FISHERY BULLETIN: VOL. 71. NO. 3 



controlled but was unableto explain satisfactorily 

 the variability in behavior of outlet fry. 



The nature and mechanisms controlling these 

 movements of sockeye salmon fry have been in- 

 vestigated by Brannon (1967, 1972) who found 

 that these movements are basically genetically 

 controlled though they are subject to environ- 

 mental factors which require flexibility in the 

 response of the newly emerged fry. The genetic 

 nature of these movements was proven by cross- 

 ing two races, one in which the fry moved down- 

 stream to their nursery lake with another in 

 which the fry must move upstream. The progeny 

 of these crosses showed intermediate behavior. 

 Thus the behavior of fry in the Karluk River 

 agrees with that shown by progeny of the Little 

 River and Chilko races in the Fraser River. The 

 immediate movement of some fry into Karluk 

 Lake followed some time later by the upstream 

 movement of larger fingerlings is in complete 

 agreement with similar activities described by 

 Brannon (1972) for the Fraser River sockeye 

 salmon. 



The more complicated behavior of Weaver 

 Creek fry mentioned by Ricker (1972) and .studied 

 in detail by Brannon (1972) is only an extended 

 variation of behavior found in similar form but 

 under slightly different conditions in Chilko 

 River, Little River, and the South Thompson 

 in the Fraser River system and in the Karluk 

 River. Those fry which move or are carried 

 downstream on emergence are found resting in 

 more quiet water below and later are seen as a 

 wave of larger fingerlings which move upstream 

 along the margins of their spawning grounds 

 into their nur.sery lakes. 



The genetic basis of this behavior leaves no 

 basis for doubt that the Karluk watershed is in- 

 habited by several races of sockeye salmon. The 

 insistence of adults upon entering a particular 

 stream indicates that the fish spawning in the 

 different streams must belong to separate races. 

 The behavior of sockeye salmon in the Brooks 

 River and the evidence presented by the progres- 

 sive elimination of sockeye salmon spawning 

 in different sections of the Birkenhead River 

 gives further substance to the probability of the 

 existence of individual races of sockeye salmon 

 that spawn in different sections of the Karluk 

 River itself. 



Ricker (1972:30) also discussed the "over- 

 shooting" and "proving" of spawning grounds 

 by salmon, an aspect of homing that has been 

 noted in the past but which has received little 

 attention until recently. He includes in this 

 activitiy the return of most transplanted sock- 

 eye salmon to the vicinity of the stream from 

 which they were removed by Hartman and 

 Raleigh (1964) and the overshooting of Sweltzer 

 Creek sockeye salmon up the Chilliwack River 

 at least 0.5 km reported by Ricker and Robertson 

 (1935) as well as the movement of sockeye salm- 

 on through the Brooks River into Brooks Lake, 

 described by Hartman and Raleigh (1964), 

 where they remained until ready to sjmwn when 

 they dropped downstream into the Brooks River. 

 The latter incident is mentioned above as it was 

 first discovered by W. F. Royce. Moreover, simi- 

 lar movements of sockeye salmon have been 

 noted in Chilko Lake and in the Karluk River 

 (see above). The establishment of a barrier, such as 

 a weir, in the middle of a spawning ground or 

 above it, as at Karluk, in limiting this movement 

 of sockeye salmon therefore must reduce the 

 productivity of those grounds. 



Whatever the causes, by the time the size of 

 the Karluk River race of spawners was 

 estimated by tagging in the late 1950's it had 

 been reduced to 10% of the total escapement to 

 the Karluk watershed. The total returns from 

 each year's s])awning computed from ages ob- 

 tained from samples taken at the weir are shown 

 in Figure 3 and indicate that these returns 

 reached a low point in 1950 and 1951 of about 

 250,000 fish. While the possible errors involved 

 in using samples taken at the weir to determine 

 the distribution of ages offish taken in the catch 

 as well as the errors in estimating age men- 

 tioned by Walker are unknown, the dominance of 

 5- and 6-yr-old fish is sufficient to give credence 

 to these returns. 



As a result of these recommendations attempts 

 were made to protect these midseason sockeye 

 salmon especially in the odd-numbered years 

 when pink salmon are scarce. Apparently the 

 total returns from the spawning years of 1950 to 

 1957 responded to these regulations as well as 

 to efforts to improve passage at the weir, and 

 increased from about 250,000 in 1951 to 1,100,000 

 in 1957. But to accomplish this the catch was 



640 



