RESUMO
Understanding the characteristics of the production of nitrogen gases (N2, N2O and NO), CO2 and CH4 in anaerobic paddy soils is not only a prerequisite for an improved mechanistic understanding of key microbial processes involved in the production of atmospheric greenhouse gases (GHG), but might also provide the basis for designing greenhouse gas mitigation strategies. Moreover, quantifying the composition fractions of denitrification gaseous products is of key importance for improving parameterization schemes of microbial processes in process-oriented models which are increasingly used for assessing soil GHG emissions at site and national scales. In our experiments we investigated two sandy loam soils from two paddy fields. The initial concentrations of soil nitrate and dissolved organic carbon (DOC) were set at approximately 50 mg.kg-1 and mg.kg-1, respectively, by adding a mixture solution of KNO3 and glucose. The emissions of N2, N2O NO, CO2 and CH4, as well as concentrations of carbon and nitrogen substrates for each soil sample were measured simultaneously, using a gas-flow-soil-core technique and a paralleling substrate monitoring system. The results showed that the accumulative emissions of N2, N2O and NO of the two soil samples for the entire incubation period were 6 - 8, 20, and 15 - 18 mg.kg-1, respectively. By measuring the cumulative emissions of denitrification gases (N, = N2 + N2O + NO) we were able to explain 95% to 98% of observed changes in s1ifr nilrate concentrations. The mass fractions of N2, N2O and NO emissions to Nt were approximately 15% -19%, 47% -49%, and 34% -36%, respectively. Thus, in our experiments N2O and NO were the main products of denitrification for the entire incubation period. However, as the temporal courses of hourly or daily production of the denitrification gases showed, NO production dominated and peaked firstly, and then N2O, before finally N2 became the dominant product. Our results show the high temporal dynamic of denitrification end products and this knowledge is of crucial importance for model development, since so far existing models assume a fixed fraction of denitrification end products.