VAN CLEVE and BEVAN: DECLINE OF KARLUK SALMON RUNS 



of upper Fraser sockeye salmon which Talbot 

 estimated must have been as high as 70 to 80% 

 of the total run reaching Hell's Gate in some 

 years. Thompson (1945) also felt that the ef- 

 fects of delays in migration would increase as 

 the time of spawning was reached. Thus, as 

 each run approached its spawning grounds, and 

 as the time for spawning approached, a shorter 

 delay would be fatal. Even a short delay at the 

 Karluk weir could have been as fatal to some 

 sockeye salmon races as the longer delay at 

 Hell's Gate on the Fraser River was to the up- 

 river spawners which in the Fraser had several 

 hundred miles to go to the nearest spawning 

 ground. 



A more important effect of the weir, located 

 just below the outlet of Karluk Lake could have 

 been the same as that observed by W. F. Royce 

 in 1957 at the Brooks Lake weir (Royce, pers. 

 comm.). In 1957, Royce noted that a school of 

 salmon that had migrated into Brooks Lake, 

 and had been holding in the lake not far from 

 the outlet, moved down against the upper side 

 of the Brooks Lake weir. On removal of some 

 pickets in the weir this entire school of salmon 

 moved down into the Brooks River where it 

 spawned. Nothing similar to this has been re- 

 ported in the literature on the Karluk River, 

 but Bevan has observed similar behavior of 

 schools of adult sockeye salmon above the Kar- 

 luk weir. The same behavior has been observed 

 by John Roos (pers. comm.) in Chilko Lake 

 and River. The Karluk weir thus could have 

 prevented salmon from returning down river to 

 spawn after moving into the lake. Richard Gard 

 has informed us (pers. comm., 1972) that some 

 time after 1957 sockeye salmon were being 

 counted downstream through the Karluk weir 

 during the end of the season. Nevertheless the 

 weir has been a barrier to the free movement 

 of sockeye salmon adults up and downstream 

 and has prevented easy access of the young to 

 the lake. Thus it must have reduced the pro- 

 ductivity of the Karluk River spawning grounds. 



Gard and Drucker (see footnote 5) determined 

 the number of "red salmon" spawning in the 

 upper Karluk River by marking sockeye salmon 

 at a weir installed at the portage and recovering 

 the tags and establishing a ratio of tagged to 

 untagged fish as they passed through the Kar- 



luk weir. The number spawning in the river 

 was calculated to be about 10% of the total es- 

 capement to the Karluk watershed. 



The chain of events noted on the Birkenhead 

 River and at Hell's Gate in the Fraser River 

 system indicate that sockeye salmon which are 

 prevented from spawning in their ancestral 

 grounds do not do so successfully elsewhere. 

 While all of the young derived from eggs taken 

 from the Birkenhead run were ])lanted in other 

 streams, the absence of fish from the upper Bir- 

 kenhead River for a number of years after the 

 weir had been removed indicates that the fish 

 which formerly spawned above it were not casual 

 strays which spawned in the upper Birkenhead 

 because it happened to be a good spawning ground 

 and was available for occupation. While there 

 seems to have been no occasion for study of the 

 degree to which the homing of sockeye salmon 

 is specific for different sections of any spawning 

 ground, the recent work of Hara, Ueda. and 

 Gorbman (1965) supports the investigations of 

 the International Pacific Salmon Fisheries 

 Commission on the Fraser River, which indi- 

 cate that specific streams are occupied by indivi- 

 dual races of sockeye salmon. 



This was also indicated by the experiments 

 reported by Hartman and Raleigh (1964) in Mea- 

 dow Creek, a tributary of Karluk Lake which 

 proved that the sockeye salmon refused to spawn 

 in other streams when transported from Meadow 

 Creek and died unspawned when denied access 

 to what was obviously their natal spawning 

 ground. 



The adaptation of sockeye salmon to a speci- 

 fic type of spawning ground has been further 

 verified by Raleigh (1967) who pointed out the 

 different migratory behavior of fry si)awned in 

 the outlet stream, Karluk River, and in Meadow 

 Creek, a tributary of Karluk Lake. When tested 

 in an experimental apparatus only 1% of the 

 Meadow Creek fry moved upstream while 98% 

 moved downstream which would take them into 

 Karluk Lake, their nursery lake. When tested, 

 30% of fry from the Karluk River moved up- 

 stream while 66% moved downstream. Moreover 

 the Meadow Creek fry were found to move al- 

 most entirely at night while Karluk River fry 

 moved in either the day or night. Raleigh con- 

 cluded that this behavior must be genetically 



639 



