Control methods

The type of control or ways of managing aquatic weeds can be conveniently grouped into the following:


Cutting and harvesting

Weed cutting costs can be low. It usually does not require a consent and it can target a specified area and cut to a nominated depth though this is usually only possible to a maximum of <2.0 m below the surface.

Weed harvesting (cutting and removal) can be beneficial if it removes significant amounts of nutrients from the system.

Weed harvesting – cutting, collection and dumping – is often priced around $2000–4000 per hectare. The price of operating a weed harvester is usually charged by an hourly rate for the machine and so costs vary considerably with the density of weed and the distance to a dump site.

Clear water to view cutting lines is often essential for efficient cutting, unless carried out in a drain. Weed cutting is like mowing the lawns as it may have to be repeated two or three times in a growing season when re-growth occurs.

Some sites are unsuited to cutting because of uneven bottom contours, obstacles or high flows and the weed harvester is also affected by wind. Weed harvesting also results in capture of a wide range of aquatic organisms including many small fish that take refuge in the weed. However, the absence of weed control can lead to habitat decline (such as deoxygenation in dense weed beds), while alternative weed control methods can also result in loss of habitat.

Weed cutting machines have been developed locally for the Hawke’s Bay and Bay of Plenty regions. Marginal or submerged weeds in drains and streams are simply cut and left to drift downstream to a convenient collection point for a digger. This option can be very cost effective ($130/km).


Rototilling (underwater rotary hoeing) is another mechanical option for weed control and can be used in water depths of between 1.5 and 4 m.

The depth of sediment penetration affects the results. Deep rototilling (to ca. 3–5 cm sediment depth) is more costly (ca. $5000/ha) but provides a greater duration of control (1–2 years in Lake Dunstan), while shallow rototilling (blades set at surface of sediment) is more rapid and less costly (ca. $1000–2000/ha) but provides only about six months control. Shallow rototilling is similar to weed cutting (when cut at equivalent water depths) in efficiency and cost.

The texture of sediment and presence of obstacles have a large bearing on the ease of operation and success of the outcome. The largely soft lake sediments and widespread occurrence of submerged rocks and trees in the Waikato hydro-lakes would be largely incompatible with this option.


Suction dredging contractors have been used in the Rotorua lakes to clear dense patches of weed over small high use areas.

The costs have been estimated at ca. $15–20,000/ha; $1000/day for equipment x 2 divers plus 1 surface operator and suction equipment, with a clearance rate of up to 20 days/ha in dense weed.

Suction dredging can be effective for up to three years in lagarosiphon beds, but re-establishment can be as short as two months for hornwort. However, there is the potential to remove targeted infestations of new invasive weed species if they are identified at an early stage of establishment: all fragments dredged can be collected and removed from the lake.

Mechanical Diggers

New Zealand is primarily an agricultural country, with medium to high rainfall. It therefore has a vast network of drainage and irrigation systems, many of which present an ideal opportunity for aquatic plants to grow. Unfortunately this often gives rise to conflict with farming interests where free unimpeded drainage or water supply is the key function of a drain or irrigation canal.

Weed clearance in drains and canals is most commonly carried out by diggers, at a cost of ca. $1000/km. Although diggers may be necessary to remove silt build-up, they also widen and deepen drains, providing more habitat for weed growth. They have also been linked to the spread of weeds from one waterway to another.

Alternatively cheaper weed cutting machines have been developed where marginal or submerged weeds are simply cut, but left to drift downstream. This option can be very cost effective ($130/km).

Habitat manipulation

Aquatic plants require water and suitable nutrients, oxygen, light, temperature, pH, and substrates (rooted plants only) for growth. Variations outside a definable range can destroy plants and deliberate manipulation techniques have been used to control plants.

Nutrient Control

Reduction of nutrients entering a water body can be achieved by inflow diversion and by nutrient removal technology such as flocculation. Nutrient precipitation and aeration methods have also been applied directly to open water bodies to reduce nutrient content of the water. These methods are generally expensive and more relevant for minimising eutrophication in small waterbodies or controlling planktonic algae.

Lake Tikitapu (Blue Lake) has the lowest nutrient content and alkalinity of all the Rotorua Lakes. This is associated with an unusual growth form of Lagarosiphon, which has low stature and is slow-growing. Lagarosiphon in Lake Tikitapu is of little nuisance value compared to other Rotorua Lakes.

Low alkalinity has also been noted in the Northland sand dune lakes (Taharoa, Waikere and Kai-Iwi) and growth studies on sediment core and water samples from these lakes showed that a substantial increase in Lagarosiphon growth followed when alkalinity was increased.

Unfortunately, it is not generally feasible to restore enriched lakes to a nutrient-limited condition, but in the case of Lake Tikitapu it is feasible to prevent enrichment beyond the threshold required for normal Lagarosiphon growth. This will require deliberate implementation of catchment and land management policies, such as the removal of all sewage from the catchment and careful control of logging practices.


Water level drawdown is a well establish technique for submerged weed control and was practiced on New Zealand hydro-electric lakes over a number of years. Deliberate drawdown for weed control has been discontinued primarily on account of cost (lost hydro-generation potential) and adverse environmental impacts. It has relevance only those few waterbodies and waterways with controlled outlets that allow water level manipulation.

Shading and substrate modification

Dyes (such as aqua shade and nigrosine) have been used to suppress plant growth by absorbing light passing through water. Their application is limited to smaller utility waterbodies and is commonly seen on golf courses.

Light can also be prevented from reaching submerged plants by using various covers including polyethylene, PVC, polypropylene, nylon, synthetic rubber materials and fibre glass screens (aqua screen). Covers are commonly placed on the bottom as opposed to surface-floating techniques. Negatively buoyant products are generally more expensive but require less labour to put in place.

Polythene sheeting has been used in New Zealand to control localised growths of Lagarosiphon in Lake Wanaka, South Island, over the last three years, in Lake Rotoiti, North Island, adjacent to a jetty, for over twenty-five years, and more recently in Lake Waikaremoana in an attempt to eradicate Lagarosiphon from Rosie Bay. Higher material costs may be offset by the greater durability of weed mat material compared to polythene. Lining methods are limited to localised areas that are moderately wave-protected or sheltered. They are also not suitable where high siltation or unstable gradients occur.

Biological control

A range of biological control agents of submerged aquatic weeds have been researched overseas and some in New Zealand. Most are unsuitable for use in New Zealand.


Weed-eating snails Marisa cornuarietis were imported by Professor V.J. Chapman, but in trials with an initial shipment they were found to eat smaller snail species and they were not able to consume weed in the same quantity as grass carp. They were therefore considered unsuitable for weed control in New Zealand.


Insects have been used to control floating and marginal species. In New Zealand the alligator weed beetle (Agasicles hygrophila) has been introduced and released to control Alternathera philoxeroides a marginal aquatic weed.

More information on Alternathera philoxeroides


In some lakes in New Zealand submerged weeds are controlled to approximately 1 m below the surface where large populations of black swans (Cygnus atratus) intensively browse. They may even have contributed to macrophyte collapses in some lakes (for example, Lake Whangape). The level of control is variable and dependent on plant growth rates and swan feeding rates. Lake Rotoehu presently supports a large population of swans and other wildlife which undoubtedly exert some control on Elodea canadensis growths at the south-east end of this lake.

Ducks, geese and swans have been used to control aquatic weeds in small water bodies overseas, but large numbers are required and they would be difficult to manage for general weed control in the Rotorua Lakes. Furthermore, Lagarosiphon major and Ceratophyllum demersum are less preferred by black swans. Large swan populations also result in an often unacceptable level of fouling in areas that public frequent. Swans in Lake Taupo have been known to move to nearby bays and uproot desirable native plant species.

Grass Carp

Grass carp (Ctenopharyngodon idella) are probably the best known and widely used method for biological control of submerged aquatic plants. They have been the subject of intensive investigation overseas and in New Zealand. Grass carp are now commercially available to water body managers for the control of aquatic weeds in New Zealand. Their use is subject to approval being granted from the Minister of Conservation, the Ministry of Fisheries, and from the regional Fish and Game Council (if release is to the habitat of sportsfish or gamebirds), and with consultation with relevant Iwi and the public.

Grass carp are often confused with koi carp. The latter are common throughout the Waikato where they (together with rudd) are implicated in the loss of much aquatic vegetation and habitat in lakes. Koi carp can breed in our waterways, whereas grass carp are extremely unlikely to do so; for this reason diploid grass carp are now available for use.

Risks include:

  • The site may not be suitable for containing the fish. It might be prone to flooding or not suitable for constructing and maintaining fish barriers in the longer term. Grass carp have a strong migratory instinct and there have been several surprising escapes. Escapes represent a financial loss in that re-stocking is required and they could present an undesirable environmental risk within the catchment.

  • There may be ecological impacts within the release area with the anticipated loss of vegetation and the potential for damage within the catchment should the fish escape. An evaluation of the risks is required by the Minister of Conservation (through the Department of Conservation) in the form of an Environmental Impact Assessment (EIA) report. Partial control of macrophytes for extended periods of time has not been demonstrated. Once grass carp consume macrophytes at a rate that begins to reduce standing crop, then total loss of macrophytes is a likely outcome as consumption outstrips macrophyte production.

In New Zealand, nearly all grass carp stocked for aquatic plant control have been in smaller water bodies (less than 100 ha), for example Parkinson's Lake in South Auckland (2 ha) and Elands Lake (4 ha). To date, Lake Omapere is New Zealands biggest example with an area of 11.62 km2 and a maximum depth of 2.6 m. Grass carp releases into the lake in August and again in December 2000 have totalled ca. 40,000 fish for the control of the aquatic weed Egeria densa.

The use of grass carp and associated costs are largely dependent on the desirable stocking rate and a fish price of ca. $25 each. Furthermore it may become desirable to remove fish from a waterbody to avoid total devegetation. This is difficult to achieve but possible through the use of netting, or with the fish toxin Prentox. However, the use of Prentox on large waterbodies (e.g. the size of Lake Omapere) in New Zealand has not been attempted.

Chemical control


Diquat is one of two herbicides registered in New Zealand for use in water. Diquat is a short-lived contact herbicide controlling aquatic weeds. The fate and environmental impact of diquat application to aquatic weed beds has been the subject of intensive studies in New Zealand and overseas. Aquatic plant species range in their susceptibility to diquat from tolerant (for example, native characeans) to highly susceptible (for example, oxygen weed species, especially Egeria densa and Elodea canadensis). Lagarosiphon major and Ceratophyllum demersum are moderately susceptible. The selective action of diquat can be utilised to enhance low-growing desirable native characean species, while controlling tall nuisance weed beds. Following the die back of tall oxygen weed beds, native pondweeds may also re-establish from seed and form temporary dense growths until they are once again competitively displaced by the taller and denser growths of exotic weed beds. Chemical control is commonly used where there are extensive weed beds to be controlled and rapid reduction of plant biomass is needed. This method of weed control is generally the most cost effective and efficient, particularly when the time-frame to complete the job is considered. Cost estimates for chemical control of submerged weeds have been reliably established in the North Island based on annual applications to a wide range of waterbodies. Chemical and application costs are typically in the order of $1600/ha, including herbicide, spraying equipment and applicator. Weed beds take one week or more (longer in winter) before showing effects from diquat control. The level of control can be variable and may not always be consistent, since results can be affected by water turbidity, epiphytic growths and sediment deposits on plant surfaces or water movement. The latter may cause rapid dispersion of herbicide, reducing contact time with targeted weed beds and possibly causing effects on non-target areas. This has been a problem particularly with aqueous applications.

More information


Endothall, in the form of Aquathol® K and Aquathol® super K was registered for use in New Zealand in 2004. Endothall is a short lived contact herbicide used to control aquatic weeds. When the tissues of susceptible plants come into contact with these products they usually die within 4 – 6 weeks. The compound that causes this reaction is the dipotassium salt of endothall which contains, potassium, carbon, hydrogen, and oxygen. It is broken down rapidly by soil microbes to CO2 and other non-toxic naturally occurring products. Depending on the amount of suitable microbes and the dilution and mixing rates present in a stream, lake or river the dipotassium salt of endothall can take 1 day to 30 days to degrade. Susceptible aquatic weed species for control with endothall include C. demersum, H. verticillata, and L. major. Like diquat, not all plants are equally susceptible to endothall, which enables selective control of target weeds to be achieved. For example, native characeans were not affected by endothall at rates used to control weeds in NIWA trials.

More information

Integrated control

Rarely does one type of weed control remedy a weed problem. A combination of the above options is often required or provides the best strategy. For example if a leave-alone approach is adopted for a particular weed problem it may still be beneficial to implement some containment measures such as signage and public education to minimise the spread of weeds to other local waterways. Or it may be beneficial to use grass carp, but only after herbicides have reduced the initial biomass of weeds.

More importantly, it should be noted that although a number of control options have been outlined here, these are not the entire solution. There is a need to carefully evaluate the problems, identify the issues, and define the options. Monitoring usually provides important feedback, enabling more effective management for the outcomes sought.

Revised October 2002

Weed cutter at Kinloch. (Photos: R. Wells)
Weed harvester at Kinloch. (Photos: R. Wells)
Rototiller on Lake Wanaka. (Photo: J.Clayton)
Rototiller results one year later. (Photo: J.Clayton)
Suction dredge underwater. [NIWA]
Suction dredge Lake Tarawera. (Photo: R. Wells)
Mechanical digger. (Photo: R. Wells)
Drain weed harvester. (Photo: J. Clayton)
Research subject: Biosecurity