counted tagged and untagged females in each 

 area and recorded their location. Only females 

 were counted because they determine potential 

 egg deposition. 



The females were easy to count on the spawn- 

 ing ground because they remained near the site of 

 their redds from the beginning of spawning until 

 they died. I evaluated this behavior, which is typ- 

 ical of spawning females, by observing 14 tagged 

 females that were spawning in a lOO-m.-long sec- 

 tion of Sashin Creek. The locations of the 14 fe- 

 males were determined twice daily with a transit 

 and stadia and were plotted on a detailed map. 

 Tlie average longevity on the spawning ground 

 of the 14 females was 11.5 days (range 3 to 20 

 days). The average size of the area occupied was 

 3.8 m.= (1.4 m. wide by 2.7 m. long). The smallest 

 area occupied, 0.6 m.-, was for a fish that lived 

 only 3 days after establishing a site, and the larg- 

 est, 17.3 m.=, was for a fish that lived 9 days. All 

 died near their redds. 



One method of estimating the number of fe- 

 males that spawned in an area was to sum the 

 daily counts of untagged females and divide by 

 their average longevity. The daily counts were 

 summed by fitting a curve to the ninnber of un- 

 tagged females counted each day and measuring 

 the area under the curve (examples are given by 

 McXeil, 1964a and 1964b). Average longevity was 

 estimated from daily observations of taggefl fe- 

 males. One day was added to the number of days 

 individual tagged females were observed because I 

 assumed they occupied the spawning ground one- 

 half day before they were first observed and one- 

 half day after they were last observed. 



Estimates of tlie number of females in each 

 area based on summed daily counts and average 

 longevity were 2,040 in the upper area, 3,095 in 

 the middle, and 3,051 in the lower. I assumed that 

 97 percent of the total number of females spawned 

 in the three area.s, and my estimate for the whole 

 stream was 8,439 females, or 118.7 percent of the 

 number counted at the weir. The estimates for 

 each area were, therefore, adjusted by dividing 

 by 1.187. The resulting estimates were 1,719 fe- 

 males in- the upper area, 2,607 in the middle, and 

 2,570 in the lower. 



A second method of estimating the niunber of 

 females in each area was based on the occurrence 

 of tagged females. I assumed that tagged and un- 

 tagged females were distributed similarly. Of the 



Table 3. — Density of female pink salmon spawning in three 

 areas of Sashin Creek, based on observations of untagged 

 females {summed daily counts) and tagged females 



200 tagged females released, 184 (92 percent) ^ 

 were recorded in the study areas : 45 in the upper 

 area, 64 in the middle, and 75 in the lower. On the 

 further assumption that 97 percent of the females 

 counted at the weir spawned in the three areas 

 and were distributed in the same proportion as 

 the 184 tagged females, I estimated that 1,689 

 females spawned in the upper area, 2,399 in the 

 middle, and 2,813 in the lower. 



The density of females spawning in each area 

 was calculated by dividing the total number of 

 females by the area of spawning bed. The esti- 

 mates of density of females in each area by each of 

 the two methods for estimating the number of fe- 

 males agreed closely (table 3) and indicated that 

 density of spawners was about the same in all areas. 

 Although the obsei-\'ed number of tagged females 

 in each area was not significantly ditferent from 

 the expected number calculated from an assmued 

 uniform density of tagged females (table 4), the 

 conclusion that the average density of females was 

 identical among the areas is less attractive than the 

 conclusion that small differences existed. I will use, 

 therefore, the mean of the two estimates of density 

 for each area (table 3) as the best (most probable) 

 estimate of density in my calculations of potential 

 egg deposition. 



' Additional tagged females may have spawned in the study 

 areas but were not seen. 



Table 4. — Numbers of tagged female pink salmon observed 

 in three areai in Sashin Creek and the expected number, 

 based on an assiuned equal density 



PINK SALMON SPAWNERS IN SASHIN CREEK 



577 



