FRENCH and DUNN: LOSS FROM HIGH-SEAS GILLNETTING 



Table 9. — Number of salmon marked and observed in gill nets over varying periods of 



time. Julv 1969. 



Total 



Percentage loss and 

 95°o confidence interval 



Monofilament nets ( 1 14-mm mesh): 



17 7/21 Lat.51°32'N 



Long. 176°22'W 



21 7/25 Lat.50°10'N 



Long. 176°22'W 



25 7/29 Lat.5I°35'N 



Long. 176°22'W 



Total 



Percentage loss and 

 95°'o confidence interval 



Grand total 



Percentage loss and 

 95° o confidence interval 



18 



16 



(17) 



5.9± 12.0 



8 



(9) 



11.1±23.7 



14 



(18) 



22.2 ± 11.3 



' A zero (0) indicates that the net was examined, but no marked fish were observed; a dash 

 ( — ) indicates that the net was not examined. Numbers in parentheses indicate number of possible 

 observations for the time period; difference from the number marked is because, some observations 

 were not mode in some periods. 



The results showed that for many of the 1-h 

 and 2-h time periods the 95% confidence inter- 

 val included zero indicating that the estimated 

 loss rates were not significantly greater than 

 zero. Similarly, the width of the confidence 

 intervals indicated that the estimated losses for 

 some adjacent time periods would not be con- 

 sidered to be significantly different. The point 

 estimates, however, showed an increase in loss 

 rates with time indicating that in general the 

 longer the fish were in the nets the greater like- 

 lihood there was for the fish to drop out. 



Because we could not determine species 

 readily during the process of marking the loca- 

 tion of salmon in the nets, we could not identify 

 dropouts as to species. During the 1966-69 

 experiments, sockeye and chum salmon made 

 up over 70% of the catches and thus these 

 species most likely were the species escaping 



the nets. Undoubtedly, in a salmon gill net 

 fishery, dropout rates would be applicable to 

 those species which make up the bulk of the 

 catches. 



The above summary of dropout rates for all 

 experiments includes losses due to predation as 

 well as to dropouts. The effect on predation on 

 dropout loss is examined in the next section. 



Effects of Predation on Dropout Rates 



The methods used in this study to estimate 

 the loss of salmon due to gillnetting fail to 

 differentiate between losses due to predation 

 and losses due to disentanglement. The subject 

 of losses due to predation apparently has not 

 been considered by other authors (e.g., Doi, 

 1962; Konda. 1966; Ishida et al.. 1969). 



There is evidence, however, that predators 



861 



