Nitric oxide flux from soil during the growing season of wheat by continuous measurements of the NO soil–atmosphere concentration gradient: A process study

Abstract

The surface flux of nitric oxide from a wheat field was investigated from 23 March to 29 May 1997 in the Kerzersmoos, Switzerland. A plot fertilised with 19 kg N ha-1 in cattle slurry and 40 kg N ha-1 in mineral NH4NO3 fertiliser and a plot receiving no nitrogen containing fertiliser were compared. The flux was calculated based on hourly measurements of the NO soil–atmosphere concentration gradient using the one-dimensional soil diffusion model of Galbally and Johansson (1989). The soil bulk diffusion coefficient was determined from measurements of the 222Rn surface flux and the activity gradient between 10 cm depth and the surface. It ranged between 79% and 0.3% of the NO diffusion coefficient in air and was parameterised by air filled soil pore space. The indirectly determined NO flux agreed well with standard flux measurements using dynamic chambers. The largest NO emission was found following fertiliser application and irrigation. The emission occurred in pulses, which lasted for 4 days up to 3 weeks coinciding with elevated soil ammonium concentrations. Nitric oxide emission in 5 days following application of cattle slurry were 31 g NO-N ha-1 and 5 g NO-N ha-1 from the non-fertilised plot, respectively. Nitric oxide emission in 15 days following application of NH4NO3 was 95 g NO-N ha-1 and 10 g NO-N ha-1 from the non-fertilised plot, respectively. NO emission in 4 days following irrigation on 21 April were 36 g N ha-1 from the fertilised and 39 g N ha-1 from the non-fertilised plot. The daily NO emission before and after fertiliser and irrigation pulses was between 0.3 and 0.7 g NO-N ha-1 d-1. NO production and NO uptake of the soil was measured regularly. No systematic influence of management or climate on NO uptake was found. NO production was strongly stimulated by fertiliser input and soil moisture content. The simulation of NO production could be reproduced using a nitrification algorithm (Riedo et al., 1998) driven by soil temperature, moisture and ammonium concentration. A NO production rate constant of 1.1ċ10-3 h-1 at 15 °C was derived from a linear regression between nitrification and NO production. Introducing the parameterisation of NO production into the model of Galbally and Johansson (1989) the duration and the strength of the NO emission pulses could be reproduced and the total NO emission during the experiment was approximated within a factor of two.