ROUNSEFELL: COMMENTS ON DECLINE OF KARLUK SALMON RUNS 



early September. Van Cleve and Bevan cite 

 Walker and Bevan (1955),- 



.... They also maintained that the decrease in Kar- 

 luk River spawning could explain the increase in rela- 

 tive numbers of 4-yr smelts, since the river spawners 

 were predominantly of 53 age. 



To refute this claim one has only to look at 

 Figure 2 showing the seasonal occurrence of 

 the total runs and of the 53 age group during 

 the 30-yr period in which this alleged reduction 

 in Karluk River spawners was supposed to have 

 taken place. In other parts of their paper they 

 state that the center of the run, which they 

 claimed spawned in the main river, was sup- 

 posed to have been depleted before 1921; you 

 cannot have it both ways. Very obviously the 

 decrease in the proportion of fish of 53 age has 

 been occurring almost equally in all segments 

 of the runs. 



The authors place undue emphasis on a quo- 

 tation from Chamberlain (1907), which they 

 suggest proves that in the early years there was 

 only one peak in the run. As I shall mention 

 later, Cloudsley Rutter actually said that there ^ 

 were two runs, but that only one occurred in 

 the year (1903) he visited the lake. Chamberlain 

 was not personally acquainted with Karluk but 

 was paraphrasing field notes by Rutter. In dis- 

 cussing the egg take of the Karluk hatchery for 



1898 (the parent brood year for most of the 

 1903 run described in Rutter's notes) Moser 

 (1901:342) says, 



Of the season's take the spring run therefore amounted 

 to 50.4 per cent as against 49.6 per cent for the fall 

 run ... . the monthly percentages of fish spawned are 

 as follows: June, 0.5 per cent; July, 47.9 per cent; 

 August, 2.9 per cent; September, 41.5 per cent; Oc- 

 tober. 6.8 per cent; November, 0.4 per cent. 



In discussing the hatchery operations for 



1899 Moser (1901:343) lists the eggs taken as 

 coming from the spring run and from the fall 

 run. Again, on page 344 Moser states, 



It would appear from the above that the eggs eye 

 very much faster with the spring run. and that the 



2 Walker, C. E., and D. E. Bevan. 1955. Observations 

 on the biology of the red salmon in the Karluk watershed. 

 Unpubl. manuscr. Univ. Wash., Fish. Res. Inst., Seattle, 

 W A 98195. 



hatching range covers a much longer period. It is also 

 apparent that in considering the hatching of redfish at 

 Karluk the two runs must be treated separately — ihe 

 runs arc so marked and the prevailing conditions so 

 radically different .... the early run in 1899. under 

 natural conditions of temperature, hatched in an aver- 

 age of 129 days, whereas the fall run required 198 days. 

 [Italics mine.] 



The constant reiteration by Van Cleve and 

 Bevan of the notion of a former run with only 

 one mode is quite contrary to their insistence 

 that the run consists of subpopulations and that 

 only the main river spawners are of any impor- 

 tance in maintaining the run. 



If we concede for the moment that there are 

 strong tendencies for most sockeye salmon to 

 attempt to spawn in their natal area (within 

 reasonable limits), then we can make a separate 

 assessment of the relative reproductive success 

 of different portions of the run. For the 25-yr 

 period from 1921 to 1945 I have compared the 

 spring, summer, and fall escapements (only 

 spring escapement was available for 1934) with 

 the size of the run returning 5 yr later during the 

 same respective seasons. The resulting returns 

 per spawner for all three periods (73 compari- 

 sons) are shown in Figure 3. Table 1 shows 

 which of the three groups of spawners was most 

 successful for the same number of spawners, 

 and on odd- and even-numbered years. 



This table shows that if the theory of inde- 

 pendent seasonal subpopulations at Karluk 

 advanced by Van Cleve and Bevan is correct, 

 the spring and summer spawners are about 

 equally successful; the fall spawners are much 

 less successful. It also shows that fall spawners, 

 while successful on the odd-numbered years, 

 are very unsuccessful on the even-numbered 

 years. This analysis therefore strongly suggests 

 that it is chiefly the fall spawners (not the 

 summer spawners as Van Cleve and Bevan 

 contend) that use the main river and are thus 

 in direct competition with the large even-year 

 pink-salmon, 0. gorbuscha, runs. 



Concerning the question of races and season- 

 al races I should like to emphasize the lack of 

 factual data in the report of Van Cleve and 

 Bevan. They quote liberally and at length from 

 various unpublished manuscripts and inhouse 

 mimeographed memoranda. Thus they cite an 

 unpublished manuscript by Card and Drucker 



653 



