much the same thing, these areas proved to be of 

 little value in evaluating relations between 

 mortality and oxygen supply. 



Differences in mortality level between intertidal 

 Harris River and Indian Creek could not be 

 demonstrated with estimates of M T and M t . 

 However, estimates of p made in mid-October 

 showed that samples containing 35 or more dead 

 eggs per m. 2 occurred with almost equal frequency 

 in intertidal Harris River and Indian Creek 

 despite much lower spawning density in intertidal 

 Harris River (about 53 percent of intertidal 

 Indian Creek). This evidence suggested that egg 

 mortality was higher in intertidal Harris River 

 than in intertidal Indian Creek. 



I960 Brood Year 



Egg mortality was studied in intertidal Twelve- 

 mile Creek, Harris River, and Indian Creek 

 (tableS). 



High early mortality from causes associated 

 with the direct removal of affected specimens from 

 the spawning bed occurred in intertidal Twelve- 

 mile Creek for the third year. In 1960, however, 

 most of this high mortality occurred during spawn- 



A 



ing (M t =0.71 in late September). 7 Since mor- 

 tality from causes associated with spawning (e.g., 

 redd superimposition) is beyond the scope of this 

 paper, these causes will not be considered further 

 here. By late November there was evidence of 

 increased egg mortality in intertidal Twelvemile 

 Creek, but this mortality was relatively low. 



The density of females spawning in intertidal 

 Harris River and Indian Creek was relatively high 

 and nearly equal. Late in September at the con- 



7 This same phenomenon was also observed in 1961 (unpublished data, 

 FRI, University of Washington, Seattle). 



elusion of spawning, estimates of M T and p were 

 very nearly the same for these two areas. At 

 hatching, however, 35 or more dead eggs per m. 2 

 were found at 26 percent of the points sampled in 

 intertidal Harris River and at only 10 percent of 

 the points sampled in intertidal Indian Creek. 

 Estimates of M T also indicated that mortality of 

 eggs remaining in the streambed was higher in 

 intertidal Harris River than in intertidal Indian 

 Creek. 



STABILITY OF THE SPAWNING BED AND 

 MORTALITY 



Two factors causing gravel to shift in spawning 

 beds are flooding and females digging redds. The 

 importance of redd superimposition as a factor 

 limiting production of salmon fry is beyond the 

 scope of this paper and will not be considered. 

 My discussions will be limited to the influence of 

 flooding and debris movement on egg and larval 

 mortality. 



Mortality caused by gravel shift would make 

 itself evident by complete disappearance of eggs 

 and larvae from spawning beds. Changes in 

 abundance and distribution of eggs and larvae 

 will be examined to obtain evidence of mortality 

 caused by gravel movement. Three population 

 parameters will be considered in evaluating the 

 stability of spawning beds: (1) Total mortality 

 (M t ), (2) mean abundance of eggs and larvae per 

 m. 2 (live plus dead)(P), (3) fraction of points 

 containing fewer than 35 live plus dead eggs per 

 m. 2 {&'). 



High discharge occurs most frequently in South- 

 eastern Alaska streams during October, Novem- 

 ber, and December. Autumn storms are often 

 accompanied by heavy rain which sometimes 



Table 8. — Estimates of M t , M r , and p (1 used to evaluate time and magnitude of mortality of I960 brood year eg 



Spawning area and date 



Intertidal Harris River: 



Sept. 28 



Dec. 2 



Intertidal Indian Creek: 



Sept. 22 _. 



Nov. 22 



Intertidal Twelve-mile Creek 



Sept. 30 



Nov. 30 



Mi 



.44 

 .50 



.71 



.75 



90-percent 



confidence 



limits of 



the mean 



±0.18 

 ±.09 



±.20 

 ±.22 



M, 



0.03 

 .18 



90-percent 



confidence 



limits of 



the mean 



±0.01 

 ±.07 



Pa l 



0.92 

 .74 



90-percent 



confidence 



limits of 



the mean 



±0.05 

 ±.07 



±.05 

 ±.06 



±.02 

 ±.03 



1 Po is the fraction of points containing less than 35 dead eggs per square meter. 

 - This estimate of M, was obtained with log-transformed data. 



SPAWNING BED ENVIRONMENT OF PINK AND CHUM SALMON 

 774-711 O— 66 15 



513 



