466 
Fishery Bulletin 1 10(4) 
We also considered the possible emigration of larger, 
older red snapper from other reef sites or an effect of 
nearby reefs not sampled in our study. Proximity to 
other natural or artificial reefs has been shown in other 
studies to be an important factor that can affect density 
of reef fishes (Jessee et ah, 1985; Sogard, 1989; Strel- 
check et al., 2005; Shipley and Cowan, 2010). In our 
study, no significant relations were detected between 
proximity of our study reefs to other reefs and red snap- 
per ages or abundance. In general, other artificial reefs 
were, for the most part, evenly distributed across our 
overall study area (Fig. 1) and would not be expected 
to bias red snapper age distribution to either younger 
or older reef sites in our study (Fig. 4). 
Comparison of collection methods 
This study supports previous studies on the importance 
of using several collection methods to adequately esti- 
mate size and age distribution of red snapper on artificial 
reefs (Myers and Hoenig, 1997; McClanahan and Mangi, 
2004; Szedlmayer, 2007; Wells et ah, 2008a; Gallaway 
et ah, 2009). Hook-and-line and fish-trap methods are 
known to be size selective, and red snapper caught in our 
study were consistently larger with hook-and-line than 
with fish traps. This difference occurred mostly because 
larger fish are able to swallow whole bait and smaller 
fish consume smaller portions. In addition, smaller fish 
are more likely to enter a trap, and larger fish may be 
limited by the size of a trap opening. The distinct size 
differences observed in our study also could have been 
influenced by differences in bait: the fish traps had squid 
in addition to Gulf menhaden, but the hook-and-line bait 
was strictly Gulf menhaden. 
The divers’ visual counts were used to estimate the 
red snapper remaining present on the reef after hook- 
and-line and trap sampling. At 2 sample sites, visual 
surveys were conducted a maximum of 30 days after 
our initial sampling because sharks were present dur- 
ing our initial sampling. Although it is possible that 
additional red snapper immigrated to the reef at these 
2 sites within that 30-d time period, it is unlikely that 
enough fish recruited to cause a bias in our abundance 
estimates. This notion is supported by evidence from 
telemetry studies of high site fidelity for red snapper 
(72% residency rate per year; Topping and Szedlmayer 
2011a) and by the fact that diver counts typically under- 
estimate abundance. In comparison with results from 
visual surveys, counts were significantly lower from 
the video and photographic methods. These differences 
mostly were due to fish swimming throughout the part 
of the water column that was not within the field of 
view of the cameras. A bait jar was used with the intent 
to attract fish closer to the cameras and reduce these 
differences, but it had only limited success. Compari- 
sons of counts from remote, underwater, baited cameras 
with those from scuba-diver surveys have shown similar 
results, with diver visual surveys showing the greatest 
abundance and diversity of fishes among the methods 
compared (Tessier et al., 2005; Langlois et al., 2006). 
Because counts from photographs and video recordings 
were lower, we used the divers’ counts in our estimates 
of red snapper density for each reef. However, the pho- 
tographs and video recordings were still important in 
verifying species identification. 
Artificial reef succession and red snapper densities 
The reefs in our study supported higher densities of red 
snapper than reefs sampled in previous studies. In a 
demolition study of 9 offshore oil platforms, mean den- 
sity of red snapper was 0.24 individuals/m 3 (Gitschlag 
et al. 3 ). In another study of platforms where stationary 
hydroacoustics and visual diver counts were used, mean 
density was 0.16 individuals/m 3 (Stanley and Wilson, 
1997). Substantially higher than these platform esti- 
mates, the estimates from our study of total density of 
red snapper were 1.6-47.9 individuals/m 3 , with a mean 
of 15.7 individuals/m 3 . One difference between our study 
and these previous studies was the larger size of the 
platforms surveyed which also encompassed the entire 
water column. The volume of these platforms varied: 
1037-29,860 m 3 (Gitschlag et al. 3 ) and 19,800 m 3 (Stan- 
ley and Wilson, 1997); in contrast, all reefs in our study 
had a volume of 6.9 m 3 . However, even if the volume 
estimates of these platforms were reduced by two-thirds 
(to account for the habitat in the upper water column 
that red snapper typically do not use), mean densities 
of red snapper on platforms would be 0.73 individuals/ 
m 3 (Gitschlag et al., 2000 3 ) and 0.47 individuals/m 3 
(Stanley and Wilson, 1997) — levels that would still be 
considerably less than the estimates from our study of 
artificial reefs formed from metal cages. 
These differences in the density of red snapper among 
artificial habitats may be due to higher habitat complex- 
ity and associated enhanced protection from predation, 
additional prey resources, and fewer resident larger 
predators at cage reefs than at platforms. The densi- 
ties of lemon damselfish ( Pomacentrus moluccensis ) 
found on highly complex coral reefs with predators 
were similar to densities found on reefs where preda- 
tors were excluded, indicating that these complex coral 
habitats provided protection for this species (Beukers 
and Jones, 1997). Similarly, higher densities of young 
(age-0 and age-1) red snapper were shown to inhabit in- 
creasing complex reef structure (Lingo and Szedlmayer, 
2006; Piko and Szedlmayer, 2007) with an absence of 
predators (Mudrak and Szedlmayer, 2012). At large 
structures, such as platforms, complexity probably is 
lower and the abundance of potential predators likely 
is higher than at the smaller artificial reefs used in our 
study. Therefore, these larger reefs may not support as 
3 Gitschlag, G. R., M. J. Schirripa, and J. E. Powers. 2001. Esti- 
mation of fisheries impacts due to underwater explosives 
used to sever and salvage oil and gas platforms in the U. S. 
Gulf of Mexico. OCS Study MMS 2000-087, 94 p. Final 
report prepared by the National Marine Fisheries Service 
for the U.S. Dept, of the Interior, Minerals Mgmt. Service, 
Gulf of Mexico OCS Region, New Orleans, LA. [Available 
from http://www.boemre.gov/itd/abstracts/2000-087a.html.] 
