Summary

Introduced carp are widely distributed in NSW.  They dominate many of the fish communities within their range, particularly in the Murray-Darling Basin.  The successful establishment and spread of carp is aided by their tolerance to a wide range of environmental conditions, good mobility and high reproductive output.  These characteristics make carp persistent pests that are difficult to eradicate.  However, there are a number of possible ways to reduce carp numbers including biological control, habitat modification, physical control such as harvesting and chemical control.  Controlling carp requires a detailed assessment of the effectiveness and short-falls of alternative techniques.

This report documents the findings of three separate studies.  The first study assessed different fishing gear types and provided information on the most cost-efficient and effective harvest techniques, particularly those suited for use by community groups wishing to remove carp from their local waterways.  The second study evaluated the effectiveness of poison baits (rotenone) on carp eradication and the impacts of the baits on non-target native fish.  The third study investigated ecological differences between large and small carp in enclosed waterways.

The effectiveness of four fishing methods was trialed in the Menindee Lakes in Western NSW.  Small mesh fyke nets caught the most carp from the widest range of size classes, were the easiest to use, caused least mortality of by-catch and were the second most cost effective gear type.  Therefore, they are recommended as the most suitable gear type for community groups to use for harvest of carp from local waterbodies.

Floating pellet baits containing a lethal dose of rotenone were tested in billabongs, hatchery ponds and aquaria.  Rotenone pellets in billabongs killed only 12 carp from an estimated population of over 2,500, but a large number of Australian smelt were killed in one billabong.  In aquarium experiments, Australian smelt were the only species that ate the pellet baits, suffering 95% mortality.  Other native fish species suffered high mortality if uneaten pellets were left in the water, suggesting that leaching of the poison from uneaten pellets can harm non-target fish species.  In the pond experiment, substantial numbers of bony herring (39%) but very few carp (5%) died after application of rotenone pellets.  Based on these results, pellet baits would require further development and extensive additional testing before they could be used on carp in NSW waterways.

A previously proposed recruitment model suggests that success of juvenile carp is greatest at low to intermediate densities of adult fish.  Therefore, control techniques that only target adult carp may result in improved survival of juvenile fish.  It may not be desirable to replace a given number of large adult carp with potentially many more smaller ones.  A field experiment was set up to examine growth and survival of small carp in the presence of varying densities of large carp.  Unfortunately, the system of isolating the different treatments in each billabong was not effective and individual carp moved between them, sometimes by jumping between pens, and sometimes by burrowing under the mesh walls.  Future attempts at such an experiment should use solid sheeting rather than mesh, construct double fencing to create effective buffer areas between enclosures and ensure that carp do not jump the fences.