NITROGEN AND WATER FLOWS UNDER PASTURE-WHEAT AND LUPIN-WHEAT ROTATIONS IN DEEP SANDS IN WESTERN-AUSTRALIA - 1 - NITROGEN-FIXATION IN LEGUMES, NET N MINERALIZATION, AND UTILIZATION OF SOIL-DERIVED NITROGEN

Detailed studies on the efficiency with which pastures and crops use s oil-derived nitrogen (N) in southern Australia are limited. Inefficien cies in the N cycle are indicated by widespread soil acidification and low N status in wheat grain. The aims of this study were to document rates of Nz fixation by subterranean clover-based pastures and narrow- leaf lupin, plant uptake of soil-derived N, mineralisation of organic N during legume and cereal phases, and export of N from pastures, lupi n, and wheat in relation to climate and soil water. These measurements were undertaken in a rotation experiment conducted on a deep sand loc ated in the northern wheatbelt of Western Australia at a site with a l ong-term average rainfall of 460 mm. The rotations examined over 3 yea rs were 2 years pasture-wheat and lupin-wheat. The N-15 natural abunda nce technique was used to differentiate soil-derived N from atmospheri c N in legumes. Biomass production, grain yields, and N contents were standard plant measurements in all treatments. Net N mineralisation be tween growing seasons was ascertained by measuring changes in soil ino rganic N to 1.5 m. Growing season net N mineralisation was determined using an in situ method in which soil cores were isolated from plant r oots. Anion exchange resin was used to trap leached below the depth of the soil cores. Nitrogen fixation by subterranean clover in a mixed p asture ranged from 29 to 162 kg N/ha, whereas Np fixation by lupins wa s less variable, ranging from 90 to 151 kg N/ha. Pastures were large c onsumers of soil-derived N (range 58-154 kg N/ha), with capeweed being the most important sink (range 38-120 kg N/ha). In comparison, wheat and lupins were inefficient users of soil N, removing 29-51 kg N/ha wi thin a season. Another 31-67 kg N/ha of inorganic N in soil was not ut ilised by wheat or lupin. Annual net N mineralisation ranged from 80 t o 130 kg N, confirming the high rate of decomposition of organic matte r in the sandy soil. Mineralisation over summer and autumn, when crop and pastures were not grown, supplied similar to 25% of the inorganic N produced in soil profiles in 1995 and 20-40% in 1996. The study indi cated that legumes used in rotations with cereals on deep sands were a ble to add adequate organic N to soil to insure rates of net N mineral isation sufficient to support cereal yields in excess of current shire averages. However, in practice, the asynchrony in supply and demand f or N resulted in the inefficient use of soil-derived N by wheat.