NOTE Gobert et al.: Biology of Etelis oculatus in the Caribbean 



421 



5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 



6 i 



5 



4 - 



3 



2 - 



1 







5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 



5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 

 FL (cm) 



Figure 3 



Length-frequency distributions for queen snapper {.Etelis 

 oculatus) catches. (A) Semi-industrial deepwater Hon- 

 durian line fishery (rc=3415). (B) Semi-industrial shal- 

 low-water Hondurian line fishery (n=387). (C) Artisanal 

 line fishery of Roatan (Honduras) (ra=52). (D) Gillnet 

 exploratory fishing in Guadeloupe (rc=779). (E) Trammel- 

 net exploratory fishing in all areas: males (h=231). (F) 

 Trammel-net exploratory fishing in all areas: females 

 (n=227). 



5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 



5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90 

 FL (cm) 



Figure 3 (continued) 



was particulary clear for trammel nets (Fig. 3, E and 

 F): the mode corresponding to fish gilled in the small- 

 mesh central panel had similar characteristics for both 

 sexes (range 25-45 cm and peak at 36 cm), as opposed 

 to the diffuse mode for fish >45 or 50 cm (predominantly 

 females) entangled in the large-mesh outer panels. With 

 many fewer fish («=23 for both sexes), the longline sam- 

 ples showed a similar difference between sizes of males 

 (maximum 55 cm, mean 43.8 cm) and females (maximum 

 71 cm, mean 51.8 cm). In Guadeloupe, the sex offish was 

 not determined, but the existence of two modes in the 

 overall size structure of gillnet catches (Fig. 3D) could 

 possibly be related to this sex-related length difference. 



Growth and mortality 



The data collected in the various surveys did not 

 allow any reliable analysis of the growth of E. ocula- 

 tus. Because growth may be different for males and 

 females, the length-frequency distributions of the large 

 samples (where sex was not determined) from Honduras 

 could not be processed rigorously to estimate life-history 

 population parameters. However, in order to provide 

 preliminary information on such a little known species, 

 the regression method of Wetherall et al. (1987) was used 

 in the modified version of FiSAT (Gayanilo et al., 1996) 

 to estimate L x and Z/K. With a satisfactory fit of the 

 regression line (r=0.986), the estimates were L, = 90.57 

 cm and Z/K = 3.73. For the reason mentioned above 

 (together with other weaknesses related to possible 

 violations of the hypotheses underlying the regression 



