Precipitation and soil N fertilization are factors that drive emissions of the greenhouse gases (GHG) nitrous oxide, a major contributor to climate change from agriculture (IPCC, 2007), and carbon dioxide. The C and N exchange with the atmosphere following rewetting by precipitation or irrigation after the period of sustained drought is known collectively as the Birch effect (Jarvis et al., 2011). Both N₂O and CO₂ emissions from soil are affected by factors such as temperature and water regimes, soil N status, and microbial activity due to carbon and nitrogen moving through ecosystems in coupled biogeochemical cycles (Li et al., 2005). Complexity of interactions between these factors results in high spatial and temporal variability of GHG emissions (Choudhary et al., 2002). Furthermore, management histories such as tillage practices can significantly impact all of these factors, as well as N and C cycling and storage by influencing inputs of N and C materials and the soil organic matter decomposition regime (Li et al., 2005; Smith et al., 2008). Assessment of factors affecting N₂O flux from soils is therefore fundamental to determining the environmental conditions promoting N₂O emissions (Forster et al., 2007).

Nitrogen fertilization and water additions have been reported to increase N₂O emissions (Barton et al., 1999). Nitrogen fertilizer is often applied to meet the season-long demand of crops, consequently, N levels are commonly in excess of crop demand at the time of application. Application of N in excess of crop N requirements tends to exponentially increase N₂O emissions (Shcherbak et al., 2014). Soil aeration status is related to the amount of water filled pore space (WFPS), as well as soil texture and structure, which regulate the movement of water and oxygen, thereby affecting the relative rates of nitrification and denitrification in soil (Wrage et al., 2001; Saggar et al., 2013). Emissions of N₂O generally increase at 60–80% WFPS depending on soil texture and structure, with coarse-textured and well-structured soil requiring greater water additions than finer textured soils with little structure (Barton et al., 1999; Barton and Schipper, 2001). Emissions of CO₂ increase immediately following water additions; however, N fertilization can have short-term reducing effects on CO₂ emissions due to reduced enzyme activity and pH (Kowalenko et al., 1978; Wilson and Al-Kaisi, 2008; Gagnon et al., 2016). Previous research showed single events of very high fluxes of N₂O and spikes of CO₂ fluxes after N additions following rewetting, demonstrating importance of short-term events on overall GHG budgets (Nobre et al., 2001; Gelfand et al., 2015). Episodic production of N₂O and CO₂ associated with artificial or natural rainfall and N fertilization events emphasize the need to assess short-term emissions of N₂O and CO₂ from agricultural soils as influenced by different management histories.

No-tillage (NT) management is a conservation practice, which can sequester soil carbon, act as a sink of atmospheric CO₂, preserve soil water and reduce erosion, while its effects on greenhouse gas emissions is less well-known. The N₂O emissions were shown to generally stay the same following conversion from conventional tillage (CT) to NT in coarse-textured and fine-textured well-aerated soils, but tend to increase in fine-textured, poorly aerated soils, primarily due to denitrification (Six et al., 2002; Rochette, 2008). Longer-term adoption of NT has been reported to result in reduced contributions to GHGs through lower N₂O emissions from soil (Six et al., 2004). Effects of tillage on CO₂ emissions vary (BilandŽija et al., 2016), while being mostly dependent on soil moisture and temperature (Chavez et al., 2009). Due to existing controversy on the long-term effects of no-till on climate mitigation (Halvorson et al., 2008; Rochette, 2008; Stockle et al., 2012; Sainju et al., 2014; Bayer et al., 2016), field-level analysis is needed to address the vulnerability of GHG emissions to precipitation and N fertilization events as impacted by tillage management.

We conducted a study on short-term GHG emissions (CO₂ and N₂O) immediately following water and N additions in long-term CT and NT research sites representing dryland cropping as well as a more recently established irrigated site within the inland Pacific Northwest (iPNW). Our objectives were to (i) compare short-term N₂O and CO₂ emissions in CT and NT immediately following water and N fertilization additions; and (ii) assess factors affecting event-driven N₂O emissions in cropping systems representative of the iPNW.

Field experiments were conducted between July 3 and July 28, 2013 in the long-term studies with tillage treatments, located in dryland and irrigated cropping regions of the iPNW. The summer timing for experiments was chosen as the highest levels of soluble soil C and total C are found in soils during that period, as well as potential denitrifier activity is increased (Shrewsbury et al., 2016). Summers are typically hot and dry in this Mediterranean-like climate and soils are vulnerable to high spikes of GHG emissions during rewetting events, which maximize expected treatment effects in contrasting long term tillage sites. The sites were located in Sherman Research Station, located in Moro, OR (45.48° N, 120.69° W), Columbia Basin Agricultural Research Center in Pendleton, OR (45.44° N, 118.37° W), Irrigated Agricultural Research Center in Prosser, WA (46.29° N, 119.74° W), and Kambitsch Farm of the University of Idaho in Genesee, ID (46.58° N, 116.95° W). The sites represent four agro-ecological classes (Kaur et al., 2017): (i) annual cropping at Kambitsch Farm in Genessee, ID, (ii) annual crop-fallow transition at Columbia Basin Agricultural Research Center in Pendleton, OR, (iii) grain-fallow at Sherman Research Station in Moro, OR, and (iv) irrigated at Irrigated Agricultural Research Center in Prosser, WA, and included tillage and no-tillage treatments established between 1982 and 2011 (Morrow et al., 2016) and were all planted to winter wheat (Triticum aestivum L.) at the time of monitoring.

We implemented an automated CO₂ and N₂O measurement mobile setup, which allowed the flexibility of moving equipment between sites. The system included Li-Cor 8100A automatic chambers coupled with LGR 23r N₂O analyzer for continuous monitoring of CO₂ and N₂O emissions in a short-term micro-plot study. More details on the instrumentation setup can be found in Kostyanovsky et al. (2017). The experimental design at each site was a split-plot design with 3–4 replications. The main plot was the tillage treatment (CT and NT), and subplots 1 × 1 m² were water addition and fertilizer treatments. The treatments applied at each site were: (1) no water (DRY); (2) water added approximately to 80% WFPS (H2O); and (3) water added to 80% WFPS and amended with 150 kg N ha⁻¹ as NH₄NO₃ (H2O+N). The water addition mimicked a 59 mm rainfall event, achieved by distributing 59 L of water over the 1 × 1 m plot area. No observable runoff occurred during the water addition. Ammonium nitrate fertilizer was applied to ensure ammonium and nitrate are not limiting factors to the N₂O emissions from treatments by allowing nitrification and denitrification processes to occur at the same time. Application of N and water took place at 9:00 a.m. and the measurements continued from that time until 7:00 a.m. the following day for a total of 22 h at the 2 h interval, for the total of 11 measurement cycles per plot. Our studies showed that GHG emissions were highest during the first day and decreased on the second and third days following water and N additions (Figure 1); therefore, the first 22 h appeared adequate for capturing diurnal dynamics and for treatment comparison, as it captured the highest N₂O flux (Trost et al., 2013; Leitner et al., 2017; Mumford et al., 2019). Soil temperature (T, °C) and volumetric water content (θ, m³ m⁻³) at the depth of 20 cm was measured continuously at each plot with 5TM temperature and moisture sensors (METER, Pullman, WA, USA) to monitor any changes in soil conditions that could affect results. The study was conducted at the end of the vegetative period for wheat (Triticum aestivum, L) during the dry summer period with high daytime temperatures typical of the Mediterranean climate in July 2013. Greenhouse gas response to applied N and water would be expected to be maximal at this time (Shrewsbury et al., 2016).

Soil sampling and analyses were conducted to aid assessment of factors that could affect N₂O and CO₂ emissions. Soil analytical methods and results are fully described and presented in Morrow et al. (2016). Briefly, soil was sampled at each site in CT and NT treatments at 0–10 cm depth in July 2013, just prior to the GHG chamber study, and analyzed for bulk density, pH in 1:1 DI water extract, total organic carbon (TOC), and total nitrogen via dry combustion, and 1M KCl extracted nitrate (NO3--N) and ammonium (NH4+-N) (Table 1). Furthermore, soil C and N fractions were analyzed including non-hydrolyzable C and N, hydrolysable C and N, microbial biomass C and N, and water extractable C and N. Potential carbon mineralization was measured at 3, 10, 17, and 24 days, anaerobic N mineralization was determined at 28 days, and ion exchange membrane NH4+ and NO3- were measured after 24 h (Morrow et al., 2016).

The change in concentration of the N₂O mole fractions was determined via linear regression, and the R² values were computed in Matlab R2011a (The Mathworks, Inc., Natick, MA). The first 30 s of N₂O concentrations in the 8 min (480 s) measurement were eliminated from regression analysis, following the standard Li-Cor 8100A algorithm for calculation of the CO₂ fluxes.

The GHG emissions and soil water and temperature data were analyzed for the effects of time, treatment and tillage system by the proc mixed procedure using Satterthwaite method in SAS (SAS 9.3, Cary, NC). Treatment effects at each time step were determined via the Differences of Least Squares Means analysis (P < 0.05 within proc mixed procedure). The effects of soil water and temperature on N₂O and CO₂ emissions were determined via stepwise multiple regression in proc reg procedure of SAS (SAS9.3, Cary, NC). Stepwise multiple regression was also used to determine which soil carbon and nitrogen fraction, potential C and N mineralization, and ion exchange membrane nitrogen variables best describe the cumulative N₂O and CO₂ flux over 22 h in prog reg (SAS 9.3, Cary, NC). Soil pH, TOC, TN, NO3-N, and NH4+-N were analyzed for treatment and tillage effects in SAS by proc mixed procedure with Tukey-Kramer difference of least squares means adjustment (SAS 9.3, Cary, NC).

Overall, the current study showed that emissions of CO₂ increase during short-term events with increase in moisture and rise of temperature during daytime, and decrease under NH₄NO₃ application; however, the spike in N₂O emissions is likely to eliminate the partial reduction of GHG emissions. Our results suggest that NT could reduce N₂O emissions compared to CT during the short-term irrigation and N fertilization events; however, this effect was likely compounded by the site effect, which significantly includes crop rotation and the soil tillage intensity rating (STIR). The active fraction of soil organic matter plays a crucial role in soil's ability to reduce the N₂O emissions. Our data shows that further studies are needed to investigate the effects of the early stages of no-till adoption as well as long-term no-till on spikes of N₂O in dry conditions followed by short-term artificial or natural rainfall events.

KK designed the study layouts at the sites, assembled the N₂O and CO₂ analyzer setups, conducted greenhouse gas emissions measurements, analyzed data and wrote the manuscript. DH as a principal PI on the USDA-NIFA project, obtained funds to supply instrumentation and workforce to conduct the study, located the long-term tillage, no-till sites within the irrigated and dryland management systems in inland Pacific Northwest, discussed the details of the study design, and data analysis with the corresponding author, edited the manuscript. CS as a PI on the USDA-NIFA project, obtained funds to supply instrumentation and workforce to conduct the study, discussed the treatments in order to obtain contrasting N₂O and CO₂ emissions effects during the study, provided suggestions for data analysis, and edited the manuscript. JM assisted in conducting greenhouse gas emissions measurements, collected and analyzed soil samples from respective treatments, discussed data analysis and edited the manuscript. IM designed the irrigated long-term sites, assisted in conducting the GHG measurements, discussed data analysis and edited the manuscript.