FISHERY BULLETIN: VOL. 70. NO. 2 



18 



16- 



U- 



1959, N= 306 

 I960,N=20I 



I96I.N=II2 



n — 



10 

 DIS 



1 1 1 — I—^ 1— 



30 40 50 60 70 80 90 

 TANCE FROM STREAM MOUTH (PERCENT) 



100 



Figure 4. — Distribution of redds occupied by sockeye 

 salmon in One Shot Creek, Brooks Lake, Alaska, during 

 the 1959, 1960, and 1961 spawning seasons. 



ers in great detail. For example, some had so 

 many steep runs and falls that the average 

 gradient appeared to be too high for spawning, 

 although the sections contained one or two short 

 reaches of stream with gradients acceptable to 

 spawning fish. Other sections had long reaches 

 of generally too low gradient but had one or two 

 short steep falls, so that little acceptable spawn- 

 ing ground was actually available even though 

 the average gradient appeared suitable for 

 spawning. Despite these problems in methodol- 

 ogy, after comparing the density of spawning 

 females and stream gradients section by section 

 in the three study streams for all 3 years, I 

 found that little or no spawning occurred in 

 areas having average gradients of less than 0.5% 

 or more than 2.0%. 



LATE 



10 20 30 40 50 60 70 80 90 

 DISTANCE FROM STREAM MOUTH ( PERCENT) 



Figure 5. — Temporal and spatial distribution of redds 

 occupied by sockeye salmon during the early, middle, and 

 late periods of the 1959 spawning season in Hidden 

 Creek, Brooks Lake, Alaska. 



SIZE OF STREAMBED MATERIALS 

 IN SELECTED SECTIONS 



I made detailed studies of bottom composition 

 in eight of the sections surveyed for gradient.* 

 Sections were selected for these detailed studies 

 on the basis of their relative amounts of use by 

 spawning salmon so that areas having high and 

 low spawning densities could be compared. Two 

 sections with high and two with low densities 

 of spawners were selected in Hidden Creek, and 

 one each was selected in Up-a-tree and One Shot 

 Creeks. 



From six to nine samples of the streambed 

 were collected in each of the eight sections. The 

 samples were taken from both pools and riffles 

 at locations where salmon spawned in 1959. 

 Each sample consisted of the top 20 to 30 cm of 

 bottom materials (gravel and sand) from an 

 area 1 m^. The material was removed from the 



* Observations were also made on water depth and 

 velocity and availability of streambank cover. 



452 



