FISHERY BULLETIN: VOL. 71. NO. 3 



filament nets are also listed in Table 6. Loss 

 rates for fish observed up to 10 h were 86% 

 in the large multifilament nets as opposed to 

 41% for monofilament nets of the same mesh 

 size. 



1968. — Studies on the loss of salmon from 

 gill nets in 1968 were concentrated on matur- 

 ing fish in May and June, south of Unalaska 

 Island (long. 164 °W). The procedures used 

 to estimate loss of salmon were the same as used 

 in 1966 and 1967. The net string consisted 

 of five mesh sizes (64-, 83-, 98-, 114-, and 

 133-mm mesh) of multifilament net, all of 

 which were observed for dropouts. Two 

 shackles of monofilament nets (133-mm mesh) 

 were also observed for comparison with multi- 

 filament nets. 



Overall loss rates were zero up to 1 h, 4.7% 

 (±6.3) up to 2 h, and 20.4% (±7.8) for up to 

 10 h (Table 7). Rates of loss for periods up to 

 10 h were nearly 18% ( ± 9.9) for multifilament 

 nets and 22% (± 14.5) for monofilament nets; 

 the difference was not significant. This similar 

 loss rate for monofilament nets as opposed to 

 multifilament nets contrasts to the loss rate of 

 the two types of gear in 1967 on immature fish 

 (41% loss for monofilament nets and 86% for 

 multifilament nets). 



Compared to the 1966 and 1967 studies of 

 losses of primarily immature salmon, the drop- 

 out rates of maturing fish in 1968 were much 

 smaller. Possibly this reduction in dropout 

 rate in 1968 was due to unusually calm seas 

 and partly to the fact that the experiments were 

 on maturing fish taken in large mesh nets as 

 opposed to immature salmon in large and small 

 mesh nets in 1966 and 1967. 



1969. — Studies of the loss of salmon from 

 gill nets in 1969 were conducted in April and 

 May on maturing salmon and in July on im- 

 mature salmon. 



Because of inclement weather in April and 

 May, we were able to patrol the net string and 

 mark the position of salmon in the nets in only 

 four different sets. Eight nets were observed: 

 one each of 114- and 133-mm monofilament 

 nets and two each of 98-, 114-, and 133-mm 

 multifilament nets (in one set, a single 64-mm 



multifilament net was also observed). Only 13 

 salmon were marked in the four sets; no loss 

 was observed up to 1, 2, or 5 h (Table 8). A 

 loss of two fish (15.4% ) was recorded for periods 

 up to 10 h. 



In July, experiments on immature salmon 

 were conducted south of Adak Island (long. 

 176 °W) during three sets. Eight nets were 

 patrolled: two 114-mm monofilament nets and 

 three each of 114- and 133-mm multifilament 

 nets. Comparison of loss rates was made 

 between monofilament and multifilament nets 

 (Table 9). 



The loss rate in the monofilament nets was 

 over twice that of the multifilament nets for the 

 relatively small sample sizes (46% vs. 22% for 

 periods up to 10 h). The combined loss was 

 35% (±14.8) up to 10 h; it was 15% (±13.6) 

 up to 1 h and 25% (±21.4) up to 2 h. The 

 total loss rate in the large nets was smaller 

 than that observed in 1966 and 1967. 



SUMMARY.— The total number of salmon 

 marked and observed by direct methods over 

 varying periods of time, by mesh size and year, 

 are presented in Table 10. In 4 yr of experi- 

 ments. 534 fish were marked. Overall loss 

 rates for periods up to 11 h were about 41% 

 (±4.0). Loss rates were about 6% (±3.4) up to 

 1 h and 14% (±4.4) up to 2V2 h. The loss rate 

 of about 52% ( ± 16.2) for periods up to 5 h 

 was greater than the rates for up to 11 h. How- 

 ever, this loss rate for up to 5 h was based on 

 only 21 fish marked, whereas the loss rate for 

 up to 11 h was based on 524 fish marked. The 

 data indicated that losses of salmon continued 

 with time in the net. 



Losses from small mesh (64- and 83-mm) 

 multifilament nets, which capture mainly the 

 small age .1 immature salmon, were similar 

 to losses from the large mesh multifilament 

 nets (45% vs. 41%). Loss rates from the small 

 mesh nets were similar to those of large mesh 

 nets for periods up to 1 h (5 and 6% , respec- 

 tively) and were greater than for the large mesh 

 multifilament nets up to 2V2 h (21 and 7%, 

 respectively). 



Comparisons of the loss rates of large mesh 

 multifilament nets as opposed to large mesh 

 monofilament nets indicated that, for periods 



858 



