Comparative soil water, pasture production, and crop yields in phase farming systems with lucerne and annual pasture in Western Australia

Two field experiments in the Great Southern region of Western Australia compared the soil water content under lucerne (Medicago sativa) with subterranean clover (Trifolium subterranean) and annual medic (Medicago polymorpha) over a 2-year period. Lucerne depleted soil water (10-150 cm) between 40 and 100 mm at Borden and 20 and 60 mm at Pingrup compared with annual pasture. There was also less stored soil water after wheat (Triticum aestivum) and canola (Brassica napus) phases which followed the lucerne and annual pasture treatments, 30 and 48 mm after wheat, 49 and 29 mm after canola at Borden and Pingrup, respectively. Lucerne plant densities declined over 2 seasons from 35 to 25 plants/m(2) (Borden) and from 56 to 42 plants/m(2) (Pingrup), although it produced herbage quantities similar to or greater than clover/medic pastures. The lucerne pasture also had a reduced weed component. Wheat yield at Borden was higher after lucerne (4.7 t/ha) than after annual pasture (4.0 t/ha), whereas at Pingrup yields were similar (2 t/ha) but grain protein was higher (13.7% compared with 12.6%). There was no yield response to applied nitrogen after lucerne or annual pasture at either site, but it increased grain protein at both sites. There was no pasture treatment effect on canola yield or oil content at Borden (2 t/ha, 46% oil). However, at Pingrup yield was higher (1.5 t/ha compared to 1.3 t/ha) and oil content was similar (41%) following lucerne-wheat. The results show that lucerne provides an opportunity to develop farming systems with greater water-use in the wheatbelt of Western Australia, and that at least 2 crops can be grown after 3 years of lucerne before soil water returns to the level found after annual pasture.