Over the last century, New Zealand's natural habitats have come under great pressure from a wide range of introduced plants and animals. Much has been written about the demise of our forests as a result of browsing animals such as possums and deer. Dwindling numbers of native species such as kokako, kaka and kereru (pigeon) have been directly attributed to the voracious appetites of stoats, cats, possums and rats. Introduced plants, some of them garden escapees, are choking large tracts of bush and wetlands. Many species, including Clematis vitally and Kahili ginger, are extremely difficult to eradicate. Sadly, this phenomenon is no longer restricted to terrestrial habitats. Increasingly, many estuaries and saline wetlands are being invaded by the insidious Spartina, or cord grass. Spartina is a perennial rhizomatous grass (meaning that it has a horizontal, underground stem that sends out roots and shoots from its nodes) which can grow up to a metre tall. Two species and one hybrid exist in New Zealand. Of these, the one with the greatest ability for natural spread is S. anglica as it readily disperses by seed. Widely promoted for its sediment-trapping ability, Spartina's potential use for estuary reclamation created a worldwide enthusiasm for planting in the 1900s. By using Spartina it was envisaged that reclamation costs would he reduced. Spartina was introduced to New Zealand around 1913. Plants of a sterile hybrid obtained from Southampton were planted in the Manawatu River estuary near Foxton in an effort to reclaim land for grazing. This planting was extended in 1924, when culms of fertile seed of S. anglica were also planted. These Foxton plantings formed the stock later used to colonise other areas. S. alterniflora from the eastern United States was introduced in the 1950s, and Spartina is now established in many estuaries throughout the country. But there arc substantial concerns about the plant. Water catchment authorities arc worried about how Spartina perturbs natural water flows. Its dense form of growth and associated sediment-trapping can result in raised flood plains and restricted waterways. Fields of densely matted knee-high Spartina significantly reduce the intertidal area available to wading birds wanting to feed, and are also unsuitable for rails and crakes that prefer more open native vegetation. Even flatfish habitat is reduced, and the coarse grass encourages incursion (and contamination) by grazing cattle. Spartina also invades and alters the character of indigenous plant communities such as those found in mangrove and saltmarsh. In the Kaipara Harbour, single plants have been recorded as spreading at a rate of 50 square metres in four years. Spartina is now classified as a noxious weed and legislation is in place to prevent it from being planted. The problem, though, is how to control it. A number of methods have been tried. Herbicides have been applied by hand and air and from boats. Other forms of control have including digging out, steam treatment and the use of weed matting. Gallant, a grass-selective herbicide, has been found to be the most cost effective. Moderately toxic to marine invertebrates in a laboratory situation, field studies have revealed that it is quickly diluted by the tide, thereby minimising actual toxicity problems. However, researchers caution against the widespread use of Gallant until its effects on non-target indigenous plants have been fully determined. Few other techniques tried to date have proven very successful. If plants disintegrate after control measures have been applied, viable floating rhizomes may quickly re-establish. Control work has to be carried out between tides, limiting the time which can be spent on the task. One area where Spartina eradication is being vigorously pursued is the Bay of Islands. Volunteers from the Bay of Islands Coastal Watchdog group began trying to clear the weed from Uruti Bay, near Russell, in 1995. Spartina alterniflora had infested 300-400 square metres of mudflats, probably seeded from floating rhizomes arriving from elsewhere in the Bay. Herbicides were ruled out by the presence of shellfish beds and oyster farms nearby, so the first stage of the control programme involved the assessment of all known nonchemical control methods, including steam treatment, weed matting, burning, suction, dredging and hand removal. A combination of cutting the leaves and burning the base of the shoots near the mud was found to be the most effective. The treatment leaves no toxic residue, and the burner causes only incidental damage to non-target species. The second phase of the project will involve determining how often the burning needs to be done to prevent regrowth and kill the underground rhizomes. Unlikely as it sounds, burning a maritime weed may prove to be the best control method in some situations—a method which could be gainfully adopted elsewhere in the country.