Koslow et al.: Catch and effort analysis of reef fisheries 



741 



The fishing effort and catch rates of the two coun- 

 tries differed considerably. The mean fishing effort 

 per unit area on the Jamaican grounds was 

 fifteenfold higher than off Belize: 3,098 hook h/km 2 yr 

 (equivalent to 527 trap hauls/km 2 ) in Jamaica and 

 210 hook h/km 2 -yr in Belize (KW: x 2 =9.00, P<0.005). 

 However, catch rates were ninefold higher in Belize: 

 1.61 kg/hook h compared with 0.18 kg/hook h (equiva- 

 lent to 1.06 kg/trap haul) in Jamaica (KW: x 2 =9.00, 

 P<0.005). 



fishing grounds (Fig. 2). Landings of the 

 main species groups were approxi- 

 mately log-normally distributed among 

 regions within each country, especially 

 in Belize, where landings per unit of 

 area generally varied among fishing 

 grounds by two to three orders of mag- 

 nitude. In Jamaica the differences were 

 generally closer to two orders of magni- 

 tude. Thus in Belize, landings per unit 

 of area of lutjanids were highest in ar- 

 eas 4 (west Ambergris Cay) and 7 

 (Placencia); of serranids in areas 3 and 

 4 (Gallows Point and west Ambergris Cay); and of 

 haemulids in areas 3 and 5 (Gallows Pt. and east 

 Ambergris Cay). Several of these areas were sites of 

 major spawning aggregations (S. Auil, unpubl. data). 

 The outer atolls, Halfmoon Cay and Turneffe (areas 

 1 and 6), did not appear to be intensively fished for 

 finfish. In Jamaica, lutjanid and haemulid landings 

 were higher in areas 1 and 2; serranid landings were 

 higher in areas 2 and 3. The catches of low-valued 

 fish were more evenly distributed. 



Catch composition 



The composition of the fishery also was substantially 

 different in the two countries (Fig. 2). Prime com- 

 mercial fishes from the Lutjanidae (snappers) and 

 Serranidae (groupers) dominated the Belizean fish- 

 ery, representing 74% and 11% of the catch, respec- 

 tively. In contrast, lutjanids represented 23% of the 

 Jamaican catch and serranids only 2%. Of the land- 

 ings in Jamaica, 62% were of low-value species, fishes 

 in the families Scaridae, Sparidae, Labridae, 

 Mullidae, Holocentridae, and Acanthuridae. Another 

 14% were haemulids, which composed only 2% of the 

 catch in Belize. When the data were aggregated by 

 area, the differences in catch composition between 

 the countries were all significantly different (Table 

 3), as were the differences in actual 

 catch for all groups except lutjanids. 

 The catch of serranids was significantly 

 higher in Belize and that of haemulids 

 and 'other' fish was higher in Jamaica. 

 When the data were examined on the 

 basis of individual landings, the num- 

 ber of degrees of freedom was greatly 

 increased. Differences were highly signifi- 

 cant for all groups: landings of serranids 

 and lutjanids were again higher in Belize; 

 landings of haemulids and 'other' fishes 

 were higher in Jamaica (Table 3). 



Within each country, species compo- 

 sition also varied significantly among 



Surplus production models 



Because of the heterogeneity of the fishery, we ex- 

 amined catch-effort relationships for species groups, 

 both individually and combined. The slopes of the 

 relationships of CPUE with effort (f) were nonsig- 

 nificant but were positive in sign for the Jamaican 

 and Belizean reef fisheries considered separately 

 (Table 4). When the data for the two countries were 

 combined, the relationship between CPUE and /"was 

 negative (Table 4, Fig. 3A). A linear relationship, 

 obtained after log-transforming the data on f, was 

 due largely to the substantial difference in f and 

 CPUE between the two countries (Fig. 3B). Based 

 upon these relationships, MSY for the total reef fish- 

 eries was estimated to be 1,046 kg/km 2 of productive 



Table 3 



Results of Kruskal-Wallis one-way ANOVA to test for differences in 

 catch composition between Jamaica and Belize for the data shown in 

 Figure 2. (A) test for differences in proportion of catch by fish groups 

 aggregated by area (n = 13); (B): test for differences in landings offish 

 groups aggregated by area (n = 13); (C): test for differences in land- 

 ings by individual landing (n=503). The statistic shown is the x 2 value. 

 *P< 0.05; **P< 0.01; ***P< 0.001. NS = not significant. 



Lutjanidae 



Serranidae 



Haemulidae 



Other 



A 

 B 



C 



7.37** 



2.47 NS 



19.42*** 



6.61* 

 5.22* 



38.2*** 



4.00* 



4.00* 



52.6*** 



7.37** 



4.00* 



12.2*** 



