the gill-netters is 26 -feet long, gasoline power- 

 ed, and well known for its seaworthiness . 

 Usually gill-net fishing occurs at night after 

 which the fishermen make their deliveries to 

 the various receiving stations in the morning. 

 In spring the fishing may take place in daylight 

 when the river is turbid. 



Certain seasonal restrictions have been 

 imposed on commercial fishing in the Columbia 

 River to obtain larger escapements for main- 

 taining future runs of salmon. In 1877 weekend 

 closures were first established. By 1909, closed 

 seasons existed from March 1 to May 1, from 

 August 25 to September 10, plus a 24-hour 

 closure on weekends during spring and summer. 

 During the years 1943 to 1947, an additional 

 closed period of May 20 to June 10 was estab- 

 lished. In 1948 and 1949 the latter period was 

 changed to June 15 - July 6 below Bonneville 

 Dam, with a similar period established above 

 Bonnevile Dam so staggered as to protect the 

 fish, safeguarded by the closure below. Clo- 

 sures of similar length but slightly different 

 timing have been continued in the May -June 

 period. In addition, a July 15 to 29 closed sea- 

 son was established below the dam with a 

 staggered period above, beginning in 1952. 



CATCH -PER -UNIT-OF -EFFORT 



Records of individual fisherman's catches, 

 copied from the ledgers of a large fish packing 

 company on the Columbia River, constitute the 

 basic data from which the catch -per-unit -of - 

 effort was calculated for each year. Only catch 

 records from the season between May 1 and 

 October 1 which normally accounts for 95 per- 

 cent of the yearly catch of chinook salmon were 

 used. These five months were divided into 22 

 weekly periods: weeks 1-4 involved fishing be- 

 tween May 1 to May 28; weeks 5-13 fishing be- 

 tween May 29 to July 30; and weeks 14 to 22, 

 fishing between July 31 and October 1 . 



The catch records of only full-time 

 fishermen were considered in the analysis, with 

 the belief that they would sample the chinook 

 salmon population better than part-time fisher- 

 men. The fishermen were divided into two 

 groups. One group must have made one delivery 

 each month in the period between May 1 and 

 August 26. The other fished between September 



10 and October 1 . A total of thirty or more in- 

 dividual fisherman' s records was the attempted 

 goal, but in a few instances that many were not 

 available. 



The analysis is based on the assumption 

 that a man's ability to sample the salmon popula- 

 tion by fishing remained approximately the same 

 for two comparable years . An ever changing 

 list of fishermen's records was found necessary 

 because fishermen were constantly dropping out 

 of and entering the fishery. Use of such records 

 was made possible by a chain -link system as 

 used by Silliman (1950) in which catch records 

 were paired by successive years, permitting a 

 continuous analysis. Of interest are the records 

 of two fishermen who were found to have fished 

 for 25 years. Their catch records are plotted 

 in figure 1 and resemble each other so closely 

 as to indicate their fluctuations may stem from 

 a common cause; the most likely such cause is 

 fluctuation in the size of runs of salmon. 



In comparing similar weeks of successive 

 years, a non-fishing week was not weighted 

 against a week in which a fisherman was engaged 

 in fishing. However, the fact that a fisherman 

 may have fished unsuccessfully, must be taken 

 into consideration. In an attempt to decide 

 whether a fisherman had fished for chinooks un- 

 successfully or whether he had simply not fished, 

 certain assumptions were made. First, it was 

 decided that a fisherman would not continue to 

 fish for three successive weeks unsuccessfully, 

 but instead would quit until fishing improved or 

 use a smaller meshed net for catching other 

 species. Therefore, the rule adopted by Silliman 

 (1950) was used: A fisherman is considered to 

 have fished during a given week, even though he 

 made no deliveries, if he made one or more de - 

 liveries during the preceding or succeeding week . 

 In addition to this, it was felt that unless at least 

 10 percent of all the fishermen had made only 

 one delivery during the week in question, it was 

 not very likely that anybody had been unsuccess- 

 ful. The following was therefore added to the 

 rule above: ( and ) at least 10 percent or more of 

 the fishermen making deliveries during the given 

 week made only one delivery . 



This rule was used with the full realiza- 

 tion that some error would be introduced, but 

 it did supply an approximation which would suffice 

 for an analysis based on averages. 



