FISHERY BULLETIN: VOL. 71. NO. 3 



skates jointed together as a longline varies but 

 probably average 30 to 40. Anchovy, herring, 

 and squid are preferred bait. The gear is usually 

 set just before dawn and hauled after about 45 

 min of fishing. 



The purse seine was used to capture fish for 

 tagging and migration studies by FRI and by 

 the Fisheries Research Board of Canada 

 (FRBC). The net and method of fishing were 

 described by Hartt (1962). Basically, the net 

 measures 732 m (400 fm) long and 36 m (20 

 fm) deep and has a bunt section of 50.8-mm 

 mesh nylon near one end. A power skiff holds 

 one end of the net while the vessel sets the net 

 in a wide arc. After about 30 min the net is 

 closed and hauled on board the vessel. The bunt 

 section containing the fish remains in the water 

 until the fish are brailed from the net. The opera- 

 tion may be repeated up to 4 times a day if the 

 catches are small and the seas relatively calm. 



The catch data obtained by the various fish- 

 ing gears were reviewed for information on the 

 distribution and abundance of steelhead trout 

 on the high seas. Biological data — age, length, 

 and weight — were taken from specimens col- 

 lected by NMFS. Some of the observations 

 were made at sea on fresh specimens and others 

 on frozen specimens returned to the laboratory 

 for subsequent analysis. Information on steel- 

 head trout migrations was obtained from tag- 

 ging and recovery data provided by FRBC, 

 FRI, and state agencies. 



The catch data were not obtained in a man- 

 ner that precludes objectionable sources of 

 bias. The majority of sets, for instance, were 

 made within 10 m of the surface, and the 

 effort was primarily in the spring and summer. 

 Set and haul times varied daily and seasonally. 

 Loss of gilled fish by dropout and predation 

 varied with sea state, type of enmeshment, 

 time of day, and distance from shore. In near- 

 shore gill net sets, predation by sea lions 

 occasionally reduced the catch per set to zero.- 



Interpretation of the gill net catch data is 

 particularly difficult because of the variations 

 in the gear. The U.S. research vessels have 



- D. R. Craddock. Comparison of gill net and purse 

 seine catches of salmon in the North Pacific Ocean. 

 Northwest Fish. Cent.. Natl. Mar. Fish. Serv., NOAA, 

 Seattle, Wash. [Unpubl. Manuscr.] 



routinely fished with mesh sizes of 64, 83, 115, 

 and 133 mm, but occasionally fished with 51- 

 and 98-mm mesh sizes. The total shackles in a 

 net string have varied from 4 to 40. Monofila- 

 ment and multifilament nylon has been used in 

 net construction. The Japanese gill nets have 

 similarly varied in mesh size, in length and 

 number of shackles, and in type and color of 

 net materials. For this reason Mason (1965), 

 after reviewing the Japanese mothership and 

 research vessel catch data for distribution 

 and abundance of chinook salmon, O. tshawy- 

 tscha, concluded that trends and conclusions 

 can be indicated only very generally. 



Much of the above criticism of gill net data 

 also applies to longline catch data. Further- 

 more, the efficiency of longlines depends on 

 the surface feeding activity of the fish, kind of 

 bait used, size of hooks, and number of baited 

 hooks available to passing fish. 



Most of the purse seine fishing was done 

 within a day's cruising range offshore. This fact 

 alone limits the value of the data for describing 

 either the distribution or abundance of steel- 

 head trout on the high seas. But, since much of 

 the effort centered about the Aleutian Islands, 

 the data provide some valuable information on 

 the relative abundance of steelhead trout in 

 that area. 



The selection of a standard unit of effort 

 involved consideration of the many problems 

 which have been discussed. Ideally, a popula- 

 tion should be sampled with one kind of gear so 

 that all units of effort would be of comparable 

 value (Ricker, 1958). Because of the relatively 

 small steelhead trout catches and the variations 

 in gear, the catch per set was chosen as the 

 comparative unit of effort. 



OFFSHORE FISHING EFFORT AND 

 CATCHES OF STEELHEAD TROUT 



The catches of steelhead trout by research 

 vessels of the United States. Jai)an, and Canada 

 are summarized in Table 1. A summary of the 

 salmon catches and the ratio of salmon to steel- 

 head trout is also given. Assuming equal catch- 

 ability, the apparent abundance of steelhead 

 trout in relation to salmon is provided by the 

 catch data. As indicated, the greatest ratio of 



790 



