Control of St. John's wort (Hypericum perforatum var. angustifolium) on non-arable land

Methods for controlling St. John's wort (Hypericum perforatum var. angustifolium) on non-arable land were investigated in two experiments near Tuena, New South Wales. The effects of herbicides, seed, fertilizer and the timing of the application of these inputs on the establishment of surface-sown legumes and grasses, and on the control of St. John's wort were ascertained. The herbicides, amitrole, 2,4-D, 2,4-D + 2,4,5-T, and picloram were applied with and without 2,2-DPA (sodium 2,2-dichloropropionate) ; glyphosate (N-phosphonomethyl glycine) was applied alone. Establishment, survival and ground cover of sown species and the control of St. John's wort were recorded for up to two years after treatment. Herbicide application was essential for the establishment of cocksfoot (Dactylis glomerata) and phalaris (Phalaris tuberosa), but not for subterranean clover (Trifolium subterraneum), on land infested with St. John's wort. Establishment of grasses and legumes was achieved in one experiment by spraying in late summer and sowing in late autumn, and in another experiment by spraying in late autumn and sowing in early winter. Glyphosate was the most effective herbicide in controlling St. John's wort and associated weeds and in promoting the establishment and development of grasses and legumes. However, good establisment of improved species and good control of St. John's wort was also attained on the 2,4-D and 2,4-D f 2,4,5-T treatments. Thus, until glyphosate becomes commercially available, the most economical method for control of St. John's wort on non-arable land will be the aerial application of 2,4-D + 2,2-DPA three weeks after a distinct autumn break followed by seed and fertilizer three weeks later.