Romanov Bycatch in the tuna purse seine fisheries of the western Indian Ocean 



91 



111 th(_' \v(-stern Indian Ocean iW'IOi. lima-clolpliin as- 

 sociations are well known in coastal pelagic zones, e.g. 

 Gulf of Aden (Deniidov') and Sri-Lanka (do Silva and Bon- 

 iface-). They are often used in small-scale troll and pole- 

 and-line fisheries for locating yellowfin tuna. In offshore 

 regions of the WIO tuna-dolphin associations are rare, 

 purse seining for them is not practiced, and there is no dol- 

 phin bycatch problem. Perhaps for this reason, the magni- 

 tude of bycatch in the WIO is unknown, except for recent 

 information on species composition (Santana et al.. 1998). 

 Bycatches are not recorded for tuna seiners operating in 

 the WIO, except bycatches of nontarget tuna species. This 

 paper represents a first attempt to estimate catches of as- 

 sociated species by tuna purse seiners in the WIO, based 

 on scarce information collected bv scientific obsen'ers. 



Materials and methods 



Bycatch assessments were based on data collected by Yug- 

 NIRO scientific observers aboard Soviet (since 1992 — Rus- 

 sian) tuna purse seiners in the WIO, during 1987, and 

 1990-91. The vessels were the "Rodina" type.-^ In addition, 

 observer data collected in the same area aboard sister- 

 ships by AtlantNIRO^ and "Zaprybpromrazvedka""' during 

 1986-90 and data by TINRO'^ and TURNIF' during 1990 

 and 1992 were used. The fishing vessels all used purse 

 seines of 1800 m in length, 250-280 m in depth, and 

 90-100 mm mesh size in the bunt. 



The principal goal of the observer sampling program 

 was an estimation of the species composition of catches in 

 this fisheries, biological analysis of the principal species, 

 and estimates of the length and weight compositions of 

 these principal species in the catches. The observers were 

 placed on board opportunistically (i.e. if a vessel had a free 

 sleeping bed and if there was available funding), without 

 a sampling scheme and without preference to any vessel 

 type. Thus, the sampling could be considered as random. 



' Demidov, V. F. 1998. Personal commun. Southern Scien- 

 tific Research Institute of Marine Fisheries and Oceanography 

 (YugNIRO), 2. Sverdlov St., 98300, Kerch, Crimea Ukraine. 



-' de Silva, J, and B. Boniface. 1991. The study of the handline 

 fishery on the west coast of Sri Lanka with special reference to the 

 use of dolphin for locating yellowfin tuna I Thimnuti albacares I. /;i 

 Indo-Pacific Tuna Development and Management Pi'ogramnie 

 (IPTPl Coll. Vol. Work. Doc TWS/90/18., Vol. 4, p. 314-324. Food 

 and Aginculture Organization of the United Nations (FAO), Viale 

 delle Terme di Caracalla, 00100, Rome, Italy. 



■'* Length overall: 85 m; CRT (gross tonnage): 2634; carrying 

 capacity: -1600 mV 



■> AtlantNIRO— The Atlantic Scientific Research Institute of 

 Marine Fisheries and Oceanography, 5 Dmitry Donskoi St., 

 2.36000 Kaliningrad. Russia. 



■'' The Department of Searching and Scientific Research Fleet of 

 the Western Basin "Zaprybpromrazvedka," ^" Dmitry Donskoi 

 St., 236000 Kaliningi-ad. Russia. 



" TINRO— The Pacific Scientific Research Institute of Marine 

 Fisheries and Oceanography, 1 Shevchenko Alley, 690600 VHad- 

 ivostok, Russia. 



' TLIRNIF — The Pacific Department of Fish Searching and Sci- 

 entific Research Fleet, 2 Pervogo Maya St., 690600 Vladivostok, 

 Russia. 



Two other types of Soviet fishing vessels, "Tibiya"*' and 

 "Kauri,""' which took part in the Indian Ocean fisheries 

 during 1985-87 and since 1991 (under the Liberian fiag), 

 were not sampled. In this study coverage rate was esti- 

 mated as percentage of sampled catch to total catch. 



The obsei-vers recorded the results of each set. The type of 

 school, according to Scott ( 1969) and Petit and Stretta ( 1989), 

 of each set was recorded. I considered sets ftir which an ob- 

 server recorded catch in any quantity as positive sets. The 

 average bycatch level was estimated for all positive sets. 



For the positive sets, species composition, total weights, 

 and numbers of each species in the catch were recorded. 

 In the vessels of the "Rodina" type, the retained catch 

 was frozen and stored separately. The retained catch was 

 weighed after freezing while being moved to the ship's 

 holds. In nine cases, the weight of some of the catch was es- 

 timated by the ship masters because the holds were over- 

 loaded and some catch was stored in the freezers till land- 

 ing. Therfore estimates of retained catch are presented in 

 this study as frozen weights rather than wet weights. The 

 bycatch was estimated as wet weight. CJnly bycatch taken 

 on board was sampled. The sets when bycatch was not 

 taken onboard but discarded alive (usually with negligible 

 target species catch) and malfunction sets, which do not 

 produce any catch, were not analyzed in this study. Large 

 species, sharks and billfishes generally, were weighed and 

 counted. The weights of specimens heavier than 200 kg 

 (i.e. Mobulidae) were estimated. When the bycatch was 

 more than 200-300 kg, species composition and weight 

 were estimated by using representative samples. 



Sometimes the obsei-ver recorded the bycatch in num- 

 bers. In these rare cases, the total weights of the fishes 

 were estimated from the average weights of these species 

 in previous catches. 



The obsei-vers had free access to every fish in the catch. 

 Nevertheless, some obsei-vers had difficulties identifying 

 some billfishes, sharks, and Mobulidae species. Therefore, 

 I pooled the records with doubtful species identification 

 into these three groups for my analysis. These are marked 

 by "?" in the tables. 



The data were gi'ouped and analyzed by free-swimming 

 schools (including associations between schools of differ- 

 ent species of tuna) and associated schools. The latter in- 

 cluded whale-associated schools and log-associated schools 

 (associated with floating objects). 



Schools caught in the area of seamounts and shoals — at 

 the peaks of the Equator Seamount and at Saya-de-Malha 

 bank — were considered free-swimming schools. Some ob- 

 servers did not record the type of floating objects that were 

 set on; therefore the sets on natural floating objects (509f to 

 90% of the log sets sampled) and on fish aggregation devices 

 (FADs) (10-50%) were grouped. Several log sets were made 

 in areas with surface evidence of water masses or current 

 interactions (rips). A set that could not be clearly identified 

 as to set tjqpe was made in such an area and was treated as 

 a log set because of the species composition of the catch and 

 the occurrence of small scattered debris in the rips. 



^ Length overall: 55.5 m, GRT: 736, carrying capacity: -361 m-'. 

 '' Length overall: 79.8 m. GRT: 2100. carrying capacity: -1200 m-'. 



