Case study: Carp removal in Tasmania

Carp were first detected in Tasmania in 1975 when they were found in more than 30 small farm dams on the northwest coast1. Specific noxious fish legislation was enacted in response and an eradication campaign using the fish poison rotenone was initiated. The populations in these dams were successfully eliminated. In 1995, carp were again found in Tasmania in two popular recreational trout fishing waters (lakes Sorell and Crescent), most likely introduced by people illegally using carp as live bait.

Lakes Sorell (5,300 ha) and Crescent (2,300 ha) are large, shallow, interconnecting freshwater lakes in the southeast corner of the Tasmanian Central Plateau. In addition to recreational fishing, these waters are used as a stock and domestic water supply for the downstream townships of Bothwell and Hamilton. They also supply irrigation water for farmers in these districts. The lakes contain a small endangered native fish species (Galaxias auratus) and extensive wetlands, including some of international importance (RAMSAR sites). The lakes also support a small commercial fishery for short-finned eels.

Incursion response

The Inland Fisheries Service acted quickly after finding carp in Lake Crescent in 1995:

  • Public access to the lake was closed.
  • Carp were contained by closing off water flow into the lake from Lake Sorell as well as from the lake to the Clyde River using existing structures.
  • Upstream and downstream surveys were undertaken see if carp were more widely distributed.
  • The population structure of carp within Lake Crescent was determined.
  • Contingency plans aimed at eradicating the carp were initiated.

Partners and Management

The partners involved in the carp eradication project included: Inland Fisheries Service (Tasmania), Invasive Animals Cooperative Research Centre (IA CRC), Fisheries Research and Development Corporation (population modelling project) and the National Heritage Trust (water management plans and environmental studies). Victorian and New South Wales fisheries agencies provided electrofishing boats and staff to support the survey work.


  • Once it was determined that carp were restricted to Lakes Crescent and Sorell, the following processes were put in place:
  • New containment screens and water management structures were built on the outlets of both lakes Sorell and Crescent.
  • A water management plan to protect water supplies for downstream users was developed with the Clyde Water Trust.
  • Removal of carp was begun in both lakes using all available methods.
  • The feasibility of poisoning the lakes with rotenone was evaluated.
  • Information on the habitat and biology of carp was collected.
  • Feedback was provided to the public about carp and the removal program on an ongoing basis.

Containment and water level management

Containing the carp to prevent escape or transfer of carp from the Lake Sorell and Lake Crescent populations to other areas was a priority. Effective containment was quickly achieved by placing a weir with a series of mesh screens at the outlet of Lake Crescent, the downstream lake. Mesh sizes were small enough to prevent eggs and juveniles leaving the lake. Lake levels were manipulated to reduce the chance of carp successfully spawning and to eliminate any chance of carp moving down the system through uncontrolled overflow.

The screens on the control structures became critical for management of water levels, limiting the possibility of an uncontrolled spill causing carp to spread. These structures had to be made larger than usual to effectively manage the water flow. They also required regular (around the clock) cleaning to prevent blockages.

The following construction activities were undertaken to maximise water level control and contain the carp:

  • The capacity of the canal between the two lakes was deepened and widened to allow Lake Sorell to be lowered if needed.
  • Four new control gates were installed on the outlet of Lake Sorell to increase flow and to screen for carp.
  • A levee bank was built at Lake Crescent to direct all floodwaters through a screened weir and overflow channel.
  • Four outlet bays were built in Lake Crescent with a variety of screen heights to increase the ability to respond to different lake levels.
  • The canal above the screens was narrowed to increase the speed of water flow and decrease sediment build-up.
  • A stop log structure was built at the canal entrance to reduce silt build-up in the outlet canal.

Carp population reduction

Population reduction was achieved by directly fishing numbers down and by limiting recruitment success by preventing carp access to spawning sites. A combination of fyke nets, seine nets, gillnets, traps and electrofishing (both backpack and boat) were used for fishing.

Meshed fences were used to prevent carp from accessing marshy areas favoured for spawning. Fish traps were placed in the mesh fences at some marsh drainage points. Some carp were implanted with a slow-release pheromone insert and placed in a secure cage upstream of the trap, to try and attract mature carp.

A ‘mark-recapture’ model was developed to estimate the carp population size in each of the lakes. This involved an initial survey where all carp captured were tagged and re-released. A follow-up survey was then done and an estimate of the total population was calculated from the proportion of tagged to untagged carp.

Fish tracking

To achieve total eradication, all female carp needed to be caught to prevent any chance of future spawning. Male fish were radio tagged and released back into the lakes to help locate the carp aggregations, identify habitat preferences and estimate the remaining carp population size. This is known as the ‘Judas carp’ technique.

Lake closures

Both lakes were initially closed to the public following the discovery of carp. Lake Sorell was subsequently seen as low risk and re-opened in August 1995. Lake Crescent was re-opened in 2004. In 2010, Lake Sorell was again closed due to a significant carp spawning event.


The program has had some success at eradicating carp from the target areas, although some carp still remain in Lake Sorell. The following is a brief summary of the status of carp in each lake at the end of the 2010–11 (spring/summer) fishing season.

Lake Crescent

The last female carp captured in Lake Crescent was in December 2007. The male fish with transmitter implants previously used to find any remaining females have all been removed. Although no female carp have been captured in Lake Crescent since 2007, intensive surveys for juvenile carp (including trapping, fyke netting, gill netting, electrofishing and spot rotenone treatments at known aggregation sites) have been done each year as a precaution. Recently, additional survey support has been provided by a commercial eel fisherman using a large number of fyke nets.
This collective effort has yielded no carp, suggesting that if any carp remain in Lake Crescent, they are few and far between, unable to successfully spawn and therefore not a viable population.

Lake Sorell

Spawning events occurred in Lake Sorell in 1996, 1998, 2000, 2003 and 2005, however most of these carp were removed. Juvenile carp were again found in Lake Sorell in December 2009. A concerted effort was made to remove as many of these juveniles as possible before they moved from marsh areas to deeper water. Over 14,000 fish were captured in six weeks. The remaining fish were actively targeted during 2009–10 and a further 9000 carp were removed. Estimates indicate over 90% of the carp have now been removed from Lake Sorell.

The ‘Judas’ carp technique has been very successful and provided extensive knowledge of carp behaviour and movement through a range of environmental conditions. It has also allowed refinement of a range of carp capture techniques.

Currently, a sterilisation technique is being trialed. All carp implanted with radio transmitters to be released into Lake Sorell have undergone this procedure to minimise the risk of unwanted spawning.

Efforts to prevent any further spawning events have intensified. Despite ideal spawning conditions in 2010–11, it appears that spawning was successfully prevented by blocking and capturing 12 mature females.

What worked and why

Extensive knowledge gained through this control program led to greater confidence in managing carp at a variety of water levels. This enabled the lakes to be maintained at higher water levels and ensured a water supply for downstream users.

The release of radio-tagged male carp that can be targeted when they aggregate proved to be a successful Judas technique. It increased knowledge of carp movement and behaviour under the local conditions and greatly assisted with the eradication of carp from Lake Crescent. A potential refinement to this technique may be remote monitoring so fish could be located without physically tracking them from a boat.

The use of fish traps and exclusion fences to block carp access to spawning areas was highly effective, particularly when used with pheromone attractants and positioned to take advantage of environmental stimuli.

The population estimate method using the mark-recapture model also proved to be a valuable means of tracking progress. Data collected on carp numbers and movement showed the variability of behaviours in the carp populations and has important implications for ongoing management. This model will also allow generalisations to be made for other lakes.

Netting techniques were refined for particular situations and growth stages. For example, seine nets were used for adult feeding aggregations. Fyke nets were found to be very efficient for assessing population size and fishing for juveniles and passive capture of adult fish during spawning. Gillnets of various sizes (when used in combination with radio tracking and electrofishing) were very effective in capturing carp during different life stages.

The Inland Fisheries Service has worked hard to keep the public informed of progress throughout, and the project has enjoyed strong public support since its inception. This has no doubt helped with ongoing funding.

What didn’t work and why

An initial goal was to test the extent to which ‘Judas’ carp could contribute to the eradication of female carp from Lake Crescent, and to transfer this technique to Lake Sorell. However, some problems occurred early on in Lake Crescent:

  • Despite being able to actively follow and monitor carp to spawning aggregations, it was difficult to prevent eggs being released and fertilised.
  • The last spawning event in November 2000 in the Clyde Marshes and on debris in the canal joining Lake Crescent to Lake Sorell led to juveniles entering Lake Crescent (carp are very opportunistic).

Before the 2009 recruitment event in Lake Sorell, it was thought there were less than 50 remaining adult male carp. There was concern that the virile transmitter fish used to target the population might have actually fertilised carp eggs. This is a serious issue when numbers are low and the objective is eradication. To minimise the risk of further recruitment during subsequent seasons, all transmitter fish needed to be sterilised or removed from Lake Sorell before the peak spawning period.

The final few carp appear to be increasingly difficult to catch with current fishing methods (potentially due to learned behaviour). More effective fishing techniques will need to be developed — this is an important area for future research.


The control of carp in these two Tasmanian lakes is a good example of what scientists call ‘adaptive management’ or ‘learning by doing’. Control began in 1995 and is still ongoing. The cost of removing carp so far (as at the end of 2013) has been estimated at $9.6 million. This highlights the importance of preventing the introduction of carp in the first place, and of containment as a control strategy.

The knowledge, skills and infrastructure assembled since control efforts began in 1995 have enabled carp eradication in Lake Crescent. However, physical removal of the last remaining carp in that lake was challenging due to their reduced catchability. The eradication of carp from Lake Sorell has been even more difficult because of the larger size of the lake, its wetland spread and response to rainfall, and the effect of several spawning events.

Unless carp can be stimulated to form spawning aggregations, it may prove difficult to remove the last few carp from Lake Sorell. The use of pheromone attractants5 combined with strategically located traps could enable the last remaining fish to be caught.

Whether or not carp can be eradicated from Lake Sorell will depend on the rate at which the population can be fished down, and the number of years for which spawning can be prevented. Because the final carp might only be catchable during a spawning aggregation, the removal of these last fish is likely to prove both difficult and expensive.

Author Centre for Invasive Species Solutions
Year 2014
Publisher Centre for Invasive Species Solutions
Pages 4
ISBN/ISSN PestSmart code: CPCS5
Control method Integrated Pest Management
Region TAS
Links PestSmart Toolkit carp page
Documents CPCS5 PestSmart Case Study: Carp removal in Tasmania   [470kb PDF]