BRISTOL BAY AND ALASKA PENINSULA SALMON STATISTICS ' 67 



is the possibility that cyclic fluctuations in the abundance of fish might follow cyclic 

 changes in environmental conditions, such as those that have been shown to accom- 

 pany the periodicity of sim spots;* but it does not seem lil^ely that such a factor 

 could cause such cycles as we observe in the salmon. We have also the remarkable 

 four-year cycles of the Frazer River sockeyes, which existed for an unknown number 

 of years before the white man came and recorded the phenomenon. In this case, 

 as is well known, the tremendous runs that came every fourth year consisted of two 

 races (or groups of races), one spawning in the lakes tributary to the lower course of 

 the Frazer and the other in the higher lakes above the Frazer River Canyon. The 

 first race entered the river yearly, but it was only every fourth year that the second 

 and much more important race entered the river. Tlie latter was a race in which 

 4-year fish (especially among the females) predominated to a remarkable extent.' 

 The last "big" year was in 1913, and in that year a slide in the Frazer River Canyon 

 prevented the ascent of the fish to the upper spawning grounds, the race died out, 

 and the four-year cycle became virtually obliterated. One can hardly doubt that 

 originally the runs of salmon in the Frazer were "big" every year; that every year 

 saw the upper spawning grounds as well covered with spawning fish as they were in 

 the "big" years that we liave known. At some more or less remote lu'ehistoric time, 

 however, a slide probably blocked the river for a period of three years and destroyed 

 the race that spawned in the upper lakes. In the fourth year the obstruction was 

 removed and the fisli were able to proceed as usual to the spawning grounds. On 

 account of the great predominance of 4-year fish in this race this one year was 

 perpetuated, perhaps for centuries, until the disaster of 1913. In this case the most 

 remarkable cycle known developed under natural conditions, quite unaffected by 

 exploitation, but we have a sufficient understanding of the circumstances so that an 

 adequate explanation can be given. Under ordinary circumstances it seems probable 

 that marked cj^cles occur most commonly under the conditions resulting from 

 exploitation. 



Cycles may become established by the occurrence of various unusual condi- 

 tions, such as an expeciaUy large or small spawning escapement or the effect of 

 environmental conditions that make for a high or low rate of mortality during the 

 life of a brood. Such conditions may be expected to occur only occasionally and at 

 irregular intervals, and the effect will tend to be perpetuated more or less strongly in 

 future generations by the dominance of certain age groups in the race in question. 

 If a single age group is dominant the effect may last indefinitely, but if two or more 

 age groups occur in fairly large percentages the effect will be spread out gradually 

 and the cycles will lose their sharpness and become obscured, or they may be de- 

 stroyed entirely or modified by the incidence of another set of unusual conditions, 

 whicli in turn may give rise to an entirely different cycle. Overfishing, especially 

 at critical times, may be an important determinant of such cycles, although undoubt- 

 edly they are frequently caused by natural conditions about whicli we know very 

 little at present. 



' Climatic Cycles and Tree Growth. By A. E. Douglass. Carnegie Institution of Wasliington, Publication No. 289. 

 Washington. 



' Contributions to the Life History of the Sockeye Salmon. Nos. 1 to 9. By Charles H. Gilbert. Reports of the Commissioner 

 of Fisheries for British Colombia, 1912 to 1923. 



