The mineralization of N from shoot residues of two legume species and a common weed component of Western Australian pastures was measured after a simulated 'summer rainfall' in a laboratory experiment using undisturbed soil cores of a loamy sand. Water was added to the surface of the cores equivalent to a single rainfall event of 10 mm water (treatment 1); or followed 48 h later by a further application of 10 mol water (treatment 2). Net mineralization and microbial biomass N were measured over 144 h following the initial rewetting of the soil cores. Three types of 15N-labelled plant residue, namely (i) lupin leaf (Lupinus angustifolium), (ii) clover (Trifolium subterraneum) shoot, burr and petiole, and (iii) capeweed (Arctotheca calendula) shoots and leaves were placed on the soil surface as large fragments and examined to assess their mineralization and incorporation into microbial biomass. The soils dried rapidly after rewetting and net mineralization of N was very low. The proportion of N as nitrate was increased in the rewetted soils with the major changes occurring in the top 5 cm of soil. However, there was very little mineralization of the surface-applied plant residues, with more than 90% of the mineral N being derived from the native organic matter. Microbial biomass showed fluctuations in both the rewetted treatments, but no consistent increase or decrease and no significant immobilization of 15N. Gross N mineralization was measured using an isotopic dilution technique involving the injection of 15N-labelled solutions into the soil. Estimates of gross N mineralization after a single rainfall event of 10 mm water were, on average, four times greater than measures of net N mineralization. Gross N mineralization rates declined as the soil dried, with three times more mineralization occurring in the 0-5 cm depth (4.73-8.93 �g g-1 day-1), compared with the 5-10 cm depth (0.86-2.38 �g g-1 day-1). A major disadvantage with the isotopic dilution method was that injection of the solutions into soil greatly increased the soil moisture content. Gross mineralization in the injected cores is likely to have been overestimated because of N movement below the sampling zone and increased microbial activity relative to non-injected cores.
