Influence of weed competition on autumn-sown lucerne in south-eastern Australia and the field comparison of herbicides and mowing for weed control

Abstract

A field experiment in autumn-sown lucerne was conducted at Tatura to quantify the growth of winter weeds and the effects of weeds on lucerne seedling establishment, plant densities and hay production over the first year of cutting. In addition, mowing treatments and a range of herbicides were evaluated with respect to weed control. Weed growth rates in spring of up to 114 kg ha-l day-1 DM and high yields of annual winter weeds excluded lucerne (Medicago sativa L. cv. Delkalb 167) almost entirely from the first hay cut and significantly reduced seedling vigour and plant densities. The unsprayed control plots yielded 9.3 t ha-1 at the first hay cut and comprised 52% capeweed, 41% ryegrass and less than 1% lucerne. In other treatments where either capeweed or ryegrass remained uncontrolled, the proportion of lucerne in the hay was low. Weeds significantly reduced lucerne yields (P < 0.01) at the first and second hay cuts of 5 cuts taken in the first season. Combinations of herbicides gave the best weed control and highest lucerne yield at the first and second cuts and significantly higher total lucerne yield. The best herbicide combination was trifluralin and bromoxynil, which controlled all weed species on the site, reduced the weed yield at the first cut by 87% below that of the control and produced the highest total lucerne yield of 14 t ha-l. The hay composition of this treatment at the first cut comprised 59% lucerne, 27% ryegrass and 14% other weed species, while at the second hay cut lucerne comprised 95% of the DM. Lucerne was the highest yielding species in all treatments at the second cut. Lucerne plant densities were lowest on the unsprayed control plots, with 56% fewer plants than for the herbicide combination treatments. Mowing in August was effective in preventing the decline in lucerne plant density that occurred on the unsprayed control plots.