The effects of past grazing management practice on subsequent seedbed condition, draft requirements, fuel consumption, crop establishment and growth, and grain yield and quality were examined using three tillage systems on two sowing dates. The crop was wheat (Triticum aestivum), sown on a fragile sandy clay loam (red duplex soil) in a dryland agricultural area (307 mm average annual rainfall) of Western Australia. The three tillage-sowing systems investigated were: (i) scarifying followed by sowing with wide (180 mm) points; (ii) direct drilling with wide (180 mm) points; (iii) direct drilling with narrow (50 mm) inverted 'T'-shaped Super-Seeder points. The two sowing dates provided differences in seedbed water content at sowing time. The three grazing management strategies practised in the previous pasture year were: (i) traditional set-stocking (where sheep were grazed continuously for 17 weeks, beginning soon after the start of the early winter rains); (ii) controlled grazing (where sheep were temporarily removed from the enclosure when the topsoil was close to its plastic limit); (iii) no grazing (where the pasture was mown to simulate grazing without trampling). Tillage prior to sowing with wide points reduced the mechanical impedance of the soil following set-stocking and provided a good seedbed for successful crop establishment and growth. In both the controlled-grazing management treatment and the treatment where the pasture had been mown the soil was suitable for direct drilling with both wide and narrow points (i.e. no pre-sowing tillage was required). The use of narrow points had the added advantage of requiring less fuel, but the need for a suitable implement to cover seeds was more critical than for wider sowing points. There were no advantages with respect to grain yield from adopting a controlled-grazing management practice owing to the lack of finishing rainfall. However, grain protein levels were higher in both the controlled and ungrazed treatments compared with the set-stocking treatment.
