92 



Fisher-y Bulletin 100(1) 



Table 1 



Numbers of sets sampled by year. Positive sets are sets m which an obsei-ver registered catch in any quantity . 



1986 



1987 



1988 



1989 



1990 



1991 



1992 



Total 



Total number of sets 

 Number of positive sets 

 Percentage of sets with catch 



Because tuna purse-seine fishing in the WIO is clearly 

 seasonal (monsoons governing fishing techniques and op- 

 erations), the data were analyzed by season. I followed 

 Romanov's (1982) seasonal divisions, in accordance with 

 long-term average seasonal variations in the monsoon 

 atmospheric circulation for the WIO. The winter season 

 (northeastern monsoon) lasts from December to March. 

 the spring intermonsoon period falls during April and 

 May, the summer (southwestern monsoon) lasts from June 

 to August, and the autumn intermonsoon period lasts 

 from September through November. The wind regime de- 

 termines the onset and duration of the hydrological sea- 

 sons, which do not quite coincide with seasons of atmos- 

 pheric circulation owing to a considerable time lag of the 

 processes occurring in the ocean. However, the wind re- 

 gime is instrumental in determining the tactics of purse 

 seining for tuna; therefore I used seasonal strata based on 

 atmospheric rather than on hydrological processes. 



The spatial and temporal distribution of catch and ef- 

 fort for the Soviet tuna purse-seine fishery in the Indian 

 Ocean was determined from data in the YugNIRO data- 

 base, a collection of daily radio reports from vessels fishing 

 in the area from 1983 until the mid- 1990s. i" The catches 

 reported by the author's estimates varied by 96-99'7f dur- 

 ing 1985-91, decreasing to 71% in 1992. This study did 

 not take into account reflagging of some Soviet (from 

 1992 — Russian) vessels with the Liberian flag, and the 

 vessels' nationality was defined in this study by the loca- 

 tion of their shipowners. Analysis of fleet activity and ex- 



'" Daily information on fishing activity of these vessels in the 

 Indian Ocean in 1983-84 and since 1995 is not available. 



trapolations of results were made on the assumption that 

 the operations and procedures on vessels that did not car- 

 ry observers did not differ from the operations and proce- 

 dures on vessels with an obsei"ver aboard; similarly it was 

 assumed that the species composition of the catch from 

 these vessels did not differ. 



Some of the bycatch was retained on board the fishing 

 vessels. Unused bycatch was discarded in the ocean. The 

 observers usually did not record the levels of discards, and 

 it was not possible to assess quantitatively the discards of 

 tuna and associated species. 



Average values are presented as arithmetic means, plus 

 or minus 95% confidence intervals for estimated values. 

 Estimates of unrecorded bycatches for all fishes, except tu- 

 nas, are provided in numbers and metric tons per positive 

 set and per 1000 t of target species. 



Results 



Primary data and adequacy of samples 



A total of 494 purse-seine sets were sampled and 377 posi- 

 tive sets were analyzed. The total catch in the sets that 

 were sampled amounted to 7713 t. The distribution of 

 sets sampled by years, seasons, and the types of schools 

 is given in Tables 1 and 2. The catch sampled by type of 

 school is presented in Table 3. 



The obsei-ver coverage rate varied from 0% (no obsei-v- 

 ers at sea) to 75% and averaged 14% during 1986-92. Dur- 

 ing the periods when observers were on board, the cover- 

 age rate averaged 30% and varied from 5% to 75%. The 

 spatial distribution of sampled sets agi-eed quite well with 



