476 U. S. BUREAU OF FISHERIES 



Three trips were made to Karluk Lake to obtain limnological data 

 and to observe conditions on the spawning grounds. This investiga- 

 tion lias been conducted under the general direction of Dr. Willis H. 

 Rich, of Stanford University. 



Chignik red sahnon. — The season of 1931 proved to be very favor- 

 able for the investigation of the red-salmon runs at Chignik, Alaska, 

 which Harlan B. Holmes and assistants have been conducting for the 

 past four years. The run of mature fish, amounting to approxi- 

 mately 1,500,000, was sufficient to permit fishing by the greater part 

 of the traps throughout the season. As a result of this condition it 

 was possible to collect numerous data for age analysis of the run 

 and to study the adequacy of sampling. During the two preceding 

 seasons studies of sampling were undertaken, but a poor run in 1930 

 and the dominance of a single-age class in 1929 made these attempts 

 of little value. The run of 1931 was very favorable, not only as a 

 result of continuous fishing but also as a result of a good representa- 

 tion of at least four age classes. A preliminary study of the data 

 has shown an orderly progression of runs of the several age classes, 

 the fish of a single class starting with few individuals, gradually in- 

 creasing to a maximum of abundance, then gradually declining until 

 it finally disappears. The overlapping of two or more such inde- 

 pendent runs gives the general run a complex and constantly chang- 

 ing age composition. This complexity apparently is increased by 

 the existence of two independent races of red salmon supported by 

 the spawning grounds of the two lakes in the Chignik system. Each 

 of these two races is represented by the four age classes, making a 

 total of eight such classes to be considered in the analysis of the run. 



Comparison of the catches of traps on the two sides of Chignik 

 Bay, separated by a distance of approximately 6 miles, shows that 

 traps on one side catch a significantly greater proportion of larger 

 and older fish than those on the other. The reason for this difference 

 is not apparent unless it be that the fish of different ages tend to 

 follow opposite shores of the bay. This evidence of selective fishing 

 further increases the needs for extensive sampling. 



Observations of the young fish in fresh water were continued in 

 1931. Samples of migrants and lake residents were collected at fre- 

 quent intervals. Approximately 70,000 of the seaward migrants 

 were marked by the removal of fins. 



The problem of independent races is proving to be very important 

 at Chignik. It was at first supposed that the spawning fish scattered 

 at random over the entire spawning area and that as a result the 

 entire population represented a single race. More recent observa- 

 tions indicate that the two lakes support separate groups, or races, of 

 fish. Those fish whose parents spawned in the tributaries of one of 

 the lakes seem to return to that lake to spawn, thus perpetuating the 

 independence of the two races. The two races seem to have different 

 habits ; for example : One starts its spawning migration earlier in the 

 season. These racial differences probably are responsible for a great 

 part of the complexity of the problems encountered at Chignik. 

 Special attention is to be given to the races during the next field 

 season. 



Copper River red sahnon. — An investigation of the red-salmon runs 

 of Copper River, Alaska, has been continued by Seton H. Thompson, 



