224 
BULLETIN OF THE BUREAU OF FISHERIES 
were made, but chums and pinks were taken in fair quantities. A trap operated at 
the mouth of the bay probably accounts for the better catch in 1920 and in subsequent 
years. During the earlier years of its operation, the trap was driven directly in the 
entrance of the bay, but in later years was moved to a point north of the entrance, and 
doubtless has taken salmon which were not strictly Sawmill Bay fish. Accepting the 
data as given and disregarding the years of 1921 to 1923 when the fishing effort was 
reduced, it is found that the catch of chums has declined markedly and gives rather 
positive evidence of a depleted fishery at Sawmill Bay. The pink-salmon catch 
fluctuated widely between 1920 and 1927, reaching the lowest level in 1924, a year 
that shows large production in other localities of the sound. It improved, however, in 
the next three years — -a fact which indicates that the early apparent reduction in catch 
was not due to depletion but either to faulty data or variations in fishing intensity of 
which there is no record. 
Valdez Arm . — Before 1923, comparatively few salmon were taken directly in 
Valdez Arm. Fishing was confined largely to the smaller bays. However, as traps 
came into general use, locations were established in the more open waters and profit- 
ably operated. Due to this change in the character of fishing and the probable 
incorrect allocation of catches, fairly large numbers of pinks were reported from the 
arm in 1924 and 1926, while the catch in 1925 was a complete failure. The trend of 
the pink-salmon catches in Valdez Arm proper seems to be downward, especially in 
the even years when the runs were universally heavy. On the other hand, chums 
were taken in larger numbers than ever before, and, though the catches were small by 
comparison, they show a steady upward trend, except for the poor catch recorded for 
1925. The best catch of reds was made in 1920, the poorest in 1927. Kings are 
rarely taken in any number, while the catch of cohos was extremely variable. 
Port Valdez . — The upper part of Valdez Arm, inside Valdez Narrows, is Port 
Valdez. At its head are several streams, all fed at least in part by glaciers. Robe 
River, the outlet of Robe Lake, carries less glacier water than the others and may be 
considered as the only tributary of the port that supports a run of red salmon. 
Although reds have been in general more abundant than any other species, the catch 
has declined quite steadily from 1917 to 1927. In 1924, the largest catch of pink and 
chum salmon ever made in Port Valdez was reported, due perhaps to the inclusion of 
the catch of the trap at Potato Point. This record year was followed by three years 
of poor catches. In general it is apparent that the catch of all species in Port Valdez 
was rapidly declining during the years just preceding 1927 — the last year considered 
in this report. 
Table 6 shows in graphic form the catch of the salmon fisheries of the Valdez Arm 
district. The curves for both cohos and chums are very similar, showing good catches 
during the early and late years with a period of poor catches between. So far as these 
data indicate, the productivity of the coho and chum fisheries of this district may be 
viewed with uncertainty in the next few years. The red-salmon fishery, centered 
mainly in Port Valdez, is undoubtedly failing. The supply of pinks alone appears to 
be unaffected, the catches becoming better in the odd years as well as in the even. 
King salmon do not constitute an important fishery, the largest catch being only 451 
in 1923. 
