LIFE-HISTORY OF THE COHO 41 
SESSIONAL PAPER No. 38a 
the creek is the cut-throat trout, with an occasional small sculpin or fresh-water bull- 
head. .The cut-throat of the same year is not hatched for some time after the coho 
appears, and those of the preceding year are large enough to look after themselves. 
The young fry, therefore, have no fish as small as themselves to attack, and hence 
insect larve, with a few fresh-water crustacea must supply the demand. It is possible 
that those earlier hatched may attack those later hatched and that both may attack 
the cut-throat fry when they come out, but by this time they must have attained greater 
size. It is possible, too, that the yearling coho attack the firy, and the cut-throat a year 
or more old may do so also, as all the Salmonide eat fish when other food is not avail- 
able, if not at other times. In this creek the cohos and the trout seem to live in har- 
mony, as both are commonly found in the same small group. : 
It is a fact that when large numbers of fry are put out in the creeks from the 
hatchery that the older ones may be seen devouring the younger ones, but in such cases 
thousands, sometimes hundreds of thousands, are put out in the one creek within com- 
paratively narrow limits so that before they become well distributed insect food must 
be at a premium. As the younger fry offer the only food for the older ones, very hun- 
gry by this time, they are devoured. If there are trout in the same stream they prop- 
ably assist in the operation. 
The statement that coho remain in the rivers for two or three years feeding on the 
trout is evidently absurd. In the first place, the coho does not live to be three years 
old, or at least there has been no evidence adduced that it does. In the second place, 
there is a similar lack of evidence that any of them remain in fresh water for two 
years. Furthermore, as the yearling coho is seldom more than 5 inches long when it 
migrates, and more often is considerably short of that, the injury done to the trout by 
it must be very much exaggerated. In reality the coho has a much stronger case against 
the trout, the steelhead, the cut-throat and dolly varden or char. These fish follow the 
coho to the spawning beds and devour so many of the eggs as soon as they are spawned 
that the possible number of coho fry is at once very much reduced. No matter how 
often the male coho turns to chase them, they follow him back, as soon as he turns, 
to gorge themselves once more. After the eggs are hatched the fry are attacked. and 
it is there that the dolly varden does the most damage. It is the general opinion of 
observers all the way from the Aleutian islands to California that the dolly varden 
does more harm to the salmon fry than any other agency, and many will go so far as 
to say that it does more harm than all the other agencies put together. Therefore, 
instead of protecting the dolly varden by a close season, it would ba very much better 
for the salmon fisheries if everything possible were done to reduce their numbers. The 
case against the other trout is not so strong, but as they remain in the fresh water for 
a much greater portion of their lives than the coho, the balance of destruction is prob- 
ably in their favour. 
The food of the coho in the sea has been indicated. Pelagic crustacea form the 
bulk of it. Schizopods predominate if the whole year is considered but, at certain 
times, larval barnacles and larval decapods form an important portion. Of the fish 
used, reference has been made to the small herring fry. The older fry and even the 
herring a year or more old are eaten later in the season. Apparently they have pre- 
ference over other fish. Salmon fry, sand launces and capelin are the only other fish 
that have to be observed. For a short period about October the 1st the capelin are taken 
in large numbers as they come inshore to spawn. 
The mature fish feed actively until they come to the mouth of the streams up 
which they go to spawn, or possibly until they enter these streams. Consequently, 
they must increase in weight almost until spawning time. 
The general rate of growth has been considered and some remarks made about the 
age of the coho. A more complete analysis of the relation of growth to age, depending 
on the examination. of seales, will now follow. The method of growth determination 
