by the drop in lake level was a 10° F. de- 

 crease in stream temperature at the outlet 

 of the lake. These factors apparently tem- 

 porarily upset the normal behavior of the 

 spawning salmon. No relationship could be 

 demonstrated between spawning peaks and daily 

 water level, flow, or temperature. 



As the season progressed, the number of 

 spawned females unaccompanied by males 

 increased consistently, reaching a peak by 

 October 1, 7 days after the final spawning 

 peak. This phenomenon was a reflection of 

 the longer life of females after completion 

 of spawning. Females continued to guard the 

 redd aggressively after completing spawning 

 and until they died. 



The upstream counts of sockeye salmon at 

 the weir (fig. 3) indicate a trimodal pattern 

 similar to that indicated by spawning activity 

 in the observation area (fig. 9). Tagging data, 

 however, show that the first wave in weir 

 counts included very few Brooks River fish. 

 Most Brooks River fish that subsequently 

 returned downstream through the weir were 

 from the second and third waves of upstream 

 weir counts. 1 conclude that migration Into 

 Brooks Lake, as indicated by weir counts, 

 was not closely related to the spawning waves 

 in Brooks River, but a definite relationship 

 was detected between migration downstream 

 from the lake at the weir and the spawning 

 waves. The August 26 spawning wave was 

 preceded by a peak downstream movement 

 un August 6; the Septemlier 7 wave by the 

 September li peak; and the September 24 wave 

 by the September 18 peak. Neither the period 

 of precedence nor the comparative magnitude 

 are consistent, liui tiiis relatit)naliip lieiween 

 downstream migration from the lake and 

 peaks of spawning in Brooks River (and 

 results of tagging at the weir) prove that 

 many of the sockeye salmon included in tlie 

 counts at Brooks weir are from the Brooks 

 River stock. 



l^'rom the 1957 data I hypothesize two pos- 

 sible kinds ol causative factors responsible for 

 this wave pattern of nuiss spawning: The time 

 of spawning Is governed by an endogenous fac- 

 tor somehow related to the time spt-nt In fresh 

 water or by an exogenous environmental factor 



that induces sockeye salmon to spawn during 

 successive periods. If the first hypothesis is 

 true, fish must migrate from the ocean in waves 

 timed similarly to spawning waves. In 1957 

 we had no means of verifying this. If the second 

 hypothesis is true, we should be able to recog- 

 nize the influencing factors by careful study 

 and observation. In 1957 none of the exogenous 

 environmental factors measured appeared to 

 influence spawning, with the possible exception 

 of competition for spawning space. 



The behavior of individual sockeye salmon 

 in the observation area was found to follow 

 certain well-defined patterns closely asso- 

 ciated with the three spawning waves. The 

 sequence of behavior through a spawning was 

 generally as follows: 



1. Groups of 10 to 50 sockeye salmon, 

 predominantly males, moved upstream from 

 deep pools onto spawning riffles. The most 

 distinctive behavior pattern noted was what 

 we termed a "sidling bluff." This started 

 when two males about the same size moved 

 toward each other until they were side by 

 side. They then swam diagonally across the 

 stream, eye-to-eye. When one bank was 

 reached, they quickly reversed positions and 

 sidled back to the other bank where the process 

 was repeated. After several trips back and 

 forth, one of the males began to give ground. 

 The bluffing was ended with the victor crowding 

 the loser until the loser fled downstream to 

 escape. No biting or other violence was evi- 

 dent during the sidling bluff nor were females 

 ever involvetl. 



2. A pair of fish would establish a redd 

 in a location not strongly defended by another 

 female. When many fish were present, a 

 female was often chased from place to place 

 before finding an undefended site. 



3. Small males (jack salmon) actively 

 tried to establish iheniselves with a female 

 during the redd site-seeking stage. They 

 were never observed succeeding, always being 

 forced away by a larger male. 



4. Large females were most sought by 

 tlu- males, resulting in ihe largest and most 

 powerful males spawning with the largest 



15 



