318 Seventeenth Australasian Weeds Conference Summary Manchurian wild rice (Zizania latifolia) is a giant invasive grass that forms dense monospecific stands of vegetation on the margins of water bodies, also invading flood-prone pasture and cropping land. A national eradication programme has been initiated for this species. Biomass for component parts of the plant was measured in pot and field trials. In the field Man- churian wild rice produced a total biomass of 109.9 tonnes dry weight ha 1 , with ~80% of this biomass below ground (mostly rhizomes). Of the shoot bio- mass, 41.4% is dead and in pot trials this percentage was higher. The dead biomass was predominantly comprised of shoot bases with large amounts of aeren- chymous tissue that, therefore, could act like a sponge to herbicide application. This may explain the poor control obtained with various herbicide treatments. This paper describes proposed field trials to pre-treat stands of this plant by burning, crushing or mowing prior to herbicide treatment. After these pre-treatments the efficacy of haloxyfop, imazapyr and glyphosate will be evaluated. Keywords Biomass, eradication programme, herbicide trial. INTRODUCTION Manchurian wild rice (Zizania latifolia (Griseb.) Turcz. ex Stapf) is a giant emergent perennial rhi- zomatous grass, growing up to 4 m tall. Plants are comprised of basal fans of leaves (up to 2 cm wide) arising from a bulky, deep rooting and far-spreading rhizome system. Plants form dense, monospecific veg- etation in wetlands and on the margins of water bodies, also invading flood-prone pasture and cropping land. Obstruction of drains by this plant promotes flooding and expands the habitat available to it. It has a limited distribution in New Zealand mostly centred around Dargaville in Northland, which was its site of introduction in the 1900s. Here it occupies approximately 338 ha along 60 km of the Northern Wairoa River, with sixteen outlier populations in west- ern Northland (Joynt and Newby 1998). Additional sites are found in other parts of Northland (Kauri and Whangarei), Auckland (Lakes Kereta and Karaka on Manchurian wild rice ( Zizania latifolia) biomass allocation and implications for control Paul D. Champion and Deborah E. Hofstra National Institute of Water and Atmospheric Research (NIWA), PO Box 11-115, Hamilton 3251, New Zealand Corresponding author: p.champion@niwa.co.nz the South Head of the Kaipara, Helensville, Laingholm and Mangere), Waikato (Waihou, Piako and Awaiti Rivers on the Hauraki Plains and Horohoro and Lake Te Koutu near Cambridge) and Wellington (Waikanae) (Champion and Hofstra 2008). MAF Biosecurity New Zealand has instigated a National Interest Pest Response (NIPR) Programme on this species, with the initial aim of eradicating all sites outside of the Northern Wairoa River, with national eradication the long-term goal (Anon. 2009). Manchurian wild rice is notoriously difficult to control using herbicides, with glyphosate, or a com- bination of dalapon and amitrole, the most commonly used to date (P. Joynt, Northland Regional Council, pers. comm.). Neither gave long-term control even when used at very high rates, with strong regrowth per- sisting after many re-applications of these herbicides. NIWA trials have led to the use of haloxyfop-P methyl in the current eradication programme (Champion et al. 2001), with twice annual application (late spring, early autumn) at 700 g a.i. ha 1 . This paper details the estimation of Manchurian wild rice biomass, including above- and below-ground biomass, and the amount of dead shoot material (trash) present in natural stands of this species, in order to optimise effective control of this weed. Proposed pre- treatment of Manchurian wild rice stands to reduce trash, and a herbicide trial comparing haloxyfop-P methyl with imazapyr and high rates of a new for- mulation of glyphosate (Roundup™ Transorb ® ) are also outlined. MATERIALS AND METHODS A large area of unmanaged Manchurian wild rice was located on the Northern Wairoa Experimental Farm, on the margin of the Kaihu River, near Dargaville. A block of ten 1 m 2 quadrats was established within the stand, over 5 m from the edge. All shoot mate- rial within each quadrat was harvested and divided into living and dead matter. A digger was used to excavate soil over the entire 10 m 2 block, sampling in 20 cm depth increments. All root and rhizome material was separated from soil and subsequently washed to remove any additional soil. Plant material