Biol Fertil Soils (1995) 20:213-220 9 Springer-Verlag 1995 ORIGINAL PAPER M.J. Connell 9 R.J. Raison 9 P.K. Khanna Nitrogen mineralization in relation to site history and soil properties for a range of Australian forest soils Received: 1 December 1993 Abstract Rates of N mineralization were measured in 27 forest soils encompassing a wide range of forest types and management treatments in south-east Australia. Undis- turbed soil columns were incubated at 20 ~ for 68 days at near field-capacity water content, and N mineraliza- tion was measured in 5-cm depth increments to 30 cm. The soils represented three primary profile forms: grada- tional, uniform and duplex. They were sampled beneath mature native Eucalyptus sp. forest and from plantations of Pinus radiata of varying age (< 1 to 37 years). Several sites had been fertilized, irrigated, or intercropped with lupins. The soils ranged greatly in total soil N concentra- tions, C:N ratios, total P, and sand, silt, and clay con- tents. Net N mineralization for individual soil profiles (0-30 cm depth) varied from 2.0 to 66.6 kg ha -1 over 68 days, with soils from individual depths mineralizing from < 0 (immobilization) to 19.3 kg ha-1 per 5 cm soil depth. Only 0.1-3.1% of the total N present at 0-30cm in depth was mineralized during the incubation, and both the amount and the percentage of total N mineralized de- creased with increasing soil depth. N fertilization, addi- tion of slash residues, or intercropping with lupins in the years prior to sampling increased N mineralization. Sever- al years of irrigation of a sandy soil reduced levels of total N and C, and lowered rates of N mineralization. Consid- ering all soil depths, the simple linear correlations be- tween soil parameters (C, N, P, C : N, C : P, N : P, coarse sand, fine sand, silt, clay) and N mineralization rates were generally low (r<0.53), but these improved for total N (r = 0.82) and organic C (r = 0.79) when the soils were grouped into primary profile forms. Prediction of field N-mineralization rates was complicated by the poor cor- relations between soil properties and N mineralization, and temporal changes in the pools of labile organic-N substrates in the field. M.J. Connell (~) 9 R.J. Raison . R K. Khanna CSIRO Division of Forestry, PO Box 4008, Queen Victoria Terrace, Canberra, ACT 2600, Australia Key words N mineralization 9 Forest soils 9 Forest management 9 N cycling 9 N modelling Introduction N mineralization is an important process in forest soils, particularly in unamended soils where it may be the ma- jor source of N available for plant uptake. N-mineraliza- tion rates are usually low in undisturbed soils under ma- ture forests, but estimates range from 1 to 100kg ha -~ year -1 (Binkley and Hart 1989). Low rates of N mineralization often limit forest growth (Pastor et al. 1984; Nadelhoffer et al. 1985). The rates depend on many factors, including management practices such as clear felling, ripping, mounding, slash burning, and fertiliza- tion, which alter soil litter or slash horizons (Raison et al. 1987; Smethurst and Nambiar 1990). Such treatments af- fect the input of organic residues, and the biomass and activity of the microbial populations which are largely re- sponsible for N mineralization. A better understanding of N-mineralization processes can guide management in maintaining soil fertility and stability, and the efficiency and management of N-fertilizer programmes would be enhanced by a better ability to predict rates of N release from the soil and the forest floor. In situ incubation of soils can provide realistic mea- sures of N mineralization (Raison et al. 1987; Binkley and Hart 1989; Raison et al. 1992) but these methods are la- borious and therefore unsuitable for routine predictions of soil N availability. To be useful, any assay of soil avail- ability must be rapid and available. Much of the research to date has been performed by using either soil chemical properties or short-term biolog- ical measures (usually in disturbed samples) as indices of the soil N-mineralization potential (Powers 1980; Stan- ford 1982). In most of these studies the research has been concentrated on the surface few centimeters of agricultur- al soils, thus ignoring the contribution that deeper layers of the solum make to N cycling. Forest soils differ from agricultural soils in that, apart from the initial forest es-