Field Application of Organic Fertilizers Triggers N2O Emissions From the Soil N Pool as Indicated by 15N-Labeled Digestates

Anaerobic digestion (AD) can generate biogas while simultaneously producing digestate which can be used as fertilizer. Feedstocks used for AD influence digestate composition, which in turn may affect carbon (C) and nitrogen (N) turn-over in soils and subsequently influence nitrous oxide (N₂O) emissions after soil application. Assessment of greenhouse gas emissions from digestates can help to evaluate the overall sustainability of an agricultural production system. The objective of this study was therefore to evaluate and understand the effect of differences in digestate composition on in situ N₂O emissions within the 1st weeks after application of seven digestates. The digestates were derived from different feedstocks and ¹⁵N-labeled, either in total N or only in ammonium-N. Therefore, the experimental design enabled us to differentiate between potential N₂O-N sources (i.e., digestate N or soil N). Furthermore, it allowed to distinguish to some extent between organic-N and ammonium-N as potential N sources for denitrification. Digestates were homogeneously incorporated into the upper 5 cm of microplots in an arable Haplic Luvisol in South Germany at a rate of 170 kg N ha⁻¹. After application, N₂O fluxes were measured for ~60 days (May-July) using the closed chamber method in 2 experimental years. Mainly due to higher precipitations in the 1st year, cumulative N₂O emissions were higher (312–1,580 g N₂O-N ha⁻¹) compared to the emissions (133–690 g N₂O-N ha⁻¹) in the 2nd year. Between 16–33% (1st year) and 17–38% (2nd year) of N₂O emissions originated from digestate N, indicating that digestate application triggered N₂O production and release mainly from soil N. This effect was strongest immediately after digestate application. It was concluded that the first (short term) peak in N₂O emissions after digestate application is largely related to denitrification of soil-N. However, the experimental setup does not allow to differentiate between the different denitrification pathways. Weather conditions showed a substantial effect on N₂O emissions, where the correlation between N₂O and CO₂ flux rates hinted on denitrification as main N₂O source. The effect of digestate composition, particularly organic N from the digestate, on soil N₂O emissions seems to be of minor relevance.

Introduction

In the EU, about 180 million tons of anaerobic digestate are estimated to be produced per year, most of which is used as organic fertilizer (Corden et al., 2019). Digestates have been shown to have the potential to substitute mineral fertilizers and contribute to a sustainable soil management (Gutser et al., 2005; Cavalli et al., 2016; Verdi et al., 2019). However, application of organic as well as mineral nitrogen (N) fertilizers is also known to increase greenhouse gas (GHG) emissions from soils. Globally, agriculture contributes up to 20% to carbon dioxide equivalents (CO₂-eq.) from all human activities (2010–2017), with nitrous oxide (N₂O) and methane (CH₄) as main GHGs (FAO, 2020). About 60% of anthropogenic N₂O emissions are emitted by agricultural soils (Ciais et al., 2013), thus it is of high relevance to assess N₂O in relation to fertilizer application.

Studies have shown that digestates might lead to a higher risk of N₂O formation than manures, which is related to the higher share of ammonium (${\text{NH}}_4^{+}$-N) after AD (Möller and Stinner, 2009). Ammonium is quickly nitrified to nitrate (${\text{NO}}_3^{-}$), which can further be denitrified to dinitrogen gas (N₂). Both processes, as well as nitrifier denitrification, bear the risk of producing N₂O and are considered as main N₂O source from soils (Granli and Bøckman, 1994; Bremner, 1997; Koola et al., 2010). Application of liquid manures like slurry or digestates provides available N and carbon (C), which in turn promotes heterotrophic activity (oxidation of C, N, S, etc.), depleting oxygen (O₂) availability in soil, and thus favors creation of anaerobic microsites that ultimately trigger N₂O production and release via denitrification (Chadwick et al., 2000; Petersen et al., 2003). Hence, N₂O emissions largely depend on the availability of labile organic C (C_org), mineral N, O₂ and water in the soil and their subsequent effect on soil microbial processes (Flessa and Beese, 2000; Ruser et al., 2001). However, AD has also been reported to reduce the N₂O potential compared to the initial feedstock e.g., by decreasing slurry viscosity or increasing the recalcitrance of organic matter (OM) (Petersen, 1999; Möller, 2015).

The different organic substrates that are used as feedstock for anaerobic digestion (AD) affect the physico-chemical characteristics of the digestate (Fouda et al., 2013; Zirkler et al., 2014). For example, comparing food wastes and maize silage, food wastes are already processed goods with a high degradability and high protein content. Thereby, food waste-based digestates tend to have a higher OM degradability and a higher share of ${\text{NH}}_4^{+}$-N than maize silage, that could enhance soil microbial activity (Möller and Müller, 2012; Guilayn et al., 2020).

Based on compositional differences, such as N content, C/N ratio and OM degradability [C_org/organic N (N_org)], it can be assumed that digestates from different feedstocks will show differences in N₂O emissions after field application. However, a differentiated consideration of the GHG emission potential for digestates from different feedstocks is currently scarce, and therefore will be the main research focus of this study.

The largest share of N₂O release during the growing season usually occurs shortly after field application, with further peaks correlated to rainfall-events (Guzman-Bustamante et al., 2019; Herr et al., 2019) or freeze-thaw periods (Flessa et al., 1995; Rochette et al., 2008). For this reason, the following experiment was conducted to evaluate digestates regarding short-term N₂O emissions on fallow land. To calculate the amount of N₂O derived from the digestate, ¹⁵N-stable isotope labeling was used. The following hypotheses were tested:

(1) Digestates with varying physical and chemical properties will show different temporal N₂O and ¹⁵N-N₂O flux patterns.

(2) Application of these digestates will also result in different cumulative N₂O emissions and N₂O emission factors.

(3) The amount of N₂O-N directly derived from the digestate will differ among the digestate types.

Materials and Methods

¹⁵N Labeling and Digestate Production

Labeled anaerobic digestates were prepared by cultivation of ¹⁵N-enriched plants in a comparable approach as applied by Schouten et al. (2012). Maize (Zea mays L. cv. Ronaldinio), ryegrass (Lolium perenne L. cv. Kentaur), and sugar beet (Beta vulgaris subsp. vulgaris, Altissima Group) were ¹⁵N labeled, by addition of ¹⁵N ammonium sulfate ((NH₄)₂SO₄) as fertilizing solution. Ryegrass was cultivated in sand culture in 12 kg boxes of 10 cm height. For fertilization, 96.2 mg N kg ⁻¹ as (NH₄)₂SO₄ (30 atom% ¹⁵N) solution was applied before sowing ryegrass. We cut the ryegrass three times in 30-days intervals. Sugar beet was grown in Mitscherlich pots with 12 kg sand. After pre-growing sugar beet seedlings in peat, two plants were set for each pot. Four rates of (NH₄)₂SO₄ (50 atom% ¹⁵N) solution were applied during growth (in total 1.5 g N per pot). Maize was grown in a hydroponic system with two plants per 10-liter pot. Nutrient solution adapted after Engels (1999) with modified N concentration was exchanged twice a week. Within the first 5 weeks of growth, ${\text{NH}}_4^{+}$-N concentration was gradually increased, while ${\text{NO}}_3^{-}$-N supply was decreased to acclimate maize plants to primary ${\text{NH}}_4^{+}$-N nutrition. After this adaption phase for the plants, the N concentration was kept stable at 0.5 mM ${\text{NO}}_3^{-}$-N and 3 mM ${\text{NH}}_4^{+}$-N, in the form of calcium nitrate and ammonium sulfate. For ¹⁵N labeling, four additions of ${\text{NH}}_4^{+}$-N were substituted by 50 atom% ¹⁵N- ${\text{NH}}_4^{+}$ and applied at BBCH stages: 16–19, 30–33, 51–55, and 71. As commonly done for maize, as energy crop, it was harvested at the dough-ripe stage. After harvest, ryegrass, maize sugar beet, as well as sugar beet leaves were immediately cut and homogenized by short blending (Thermomix TM31, Wuppertal, Gemany). The ¹⁵N enrichment of crops and harvest residues was determined by IRMS with previous freeze-drying, leading to 19.3 atom% ¹⁵N in maize, 26.1 atom% in ryegrass, 43.8 atom% in sugar beet, and 45.3 atom% in sugar beet leaves. After weighing the ¹⁵N-plant biomass into small portions, they were frozen at −20°C until anaerobic digestion in a batch reactor as previously described by Brulé (2014), Mönch-Tegeder et al. (2014).

Anaerobic digestion of the ¹⁵N-labeled plants and plant residues was carried out at the State Institute of Agricultural Engineering and Bioenergy, at the University of Hohenheim. Before, digestates from maize, grass silage and sugar beet were collected from biogas plants in southern Germany to be used as inoculum for AD of the ¹⁵N-feedstocks (maize, ryegrass, sugar beet, and leaves). Digestates were “starved” for 10 days according to the German standard VDI 4630 guideline (2016) to minimize residual gas production. During this starvation phase, the vessels were kept open and stirred to volatilize ammonia (NH₃) from the inoculum. By decreasing ${\text{NH}}_4^{+}$-N in the digestate, hence total N concentration, a high ¹⁵N-signature could be assured, with only marginal N dilution of the added ¹⁵N-feedstock. Prior to AD, the inoculum was sieved to produce a homogeneous slurry. ¹⁵N-labeled ryegrass, maize, sugar beet, and sugar beet leaves were separately added to the substrate-specific inoculum in a ratio of 1:2.5 organic total solids (oTS) (VDI 4630, 2016). Anaerobic digestion was carried out in 2 liter fed-batch systems under mesophilic temperature at 37.5 ± 1°C for 60 days. During digestion, three feeding portions of ¹⁵N enriched plant substrates were added: at the start of the experiment, after 20, and after 40 days, respectively. Due to the amount of oTS added by the digestates, the ¹⁵N amount of the feedstocks was diluted by N contained in the inoculum, leading to a lower labeling of the ¹⁵N-digestates compared to the initial plant feedstock (Table 1).

TABLE 1

Table 1. Physico-chemical digestate properties, ¹⁵N-labeling and ¹⁵N abundance.

Additionally, digestates from existing biogas plants were included and the mineral ${\text{NH}}_4^{+}$-N fraction was labeled: organic waste digestate, food waste digestate, and cattle slurry digestate. The digestates were analyzed for total N and ${\text{NH}}_4^{+}$-N. Each digestate was filled into a glass beaker and put into a rotating water bath for 12 h at 37°C, to volatilize a small amount of NH₃. Afterwards the digestates were analyzed again for total N and ${\text{NH}}_4^{+}$-N to assess the amount of N that was emitted. The lost N was substituted by addition of ¹⁵N-enriched (NH₄)₂SO₄ solution to 5 atom% ¹⁵N excess. If more N was lost than resupplied by ¹⁵N-${\text{NH}}_4^{+}$, ammonium chloride solution was added.

Experimental Design

The experiment was performed at the research station “Heidfeldhof” at the University of Hohenheim, 13 km south of Stuttgart, in South-Germany. The research station has a mean annual precipitation of 686 mm and a mean annual temperature of 8.8°C, monitored by a local meteorological station. The soil type of the arable field was a Haplic Luvisol (IUSS Working Group, 2015) with a silty loam soil texture (2% sand, 68% silt, and 30% clay), a bulk density of 1.24 g cm⁻³ in the upper 30 cm. Soil analytical results are presented in Table 2. The micro-plot field experiment (1 × 1 m plot size) was conducted as randomized block design with four replicates per treatment in 2 years 2016 (1st) and 2017 (2nd year) from May to July. The treatments consisted of one unfertilized control and seven ¹⁵N-labeled digestates based on maize (M), grass (G), sugar beet (SB), sugar beet leaves (SBL), organic waste (OW), food waste (FW), and cattle slurry (CS) (Table 1).

TABLE 2

Table 2. Soil characteristic at the beginning of the experiment spring 2016 (1st year) and 2017 (2nd year), mean mineral N (N_min) (± standard deviation (n = 2).

Gas Measurements and Analysis

The closed chamber system was used to monitor N₂O, CO₂, and methane (CH₄) soil fluxes (Hutchinson and Mosier, 1981). The system consisted of a polyvinyl chloride (PVC) base ring (30 cm inner diameter) and a corresponding chamber (Pfab et al., 2011). Within the center of the 1 m² micro plot, the PVC base ring was embedded 10 cm deep in the soil. The ¹⁵N-labeled digestates (Table 1) were applied at a rate of 170 kg N ha⁻¹, meaning 1.2 g N per base ring and quickly incorporated into the upper 5–10 cm of the soil. In order to do so, a 10 cm deep furrow was dug across the ring, using a spade. Digestate was filled into the furrow, covered by soil and mixed. The same procedure was done for the unfertilized control using water. The amount of water (290 ml) corresponded to the average volume of digestate application. Directly after application, the first gas measurement was performed. For gas sampling between 8.00 and 12.00 am, the base ring was covered with the dark, vented PVC chamber, sealed by a rubber ring to collect the trace gas. The chambers were closed for 45–60 min. The first gas sample was directly taken after closure, followed by additional sampling every 15–20 min using a syringe and transferred into evacuated 20 ml gas vials. At the same time, two additional gas samples were collected into 100 ml vials at the start and end of each measurement for ¹⁵N-N₂O determination. Soil and chamber temperature were recorded within each block from two random plots at beginning and end of sampling. Within the 1st month, gas samples were taken 3–4 times a week. Afterwards the sampling frequency was reduced to once or twice per week, with additional samplings after strong rainfall events. In both years 20 gas samplings were performed and measured for N₂O, CO₂, and CH₄, whereas 13–15 out of 20 ¹⁵N-N₂O gas samples could be measured due to cost and time reasons in the 1st and 2nd year, respectively.

Gas samples were measured with a gas chromatograph (GC 450 Greenhouse Gas Analyzer, Bruker Daltonic, Bremen, Germany) equipped with electron capture detector (ECD) and flame ionization detector (FID) and an automatic sampler (GX-281, Gilson, Limburg, Germany). During GC measurements, concentrations of N₂O and CO₂ were analyzed with a ⁶³Ni ECD and CH₄ concentrations were determined with the FID. Fluxes of N₂O, CO₂, and CH₄ were calculated by an extended version of the R (R Core Team, 2016) package “gasfluxes” (Fuß and Asger, 2014).

Analysis of Digestate and Soil

Digestates were dried at 105°C for dry matter (DM) analysis. Total C (C_t) was measured by Dumas combustion via elemental analysis (Elementar vario MAX CN, Analysensysteme GmbH, Hanau, Germany). Carbonate content was determined volumetrically using the Scheibler method according to DIN 10693 (2014). Thus, organic C can be calculated as the difference between C_t and Carbonate-C. Total N (N_t) and ${\text{NH}}_4^{+}$ of fresh matter (FM) digestate sample was determined by Kjeldahl method. Organic N was derived by the difference of ${\text{NH}}_4^{+}$ from N_t. The pH value was measured in FM digestate using 0.01 mol L⁻¹ calcium chloride solution (1:10 w/w).

Soil mineral N (N_min)was determined by extraction with 0.5 M potassium sulfate solution (1:4) and measured colorimetrically with a photometer (Flow-injection-analyzer 3 QUAAtro, SEAL Analytical, UK). Bulk density of the top soil was determined using 100 ml stainless steel cylinders in the field.

Total N and C, and ¹⁵N-signature of ¹⁵N-labeled plant substrates, soil and digestate was measured with a CN-elemental analyzer (EuroVector, HEKAtech, Wegberg, Germany) with Isotope Ratio Mass Spectrometer (IRMS) (Delta plus Advantage, Thermo Finnigan, Bremen, Germany). For the determination of the ¹⁵N abundance in N₂O we used an IRMS delta plus (Finnigan MAT, Bremen, Germany) coupled with an automated PreCon-Interface (Brand, 1995).

Statistics and Calculations

Trapezoidal linear interpolation of daily gas fluxes (N₂O and CH₄) was used to calculate total cumulative emissions for the 55–58 days of the experiment in the 1st and 2nd year, respectively. The percentage of N₂O-N originating from digestate N (Nd) was calculated by equation (1), with digestate i at sampling time t. Atom%¹⁵N excess was calculated by subtraction of the natural abundance of N₂O in the atmosphere (0.369 atom%) from the measured ¹⁵N. The daily N₂O flux rate (μg N₂O–N m⁻² h⁻¹) was multiplied with Nd in equation (2) to determine the amount of N₂O derived from digestate (¹⁵N-N₂O) as reported by Senbayram et al. (2009). We calculated the recovery (%) of ¹⁵N applied by summing up ¹⁵N content of the soil at the end of the experiment and cumulative ¹⁵N-N₂O loss. This sum was then divided by the amount of ¹⁵N applied by the digestate as described by Pfab (2011).

$\begin{array}{ll}N{d}_{i,t}(\%)=\frac{atom{\%}^{15}Nexcessdigestat{e}_i}{atom{\%}^{15}Nexcess{N}_2{O}_{i,t}}& (1)\end{array}$

$\begin{array}{l}{}^{15}N-N_2{O}_{i,t}(\mu g{N}_{2}O-N{m}^{-2}{h}^{-1})\\ =N{d}_{i,t}(\%)*N_{2}Oflu{x}_{i,t}& (2)\end{array}$

Digestate derived fluxes (¹⁵N-N₂O) were also linearly interpolated to calculate cumulative (cum) ¹⁵N-N₂O emissions. The total share of N derived from digestate (total Nd) in cumulative N₂O was calculated by Equation (3). As suggested by Schleusner et al. (2018), the amount of primed N₂O-N lost by fertilizer application was calculated by a simplified approach accounting for cumulative N₂O-N emissions of the unfertilized control treatment (Equation 4), without considering other gaseous losses via NH₃ or N₂.

$\begin{array}{ll}totalN{d}_i(\%)=\frac{{{cum}^{15}{N-N}_{2}O}_i(g{N}_{2}Oh{a}^{-1})}{{cum{N}_{2}O}_i(g{N}_{2}Oh{a}^{-1})}*100& (3)\end{array}$

$\begin{array}{l}Primed{N}_2{O}_i(g{N}_{2}Oh{a}^{-1})=\left({cum{N}_{2}O}_i-{{cum}^{15}{N}_{2}O}_i\right)\\ -cum{N}_{2}Ocontrol& (4)\end{array}$

$\begin{array}{ll}{EF}_i(\%)=\frac{\left({cum{N}_{2}O}_i-cum{N}_{2}Ocontrol\right)g{N}_{2}Oh{a}^{-1}}{170{kgha}^{-1}*1000}*100& (5)\end{array}$

N₂O emission factors (EFs) were calculated according to the IPCC guidelines for direct emissions (Equation 5), meaning total cumulative N₂O-N emissions accounted for the control, per applied N (IPCC, 2019). The disaggregated IPCC N₂O EF for “other N inputs in wet climates” with the default value of 0.6% was applied for comparison, where other N inputs refer to organic amendments such as digestates. Field conditions of the experimental site showed a positive water balance and fit with IPCC conditions for wet climates (IPCC, 2019). Greenhouse gas emissions of N₂O and CH₄ were transformed to CO₂-eq. to assess the total global warming potential (GWP) of each digestate. The default values of 296 g g⁻¹ CO₂ for N₂O and 24 g g⁻¹ CO₂ for CH₄ were applied to the measured emissions. Ammonia (NH₃) volatilization within the first 72 h was derived from the ALFAM2 model to calculate potential indirect N₂O emissions (Hafner et al., 2019). The model is used to predict NH₃-N losses within the first 72 h from animal slurry and therefore holds a higher uncertainty for digestates. Digestate NH₃-N losses mainly served as an indicator for the amount of indirect N₂O emissions. Indirect emissions from ${\text{NO}}_3^{-}$ leaching were not accounted for. According to IPCC (2019) 1% of NH₃-N losses was assumed to be re-deposited as N₂O-N. Soil organic C stocks were presumed to be stable over the experimental period, thus, CO₂ fluxes were not considered for the calculation of total GWP (Herr et al., 2020).

Water filled pore space (WFPS) was calculated by Equation (6) using the measured volumetric water content (WC_vol) and porosity (P),

$\begin{array}{ll}WFPS(\%)=\frac{W{C}_{vol}}{P}*100& (6)\end{array}$

where P is depicted as soil bulk density (ρd) and solid particle density (ρs) (Equation 7). For ρs the density of quartz (2.65 g cm⁻³) was assumed.

$\begin{array}{ll}P(\%)=\left(1-\frac{\rho d}{\rho s}\right)*100& (7)\end{array}$

For each year, a regression analysis of N₂O fluxes was calculated, using a stepwise forward selection in a multiple linear regression approach. Air temperature (2 m height), WFPS and CO₂ fluxes were included as independent variables within the model (8). Only significant variables remained in the model (α = 0.05) and the square root of the partial R² was determined. Same regression procedure was applied for cumulative N₂O and ¹⁵N-N₂O emissions within each year separately. For this approach the effect of digestate composition was determined, using ${\text{NH}}_4^{+}$-N share, and the ratios C/N and C_org/N_org in model (9).

$\begin{array}{ll}{y}_{it}=\mu +{\beta }_{1}tem{p}_t+{\beta }_{2}WFP{S}_{it}+{\beta }_3{C{O}_2}_{it}+b_{it}+e_{it}& (8)\end{array}$

$\begin{array}{ll}{y}_i=\mu +{\beta }_{1}C{N}_i+{\beta }_2{C_{org}{N}_{org}}_i+{\beta }_3{N{H}_4}_i+b_i+e_i& (9)\end{array}$

where y_i is the observation of the i^th digestate treatment, μ represents the average response, β_n are the parameters of fixed effects, b_i is the complete block effect and e_i is the error of y_i.

Significant differences among treatments for cumulative N₂O, CH₄, CO₂, as well as ¹⁵N-N₂O and total Nd (%) were determined by the Proc MIXED procedure and the Tukey test (α = 0.05). The MIXED procedure can fit various mixed linear models to data and produces the appropriate statistics (SAS Institute Inc, 2015). All statistical analyses were performed with SAS 9.4 (SAS Institute, Cary NC, USA). Graphics were produced with SigmaPlot 11.0 (Systat Software GmbH, Erkrath, Germany).

Results

Meteorological Conditions

Weather conditions showed distinct differences in precipitation between the 2 experimental years. Over the 1st and 2nd year, 183 and 178 mm of precipitation were measured during the 55 and 58 days when the experiment lasted, respectively (Figure 1). Within the first 30 days of measurements the rainfall pattern differed, showing 155 mm in the 1st compared to 86.2 mm precipitation in the 2nd year. In the 1st year, two strong rainfall events occurred on day 12 (33.2 mm) and day 22 (39.5 mm). In contrast, the 2nd year showed lower rainfall events on day 6 and 27 with 18.8–22.2 mm, and two stronger events at the end of the experiment on day 51 and 52 (32.2–27.4 mm). The mean air temperature over the experimental period was 17.1°C in the 1st and 18.0°C in the 2nd year.

FIGURE 1

Figure 1. Mean daily N₂O fluxes after application of different digestates based on organic waste (OW), food waste (FW), cattle slurry (CS), sugar beet (SB), sugar beet leaves (SBL), maize (M), grass (G), and unfertilized soil (control) measured from May to July 2016 and 2017 (1st and 2nd year) at Heidfeldhof. Digestate application of 170 kg N ha⁻¹ on day 1. Standard error of measurements was excluded due to improved clarity, but shown in ¹⁵N fluxes (Figures 2, 3). Climate data with mean daily precipitation on the left y-axis and mean daily temperature on the right y-axis.

Temporal N₂O Fluxes

Nitrous oxide fluxes measured in the 2 experimental years showed distinct differences in peak number and flux magnitude. In both years N₂O pulses occurred directly after digestate fertilization and after strong rainfall events (Figure 1). Three major peaks were detected in the 1st year: one directly after digestate application, the second and third peak after 13 and 24 days, and a minor peak after 1 week, following strong rainfall events on day 12 and 23. Highest N₂O flux rate in the 1st year was measured with the SBL treatment on day 13 (1,260 μg N₂O-N m⁻² h⁻¹). In the 2nd year, the N₂O pulse developing directly after N fertilization did not reach the same magnitude as in the 1st year and appeared 1 day later. Highest N₂O flux in the 2nd year followed a rainfall event 1 week after digestate application reaching up to 424 μg N₂O-N m⁻² h⁻¹ with SB (Figure 1). The peak decreased sharply in case of SB, and gradually until day 14 for the other digestates. After another strong rainfall event on day 22 (2nd year), only FW showed a slight N₂O rise (38 μg N₂O-N m⁻² h⁻¹). Approaching the end of the experiment, 50 days after digestate application, a small peak (5.22–21.9 μg N₂O N m⁻² h⁻¹) was noted within 4 days of continuous rainfall (Figure 1).

In both years, WFPS showed a significant positive linear correlation with N₂O flux rates, r = 0.400 (p < 0.001) and r = 0.454 (p < 0.001) in the 1st and 2nd year, respectively. Similarly, CO₂ fluxes correlated with N₂O fluxes, exhibiting r = 0.233 (p < 0.001) in the 1st year, and a weaker coefficient of correlation in the 2nd year (r = 0.144, p < 0.001). Soil temperature showed a negative correlation with N₂O fluxes (r = −0.340; p < 0.001) in the 2nd year, but no significant correlation in the 1st year. All parameters (WFPS, soil temperature and CO₂) combined in a linear regression model (model 8) could account for 28.1 to 20.9 % of the prediction of N₂O fluxes in the 1st and 2nd year, respectively (Table 3).

TABLE 3

Table 3. Output of (stepwise forward) regression analysis for N₂O fluxes in the 1st and 2nd year, testing for the inclusion of parameters CO₂ flux, water filled pore space (WFPS), and soil temperature into the model (model 8).

Temporal ¹⁵N-N₂O Fluxes

Total N₂O and digestate derived ¹⁵N-N₂O fluxes in the 1st year are shown in Figure 2 and the 2nd year data are shown in Figure 3. A comparable trend was observed for both ¹⁵N_t- and ¹⁵${\text{NH}}_4^{+}$-N-labeled digestates in each year, with variations in flux magnitude of ¹⁵N-N₂O among digestates (Supplementary Tables 2, 3).

FIGURE 2

Figure 2. Mean daily N₂O fluxes (total) and digestate-derived ¹⁵N-N₂O fluxes (¹⁵N) within the 1st year Digestates from maize (M), grass (G) sugar beet (SB), and sugar beet leaves (SBL) were ¹⁵N_t-labeled (mineral and organic N) and digestates based on cattle slurry (CS), organic waste (OW), and food waste (FW), were ¹⁵N-labeled only in the ${\text{NH}}_4^{+}$-N pool. Error bars indicate the standard error (n = 4).

FIGURE 3

Figure 3. Mean daily N₂O fluxes (total) and digestate-derived ¹⁵N-N₂O fluxes (¹⁵N) within the 2nd year Digestates from maize (M), grass (G) sugar beet (SB), and sugar beet leaves (SBL) were ¹⁵N_t-labeled (mineral and organic N) and digestates based on cattle slurry (CS), organic waste (OW), and food waste (FW), were ¹⁵N-labeled only in the ${\text{NH}}_4^{+}$-N pool. Error bars indicate the standard error (n = 4).

In the 1st year, the emerging peak directly after digestate application showed a low ¹⁵N signature, indicating that 92.4–96.5% of N₂O was derived from soil internal N sources (Supplementary Table 2). Within the first 10 days after digestate application, ¹⁵N-N₂O fluxes showed no significant differences among treatments (Supplementary Table 2). Only on day 7, a small peak in ¹⁵N-N₂O (13–87 μg ¹⁵N-N₂O-N m⁻² h⁻¹) appeared, and SB showed significantly higher emissions than M, OW, FW, and CS. On that day, ~18–30% of N₂O-N was derived from digestates (Figure 2 and Supplementary Table 2). During the highest peak on day 13, a significant proportion of digestate-based N₂O-N (31–59%) was emitted. Highest ¹⁵N-N₂O among digestates was measured with SBL (539 μg ¹⁵N-N₂O-N m⁻² h⁻¹), not significantly different from G (338 μg ¹⁵N-N₂O-N m⁻² h⁻¹). Following the two peaks on day 7 as well as on day 13, lower total N₂O and ¹⁵N-N₂O fluxes were measured, but the share of digestate-derived N was still relatively high (Supplementary Table 2). The last major peak appeared on day 24 (10–114 μg ¹⁵N-N₂O-N m⁻² h⁻¹), with 10–27% N₂O-N originating from digestates. Flux rates peaking on day 24 were comparable among most treatments, and only OW exhibited significantly higher flux rates than SB. Prior to peaks of day 24, OW already indicated a rising flux rate on day 20, being significantly higher than all digestates, except G and FW. The flux rate further increased on day 22, where OW significantly exceeded all other treatments (Figure 2 and Supplementary Table 2). Within the first 3 weeks of measurements, flux pattern of CS digestate significantly differed from the other treatments, where ¹⁵N-N₂O gradually increased after 7 days and reached its maximum on 13 (Figure 2). At both peaks, on day 13 and 24, ¹⁵N-N₂O flux rates of CS were in a comparable range (48–40 μg ¹⁵N-N₂O-N m⁻² h⁻¹). The 2nd year showed a similar temporal pattern in ¹⁵N-N₂O abundance over the duration of the experiment. The first pulse after digestate application was observed 2 days after application with more than 80% soil-borne N₂O-N. Only CS showed lower soil-borne N₂O-N, thus highest digestate derived N₂O-N (45%) among digestates (Supplementary Table 3). The major peak in total N₂O and ¹⁵N-N₂O appeared after 1 week (Figure 3), with highest ¹⁵N-N₂O flux measured in G (189 μg ¹⁵N-N₂O-N m⁻² h⁻¹) showing 67% digestate-derived N. In contrast, 28–36% of N₂O-N was emitted from the other digestate treatments on that day (Supplementary Table 3). From day 7 to day 18, ¹⁵N-N₂O gradually decreased for all digestates to 0.3–20 μg ¹⁵N-N₂O-N m⁻² h⁻¹, except for M being significantly higher (66.1 μg ¹⁵N-N₂O-N m⁻² h⁻¹ on day 18). From day 18 through 29, the M treatment continued to show higher ¹⁵N fluxes compared with the other digestates, even though these emission rates were quite low (from 1.1 to 2.7 μg ¹⁵N-N₂O-N m⁻² h⁻¹).

Cumulative N₂O and ¹⁵N-N₂O Evolution and Emission Factors

Total cumulative N₂O emissions in the 1st year (302–1,345 g N₂O-N ha⁻¹) were more than twice as high as in the 2nd year (124–613 g N₂O ha⁻¹) (Figure 4 and Table 4). In both years, digestates lead to significantly higher N₂O emissions than the unfertilized control. Differences among digestates were observed only in the 1st year, with significantly higher N₂O emissions for SBL compared to CS and FW (Figure 4). Compared to total N₂O emissions, digestate-based ¹⁵N-N₂O emissions indicated larger differences between the different treatments in both years (Figure 4). In the 1st year, G and SBL emitted significantly more ¹⁵N-N₂O than CS, while all other treatments did not differ significantly. In the 2nd year, highest ¹⁵N-N₂O emission was measured with G and M, while all ${\text{NH}}_4^{+}$-N-labeled digestates OW, FW and CS, as well as SB and SBL were comparable. Calculated amounts of primed N₂O-N showed that significant higher N₂O-N losses were induced by SBL compared with CS in the 1st year (Table 4). In the 2nd year, there were no significant differences among digestates. Total Nd emitted by the digestates was 16–33% in the 1st, and 17–38% in the 2nd year (Table 5). Grass digestate tended to show the highest share of digestate derived Nd in both years.

FIGURE 4

Figure 4. Total cumulative N₂O-N emissions derived from digestate (¹⁵N-N₂O) 55 (1st year) and 58 days (2nd year) after digestate application. In both years, measurements were conducted from May to July. Unfertilized soil served as control. Digestates from cattle slurry (CS), organic waste (OW), food waste (FW) were ¹⁵N-labeled only in ${\text{NH}}_4^{+}$-N; digestates based on grass (G), maize (M), grass (G), sugar beet (SB), and sugar beet leaves (SBL) were ¹⁵N_t-labeled (mineral and organic N). Error bars show standard error (n = 4). Different letters indicate significant differences at p < 0.05 (Tukey Test); Capital letters refer to total N₂O and small letters to ¹⁵N-N₂O.

TABLE 4

Table 4. Total cumulative N₂O-N, ¹⁵N-N₂O-N emissions, and primed N₂O-N emissions after 55 (1st) and 58 days (2nd year).

TABLE 5

Table 5. Share of digestate derived N₂O-N on total N₂O emissions (Total Nd) and N₂O-N emission factors after 55 (1st) and 58 days (2nd year).

There was no correlation of digestate properties (C/N, C_org/N_org, ${\text{NH}}_4^{+}$-N/N_t) with N₂O emissions. The respective digestate characteristics did not help to predict cumulative N₂O or ¹⁵N-N₂O emissions in a multiple (stepwise forward) linear regression model (model 9).

According to IPCC guidelines, ~0.6% of the annual amount of total N of organic amendments applied as fertilizer is lost as N₂O-N in wet climates (IPCC, 2019). In the 1st year, most digestates approached this IPCC EF within only 55 days and SBL even exceeded it with 0.75% (Table 5). Only FW and CS indicated lower EFs than the IPCC default value in the 1st year, with 0.44 and 0.30%, respectively. Related to the overall lower cumulative N₂O emissions of the 2nd year, mean N₂O EFs were below 0.33% and in a comparable range for all digestates.

Total ¹⁵N recovery within cumulative N₂O and soil N at the end of the experiment was 10–57% and 27–64% in the 1st and 2nd year, respectively (Supplementary Table 4). The largest share of digestate ¹⁵N remained in the soil.

Total Global Warming Potential

Cumulative CH₄-C emissions were significantly higher in the 1st year compared to the 2nd. In both years, unfertilized soil served as CH₄ sink (−147 to −184 g CH₄-C ha⁻¹) (Table 5). Within the 1st year, emissions among digestates ranged between 0.26 and 1.82 kg CH₄-C ha⁻¹ and decreased in the following order SBL ≥ CS ≥ FW, OW, SB ≥ M, G ≥ control. In the 2nd year, digestates as well as unfertilized soil were comparable and digestates indicated CH₄-C emissions close to zero (Table 6).

TABLE 6

Table 6. Modeled NH₃-N losses over the first 72 h after application (ALFAM2 model) and total cumulative CH₄ fluxes of digestates and unfertilized soil (control) after 55 (1st) and 58 days (2nd year).

In both years, the release of CO₂-eq after digestate application was significantly higher than in the control (Table 7). Significant differences among digestates were only noted in the 1st year, where SBL caused significantly higher total CO₂-eq. than M, FW, and CS. In both years, N₂O emissions made up the largest share in total GHG emissions, based on CO₂-eq, above 85.6% in the 1st and almost 100% in the 2nd year.

TABLE 7

Table 7. Carbon dioxide equivalents (CO₂-eq.) of unfertilized soil (control) and soil after application of different digestates, originating from maize, grass, sugar beet, sugar beet leaves, organic waste, food waste, and cattle slurry, based on cumulative N₂O and CH₄ emissions (kg ha⁻¹) after 55 (1st) and 58 days (2nd year); and indirect N₂O emission as NH₃-N volatilization over 72 h after application.

Discussion

Temporal N₂O and ¹⁵N-N₂O Fluxes

The high temporal variability of N₂O fluxes in this study, with increased flux rates after application of crop residues or organic fertilizers and after rainfall events, was similarly documented in other experiments (Pfab et al., 2012; Herr et al., 2019). Ultimately, N₂O fluxes leveled off 30 days after digestate application in both years with drying of the soil during warm periods with low rainfall. Dry conditions with low WFPS have often been reported to result in low N₂O emissions from arable soils even if these soils were well-provided with microbial easily degradable C and available N (Möller and Stinner, 2009; Pezzolla et al., 2012). However, in the 2nd year strong rainfall events were recorded 50 days after digestate application and only caused a minor increase in N₂O fluxes (Figure 1). Hence, digestate-related effects were short-term and had the highest impact on N₂O release within the first 30 days after application.

First Peak After Digestate Application

Peaks evolving shortly after organic N fertilizer application, such as digestates or manures, have been reported by several studies (Wulf et al., 2002; Johansen et al., 2013; Holly et al., 2017). As indicated by ¹⁵N measurements in both years, the first N₂O peak after digestate application showed a low ¹⁵N abundance, demonstrating that more than 90% of N₂O-N was derived from soil N (Figures 2, 3). However, the experimental setup does not allow for a differentiation between nitrification and denitrification. Therefore, we can only conclude that the first N₂O peak was mainly derived from soil N. The high share of soil-borne N suggests that the addition of OM positively affected microbial activity which further enhanced the turnover of native soil-N, as also stated by Schleusner et al. (2018). Furthermore, digestate or slurry application moistened the soil close to the applied fertilizer, another factor that has been shown to promote denitrification of ${\text{NO}}_3^{-}$-N (Comfort et al., 1990). Moreover, CO₂ flux rates were elevated directly after digestate fertilization (Supplementary Figure 1), supporting the assumption of increased microbial activity which further stimulated denitrification of ${\text{NO}}_3^{-}$ by O₂ depletion (Buchen-Tschiskale et al., 2020). However, these are only speculations, as soil N_min and its ¹⁵N-signature was not measured during the experiment. It should also be considered that digestates contain carbonate-C (${\text{HCO}}_3^{-}$ and ${\text{CO}}_3^{2-}$): the higher the total N content, the higher the carbonate-C content (Georgacakis et al., 1982). Carbonate-C in the digestates can also contribute to soil CO₂ release within the 1st days after application (Chen et al., 2011). For example, carbonate-C release from digestates can occur after application to acidic soils (Chen et al., 2011), which is not the case in present study, or due to microbial turnover processes (Tamir et al., 2013). Therefore, the immediate effects of digestate application on soil microbial activity and the related CO₂ release might be masked by decomposition of carbonate-C to CO₂. In order to elucidate the driving processes related to the N turnover processes in the soil shortly after digestate application, a more detailed measurement of the pathways of the different fractions of soil and digestate N (${\text{NH}}_4^{+}$, N_org), as well as digestate C (C_org, carbonate-C), is necessary.

Rainfall-Induced Peaks

The emission pattern found in present study strongly coincided with the precipitation pattern, providing a major indication that the environmental conditions are the main driving factor for soil N₂O fluxes. Also the unfertilized control showed a significant increase in N₂O flux rates after rainfall, whereas almost no fluxes were observed in dry periods. The occurrence of increased N₂O fluxes in conjunction with heavy rainfall events, hence a high soil WFPS, is typical for arable fields and has extensively been described in the literature (Pfab et al., 2011; Senbayram et al., 2014; Ruser et al., 2017).

Contribution of Digestate N and Soil N Pool to N₂O Emissions

The largest rainfall-induced N₂O peaks in both years, had also the highest ¹⁵N abundance, with up to 56–66% of N₂O-N derived from the digestate (Figures 3, 4 and Supplementary Tables 2, 3). Although it was shown that even at a high soil moisture of 70% WFPS nitrification may also contribute to the N₂O-release from soils (Ruser et al., 2006), the positive correlations between N₂O flux rates and CO₂ flux rates as well as between N₂O fluxes and soil moisture (Table 3) indicate that denitrification is the driving process releasing N₂O after rainfall. The contribution of denitrification to the N₂O release generally increases with increasing soil moisture (Davidson, 1991). When compared to soil air, the ~10⁻⁴ lower diffusion coefficient for atmospheric O₂ in soil water (Heincke and Kaupenjohann, 1999) restricts O₂ delivery, the creation of anaerobic conditions is favored. Similarly, the turn-over of fresh OM, as indicated by the increased CO₂ fluxes, further depletes O₂ availability and thus fuels anaerobiosis (Flessa and Beese, 1995).

The largest peaks evolved 13 (1st year) and 7 days (2nd year) after digestate application, where presumably digestate ${\text{NH}}_4^{+}$-N was already nitrified (Johansen et al., 2013) and available for denitrification, thus, explaining the high share of digestate-based N₂O-N. Senbayram et al. (2009) observed that nitrification of ¹⁵N-labeled digestate rapidly increased 1 week after application in a pot experiment initiating a rise in N₂O flux rates, as also noted in our study. It cannot be excluded that beside denitrification also nitrification contributed a share of the measured N₂O. Yet, Köster et al. (2011) measured the intramolecular ¹⁵N distribution in N₂O within a 43-days incubation experiment, showing that bacterial denitrification was the main process emitting N₂O after application of food waste digestate, driven by C availability. This is in line with other studies, reporting that the largest N₂O-N contribution of digestates was caused by denitrification, even at 65% WHC (Senbayram et al., 2009). Later N₂O peaks (1st year, day 24) showed lower ¹⁵N-N₂O fluxes, hence indicating an increasing share of N₂O from soil-internal N. This shift in ¹⁵N-N₂O abundance over the measuring period indicates increased effects of the soil microbial processes, affecting N availability and N₂O emissions, and might result from mineralization of digestate ¹⁵N_org-N and subsequent processes. A comparable shift was observed by Senbayram et al. (2014).

For both, ${\text{NH}}_4^{+}$-N or N_t-labeled digestates, the low shares of fertilizer-derived N₂O-N supported the notion that the largest source of N₂O was native soil N (> 62%, Table 5). The open hypothesis of an “enhanced soil-derived N₂O” stated by Senbayram et al. (2014), regarding the low share of emitted digestate-N, can therefore be confirmed. This triggering effect on N₂O emissions due to digestate application was accounted for by a simplified calculation via equation 4. The amount of triggered N₂O-N reflects the high share of soil-derived N₂O, and was approximately half of total cumulative N₂O emissions (Table 4). Significant differences in primed N₂O-N among digestates followed the same trend as N₂O emissions. For ${\text{NH}}_4^{+}$-N-labeled digestates, N₂O-N losses which might originate from digestate N_org, were not accounted for. Therefore, the amount of primed N₂O-N might be overestimated.

In general, the rather comparable share of total digestate-derived N₂O-N losses among the digestates with different labeling approaches indicates that digestate-N_org plays only a minor role in short-term N₂O formation. Senbayram et al. (2014) labeled only the mineral N fraction of a digestate and found 31% of N₂O-N was derived from the digestate mineral fraction. The share of digestate-derived N₂O-N losses among ${\text{NH}}_4^{+}$-N-labeled digestates FW, OW and CS ranged from 15.7 to 24.2% over the 2 years. For these digestates as well as for the digestates in the study of Senbayram et al. (2014) it cannot be excluded that non-labeled organic N was mineralized and emitted as N₂O. However, the N_t-labeled digestates, M, G, SB, and SBL showed a rather comparable range with 18.2–37.8% digestate-N being emitted as N₂O over the 2 years. Similar to our findings, also other studies reported a higher share of N₂O-N originating from the soil N pool than from fertilizer N. For instance, only 22% of N₂O-N was derived from ¹⁵N-labeled manure after 22 days (Ingold et al., 2018) or 40.4% from ¹⁵N-urea after 35 days (Roman-Perez and Hernandez-Ramirez, 2020) in incubation experiments. In a field study, ${\text{NH}}_4^{+}$-N-labeled cattle slurry was applied, which produced higher fertilizer-derived N₂O emissions within the first 10 days, but higher soil-derived N₂O 11–22 days after application (Dittert et al., 2001). However, the study was carried out on grassland and using the injection technique (Dittert et al., 2001), which has been reported to increase N₂O emissions compared to trail hose application with immediate incorporation (Herr et al., 2019).

N₂O Emissions Affected by Fertilizer Type

As previously described, N₂O fluxes were shaped and influenced by weather conditions and soil microbial processes. Environmental conditions and soil type may play a more important role than the fertilizer type, as previously suggested by Senbayram et al. (2014): the authors noted no significant differences in N₂O emissions between mineral and organic N fertilization. However, in both years, significant differences among digestates were noted on several sampling dates, for N₂O as well as ¹⁵N-N₂O fluxes (e.g., flux rates from M digestate vs. fluxes from SB digestate in Figure 2 and Supplementary Tables 1–3), indicating that digestate composition affects N₂O emissions. This supports hypothesis (1), that digestates from different feedstocks will differ in N₂O flux rates.

However, regarding cumulative effects, there was no clear indication that the digestate type influenced total N₂O emissions. This was supported by the lack of a significant correlation between digestate composition (${\text{NH}}_4^{+}$/N, C/N, C_org/N_org) and cumulative N₂O or ¹⁵N₂O. Only measurements of the 1st year showed significant differences among digestates. Therefore, hypothesis (2) had to be rejected for the 2nd year and could be partly accepted for the 1st year. Yet, when separating cumulative ¹⁵N-N₂O data into N_t-labeled and ${\text{NH}}_4^{+}$-N-labeled digestates, there was a significant effect of C/N ratio in the 1st year, predicting 22.1% of ¹⁵N-N₂O emissions of ¹⁵N_t-labeled digestates (R² = 0.36, F-value = 3.81, p-value = 0.077). For the 2nd year, C_org/N_org accounted for 39.2% (R² = 0.56, F-value = 9.85, p-value = 0.0094) of ¹⁵N-N₂O emissions among ¹⁵N_t-digestates (M, G, SB, and SBL). Also Abubaker et al. (2013) noted significantly different cumulative N₂O emissions after 24 days between two types of urban waste digestates, which were low or high organic C. For ${\text{NH}}_4^{+}$-N-labeled digestates, there was no significant relation of digestate properties to ¹⁵N-N₂O emissions. Hence, the correlation between digestate properties and N₂O emissions seems more strongly related to the total N and N_org content of digestates, than ${\text{NH}}_4^{+}$-N. Regarding the total share of digestate-derived N₂O-N (Nd), significant differences among digestates (Table 5) could support hypothesis (3).

Ultimately, the results of the present study suggest that the different digestate types influenced cumulative N₂O, flux rates and digestate derived N₂O-N only marginally. Hence, N₂O emissions were more strongly affected by environmental conditions (Table 3). The effect of digestate properties on total N₂O emissions was overlaid to some extent by the high amount of N₂O from the native soil N pool. Abubaker et al. (2013) incubated two digestates in three soil textures and noted considerable differences regarding emission peaks and cumulative N₂O-N emissions among digestates, particularly in the sandy soil. In loam, digestates showed comparable total N₂O emissions (Abubaker et al., 2013). Therefore, N₂O emissions discussed in this study might differ on soils with different soil textures or amendment history (Rosace et al., 2020). In this context, soil texture, soil amendment history and fertility status, especially OM content, plays a crucial role, exceeding the effect of digestate properties.

Digestate Emission Factors and Practical Consequences

Digestate EFs determined in this study (0.21–0.75%) were all within the range of the IPCC default value, except for SBL in the 1st year (Table 5). However, these EFs will not cover the whole year and might underestimate the total EF of the digestates. Shang et al. (2020) determined 10–30% lower EFs when only the growing season and not the whole year N₂O emissions were considered. Moreover, the authors found that the differences between EFs of the whole year and growing season were higher with higher precipitation (Shang et al., 2020). The experimental design of the present study used bare soil, hence there were no crops removing the applied digestate N. Crop N uptake could have decreased available N from the soil as well as soil moisture, which could have lowered digestate-derived N₂O emissions and EFs. Thorman et al. (2020) determined annual N₂O EFs from different organic amendments, top-dressed to a cereal crop (0.15–0.73% in 2011 and 0.27–0.51% in 2012), which were in a comparable range with our EFs. Most digestate EFs did not show significant differences, except SBL compared with CS in the 1st year, thus hypothesis (2) cannot be fully confirmed.

Soil derived N₂O-N contributed to a large extent to digestate EFs. As a consequence of the high share of N₂O-N from the native soil pool within the first 30 days after digestate application, crop cultivation should be synchronized with available soil N. In particular, mineral N from the soil pool should be taken up by the crops, before digestates are applied. Thereby, the triggering effect of short-term soil-enhanced N₂O emissions by digestates could be decreased. For example, N_min supply in the present study would be sufficient for maize cultivation in the early growth stage. Digestates could then be top-dressed ~1 month after emergence when most of soil M was already taken up by the crop. Also de Neve (2017) emphasized that in ideal cropping systems fertilizer availability and soil mineral N should be synchronized with crop demand, which could mitigate potential N losses.

Experimental Limitations

Determination of N₂O isotopomers in the present study, including the δ¹⁸O and site preference of ¹⁵N in the N₂O molecule, could have helped to understand the underlying soil microbial processes, differentiating between denitrification and nitrification (Köster et al., 2015). Yet, distinguishing nitrifier denitrification from nitrification is not possible using site preference (Köster et al., 2011). A dual isotope labeling approach of ¹⁵N and ¹⁸O-labeled water would be required (Koola et al., 2010), which is not feasible in field studies (Baggs, 2008). Also the N₂O/N₂O+ N₂ product ratio could have provided a better indication of denitrification in the study (Buchen-Tschiskale et al., 2020). However, measuring N₂ in the field is rather difficult due to the high N₂ background level in the atmosphere, as well as its spatial and temporal heterogeneity (Groffman et al., 2009). Instead N₂ is often studied in incubation experiments using an artificial helium–oxygen atmosphere (Scholefield et al., 1997). Regular soil N_min and ¹⁵N_min analysis at the sampling dates could have given a hint for respective microbial processes, but would not have completely identified them. Thus, allocating the specific N₂O pathways after digestate application in the field is still challenging and needs further research and suitable methods to provide accurate measurements (Well et al., 2019).

Conclusion

The major finding of this study was the large share of N₂O-N from the soil pool, showing that digestate application triggers “enhanced soil-derived N₂O.” The major driving forces of the emission pattern are the weather conditions, the specific chemical composition of digestates do have only minor effects on the denitrification. The different ¹⁵N-labeling approaches of the digestates indicate that contribution of the organic fraction seems to be of very low significance for short-term N₂O emissions. The ¹⁵N labeling approach helped to determine the source of N₂O emissions, but not the underlying processes (nitrifier denitrification or heterotrophic denitrification). Analysis of isotopomers and N₂ is needed to further identify the N₂O-releasing microbial processes in the soil. Emission factors were comparable for most digestates, but reached and even exceeded the default IPPC EF (0.6%) within only 60 days in the 1st year.

Data Availability Statement

The original contributions presented in the study are included in the article/Supplementary Materials, further inquiries can be directed to the corresponding author/s.

Author Contributions

FH prepared original draft, including statistical analysis, and graph production. RR supported calculation of global warming potential, provided guidance, and new input for determination of indirect N₂O emissions. The basis for the experimental idea was based on KM, with further contribution by RR regarding design of the field experiment, sampling frequency, and measurement technique. IC-M supported calculation of ¹⁵N abundance in N₂O from IRMS data. FH and IC-M conducted the experiment. All authors reviewed and proofread the whole manuscript and gave critical feed-back to all sections.

Funding

The experiment was conducted with financial support of the BMEL and the Fachangentur Nachwachsende Rohstoffe (FNR) e.V. within the project FNR Gärresteigenschaft.

Conflict of Interest

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

Acknowledgments

The anaerobic batch digestion of ¹⁵N-labeled plant substrates was performed by the State Institute of Agricultural Engineering and Bioenergy at the University of Hohenheim with the support of the director Dr. Oechsner and the scientist Natalie Böttcher. Analysis of substrates and soil via IRMS, was done by Elke Dachtler, and ¹⁵N-gas measurements with support by Wolfgang Armbruster and IC-M. Gas sampling in the field was supported by IC-M, Regina Roth, and student assistants. We are grateful for the help with flux calculation by Christina Herr and RR, as well as the experimental support by Heidi Zimmermann and the working group 340i. Our gratitude to Hans Buegger, from Helmholtz Zentrum Muenich for providing ¹⁵N-N₂O standards for IRMS gas measurements, and supporting calculation of ¹⁵N-N₂O.

Supplementary Material

The Supplementary Material for this article can be found online at: https://www.frontiersin.org/articles/10.3389/fsufs.2020.614349/full#supplementary-material

References

Abubaker, J., Odlare, M., and Pell, M. (2013). Nitrous oxide production from soils amended with biogas residues and cattle slurry. J. Environ. Qual. 42, 1046–1058. doi: 10.2134/jeq2012.0247

PubMed Abstract | CrossRef Full Text | Google Scholar

Baggs, E. M. (2008). A review of stable isotope techniques for N2O source partitioning in soils: recent progress, remaining challenges and future considerations y. Rapid Commun. Mass Spectrom. 22, 1664–1672. doi: 10.1002/rcm.3456

CrossRef Full Text | Google Scholar

Brand, W. A. (1995). Precon: a fully automated interface for the PRE-GC concentration of trace gases in air for isotopic analysis. Isotopes Environ. Health Stud. 31, 277–284. doi: 10.1080/10256019508036271

CrossRef Full Text | Google Scholar

Bremner, J. M. (1997). Sources of nitrous oxide in soils. Nutr. Cycle Agroecosyst. 49, 7–16. doi: 10.1023/A:1009798022569

CrossRef Full Text | Google Scholar

Brulé, M. (2014). The Effect of Enzyme Additives on the Anaerobic Digestion of Energy Crops. (Dissertation), University of Hohenheim, Stuttgart, Germany.

Google Scholar

Buchen-Tschiskale, C., Hagemann, U., and Augustin, J. (2020). Soil incubation study showed biogas digestate to cause higher and more variable short-term N₂O and N₂ fluxes than mineral-N. J. Plant Nutr. Soil Sci. 183, 208–219. doi: 10.1002/jpln.201900075

CrossRef Full Text | Google Scholar

Cavalli, D., Cabassi, G., Borrelli, L., Geromel, G., Bechini, L., Degano, L., et al. (2016). Nitrogen fertilizer replacement value of undigested liquid cattle manure and digestates. Eur. J. Agron. 73, 34–41. doi: 10.1016/j.eja.2015.10.007

CrossRef Full Text | Google Scholar

Chadwick, D. R., Pain, B. F., and Brookman, S. K. E. (2000). Nitrous oxide and methane emissions following application of animal manures to grassland. J. Environ. Qual. 29, 277–287. doi: 10.2134/jeq2000.00472425002900010035x

CrossRef Full Text | Google Scholar

Chen, R., Senbayram, M., Lin, X., and Dittert, K. (2011). Origin of positive δ13C of emitted CO2 from soils after application of biogas residues. Soil Biol. Biochem. 43, 2194–2199. doi: 10.1016/j.soilbio.2011.07.003

CrossRef Full Text | Google Scholar

Ciais, P., Sabine, C., Bala, G., Bopp, L., Brovkin, V., Canadell, J., et al. (2013). “Carbon and other biochemical cycles,” in Climate Change 2013: The Physical Science Basis. Contribution of Working Group I to the Fifth Assessment Report of the Intergovernmental Panel on Climate Change, eds T. F. Stocker, D. Qin, G.-K. Plattner, M. Tignor, S. K. Allen, J. Boschung, A. Nauels, Y. Xia, V. Bex and P. M. Midgley (Cambridge; New York, NY: Cambridge University Press).

PubMed Abstract | Google Scholar

Comfort, S. D., Kelling, K. A., Keeney, D. R., and Converse, J. C. (1990). Nitrous oxide production from injected liquid dairy manure. Soil Sci. Soc. Am. J. 54, 421–427. doi: 10.2136/sssaj1990.03615995005400020021x

CrossRef Full Text | Google Scholar

Corden, C., Bougas, K., Cunningham, E., Tyrer, D., Kreißig, J., Zetti, E., et al. (2019). Digestate and Compost as Fertilisers: Risk Assessment and Risk Management Options. 121, 128, B2, D13. Available online at: https://ec.europa.eu/environment/chemicals/reach/pdf/40039%20Digestate%20and%20Compost%20RMOA%20-%20Final%20report%20i2_20190208.pdf (accessed December 21, 2020).

Davidson, E. A. (1991). “Fluxes of nitrous oxide and nitric oxide from terrestrial ecosystems,” in Microbial Production and Consumption of Greenhouse Gases: Methane, Nitrogen Oxides and Halomethanes, eds J. E. Rogers, and W. B. Whitman (Washington, DC: American Society of Microbiology), 219–335.

Google Scholar

de Neve, S. (2017). “Organic matter mineralization as a source of nitrogen BT -advances in research on fertilization management of vegetable crops,” in, eds F. Tei, S. Nicola, and P. Benincasa (Cham: Springer International Publishing), 65–83. doi: 10.1007/978-3-319-53626-2_3

CrossRef Full Text

DIN 10693 (2014). Soil quality - Determination of carbonate content - Volumetric method (ISO 10693:1995). German version EN ISO 10693:2014.

Dittert, K., Bol, R., King, R., Chadwick, D., and Hatch, D. (2001). Use of a novel nitrification inhibitor to reduce nitrous oxide emission from 15N-labelled dairy slurry injected into soil. Rapid Commun. Mass Spectrom. 15, 1291–1296. doi: 10.1002/rcm.335

PubMed Abstract | CrossRef Full Text | Google Scholar

Engels, C. (1999). Regulation of xylem transport of calcium from roots to shoot of maize by growth-related demand. J. Plant Nutr. Soil Sci. 162, 287–294.

Google Scholar

FAO (2020). FAOSTAT Emissions Shares. Available online at: www.fao.org/faostat/en/#data/REM (accessed August 20, 2020).

Flessa, H., and Beese, F. (1995). Effects of sugarbeet residues on soil redox potential and nitrous oxide emission. Soil Sci. Soc. Am. J. 59, 1044–1051. doi: 10.2136/sssaj1995.03615995005900040013x

CrossRef Full Text | Google Scholar

Flessa, H., and Beese, F. (2000). Laboratory estimates of trace gas emissions following surface application and injection of cattle slurry. J. Environ. Qual. 29, 262–268. doi: 10.2134/jeq2000.00472425002900010033x

CrossRef Full Text | Google Scholar

Flessa, H., Dörsch, P., and Beese, F. (1995). Seasonal variation of N2O and CH4 fluxes in differently managed arable soils in southern Germany. J. Geophys. Res. Atmos. 100, 23115–23124. doi: 10.1029/95JD02270

CrossRef Full Text | Google Scholar

Fouda, S., Von Tucher, S., Lichti, F., and Schmidhalter, U. (2013). Nitrogen availability of various biogas residues applied to ryegrass. J. Plant Nutr. Soil Sci. 176, 572–584. doi: 10.1002/jpln.201100233

CrossRef Full Text | Google Scholar

Fuß, R., and Asger, R. P. (2014). Greenhouse Gas Flux Calculation From Chamber Measurements. R Packag. Version 0.1. Available online at: https://bitbucket.org/ecoRoland/gasfluxes/src/master/ (accessed August 8, 2016).

Georgacakis, D., Sievers, D. M., and Iannotti, E. L. (1982). Buffer stability in manure digesters. Agric. Wastes 4, 427–441. doi: 10.1016/0141-4607(82)90038-5

CrossRef Full Text | Google Scholar

Granli, T., and Bøckman, O. C. (1994). Nitrous oxide from agriculture. Nor. J. Agric. Sci. Suppl. 12, 1–128.

Google Scholar

Groffman, P. M., Butterbach-Bahl, K., Fulweiler, R. W., Gold, A. J., Morse, J. L., Stander, E. K., et al. (2009). Challenges to incorporating spatially and temporally explicit phenomena (hotspots and hot moments) in denitrification models. Biogeochemistry 93, 49–77. doi: 10.1007/s10533-008-9277-5

CrossRef Full Text | Google Scholar

Guilayn, F., Rouez, M., Crest, M., Patureau, D., and Jimenez, J. (2020). Valorization of digestates from urban or centralized biogas plants: a critical review. Rev. Environ. Sci. Biotechnol. 19, 419–462. doi: 10.1007/s11157-020-09531-3

CrossRef Full Text | Google Scholar

Gutser, R., Ebertseder, T., Weber, A., Schraml, M., and Schmidhalter, U. (2005). Short-term and residual availability of nitrogen after long-term application of organic fertilizers on arable land. J. Plant Nutr. Soil Sci. 168, 439–446. doi: 10.1002/jpln.200520510

CrossRef Full Text | Google Scholar

Guzman-Bustamante, I., Winkler, T., Schulz, R., Müller, T., Mannheim, T., Laso Bayas, J. C., et al. (2019). N2O emissions from a loamy soil cropped with winter wheat as affected by N-fertilizer amount and nitrification inhibitor. Nutr. Cycl. Agroecosystems 114, 173–191. doi: 10.1007/s10705-019-10000-9

CrossRef Full Text | Google Scholar

Hafner, S. D., Pacholski, A., Bittman, S., Carozzi, M., Chantigny, M., Génermont, S., et al. (2019). A flexible semi-empirical model for estimating ammonia volatilization from field-applied slurry. Atmos. Environ. 199, 474–484. doi: 10.1016/j.atmosenv.2018.11.034

CrossRef Full Text | Google Scholar

Heincke, M., and Kaupenjohann, M. (1999). Effects of soil solution on the dynamics of N₂O emissions: a review. Nutr. Cycle Agroecosyst. 55, 133–157. doi: 10.1023/A:1009842011599

CrossRef Full Text | Google Scholar

Herr, C., Mannheim, T., Müller, T., and Ruser, R. (2019). Effect of cattle slurry application techniques on N₂O and NH3 emissions from a loamy soil. J. Plant Nutr. Soil Sci. 182, 964–979. doi: 10.1002/jpln.201800376

CrossRef Full Text | Google Scholar

Herr, C., Mannheim, T., Müller, T., and Ruser, R. (2020). Effect of nitrification on N₂O effect of nitrification inhibitors on N₂O emissions after cattle slurry application. Agron. J. 40, 876–884. doi: 10.3390/agronomy10081174

CrossRef Full Text | Google Scholar

Holly, M. A., Larson, R. A., Powell, J. M., Ruark, M. D., and Aguirre-Villegas, H. (2017). Greenhouse gas and ammonia emissions from digested and separated dairy manure during storage and after land application. Agric. Ecosyst. Environ. 239, 410–419. doi: 10.1016/j.agee.2017.02.007

CrossRef Full Text | Google Scholar

Hutchinson, G. L., and Mosier, A. R. (1981). Improved soil cover method for field measurement of nitrous oxide fluxes. Soil Sci. Soc. Am. J. 45, 311–316. doi: 10.2136/sssaj1981.03615995004500020017x

CrossRef Full Text | Google Scholar

Ingold, M., Khanal, G., Dyckmans, J., Wachendorf, C., and Buerkert, A. (2018). NH₃ volatilization, N₂O emission and microbial biomass turnover from 15N-labeled manure under laboratory conditions. Commun. Soil Sci. Plant Anal. 49, 537–551. doi: 10.1080/00103624.2018.1431271

CrossRef Full Text | Google Scholar

IPCC (2019). 2019 Refinement To the 2006 IPCC Guidelines for National Greenhouse Gas Inventories. eds S. Calvo Buendia, E. Tanabe, K. Kranjc, A. Baasansuren, J. Fukuda, M. Ngarize, et al. (Switzerland: IPCC).

PubMed Abstract | Google Scholar

IUSS Working Group (2015). World Reference Base for Soil Resources 2014, Update 2015. International Soil Classification System for Naming Soils and Creating Legends for Soil Maps. World Soil Resources Reports No. 106. Rome: FAO.

Johansen, A., Carter, M. S., Jensen, E. S., Hauggard-Nielsen, H., and Ambus, P. (2013). Effects of digestate from anaerobically digested cattle slurry and plant materials on soil microbial community and emission of CO₂ and N₂O. Appl. Soil Ecol. 63, 36–44. doi: 10.1016/j.apsoil.2012.09.003

CrossRef Full Text | Google Scholar

Koola, D. M., Wragec, N., Zechmeister-Boltensternd, S., Pfefferd, M., Brusa, D., van Groenigen, J. W., et al. (2010). Special issue article Nitrifier denitrification can be a source of N²O from soil: a revised approach to the dual-isotope labelling method. Eur. J. Soil Sci. 61, 759–772. doi: 10.1111/j.1365-2389.2010.01270.x

CrossRef Full Text | Google Scholar

Köster, J. R., Cárdenas, L., Senbayram, M., Bol, R., Well, R., Butler, M., et al. (2011). Rapid shift from denitrification to nitrification in soil after biogas residue application as indicated by nitrous oxide isotopomers. Soil Biol. Biochem. 43, 1671–1677. doi: 10.1016/j.soilbio.2011.04.004

CrossRef Full Text | Google Scholar

Köster, J. R., Cárdenas, L. M., Bol, R., Lewicka-Szczebak, D., Senbayram, M., Well, R., et al. (2015). Anaerobic digestates lower N₂O emissions compared to cattle slurry by affecting rate and product stoichiometry of denitrification—An N²O isotopomer case study. Soil Biol. Biochem. 84, 65–74. doi: 10.1016/j.soilbio.2015.01.021

CrossRef Full Text | Google Scholar

Möller, K. (2015). Effects of anaerobic digestion on soil carbon and nitrogen turnover, N emissions, and soil biological activity. A review. Agron. Sustain. Dev. 35, 1021–1041. doi: 10.1007/s13593-015-0284-3

CrossRef Full Text | Google Scholar

Möller, K., and Müller, T. (2012). Effects of anaerobic digestion on digestate nutrient availability and crop growth: a review. Eng. Life Sci. 12, 242–257. doi: 10.1002/elsc.201100085

CrossRef Full Text | Google Scholar

Möller, K., and Stinner, W. (2009). Effects of different manuring systems with and without biogas digestion on soil mineral nitrogen content and on gaseous nitrogen losses (ammonia, nitrous oxides). Eur. J. Agron. 30, 1–16. doi: 10.1016/j.eja.2008.06.003

CrossRef Full Text | Google Scholar

Mönch-Tegeder, M., Lemmer, A., Jungbluth, T., and Oechsner, H. (2014). Effects of full-scale substrate pretreatment with a cross-flow grinder on biogas production. Agric. Eng. Int. CIGR J. 16, 138–147.

Google Scholar

Petersen, S. O. (1999). Nitrous oxide emissions from manure and inorganic fertilizers applied to spring barley. J. Environ. Qual. 28, 1610–1618. doi: 10.2134/jeq1999.00472425002800050027x

CrossRef Full Text | Google Scholar

Petersen, S. O., Nissen, H. H., Lund, I., and Ambus, P. (2003). Redistribution of slurry components as influenced by injection method, soil, and slurry properties. J. Environ. Qual. 32, 2399–2409. doi: 10.2134/jeq2003.2399

PubMed Abstract | CrossRef Full Text | Google Scholar

Pezzolla, D., Bol, R., Gigliotti, G., Sawamoto, T., López, A. L., Cardenas, L., et al. (2012). Greenhouse gas (GHG) emissions from soils amended with digestate derived from anaerobic treatment of food waste. Rapid Commun. Mass Spectrom. 26, 2422–2430. doi: 10.1002/rcm.6362

PubMed Abstract | CrossRef Full Text | Google Scholar

Pfab, H. (2011). Nitrous oxide emission and mitigation strategies—measurements on an intensively fertilized vegetable cropped loamy soil. (Dissertation). University of Hohenheim, Stuttgart, Germany.

Google Scholar

Pfab, H., Palmer, I., Buegger, F., Fiedler, S., Müller, T., and Ruser, R. (2011). N₂O fluxes from a Haplic Luvisol under intensive production of lettuce and cauliflower as affected by different N-fertilization strategies. J. Plant Nutr. Soil Sci. 174, 545–553. doi: 10.1002/jpln.201000123

CrossRef Full Text | Google Scholar

Pfab, H., Palmer, I., Buegger, F., Fiedler, S., Müller, T., and Ruser, R. (2012). Influence of a nitrification inhibitor and of placed N-fertilization on N₂O fluxes from a vegetable cropped loamy soil. Agric. Ecosyst. Environ. 150, 91–101. doi: 10.1016/j.agee.2012.01.001

CrossRef Full Text | Google Scholar

R Core Team (2016). R: A Language and Environment for Statistical Computing. Vienna: R Foundation for Statistical Computing.

Google Scholar

Rochette, P., Angers, D. A., Chantigny, M. H., Gagnon, B., and Bertrand, N. (2008). N₂O fluxes in soils of contrasting textures fertilized with liquid and solid dairy cattle manures. Can. J. Soil Sci. 88, 175–187. doi: 10.4141/CJSS06016

CrossRef Full Text | Google Scholar

Roman-Perez, C. C., and Hernandez-Ramirez, G. (2020). Sources and priming of N₂O production across a range of moisture contents in a soil with high organic matter. J. Environ. Qual. doi: 10.1002/jeq2.20172. [Epub ahead of print].

PubMed Abstract | CrossRef Full Text | Google Scholar

Rosace, M. C., Veronesi, F., Briggs, S., Cardenas, L. M., and Jeffery, S. (2020). Legacy effects override soil properties for CO₂ and N₂O but not CH₄ emissions following digestate application to soil. GCB Bioenergy 12, 445–457. doi: 10.1111/gcbb.12688

PubMed Abstract | CrossRef Full Text | Google Scholar

Ruser, R., Flessa, H., Russow, R., Schmidt, G., Buegger, F., and Munch, J. C. (2006). Emission of N₂O, N₂ and CO₂ from soil fertilized with nitrate: effect of compaction, soil moisture and rewetting. Soil Biol. Biochem. 38, 263–274. doi: 10.1016/j.soilbio.2005.05.005

CrossRef Full Text | Google Scholar

Ruser, R., Flessa, H., Schilling, R., Beese, F., and Munch, J. C. (2001). Effect of crop-specific field management and N fertilization on N₂O emissions from a fine-loamy soil. Nutr. Cycl. Agroecosyst. 59, 177–191. doi: 10.1023/A:1017512205888

CrossRef Full Text | Google Scholar

Ruser, R., Fuß, R., Andres, M., Hegewald, H., Kesenheimer, K., Köbke, S., et al. (2017). Nitrous oxide emissions from winter oilseed rape cultivation. Agric. Ecosyst. Environ. 249, 57–69. doi: 10.1016/j.agee.2017.07.039

CrossRef Full Text | Google Scholar

SAS Institute Inc (2015). SAS/STAT^® 14.1 User's Guide. Cary, NC: SAS Institute Inc.

Google Scholar

Schleusner, P., Lammirato, C., Tierling, J., Lebender, U., and Rütting, T. (2018). Primed N2O emission from native soil nitrogen: a 15N-tracing laboratory experiment. J. Plant Nutr. Soil Sci. 181, 621–627. doi: 10.1002/jpln.201700312

CrossRef Full Text | Google Scholar

Scholefield, D., Hawkins, J. M. B., and Jackson, S. M. (1997). Development of a helium atmosphere soil incubation technique for direct measurement of nitrous oxide and dinitrogen fluxes during denitrification. Soil Biol. Biochem. 29, 1345–1352. doi: 10.1016/S0038-0717(97)00021-7

CrossRef Full Text | Google Scholar

Schouten, S., van Groenigen, J. W., Oenema, O., and Cayuela, M. L. (2012). Bioenergy from cattle manure? Implications of anaerobic digestion and subsequent pyrolysis for carbon and nitrogen dynamics in soil. GCB Bioenergy 4, 751–760. doi: 10.1111/j.1757-1707.2012.01163.x

CrossRef Full Text | Google Scholar

Senbayram, M., Chen, R., Mühling, K. H., and Dittert, K. (2009). Contribution of nitrification and denitrification to nitrous oxide emissions from soils after application of biogas waste and other fertilizers. Rapid Commun. Mass Spectrom. 23, 2489–2498. doi: 10.1002/rcm.4067

PubMed Abstract | CrossRef Full Text | Google Scholar

Senbayram, M., Chen, R., Wienforth, B., Herrmann, A., Kage, H., Mühling, K. H., et al. (2014). Emission of N₂O from biogas crop production systems in Northern Germany. Bioenergy Res. 7, 1223–1236. doi: 10.1007/s12155-014-9456-2

CrossRef Full Text | Google Scholar

Shang, Z., Abdalla, M., Kuhnert, M., Albanito, F., Zhou, F., Xia, L., et al. (2020). Measurement of N₂O emissions over the whole year is necessary for estimating reliable emission factors. Environ. Pollut. 259:113864. doi: 10.1016/j.envpol.2019.113864

PubMed Abstract | CrossRef Full Text | Google Scholar

Tamir, G., Shenker, M., Heller, H., Bloom, P. R., Fine, P., and Bar-Tal, A. (2013). Organic N mineralization and transformations in soils treated with animal waste in relation to carbonate dissolution and precipitation. Geoderma 209–210, 50–56. doi: 10.1016/j.geoderma.2013.05.028

CrossRef Full Text | Google Scholar

Thorman, R. E., Nicholson, F. A., Topp, C. F. E., Bell, M. J., Cardenas, L. M., Chadwick, D. R., et al. (2020). Towards country-specific nitrous oxide emission factors for manures applied to arable and grassland soils in the UK. Front. Sustain. Food Syst. 4:62. doi: 10.3389/fsufs.2020.00062

CrossRef Full Text | Google Scholar

VDI 4630 (2016). Fermentation of Organic Materials—Characterisation of the Substrate, Sampling, Collection of Material Data, Fermentation Tests. Berlin: The Association of German Engineers (VDI).

Verdi, L., Kuikman, P. J., Orlandini, S., Mancini, M., Napoli, M., and Dalla Marta, A. (2019). Does the use of digestate to replace mineral fertilizers have less emissions of N₂O and NH₃? Agric. For. Meteorol. 269–270, 112–118. doi: 10.1016/j.agrformet.2019.02.004

CrossRef Full Text | Google Scholar

Well, R., Maier, M., Lewicka-Szczebak, D., Köster, J. R., and Ruoss, N. (2019). Underestimation of denitrification rates from field application of the 15N gas flux method and its correction by gas diffusion modelling. Biogeosciences 16, 2233–2246. doi: 10.5194/bg-16-2233-2019

CrossRef Full Text | Google Scholar

Wulf, S., Maeting, M., and Clemens, J. (2002). Application technique and slurry co-fermentation effects on ammonia, nitrous oxide, and methane emissions after spreading. J. Environ. Qual. 31, 1795. doi: 10.2134/jeq2002.1795

CrossRef Full Text | Google Scholar

Zirkler, D., Peters, A., and Kaupenjohann, M. (2014). Elemental composition of biogas residues: variability and alteration during anaerobic digestion. Biomass Bioenergy 67, 89–98. doi: 10.1016/j.biombioe.2014.04.021

CrossRef Full Text | Google Scholar