STUBBLE RETENTION AND TILLAGE IN A SEMIARID ENVIRONMENT .3. RESPONSE OF WHEAT

Wheat crops were studied over 4 years at two sites in north-west Victo ria, Australia in response to fallow treatments. The four combinations of stubble management (with and without) and tillage (with and withou t) in 18-month-long fallows of a fallow-wheat rotation were compared w ith a tilled summer fallow prior to field pea and wheat crops in rotat ion. Grain yield responded diversely to fallow method. At Dooen, wheat yield from the tilled fallow without stubble increased over that in t he field pea-wheat system by an average of 0.9 Mg ha(-1) (from 2.7 to 3.6 Mg ha(-1)), Stubble retention further increased yield over the no- stubble treatments in three of the four years, averaging an additional 0.6 Mg ha(-1). At Walpeup the tilled fallow without stubble increased yield above the field pea-wheat system in two years, by 0.7 Mg ha(-1) from 2.6 to 3.2 Mg ha(-1) in 1990 and by 0.6 Mg ha(-1) from 1.7 to 2. 3 Mg ha(-1) in 1991. Stubble retention increased yield in only one yea r, 1988, by 0.5 Mg ha(-1) and that was with tillage, Zero tillage had no independent effect on yield, Crop water use varied from 173 to 317 mm at Walpeup and from 247 to 435 mm at Dooen, At Walpeup, total crop water use was largely determined by its pre-anthesis component whereas at Dooen it was strongly related to water use during grain filling, i ndicative of the importance of water reserves deep in the profile, par ticularly under stubble retention with zero tillage. Total N uptake wa s strongly related to yield and to a lesser extent to soil mineral nit rogen at sowing. At Walpeup, high wheat grain [N] was observed in the field pea-wheat rotation (2.3 to 2.7% N) and, over both sites and all treatments, high grain [N] was loosely associated with low yield (Walp eup: R-2 = 0.51, Dooen: R-2 = 0.26). The response of wheat to conserva tion tillage was strongly driven by water availability and to a lesser extent N availability. (C) 1997 Elsevier Science B.V.