries Service, Office of Science 
and Technology, through the Untrawlable Habitat Stra- 
tegic Initiative. We thank the captain and crew of the 
RV Pelican for their efforts during the field experiment. 
In addition, we thank S. Stienessen, S. Grasty, and S. 
Murawski for reviewing the manuscript and providing 
helpful suggestions. 
Literature cited 
Abrahams, M. V., and M. G. Kattenfeld. 
1997. The role of turbidity as a constraint on predator- 
prey interactions in aquatic environments. Behav. Ecol. 
Sociobiol. 40:169-174. 
Adams, P. B., J. L. Butler, C. H. Baxter, T. E. Laidig, K. A. 
Dahin, and W.W. Wakefield. 
1995. Population estimates of Pacific coast groundfishes 
from video transects and swept-area trawls. Fish. Bull. 
93:446-455. 
Belcher, E. 
2002. Dual-frequency identification sonar. DIDSON opera- 
tion manual, 4.47.15, 44 p. Sound Metrics, Seattle, WA. 
Bouguet, J.-Y. 
2014. Camera Calibration Toolbox for Matlab. [Available 
from website, accessed November 2014.] 
Campbell, M. D., A. G. Pollack, C. T. Gledhill, T. S. Switzer, 
and D. A. DeVries. 
2015. Comparison of relative abundance indices calculated 
from two methods of generating video count data. Fish. 
Res. 170:125-133. 
Clarke, M. E., N. Tolimieri, and H. Singh. 
2009. Using the Seabed AUV to assess populations of 
groundfish in untrawlable areas. In The future of fish- 
eries science in North America. Fish and Fisheries Se- 
ries, vol. 31 (R. J. Beamish and B. J. Rothschild, eds.), 
p. 357-372. Springer Science + Business Media B.V., 
Dordrecht, Netherlands. 
De Robertis, A., and N. O. Handegard. 
2013. Fish avoidance of research vessels and the efficacy 
of noise-reduced vessels: a review. ICES J Mar. Sci. 70: 
34-45. 
Domenici, P. 
2010. Context-dependent variability in the components of 
fish escape response: integrating locomotor performance 
and behavior. J. Exp. Zool. 313A:59-79. 
Engas, A., and O. R. Godp. 
1989a. The effect of different sweep lengths on the length 
composition of bottom-sampling trawl catches. J. Cons. 
45:263-268. 
1989b. Escape of fish under the fishing line of a Nor- 
wegian sampling trawl and its influence on survey re- 
sults. J. Cons. Int. Explor. Men 45:269-276. 
Grimes, C. B. 
1979. Diet and feeding ecology of the vermilion snapper, 
Rhomboplites aurorubens (Cuvier), from North Carolina 
and South Carolina waters. Bull. Mar. Sci. 29:53-61. 
Handegard, N. O., and D. Tjpstheim. 
2005. When fish meet a trawling vessel: examining the be- 
havior of gadoids using a free-floating buoy and acoustic 
split-beam tracking. Can. J. Fish. Aquat. Sci. 62:2409- 
2422. 
