MenuMain navigation


On this page

Decontamination of fishing nets and other equipment to prevent transfer of freshwater pest organisms

New Zealand’s freshwater environments support many species of introduced aquatic plants and animals, often to the detriment of indigenous flora and fauna. The spread of these species throughout New Zealand is largely the result of human activities. One common means of transfer is via contaminated fishing equipment used by commercial eelers, and by science and management agencies in the course of their work. Currently, to reduce this risk, fishing nets are usually hung out to dry, desiccating organisms that remain attached to the nets. This is effective for most aquatic life, but it requires a period of drying that cannot always be achieved when nets are re-used within short time periods, particularly in wetter and cooler times of the year. Additionally, some plant material can also remain viable after long periods of desiccation.

NIWA (with funding from the Department of Conservation and the Foundation for Research, Science and Technology) conducted a study to identify potential treatments that could be used to decontaminate fishing nets, and so prevent the accidental spread of plants and animals on fishing equipment.

Eight compounds were selected for evaluation based on their potential toxicity to target plants and animals and on their relative ease of use. The eight compounds were salt, lime, copper sulphate, alcohol, chlorine, quaternary ammonium compounds (QAC), alum, and detergent. Each product was assessed for its toxicity to submerged aquatic weeds (egeria and hornwort), marginal aquatic weeds (parrots feather and alligator weed), aquatic snails (Lymnaea stagnalis and Physa acuta), and fish (brown bullhead catfish). The plant species were chosen because they are high-priority weed species, the snails to represent a broader range of organisms, and catfish because they are capable of surviving out of water for long periods and pose a risk of dispersal between waterbodies by fishnets.

Salt (35g/L) was the most promising treatment with regard to efficacy (though of limited effect on emergent plants), environmental impact, safety, availability, and ease of use. Further tests are underway to more closely determine efficacy under a range of exposure times and temperatures, and with a wider range of test organisms and life stages.

Tony Dugdale and Rohan Wells

National Centre for Aquatic Biodiversity and Biosecurity

Hornwort: a serious threat to South Island lakes

Hornwort, an aquatic weed capable of growing taller than a 3-storey building, is the latest aquatic strangler to invade our waterways and lakes.

Hornwort is currently considered New Zealand’s worst submerged weed. The reasons for this include its ability for rapid spread, growth in a wide range of waterbodies and growth to deeper depths than other weed species. It’s also the tallest-growing weed, recently found to be growing to 10 m in height in Lake Maraetai as a very dense surface-reaching weed bed. Weed beds such as this can be both an aesthetic nuisance and a serious recreational hazard.

Hornwort's vital statistics
Name: Hornwort or coontail (Ceratophyllum demersum L.)
Family: Ceratophyllaceae
Origin: Native to North America, probably introduced to NZ via the aquarium trade
Key attributes: Ranked NZ's worst submerged weed
NZ's tallest growing submerged weed (up to 10 m)
NZ's deepest growing submerged weed (to depths below 12 m)
NZ's only non-rooted submerged weed
Sale and distribution prohibited under the Noxious Plant Act 1979

Like most other submerged weeds, hornwort does not set seed in New Zealand, but reproduces from broken off fragments that are easily moved to new sites on boats, boat trailers, fishing gear, float planes, and weed harvesters. Just a small fragment of the plant can be enough to fill another lake.

From an ecological perspective, hornwort weed beds outcompete and smother all other aquatic plants. This includes native submerged species as well as other noxious invaders such as the oxygen weeds, egeria (Egeria densa) and lagarosiphon (Lagarosiphon major). While overseas, hornwort is often regarded as a weed of more nutrient-enriched waters, the spread of this weed into Lakes Taupo and Tarawera is evidence that it can also do well here in New Zealand in more pristine lakes, particularly in sheltered bays and at deeper depths.

Another unique feature of hornwort is it’s non-rooted growth form. While it can anchor itself to bottom sediments with buried stems, taller plants are easily dislodged by strong currents or waves and can form large drifting weed rafts. These are commonly seen in Lake Rotoaira and the Waikato River lakes, particularly during autumn, where they obstruct hydroelectric power station intakes resulting in substantial costs associated with removal and lost generation.

History of spread

Hornwort is a popular aquarium and ornamental pond plant overseas and this is the likely means by which it was first introduced into New Zealand. The plant was first recorded in natural waters near Napier in 1961. A couple of years later it was recorded at Mihi Bridge on the Waikato River from which it rapidly spread with masses of the weed causing the closure of the Lake Ohakuri power station in 1965. From these initial sites, hornwort has since spread to sites from Northland to Manawatu, and is currently known to be present in over 30 North Island lakes.

Until recently hornwort had not managed to spread to the South Island, but in 2002 the plant was found in a stream and several ponds near Motueka. Concerns about protecting the many pristine South Island lakes from the ravages of this plant prompted the Department of Conservation and local authorities in the Nelson area to act swiftly to contain and eradicate it. Evidence from this incursion suggested that the hornwort had probably been planted in an ornamental pond and had escaped to a nearby stream. However some people had the misguided belief that this weed would benefit fisheries and wildlife values of waterways, and it may have been introduced along with coarse fish.


Hornwort's Invasion Timeline
1961 Hawkes Bay Drains, Tukituki River
1963 Waikato Mihi Bridge, Waikato River
1981 Bay of Plenty Lake Rotoiti
1982 Manawatu Manawatu River
1985 Northland Lake Ngakeketa
1997 Wanganui Lake Whiritoa
1999 Auckland Lake Kereta
2002 South Island Moutere Stream and ponds, Motueka (possibly now eradicated)

Ongoing threat

Hornwort poses an ongoing, serious environmental and economic threat to New Zealand’s lakes and rivers, particularly the South Island’s pristine lakes which includes many that are currently untouched by any exotic weed species. Cooler temperatures in the South Island will be no barrier to hornwort as it overwinters under ice in the northern hemisphere, and in high-altitude Lake Rotoaira (Central North Island), where winter water temperatures can drop to 4 °C, it has proved to be particularly troublesome.

Of particular concern is the threat posed by increased boat traffic and eel fishers between the North and South Islands, which will heighten the risk of hornwort spread. Boat users and eel fishers need to be extra vigilant about removing weed from their boat and trailer as they leave lakes, particularly if moving their boat to 'hornwort-free' areas. Although hornwort, and other weeds, will eventually dessicate and die if kept out of water for long enough, fragments can survive for months in wet spots underfloor or in the anchor well of boats.

Hornwort weed impacts on lake bed sediments

Invasive aquatic weeds have a major impact on habitat quality in New Zealand’s freshwater environments. Weeds such as hornwort (Ceratophyllum demersum), egeria (Egeria densa) and lagarosiphon (Lagarosiphon major) can form dense, unsightly and hazardous weed beds (Fig. 1), displacing valuable native plant communities and removing essential habitat for fish and other important freshwater animals. Recent research by NIWA scientists also indicates that the litter laid down beneath hornwort weed beds might pose a potent barrier in efforts to restore native habitat.

Experiments using sediments collected from two Central North Island lakes (Tarawera and Rotorua) have shown that the growth of native charophyte plants (Chara globularis) is greatly reduced on sediments collected from beneath hornwort weed beds compared to sediment from beneath beds of native charophytes while growth of the invasive weed egeria was not (Figs. 2-4). Other native plant species may be similarly affected, and the sediments from beneath weed beds of other weed species that grow with equivalent vigour to hornwort, may have the same detrimental impact on charophyte growth. However, it is suspected that the impact of hornwort weed may be greater than with other weeds given the unique growth habit of this species.

Hornwort weed usually anchors itself to bottom sediments with buried stems and doesn’t form true roots. This could make it immune to any unfavourable sediment conditions (eg. anoxia, toxins) that develop as a result of the high levels of detritus that appear to accumulate beneath these and other weed beds. Plants with roots are more susceptible but some, with robust, vascular root systems, may have the ability to tolerate, or even remediate, sediments to a certain extent by releasing oxygen from their roots. This includes egeria and may offer an explanation as to why growth of this plant was not inhibited on hornwort sediment as outlined above. In contrast to this plant and other vascular species, native charophytes have much smaller, non-vascular root systems (rhizoids) which cannot transport bulk oxygen and may therefore be much more susceptible to unfavourable sediment conditions.

Ongoing research at NIWA continues to address this issue, with experiments planned to study the impact of hornwort weed beds on native aquatic plant communities and sediment and water quality in Lake Taupo.

Fleur Matheson

Mary de Winton

John Clayton

Tracey Edwards

This research is funded by the New Zealand Foundation for Research, Science and Technology under contract no. C01X0221 Aquatic Plant Management.

Effects of grass carp on the biota of a Waikato drain

In 1971 grass carp (Ctenopharyngodon idella) were imported from Hong Kong into New Zealand to evaluate their potential for the control of aquatic weeds. Various trials demonstrated that the carp could eliminate virtually all aquatic plants in discrete water bodies and streams. However, trials in drains have had mixed success because grass carp in these systems often suffer from predation and from low oxygen and pH levels, and there have been limited studies on the impacts of grass carp on other biota in drains. This information gap is now becoming important because grass carp are commercially available and promoted in all kinds of waterways for "weed" control in New Zealand. Managers require the approval of the Minister of Conservation to introduce the grass carp to a new site and this is dependent on the evidence being provided that the risk grass carp pose to the natural values in the proposed area is acceptable.

The aim of the study was to assess the potential of grass carp as a biological control agent of aquatic weeds in Waikato drains, by measuring changes in aquatic vegetation and biodiversity of macro-invertebrates, fish and waterfowl in drains stocked (with grass carp) and in unstocked (reference) drains.

The drains selected for the study have low overall species diversity and a history of aquatic macrophyte problems, the dominant weed species being Ceratophyllum demersum (hornwort). Prior to this study, weeds were controlled by mechanical clearing. The limited range of macrophyte species present in both drains is typical of a large number of drains throughout the North Island. In this instance nine species were identified – seven were alien species common in the North Island (and these dominated the vegetation cover), and two were native species Potamogeton ochreatus and Polygonum salicifolium, which are common and widespread throughout New Zealand, but were in low abundance in the drains.

Effects on vegetation. Prior to releasing grass carp in spring, both drains had little surface vegetation cover (less than 10%), with only about 30% cover when subsurface vegetation is included. Six months after grass carp release, the drains were noticeably different in their vegetation composition and structure. In the drain with no grass carp there was an increase in vegetation to about 80% cover, dominated by C. demersum, because the vegetation was recovering to levels that existed prior to mechanical weed clearance. In contrast, the drain with grass carp had virtually no submerged vegetation, and total cover was only about 17%, comprising mainly Glyceria maxima (a marginal species) growing in from the edges of the drain (Figure 1). Grass carp reduced the total number of macrophyte species by four, and this is likely to increase as the fish age and include fibrous plants species (such as G. maxima) in their diet.

Effects on invertebrates. The two drains had only four macro-invertebrate taxa in winter and six in summer. The low species richness shows that these drains are an unfavorable environment in which only a few tolerant invertebrates species can survive. In the drain with grass carp, vegetation cover was 80% prior to mechanical clearing of weed beds, which means that during the experiment the grass carp ate away 2–3 ha of habitat for invertebrates, with a consequent reduction in invertebrates.

Moderate increases in benthic tubificid worms, which occurred with removal of the weed, would not have compensated for the overall loss of invertebrate numbers or biomass.

Effects on other fish. Grass carp appeared to have had no effect on the two major fish species in the treated drain (brown bullhead catfish and shortfinned eels), either directly or through removal of macrophytes – catch rates showed seasonal trends in abundance only. The catfish were generally small because there were few older fish present and strong recruitment was evident from catfish spawning in the drain with grass carp. Eel numbers and size varied little during the trial in both the stocked and reference drains.

Effects on birds. The range of bird species present was consistent with the pastoral setting of the drains. Nineteen species of bird were observed within 20 m of the drains, and of these nine were birds normally found near the water, including mallards, grey duck, grey teal, spurwing plover, kingfisher, white-face heron, pukeko, paradise shelduck and feral goose. Concerns have been expressed that low numbers of invertebrates would mean fewer surviving young waterfowl, and this trial has shown that grass carp reduced macroinvertebrates. However it is a difficult task (without further replication of the experiment) to assess the extent to which grass carp affect survival of young birds by reducing invertebrate numbers and diversity, when so many other variables also affect their survival.

Rohan Wells and Deborah Hofstra

This work was also supported and funded by the Department of Conservation.

The study was undertaken by R.D.S. Wells, B.J. Hicks, H.J. Bannon, J. Dyer and P. Teal.

For further reading see Wells, R.D.S. (1999). Grass carp: an effective option for aquatic weed control? Water & Atmosphere 7(2): 13-15

Controlling hydrilla

Hydrilla verticillata (Lf) Royle (hydrilla) is an invasive submerged aquatic plant, which has earned worldwide recognition as one of the worst aquatic weeds. Hydrilla was first recorded in Hawke’s Bay lakes in the 1960s. Currently its distribution is limited to four lakes in the Hawke’s Bay region. In two of these, Lakes Tutira and Waikopiro, hydrilla forms extensive weed beds, whilst in Lake Opouahi it is less abundant, and in Lake Eland there are only a few plants.

Find out more about Hydrilla verticillata

Hydrilla is a much-branched perennial plant, which can grow from the water’s margins to depths of about 7 m, and it often forms mono-specific communities that can reach to the water surface from depths of 4 m. It can occur in both still and flowing waters, and water quality is rarely limiting. Hydrilla has been recorded in waters that are acidic to alkaline and oligotrophic to eutrophic. Its ability to cover a wide depth range, its dense canopy and tolerance of a wide range in conditions mean that it readily displaces native macrophytes, which typically have a shorter and more open growth form, and can lead to the degradation of fish and wildlife habitat. The presence of large mono-specific communities of hydrilla also affect the recreational and aesthetic value of lakes and waterways. Weed beds of hydrilla are considered a direct nuisance to lake users such as bathers, anglers and boaties, and stranded plant material on the beaches reduces the aesthetic value of the lakes and access to the water.

Of particular concern is the threat hydrilla poses to other lakes and waterways. All of the problems that have been encountered in the Hawke’s Bay, and more, could be expected to occur elsewhere. If hydrilla were to invade a lake in which other introduced submerged weeds dominate, hydrilla can also be expected to displace these species. Tank studies on the competitive growth interactions between hydrilla and the submerged weeds Ceratophyllum demersum, Egeria densa, Elodea canadensis and Lagarosiphon major, indicate that hydrilla is capable of displacing these species. This has far-reaching implications because, unlike hydrilla, these other weed species can be controlled by diquat, the only herbicide currently registered for submerged aquatic use in New Zealand. Furthermore, any escape and subsequent invasion by hydrilla is compounded by the production of vegetative propagules (tubers and turions). Hydrilla is dioecious and with only male plants present in New Zealand it does not reproduce sexually, but asexually by stem fragmentation and two types of specialised vegetative propagules. Tubers (produced underground) and turions (produced in the axils of leaves) are long-lived propagules that enable hydrilla to colonise new sites, and regrow after periods of adverse conditions or after control measures have been implemented. Although turions generally expire after 1–2 years, the longevity of tubers from New Zealand’s hydrilla substantially exceeds the 4 years recorded for tubers of the US hydrilla strain.

Find out more about Ceratophyllum demersum

Find out more about Egeria densa

Find out more about Elodea canadensis

Find out more about Lagarosiphon major

Several strategies (including containment measures and control/eradication trials) have been used in the past to minimise the problems and threats posed by hydrilla. Containment measures have included the use of signage beside lakes with hydrilla to ensure public awareness of the plant, and weed mat has been used in selected areas of public use (i.e. access sites) to minimise the risk of hydrilla spread to other waterbodies. For example, weed matting was laid down in 1988/89 in Lake Opouahi to control the main infestation of hydrilla in the lake, and results indicate that for several years this reduced the area of weed bed and slowed down its spread. The confinement of hydrilla to the Hawke’s Bay Region has no doubt also been facilitated by the prohibition of motorised boats on Lakes Tutira and Waikopiro, the most publicly accessible of the hydrilla-infested lakes.

In an effort to determine whether hydrilla could be controlled and/or eradicated, Lake Eland, which is situated on private farmland, became the site of a grass carp (Ctenopharyngodon idella) trial in 1988 (see 'Control of hydrilla'). In 1988 there was a reported 1 ha of hydrilla covering the 1.5 m–4.5m water depth zone. Two and a half years after the original release of the grass carp (April 1991) there was no trace of hydrilla weed beds in the lake. But in November of the same year, there was occasional spring growth from turions, tubers and stem fragments. More importantly, a newly formed tuber was discovered on a small plant in Lake Eland in April 1996, and in subsequent years (1997–2002) remnant hydrilla plants have still been located and viable tubers are sometimes found in sieved lake sediment during annual lake surveys.

See 'Control of hydrilla'

Herbicides have also been evaluated for their potential to control hydrilla and their potential for use in either whole or partial lake treatments (see 'Control of hydrilla'). The best results were obtained with endothall (dipotassium), a selective contact herbicide that is a desiccant and defoliant with a long history of use in the USA. Effective hydrilla control (plant death) was achieved in tank trials at almost all concentrations (up to maximum label rate) and contact times tested, without damage to the native charophyte species that were also included in the treatments. A field trial was also undertaken to evaluate the effectiveness of endothall to control hydrilla in Lake Waikopiro. Endothall treatment resulted in a significant reduction in hydrilla biomass and height, while native charophytes and shallow-water plant species were maintained. Endothall is potentially the first management tool in New Zealand to achieve targeted hydrilla control.

See 'Control of hydrilla'

Deborah Hofstra, Paul Champion and John Clayton

Further reading

Hofstra, D.E.; Clayton, J.S. (2001). Evaluation of selected herbicides for the control of exotic submerged weeds in New Zealand: I. The use of endothall, triclopyr and dichlobenil. Journal of Aquatic Plant Management 39(1): 20–24.
Hofstra, D.E.; et. a;. (1998). Competitive performance of Hydrilla verticillata in New Zealand. Aquatic Botany 63: 305–324.

Control of hydrilla

Hydrilla (Hydrilla verticillata (Lf) Royle) is an invasive submerged macrophyte that diplaces native flora by the dense canopy it produces. It has been in New Zealand since the 1960s, and is established in four lakes (Tutira, Waikapiro, Opouahi and Eland) in the Hawke's Bay Region.

Longevity of hydrilla tubers and grass carp trial in Elands Lake

Elands Lake is the smallest hydrilla-infested lake at just 4ha and 7m deep. It became a grass carp trial site in 1988 to test the feasibility of eliminating hydrilla from one of the four infested Hawkes Bay lakes using triploid grass carp (Ctenopharyngodon idella). An initial stocking of 400 fish followed a lake vegetation survey using photographic and scuba methods, and a sediment sampling and sieving programme to estimate tuber and turion density.

Tubers and turions are over-wintering vegetative buds or propagules. Tubers, also called subterranean turions, form on underground stems, and turions (also called axillary turions) are formed between the leaf and stem. Both of these propagules provide a source from which the plants can re-grow after natural catastrophic events or after control measures have been implemented, much in the same way that seeds ensure a species survives long after an individual plant has died (hydrilla in New Zealand does not produce seed).

Definition of axillary

Prior to the introduction of grass carp hydrilla formed a continuous band of vegetation around the lake between 1.5 and 4.5m water depth, and occupied an estimated 1ha, with dense subsurface canopy forming beds up to 3m tall.

The initial vegetation survey was followed by at least annual surveys, which revealed that seventeen months after the introduction of grass carp, the density and area occupied by the hydrilla had been significantly reduced. Only one patch (ca. 100m2) was located at the western end of the lake. A year later there was no evidence of any of the original hydrilla weed beds, although extensive searching revealed occasional hydrilla plants that were less than 10cm in height growing from tubers of buried stems.

As recently as April 2000 small hydrilla plants have still been found in Elands Lake. This raises questions about the time frame that may be required to eliminate hydrilla, and whether additional measures may be required to achieve this goal. Based on propagule viability information from the USA, turions could be expected to expire after a year, whereas tubers may survive for up to four years in undisturbed sediment. It is interesting to note that the USA plant is monoecious hydrilla, which differs from the New Zealand dioecious male strain. It now appears that the New Zealand hydrilla is more durable and that it would take at least 10 years without hydrilla weed beds before the tuber and turion bank would be sufficiently depleted to have any chance of eliminating this species from a waterbody. One concern is whether there is a potential for hydrilla to develop new tubers before the plants are actively controlled or grazed, and thereby continue the cycle.

Definition of monoecious

Definition of dioecious

For further reading see: Hofstra, D.E.; Clayton, J.S.; Champion, P.D.; Green, J.D. (1999). Distribution and density of vegetative propagules in the sediments of two New Zealand lakes. Journal of Aquatic Plant Management 37: 41-45.

Effect of fluridone on Hydrilla

Fluridone is an aquatic herbicide that targets the photosynthetic pathway. Specifically it interferes with a desaturase enzyme complex that converts phytoene to gamma-carotene. The consequences of carotenoid inhibition are that chlorophyll although formed will not accumulate, because without carotenoids the chlorophyll is susceptible to photooxidation. Consequently leaves emerging after herbicide treatment will have a characteristic bleached appearance (see photos below). Since its registration in the USA in 1986 fluridone has been used successfully to control hydrilla and other nuisance aquatic macrophytes.

The potential of fluridone to control aquatic weeds in New Zealand was evaluated in the 1980s in field sites and in mesocosm studies (Wells et al. (1986) Journal of Aquatic Plant Management). However fluridone produced only transient albescence in growing tissue, with little damage to older shoots and did not provide effective control. This has largely been attributed to insufficient contact times.

Definition of albescence

In recent New Zealand trials fluridone was re-evaluated as a potential herbicide at a range of concentrations and contact times (exceeding current label contact times). Although differences in plant appearance and biomass were apparent during the five-month study and growth was inhibited, complete control was not achieved under these conditions. In addition the production of vegetative propagules in the presence of fluridone may limit the use of fluridone in New Zealand against hydrilla. However, fluridone is a potential control agent for weed species that do not produce long-lived propagules.

Deborah Hofstra, John Clayton and Paul Champion

Manchurian wild rice: the alien invader can be stopped

Zizania latifolia, commonly known as Manchurian wild rice or Manchurian ricegrass, is a giant semi-aquatic grass that has smothered riverbanks, invaded pastures, and run rampant through drainage channels as it continues its invasion of our waterways.

A native of Asia, Manchurian wild rice was originally introduced to New Zealand around the turn of the last century in the ballast carried by timber ships, which was discarded on the banks of the Northern Wairoa River. Although introduced accidentally (one of the few aquatic weeds not introduced deliberately – for example, as an ornamental pond plant), it has also been deliberately planted in the Hauraki Plains area, supposedly to stabilise stop-banks. However, rather than stabilise banks, Manchurian wild rice can in the longer term cause them to slump and encourage erosion of bank material. Commonly found growing in soft mud, its growth intensifies the wet soft soil conditions that may cause the deterioration of stop-banks. In addition to stop-bank slumping, Manchurian wild rice causes a host of other problems wherever it is present in New Zealand. For example, it invades drainage channels, preventing access to them and impeding water flow and in turn increasing the likelihood of flooding. Unless intensive grazing is maintained in pastures adjacent to Manchurian wild rice-filled drains, it will also invade these areas, encouraged by the flooding it causes by blocking the drains. This plant is extremely invasive in native vegetation and appears to reduce the diversity of vegetation it invades, displacing small-stature species and enveloping taller vegetation. In general, species enveloped by dense growths of this grass are unable to reproduce and sustain themselves under those conditions, resulting in long-term monocultures of Manchurian wild rice.

Within its native range in Asia (Taiwan, Eastern China, and Southeast Asia) there are no reports of nuisance growths of Manchurian wild rice. This may be attributed to the intensive landuse practises surrounding its cultivation as a food plant. Manchurian wild rice is cultivated for its edible seed, rhizomes, young shoots and stem bases. In addition galls induced by the smut fungus (Ustilago esculenta) on Manchurian wild rice are cultivated and used as a summer and autumn vegetable.

In New Zealand, Manchurian wild rice is typically found on the berm of waterways, where it is tolerant of both fresh and brackish water, and is commonly found on the tidal reaches of rivers. It forms dense stands of around 3–4 m height, with a strong deep root system with bulky spreading rhizomes that extend several metres down into soft sediment. Established plants increase in area due to rhizome extension, which can grow to over 10 m from the nearest shoot. Dispersal to new sites is by water movement of seeds and rhizome pieces, as well as transfer on contaminated drainage machinery, which is recognised as a major factor in the spread of Manchurian wild rice between catchments.

The current distribution of Manchurian wild rice is predominantly in the Kaipara District of Northland centred around its site of introduction, the Northern Wairoa River (near Dargaville) and associated waterways. Smaller infestations occur within the Whangarei and Far North Districts, as well as in Rodney and Waitakere Districts (Auckland), Hauraki Plains (Waikato), and Kapiti Coast (Wellington). Potentially Manchurian wild rice could infest any lowland wetland, especially the margins of still or flowing water bodies in New Zealand.

To stem its progress a combination of both physical and chemical control options have been investigated in the past with varying results. Physical control methods have varied, depending on the site in which the Manchurian wild rice is growing. Mechanical diggers have commonly been used, but pose the risk of transferring rhizome fragments to new sites. Northland Regional Council has identified this as the main method of dispersal and actively promotes cleaning of drainage machinery before use in uninfested areas. Mowing, grazing, burning and a combination of these have been used to control Manchurian wild rice that has spread to pastures, but must be constantly maintained to prevent plants from becoming large and unpalatable, because stock will only graze new Manchurian wild rice shoots.

Past herbicide trials in New Zealand have evaluated sodium chlorate, sodium TCA, dalapon (2,2-dichloropropionic acid) in combination with amitrole, paraquat, and glyphosate. None of these products will eradicate Manchurian wild rice, although some do demonstrate herbicidal activity, reducing the height and/or cover of Manchurian wild rice and preventing it from flowering and thus eliminating its chances of dispersal from seed. More recent use of grass-specific herbicides showed some promise.

NIWA included Manchurian wild rice in their Aquatic Plant Management research programme (FRST-funded) because there were limited effective control options available and it is a highly ranked weed species. Its weediness was assessed using the NIWA Aquatic Weed Risk Assessment Model and based on the concerns of water body managers. This research evaluated new tools for the control and management of Manchurian wild rice.

The tools trialled were three herbicides that had previously been used with some success in different regions in New Zealand for the control of nuisance rhizomatous marginal grasses including Manchurian wild rice, phragmites and spartina. Two of these – haloxyfop (Gallant®) and quizalofop (Targa®) – were evaluated because they are grass-selective products. The third product – imazapyr (Arsenal®) – is a broad-spectrum herbicide that had been used to successfully control the nuisance marginal aquatic species phragmites.

We conducted trials to control Manchurian wild rice in containers at our experimental facility at Ruakura and in field plots near Dargaville (in conjunction with Northland Regional Council (NRC)). Trials were monitored for over a year and each product was evaluated at several different rates. In containers both haloxyfop and imazapyr were successful in significantly reducing the leaf biomass of Manchurian wild rice. In the field plots the best results were also achieved with haloxyfop at rates as low as 0.5 kg/ha using very high water rates (1600 L/ha), which reduced the cover of Manchurian wild rice to below 10% for over a year. This rate is equivalent to a ca 40% reduction in the amount of haloxyfop required to control Manchurian wild rice than that previously recommended by NRC.

From the information resulting from these trials and ecotoxicological studies carried out at NIWA, NRC have obtained a consent to control all Manchurian wild rice within their region and have a programme to progressively control it, beginning with isolated areas outside of the main infestation zone.

Deborah Hofstra and Paul Champion
National Centre for Aquatic Biodiversity and Biosecurity

We would like to thank Northland Regional Council for funding part of this research and especially Peter Joynt, who assisted with organisation of the field trial.

Managing marginal mayhem

Two sprawling emergent weeds – alligator weed (Alternanthera philoxeroides) and parrot’s feather (Myriophyllum aquaticum) – have invaded the margins of our streams, ponds and drainage systems and are shrouding our waterways with tangled mats of vegetation. Both species are perennial and are aliens to our shores, characterised by their rapid growth rates during warmer times of the year.

As aquatics, these plants root in soil near the water’s edge and extend their stems over the surface of the water, forming buoyant mats. In addition to marginal aquatic sites, alligator weed will grow as a terrestrial plant, where its weed impact and rate of invasion is dependent on rainfall. It is also tolerant of saline conditions. Parrot’s feather is restricted to freshwater habitats and margins, to which it is well adapted, exhibiting two different leaf forms depending on whether it is growing as a submerged plant or as an emergent.

A native of South America, alligator weed was first recorded in New Zealand in 1906 on the banks of the Northern Wairoa River near Dargaville being introduced into this country via ballast in the days of kauri logging and export. Today it is widely distributed through much of Northland and parts of Auckland, with patchy distribution in the Waikato, and a few sites in the Bay of Plenty, Taumarunui and at least two sites in the South Island. Spread within New Zealand has been mostly due to movement of drainage machinery, or produce contaminated with rhizome fragments, with subsequent spread by water movement. Alligator weed has also been deliberately spread in Australia and New Zealand. Its use as a cultivated vegetable by some ethnic communities (a case of mistaken identity for the traditional herb mukuna-wenna, A. sessilis) has resulted in many new locations in Australia and a few sites within New Zealand. In New Zealand, several of the southernmost sites are found within constructed wetlands used to process wastewater, with alligator weed presumably contaminating the rhizomes of sedges used to treat effluent.

Parrot’s feather occurs in the North Island of New Zealand from Northland through to Wellington, with scattered infestations in Nelson and Marlborough, and one site in Westland. Like alligator weed it is native to South America, and has been widely naturalised elsewhere. Until its inclusion on the national list of plants banned from sale and distribution under the Biosecurity Act (1993), it had often been cultivated in ornamental garden ponds from which it has escaped, with subsequent spread by contaminated drainage machinery. Its ability to grow either as floating emergent mats, submerged, or as a combination of both of these forms, enables it to inhabit a wide range of aquatic conditions.

Current control methods for both parrot’s feather and alligator weed include physical excavation and removal, the use of herbicides, and the introduction of biological control agents (in the case of alligator weed). Mechanical excavation or harvesting provides immediate and localised clearance but also results in fragmentation and potential spread of the weed.

Chemical control has largely been more effective for long-term control than mechanical techniques. Extensive trials both in the USA and Australia have evaluated a large number of herbicides including, diquat, paraquat, glyphosate, metsulfuron, imazapyr, dichlobenil and various hormonal formulations for the control of alligator weed. However, control of aquatic alligator weed was relatively unaffected by those herbicides registered for use in aquatic situations (diquat and glyphosate), which only damage the foliage and new shoots but do not affect older stems and rhizomes. Thus they only provide temporary control (usually less than six weeks). In New Zealand the herbicides metsulfuron (Escort®) and Tordon Gold® (picloram and triclopyr amine), are successfully used to control terrestrial plants, whereas only short-term control is achieved with permissible herbicides in aquatic situations.

Excellent control has been reported when using endothall (Aquathol K®) and 2,4-D against parrot’s feather in the USA, and more recently triclopyr (Garlon 3A®) has provided promising results in a lake in California. In some areas glyphosate (Roundup®) is reported to provide effective control, but equally in others it is not recommended, whilst dichlobenil and diquat (Reglone®) have also shown some efficacy for parrot’s feather. In New Zealand, glyphosate has been used to control parrot’s feather, but several repeat applications are required to give adequate control over the growing season.

The bio-control agents Agasicles hygrophila, Vogtia malloi, and Disonchya argentinensis have been released in New Zealand with successful localised control of alligator weed. A. hygrophila (flea beetle) is an aquatic insect that is very effective at reducing floating mats of alligator weed, however, it is unable to overwinter in cooler regions. Additionally the flea beetle has little effect on the terrestrial form of alligator weed. So, another beetle, D. argentinensis, was introduced into Australia and New Zealand to control in particular alligator weed growing in terrestrial habitats. However, attempts to establish this beetle have been unsuccessful. V. malloi (a moth) was also introduced to Northland from Australia to complement the effects of the flea beetle, as it is apparently more tolerant of flooding and cold temperatures than A. hygrophila. Although both A. hygrophila and V. malloi have established in most areas where alligator weed occurs, neither species appears to be halting the further spread of this plant, or decreasing its impact.

Due to the current limitations of these control methods, alligator weed and parrot’s feather have been identified as appropriate targets to research new control options. In association with Waikato, Bay of Plenty and Northland Regional Councils, NIWA evaluated further chemical options for the control of both species in field sites as well as in tank studies.

For parrot’s feather and alligator weed a range of herbicides were evaluated for their ability to control or kill the weeds when grown in tubs. The best control of parrot’s feather was achieved with triclopyr amine, dichlobenil and endothall, so these three products were subsequently evaluated in field plots. Results from the field trials showed that the best control of parrot’s feather was achieved with triclopyr amine, which was then evaluated further to determine the best (minimum) rate that could be used to successfully control parrot’s feather.

In tank studies successful control of alligator weed was achieved with triclopyr amine, metsulfuron, imazapyr, and triclopyr with picloram. However although some of these products may be acceptable for use on alligator weed growing on land, they should not be applied over waterways. By comparison triclopyr amine is registered in the USA for use in aquatic situations, so this product was evaluated further in tank studies to determine rates that could control alligator weed without the need for additional treatment.

The results from these trials have shown that triclopyr amine can be used to successfully control both parrot’s feather and alligator weed in New Zealand.

Deborah Hofstra, Paul Champion and Tony Dugdale
National Centre for Aquatic Biodiversity and Biosecurity

We would like to acknowledge funding from the Foundation for Research, Science & Technology Contract No. C01X0021. Peter Joynt (Northland Regional Council), Phillip Mabin (Environment Waikato) and Bruce Crabbe and Kirsty Brown (Environment Bay of Plenty) all assisted with field trial selection and these organisations also provided co-funding.

Endothall trial for the control of Ceratophyllum demersum in the Wairarapa

Hornwort (Ceratophyllum demersum) is a major submerged weed in New Zealand, presently found in lakes and waterways in scattered locations throughout the North Island. Available control options for this species are currently limited to mechanical harvesting (short-term control with potential increased risk of spread), the use of grass carp (non-selective control and hornwort is not a preferred species), or the herbicide diquat. Diquat, the only herbicide registered for submerged aquatic vegetation control, can be effective against hornwort and is used in some lakes for the control of this plant. However diquat can be largely ineffective in turbid water because it binds strongly to suspended sediments in the water column, which renders it herbicidially inactive.

Recently evaluations on a range of herbicides that are registered for use in aquatic situations overseas have been carried out to determine their efficacy on hornwort and a range of other waterweeds. Endothall (dipotassium endothall) is one product that gave effective control of hornwort in outdoor tank studies. Endothall is a contact-type herbicide that produces symptoms of defoliation and stem die-off, which has been used in the US for aquatic weed control for over 40 years. Unlike diquat, endothall is not inactivated by suspended sediments, and so potentially provides an effective management technique for the control of hornwort in turbid water and an alternative to diquat. In addition tank studies have shown that endothall can be used at selected rates to control problem species whilst minimising impacts on non-target vegetation.

Following on from successful tank trials, a field trial was undertaken in the Wairarapa region in association with the Wellington Regional Council. Endothall was applied to drainage channels that were choked with hornwort (like that pictured above left). By five days post-treatment clear areas were appearing in treated drains, and within fourteen days the main area of the waterway was clear of weed (see picture above right). Thus endothall provides great promise for future control of this plant.

Deborah Hofstra and Paul Champion

Decision support system for managing aquatic weeds

Severe aquatic weed problems appear “overnight” and usually come with intense public concern and lobbying for an instant solution. Local body managers may have little experience dealing with these situations and require some guidance.

The Aquatic Plant Group at NIWA has put together a decision support system to assist with the management of aquatic weeds (see flow chart).

The steps are:

  1. Define the problem by determining when, where and to whom the weed is a problem. If possible, quantify the problem by defining the area of weed causing the problem, and determining how much the problem is costing directly and indirectly. Consideration also needs to be given to its impacts on the ecology of the area.
  2. It is important to identify the species (see our online species guide) causing nuisance and check if it is a National Pest Plant, or listed in the Plant Pest Accord or a Regional Plant Pest.

See our online species guide

  1. A Weed Risk Assessment can then be carried out by assessing its invasiveness (ability to spread and establish, including versatility, competitive ability and effective dispersal) to determine its potential distribution by considering available habitat. The Risk Assessment also involves elements from the problem definition – its potential economic, recreational and ecological threat. For more on weed risk assessment, see “Aquatic plant pests” in Aniwaniwa no. 18.
  2. If intervention is thought desirable then the outcomes sought need to be stated and the range of management options available considered in relation to their practicality and the budget available. Where business is affected, a cost–benefit analysis can be done. A wide range of options are available, but no one solution is best for all problems or one problem at all times. Methods include mechanical control, chemical control, biological control, habitat modification, sediment lining, shade, and water level fluctuations. Integrated management (integrating several methods) is often more effective. (See our weed management guide online and “Grass carp: an effective option for aquatic weed control?” in Water & Atmosphere 7(2): 13–15.)

See our weed management guide online

  1. A Weed Management Plan can now be formulated and it is usually necessary and desirable to involve and consult stakeholders and the community.
  2. Once the plan is implemented it is important to monitor the results in an objective way and fine-tune the plan as appropriate.

It is always useful to document the process so that future managers can learn from past experience, particularly as every water body is unique and can respond in different ways. Too often managers have gone straight to Step 4, not evaluated outcomes, and kept poor documentation.

For regional and national managers, management of inter-lake / waterway spread is also important. In this context it is necessary to have baseline lake surveys to know what resources are being managed, which water bodies are at risk, and to prioritise their ecological value (link to LakeSPI web reference). This information needs updating periodically, depending on the rate of change. Containment of weed problems and protection of high-value / highly prioritised water bodies by restricting boat access is a common strategy. Public education by various means (signage, school programmes and public campaigns) has been commonly used to minimise inter-lake spread.

Rohan Wells


Freshwater Ecologist
Principal Scientist - Freshwater Ecology
Hornwort specimen (Photo: R. Wells)
Current distribution of hornwort in New Zealand waterways, showing the potentially eradicated South Island site. [NIWA]
Lake weed caught on a boat trailer - this is a typical way for weeds like hornwort to spread to new sites. (Photo: J. Clayton)
Hornwort can form very dense, tall weed beds in the littoral zone of lakes. [NIWA]
Chara globularis plants in a water holding tank awaiting transplant in our growth experiments. [NIWA]
Experiments testing growth of Chara globularis plants on different sediment types. [NIWA]
Biomass of Chara globularis and Egeria densa plants on charophyte and hornwort sediments after 12 months. [NIWA]
Stylised profiles showing differences between drain vegetation abundance and structure - Click to enlarge.
Key to diagram - Click to see enlarged version of diagram
Hydrilla. [NIWA]
Elands Farm Lake (photo by J. Clayton)
Hydrilla tubers and turions (small, green). Click on image for enlargement. (photo by J. Clayton)
Hydrilla shoots. Click on image for enlargement. [NIWA]
Manchurian wild rice in Northland. [NIWA]
Manchurian wild rice in flower. [NIWA]
Manchurian wild rice herbicide trial. [NIWA]
Alligator weed (Alternanthera philoxeroides) shoot and flower. (Photo: P. Champion)
Alligator weed in a drainage ditch. (Photo: Coffey and Clayton 1989)
Emergent shoot of parrot’s feather (Myriophyllum aquaticum) showing minute flowers in the axils. (Photo: Coffey and Clayton 1989)
Parrot’s feather in a drainage ditch in the Waikato. (Photo: R.D.S. Wells)
Drainage channel choked with hornwort Ceratophyllum demersum (Photo by A. Taumoepeau)
Drainage channel fourteen days after treatment with endothall, clear of Ceratophyllum demersum. (Photo by D. Bayly, WRC).
Weed management decision support system. [NIWA]


This page has been marked as archived, and is here for historical reference only.

Information provided may be out of date, and you are advised to check for newer sources in this section.

This content may be removed at a later date.