36 



BULLETIN OF THE BUEEATJ OF FISHERIES 



The secondary object is to conduct these investigations over a sufficient period 

 of time that we may determine to what extent results will vary in different years, 

 due to natural influences, favorable or unfavorable, which shift in effectiveness from 

 year to year. This will constitute a quantitative study of fluctuations in abund- 

 ance due to natural causes. 



The varying productivity from year to year in a given salmon fishery is due to 

 the interaction of the two factors that we here propose as objects of our investiga- 

 tion. An unusually small run of salmon in any river in a given year may be due to 

 insufficient spawning in the corresponding year of the preceding cycle, or it may be 

 due to unusually unfavorable conditions that have increased mortality among the 

 members of the brood beyond the ordinary, or it may be due to a combination of 

 the two factors. We have at present no means of making a direct study of fluctu- 

 ations in abundance due to natural causes. Many of these are unknown to us, and 

 with regard to the rest their effect can in no instance be quantitatively determined; 

 but when we shall have eliminated from our problem variations in yield due to dif- 

 fering intensities of spawning, as we hope to do in the prosecution of these investi- 

 gations, all fluctuations that can not be accounted for by this primary cause must 

 be referred to the total effect of shifting natural conditions. The relative impor- 

 tance of the two can then be determined, and we can ascertain how high a degree of 

 correlation exists between the intensity of spawning and the productivity of the 

 fishery. If normally a high degree of correlation exists, the problem of an effective 

 administration of the salmon fisheries will be greatly simplified. To produce con- 

 sistently large runs, it will be necessary only to provide spawning reserves of suffi- 

 cient size, which it will be within our ability to determine. If, on the other hand, 

 it shall prove that fluctuations due to causes beyond our control are of preponderat- 

 ing influence and largely mask the effects of variable spawning, a successful admin- 

 istration of the fisheries will be rendered more difficult and uncertain. A larger 

 safety factor would then be called for. Larger spawning reserves would have to be 

 provided in order that their more numerous progeny can neutralize in some measure 

 the disastrous effects of unfavorable years. It is not necessary to point out that 

 predictions of the size of future runs, based on the size of spawning colonies year by 

 year, will have value in such degree only as we establish correlation between numbers 

 of spawning fish and the number of their progeny that attain maturity and enter 

 the runs. Such predictions will be unreliable in such degree as fluctuations due to 

 natural causes interfere with the expected results. 



Our primary line of investigation calls first for an accurate determination each 

 year of the total returns of red salmon to the Karluk River and the division of these 

 between commercial catch and spawning escapement. The commercial catch is 

 obtained directly from the companies that operate in the Karluk district, and the 

 spawning escapement is obtained at a counting weir operated by the Bureau of 

 Fisheries near the mouth of the river. Returns from these two sources are now 

 available for each year from 1921 to 1926, inclusive. 



Having, then, the complete census of each annual run, as well as the number 

 of salmon that each year comprise the spawning reserve, it remains only to recog- 

 nize and enumerate the progeny of the various spawnings as they return at maturity, 

 in order to establish the ratio of increase for which we are seeking. This could be 



