Entrapment of Nitrogen-15 Dinitrogen during Soil Denitrification

Denitrification has been proposed to be a major N loss mechanism from flooded systems, but direct measurement of the N, product has met with little success. A laboratory experiment was undertaken to determine if N-labeled dinitrogen gas was encapsulated in the saturated soil after applied N was denitrified in a Crowley silt loam (Typic Albaqualfs) soil. Under laboratory conditions, 28% of the applied "N-urea and 40% of the N-KNO3 was trapped as N2 in the soil 33 d after N application. In addition, 77% of the applied "NO;r and 44% of the "N-urea was denitrified and recovered as N2. Additional Index Words: Stable N isotopes, N loss, Oryza saliva L., N transformations. F RICE (Oryza saliva L.) uses applied N inefficiently and recovery of fertilizer N rarely exceeds 40% (Craswell and Vlek, 1979; DeDatta, 1981). Ammonia (NH3) volatilization and/or denitrification can be major N loss mechanisms from the flooded soil-rice system (Patrick, 1982; Fillery et al., 1984). In the past, N losses due to NH3 volatilization and denitrification (N2O, N2) have been determined from the difference between fertilizer N additions, N uptake by crops, N leaching and residual soil N. Nitrogen balance studies were used as an indirect measure of N losses. Direct measurement of NH3 volatilization and denitrification have been restricted due to problems involved in making accurate field measurements and availability of high precision instrumentation (Freney et al., 1983; Ryden and Rolston, 1983; Parkin et al., 1984). Micrometeorological techniques are now being used to quantify NH3 volatilization from flooded rice systems. These techniques are preferred because they do not disturb soil or environmental processes which affect gas exchange (Denmead, 1983; Fillery et al., 1984; Fillery and S.K. DeDatta, 1986). Very few field experiments have quantified the rates and absolute amounts of denitrification. Direct measurement of N2 and N2O has been achieved in nonflooded systems using gas chromatographic techniques (N2O) and isotope ratio mass spectrometers to measure the N2, N2, and N2 intensities (R.olston et al., 1976; Mulvaney and Kurtz, 1982; Mosier et al., 1986; Mulvaney and Boast, 1986). Published results on direct measurement of measurable N2 denitrification fluxes from flooded field systems is severely lacking at this time. Lindau and Patrick (unpublished data, 1987) were unable to measure denitrification N2 fluxes from C.W. Lindau, W.H. Patrick, Jr., R.D. DeLaune, Lab. for Wetland Soils and Sediments (LWSS), Louisiana State Univ., Baton Rouge, LA 70803-7511; K.R. Reddy, Univ. of Florida, Gainesville, FL; and P.K. Bollich, Rice Res. Stn., Louisiana. Agric. Exp. Stn. Joint contribution from the Lab. for Wetland Soils and Sediments, Louisiana State Univ., and the Univ. of Florida. Received 14 August 1987. 'Corresponding author. Published in Soil Sci. Soc. Am. J. 52:538-540 (1988). flooded rice systems in field experiments conducted during the 1985, 1986, and 1987 growing seasons using N mass spectrometer techniques (Siegel et al., 1982). No detectable increases in the N2 and N2 concentrations above natural background levels were observed. Closed chamber techniques using highly labeled N-urea (30, 64, and 99 atom % N), collection times up to 24 h and headspace heights minimized to 2-3 cm were tried. The objective of this study was to assess the possibility some fraction of the N2 being produced during denitrification was being trapped beneath the floodwater in the saturated soil layers. Materials and Methods The soil used for the laboratory experiment was a Crowley silt loam (Typic Albaqualfs). It contained 7.0 g total C kg~' and 0.8 g total N kg '. Its cation exchange capacity was 9.4 cmolc kg" of soil, and it had a pH of 5.8 (1:1, soil/water). The soil contained 10.8% clay, 70.7% silt, and 9.5% sand. Laboratory Procedures Large Pyrex® tubes (45.5-cm length by 5.0-cm i.d.) were fitted at the top with 34/45 ground glass joints and a stopcock and rubber septum for gas sampling (Fig. 1). Two hundred and fifty g of air-dried soil ground to pass through a 0.50-mm mesh sieve were placed in the tubes and lightly tapped until a 13-cm soil column formed. Distilled water was added to the soil from the bottom to prevent air entrapment, and a 2.0-cm floodwater layer established. The soil was incubated in the dark under atmospheric conditions for two months until a distinct aerobic-anaerobic interface developed (Reddy and Patrick, 1984). Labeled urea (64 atom % N) and KNO3 (71 atom % N) were dissolved and added to the column floodwater at a rate of 120 kg N ha~'. The column tops were sealed and the soil allowed to incubate for periods of 5, 10, 20, or 33 d at room temperature (2426 °C). A manometer was fitted to one column to monitor pressure changes within the sealed system over the incubation period. Only slight variations in pressure were observed and were attributed to minor changes in laboratory temperature. At the end of each incubation duplicate columns were removed for analysis. A gas-tight syringe was