Use of pine oil and sugar to reduce weed seed banks under cane needle grass ('Nassella hyalina') patches

Cane needle grass (Nassella hyalina (Nees) Barkworth) is perennial, exotic, unpalatable weedy grass from South America threatening critically endangered grasslands within Victorian Basalt Plains in Victoria. Recent advances in weed management now place greater emphasis on eradication of small manageable 'locally eradicable' weed populations before their incursions are beyond eradication. A key component of weed eradication is destroying or controlling the weed seed bank. Essential oils have been shown to exhibit herbicidal activity and have been previously used for weed seed bank control. Carbon (sugar) has been shown to increase microbial activity and has also been linked to reducing seed bank germination. This trial examined applications of an essential oil (pine oil) at a range of rates (0, 5, 10 and 20% applied at 20 000 L ha-1), with and without carbon (sugar at 0.22 kg C ha-1) to assess their effects on cane needle grass seed banks in the field. Prior to treatment application, forty patches of cane needle grass were identified and individually treated with spot applications of glyphosate (10 mL L-1) to kill mature standing cane needle grass plants. Thirty six days after glyphosate application, two 1 m2 patches of cane needle grass were pegged out individually at each cane needle grass patch with one plot having all vegetation dug out and cane needle grass plants counted. The other 1 m2 cane needle grass plot was left intact. Before pine oil and carbon applications, a single 10 cm diameter x 5 cm deep soil core was collected from each cane needle grass patch and used to estimate the initial seed bank via germination from this core. Field cane needle grass germinations were assessed by counting individual plants and by making visual % density assessments. Associated species were assessed by making visual % density assessments. The seed bank below cane needle grass patches were dominated by cane needle grass, annual ryegrass, onion weed and broadleaf weeds while only traces of native grass (Austrostipa sp. and windmill grass, Chloris truncata R. Br) and native herbs were found. Cane needle grass patches had an average density of 19.5 cane needle grass plants m-2 and an estimated seed bank of 3 800 seeds m-2. More than a third of bared plots had seeding cane needle grass plants six months after treatment. Pine oil and/or carbon (sugar) did not reduce cane needle grass from treated plots but did reduce annual ryegrass and increased bare ground 6 months after treatment. Reducing annual ryegrass competition and increasing bare ground could be a useful mechanism for the establishment of indigenous grassland species. Application of herbicide and physical removal of cane needle grass plants did not prevent cane needle grass densities returning to close to pre-treatment densities 28 months after treatment.