the major cause of declining abundance in 

 the gravel between March and July 1962 was 

 clearly determined to be the emergence and 

 migration of fry. 



Dea d Eggs and Larvae . — Although the abund- 

 ance of dead 1961 brood year eggs declined 

 after hatching, sizable numbers of dead eggs 

 were present in spawning beds 18 months 

 after spawning (table 2). 



About 14 percent of the eggs collected in 

 the October 25, 1961, samples were classified 

 as dead, but no fragments were recorded. All 

 pink and chum salmon eggs remaining in March 

 1962 were dead because hatching had termi- 

 nated at an earlier date. During the period 

 October 1961 to March 1962, the abundance 

 of dead eggs increased from an estimated 

 mean of 156 to 420 per square meter. By mid- 

 July 1962 the mean density of dead eggs had 

 declined to an estimated 159 per square meter, 

 and in August it declined further to 98 per 

 square meter. Between August and November 

 1962, the declining trend reversed itself; and 

 mean egg densities increased to 153 per 

 square meter, the difference being significant 

 at about the 10-percent level of probability. 



Two factors could have accounted for this ap- 

 parent increase: (1) Early mortality of 1962 

 brood year chum salmon eggs which potentially 

 added 20 eggs per square meter to the stream- 

 bed and (2) disintegration of 1961 brood year 

 eggs which increased the numbers of fragments 

 counted and biased our estimates upward. 

 In March 1963 the estimated mean density of 

 dead eggs declined to 70 per square meter. 

 Of these, 18 per square meter were whole 

 eggs and were thought to be 1962 brood year 

 eggs. 



Dead larvae disappeared rapidly in com- 

 parison with eggs (table 2). In late March 

 1962, the estimated mean density of dead 1961 

 brood year larvae was 104 per square meter. 

 By early July this number had declined to 

 31 per square meter. In mid-August 1962 no 

 remains of larvae were found. 



The estimated average abundance of dead 

 1961 brood year eggs and larvae (table 2) is 

 plotted in figure 2. The curves are sketched 

 in by eye. The curve showing changes in 

 abundance of dead 1961 brood year eggs was 

 drawn by assuming a zero point coincident 



Table 2 . --Estimated abundance of dead 196l brood year eggs and 

 larvae in the Sashin Creek study area 



— About 20 dead eggs per square meter were thought to be of the 1962 

 brood year. 



