Methodology and quantification of nitrous oxide emissions from grain sorghum production in humid and semi-arid environments using an automated trace gas analyzer system

Abstract

The Intergovernmental Panel on Climate Change (IPCC) approximates the global warming potential of N₂O at 300 times that of carbon dioxide. More than one third of all N₂O emissions are anthropogenic and are primarily due to agriculture. This study aims to compare N₂O emissions from sorghum fields under varying precipitation regimes and fertilizer management, to assess whether the IPCC default emissions factor of 1% overestimates these emissions and to evaluate how disaggregating this factor based on environmental and management factors could improve its accuracy. Field trials were conducted over six site years with a range of annual precipitation (350 mm -700 mm). All analyses evaluated six nitrogen application rates of streamed UAN (including a no nitrogen control), along with three additional product & placement combinations, evaluated both at planting and one month prior to planting. N₂O fluxes were measured using an N₂O / H₂O trace gas flux analyzer (Model 7820, LI-COR) and a portable automated chamber system (Model 8200-01, LI-COR). Data collected from chambers were processed using Soil Flux Pro software (Version 5.3.1, LI-COR). Nitrous oxide (N₂O) emissions have high spatial and temporal variability across the landscape, and agricultural management can introduce additional spatial variability, making quantification even more difficult. In addition to the main analysis of N₂O emissions, this study also focused on the methods used to collect N₂O flux estimates by evaluating sampling time and chamber placement. A study was conducted to identify the optimum sampling duration for different time intervals: 180 seconds, 120 seconds (recommended by the manufacturer), and 90 seconds, to identify an optimum sampling time and minimize time spent obtaining N₂O measurements while maintaining data quality. Chamber fluxes were analyzed using SAS Version 9.4 using PROC REG procedure. Regression analysis with all sites combined resulted in high r² (r² > 0.97) in each comparison group. However, with all sites combined, all slope values were different from one (p < 0.0019). Under the assumption that the longer duration is most reliable, the low RMSE values indicated that samples can be collected at the shorter sampling duration of 120-seconds without compromising data quality. The manufacturer’s recommended sampling duration of 120-seconds is likely a reliable balance between data quality and labor required for measurements. The analysis of chamber placement’s impact N₂O flux in banded (38 cm centers) and non-banded nitrogen fertilizer treatments was conducted by collecting N₂O flux from two chambers in each plot, one adjacent to the row and over the fertilizer band (on-band), and one centered in the inter-row area between the fertilizer bands (inter-row). In non-banded treatments, two chambers were placed in the same positions. Data were evaluated in two subsets, rate treatments and placement treatments. Statistical analysis was conducted using SAS Version 9.4. PROC MIXED procedures. When evaluating rate treatments, there was an interaction between chamber position and nitrogen fertilizer rate (p = 0.0043). At high rates (>100.8 kg ha⁻¹), the on-band chambers had higher flux values than the inter-row treatments (p < 0.05). On sampling days with high N₂O flux levels, differences between the two chamber placements were the largest. Since these two chamber placements had significant differences from one another, plots missing on-band chambers should be omitted from daily flux calculations to avoid underestimating or overestimating flux. Across all fertilizer placement treatments (broadcast urea with urease inhibitor – NBPT (N-(n-butyl) thiophosphoric triamide), streamed UAN, coultered UAN, and no nitrogen control) and at all sites, the interaction between fertilizer placement and chamber position was not significant at [alpha] = 0.05 but was significant at [alpha] = 0.1. Streamed UAN placements applied in a band at both application timings did show differences between flux values in the inter-row and on-band chambers (p = 0.0768). There was no statistical difference between the on-band and inter-row chambers in the broadcast urea with urease inhibitor - NBPT and no nitrogen control treatments. However, the mean flux for the on-band chamber overall was 0.45 nmol N₂O m⁻² s⁻¹ and the mean flux for the inter-row chamber overall was 0.29 nmol N₂O m⁻² s⁻¹ indicating a strong overriding influence of chamber position (p = 0.0001). Given the variability of collected results, plots with banded fertilizer application where either the on-band or inter-row chamber had missing data, should be omitted from analysis because of the potential of influence from the high concentration of fertilizer in one area. Chamber position as a main effect was highly significant when all sites were evaluated, meaning that regardless of fertilizer rate or placement relative to the chamber, there is enough inherent spatial variability that plots missing one of the two chambers should be omitted from the dataset. Analysis of cumulative N₂O emissions data show an influence of nitrogen application rate on cumulative N₂O flux at all evaluated site years (p < 0.0001). Increasing nitrogen fertilizer application rates resulted in higher cumulative N₂O emissions (p < 0.05). Overall, nitrogen fertilizer applied one month prior to planting did not show any differences between nitrogen fertilizer application methods. Sorghum grown in dryer environments generally did result in lower cumulative N₂O emissions compared to sorghum grown in wetter environments with similar nitrogen treatments. N₂O emissions from grain sorghum did not always peak within the first two weeks following nitrogen application, but a peak was observed within the first month following nitrogen application, typically following a rainfall event. Overall, increasing nitrogen rate did not result in increased N₂O emission factors and subsurface placement with a coulter or addition of a urease inhibitor did not clearly reflect a reduction in emissions factors in evaluated environments. The EF values calculated, specifically those from the more characteristic rainfall year, 2025, are largely in-line with IPCC’s disaggregated EF value for arid and semiarid environments of 0.5 percent (IPCC, 2019).