ROUNSEFELL: COMMENTS ON DECLINE OF KARLUK SALMON RUNS 



anatomical position) to the back of the head in 

 36 days increased in percentage of body length 

 (measured from the back of the head) by 5.08% . 

 Thus we do not need to postulate races on the 

 basis of the unpublished findings of Gard and 

 Drucker. 



If we were to assume (as done by Van Cleve 

 and Bevan) that the early and late spawners in 

 the same tributary are of separate races, then 

 there is no plausible explanation of why large 

 numbers of late-running 43 grilse should presage 

 a large total run of 03 age fish in the following- 

 year, yet the coefficient of correlation between 

 them over a 25-yr period is 0.956 and highly 

 significant. Rich describes these 43 grilse 

 (Gilbert and Rich, 1927:32-33), 



A larger size of grilse, belonging 10 group 3, is one 

 year older than those above mentioned and returns in 

 its second season in the sea as 4-year fish. These are 

 still largely, but not exclusively, males, and are under- 

 sized fish of little value .... conspicuously deficient in 

 color of flesh and amount of oil 



I question the quotation concerning fry from 

 river spawners (the report says it was a 1951 

 observation but from what field notes or unpub- 

 lished manuscript is not divulged) in which 

 Walker is purported to state concerning these 

 river-derived fry that, 



.... These fish were, on the average, three milli- 

 meters longer and had heavier bodies than those seen 

 elsewhere. [Italics mine.] 



Such a supposedly exact comparison between 

 two or more sets of data without divulging the 

 source is questionable. Did he compare actual 

 measured and weighed samples with fry from 

 spring, summer, or fall spawners in lateral or 

 terminal tributaries of Karluk Lake? Or with 

 beach spawners, or with fry from Thumb or 

 O'Malley Lakes? 



Van Cleve and Bevan follow this by stating, 



. . . .Walker and Bevan maintained that since fry from 

 the Karluk River are largest at the time of emergence, 

 they should also be largest at the time of seaward mi- 

 gration if they retain their size difference during lake 

 residence. 



This statement apparently infers that biolo- 



gists should believe that fry from eggs spawned 

 in June are smaller when they emerge from the 

 gravels in the spring than are fry from eggs 

 spawned in the fall. Such a statement requires 

 carefully gathered data for substantiation. It is 

 more likely that if data e.xists showing smaller 

 fry descending into the lake from tributary 

 streams, the fry were from the late spawners in 

 those streams and that the fry from the early 

 spawners descended into the lake in the very 

 early spring, probably unobserved. 



The probability of a very early lakeward 

 migration of the fiy from the early spawners 

 is suggested by records of the Karluk hatchery 

 (Moser 1901:345-346), 



According to a report from the hatchery, under date 

 of November 3, 1900, all the June eggs and part of 

 the July eggs had hatched out, producing an excellent 

 lot of healthy fry. It was found (1900) that the earlier 

 eggs and the September eggs were the- best, while a 

 portion of those taken during the middle of the season 

 were of indifferent quality. . . . 



After the fry are hatched out they escape to the 



bottom of the trough As they age they require 



more space, but they are usually held in the parent 

 trough until the egg — or umbilical — sac is absorbed, 

 a period of about ten weeks, depending upon the tem- 

 perature of the water. 



This suggests that the spring fry would be 

 ready to emerge from the gravel of the redd by 

 the middle of January. 



At another point Van Cleve and Bevan para- 

 phrase an additional unpublished manuscript 

 by Walker and Bevan saying, 



.... Walker and Bevan (1955)** noted that the largest 

 spawning population of sockeye salmon in the Karluk 

 watershed was found in the Karluk River. 



This type of generalized statement, without 

 presentation of data is unconvincing, especially 

 as in the second paragraph of the whole paper 

 Van Cleve and Bevan used the factor of 10% 

 of the total run to estimate the numbers spawn- 

 ing below the lake in 1971, as determined by 

 the National Marine Fisheries Service in 1963. 

 In fact, it means almost nothing for them to say 

 "largest" when in another place they say 

 400,000 (out of an escapement of 2,600,000) 

 was "more fish than were recorded for any other 

 part of the watershed." 



655 



