Table 1. — Size and location of six spawning riffles sampled 

 in 1956, 1957, and 1958 and five spawning areas sampled 

 in 1958, 1959, and 1960 



Study area and years sampled 



1956, 1957, 1958 



Rime A, Harris River 



Riffle B, Indian Creek 



Riffle C, Indian Creek 



Riffle D, Twelvemile Creek. 

 Riffle E, Twelvemile Creek.. 

 Riffle F, Twelvemile Creek. 



1958, 1959, 1960 



Intertidal Harris River 



Upstream Harris River 



Intertidal Indian Creek 



Intertidal Twelvemile Creek.. 

 Upstream Twelvemile Creek. 



Tide level 



Meters 



3.4 

 4.0 

 5.2 

 3.7 

 4.3 

 4.9 



3. 7-1. 3 



3. 4-5. 2 

 3. 7-1. 9 



Area 



Meters 



305 

 386 

 260 

 648 

 372 

 486 



7,800 

 13, 400 

 3,400 

 5,580 

 6,130 



and distribution of spawners, mortality of eggs 

 and larvae, quality of spawning beds, and quality 

 of intragravel water. 



To insure random sampling, FRI personnel drew 

 maps of areas sampled to scale on cross-sectional 

 paper having 100 squares per square inch. The 

 scale selected made each square representative 

 of not more than a 0.4 m. 2 area in the stream. 

 Co-ordinate axes were established for each map, 

 and sampling points were selected by the following 

 procedure. A pair of random numbers was 

 obtained from a random number table — one 

 number for the abscissa and the other number for 

 the ordinate. Distance of the selected square 

 from a reference point and an angle of the selected 

 square from a reference line were measured on the 

 map and recorded for use in the field. When a 

 sample was to be taken, the angle and distance 

 of the sample area were measured from the base 

 line and reference point. 



MEASUREMENT OF ENVIRONMENTAL FACTORS 



The Northern Forest Experiment Station, U.S. 

 Forest Service, operated a weather station at 

 Hollis, a logging community located within 3 

 miles of Harris River and Indian Creek spawning 

 areas and within 12 miles of Twelvemile Creek 

 spawning areas (see fig. 4). Forest Service per- 

 sonnel obtained continuous records of air tem- 

 perature and precipitation during the study. 

 Instruments installed and operated by the North- 

 ern Forest Experiment Station recorded water 

 level and temperature of each study stream. 

 Forest Service personnel also established discharge 

 rating curves for each stream. James (1956) gives 



further information on hydrological studies made 

 by the Forest Service. 



Physical quality of the spawning bed was 

 evaluated in each area where observations were 

 made on egg and larval mortality. Environ- 

 mental attributes measured included the dissolved 

 oxygen content of intragravel water and the size 

 composition and permeability of bottom materials. 

 McNeil (1962b) described the techniques used to 

 measure dissolved oxygen levels and McNeil and 

 Ahnell (1964) described the methods of measuring 

 size composition and indexing permeability of 

 bottom materials. 



ESTIMATION OF SPAWNING DENSITY AND 

 POTENTIAL EGG DEPOSITION 



Fisheries Research Institute (FRI) personnel 

 estimated the number of female pink and chum 

 salmon occupying study areas by means of daily 

 foot survey censuses when water conditions per- 

 mitted. Institute personnel also observed tagged 

 females daily to estimate average life on the 

 spawning ground (redd life) . Females were tagged 

 before they entered the spawning ground and FRI 

 workers calculated the total number spawning by 

 summing daily abundance and dividing by average 



redd life; i.e., 



2]daily abundance in female 

 days 



Number spawmng= average redd i ife in days 



A summation of daily abundance was obtained 

 by constructing an eye-fitted curve of the daily 

 counts of females and determining the area under 

 the curve in female days. In the determination 

 of average redd life, daily observations on tagged 

 females have been adjusted to account for the 

 periodicity of observations. Assuming that each 

 tagged female occupied the spawning bed one-half 

 day before being observed the first time and one- 

 half day beyond the date of the last observation, 

 I added 1 day to the duration each tagged female 

 was observed. 



Two observers made most of the survey censuses; 

 periodic comparisons of their counts showed con- 

 sistently good agreement. Although it was not 

 feasible to determine bias in estimates, a recent 

 unpublished study conducted by the author at 

 Sashin Creek revealed good agreement between 

 the number of spawning female pink salmon esti- 

 mated by this method and the number counted 

 into the stream. 



SPAWNING BED ENVIRONMENT OF PINK AND CHUM SALMON 



503 



