FISHERY BULLETIN: VOL. 69. NO. 4 



In an effort to reverse this downward trend 

 by providing for larger escapements, fishing ef- 

 fort was reduced by restrictions on fishing time, 

 gear, and location. The effect of these measures 

 can be gauged from three principal sources of 

 information: (1) A counting weir was oper- 

 ated in the Wood River of the Nushagak District, 

 the principal trunk stream, during the years 

 1908-1919. (2) Biological studies were con- 

 ducted in subsequent years that provided data 

 on the age, length, and size composition of the 

 catch and in jiart of the escapement. (3) The 

 salmon canning iiidusti'y itself has kept meticu- 

 lous records on daily catches, number of fishing 

 units, and type of gear operated. 



The various sources of data indicated above 

 were utilized to reconstruct the levels of escape- 

 ments in the Nushagak District during the last 

 50 years in an effort to determine whether the 

 magnitude of the yearly escapements is corre- 

 lated with the declining salmon production in 

 Nushagak Bay. If this were not the case, the 

 fishery may have changed the age and size com- 

 position of the stock or the distribution of the 

 various stocks in time and space. These factors 

 will be examined in a search for a logical expla- 

 nation of the decline of the Nushagak fishery. 



NUSHAGAK BAY AND WATERSHED 



Nushagak Bay includes the waters between a 

 line drawn from Nichol's Spit to Etolin Point 

 and the confluence of the Wood and Nushagak 

 Rivers (Figure 1). These streams serve as the 

 trunk streams of the Wood River lakes and the 

 Tikchik lakes, respectively. Two other trunk 

 streams drain into Nushagak Bay, namely, the 

 Snake River and the Igushik River. The entire 

 watershed comjn-ises a drainage basin of 

 10,207 km-. The morphometric parameters of 

 some of the more important salmon-producing 

 lakes are given by Gadau (1966). 



Although sockeye salmon occur in more north- 

 erly latitudes, the Nushagak River system rep- 

 resents the northern boundaries of large sockeye 

 salmon runs. The reason may be the absence 

 of large lakes in more northern stream systems, 

 which would i)rovide sullicient nursery grounds. 

 Thus, in the Tikchik system there are six lakes 



with five accessible to the salmon, but only the 

 three lower ones, indicated on Figure 1, are im- 

 portant for sockeye salmon production. 



NUSHAGAK SOCKEYE CATCHES, 

 1884-1966 



The commercial fishery for sockeye salmon in 

 Bristol Bay began in Nushagak Bay in 1884 after 

 the schooner Neptune made an exploratory salt- 

 ing expedition (Moser, 1902). Prior to that 

 time, some salting, from 800 to 1200 barrels each 

 year, was done by fishermen operating a simple 

 trap in the Wood River. 



The most recent account of catch data was pub- 

 lished by Kasahara (1963). His figures differ 

 in some years from those given in Tables 1 and 2 

 of this paper, compiled in part from original 

 sources, but the discrepancies are mostly minor 

 in nature, and they do not change the overall 

 picture in catch level and trend. Derivation of 

 the Nushagak catch figures used in this report 

 is given in the footnotes and comments to the 

 mentioned tables. 



When the Nushagak catches are plotted, they 

 exhibit strong annual variations, as in most sock- 

 eye salmon runs (Figure 2). A small part of 

 the variability can be explained by differences in 

 fishing effort, which reflected economic condi- 

 tions or inaccurate predictions by the cannery 

 superintendents as to the actual size of the run. 

 Viewed over longer time periods, however, there 

 can be no doubt that the annual catches reflect 

 changes in stock strength. This conclusion is 

 amply brought out by the construction of a trend 

 line by a moving average of 5's because of 5-year 

 cycle. 



Three distinct periods are discernible. The 

 first period spans the years 1900-1918, the sec- 

 ond one covers the years 1921-1945, and the last 

 period includes the years 1946-1966. The aver- 

 age annual catches during these jieriods were 

 5,134,156; 2,888,726; and 1,183,485 salmon, re- 

 spectivel.y.'' 



Transition from one level to the next took 



' If the estimated forpipfn catches made since 1956 

 were included with the domestic catches for the third 

 period, the average annual catch would be raised about 

 25%. 



748 



