214 
Fishery Bulletin 106(2) 
Figure 1 
Maps indicating the position of Turneffe Atoll, Belize (A) and the positions of crest and channel nets around Calabash 
Caye (B). The inset in map A indicates the position of Belize in Central America and the rectangle around Turneffe Atoll 
specifies the area enlarged in map B. 
rier Reef in Australia. Turneffe Atoll is a large offshore 
ring of islands bordered by coral reefs. It has a large 
central lagoon that contains many mangrove islands 
and channels. The atoll is located outside the coastal 
barrier reef, approximately 46 km west of mainland 
Belize (Fig. 1A). 
Larval collection and identification 
The definnition of “larva” will follow that of Leis (2006): 
the posthatching pelagic life history stage of demersal 
fishes (which is equivalent to the presettlement stage 
of Kingsford and Milicich, 1987). Larvae were sampled 
with crest nets and channel nets from 6 July to 26 
August 2005, 24 January to 4 March 2006, and 17 May 
to 28 July 2006. One crest net was positioned in shallow 
water directly behind the reef crest in each of three sites 
approximately 1 km apart (Fig. IB). The crest nets had 
a mean width of 5.85 m, a mesh size of 2 mm, and were 
situated in 65-90 cm of water at each site. One surface 
channel net (Shenker et al., 1993) was placed in each of 
three separate mangrove channels leading to the central 
lagoon, each net with a square mouth (2 mx 1 m) with 
1.6-mm mesh. It was not our intent to optimize the per- 
formance of either net. Therefore, although there were 
differences in net cross-sectional area, mesh sizes, and 
placement locations between crest and channel nets, 
these differences represent how each net has been typi- 
cally deployed. 
In preliminary sampling at Turneffe Atoll, near 
zero or zero catches occurred during daylight hours, 
which was consistent with the findings of Shenker et al. 
(1993). Therefore, collections were made only at night. 
Both types of nets were deployed nightly and the catch 
was retrieved and identified each morning. All indi- 
viduals of all species of larval reef fishes were counted 
each day. Where species could not be determined, the 
lowest taxonomic category that could be unambigu- 
ously determined was used. Larvae were examined live 
and identified (Humann and DeLoach, 2002; Richards, 
2005). Over the course of the study a number of speci- 
mens of all species were preserved in 95% ethanol for 
later validation. 
Environmental variables 
Mechanical flowmeters (model 2030R6, General Ocean- 
ics, Inc., Miami, FL) were deployed with each net. These 
