Dissolved oxygen concentrations of intragravel 

 water were higli in August but low in September, 

 when a drought became severe. The low values in 

 September were due partly to consumption of the 

 dissohed oxygen in the stream water by decompos- 

 ing salmon carcasses. Dead salmon are usually re- 

 moved from Sashin Creek during freshets, but no 

 freshets occurred in September 1965. The dis- 

 solved oxygen cannot be replenished without ex- 

 tensive exposure to tlie air. The upper area, which 

 has a steep gradient and turbulent water, provided 

 for more rapid replenishment of dissolved oxy- 

 gen than tlie calm water in the lower area, which 

 has a shallow gradient. Metabolites that are not 

 freely exchanged with tlie atmosphere, such as 

 ammoniacal nitrogen, also may have accumulated 

 in the lower area. 



T.\BLE 11. 



—Dissolved oxygen content of intragravel water in 

 three areas in Sashin Creek, 1965 ' 



1 Methods of sampling were described by McNeil (1962). 



- The August 31 samples were collected 1 day after a freshet; the data for 

 the lower area are omitted because 60 percent of the standpipes in the area 

 had been washed away. 



I expected to find that the percentage of eggs 

 alive at the end of spawning would be higliest in 

 the upper area and lowest in the lower area be- 

 cause of the progressively decreasing amounts of 

 dissolved oxygen in intragravel water down- 

 stream. On September 29, 8-4 percent of total eggs 

 were alive in the upper area; 86 percent in the 

 middle area; and 60 percent in the lower area 

 (table 5). Although the lowest survival of eggs 

 was in the lower area, as anticipated, I was sur- 

 prised to find survival of eggs in the middle area 

 similar to that in the upper area. The percentages 

 of live eggs did not change significantly in any 



of the areas between September 29 and Novem- 

 ber 20 (table 5). 



The development of embryos was somewhat re- 

 tarded in tlie middle and lower areas. Laboratory 

 experiments have demonstrated that oxygen pri- 

 vation during early development may retard 

 growth and development of embryos witliout 

 causing death (Silver, Warren, and Doudoroff, 

 1963; Shumway, Warren, and Doudoi'off, 1964). 

 When oxygen is deficient, it is usual for hatching 

 to be delayed. This fact was confirmed in Sashin 

 Creek in 1965 where the percentage of eggs hatch- 

 ing by November 20 was 77 jtercent in tlie upper 

 area, 30 percent in tlie middle, and 11 percent in 

 tlie lower. 



The tendency of late spawners to concentrate 

 in the lower area (table 10) may liave contributed 

 to the later hatching there, but it was not a fac- 

 tor in the middle area where tlie proportion of 

 early and late spawners was the same as in the 

 upper area.- Temperature apparently was not a 

 factor either, because repeated measurements of 

 the temperature of intragravel water failed to 

 demonstrate differences among the three areas. I 

 conclude, therefore, that hatching wiis delayed in 

 the middle area because embryos had been ex- 

 posed to low concentrations of dissolved oxygen 

 in September. Tlie late hatching in the lower area 

 may have resulted from a combination of later 

 time of egg deposition and exposure to low dis- 

 solved oxygen. 



DISAPPEARANCE OF EGGS AND ALEVINS 

 FROM SPAWNING BEDS 



The disappearance of eggs and alevins of the 

 1965 brood was characterized by ( 1 ) disappearance 

 of relatively few eggs from the upper and middle 

 areas in summer during spawning, (2) disappear- 

 ance of relatively large numbers from the up^ier 

 and middle areas in autumn after spawning, and 

 (3) disappearance of many eggs from all three 

 areas in winter. 



Fewer eggs disappeared during spawning in 

 1965 tlian in 1963 even though the densities of 

 spawners were similar (table 12). In 1963, when 

 the density was about 0.6 female 2)er square meter, 

 22 percent of the potential egg deposition was esti- 



= A chi-square comparisfin of the proportion of early and late 

 spawners in the upper and middle areas (i»", degree of freedom, 

 = 0.17) demonstrated that the proportion ot early and late 

 spawners was the same in these two areas. 



PINK SALMON SPAWNERS IN SASHIN CREEK 



583 



319-171 O - 68 - 10 



