Yield and grain protein of wheat following phased perennial grass, lucerne, and annual pastures

Abstract

The effect of using 4 perennial grasses or lucerne (Medicago sativa L.) in the pasture phase on subsequent wheat grain yield, protein, and grain hardness was investigated at 2 sites (Kamarah and Junee) in the south-eastern Australian cereal belt. The 6 perennial treatments were 5 mixtures of subterranean clover (Trifolium subterraneum L.), with one of lucerne, phalaris (Phalaris aquatica L.), cocksfoot (Dactylis glomerata L), wallaby grass (Austrodanthonia richardsonii (Cashm.) H.P. Linder), or lovegrass (Eragrostis curvula (Schrader) Nees cv. Consol), or one mixture of cocksfoot, phalaris, and lucerne. The results were compared with wheat after one of 3 annual pastures consisting of either pure subterranean clover, subterranean clover with annual volunteer broadleaf and grass weeds, or yellow serradella (Ornithopus compressus L.). The duration of the pasture phase was 3 years at the drier Kamarah site (av. annual rainfall 430 mm) and 4 years at Junee (550 mm). The effect of time of removal of the pastures in the year prior to cropping (28 August–3 September or 6–7 November) and the effect of nitrogen (N) fertiliser application were also examined. In the absence of applied N, wheat grain yields at Kamarah were highest (4.7–4.9 t/ha) and grain protein lowest (10.3–11.1%) following phalaris, wallaby grass, and cocksfoot. Grain protein levels were highest (12.9–13.9%) in wheat following the 3 annual legume swards at both sites. Previous pasture type had no effect on wheat yields at the Junee site. Wheat grain protein and total N taken up by the crop were positively related to available soil N to 100 cm measured at sowing at both sites. Grain protein was inversely related to grain yield at both sites where additional N fertiliser was added, but not in the absence of fertiliser N. There was a positive response in grain protein to delayed time of pasture removal in second year wheat at Junee. The application of additional N fertiliser increased grain protein of wheat following all 9 pasture types at the drier Kamarah site, but at the Junee site there was only a positive grain protein response following phalaris, cocksfoot, and wallaby grass. Early removal of the pasture prior to cropping increased soil water (10–130 cm) at sowing by 18 mm, delayed wheat senescence, and increased crop yield by 11% (0.44 t/ha) at the drier Kamarah site. Early removal of the pasture at Junee increased soil water by 29 mm, crop yields by 2% (0.14 t/ha), and increased grain protein in wheat following cocksfoot, wallaby grass, and phalaris, but not following the 3 annual legume treatments. The study demonstrated that perennial grasses can be successfully incorporated into phased rotations with wheat without affecting grain yield, but protein levels may be lower and timing of pasture removal will be important to limit the effect of water deficits on grain yield.