934 
FOREST AND STREAM. 
[Dec. 30, 1911. 
Columbia River Salmon Migration 
By CHARLES W. GREENE, Ph.D.* 
T liE life history of the anadromous fishes is 
one of the most interesting subjects in 
biology. The detail of facts surrounding 
the migration of the young from the fresh water 
to the sea, and the migration of the adults to 
fresh water for spawning purposes, are indeed 
little enough known of themselves. How much 
more shrouded in obscurity, therefore, must be 
the causes operating during these migrations. 
The United States Bureau of Fisheries has never 
ceased in its efforts to untangle this thread of 
piscatorial history. 
In the instance of the Pacific coast salmon 
of the genus Oncorhynchus, thanks to the labors 
of Evermann, Gilbert, Meek, Rutter, Chamber- 
lain and others, the following general facts are 
now established within a reasonable degree of 
certainty: 
1. The young of the species of Oncorhynchus, 
which have been hatched in the freshwater 
streams, migrate to the sea, where they can 
secure an abundance of food during their de¬ 
velopmental period. Evermann in 1894 and 1895 
observed many young O. tschawytscha and O. 
nerka in the Salmon River headwaters in Idaho. 
He says: “We are not yet able to say just 
when the young salmon leave the waters where 
they were hatched and begin their journey to 
the sea, but it undoubtedly occurs between Sep¬ 
tember of the first and July of the second year 
following that in which they were spawned.” 
Later Rutter followed the downward migration 
of young salmon in the Sacramento River, Cali¬ 
fornia. He found that young salmon fry “begin 
their down-stream migration as soon as they are 
able to swim.” They reach the estuary in large 
numbers in from ninety to one hundred days or 
more. He found also that many young salmon 
“summer residents’’ remained in the headwaters 
of the Sacramento until the first winter rains, 
when they all went out. 
2. The salmon feed in the ocean for a period 
of years. For the chinook salmon this period 
is believed to be from three to five years, though 
the evidence is not entirely conclusive. The 
feeding period continues until maturity. 
3. At the end of the feeding and maturing 
period the salmon migrate up the Pacific coast 
rivers to spawning grounds, which are sometimes 
only a few miles from the sea and scarcely be¬ 
yond brackish water, but often for hundreds of 
miles, apparently always into cold fresh waters 
of the streams fed by springs, lakes and moun¬ 
tain snow fields. 
4. It has long been known in a general way 
that the migration of O. tschawytscha to the 
spawning grounds is made wholly without food. 
5. The most striking and least expected cli¬ 
max to this interesting life cycle was discovered 
in 1894 by Evermann for the species O. tschawyt¬ 
scha and O. nerka, namely, the fact that death 
invariably follows the spawning act. Evermann 
states, on page 260 of his preliminary report 
upon the 1904 expedition, that on Sept. 13 he 
counted seventy-two dead salmon in a three- 
mile stretch of Salmon River and a mile or more 
*Bulletin of the Bureau of Fisheries. 
of the lower portion of Alturas Creek in Idaho. 
Only one live salmon was noted on this date. 
He quotes numerous observations and conclus¬ 
ions of local men of the region tending to con¬ 
firm the deduction expressed on page 153 of his 
final report as follows: “The chinook salmon 
which come to these waters die after spawning.” 
This brief salmon history is repeated here for 
the reason that it is the most effective way of 
presenting the setting for the problems that ap¬ 
peal to the physiologist. Of these problems I 
have in a previous paper attacked the question 
of the acclimatization of the chinook salmon to 
fresh water after its life in the sea. That study 
was based on an examination of the blood and 
other body fluids. The special interest attaches 
to the osmotic changes during the passage of 
the fish through the various degrees of brackish 
water in the journey from the salt water of the 
sea to the fresh water of the rivers. The fur¬ 
ther osmot:c change during the run up the river 
was also studied. 
The changes in the blood and body fluids are 
relatively slight and are carried on very slow 
and gradually. The osmotic changes in the body 
fluids give little or no intimation of the length 
of time consumed by the fish in the transition 
from salt to fresh water. Neither do the osmotic 
changes give any measure of the duration of the 
sojourn in fresh water. In order to arrive at 
any adequate explanation of the profound 
changes in the tissues and organs during the 
migration, it becomes almost a necessity that the 
rapidity of change in the environment and the 
total duration of the period be determined. The 
time element in this change is the most import¬ 
ant factor, yet an almost wholly unknown one. 
The present paper gives the results of a pre¬ 
liminary experiment designed to secure more 
tangible evidence as to the time element in the 
migration, especially on the Columbia River. 
I he question can be better understood when 
analyzed into the following points or questions: 
1. How long do salmon remain in brackish 
water ? Or, stated more fully, how rapidly do 
salmon pass from salt water through the degrees 
of brackish water at the mouths of the rivers? 
2. What evidence is there that salmon swim 
back and forth with the ebb and flood during the 
migration through brackish water? 
3. When once quite within the fresh water 
of the rivers, how rapidly and how continuously 
do salmon travel on their course up the rivers 
to the spawning grounds? 
4. What evidences do salmon give of special 
responses to unusual conditions, such as obstruc¬ 
tion to their course, individual injury, etc.? 
PRINCIPLE AND METHOD OF EXPERIMENT. 
This experiment is based on the principle that 
an understanding of the details of the migration 
phenomena can only be had by a study of the 
movements of individual fishes. The informa¬ 
tion derived from the movements of large schools 
of fishes, while often of extreme value as' cor¬ 
roborative evidence, can never be taken as con¬ 
clusive evidence of the movements of individ¬ 
uals. Even if it were safe to assume that the 
movements of a given school of salmon repre¬ 
sent the average of the movements of the com¬ 
ponent individuals, yet it is quite impossible to 
identify certainly any given school at different 
points along the river. 
In order to subject the above questions to a 
preliminary test, 1 arranged a salmon marking 
experiment on the lower Columbia River. The 
experiment was accessory to a physiological in¬ 
vestigation under my immediate direction during 
the summer of 1908. Fifty-nine fish were 
marked with individual tags and liberated in the 
Columbia River at the head of Sand Island, 
which is just within the mouth of the Colum¬ 
bia. The point at which the fish were liberated 
was about eight miles up the river above the 
Canby lighthouse on Cape Disappointment. This 
experiment was launched on Aug. 14, 1908. 
On the above date the trap contained a two- 
days’ catch. We reached the trap at about 9 
o’clock in the morning, just before extreme low 
tide, and the net was lifted soon afterward. The 
fish were run from the net into a special live 
car used by the Chinook hatchery crew to trans¬ 
port fish from the trap to the retaining grounds. 
The fish were later dipped from the car with a 
large dip net, lifted out of the net by hand, and 
quickly measured for total length. The mark¬ 
ing tag was next inserted and the fish turned 
loose into the current. It goes without saying 
that the utmost dispatch was used to prevent 
asphyxiation and care taken to avoid injury 
during the necessary handling. 
The tags used to mark the salmon in this ex¬ 
periment were made of aluminum and were ex¬ 
tremely light and very strong. The entire tag or 
button weighed 2.6 grams (one-twelfth ounce). 
The tag was made of two pieces on the general 
principle of a Yankee button. On one face was 
stamped the words “U. S. Fish.” 
When a salmon is caught up in a dip net he 
struggles vigorously to get away. One should 
use a relatively large dip net with a wide flat 
bottom (i. e., not the usual round or kettle¬ 
shaped bottom). With such a net it is very easy 
to manage a fish through the struggling stage 
so that it does no injury to itself. It is not 
necessary that scales should be lost, even in such 
loose-scaled fish as the silver salmon. 
In this experiment when a fish was caught it 
was held with the bottom of the net just deep 
enough in the water for the fish to struggle 
against- the resistance of the water. While this 
method resulted in a goodly quantity of water 
being thrown over the operator, it had the very 
desirable effect of quickly producing a tempor¬ 
ary fatigue of the salmon. As a result of this 
fatigue, the fish remained quiet for some seconds. 
The instant a fish stopped struggling it was 
lifted out of the water, seized by the tail with 
a strong grip of the hand, swung free of the 
net, and over the free arm of the operator. The 
next instant it was quickly but gently laid out 
on the measuring platform and its length read 
off. The measuring platform consisted of a 
broad board with an upright at one end. A 
meter stick was tacked to the board with its 
zero against the upright. Loose folds of burlap 
were laid over the board and over the meter 
stick for the greater portion of its length. A 
fold of the burlap was so arranged that it could 
be quickly thrown over the middle portion of 
the body of the salmon whenever desirable, i. e., 
occasionally with the largest specimens. 
When a fish was laid out on the measuring 
