640 U. S. BUREAU OF FISHERIES 
Completely to ascertain the yield from one year’s spawning in the Karluk 
necessitates, therefore, the full analysis of the runs of five consecutive years, 
in each of which we ascertain the proportion of the fish of each age present in 
the run, and ther actual number. As the runs are not homogenous throughout 
the season, but vary from time to time in the proportion of fish of different age, 
it becomes necessary to take adequate samples at frequent intervals. An em- 
ployee of the bureau who is stationed on the Karluk fishing grounds takes 
therefrom a sample of scales from 100 to 120 salmon on every available day 
during the fishing season. The number thus obtained for investigation during 
the season of 1926 was in the neighborhood of 9,000. Data of length and 
weight and sex are obtained for each fish, and from the scale analysis we learn 
the length of early residence in the lake and the age at maturity. Complete 
analyses thus have been secured of the runs of 1924, 1925, and 1926; and we 
have learned from these that while, as has been stated, the total range of age 
at maturity is from 3 years to 7, the great majority of the Karluk red salmon 
mature at the age of 5. The average of 5-year fish for the three years mentioned 
is approximately 80 per cent. Now, the 5-year fish from the 1921 spawning 
matured in 1926, so, although we do not as yet have complete returns from that 
spawning, we are now in possession of figures that make possible a fairly 
reliable estimate. This is of special interest as furnishing the first informa- 
tion we have ever had concerning the returns from a known number of salmon 
spawning under natural conditions. 
The spawning reserve of the Karluk in 1921 consisted approximately of 
1,500,000 salmon. The run in 1926, over 80 per cent of which was derived from 
the 1921 spawning, consisted of over 4,500,000 fish. The final figures will not 
vary widely from these and will thus show an increase of approximately 3 to 1. 
The significance of these figures we are not now in position to estimate. How 
wide the oscillation in productiveness is from year to year, due to natural 
fluctuations, we do not know. Neither can we tell whether the ratio of 38 to 1 
lies nearer the maximum yield of the more favorable years or nearer the 
minimum, or whether it most nearly approximates the mean. Some light on 
this problem may be expected in 1927 and in succeeding years. As it happens, 
the spawning escapements to the Karluk varied widely during the first few 
years of the operation of the weir. With 1,500,000 in 1921, less than 500,000 in 
1922, over 670,000 in 1923, and over 750,000 in 1924, we shall have presented to 
us a series of runs representing returns from spawning colonies of widely 
different sizes. If the 3 to 1 ratio observed as between 1926 and 1921 is sub- 
stantially reaffirmed as between 1927 and 1922, the 1927 Karluk run will be of 
such small size as to yield little if any surplus to the fishery there located. 
These disquieting facts have been communicated by the bureau by way of 
warning to the companies concerned, coupled with the qualifying statement that 
in the present early stage of our investigations our experience is too limited 
to serve as the basis for confident prediction. If the ratio of return. in 1927 
should prove much higher than in 1926, a satisfactory run might result, despite 
the very limited spawning of 1922. The result will be observed with the 
greatest interest. Whatever it may be, it will add materially to our present 
knowledge. With each successive Karluk year, further additions will be made. 
In 1927 we shall have similar returns from the Chignik, to be followed by an 
unbroken series also from that river. In the near future, therefore, we should 
have ascertained how close a correlation exists between the size of a spawning 
colony and the extent of the run that it produces. 
Mr. Serre. With reference to this 50 per cent escapement, isn’t it 
reasonable to suppose that at times the natural run, due to natural 
causes, might be much below the usual number, and by allowing a 50 
per cent escapement of this natural run you would have much less 
than the necessary amount to produce a normal run in subsequent 
years? If the normal run amounted to 5,000,000, and in an off year 
you got only 1,000,000, would not the fluctuations be continued by 
having merely a 500,000 escapement in that year, as compared with 
a normal escapement of 2,500,000 in a normal year? Would it not be 
possible, by having a fixed escapement for any certain river, to 
control to some extent the natural fluctuations? 
Doctor Giupeerrt. That point is one that occurred to us at the begin- 
ning of our Karluk experiment. It must be apparent to you that on 
