Table 10. — Chi-square test of significance of difference between proportion of 

 samples with live and dead pink salmon eggs and alevins in Middle and Lower 

 sections of Sashin Creek, September 1966 and March 1967 



Section and date of sampling 



Samples 



Samples with three or less- 



Live plus dead Live 



Dead 



Number 



Number 



Number Number 



Middle 



September 25, 1966. 



March 10, 1967 



X 2 



Lower 



September 25, 1966. 



March 10, 1967 



X 2 



2k p _ p Sk 



X 2 = ° ° ° ° ; k = 



p(l-p) 

 o o 



number of samples with three or less pink 

 salmon, P = proportion of samples with three 

 or less pink salmon, p_ = mean of p Q values. 



with live plus dead, live only, and dead 

 only were significant. I used the method and 

 notation of McNeil (1964) and also used his 

 designation of categories, i.e., samples with 

 "three or less" and "more than three" pink 

 salmon eggs or alevins. I tested combinations 

 of three or less combined live and dead, three 

 or less live, and three or less dead (table 10). 

 The test showed (1) a significant increase in 

 the proportion of samples with three or less 

 live eggs or alevins in the Middle section; 

 (2) a significant decrease in the proportion 

 of samples with three or less dead eggs or 

 alevins in the Middle section; (3) a significant 

 decrease in the proportion of samples with 

 three or less dead eggs or alevins in the 

 Lower section; and (4) no significant change 

 in the proportion of samples with three or 

 less combined live and dead eggs or alevins 

 in either section. 



I conclude that mortality of pink salmon 

 eggs or alevins was greater in the Middle 

 than in the Lower section but can offer no 

 reason for the difference. 



The population of the 1966 brood year from 

 number of eggs to number of preemergent 

 fry for each area and the entire stream is 

 shown in table 9. I used the average survival 

 (weighted by surface area) of the Middle and 

 Lower sections to calculate the number of 

 live pink salmon alevins in the Upper section 

 and the remainder of the stream. Although 

 the fry weir was not operated in the spring 

 of 1967, the number of fry counted through 

 the weir had been very close to the number 

 of preemergent fry or alevins estimated in 

 the gravel by hydraulic sampling (McNeil, 

 1968). I therefore assumed that the production 



of pink salmon fry in Sashin Creek in the 

 spring of 1967 was the stream population of 

 alevins in March--755,000. This figure repre- 

 sents 11.9 percent of the 6,255,000 eggs carried 

 into the stream by females in August and 

 September 1966. 



I calculated the regression of total fresh- 

 water survival against the date on which 50 

 percent of the escapement had entered Sashin 

 Creek for escapements of 1,000 or more for 

 17 years from 1940 through 1965. McNeil 

 (1968) presented a similar regression, but 

 he included data from smaller escapements. 

 My empirical estimate of total fresh-water 

 survival for the 1966 brood year is the same 

 as the value calculated from the date of entry 

 of 50 percent of the escapement (the calculated 

 survival for August 31 is 1 1.9 percent--fig. 4). 

 Thus, the total fresh-water survival of the 

 transplanted stock in 1966-67 was the same 

 as that expected for an escapement of the 

 native stock with the same midentry date. 



IDENTITY OF THE 1966 ESCAPEMENT 



An underlying assumption of this paper is 

 that the adult pink salmon that entered Sashin 

 Creek in 1966 were mainly progeny of the 

 adults transplanted to Sashin Creek from 

 Bear Harbor in 1964. It is not possible to 

 state positively that this assumption is true, 

 but three types of circumstantial evidence 

 support it. This evidence centers on (l)the 

 amount of straying observed in earlier years, 

 (2) the fresh- water survival of progeny of 

 strays, and (3) the similarities between the 

 transplanted adults and the adults of the 

 1966 escapement. 



11 



