LIFE HISTORY OF BLACKFISH 
95 
Australian streams (Tilzey 1976, Jackson & 
Williams 1980). The presence of S. tnitta does 
not appear to affect the abundance of G. bispi¬ 
nosus in the present study. Indeed, the recent 
study by Koehn (1987) suggests that the amount 
of instream cover is a more serious limiting fac¬ 
tor. Similarly, G. marmoratus has been found to 
cope adequately with the competitive pressure 
of coexistence with S. tnitta in some Victorian 
streams (Jackson 1978b, Jackson & Williams 
1980), and the abundance of G. marmoratus in 
some Victorian and Tasmanian streams has 
been shown to be correlated more closely with 
instream cover and low flows than with any 
other habitat attributes (Cadwallader 1979, 
Koehn 1986, Davies 1989). 
The timing of recruitment of G. bispinosus 
appears to be similar to that of G. marmoratus 
found by Jackson (1978a) and Koehn (unpub¬ 
lished data) in some Victorian streams. In both 
those studies spawning activity was found 
between mid-November and mid-December. 
Jackson (1978) concluded from aquarium obser¬ 
vations that the time from spawning to active 
feeding was about eight weeks, and suggested 
that recruitment should have occurred in Jan¬ 
uary and February in the McKenzie River in 
1976. Recruits were first detected in the King 
Parrot Creek population of G. bispinosus in 
February' 1981 (meanTL 39.9 mm) and January 
1982 (mean TL 27.1 mm). The water tempera¬ 
tures in King Parrot Creek from November to 
February' are within the range reported by Jack- 
son (1978a) for the McKenzie River, and by 
Koehn (unpublished data) for Armstrong Creek. 
Although in the present study the spawning per¬ 
iod of G. bispinosus has been determined indi¬ 
rectly from ovarian condition and from preda¬ 
tion on eggs, it is similar to that reported for G. 
marmoratus and shows a similar association 
with increasing water temperatures. The maxi¬ 
mum diameter of ovarian eggs in (7. bispinosus 
(3.5-4.0 mm) and of fertilized eggs in G . mar¬ 
moratus, and the correlation of fecundity with 
total length are also similar (Jackson 1978a, 
Koehn unpublished data). The relationship 
between total length and month of maturity has 
not been reported previously for species of Gad- 
opsis. 
At the taxonomic level employed in the pre¬ 
sent study, the diet of G. bispinosus is very simi¬ 
lar to that of G. marmoratus in the Aberfeldy 
River, as described by Jackson (1978b). Terres'- 
trial insects were more important than dipteran 
larvae in the diet of G. marmoratus in the Aber- 
ieldy River, but the influence of different sample 
dates, the size range of fish used and the avail¬ 
ability of the different prey types could all have 
influenced this result. The presence of 
freshwater crayfish, a galaxiid fish and fish eggs 
also distinguish the diet of G. bispinosus from 
that of G. marmoratus in the Aberfeldy River, 
although these items do not comprise a major 
part of the diet of G. bispinosus. The seasonal 
composition of the diet of G. bispinosus was rela¬ 
tively stable, no distinct seasonal peaks in rank 
values for any food category being evident in the 
data. The seasonal composition of the diet of G. 
marmoratus in the Aberfeldy River was also 
relatively stable. 
The concurrence of life history characteristics 
of G. bispinosus and G. marmoratus should en¬ 
able the formulation of overall guidelines for the 
management of wild populations of Gadopsis 
species. The importance of instream cover to 
both species has already been demonstrated 
(Jackson 1978a, b; Koehn 1986, 1987; Davies 
1989). Other features which appear to be 
important are a spring increase in water temper¬ 
ature which correlates with the onset of 
spawning, and the heavy reliance on benthic 
invertebrates in the diet of both species. Envi¬ 
ronmental degradation, including an artificial 
lowering of spring and early summer water tem¬ 
peratures through release of cold impoundment 
water, a decrease in the amount of instream 
cover either from de-snagging or siltation, and 
disruption of the benthic invertebrate fauna 
through substrate disturbance or the introduc¬ 
tion of toxins into streams, could thus be 
expected to have a detrimental effect on stream 
populations of Gadopsis species. 
ACKNOWLEDGEMENTS 
I thank my supervisor, Murray Littlejohn, and 
others in the Department of Zoology at the Uni¬ 
versity of Melbourne who helped with this pro¬ 
ject. The project was supported by a Com¬ 
monwealth Postgraduate Research Award, the 
Victorian Fisheries and Wildlife Division, and 
the Australian Museum Trust. John Koehn 
made valuable comments on the manuscript, as 
did the referees. The Department of Zoology, 
University of Adelaide, and the Tasmanian 
Inland Fisheries Commission are thanked for 
the use of facilities. 
REFERENCES 
Cadwallader, P. L., 1979. Distribution of native 
and freshwater fish in the Seven Creeks river 
