HOOPES; SELECTION OF SPAWNING SITES 



FACTORS AFFECTING SPAWNING 

 SITE SELECTION 



Salmon and trout usually spawn in a specific 

 type of microenvironment within a broadly uni- 

 form environment. One such type of microen- 

 vironment occurs at the downstream ends or 

 tails of pools, just upstream from the point 

 where the flow breaks over the lip of the pool 

 into the riffle below (Needham and Taft, 1934; 

 Schultz and students, 1935; Smith, 1941; White, 

 1942; Briggs, 1953; Shapovalov and Taft, 1954). 

 Riffles may also contain microenvironments at- 

 tractive to spawning salmon' (Belding, 1934; 

 Burner, 1951; Needham and Vaughan, 1952; 

 Briggs, 1953; see also footnote 6.) 



Variations in streambed slope are important 

 in the selection of spawning sites because these 

 variations cause exchange in intragravel and 

 stream water (Vaux, 1962) and affect the sta- 

 bility of the bottom. In pools, spawning sites 

 were usually located where the bottom sloped 

 upward slightly in the direction of flow, as in 

 pool tails. Even in riffles, fish observed during 

 this study seldom constructed redds where the 

 bottom sloped steeply but rather chose gently 

 sloping sites where the excavated tailspill pro- 

 vided a slight rise. The preferred streambed 

 was also nearly level across the stream, although 

 moderate to steep lateral inclines were occasion- 

 ally used in crowded areas. 



In 1959, sockeye salmon first entered and be- 

 gan spawning in Hidden Creek on August 11. 

 On August 13, I marked 14 redds in an area of 

 high spawning density after taking detailed 

 notes on the exact location of each spawning site 

 and the features of the immediate environment. 

 I assumed that the first fish in an area selected 

 the preferred sites, so that these 14 sites repre- 

 sented preferred habitat. All of the sites were 

 at or near the tails of pools and along the bank 

 adjacent to the main streamflow. I observed the 

 sites until August 25, when high water precluded 

 further study. 



' J. S. Chambers, G. H. Allen, and R. T. Pressey. 

 Washington State Department of Fisheries, Annual Re- 

 port, 1955, submitted to U.S. Army Corps of Engineers, 

 175 p. 



Observations in the immediate vicinity of the 

 14 study redds showed that late-arriving fish 

 spawned upstream from the initial redd at the 

 tail of each pool. When the pool was filled with 

 redds for its entire length and pool width per- 

 mitted two or more redds side by side, late ar- 

 rivals chose spawning sites sequentially up- 

 stream, again beginning at the tail but across 

 the stream from the original site. No salmon 

 spawned in unstable gravels usually associated 

 with the shallow sides of pools. 



Riffles made up 60 to 70% of the stream length 

 accessible to salmon in the three study streams, 

 and many fish spawned in these riffles. As in 

 the pools, spawning salmon preferred definite 

 parts of the riffles. The first fish invariably 

 chose sites along the deeper side and higher 

 bank. Salmon also preferred certain riffles over 

 others. Short riffles with water of moderate 

 depth and flow were used as readily as pool tails. 

 Long shallow riffles with some cover along adja- 

 cent banks were occupied next. The last sites 

 used were those in long shallow riffles without 

 nearby pools or cover. 



The significance of cover in the selection of a 

 spawning site by sockeye salmon is hard to eval- 

 uate. Hourston and MacKinnon (1957) found 

 that pink salmon, 0. gorhnscha, spawning in an 

 artificial channel selected sites adjacent to cover 

 rather than those distant from cover. I fre- 

 quently observed that sockeye salmon, especially 

 those that arrived early when a wide choice of 

 sites was available, selected a site that had good 

 cover. The absence of cover did not necessarily 

 prevent early spawners from selecting a par- 

 ticular site, but given two adjacent locations ap- 

 parently equal with regard to water depth and 

 flow and gravel composition, the fish tended to 

 select the site nearer to cover. Favored sites 

 were near deeply undercut banks, banks over- 

 hung with tall grass and herbaceous plants, holes 

 washed under tree roots, log and brush jams, 

 and deep pools. 



Cover that attracts alarmed fish may not ac- 

 tually afford protection, because concealment 

 alone does not insure safety. I have observed 

 brown bears, Ursus gyas, systematically search 

 for salmon under overhanging vegetation and 

 undercut banks. In my three study streams. 



455 



