48 



THE MIGRATION AND CONSERVATION OF SALMON 



behavior and on account of their great im- 

 portance in practical conservation of the 

 salmon fisheries. 



There have been two opposing theories: 

 First that the fish, on entering the ocean 

 remain throughout their lives at sea close 

 to the mouth of their home stream — perhaps 

 spreading f anwise but always within the in- 

 fluence of the home-stream water. This was 

 a theory at one time advanced by Dr. David 

 Starr Jordan. The second theory states, in 

 effect, that the fish range widely, far beyond 

 any conceivable influence of the home stream 

 to which, however, they predominantly re- 

 turn. 



It was not until an extensive series of ex- 

 periments involving the tagging of Pacific 

 salmon was started in 1922 that satisfactory 

 data began to accumulate bearing on this 

 question. In these experiments fish were 

 taken with as little injury as possible, usu- 

 ally from traps of the commercial fishery, 

 tagged with standard, serially numbered 

 stock tags and released. On recapture the 

 fishermen were asked to report the tag to 

 the fishery officials with information as to 

 where and when taken. In 1922 and 1923 

 the late Dr. C. H. Gilbert and myself tagged 

 14,000 salmon, chiefly reds (0. nerka) along 

 the western portion of the Alaska Peninsula 

 from the Shumagin Islands in the south to 

 the region of Port Moller on the north. 



When the experiments were begun it was 

 supposed that the fish taken in this region 

 were bound for local streams, although no 

 important local spawning grounds were 

 known. The results, however, showed con- 

 clusively that the red salmon were a part 

 of the important runs that spawned in the 

 rivers of the Bristol Bay region over 300 

 miles to the north and east of the narrow, 

 shallow passage between the peninsula and 

 the first of the Aleutian Islands and 400 

 miles from Unimak Pass between the first 

 and second of the islands and the first broad, 

 open passage from the north Pacific to the 

 Bering Sea. Later statistical analysis 

 showed a high correlation between the 

 catches of red salmon at Ikatan and in Bris- 

 tol Bay for the years 1913 to 1925 inclusive. 

 The coefficient of correlation (r) was + 0.79 



— a value that would not occur as a result 

 of "sampling error" alone once in a hun- 

 dred times. 



Incidentally it is this run of red salmon 

 that the Japanese have begun to exploit by 

 gill-net fishing in the open sea along the 

 route of migration between the passes of the 

 Aleutian Islands and Bristol Bay. The im- 

 portance of a knowledge of the movements 

 of fish in the ocean is well shown in this case 

 in which the results of the tagging experi- 

 ments of 1922 and 1923 showed clearly the 

 danger to the fisheries of Bristol Bay and 

 the Ikatan-Shumagin Island districts of an 

 additional and unrestricted drain on the 

 natural resources such as would most cer- 

 tainly result from the development by the 

 Japanese of a pelagic salmon fishery in this 

 district. . 



Another series of tagging experiments has 

 shown that many of the Chinook salmon 

 taken by the troll fishery off the coasts of 

 southeastern Alaska and British Columbia 

 eventually enter the Columbia River to 

 spawn. More of the tagged fish in these ex- 

 periments were taken in the Columbia River 

 than in all other streams together and the 

 Columbia is from 300 to 800 miles southeast 

 of the points of tagging. 



The question arises as to whether we can 

 assume that the fish that travel these long 

 distances in the ocean are actually return- 

 ing to their home streams from the oceanic 

 feeding grounds. I believe that this can be 

 answered in the affirmative on the basis of 

 the following argument : 



(1) The return of the adult fish, predomi- 

 nantly, to their home streams may be ac- 

 cepted as proved beyond any reasonable 

 doubt. 



(2) It is also known that large numbers 

 of Chinook salmon that eventually enter the 

 Columbia River and large numbers of red 

 salmon that eventually enter the streams of 

 Bristol Bay are to be found at sea great dis- 

 tances from the mouths of the streams that 

 they enter. 



(3) The only simple theory that can pos- 

 sibly reconcile these two facts is that the fish 

 of the Columbia and of Bristol Bay do re- 

 turn to their home streams even when they 



