SPAWNING OF GALAXIAS OLIDUS IN BRUCES CREEK 
115 
collections. On 4 October 1990 and 25 October 
1990 sampling was conducted to determine 
population densities and on 11 July 1991 an ad¬ 
ditional 43 m stream section containing 5 pools 
and 4 riffles was sampled to determine pool or 
riffle preferences. Fish were captured from one 
electrofishing run, moving upstream using 2 op¬ 
erators, one with a dip net and Polaroid glasses. 
Pools and riffles were separated by stop nets and 
the wetted perimeter of each section was meas¬ 
ured and the area calculated. Fish numbers were 
adjusted using electrofisher efficiencies pro¬ 
vided in Koehn & McKenzie (1985). 
All fish sampling was conducted using a Smith 
Root Model 12 backpack electrofisher operated 
at settings of 120 Hz and 600 or 700 V. The 
length to caudal fork (LCF) of each fish was 
measured to the nearest mm and the spawning 
condition of each fish was assessed as per 
maturity stages in Pollard (1972). 
Water temperature and conductivity were re¬ 
corded during each sample using a Yellow 
Springs Model 33 Conductivity/Temperature 
meter. Relative water level was measured at a 
fixed depth gauge on each visit. The habitat at 
each spawning site was described. In determin¬ 
ing pool or riffle preferences (11 July), stream 
length, width and depth were measured along 
with observations of the location of each fish 
collected. Water velocities were measured 
100 mm above the substrate using an Ott mini 
flow meter. Substrate particles were assessed 
visually, according to the size categories of 
Cummins (1962). 
Fecundities were determined for 7 ripe fe¬ 
males collected on 15, 22, 29 August 1990 and 
for 24 females collected on 11 July 1991 by dis¬ 
secting gonads, prising eggs apart and counting 
individually. 
Egg searches were carried out in both pools 
and riffles in the stream by the following 
methods: 
H) by inspecting instream habitat such as sub- 
u rate anc * W00 ^ debris Cor attached eggs; 
(-) by using drift nets of mouth opening 35x17 
cm mesh mm (Hellawell 1986) positioned in 
the stream for (a) 3 hours and 22 hours 
where there had been no upstream disturb¬ 
ance of the streambed, and (b) several min- 
u tes, 0.5 to 2 m downstream from where 
substrate or instream cover had been gently 
disturbed (e.g. by lifting cobbles), 
ollected samples were inspected visually for 
e ^ s ' Diameters 0 f 50 oocytes 
n ol >8 fertilized eggs collected from the field, 
n engths of 16 newly hatched larvae, were 
measured to the nearest 0.01 mm using a Nikon 
Profile model 6C-2 projector at a magnification 
of lOx. 
Eggs collected in the field by the above me¬ 
thods were transferred to hatching baskets in the 
aquaria and the time taken to hatch was rec¬ 
orded. The hatching baskets consisted of 80 mm 
lengths of 90 mm diameter PVC pipe covered at 
one end with 0.5 mm nylon mesh netting. These 
baskets were suspended vertically in a 20 L 
aquarium tank with a through flow of approxi¬ 
mately 27 L/h of filtered, dechlorinated water 
(aquarium water) (see Bacher & O’Brien 1989). 
Water temperatures were maintained between 
12.3° and 14.9°C. 
Sixty larvae hatched from collected eggs were 
placed in three 4 L tanks (20 per tank) filled with 
static, aerated aquarium water kept at ambient 
temperatures of 15.0°C to 17.2°C and the time 
taken to use up their yolk sacs observed. 
The time to first feed was determined by plac¬ 
ing 20 5-day-old larvae which had used their 
yolk sac in a tank to which 250 mL of water from 
a swamp containing abundant plankton was 
added. Larvae and water were then removed to a 
beaker twice daily and any feeding activity ob¬ 
served using a stereomicroscope. Guts were 
checked for food particles. 
RESULTS 
Habitat preferences 
It was estimated that more than 90% of G. olidus 
collected in 1990 were from pools containing 
areas where water velocities were 0-0.15 m/sec, 
depths were 0.4-0.7 m, and there was abundant 
instream cover in the form of wood debris, sub¬ 
merged tree roots and undercut banks. The sub¬ 
strate of pools varied from predominantly silt/ 
clay to boulder and cobbles. Sampling of 5 pools 
and 4 riffles on 11 July 1991 showed that 87% of 
G. olidus came from pools. Calculated on a per 
area basis, 85.5% of G. olidus occurred in pools 
and only 14.5% in riffles. 
Two sections of the stream 22 m and 95 m long 
were electrofished on 4 October and 25 October 
1990. Population densities of 1.09 and 0.84 fish/ 
m 2 respectively were removed from one electro¬ 
fishing run. Koehn & McKenzie (1985) recorded 
a 64% capture efficiency for common galaxias, 
Galaxias maculatus , from one electrofishing run 
using a Smith Root MKV1A electrofisher. These 
trials were conducted on a similar species, of 
similar size, under comparable conditions, with 
equipment considered to be of similar effi- 
