FISHERY BULLETIN: VOL. 70, NO. 2 



cover that would protect a salmon from bears 

 was quite scant, and even when such cover was 

 available, alarmed fish often sought the shaded 

 area provided by overhanging herbs and grasses 

 as refuge rather than places of comparatively 

 greater security such as undercut banks and logs. 



The selection of a site by a spawning sockeye 

 salmon seemed to be more closely related to the 

 composition of the stream bottom than to any 

 other single factor studied. Work done with fish 

 closely related to salmon supports this view 

 (Hobbs, 1940; Fabricius, 1950; Fabricius and 

 Gustafson, 1954; Needham, 1961). This pre- 

 ference by salmon for areas having bottom ma- 

 terials of intermediate composition explains why 

 the density of spawners was high in sections 

 with moderate gradient and low in sections with 

 either high or low gradient. When more desir- 

 able bottom material was lacking, fish spawned 

 in excessively coarse material (higher gradi- 

 ents) before excessively fine (lower gradients). 

 In this respect, my observations indicated that 

 gradient was related to the selection of spawning 

 sites by sockeye salmon. The observations by 

 Trautman (1942) and Huet (1949, 1959) show- 

 ing that distribution of several species of fish 

 permanently resident in lotic environments is 

 closely related to stream gradient may well have 

 resulted from a preference by those fish for a 

 particular substrate. 



Factors other than physical conditions in the 

 spawning grounds may also influence the selec- 

 tion of redd sites by sockeye salmon. Briggs 

 (1955) stated that among anadromous fishes, 

 the earliest arrivals in a particular spawning 

 tributary traveled to the most distant spawning 

 grounds and the later arrivals occupied areas 

 closer to the stream mouth. In the three streams 

 I studied, however, early arrivals did not neces- 

 sarily ascend to the uppermost spawning areas. 

 Moreover, the sites selected by later arrivals 

 often were located between sites chosen by the 

 earlier spawners. 



The female sockeye salmon selects the spawn- 

 ing site, usually near other spawners (Noble, 

 1938). The presence and stimulation of other 

 spawners may be necessary for successful 

 spawning even though much eff"ort is expended 

 in territorial defense. I seldom saw single 



spawning pairs but observed rather that fish 

 tended to form more or less isolated groups of 

 several pairs. How much of this behavior is 

 due to preference for certain physical charac- 

 teristics of the spawning grounds and how much 

 to innate social behavior is not known. 



SUMMARY 



1. Spawning site selection by sockeye salmon 

 in three small lateral tributaries entering Brooks 

 Lake, Alaska, was studied from 1959 through 

 1961. 



2. Several waves of sockeye salmon spawners 

 ascended the principal study stream in 1959 and 

 1960 when the numbers of fish were moderate 

 and high but not in 1961, a year of relatively few 

 fish. 



3. Both temporal and spatial distribution pat- 

 terns were defined and found to be relatively 

 constant despite large fluctuations in the number 

 of spawners. 



4. The distribution of spawning fish was cor- 

 related with stream gradients. Little or no 

 spawning occurred in areas having gradients of 

 less than 0.5% or more than 2.0%. 



5. In preferred spawning areas, cobbles from 

 2.5 to 7.6 cm in diameter made up about 50% 

 of the bottom, materials 1.3 to 2.5 cm about 25%, 

 and particles less than 1.3 cm the remainder. 



6. Spawner distribution was more closely re- 

 lated to the composition of the stream bottom 

 than to any other single factor studied. 



7. Given two adjacent locations equally de- 

 sirable with regard to hydrological conditions 

 and bottom composition, sockeye salmon tended 

 to select the site with the most nearby cover 

 and concealment. 



8. Early arrivals did not necessarily use the 

 uppermost spawning areas, and later arrivals 

 primarily filled in acceptable areas already in 

 use.' 



ACKNOWLEDGMENTS 



This study was conducted while the author 

 was employed as a seasonal fishery aid at the 

 National Marine Fisheries Service Field Station, 



456 



