FISHERY BULLETIN: VOL. 71, NO. 3 



The text of the report by Burgner et al., how- 

 ever, says that in the shallow torrential lateral 

 streams in the Karluk system, space require- 

 ments were often substantially less than 2 m-'. 

 They also say their figure is only an approxi- 

 mation complicated by the occurrence of suc- 

 cessive waves of spawners in successive streams. 

 They also admit incomplete information on the 

 amount of potential spawning ground, especially 

 on lake beaches, and their figure includes only 

 the beach spawning areas of Thumb and 

 O'Malley Lakes. Observations of beach spawn- 

 ing have been published, mentioning Tent 

 Point, Meadow Point, Cascade Creek, Moraine 

 Creek, Canyon Creek, Halfway Creek, off 

 Boulder Point, off Grassy Point Creek, etc. 



Fred Lucas estimated in 1924 that sockeyes 

 spawning in O'Malley River would average a 

 l)air to each square yard. (Gilbert and Rich, 

 1927:20). 



The estimate of the area (Burgner et al., 

 1969) of river spawning beds is quite meaning- 

 less in my opinion without information on the 

 location of these beds along the 30-mile course 

 of the Karluk River. Furthermore, in the esti- 

 mates of main river spawning areas for other 

 systems redd sites range from 3 to 8 m- rather 

 than the 2 m- used for Karluk. 



Burgner et al. (1969:457) also say, 



The individual spawning areas in the Karluk system 

 are occupied continuously tor about 5 weeks to 5 

 months by a succession of spawners, with the result 

 that many more spawners are accommodated than 

 could be if they all spawned in 2 or 3 weeks. This 

 occupation by successive waves of spawners introduces 

 questions as to the effect of superimposition of redd 

 sites on the success of spawning. H-V do not know the 

 answers. [Italics mine.] 



Since the sources of Van Cleve and Bevan's 

 spawning area data readily admit that they do 

 not have the answers to many of the questions 

 raised in attempting to convert spawning area 

 into redd sites, I believe that their figures need 

 considerable qualification and cannot be taken 

 seriously. 



From the above discussion I suggest that the 

 number (ignoring the main river since there is 

 no locality information given) of redd sites 

 would be more like the following: 



Total 



236,000 



That is, the lake and its tributaries should, 

 have, and can accommodate about half a million 

 spawners without overcrowding. 



It is apparent that the Karluk sockeye salmon 

 run is continuing the decline that was acceler- 

 ated by destruction, since 1921, of the natural 

 cyclic character of the runs. Restoration of the 

 runs at this stage can scarcely be expected from 

 merely assuring a more even seasonal distri- 

 bution of spawners. 



Two important factors I stressed 15 yr ago 

 were the control of both density-dependent and 

 density-independent predators in order to raise 

 the number of smolts per spawner leaving the 

 lake, and hopefully to raise the biomass of 

 smolts in the lake to a level where eventually it 

 would lower the threshold size of the migrating 

 smolts, so that we would have a reversal of the 

 trend toward more 4-yr and fewer 3-yr smolts. 



In this regard I note that Van Cleve and 

 Bevan refrain from mentioning the high mortal- 

 ity of spawning salmon by Kodiak bears. 

 Shuman (1950) reported a bear kill of un- 

 spawned salmon of 94,000 or 19.4% of the 

 1947 spawning escapement of 485,000. If this 

 quantity has been lost annually since I recom- 

 mended control 15 yr ago it means a loss of 

 1,400,000 spawning salmon. 



I also recommended (Rounsefell, 1958) that 

 an attempt be made to restore the cyclic char- 

 acter of the runs, stating, 



The attempt to stabilize the runs by obtaining a high 

 number of spawners in every year has largely destroyed 

 and obscured the former cyclic character of the runs. 

 During the period when these cycles were present the 

 number of spawners fluctuated in a more or less regular 

 manner from very high to very low (lower than most 

 recent years). This wide variation in number of spawners 

 resulted in wide oscillations in the numbers of young 

 sockeye present in the lake and therefore available as 

 food for predator fishes. These regular oscillations in 

 the supply of available prey may have acted as a control 

 on the abundance of predators. 



I then explained more fully its purpose as 

 follows: 



658 



