Weed control when establishing irrigated lucerne (Medicago sativa)

Abstract

Three experiments were done at Tamworth, New South Wales, on weed control during lucerne establishment. In experiment 1, autumn sown, trifluralin incorporated in the soil eight weeks before sowing provided better control than benefin of grasses (mainly Poa annua) and deadnettle (Lamium amplexicaule). Neither of these treatments controlled shepherds purse (Capsella bursa pastoris), reduced lucerne density nor increased its yield. Post emergent treatment with 2,4-DB or prometryne increased lucerne yield, the latter being the more effective. In experiment 2, also autumn sown, the problem weeds were deadnettle and shepherds purse. Trifluralin and benefin both increased lucerne yield when applied either one or three weeks before sowing, but hay quality, as gauged by weed contamination, was not improved. Post emergent application of prometryne increased lucerne yield and improved the quality of the hay. Experiment 3 was spring sown. The problem weed, barnyard grass (Echinochloa spp), was controlled by benefin and trifluralin, resulting in greater lucerne yield and better hay quality. Trifluralin was more effective when applied three weeks before sowing compared with one week.