Unlocking bacterial potential to reduce farmland N₂O emissions

The Haber–Bosch process removed historical nitrogen limitations to crop production but led to widespread nitrogen enrichment of agroecosystems, causing ammonia volatilization, nitrate leaching, eutrophication and increased emissions of nitrous oxide (N₂O), a potent greenhouse gas. Direct and indirect N₂O emissions from agriculture contribute substantially to the rise in atmospheric N₂O. Traditional mitigation focuses on improving nitrogen-use efficiency via policy and technology, yet further reductions may be achieved by manipulating soil microbiota. Because NosZ (N₂O reductase) is the only biological sink for N₂O, enhancing NosZ activity in soils could lower N₂O emissions. The study evaluates whether mass-inoculating soils with a non-denitrifying N₂O-respiring bacterium (NNRB), Cloacibacterium sp. CB-01, vectored and grown in biogas digestate, can effectively and durably reduce field N₂O emissions. It investigates CB-01’s respiratory phenotype, biokinetics, survival and persistence in soils, performance across soil types and pH, and the potential European-scale mitigation impact.

Soil N₂O is produced by diverse microbial processes (denitrification by bacteria and fungi, and nitrification pathways by archaea and bacteria), but only organisms expressing NosZ reduce N₂O to N₂. Many denitrifiers possess truncated pathways, prompting discussion of modular denitrification versus complete-pathway organisms. Liming acid soils can enhance NosZ expression and reduce N₂O by 10–20% but may have neutral climate benefits due to CO₂ from lime. Increasing the abundance of N₂O-respiring bacteria (NRB) can decrease emissions; NRB with full denitrification pathways can act as N₂O sinks depending on regulation and electron allocation, whereas organisms with nosZ but lacking nir genes (NNRB) are more consistently N₂O sinks under hypoxia/anoxia. Prior work showed that NRB/NNRB can be enriched ex situ in digestates and that digestate can serve as both substrate and vector, but earlier isolates often possessed complete denitrification gene sets and were better suited to digestate than soil. A dual substrate enrichment (alternating sterilized digestate and soil) yielded Cloacibacterium-dominated enrichments and isolated CB-01, an NNRB with genes for NO and N₂O reduction but lacking genes for NO₃⁻/NO₂⁻ reduction; meta-omics suggested surface attachment and complex polysaccharide utilization supports fitness in soil.

• Organism and growth: Cloacibacterium sp. CB-01 was cultivated in batch culture in GranuCult nutrient broth (pH 7.3, 23 °C). For field applications, CB-01 was grown aerobically in autoclaved biogas digestate, reaching ~6 × 10⁹ cells mL⁻¹ in ~2 days (oxygen consumption and qPCR-based yields confirmed growth). Approximately 1% of digestate organic carbon was consumed by 20 h; growth then slowed as substrates depleted.
• Respiratory phenotype and kinetics: Genome analysis showed nosZII present; genes for dissimilatory NO₃⁻/NO₂⁻ reduction absent. Experiments confirmed CB-01 reduces N₂O to N₂ under anoxia/hypoxia and does not produce N₂O from nitrite under tested conditions; minimal NO reductase activity inferred from NO kinetics. Oxygen repression experiments indicated N₂O respiration initiates only at low O₂: <1–2 µM O₂ for aerobically raised cells and 4–6 µM O₂ for cells previously exposed to anoxia (with intact NosZ). Apparent half-saturation constants were estimated by nonlinear regression: Kₘ,O₂ = 0.9 µM O₂ (s.e. 0.27) and Kₘ,N₂O = 12.9 µM N₂O (s.e. 1.2). Maximum growth rates from respiration during unrestricted growth: µmax,O₂ = 0.29 h⁻¹ (s.d. 0.006), µmax,N₂O = 0.11 h⁻¹ (s.d. 0.001). From growth yields, Vmax,O₂ ≈ 0.72 fmol O₂ cell⁻¹ h⁻¹ and Vmax,N₂O ≈ 0.66 fmol N₂O cell⁻¹ h⁻¹; electron flow: Vmax,e–O₂ ≈ 2.9 fmol e⁻ cell⁻¹ h⁻¹, Vmax,e–N₂O ≈ 1.3 fmol e⁻ cell⁻¹ h⁻¹. Growth yield on N₂O was 85% of that on O₂, consistent with higher energy conservation via NosZII. Bet-hedging behavior (only a fraction of cells express NosZ upon O₂ depletion) was inferred from gas kinetics and modeled with FNosZ ≈ 0.03 in rich medium; this effect disappeared when CB-01 was grown in digestate (all cells switched anaerobically on O₂ depletion).
• Field bucket experiments: Outdoor bucket trials with agricultural soils. Initial experiment (14 July 2021) used a clay loam (pH 6.7), sown with ryegrass, with daily water additions for 10 days. Buckets were fertilized by mixing digestate containing live CB-01 (≈6 × 10⁹ cells mL⁻¹) into the upper soil and compared to control digestate where CB-01 had been heat-killed (70 °C, 2 h) to match C and N inputs. Buckets were re-fertilized three times with autoclaved, pH-adjusted digestate without CB-01 (days 19, 33, 89). N₂O fluxes were measured with a dynamic chamber system (3 min enclosure) operated by a field robot; soil temperature (0–5.5 cm) and water-filled pore space (WFPS) were logged (n = 4). Emission reductions were computed on cumulated flux with 95% confidence intervals.
• Multi-soil bucket experiment: Replicated with four soils (17 September 2021): organic-rich clay loam (pH 5.26; 15.8% SOC), neutral clay loam (pH 6.70; 3.21% SOC), sandy silt (pH 4.15; 0.75% SOC), and low-pH clay loam (pH 4.50; 3.23% SOC). Same live vs heat-killed CB-01 digestate treatments; n = 6 replicate buckets per treatment; emissions monitored as above.
• Field plot experiment: 0.5-m² plots (limed in 2014; average pH(CaCl₂) 6.13 ± 0.10) were fertilized on 20 August 2022 by mixing digestate with live or dead CB-01 into the upper 10 cm (n = 6 plots per treatment). High-frequency measurements during the first 10 days captured diurnal flux variations; lower-frequency monitoring extended to 290 days; temperature and WFPS logged.
• CB-01 survival: Abundance quantified by qPCR with CB-01-specific primers during the long-term bucket experiment and in a separate laboratory incubation where CB-01-digestate was added to wet soil (0.53 mL g⁻¹ dry soil). In buckets, additional digestate (without CB-01) was applied two days prior to each qPCR sampling. Exponential decay models (Nt = N0 e^–δt) were fitted to estimate apparent first-order death rates and half-lives (T1/2 = ln2/δ).
• Microbiome impact and biosafety: 16S rRNA gene amplicon sequencing (excluding the CB-01 OTU) compared community composition across treatments; presence of antibiotic resistance or pathogenicity genes in CB-01 was assessed bioinformatically (none identified).
• European-scale assessment: The GAINS model estimated 2030 national anthropogenic N₂O emissions and potential reductions, assuming a uniform IPCC emission factor (1% of N applied emitted as N₂O) and a conservative 60% reduction for emissions from liquid manure applications when using NNRB; scenarios extended to other organic wastes and combined applications.

• CB-01 physiology and kinetics: Genome carries nosZII and lacks genes for dissimilatory NO₃⁻/NO₂⁻ reduction, confirming a non-denitrifying N₂O-respiring phenotype (NNRB). Kinetic parameters: Kₘ,O₂ ≈ 0.9 µM; Kₘ,N₂O ≈ 12.9 µM (≈389 ppmv at 15 °C). Vmax,N₂O ≈ 0.66 fmol N₂O cell⁻¹ h⁻¹; µmax,N₂O ≈ 0.11 h⁻¹; growth yield on N₂O ≈ 85% of that on O₂. CB-01’s catalytic efficiency (Vmax/Km) is low relative to other N₂O-respiring organisms despite average Vmax, indicating poor performance at low N₂O concentrations. Bet-hedging behavior in rich medium (FNosZ ≈ 0.03) reduces immediate N₂O sink capacity upon O₂ depletion, but when grown in digestate, all cells switched to anaerobic respiration upon O₂ depletion.
• Oxygen effect: Aerobic N₂O respiration was not observed; initiation required low O₂: <1–2 µM (aerobically raised cells) or 4–6 µM (previously anoxic cells).
• Field bucket (clay loam, pH 6.7): Digestate with live CB-01 nearly eliminated the initial fertilization-induced N₂O peak and substantially reduced subsequent peaks. Reductions in peak responses: rain event (day 12) reduced by 51%; re-fertilization peaks reduced by 31%, 67%, and 46%. Peak 1–2 days post-fertilization was reduced by ~85 mg N₂O–N m⁻² h⁻¹, representing ~8% of the theoretical maximum N₂O consumption if all added cells respired at Vmax; later peak reductions ≤4 mg N₂O–N m⁻² h⁻¹ (≤0.36% of initial potential). Effect sizes were strongest during high-emission periods.
• Multi-soil buckets: Despite overall lower emissions (cooler September weather), CB-01 significantly reduced N₂O emissions in all four soils. Strong effect was observed even in acidic sandy silt (pH 4.15), likely due to local pH elevation by digestate (pH 7.3) in a weakly buffered soil. Among clay loams, effects were stronger at higher soil pH (neutral pH 6.7 > pH 5.26 and pH 4.5).
• Field plots (pH ≈ 6.13): Over 0–10 days, CB-01 reduced cumulative N₂O emissions by 64% (95% CI: 54, 74). Over 10–290 days, the reduction was 12% (95% CI: −8, 32) and not statistically significant; low soil temperatures likely limited effect.
• Survival/persistence: In laboratory incubation (wet soil), CB-01 showed an initial growth phase (≤2–3 days), then rapid decline (T1/2 ≈ 0.8 days during days 3–7), followed by slower decline (T1/2 ≈ 18.7 days). In field buckets, CB-01 abundance remained high over 90 days with gradual decline (fitted T1/2 ≈ 23.9 days). In field plots, long-term data indicated similar average decay (δ ≈ 0.02 d⁻¹; T1/2 ≈ 34 days). Sustained populations in buckets were likely supported by periodic digestate additions and possibly lower protozoal grazing under drier conditions (greater tortuosity and refuge in small pores).
• Microbiome and biosafety: Digestate application transiently impacted indigenous microbiota regardless of CB-01 viability; no consistent differences between live vs dead CB-01 treatments, and communities converged toward pristine soil composition. No antibiotic resistance or pathogenicity genes were identified in CB-01.
• European-scale mitigation: Assuming a conservative 60% reduction of N₂O emissions from liquid manure applications via NNRB, GAINS projections yield a 2.7% reduction in total anthropogenic N₂O emissions across Europe and 4.0% for EU27 by 2030. Extending NNRB to all organic wastes and combining with mineral fertilizers could reduce EU27 agricultural emissions by ~31%. The study’s broader scaling suggests national anthropogenic N₂O emissions reductions of ~5–20% are feasible when including broader organic waste applications.

The study demonstrates that large-scale inoculation of soils with an NNRB, CB-01, using digestate as substrate and vector, can markedly lower N₂O emissions, particularly during high-emission periods following fertilization or rainfall. Despite CB-01’s inferior catalytic efficiency at low N₂O concentrations and evidence of bet-hedging in rich laboratory media, its strong mitigation effect in soil is attributed to high cell densities achieved in digestate, favorable switching to anaerobic respiration when grown in digestate, and notably its persistence in soil. Soil-type and pH strongly modulate efficacy: higher pH enhances NosZ synthesis and function, and weakly buffered acidic soils can experience local pH elevation upon digestate incorporation that enables CB-01 activity. The mitigation effect diminishes under cold conditions and during periods of low emissions, consistent with reduced hypoxic/anoxic microsites and lower N₂O supply from indigenous processes. Survival analyses indicate environmental context matters: moisture and potential protozoal grazing influence persistence; periodic organic inputs can sustain populations. Importantly, the inoculation did not induce persistent disruption of the native microbiome and raises fewer biosafety concerns given the absence of resistance or pathogenicity genes. Scaling analyses suggest that while targeting only liquid manure systems provides modest national reductions, broader deployment across organic wastes can deliver substantial agricultural mitigation, complementing existing measures like nitrification inhibitors and precision fertilization.

This work provides a proof of concept for a cost-effective biotechnology to reduce farmland N₂O emissions by enriching and delivering non-denitrifying N₂O-respiring bacteria via organic wastes. Cloacibacterium sp. CB-01, grown to high densities in digestate and incorporated into soils, reduced emissions by 50–95% in field buckets across soil types and by 64% in the first 10 days in field plots. Its mitigation effect stems from persistence and context-dependent regulation rather than superior kinetic affinity. European-scale modelling indicates meaningful national reductions, with potential for much larger gains if the approach is expanded to diverse organic wastes and combined with mineral fertilizer regimes. Future research should assemble a portfolio of NNRB strains with broader substrate use and tolerance to abiotic and, critically, biotic stresses in soil; refine vector materials beyond digestate; optimize application timing and rates across climates and soil pH; and integrate with agronomic practices to maximize durable emission reductions.

• CB-01’s kinetic profile shows low apparent affinity for N₂O (Kₘ ≈ 12.9 µM) and low catalytic efficiency relative to other NRB, limiting sink strength at low N₂O levels. Bet-hedging in nutrient broth (low FNosZ) may constrain rapid onset of N₂O respiration upon O₂ depletion, although this effect was not observed when grown in digestate.
• Efficacy is contingent on environmental conditions: strongest during high-emission events; diminished during cold periods and in drier, well-aerated soils with fewer anoxic microsites. Soil pH strongly influences NosZ functionality; very acidic soils may require local pH elevation via amendments.
• Persistence varied between lab and field, likely due to moisture and protozoal grazing; mechanisms remain speculative and require targeted study. Long-term sustainability without repeated organic additions is uncertain.
• Field plot long-term effect (10–290 days) was not statistically significant under the tested conditions, indicating limited late-season impact.
• GAINS modelling assumes uniform emission and reduction factors (IPCC 1% EF, 60% reduction for liquid manure), which may not capture spatial/management variability; national reduction estimates thus carry uncertainty.
• Current demonstration focused on digestate as vector; generalizability to other organic wastes and different farming systems requires validation and potentially new strains and processes.