Twelvemile Creek spawning areas made it im- 

 practical to evaluate winter mortality there. 



Only a small number (about 1 percent) of 1960 

 brood year larvae collected in late winter 1961 from 

 intertidal Harris River and Indian Creek were 

 dead. Of 933 larvae collected in intertidal 

 Twelvemile Creek, 9.4 percent were dead, but 

 most dead larvae were found in one redd. Factors 

 other than freezing were suspected of having 

 caused their death. 



An experiment indicated that little or no freez- 

 ing of intragravel water occurred in winter 1960- 

 61. One-half-dram vials were filled with water 

 and buried 5 and 10 inches in spawning beds 

 during autumn 1960. They were recovered in 

 spring 1961 and examined for breakage caused by 

 water expanding during freezing. The recovery 

 of a broken vial was used to indicate occurrence of 

 freezing at the point of burial. Breakage from 

 causes other than freezing was minimized by en- 

 closing vials in latex tubing. Sixty-three vials 

 were recovered from intertidal Indian Creek, 36 

 from intertidal Harris River, and 30 from inter- 

 tidal Twelvemile Creek. There were no broken 

 vials in the lot. This experiment provided further 

 evidence that intragravel water did not freeze in 

 winter 1960-61. 



DISCUSSION AND CONCLUSIONS 



There is evidence that mortality of pink and 

 chum salmon from egg deposition to fry migration 

 is seldom less than 75 percent and commonly ex- 

 ceeds 90 percent in small coastal streams. Fresh- 

 water mortality, therefore, may place more severe 

 restrictions on production of pink and chum 

 salmon than natural salt-water mortality. 



A thorough understanding of the ecological and 

 physiological requirements of pink and chum 

 salmon eggs and larvae is necessary for evaluating 

 the potential of spawning beds to produce fry. 

 It has been a goal of these studies to determine the 

 relative importance of certain factors affecting 

 quality of the spawning bed environment and fry 

 production. 



OXYGEN SUPPLY AND RELATED FACTORS 



It is evident from a review of the literature that 

 the oxygen supply rate to an egg or larva is a 

 function of oxygen content and velocity of flow of 

 intragravel water, both of which are affected by a 



complex of interacting factors. It would appear 

 that oxygen privation is more critical in embryonic 

 than in larval stages. For evaluating mortality 

 possibly resulting from an oxygen deficiency, 

 observations on mortality were based primarily 

 on the population of eggs present in the spawning 

 bed at the time of sampling. Where large numbers 

 of eggs had disappeared from spawning beds before 

 sampling, or where density of females spawning 

 was light, it was not possible with the sampling 

 scheme used to obtain data adequate to evaluate 

 prehatching mortality possibly associated with 

 the availability of oxygen. Inadequate data for 

 the intertidal Twelvemile Creek and upstream 

 Harris River and Twelvemile Creek spawning 

 areas were of limited value in evaluating mortality 

 possibly due to oxygen privation, and observations 

 were made mostly in intertidal Harris River and 

 Indian Creek. 



Data on (1) dissolved oxygen content of intra- 

 gravel water, (2) streambed gradient, and (3) 

 organic content, size composition, and permeability 

 of bottom materials showed more favorable oxygen 

 supply in intertidal Indian Creek than in intertidal 

 Harris River. Furthermore, results of mortality 

 studies suggested that prehatching mortality not 

 associated with the disappearance of eggs was 

 lower in intertidal Indian Creek than in intertidal 

 Harris River. Hence, there was general agree- 

 ment between observations on physical quality of 

 the spawning bed environment and egg mortality. 



An opportunity arose in 1957 to evaluate the 

 relation between prolonged low streamflow during 

 spawning and egg mortality. In this instance, 

 unusually low levels of dissolved oxygen in intra- 

 gravel water were observed during the spawning 

 period. The evidence relating an exceptionally 

 high egg mortality to low streamflow corroborated 

 the findings of other workers (Brett, 1951 ; Neave 

 and Wickett, 1953; Wickett, 1958). 



In these studies it was not possible to demon- 

 strate a direct dependence of egg mortality and 

 population size. In 1957, when spawning den- 

 sities were extremely low and egg mortality was 

 unusually high, there appeared to be some justifi- 

 cation for suggesting that a minimum number of 

 spawners is required to reduce the overall bio- 

 chemical oxygen demand through the removal of 

 organic detritus from spawning beds (Ricker, 

 1962). Additional research will be required to 

 resolve this question. 



SPAWNING BED ENVIRONMENT OF PINK AND CHUM SALMON 



519 



