-133- 



Indian Ocean. There are few definitive data to support either a single or 

 multiple stock hypothesis, but similar types of data suggest the 

 possibility of multiple stocks for striped marlin, black marlin, swordfish, 

 and sail fish in this area. 



III.B. Stock Assessment and Fishery Evaluation 



III.B.l. Trends in Catch-per-Unit-Effort 



There was a decline in the catch-per-unit-effort over the 

 1952 to 1976 period in the Japanese longline fishery for blue marlin 

 (Figure 3), striped marlin (Figure 6), and black marlin (Figure 5). Catch 

 rates for swordfish have not declined significantly (Figure 6), and catch 

 rates for sail fish (including small amounts of shortbill spearfish) have 

 been quite variable although a general increase over the 1952 to 1976 

 period is evident (Figure 7). 



III.B. 2. Results of Production Model Analysis 



The following production model analyses were applied to 

 available data for Indian Ocean bill fishes to obtain a better indication of 

 the status of the stocks (FAO, in press): 



The production model analysis for blue marlin suggested 

 a MSY (maximum sustainable yield) of 3,400-3,600 mt achievable with 

 substantially less effort than has recently been expended (Table 4). 

 However, the curve fits the data points poorly (Figure 8), suggesting that 

 there may be significant errors in the total catch estimates. 



For striped marlin, assuming a single stock, the MSY 

 from the production model analysis is 3,500 mt at a level of effort 

 considerably more than the highest effort expended to date (Table 4). 

 Annual catches at effort levels of 150 to 300 million hooks are widely 

 variable about the equilibrium yield curve (Figure 9). 



The estimated MSY for black marlin is 1,400 to 1,500 mt 

 (Table 4). Optimum effort is less than those of recent levels but the 

 yield-effort curve (Figure 10) is quite flat. Effort levels of 40 to 60 

 million hooks have resulted in catches with considerable variation about 

 the equilibrium yield curve. The catch data points for 1972 to 1976 are 

 below the curve, whereas the points for 1968 to 1971 are all above the 

 curve, suggesting the possibility of some systematic bias in the above 

 estimates. 



The data for swordfish suggest that longlining has had 

 no significant impact on the swordfish population, so no production model 

 analysis was attempted. 



