Table 4-. — Number of recoveries from 1,142 pink 

 salmon tagged in envuneration study, Olsen 

 Bay, 1961 



than the weir count. It should be kept in mind, 

 however, that the weirs were taken out several 

 times because of floods, and the counts repre- 

 sent only minimum numbers. 



The intertidal estimates of 100,147 pink 

 salmon represent 74 percent of the total 

 escapement at Olsen Creek in 1961. 



In 1960, 72,617 pink salnnon were estimated 

 to have spawned in the intertidal zone, repre- 

 senting an identical 74 percent of the total 

 migration. The 1960 estimate was obtained by 

 taking the maximum live count of 30,395 on 

 Septennber 4 and adding to it a figure of 

 42,222, which represents the accunnulated dead 

 count to that date. This is a minimum figure 

 because a substantial number of carcasses 

 may have been removed by bears, and an un- 

 known number were washed out of the inter- 

 tidal zone by high water, coincident with peak 

 counts of dead fish. Also, emigrants which 

 may have entered the stream after the nnaxi- 

 mum live counts have not been included. It is 

 therefore concluded that the number of pink 

 salmon that spawned in the intertidal zone in 

 1961 was probably more than 74 percent of 

 the total spawning population. 



Analysis of the tagging data demonstrated 

 that population estimates of sufficient accuracy 

 could be obtained by sample tag ratios; con- 

 sequently, the use of weirs, which are more 

 costly in time and money and which at best 

 only give estimates at Olsen Creek, will be 

 discontinued. 



Duration of Intertidal Stream Life 



To determine their length of life after they 

 entered the streann, pink salmon in 1961 were 

 tagged at the stream mouth and were sub- 

 sequently recovered dead on daily surveys of 

 the intertidal area (table 5). Although the 

 ranges of the duration of stream life overlap. 



the means exhibit a definite trend (fig. 12). 

 A progressive seasonal reduction in stream 

 life is evident, ranging from over a month 

 early in the season to about 5 days at the end 

 of the season. The experiments on July 24, 

 August 1, August 6, and August 9 deviate 

 fronn the trend; however, pink saln-ion tagged 

 during these experinrients left Olsen Creek 

 and were not observed again in the stream 

 until 3 or 4 weeks later. 



To nneasure the relation between length of 

 stream life and season, the data were tested 

 by a regression analysis (Snedecor, 1956). 

 The four aberrant experiments mentioned 

 above were omitted. A negative regression is 

 demonstrated (fig. 13), the regression coef- 

 ficient being (b = -0.049) which is significant 

 at the 90-percent level (t = -1.70). The mean 

 duration of streann life for the entire season 

 was 11.1 days. 



SEPARATION OF RACES 



The Olsen Bay connplex offers several 

 possibilities for detecting the existence of 

 pink salnnon races, if they are present. The 

 necessary degree of genetic separation could 

 conceivably occur between early and late 

 spawners or between stocks consistently 

 homing to specific areas such as the East 

 Fork and the West Fork upstream areas or 

 the Little Creek, Middle Slough, and Olsen 

 Creek intertidal areas (fig. 1). Preliminary 

 attennpts to discern racial identities of Olsen 

 Bay spawners involved studies of size and 

 sex ratios. 



Size Differences 



Size differences associated with certain 

 portions of the Olsen Bay complex or tinne 

 of spawning season could be indicative of nnore 

 than one race of pink salmon or could be the 

 result of varying oceanic histories of more 

 or less discrete groups of nnigrants. In either 

 case, knowledge of such groups is of value to 

 research and managennent of pink salmon. 



Length measurements were taken from nnid- 

 eye to the posterior end of the hypural plate 

 (fig. 14) of 2,549 adult pink salmon in 1960 

 and 3,117 in 1961. Randomly collected fresh 

 carcasses from each weir and from the main 

 streann intertidal area were nneasured at 

 weekly intervals. 



For convenience in analyzing size nneasure- 

 ments with respect to time, adults nneasured 

 on August 13 or earlier were ternned "early" 

 spawners and those nneasured on August 14 or 

 later were termed "late" spawners. This time 

 break coincided with observed changes in 

 timing of the runs and spatial distribution of 

 the spawning adults. 



Because of certain differences in size char- 

 acteristics of nnales and fennales, length 



A4 



