Effect of gypsum on establishment, persistence and productivity of lucerne and annual pasture legumes on two grey Vertosols in southern New South Wales

Changes in pasture yield and botanical composition due to gypsum application were examined on Vertosols at two locations of differing soil sodicity, Grogan and Morangarell, in southern New South Wales. Two pasture treatments were examined. One was an annual pasture comprised of 3 annual legumes (2 subterranean clover Trifolium subterraneum L. cultivars, Clare and Riverina, and balansa clover T. michelianum Savi cv. Paradana), while the second treatment consisted of lucerne (Medicago sativa L.) cv. Aquarius sown in a mixture with the same annual legumes.Gypsum had no effect on the establishment or persistence of lucerne at either site. Gypsum increased the number of subterranean clover seedlings present in autumn in annual swards at the more sodic Grogan site in each of the 4 years, but provided no difference when the clover was in a mixture with lucerne. Annual legume seed yields in annual-only swards increased with gypsum by up to 58% at Grogan and 38% at Morangarell. Seed yields of both cultivars of subterranean clover declined as a proportion of the total annual legume seed bank when lucerne was included in the mixture, in contrast to balansa clover (at Grogan) and the naturalised annual legumes, burr medic (M. polymorpha L.) and woolly clover (T. tomentosum L.), which all increased in relative seed yield in the presence of lucerne.Total pasture production at the Grogan site increased with gypsum by up to 15% per annum in annual swards and 36% in lucerne swards depending on the season. Yield responses to gypsum by the lucerne component were observed in 10 of the 13 seasonal yield measurements taken at Grogan. However, total pasture yield and seasonal yields were unaffected by both gypsum and pasture type at the less sodic Morangarell site. It was concluded that sowing a diverse mixture of annual legumes or polycultures was conducive to maintaining productivepastures on these spatially variable soils.Lucerne dried the soil profile (0.15'1.15 m) more than annual pastures at both sites. The combination of gypsum and lucerne enhanced water extraction at depth (0.6'1.15 m) at the Grogan site increasing the size of the dry soil buffer whereas gypsum increased soil water at depth (>0.6 m) under annual swards