Carbon-sink potential of continuous alfalfa agriculture lowered by short-term nitrous oxide emission events

Alfalfa (Medicago sativa) is a nitrogen-fixing perennial forage crop grown globally and extensively across the Western United States. It has been considered a climate-friendly feedstock due to deep roots that can sequester carbon and reduced need for synthetic nitrogen fertilizers. Yet, nitrogen inputs from biological fixation and soil processes can stimulate nitrous oxide (N2O) emissions via nitrification and denitrification, potentially diminishing the net climate benefit. Few studies have integrated continuous measurements of CO2, CH4, and N2O to produce full net annual CO2-equivalent budgets in continuous alfalfa systems, which may differ from short-rotation systems. N2O production is highly dynamic and characterized by brief, high-flux “hot moments” that short or infrequent sampling can miss. In alfalfa, frequent irrigation to meet high water demand and rainfall events can create short-term anaerobic conditions that trigger N2O bursts, with acidic soils further inhibiting N2O reduction to N2. Oxygen availability is a key control on N2O production pathways. This study aims to quantify greenhouse gas fluxes and drivers in a continuously managed alfalfa agroecosystem using multi-year, continuous observations, with a focus on the role of N2O hot moments and background emissions in shaping the ecosystem-scale carbon balance.

Prior work has highlighted alfalfa’s potential for soil carbon sequestration and reduced fertilizer inputs, but comprehensive, continuous budgets including CO2, CH4, and N2O are rare in alfalfa agroecosystems. The literature identifies that N2O emissions are temporally heterogeneous, with hot moments contributing disproportionately to annual totals, and that irrigation/rewetting events and low pH can enhance N2O production by promoting anaerobiosis and inhibiting N2O reductase. Oxygen availability strongly regulates nitrification and denitrification processes. Studies also suggest plant activity and root-derived substrates can influence background N2O emissions, while upland agricultural soils often act as small CH4 sinks unless prolonged anaerobiosis occurs. Compared to rotational systems, continuous alfalfa may have distinct flux dynamics due to persistent cover, frequent irrigation, and repeated cuttings.

Study site: Sacramento–San Joaquin Delta, California, USA (38.11°N, 121.5°W), in conventional perennial alfalfa (>5 years) on degraded peat (Ryde series; fine-loamy, mixed, superactive, thermic Cumulic Endoaquolls). Mediterranean climate with hot, dry summers and cool, wet winters; regional mean annual temperature 15.1 ± 6.3°C and rainfall ~326 ± 4 mm. Nearly all alfalfa in the region is irrigated; the study field was periodically flood-irrigated during the growing season. Automated chamber measurements: Continuous surface fluxes of N2O, CH4, and CO2 were measured from January 2017 to February 2021 using nine opaque automated chambers (eosAC) connected to a multiplexer (eosMX) and a Picarro G2508 cavity ring-down spectrometer. Chambers were deployed in a 10 × 10 m grid; five placed over plant rows and four over inter-row bare soil, remaining in place throughout the study except brief management periods. Each chamber measurement lasted 10 minutes with 1.5-minute pre- and post-flushes; chambers were sampled sequentially (~80 flux measurements per day). Extended 15 cm collars prevented inundation issues. Quality control and detection: Chamber collar heights were measured weekly to determine volumes and compute conservative minimum detectable fluxes: 0.002 nmol N2O m−2 s−1, 0.06 nmol CO2 m−2 s−1, 0.002 nmol CH4 m−2 s−1. Data were processed using eosAnalyze-AC and further filtered in R; fluxes with instrument malfunctions, abnormal deployment duration (<9 or >11 min), or high model uncertainty were removed. After filtering, the final dataset included 108,638 CO2, 103,013 N2O, and 102,997 CH4 flux measurements. Statistical analyses used one-way ANOVAs with Tukey post-hoc tests (JMP Pro 15). Ecosystem-scale CO2 fluxes: Annual net ecosystem exchange (NEE), ecosystem respiration (Reco), and gross primary productivity (GPP) were obtained from a nearby (<1 km) AmeriFlux eddy covariance site in alfalfa under identical management and soils. Open-path analyzers (LI-7500 for CO2/H2O; LI-7700 for CH4) and sonic anemometers measured fluxes at 20 Hz with regular calibration. N2O and CH4 fluxes were converted to CO2-equivalents (CO2e) using IPCC AR5 100-year GWP values: 28 for CH4 and 298 for N2O. Hot moments definition and quantification: Hot moments were defined as flux measurements >4 standard deviations above the mean for a given year. Annual mean N2O fluxes were computed for (1) all data, (2) hot moments only, and (3) with hot moments removed to quantify their contribution to annual totals. Weekly soil sampling: From April 2018 to May 2019, weekly composite sampling (0–15 cm, n = 10 per week within 30 m of chambers) measured gravimetric soil moisture, pH (1:1 soil:DI water), and 2 M KCl-extractable NO3−+NO2− and NH4+. Extracts (5:1 KCl:soil ODE) were shaken for 1 h, filtered (Whatman Grade 1), and analyzed colorimetrically (AQ300 analyzer). Soil moisture was determined by oven-drying at 105°C to constant mass. Soil sensor network: From September 2018 to February 2021, oxygen (O2) and soil temperature sensors (SO-110; Apogee Instruments) and volumetric water content sensors were installed at 10, 30, and 50 cm depths. Sensors logged at high frequency, summarized to hourly/daily means (n ≈ 96 measurements per day). Satellite-based near-infrared reflectance of vegetation (NIRv) from Planet Labs daily 3 m imagery was used as a proxy for canopy photosynthetic activity and potential plant inputs to soil. Ancillary data and analyses: Annual rainfall for site years (Jan 27–Jan 26) came from a nearby AmeriFlux site. Time-series and coherence analyses (including wavelet coherence) evaluated relationships among N2O fluxes, soil moisture, temperature, O2, pH, and NIRv across diel to seasonal timescales.

- The continuous alfalfa system was a net source of N2O averaging 624.4 ± 27.8 mg N2O m−2 y−1 across four years, which reduced the net CO2e sink by up to 14% annually at the ecosystem scale. - Hot moments were rare (≤1% of measurements) but disproportionately important, contributing on average 44.4% of annual N2O emissions and up to 57% in some years (2017–2018: 56.8%; 2018–2019: 55.3%; 2019–2020: 37.5%; 2020–2021: 31.6%). Hot moments were closely associated with irrigation and rainfall events and surface soil conditions (moisture, O2) and were enhanced under acidic pH. - Background (low-magnitude), continuous N2O emissions increased in relative importance over time and were positively associated with plant activity. Significant coherence was observed between NIRv (satellite-derived photosynthetic activity) and N2O fluxes at daily timescales, implicating plant-derived carbon or ammonium as regulators of background emissions. - Seasonal patterns: N2O fluxes were highest in summer (irrigation season) and lowest in fall. Daily mean N2O fluxes correlated with weekly soil-atmosphere N2O concentrations across depths (10 cm R2 = 0.60; 30 cm R2 = 0.53; 50 cm R2 = 0.45; all p < 0.001), indicating contributions from across the soil profile to background emissions. - Soil CH4 fluxes indicated a small, consistent net sink: −53.5 ± 2.5 mg CH4 m−2 y−1 on average, equivalent to only 0.06% of the total net carbon-based CO2e uptake over four years. Sustained anaerobiosis sufficient to drive net CH4 production was not observed; periods of elevated soil CH4 concentrations did not translate to surface emissions, likely due to diffusion limits and methanotrophic consumption near the surface. - CO2 fluxes: Chamber-based soil and root respiration averaged 4925.9 ± 13.5 g CO2 m−2 y−1 and tracked soil temperature seasonally. Eddy covariance estimates indicated Reco 6451 ± 12 g CO2 m−2 y−1, GPP 8745 ± 51 g CO2 m−2 y−1, yielding NEE −2330 ± 46 g CO2 m−2 y−1 (net sink). Combining chamber N2O and CH4 with eddy covariance NEE gave a field-scale total CO2e of −2115.4 ± 54.4 g CO2e m−2 y−1 (overall sink). - Accounting for biomass harvest: Using mean annual harvested carbon of 595 ± 137 g C m−2 y−1 (2072 ± 502 g CO2 m−2 y−1), equivalent to 89% of NEE, leaves ~258 g CO2 m−2 y−1 stored belowground. Under this accounting, N2O CO2e would offset about 70% of the net CO2 sink. - Interannual context: Rainfall varied markedly (e.g., 176–447 mm y−1). Mean 0–50 cm soil moisture declined over years (site year 4 driest). Mean annual NIRv decreased significantly over the study period, and cutting events produced clear NIRv and respiration responses with recovery within 5–7 days. - Wavelet coherence analyses showed N2O fluxes lagged and were in phase with changes in soil moisture, temperature, and O2 at daily to weekly scales; surface soil moisture and O2 dominated controls of hot moments, while persistent low pH and lagged moisture/temperature responses controlled longer timescales.

The study demonstrates that continuous alfalfa agriculture, despite acting as a net CO2 sink at the ecosystem scale, emits sufficient N2O to substantially erode its climate mitigation potential. By combining multi-year, high-frequency chamber fluxes with eddy covariance and satellite indicators of plant activity, the research resolves the temporal dynamics and drivers of N2O emissions that are typically missed by infrequent sampling. Rare but intense hot moments, primarily triggered by irrigation and rainfall that transiently reduce oxygen in surface soils and interact with acidic conditions and nitrate availability, contributed nearly half of annual N2O emissions. Concurrently, background emissions were modulated by plant activity, moisture, and temperature, as indicated by coherence with NIRv. These findings address the central question by quantifying how N2O emissions—both hot moments and background fluxes—reduce the net CO2e sink of continuous alfalfa systems. The small, consistent CH4 sink provided negligible compensation. The results underscore the importance of continuous measurements for accurate greenhouse gas accounting in agroecosystems and suggest that management practices influencing soil moisture dynamics, oxygen availability, and pH could meaningfully affect N2O emissions. The observed coupling and lags between plant phenology (harvest cycles), soil respiration, and N2O also highlight the role of plant-driven substrate supply in governing background emissions.

This multi-year, continuous assessment of a flood‑irrigated, continuous alfalfa agroecosystem shows that N2O emissions materially diminish the ecosystem’s carbon-sink function, reducing the net CO2e sink by up to 14% annually and potentially offsetting 70% of the net CO2 sink when accounting for harvested biomass removal. Hot moments, though infrequent, contributed disproportionately to annual N2O budgets, while background emissions tracked plant activity and soil microenvironmental conditions. The integration of automated chambers, eddy covariance, soil sensing, and satellite imagery provides a robust framework for capturing scale-emergent drivers of greenhouse gas fluxes in working landscapes. Future research should expand continuous, multi-gas monitoring across diverse alfalfa management regimes and climates to generalize these findings, incorporate the post-harvest fate of carbon into life-cycle assessments, and evaluate mitigation strategies (e.g., irrigation scheduling, soil pH management) that may reduce N2O hot moments without compromising productivity.

- The fate of harvested carbon was not included in the ecosystem-scale flux accounting; when considering harvest removal, N2O offsets a larger fraction of the net CO2 sink. - Weekly soil nitrogen and pH measurements covered approximately one year (April 2018–May 2019), potentially missing interannual variability in soil N pools. - Soil moisture records were incomplete for site year 2 (5 of 12 months represented). - NEE and other ecosystem-scale CO2 fluxes were obtained from a nearby (<1 km) eddy covariance tower rather than co-located measurements, albeit under identical management and soil conditions. - Results are derived from a single, flood‑irrigated alfalfa site on degraded peat in a Mediterranean climate, which may limit generalizability to other soils, climates, or irrigation practices.