Dynamics of nitrogen between organic and inorganic phases in soil

Accurate prediction of mineral N dynamics in soil is important for sustainable management of land resources towards stable plant productivity, least greenhouse gas (N20) emission, and minimum water system contamination. Our ability to achieve this is largely limited by inadequate understanding of soil N cycling processes in relation to substrate quality and quantity, soil properties and the environmental conditions. This research aimed to improve our understanding of the role of the basic N turnover processes in determining mineral N accumulation in soils, and to look into the biologically labile fractions of soil organic N as estimated-with physical, chemical, biological, isotopic equilibrium, and long-term incubation techniques. Gross N mineralisation, immobilisation in and loss from soils during waterlogged and aerobic incubation were differentiated and quantified using 15N isotope dilution techniques and five different types of analytical equations. The relative magnitude of gross N mineralisation -between waterlogged and aerobic incubations was soil-dependent, whereas gross immobilisation tended to be greater under aerobic than under waterlogged conditions. The total amounts of N consumption by immobilisation and loss were equivalent to 17 - 90 % of the mineralised N during waterlogged incubation and 23 - 59 % during aerobic incubation. Several indices of N mineralisation capacity were evaluated using the gross N mineralisation as a reference criterion, including total organic N, water-soluble organic N, NaOH-hydrolysable N, hot HC1-hydrolysable N, hot KC1-hydrolysable N, microbial biomass, and N in the light organic matter fraction. All indices except light fraction N were significantly related to gross N mineralisation during the 2-wk waterlogged and aerobic incubations of rewetted dry soils. However, none of the chemically hydrolysed N fractions showed closer relationships with N mineralisation than did total or water-soluble organic N, suggesting that these chemical methods could not selectively extract a biologically labile fraction of soil organic N. The isotopic equilibrium technique for estimating active organic N was assessed in a long-term (333 d) incubation following addition of 15N-labelled NH4+ and glucose at different C/N ratios. The estimated "active organic N" in soil at each periodic sampling (OT*) and in the original soil (OT#) increased with incubation time and was affected by the amounts of labelled N and available C addition. Considerable amounts of mineralised N were derived from other than OT* fraction of organic N, but N mineralisation from the OT* pool was much faster than from the bulk soil organic matter. The results supported the concept that soil organic N exists in a continuum of degradability. A more standardised procedure and suitable formulae were proposed for determining 074 as a relative index of the active organic N in soil. Eighteen soils were incubated at 35 C and 55 % water-holding capacity for 41 wk and two soils at different temperature and moisture combinations for 29 wk to examine the concept of potentially mineralisable N (N0) and its determination procedures. It was shown that N0 and the mineralisation rate constant k estimated with the conventional model-fitting technique varied significantly with incubation time, temperature, and moisture in unpredictable manners. These problems persisted with the double exponential model that divides the mineralisable organic N into an active (Na) and a slow (Ns) pool, although it accommodated the initial mineralisation flush very well. Modifications of this technique were proposed to explicitly estimate soil-specific organic N pools and their temperature- and moisture-dependent rate constants, including (i) use of standardised incubation temperature (35 C), moisture (55 % WHC, field capacity, or 30 kPa), and length of time (>20 wk); (ii) use of fixed rate constants in the non-linear regression (k - = 0.054 wk-1 for N0, or k - a = 0.693 for Na and k - s= 0.054 wk-1 for Ns) to estimate the pool sizes with data obtained under the optimum conditions; and (iii) use of fixed N0, or Na and NS values to estimate their rate constants under non-standard conditions. The pool sizes obtained with the modified technique provided unequivocal measures of soil N mineralisation capacity, and the rate constants showed consistent and close relationships with temperature and moisture. Using the parameters. estimated from laboratory incubation, satisfactory prediction of outdoor net N mineralisation was achieved. The No pool could reliably be estimated with water-soluble organic N and microbial biomass in soil samples, but estimation of NS needs a combination of several indices including net N mineralisation from day 14-28