FISHERY BULLETIN: VOL. 71, NO. 3 



gill nets (attributable to predation and disen- 

 tanglement) based on research vessel data pre- 

 sented here would apply equally to the Jap- 

 anese mothership fishery in areas where the 

 fishery is exposed to sea lions. In other areas 

 far offshore, loss of salmon attributable to 

 predators would be minimal, but losses would 

 still be expected from fish becoming disen- 

 tangled. 



Effects of Weather on Dropout Rates 



Weather is one factor which should affect 

 the dropout rate. Normally one would expect 

 greater losses of fish from gill nets during 

 strong winds and the resultant heavy seas. 



The indirect methods of estimating losses 

 from gill nets used in 1964 and 1965 allowed 

 fishing in winds up to 35 knots. The design of 

 the experiments, however, did not provide for 

 an easy comparison of dropout loss and 

 weather conditions, nor were detailed weather 

 measurements recorded. 



The direct methods of estimating loss from 

 gill nets used in 1966-69 precluded observa- 

 tions in strong winds and heavy seas. Observa- 

 tions were normally not made in winds greater 

 than 15 knots or in other than nearly calm seas. 

 At times, however, winds would increase dur- 

 ing the dropout observations, necessitating the 

 canceling of the experiment. 



Routine weather observations made aboard 

 the vessel while setting and hauling the gill 

 nets in 1966-69 included wind speed and direc- 

 tion, wave height, sea swell conditions, and 

 barometric pressure. These data were compared 

 to the percentage loss of salmon from gill nets 

 (Table 15). 



Because of the relatively small changes in 

 most of the weather variables between set and 

 haul times, no apparent relation with percent- 

 age loss of salmon was found. The weather 

 measurement exhibiting the largest variation 

 was wind speed at haul. This measurement was 

 compared with the total loss of salmon (Figure 

 6), but little correlation was found (r — 0.125). 



In a number of sets the loss rate was high 

 (up to 40% or more) when wind speed at haul 

 was 10 knots or less-. In an attempt to determine 

 whether or not these losses were due primarily 



100 [^ 

 90- 

 80 •- 



o 



< 70 



o 60 



° 50 



30 



20 



10 



10 



15 



20 25 30 35 

 WIND SPEED (KNOTS) 



40 



Figure 6. — Percentage loss of salmon in relation to wind 

 speed at haul. 



to predation as opposed to dropouts and, there- 

 fore, more independent of weather, all sets 

 less than 145 km from shore were eliminated 

 from further analysis to reduce the effects of 

 predation. Twenty dropout sets remained for 

 analysis; the correlation between percentage 

 loss of salmon and wind speed at haul remained 

 poor(r = 0.362). 



Although it seems obvious that weather con- 

 ditions (at least sea state) should influence the 

 rate of loss of salmon from gill nets, the avail- 

 able data fail to demonstrate any relation. 

 Effects of sea state must therefore remain as 

 an unknown influence on the rate of salmon 

 loss from gill nets. If, however, we assume that 

 dropout of fish increased with increased agita- 

 tion of the nets, the rates as measured (due to 

 selectivity of experimental conditions) are 

 probably low. Ishida et al. (1969) also were 

 unable to evaluate the effect of sea state on 

 dropout rate during their experiments from 

 1963 to 1967. 



Inshore Fishing 



The loss of salmon from gill nets fished on 

 maturing salmon in inshore waters was studied 

 by direct observation in Puget Sound in 

 August 1967. The same basic experimental 

 methods used on research vessels were applied 

 to a commercial salmon fishing vessel. Because 

 the commercial fishermen in Puget Sound 



866 



