The impact of long-term rotation, tillage and stubble management on lupin (Lupinus angustifolius) productivity

A long-term field experiment commenced in 1979 at Wagga Wagga, NSW, Austral ia, to study the sustainability of a range of rotation, tillage and stubble management systems on a red earth. This paper presents results on lupin gr owth and grain yield in relation to rotation, tillage and stubble managemen t between 1979 and 1992 during which time annual rainfall ranged from 311 t o 799 mm. Mean annual grain yields of lupin varied from 0.26 to 3.15 t ha(-1). In a w heat-lupin rotation, highest mean grain yield (1.59 t ha(-1)) over all year s was obtained from direct drilling into burnt stubble. Direct drilling res ulted in significantly higher yields than conventional cultivation (three p asses) in 1979, 1980, 1981, 1982, and 1988, which was generally related to increases in total dry matter production. Retaining, rather than burning st ubble when direct drilling, usually resulted in lower plant density at emer gence and maturity, but this did not always translate into similar effects on grain yield. Grain yield was significantly reduced by retaining stubble in 1986, 1987, 1989, 1990, and 1991, but significantly improved in 1980 and 1988. When the incidence of brown leaf spot was high or when rainfall duri ng growth was excessively high resulting in waterlogging, yield benefits fr om direct drilling and stubble retention were apparent. The effect of early incorporation of stubble following summer rain with a one-way disc was not consistent, producing significant yield advantages in 3 years (1983, 1989, and 1990), a significant decrease in 1985, but no mean difference across y ears. There was no clear effect of the long rotation (wheat-wheat-lupin) on grain yield, but it increased average plant density across years. The effects of rotation, tillage and stubble management were seasonally rel ated and though often promoting major differences in plant number, the subs equent differences in grain yield were usually reduced due to compensatory effects of pod number per plant. (C) 2000 Elsevier Science B.V. All rights reserved.