420 



Fishery Bulletin 103(2) 



between 140 and 430 m. In Martinique, the 

 trammel nets were set between 100 and 300 m 

 but did not catch any E. oculatus in the shal- 

 lowest part of this range. In Guadeloupe, gill 

 nets were set down to 410 m, but the deepest 

 catch of queen snapper was 340 m. No E. 

 oculatus were caught in shallower (<80 m) 

 fishing experiments with any of the gears 

 used (traps, gill nets, trammel nets, and long- 

 lines) on the SMSBA shelf. According to some 

 local fishermen, however, queen snappers can 

 be caught from about 100 m down to 550 m 

 (Lorance 2 ). 



Depth-size relationship 



100 



150 



200 



250 300 



Depth (m) 



350 



400 



450 



No clear relationship between depth and aver- 

 age size offish was found in the fishing experi- 

 ments (Fig. 2). This is not unexpected given 

 the selectivity of some gears (gill nets) and 

 the small sample sizes in most depth strata 

 outside the main fishing range (250-300 m); 

 70% of the 456 fish caught by longlines were in the 290 

 m depth stratum, and five or fewer fish were caught in 

 most of the other strata. 



A different picture emerges from the analysis of the 

 professional fisheries of Honduras. Multivariate analy- 

 sis (principal component analysis followed by hierar- 

 chical classification) applied to the landings by species 

 revealed the two different categories of fish caught by 

 the two types of semi-industrial vessels operating from 

 Roatan (de Rodellec 6 ), the shelf-operating fleet and the 

 slope-operating fleet. The first category of fish were 

 dominated by shallow species such as Ocyurus chrys- 

 urus (59.8%), Lutjanus analis (7.8%), and several grunts 

 (Haemulidae), whereas E. oculatus accounted for only 

 2.2%. On the other hand, the second category comprised 

 mainly deep snappers: L. vivanus (39.6%), E. oculatus 

 (22.4%), R. aurorubens (6.9%) or L. buccanella (1.9%). 

 The two divisions of the fleet independently exploit the 

 continental shelf and the deep slope. Although actual 

 depth of fishing operations is unknown, the shelf-oper- 

 ating fleet probably catches E. oculatus in the deepest 

 part of its working area (i.e., at the shallowest part of 

 the species' bathy metric range), whereas the slope-op- 

 erating fleet exploits the main habitat of the deepwater 

 snappers. The size structures of Etelis catches (Fig. 3, A 

 and B) strongly indicate that only the fish up to 45-50 

 cm live on the shelf or its edge, whereas individuals of 

 all sizes, and particularly the largest ones, inhabit the 

 shelf slope. 



A similar observation was made for the island of 

 Roatan, where the artisanal fleet is the least developed 

 of the archipelago: fishermen using small (<6 m) and 

 often (57%) nonpowered canoes fish quite close to the 

 shore and catch a large diversity of coastal reef fishes, 

 a large proportion of which are juveniles. Etelis oculatus 

 is rarely caught by these small-scale fishermen but is 

 so only as individuals smaller than 50 cm, sometimes 

 as small as 16 cm (Fig. 3C). 



Figure 2 



Average fork length of queen snapper {Etelis oculatus) by depth (m) 

 strata in the fishing experiments with gill nets (circles), trammel 

 nets (squares), and lines (triangles). Sample sizes are indicated 

 by size of the symbols: empty symbol (rc<10), filled symbol by 

 increasing size (/i = ll-20, 21-50, 51-100. >100). 



Habitat of early juveniles 



Some observations were made on very small (smaller 

 than 10 cm) individuals of E. oculatus. Off Guadeloupe, 

 a few of them were entangled in gill nets at 300 m depth 

 (Fig. 3D); on the same island, previous exploratory fish- 

 ing operations with small-mesh traps caught six juve- 

 niles ranging from 5.5 to 7 cm FL at 490 m depth; off 

 Dominica, one small individual (8.5 cm TL) was found 

 in the stomach of a predator caught at a depth greater 

 than 200 m (see below). In spite of the general tendency 

 of increasing size with depth found for the larger indi- 

 viduals, these observations show that the habitat of 

 early postsettlement juveniles is not restricted to the 

 shallowest part of the species depth range. 



Morphometric relationships 



The main morphometric relations were computed from 

 the fish sampled in commercial or scientific fishing 

 operations in the Lesser Antilles (Martinique, Saint- 

 Lucia, SMSBA shelf). Because the differences between 

 relations for males and females were insignificant, only 

 global equations are given (Table 2). 



Maximum size and weight 



The largest individual caught was 90 cm FL in the 

 Lesser Antilles (Guadeloupe) and 86 cm in Honduras, 

 and the maximum weight recorded was 6280 g, in the 

 Lesser Antilles; fish were not weighed individually in 

 Honduras. 



Sex-related length differences 



When sex was recorded, the largest fish were always 

 female, and no male was found above 70 cm. The differ- 

 ence between size-structure of male and female catches 



