An experiment at 3 sites (Birchip, Elmore and Speed) in the northern Victorian cropping belt compared dry matter (DM) production of short-term (2 year) pastures and their contributions to soil mineral nitrogen (N) and subsequent wheat and barley production. The pastures included different varieties of subterranean clover, annual medic and lucerne, and these were compared with ryegrass-dominant pasture, which represented the experimental control.

More productive legume pastures generally resulted in greater accumulation of soil mineral N at sowing of the following cereal at both Elmore and Speed; however, at Birchip, soil mineral N remained high under all treatments. At Elmore and Speed, significant (P<0.10) positive relationships were observed between available N at sowing and subsequent wheat and barley production. Cereal grain yield at Birchip was not associated with available N at sowing.

The quantities of soil mineral N available at sowing (152 kg/ha) of the cereals were in excess of crop demand at Birchip. At Elmore, the soil mineral N supply (83 kg/ha) was below that required for wheat and barley to reach their water-limited potential yield (20 kg grain/mm of growing season rainfall). However, at Speed, the supply of soil mineral N (63 kg/ha) was sufficient to achieve the water-limited potential grain yield and to produce malting-grade barley, but not sufficient to elevate wheat grain protein concentrations above 11.5%.

In environments with low soil N levels, the amount of residual N following short-term pastures increased the availability of N to following cereals. Whether this is sufficient to satisfy subsequent crop demand is largely determined by water availability in the year of cropping. In cases where available N is already high, short-term pasture phases may have little effect on increasing crop production.