116 
W. G. O’CONNOR AND J. D. KOEHN 
ciency. The population densities for G. olidus 
were thus adjusted by a factor of 1.56 to give 
calculated population densities of 1.70 and 1.31 
fish/m 2 respectively. 
Population structure 
In the length-frequency histogram for the G. 
olidus population (Fig. 2), four size classes were 
recognised (Table 1). Following Fletcher (1979) 
these size classes were assigned estimated age 
classes of 04-, 1 4-, 24- and 3+ years, with 0 + 
being assigned to juveniles from the 1989 
spawning. In the 263 fish captured, the size 
ranged from 36 mm to 100 mm LCF, the largest 
fish being ripe females. 
Further analysis of the age classes (Table 1) 
shows that the population was dominated by the 
1 4- age class, although the smaller 04- age class 
was likely to be under-represented due to less 
efficient sampling of smaller fish (Koehn & 
McKenzie 1985). Only 2.3% of the population 
were 3 years or older, indicating that few fish live 
beyond their third year. 
Details of sexual maturity for each age-class 
are given in Table 1. Most G. olidus do not reach 
sexual maturity in their first year (04-) when 
only 20% of male and 5% of female fish were 
sexually mature. The smallest sexually mature 
male was 42 mm (LCF) and the smallest female 
was 47 mm LCF. All fish were sexually mature 
by their second year (14-). 
The overall sex ratio of mature fish in the 
population was 1 : 1 but the ratio of sexually 
mature individuals was not consistent through¬ 
out the age-classes. The earlier maturity of some 
04- male fish increased the apparent male : fe¬ 
male ratio in that class. The 14- class showed a 
1.1:1 ratio whereas the 2 4- class showed a 1:4.0 
ratio. All six fish in the 34- class were female. 
The 31 ripe female G. olidus analysed for fec¬ 
undity ranged between 55 mm and 88 mm LCF 
(mean 66 mm) and contained from 44 to 384 
eggs (mean 198, N = 31, SE = 15). Number of 
eggs of G. olidus is positively correlated to fish 
length (r = 0.836, p < 0.01) (Fig. 3). 
Spawning season 
The spawning season for G. olidus in Bruces 
Creek in 1990 extended from early August to 
about late October. The percentage of spent fish 
in each sample increased from 10% to 93% dur¬ 
ing the study (Fig. 4). Although spawning had 
already begun by the start of the study, this sam- 
Body Length (mm) 
Fig. 2. Length-frequency histogram for Galaxias olidus in Bruces Creek in 1990. 
