16 
Fishery Bulletin 113(1) 
Table 1 
Number of video samples ( N ) included in the analyses of reef fishes in the northern Gulf 
of Mexico, as well as the range of dates, latitudes, and longitudes covered by the samples 
in 2001-2002 and 2004-2007. 
Year 
N 
Date range 
Latitude range (°N) 
Longitude range (°W) 
2001 
42 
6/14-6/22 
27.79-28.35 
91.03-93.82 
2002 
260 
2/22-5/30 
24.50-30.00 
84.26-96.78 
2004 
169 
4/8-6/22 
24.59-30.13 
82.97-96.30 
2005 
350 
4/20-7/29 
24.51-30.13 
82.77-96.53 
2006 
333 
4/16-8/4 
24.53-30.14 
82.77-96.78 
2007 
389 
4/22-8/13 
24.50-30.13 
82.77-96.77 
the time spent surveying and the number of taxa en- 
countered for a wide variety of fish and wildlife species 
(Fuller and Langslow, 1984; St. John et ah, 1990; Bark- 
er et ah, 1993; Gledhill, 2001), and new methods can 
account for detection probabilities of <1 when estimat- 
ing species richness (Nichols et al., 1998; Johnson et 
al., 2013). Reading more frames will certainly provide 
more information but will also bring increased costs 
associated with the additional time and effort required 
(Rotherham et ah, 2007; Al-Chokhachy et ah, 2009). 
Our objective was to examine the tradeoff between 
minimizing the effort needed to read videos and maxi- 
mizing the information obtained. We focused our analy- 
ses on 2 primary response variables, each as a function 
of the number of video frames read. First, we exam- 
ined potential bias and precision of MeanCount for 3 
economically important reef fish species in the Gulf of 
Mexico. Second, we examined estimates of species rich- 
ness, defined here as the number of species observed in 
a video. The results provide general guidance regarding 
the amount of effort that should be expended to read 
underwater videos in diverse aquatic systems. 
Materials and methods 
Data 
We analyzed video data from a long-term reef fish 
monitoring program conducted within U.S. waters of 
the Gulf of Mexico. These data were collected by the 
Southeast Fisheries Science Center, National Marine 
Fisheries Service, in 2001-2002 and 2004-2007. A 
4-camera array was deployed with a soak time of 40 
min on hard-bottom habitats throughout the sampling 
range (Table 1) during the reef fish video survey of the 
Southeast Area Monitoring and Assessment Program. 
Four Sony DCR-VX2000 1 (Sony Corp., Tokyo) camcord- 
ers were mounted orthogonally on a metal array, fac- 
1 Mention of trade names or commercial companies is for iden- 
tification purposes only and does not imply endorsement by 
the National Marine Fisheries Service, NOAA. 
ing outward, at a height of 30 cm above the bottom 
(for more details, see Gledhill 2 ). Each array was bait- 
ed with approximately 0.5 kg of squid (Illex spp.) in 
a mesh bag and deployed during daylight hours only. 
These particular years were selected because they were 
the most recent years for which data were available 
before a significant methodological change occurred in 
video reading procedures. 
The reef fish video survey was developed to index 
reef fish populations and was typically conducted in 
the spring and summer on shelf-edge reefs from south 
Texas to the Dry Tortugas in Florida (Fig. 1). A 2-stage 
sampling design was used to minimize travel time be- 
tween stations because the survey area was large. For 
the first stage, we used a stratified random sampling 
design of randomly selected blocks, each of which was 
10' of latitude by 10' of longitude in size. Blocks were 
stratified by 4 geographic regions and by the amount 
of reef habitat (low or high) present in each block; each 
block was subdivided into a grid of cells that were 0.19 
km by 0.19 km. For the second stage of sampling, cells 
were randomly selected from within each block. The 
number of grid cells available for random selection 
varied depending on how much known reef area was 
contained in the sampled block. 
From each 40-min deployment, 1 of 4 videos was 
randomly selected, and 20 min of that video was ana- 
lyzed beginning at the point when video visibility was 
sufficiently clear for identification of taxa. Fish shape, 
anatomical features, coloration, and swimming behav- 
iors were used to identify individuals to genus and spe- 
cies levels by using field guides (e.g., Hoese and Moore, 
1998; McEachran and Fechhelm, 1998; Carpenter, 2002; 
Humann and Deloach, 2002; McEachran and Fechhelm, 
2005). Video frames were examined every second dur- 
2 Gledhill, C. T., G. W. Ingram Jr., K. R. Rademacher, R Felts, 
B. Trigg, and L. Lombardi-Carlson. 2006. SEAMAP reef 
fish survey of offshore banks: yearly indices of abundance 
of red grouper (Epinephelus morio). SEDAR 12-DW-6, 12 
p. [Available from http://www.sefsc.noaa.gov/sedar/down- 
load/S12%20DW06%20Video-survey.pdf?id=DOCUMENT, ac- 
cessed 31 March 2014.] 
