4S U. S. BUREAU OF FISHERIES 
from the annuli on the scales, in order to establish the relative 
abundance of the various age groups represented in the catch in 
relation to comparable data obtained in previous years. 
In 1937, the first year for which adequate data are available for 
the Kodiak region, the 1931 year-class constituted 70 percent of the 
total number of fish taken. In 1938, with the entrance of the 1935 
year-class into the catch, the year-class of 1931 declined to 49 percent 
of the total yield. Age readings of approximately 2,500 individuals 
during the 1939 season revealed that the catch was still composed of 
20 percent of the 1931 year-class even after 5 years of fishing had 
decimated its numbers, and also confirmed the previous findings that 
the spawnings of 1932, 1933, and 1934 failed to contribute any appre- 
ciable increment to the stecks. Since the spawnings of these years 
added so little to the populations, and since the 1931 year-class 
will have been nearly exhausted by 1940, the catch during that season 
will become dependent almost entirely on fish spawned in 1935 and 
1936. This condition does not bode well for the continuance of a 
high abundance level. In view of the increased fishing intensity, 
the declining return per unit of gear, and the decreased spread of 
the age composition of the stock, the need for close supervision of 
the herring fishery cannot be disregarded. 
Prince William Sound arca—TVhe catch in the Prince William 
Sound area during the 1939 season was 423,000 barrels. This total 
compares favorably with the average catch of 317,000 barrels during 
the preceding 5 years, and taken alone would indicate a healthy 
condition for this fishery. Considered, however, in relation to the 
amount of fishing effort required, and the declining return per unit 
of gear, together with the change in age composition, it 1s apparent 
that this area, too, will require careful supervision to avoid over- 
exploitation. 
Reduction capacity in this area has increased from 7 tons of raw 
fish per hour in 1934, to 66 tons per hour in 1939 and this increase 
in plant capacity has been followed by an increase in fishing effort. 
The number of vessels employed in 1934 was 7, compared to 30 in 
1939. Because of the larger vessels employed, the increase in actual 
capacity was in the ratio of 6 to 1. In the face of this increase in 
the fishing effort the catch per unit of gear, based on the catch per 
ton-day of fishing effort, as in the Kodiak area, has shown a pro- 
gressive decline from 8.1 barrels in 1934 to 3.1 barrels in 1939. 
Reduction operations in the Prince William Sound area were 
not intensive until after 1934, providing an opportunity for the 
herring stocks to build to a high abundance level. This abundance 
also was bolstered by the presence of a dominant 1931 year-class, 
as occurred in the Kodiak area. The 1937 and 1938 samples of the 
‘atch in this area were taken in the fall months only—August 25 
through September. In 1939, for the first time, age-composition 
data are available throughout the season, and age determinations 
from scales of approximately 4,600 fish were made. However, since 
there is a change in the age composition of the catch during the 
season in this area, with the older fish appearing in greater relative 
abundance during the fall months, comparisons of the age composi- 
tion of the catch with that of previous seasons has been made for 
the fall portion of the season only. This analysis shows that the 
