Table 1. --Fishing locations and numbers of salmon tagged in experiments off south shore of 

 Unimak Island and southwestern shore of Alaska Peninsula, 1961--Continued 



Recovery percentages of tagged salmon 

 were low; the maxinnum was 6.4 percent for 

 sockeye. An important factor in interpreting 

 the tag recoveries was the extreme variation 

 in recovery efforts in respect to both time 

 and place. It is possible that large numbers 

 of tagged fish migrated to uninhabited or 

 unfished spawning tributaries of the Bering 

 Sea. In Bristol Bay the fishing season for 

 sockeye salmon was closed during the last 

 part of the summer to permit adequate 

 spawning escapement, thereby eliminating 

 many potential tag recoveries in the area. 



On the other hand, it is probable that nearly 

 all of the tagged fish captured in the com- 

 mercial fishery in the tagging area were 

 reported. The California Rose , which fished 

 among the commercial fleet, was plainly 

 marked as a research vessel. Bureau em- 

 ployees, responsible for tag recoveries on 

 the fishing grounds, were aboard cannery 

 tenders operating out of False Pass and 

 King Cove. 



The advantage of using silver dollars to 

 pay for tags could not be evaluated, but the 



men and reminded them of the program 

 whenever they encountered a silver dollar 

 during business transactions. 



Full-twist Versus Half-twist Knots for 

 Tags 



The experiment to test the full-twist versus 

 the half-twist knot was made during part 

 of the cruise (June 15-28) while two tagging 

 crews were aboard. One crew tied pin knots 

 with the half-twist exclusively, while the 

 other used the full-twist. The numbers of 

 fish tagged using each knot during this period 

 and the number of recoveries are given in 

 table Z. A higher recovery from salmontagged 

 with the half-twist knot was obtained (6 per- 

 cent versus 4.6 percent). With pooled data 

 we used a chi-square analysis to test the 

 hypothesis of equal recovery for both types 

 of knots. No significant difference be- 

 tween the recovery rates of salmon tagged 

 with the two types of knots csm be demon- 

 strated: 



(X^ = 1.85, d.f. 1, P = 0,18). 



