Cottus Life History 
119 
Fecundity estimates were based on gravid females collected just 
prior to spawning in March 1977 and March/April 1978. Figure 3 pre- 
sents the relationship between TL and number of mature follicles for 
1977 and 1978. Correlation coefficients were 0.69 (N = 18, P 0.01) and 
0.82 (N = 47, P <C 0.001), respectively. Regression lines were calculated 
by Bartlett’s 3-group method for Model II regression (Sokal and Rohlf 
1969); 95% confidence limits for the slopes were 0.85 to 3.43 (1977) and 
2.08 to 3.28 (1978). Although the average total lengths for gravid females 
in 1977 (x = 63.9 mm, N = 18, 95% confidence limits = ±2.14 mm) and 
1978 (x = 63.2 mm, N = 47, 95% confidence limits = ± 2.00 mm) were 
similar, the average number of mature follicles in 1978 (x = 67.7, N = 47, 
95% confidence limits = ± 5.98) was significantly higher than in 1977 {X 
= 56.8, N = 18, 95% confidence limits = ± 6.34). Gravid females col- 
lected from North Indian Creek in February 1978 were similar in TL (x = 
63.2 mm, N = 53, 95% confidence limits = ±1.69 mm) to the Straight 
Creek samples and had an average follicle count of 55.5 (N = 53, 95% 
confidence limits = ± 3.62). 
Average follicle diameter in gravid females collected just prior to 
spawning in April 1978 was estimated by measuring five follicles each 
from nine different females. Average diameter was 3.32 mm (N = 45, 
95% confidence limits = ± 0.08 mm, range = 2.8-3. 8 mm). A clutch of 
eggs collected on 3 May 1978 had an average egg diameter of 3.73 mm (N 
= 10, 95% confidence limits = ± 0.05 mm, range = 3. 6-3. 8 mm). This 
clutch was incubated at 15°C and hatched on 8 May 1978. Sac fry at 
hatching had an average TL of 9.80 mm (N = 5, 95% confidence limits = 
± 0.22 mm, range =9.6-10.1 mm). 
Population density and recapture population estimates were 
calculated from N = (M±l) (C±1)/(R±1), where N = estimated pop- 
ulation, M = number of marked fish, R = number of marked fish recap- 
tured, and C = total sample of marked and unmarked fish in recapture 
sample (Ricker 1975:78, equation 3.7). Calculation of 95% confidence 
limits was based on the 95% confidence limits of R (Ricker 1975:78 and 
Append. II). Population estimates and supporting data are presented in 
Table 1 . 
Table 1. Mark and recapture population estimates of C. bairdi in Straight Creek, 
April 1979. 
Section 
Size Class 
(TL) 
M 
C 
R 
N 
95% Confidence 
Limits (N) 
#\ 
50 mm 
6 
5 
2 
14.0 
5.8-210.0 
1 
^ 50 mm 
30 
17 
9 
55.8 
32.6-139.5 
n 
<: 50 mm 
145 
57 
24 
338.7 
237.9-549.9 
2 
>*50 mm 
33 
16 
9 
57.8 
33.8-144.5 
