Bergsma, T.T.
Presented at the All Scientist Meeting (1999-07-20 to 1999-07-21 )
We are trying to develop a comprehensive understanding of the relationships between N2O and N2 flux from agricultural and forest soils. Both gases are products of denitrification but N2O is also a product of nitrification and perhaps other processes. Our approach is to apply 15N-labeled material to soil, and then monitor production (or consumption) of N2 and N2O in a soil cover headspace. N2O may be determined by gas chromatography (GC) or by isotope ratio mass spectrometry (IRMS). N2, however, may only be determined by mass spectrometry under these conditions. If only denitrifiers contribute to N-gas production, and if consumption of N-gas in soil is negligible, then the results of the two methods for N2O analysis should agree.Data collected over the last 16 months from agricultural soil (corn stubble, LTER bulk plots) and forest soil (deciduous, Long Woods) indicate that for N2O, fluxes measured by GC and IRMS rarely agree. Usually IRMS values are much lower than GC values, which strongly suggests (for mathematical reasons) that flux derives from a non-uniformly labeled pool of soil mineral N. Sometimes GC values are lower than IRMS values, suggesting that significant consumption of N2O is co-occurring with N2O production during the incubation.In April 1999, fertilizer solution or water (control) was added to 16 plots (4 treatments x 4 replicates) in Long Woods. In the main treatment, 15N-labeled nitrate (99%) was added at a rate of 45 kg-N ha-1 while unlabeled ammonium was added at a rate of 70 kg-N ha-1. In one variant of the main treatment, ammonium was labeled (99%) and nitrate was unlabeled. In another variant, labeled nitrate was applied at a lower rate (4.5 kg-N ha-1) but at the same enrichment. Incubations were conducted immediately after (~ 2 h) and 14 d after additions. Samples were analyzed by GC and by IRMS.Figure 1 shows apparent enrichment of the soil mineral N pool for each incubation, inferred from the distribution of masses of N2O in the sampled headspace. In most cases, IRMS flux values were less than GC flux values (Figure 2) indicating that the apparent enrichments are overestimates, perhaps representing only the labeled portion of a two-pool system. Still, it appears that N2O from labeled ammonium is present as early as two hours after addition, suggesting tightly coupled nitrification and denitrification in this soil. Enrichment of the soil N pool in labeled-nitrate plots was initially similar to added nitrate, but dropped rapidly over a two-week interval, probably due to dilution by nitrification. In contrast, enrichment of N2O in labeled-ammonium plots rose with time. Probably most of this N2O was produced by denitrification of soil nitrate which was continually being enriched by nitrification of the ammonium pool. Not surprisingly, enrichment dropped more rapidly for plots with low rates of labeled nitrate application than for plots with high rates of labeled nitrate application. The overestimation represented by flux ratios greater than 1 (Figure 2) may be an artifact of reduced analytical sensitivity due to very low enrichment in these plots.
Back to meeting | Show |