Non Technical Summary

Fifty-five freshwater fish species inhabit the river of NSW Rivers, at least 36 of them have a requirement to migrate at some stage of their life. Adult native fish have been previously observed undertaking large-scale upstream migrations of up to 1,400km whilst juveniles also perform mass upstream migrations. The presence of dams and weirs often prevent successful migration by blocking pathways to spawning habitats and also feeding areas. Since 1913, over 70 fishways have been constructed in NSW to help mitigate these effects.

A cost-effective fishway design that has considerable potential for Australian systems is the Deelder fishlock, which was first constructed on the Muese River, Belgium, in 1958. The Deelder fishlock operates in a similar fashion to a navigation lock for boats and can be best described as a lock incorporating two chambers divided by an internal weir. This report describes the operation of Australia's first (and the worlds only currently operating) Deelder fishlock that was constructed at Balranald Weir on the Murrumbidgee River, New South Wales in December 2002 and the detailed assessment that was done to determine its utility for wider application in the Murray-Darling Basin.

This study assessed the effectiveness of a Deelder fishlock at providing passage for Australian native inland fish. In 18 paired entrance and exit samples, over 3 different cycle times, a total of 7,939 fish from 9 species were sampled within the fishlock. Australian smelt, bony bream and western carp gudgeon contributing to 94% of the total catch. The maximum number of fish to pass through the fishlock in one day was 854 and the daily average was 360 (equivalent to 22 fish per hour).

The fishlock was most successful at providing passage for Australian smelt, crimson spotted rainbowfish, golden perch, the threatened silver perch and juvenile bony bream. These species were sampled in greater or equal abundance from the fishlock exit compared to the entrance. However, higher abundances of flyspecked hardhead and western carp gudgeon from the entrance suggest that these two species did not effectively use the fishlock.

The fishlock did not prevent fish of any particular size class from gaining passage and fork lengths of fish sampled from the exit trap ranged from 12-540mm. When the mesh size on the fishway trap (which was used to test for fish passage) was reduced from 10mm to 1mm, capture rates of Australian smelt, bony bream, flyspecked hardyhead, western carp gudgeon and crimson-spotted rainbowfish increased dramatically. This suggests that larger mesh sizes give an underestimate of passage by small fish because of significant escapement through the mesh. The highest levels of fish passage occurred between midday and 4pm, and few fish migrated at night.

This study proved that the Deelder fishlock was extremely effective at providing passage for Australian native fish under low-flow conditions. However, we now need to repeat the experiment under high flow conditions to test whether this design is applicable for wider application in New South Wales streams.