FISHERY BULLETIN: VOL. 74, NO. 4 



Survival of Juveniles in Sashin Creek 



We estimated the population periodically during 

 the summers from 1964 through 1967 to establish 

 curves depicting changes in the number of 

 juvenile coho salmon by brood year during their 

 freshwater life. In 1964 the numbers of fry were 

 estimated in July and August. In 1965 the 

 numbers of fingerlings (predominantly age I, the 

 balance age II) and fry were estimated in July and 

 August. In 1966 the numbers of fingerlings and fry 

 were estimated in June, July, and September. In 

 1967 estimates were made of coho salmon 

 fingerlings and fry in Sashin Creek in July and in 

 Funny Creek in July and August. 



Juvenile salmon in the stream were captured by 

 a combination of baiting and seining. A 

 homogenized mixture of salmon eggs, ovarian 

 tissue, and water was prepared with an electric 

 blender and injected into the stream at the seining 

 site (Figure 9). Underwater observations indicated 

 that several squirts of the egg solution from a 

 plastic squeeze bottle were adequate to attract 

 rainbow trout, Dolly Varden, coastrange sculpins, 

 and coho salmon fingerlings and large fry from at 

 least 30 m downstream. The downstream sides of 

 gravel bars, logs, and rocks were chosen as col- 

 lecting sites because these obstructions formed 

 slow-water areas in which the bait would linger for 

 several minutes. In some instances it was neces- 

 sary to construct a rock barrier to divert the 

 current and create a suitable site. During early 

 summer, when coho salmon fry are quite small, 

 they congregate along the shallow edges of the 

 stream and in backwaters. These small fry will not 

 travel far in response to bait, and we often had to 

 seine for them along the stream edges and back- 

 waters near the baiting site. 



Captured fish were anesthetized with MS-222 

 Sandoz"' and marked by removing part of one fin. A 

 different fin clip was used for each marking date 

 within a summer. When they recovered from the 

 anesthetic, the marked fish were released at the 

 collection site. 



To allow the marked juveniles to become redis- 

 tributed, we did not begin to recapture them until 

 several days after they were marked. To reduce 

 bias in the population estimate, we selected ran- 

 dom points as seining sites during the recapture 

 portion of the e.xperiment. Random numbers 

 between and 99 were chosen from a table of 



^Reference to trade names does not imply endorsement by the 

 National Marine Fisheries Service, NOAA. 



random numbers (Snedecor 1956) for each of the 

 30.5-m (100- foot) sections of stream. The numbers 

 chosen represented the distance in feet down- 

 stream from the origin of each section to the sites 

 that would be baited. These distances were paced 

 off, and one or several places across the width of 

 the stream at the site were baited. A site was 

 repeatedly baited and seined until only a few fish 

 could be taken in each seine haul. All fish captured 

 at a site were anesthetized and examined for 

 marks. When they recovered from the anesthetic, 

 the fish were released. The numbers of unmarked 

 and marked coho salmon juveniles were recorded 

 for each site. 



The Bailey-Petersen mark-and-recapture 

 method (Ricker 1958) was used to make all 

 population estimates, e.xcept in August 1964 when 

 a Schnable multiple mark-and-recapture method 

 (Ricker 1958) was used for fry. In 1966 and 1967 the 

 numbers of juveniles to be marked and recaptured 

 were predetermined to obtain preassigned levels 

 of accuracy and precision of population estimates 

 (Robson and Regier 1964). We tried to mark and 

 recapture enough fish to be 95'>? certain that the 

 error in estimating the population was not more 

 than lO'^r (Table 15). Confidence limits to popula- 

 tion estimates were obtained by methods given by 

 Ricker (1958) and Robson and Regier (1968). 



The number of coho salmon fry decreased 

 greatly between the first and second estimate 

 (Tables 15, 16). In the month between estimates, 

 the population dropped by 71^^ in 1964, 78*^ in 1965, 

 and 62*^ in 1966. Weir counts of emigrant coho 

 salmon fry, which were continued until mid- 

 August in 1964, showed that only about 2,000 fry 

 (4*^ of the first population estimate) left Sashin 

 Creek between mid-July and mid-August popula- 

 tion estimates. Fyke net catches indicated that 

 even fewer fry migrated from the stream in 1965 

 and 1966 than in 1964. Therefore, we attribute the 

 large decrease in number of fry each year to 

 mortality rather than emigration. Observation of 

 the activities of fish and avian predators led us to 

 believe that predation probably accounted for the 

 major portion of the mortality. The number of 

 fingerling coho salmon also decreased as the 

 summer progressed, although not as rapidly as the 

 number of fry (Table 17). 



The fry population was greater in 1964 than in 

 1965, 1966, or 1967 as a result of the large number 

 of spawners entering Sashin Creek in the fall of 

 1963. The population of fingerlings in 1965 was also 

 greater than in 1966 or 1967; the fingerlings in 1965 



914 



