Abstract
The study aimed to measure soil-atmosphere N2O fluxes and their controlling factors, as well as NH3 emissions and yields for two soils (silt loam and clay loam) in three management systems over two years under subsequent wheat and maize cultivation. The management systems were characterized as follows: (1) cash crop (C) with mineral fertilizer and conventional tillage; (2) livestock (L) with biogas residue fertilization and its incorporation prior to sowing in maize and reduced tillage; and (3) climate optimized (O) with minimum tillage, 8-year crop rotation, with biogas residue fertilization, in maize without incorporation in clay loam soil or incorporation by strip-tillage prior to seeding in silt loam soil. Stable isotope ratios of N2O and mineral N were determined to identify N2O processes. Within the organically fertilized maize treatments, cumulative N2O fluxes were highest in the O-system treatments of both sites (4.0 to 9.4 kg N ha− 1 a− 1), i.e. more than twice as high as in the L-system (1.5 to 3.1 kg N ha− 1 a− 1). Below root-strip till fertilizer application did not enhance N2O fluxes. Fluxes with mineral fertilization of wheat (1.1 to 3.1 kg N ha− 1 a− 1) were not different from those with organic fertilization. Isotopic values of emitted N2O revealed that bacterial denitrification dominated most of the peak flux events, while the N2O/(N2 + N2O) ratio of denitrification was mostly between 0.1 and 0.5. It can be concluded that, contrary to the intention to lower greenhouse gas fluxes by the O-system management, the highest N2O fluxes occurred in the O-system without biogas digestate incorporation in maize. With respect to NH3 fluxes, we could confirm that the application of digestate application in growing crops without incorporation or late incorporation in fertilization before sowing induces high fluxes. The beneficial aspects of the O-system including more stable soil structure and resource conservation, are thus potentially counteracted by increased N2O and NH3 emissions.
Avoid common mistakes on your manuscript.
Introduction
The global population is increasing, resulting in a growing demand for agricultural products. This, in turn, is leading to an increased demand for enhanced productivity of agricultural land and the global consumption of nitrogen (N) fertilizers (Erisman et al. 2011). Moreover, the current prevalence of diets with an increasing share of animal products worldwide, is leading to increased number of livestock, which in turn requires the optimization of manure fertilization in terms of nutrient efficiency and environmental impact. Soil N losses through emissions of gaseous N species such as ammonia (NH3), nitric oxide (NO), nitrous oxide (N2O), and dinitrogen (N2), and nitrate (NO3−) leaching impair crop yields and N use efficiency (NUE), but also contribute to air pollution, global warming, eutrophication and acidification of unmanaged soils and waters, and cause NO3− contamination of drinking water resources (Galloway et al. 2008). In Germany, field application of manure and digestate from biogas plants contributes to the NH3 emissions that exceed national emission caps and NO3− concentrations in groundwater recharge often exceed EU drinking water limits (Hey et al. 2015). Depending on the management system, the role of digestate fertilization vs. mineral N fertilization in terms of N emissions and NUE may differ. This may depend on the management system, where the amount and application techniques of organic fertilizers, such as manure or biogas digestate, as well as crop rotations vary between the management systems of livestock and cash crop farms (Loyon 2018). Moreover, while conventional tillage systems allow for manure incorporation, reduced or zero tillage limit manure incorporation (Webb et al. 2010), which may alter N emissions and NUE.
Nitrous oxide is a significant greenhouse gas and a contributor to the destruction of stratospheric ozone (Ravishankara et al. 2009). The majority of N2O emissions resulting from agricultural activity originate from the soil. The magnitude of these emissions is related to N management, which consists of the application and incorporation of mineral and organic fertilizers and crop residues (Flessa et al. 2014). While a multitude of N2O-producing soil processes are currently known (Butterbach-Bahl et al. 2013), the microbial processes of nitrification and bacterial denitrification are responsible for the majority of emissions (Müller et al. 2014; Wrage-Mönnig et al. 2018). The role of fungal denitrification was previously assumed to be highly relevant based on laboratory studies with microbial inhibitors (Laughlin and Stevens 2002; Ma et al. 2017). However, isotopic studies suggest that inhibitor studies may overestimate its role (Rohe et al. 2021). Nevertheless, recent isotopic studies indicated potential for impact from NO3−, litter incorporation, and rhizosphere processes (Senbayram et al. 2018, 2020). A high proportion of fungal denitrification in total N2O production may result in enhanced N2O fluxes. However, fungal denitrifiers lack N2O reductase, which may act as a limiting factor in N2O reduction (Shoun and Fushinobu 2016).
The gross formation of N2O is dependent upon the availability of N substrates for nitrification (NH4+) and denitrification (NO3−), the abundance of decomposable organic matter, and the structure of microbial community. In addition, these processes are controlled by pore space oxygen (O2) concentration (aerobic conditions enhance nitrification and inhibit denitrification; anaerobic conditions vice versa), pH, and temperature. Finally, net N2O fluxes at the soil-atmosphere interface also depend on the extent of N2O reduction to N2, which is also controlled by the combination of the above factors. N2O reduction is increased when gas exchange between soil and atmosphere is inhibited (Müller and Clough 2014) and when copper availability does not limit N2O reductase formation (Shen et al. 2020).
Fertilizer application techniques such as incorporation by tillage, banding, or injection are used to promote crop NUE and reduce NH3 losses (Petersen et al. 2004; Webb et al. 2010). The German fertilizer ordinance requires the incorporation of organic fertilizer within the first four hours on arable fields without growing crops (Bundesministerium der Justiz 2020).
In general, the restriction of the applied fertilizer to a limited volume of soil can reduce N2O fluxes, but the opposite effect cannot be excluded due to several factors. Organic fertilization, such as manure or digestate, affects N2O processes through the simultaneous supply of mineral N and mineralizable organic N, as well as labile organic C (Grosz et al. 2022). The latter process results in the consumption of O2 during the respiratory decomposition process. In the boundary zone between manure and bulk soil, this potentially leads to local co-occurrence of anoxia and high NO3− concentration from nitrification of manure-derived NH4+. Water retention in manure leads to elevated water content in manure clumps (Petersen et al. 2003). These factors can result in high denitrification rates (Petersen et al. 2016). The N2O/(N2 + N2O) ratios of denitrification will vary considerably and N2O reduction is inhibited by locally high NO3− concentrations and low pH due to acidification by previous nitrification of manure-derived NH4+. High local concentrations of manure-derived NH4+ may lead to elevated N2O yields from nitrification (Deppe et al. 2017). The same probably applies for biogas digestates. Moreover, nitrification itself represents a significant O2 sink in soil, thereby further enhancing denitrification. The concentration of NH3 may become inhibitory to bacterial nitrifiers at certain pH levels, resulting in the accumulation of NO2− and an increase in the production of N2O (Venterea et al. 2015). Several studies have thus shown that tillage and special fertilization techniques can significantly impact the emission of N2O (Ruser et al. 2001; Webb et al. 2010; Drury et al. 2012; Abdalla et al. 2013; Flessa et al. 2014).
The management systems (soil tillage) and fertilization techniques can impact the soil greenhouse gas emissions, with possible trade-offs between beneficial and adverse effects (van Kessel et al. 2013): The application of tillage, fertilizer, and crop** systems can influence the emission of N2O by increasing the mineral N content of soils, resulting from the application of either mineral or organic fertilizers. Microbial activity and organic C stocks are influenced by the amount and quality of incorporated crop residues. Reduced O2 availability can result from elevated bulk density, as for instance, in the case of no-tillage. While slurry injection has the potential to reduces NH3 losses and thus enhance nitrogen use efficiency (NUE), it is likely to increase the production of N2O through nitrification and denitrification. Conversely, the immediate incorporation of slurry prior to sowing does not result in a conflict between the emissions of N2O and NH3 (Flessa and Beese 2000; Velthof and Mosquera 2011; Flessa et al. 2014).
However, there is a limited number of field studies that have examined the combined effects of tillage, fertilizer application and crop type on yield, N2O and NH3 emissions as well as their respective factors (Sehy et al. 2003; Drury et al. 2006, 2012; Webb et al. 2010; Abdalla et al. 2013; Senbayram et al. 2014; Bayer et al. 2016). It is of the utmost importance to optimize slurry application techniques for the mitigation of NH3 and N2O emissions and the improvement of NUE. However, there is currently a paucity of experimental data for predicting how texture, moisture, pH, and organic C and mineral N dynamics interact with the spatial distribution of incorporated slurries of varying properties. While recent studies have addressed some of the aforementioned combination of activities and driving factors (ten Huf et al. 2023; Buchen-Tschiskale et al. 2023; Nyameasem et al. 2023), there is a paucity of studies that combine application techniques, several crops, NH3 and N2O fluxes and N2O processes at the field scale.
A two-year field experiment was conducted on two agricultural sites in Germany. The study focused on emissions of N2O and NH3 from these soils in three different management systems, which were represented by simulated management systems (cash crop farm – C-system, livestock farm with biogas plant – L-system, climate-optimized farm with biogas plant - O-system). The study included a variety of tillage techniques, crop rotations, fertilizers and fertilization techniques as well as two different crops (silage maize and winter wheat).
The objective of the O-system was to assess the viability of a crop** system that is more resilient to effects of climate change while simultaneously reducing greenhouse gas emissions. The objective of improved resilience to climate change effects from weather extremes, including drought and enhanced erosion, is achieved through the implementation of reduced tillage, direct seeding, and an extended crop** rotation. Moreover, it was anticipated that these features would result in reduced energy consumption and greenhouse gas emissions. This was due to the inclusion of legumes in the crop** rotation, the incorporation of crop residues, the optimized use of biogas digestates, and the reduction in the use of fossil fuels for tillage and crop protection products. Nevertheless, it is necessary to conduct further tests to ascertain the extent to which higher N2O fluxes resulting from increased nitrification and denitrification might offset the positive effects of the O-system. The overall objective of this study was to assess the extent to which O-system N2O and NH3 fluxes diverge from those of the C-system and L-system - by examining the combined effects of fertilizer type, organic fertilizer application technique, and crop type. The present study reports on the fluxes of N2O and NH3 and their regulation. The specific objectives were to (i) quantify N2O fluxes of management systems, (ii) evaluate the control factors of N2O fluxes, (iii) elucidate N2O production and reduction processes, (iv) quantify NH3 fluxes of treatments with biogas digestate application, and (v) identify the most effective management options to lower N2O and NH3 fluxes while improving or maintaining yields. The following hypotheses were proposed: (1) O- and L-systems are associated with higher N2O and NH3 fluxes than the C-system due to organic fertilizer application, (2) the L-system emits less NH3 than the O-system due to the incorporation of biogas digestate, (3) O-system treatments exhibit higher denitrification rates due to the incorporation of biogas digestate and minimum tillage with higher bulk density, (4) higher denitrification rates do not fully lead to higher N2O fluxes of the O-system. This is due to the fact that the O-system exhibits enhanced N2O reduction to N2 by heterotrophic bacterial denitrification which is a result of the higher bulk density and organic C content.
Materials and methods
Study sites
The field study was established on two sites located near Soest (North Rhine-Westphalia, Germany, 51°34’15.5”N, 8°00’06.8”E) and Braunschweig (Lower Saxony, Germany, 52°12’26.3”N, 10°34’57.0”E), in the following referred to as Soest and Braunschweig, respectively.
The soil types at both sites are representative of those typically found in their respective regions. The Soest site was established on a Stagnic Luvisol (silt loam with approximately 2% sand, 84% silt, and 14% clay) with a pH of 6.8 ± 0.1 (CaCl2), a soil organic matter content (SOC) between 1.1 and 1.48%, and a total nitrogen content (TN) of 0.14 to 0.18% in the 0–30 cm topsoil layer (Table S5). The mean air temperature at the Soest site over a 30-year period is 9.0 °C with a mean annual precipitation of 750 mm. The Braunschweig site was established on a Eutric Planosol a silty clay with a pH of 7.1 ± 0.1 (CaCl2) and a sand content of 0 to 10%, a silt content of 65 to 75%, and a clay content of 25 to 35%. The SOC was found to be between 1.19 and 1.85%, with a TN content of 0.16 to 0.20% in the 0–30 cm topsoil layer (Table S4). The mean air temperature at this site over a 30-year period was 9.7 °C, with a mean annual precipitation of 570 mm. Both sites were previously used for agricultural purposes.
In 2010, three different management systems were established on both sites as part of a simulated farm typology (Kramps-Alpmann 2017). These were continued until 2016. The experiments were initiated in June 2014, four years after the establishment of the management system. The crop rotation included all crops grown on both sites and was maintained throughout the study period.
The C-system was designed to mimic the management in an agricultural farm without livestock and arable land use of 200 ha. Fertilization was conducted solely using mineral fertilizer which is complemented by the incorporation of harvest residues. In the three-year crop rotation of winter oilseed rape, winter wheat, and winter wheat (Table S1), ploughing was conducted prior to the sowing of winter oilseed rape and prior to winter wheat following winter wheat. With the exception of the harvest residues, all commodities were removed from the farm after harvesting.
The modeled L-system included a 200-kilowatt biogas plant, 800 pig fattening places, and a land area of 200 ha. The biogas residues were used as a fertilizer, and missing nutrients were supplemented by the application of mineral fertilizer. The three-year crop rotation included winter wheat, winter barley (with hybrid rye as cover crop), and silage maize. Ploughing was also conducted prior to the sowing of winter barley. In cases, tillage was conducted as mulch-tillage. The harvest residues were retained on the site.
The modeled O-system was identical to the L-system except for the extended (8 years) crop rotation, which included seven crops (see Table S1), as well as exclusion of tillage and the implementation of direct seeding techniques. One aim of the extended rotation was a reduction in the use of mineral fertilizer through the cultivation of legumes (faba bean) and the concomitant legume N fixation.
The L- and O-system fertilizer applications were conducted using biogas digestate in accordance with the actual nutrient stocks in the model farms. Additionally, supplementary mineral fertilization was conducted to meet target values as described by the mineral set point method (LUFA, 2010). This method assumes that mineral fertilizer equivalents of manures (indicating the N-use efficiency relative to mineral fertilizer N-use efficiency) are 70% of the total N in biogas digestate (Table S2). At the Soest site, maize crop** in the O-system included a comparison between trailing hose fertilization and strip-till fertilization with below-root depot fertilization. At the Braunschweig site, strip-till was not a feasible option due to the high clay content of the soil.
The experimental plots measuring 6 × 12 m, were randomly distributed across a 48 × 100 m area, with four replicates of each treatment. The partial randomization was implemented instead of the full randomization, due to the necessity of utilizing conventional slurry application technology, which requires a plot length of > 20 m. Therefore, the control plots (without fertilization) of the trial plots had to be either in the center or at the edges of the 48 × 100 m area.
The soil chemical and physical data along with the basal respiration measurements, were obtained for all treatments in 2012 and in part also in 2016 (Kramps-Alpmann 2017) (Tables S5 to S9).
Determination of nitrous oxide fluxes and processes
N2O fluxes
In order to evaluate the impact of management systems on N2O fluxes, a comparison was conducted between full years with wheat and maize crop**. The observation period commenced at the beginning of June, 2014 and concluded in mid-June, 2016. As the study monitored two crops with different crop** periods, the observation period included data from three distinct crop** periods for each crop. The observation period for maize commenced several weeks after seeding and fertilization in 2014. The process was continued in the subsequent crop following the harvest until the seeding of the maize in the second year (2015), the full crop** season of maize in the second year, again the follow-up crop until the seeding of the maize in the third year (2016) and the beginning of the third maize crop** season until the end of May 2016. The chambers were relocated to new plots of the preceding maize crop** season in May 2015 and May 2016. Follow-up crops were planted in the L-system as hybrid rye and in the O-system as winter wheat (see Table S1). The observation period for winter wheat commenced during the ripening of the wheat harvested in 2014. The study continued throughout the post-harvest period until the seeding of the next winter wheat crop in the fall of 2014. Subsequently, the monitoring of the new wheat plots was continued until the seeding of the subsequent wheat crop in 2015, at which point the plots were again changed. Subsequent crops included winter wheat in the C-system, winter barley in the L-system, and hybrid rye in the O system (see Table S1). The final wheat crop was monitored until mid-June 2016. Hence, both crops were observed over the course of a full crop** period, as well as fragments of preceding and subsequent periods. Each observation year thus included the full or partial crop** season and the post-harvest period until the date of the subsequent seeding. The post-harvest period included tillage with incorporation of crop residues and establishment and growth of the subsequent crop. The inclusion of the post-harvest period was done to account for the impact of the first crop on factors regulating N2O fluxes such as residual mineral N, soil structure, residual soil moisture, and incorporated harvest residues. The post-harvest period thus included the after-effects of the first crop in interaction with the influence of the subsequent crop. Although this approach may mask the impact of the first crop to some extent, we chose this evaluation as optimal means to assess the effects of the management system.
The tested farm systems included a wide range of crop** rotation elements. However, within the scope of our study, it was only feasible to measure gaseous fluxes in a limited number of them, as shown in Table 1. The farm system approach allowed to investigate the combined effects of fertilizer type, organic fertilizer application technique, crop type, and site-specific factors. However, the number of treatments available for testing these factors differed. For example, we could include several comparisons for maize between O-system and L-system treatments. However, comparisons with the C-system treatments were limited to wheat because neither maize nor organic fertilization was included in the C-system. Moreover, the comparison of below-root digestate fertilization with surface application was only possible at one of the sites because the d-O-M treatment was established only at Soest.
The static closed chamber method (Hutchinson and Mosier 1981) was utilized to quantify N2O fluxes. In the maize plots, the chamber bases were inserted 5 to 10 cm deep into the soil at each tested site. The bases consisted of quadratic PVC frames measuring 60 × 60 × 15 cm (Ps-plastic, Eching, Germany). PVC chambers (60 × 60 × 70 cm, Ps-plastic, Eching, Germany) were equipped with two fans to homogenize the chamber volume after closure, and a vent tubing to allow pressure equilibration between the chamber and the atmosphere. A gastight connection was achieved by means of a rubber seal between the base and the chamber, and by the fixing of the chamber with rubber bands. In the winter wheat plots, circular PVC cylinders (30 cm diameter, 20 cm height) were used as chamber base. PVC chambers (30 cm in diameter, 30 cm in height) were equipped with vent tubing and a single fan. The sealing was achieved through the use of rubber sleeves.
In the case of maize, the frames were positioned between the plant rows, with a distance of 75 cm between rows. The width of the chamber was 60 cm, which ensured that the covered area was in close proximity (approximately 5 cm distance) to the maize plants, thereby enabling the majority of the rhizosphere soil to be covered. It was thus assumed that the exclusion of maize plants would not result in a significant bias in estimation of field fluxes, a hypothesis that had previously been demonstrated (Walter et al. 2015; Helfrich et al. 2020). In the case of winter wheat, the plants were included within the frame. When the height of the winter wheat plants exceeded the internal dimensions of the chamber in late spring, extension cylinders, with a length of 50 cm were used to maintain the plants within the chamber. Subsequently the flux calculation was adjusted to align with the revised chamber volume.
Sampling for gas analysis was conducted on a weekly basis on the same day and at the same time. The chambers were placed on the frames for a period of one hour. Immediately following the closure of the chamber and at 20, 40, and 60 min, a sample was collected in accordance with the method described by Buchen et al. (2017). This involves the flushing of septum-capped 20 mL-vials with air from the chamber. Additionally, a further sample was collected after 60 min in 100 mL vials for stable isotope analysis. The temperature of the air within the chamber was recorded at the time of each gas sample. In addition, the outside air temperature, wind speed (2 m), and precipitation were recorded at a nearby (~ 500 m linear distance) monitoring station. The concentrations of N2O, CO2 and CH4 were analyzed using a gas chromatograph (GC 2014, Shimadzu, Duisburg, Germany) equipped with an electron capture detector (ECD) for N2O and CO2 and a flame ionization detector (FID) for CH4, along with an automated sampling system (Loftfield et al. 1997). The coefficient of variation of the peak area was consistently below 3%.
Processes of N2O production and reduction
Selected gas samples from peak flux events were analyzed for isotopocule values of N2O using a Delta V isotope ratio mass spectrometer (Thermo Scientific, Bremen, Germany), coupled to an automatic preparation system with Precon + Trace GC Isolink (Thermo Scientific). N2O was pre-concentrated, separated and purified and m/z 44, 45, and 46 of the intact N2O+ ions as well as m/z 30 and 31 of NO+ fragment ions, were determined. The results were evaluated to determine the average δ15N, δ15Nα (δ15N of the central N position of the N2O molecule), and δ18O (Toyoda and Yoshida 1999; Röckmann et al. 2003; Westley et al. 2007). δ15Nβ (δ15N of the peripheral N position of the N2O molecule) was calculated as δ15N = (δ15Nα + δ15Nβ)/2 and the 15N site preference (δ15NSP) was calculated as δ15NSP = δ15Nα - δ15Nβ.
The pure N2O was analyzed for isotopocule values in the laboratory of the Tokyo Institute of Technology and was used as the internal reference gas, with the calibration procedures reported previously (Toyoda and Yoshida 1999; Westley et al. 2007) applied. Moreover, the standards utilized in a laboratory inter-comparison were used for performing two-point calibration for δ15NSP values (Mohn et al. 2014). All isotopic values are expressed as ‰ deviation from the 15N/14N and 18O/16O ratios of the reference materials (i.e. atmospheric N2 and Vienna Standard Mean Ocean Water (VSMOW), respectively). The analytical precision determined as standard deviation (1σ) of the internal standards for the measurements of δ15N, δ18O, and δ15NSP was typically 0.1, 0.1, and 0.5‰, respectively.
The isotopic signature of the soil emitted N2O was calculated as follows:
,
where δsoil_N2O represents the isotopic signature of the N2O soil flux, and δ and C represent the isotopic signatures and concentrations, respectively, of the gas mixture in the chamber (lower-case mix) and the background ambient air (lower-case bgd). For the ambient air, the mean values for the entire sampling period were utilized. The means and standard deviations of the ambient air values were considered in order to assess the uncertainty of the background values. The uncertainty assessment is described (in detail) below. At Soest, δ15Nbgd, δ18Obgd and δ15NSPbgd were 6.18 ± 0.14‰, 44.4 ± 0.15‰ and 18.55 ± 1.94‰, respectively. For Braunschweig, the respective values were 6.11 ± 0.17‰, 44.2 ± 0.24‰, and 17.74 ± 1.58‰.
For further evaluation of isotope data, we have selected the samples showing an increase in N2O concentration of at least 0.1 ppm compared to the background atmospheric N2O concentration. This represents an increase in chamber N2O concentration of over 30%, resulting in precision for isotopocule values of soil-emitted N2O that is better than 0.5‰, 0.5‰ and 2.3‰ for δ15N, δ18O, and δ15NSP, respectively. This selection ensures the minimal precision of the calculated N2O isotopic signatures of the soil flux, which is necessary for the further reliable application of isotope map** approaches and isotope modeling (see section Data Treatment below). The propagated uncertainty associated with the calculation of isotopic signatures of N2O soil flux was calculated for each sample, taking into account the analytical precision, variability of the isotopic values of N2O background values, and increase in N2O concentration in the chamber.
Determination of ammonia emissions
Ammonia emission was quantified in the maize treatments that received organic fertilizer, i.e. g-O-M and i-L-M treatments, in 2015 and 2016. Measurements were conducted using the “dynamic tube method” (DTM) (Pacholski 2016). Four cylindrical chambers, each covering a surface area of 104 cm2 were placed on the fertilized soil surface. A defined volume of ambient air was forced to pass through the chamber volume using a manual pump. The NH3 enriched air was led through Teflon tubing to an NH3−sensitive “Dräger gas analysis detector tube” which immediately allows a reading of NH3 concentrations (Drägerwerk AG, Lübeck, Germany). The coefficient of variation of NH3 concentrations was between 10 and 15%, as indicated by the manufacturer of the Dräger tubes. NH3 fluxes were calculated using the equations given by Pacholski (2016). In the calculations the chamber geometry, air temperature, barometric air pressure, and wind speed were all taken into account.
Ammonia measurements were initiated immediately following the application of organic fertilizer and were repeated at intervals of one hour for a period of four hours following fertilization for each replicate of the tested treatments (see Table 1). The following measurements were conducted at four-hour intervals until sunset, after which they commenced again immediately at sunrise the following day. This cycle was repeated two hours after sunrise and then again in four-hour intervals. The air temperature, barometric air pressure, and wind speeds were provided by a nearby weather station. Any data gaps were filled by data from the closest weather station of the German Weather Service.
Soil sampling and analysis
Soil samples for mineral N (Nmin) and water content analysis were collected from the topsoil layer (0–30 cm) on a weekly basis for all plots using a Goettinger boring rod with a diameter of 12 mm (Nietfeld GmbH, Quakenbrück, Germany). The samples were cooled to the point of transport from the field and subsequently frozen at a temperature of -20 °C within four hours. Prior to the extraction, samples were defrosted at 5.6 °C for a period overnight to prevent significant changes in Nmin content (König and Kießling 2001). For the extraction process, 50 g of defrosted soil were weighed and placed in 500 ml polyethylene bottles. The bottles were then filled with 200 ml of 0.01 M CaCl2 solution. Subsequently, the bottles were left on an overhead shaker for one hour and then filtered using a MN614 ¼ filter (Macherey and Nagel, Düren, Germany). The filtrate was analyzed for NO3−and NH4+ with a photometric continuous flow analyzer (CFA-Analyzer San++, Skalar Analytical B.V., Netherlands). A subsample of 30 g of soil was dried at 105 °C until mass equilibrium was achieved. This process was conducted over a minimum of 24 h to determine the gravimetric soil water content (SWC).
Data treatment
The volumetric water content and water-filled pore space (WFPS) were calculated from the soil water content (SWC) and bulk density. While bulk density is subject to continuous change during the year due to loosening by tillage and swelling and subsequent shrinking and compaction, continuous monitoring of bulk density in all treatments was beyond the scope of resources. Therefore, WFPS was calculated on the assumption of average bulk densities determined in 2016 (Tables S8 and S9), using values for the 10–15 cm depth layer, which represents the top soil layer in which water content was measured. Consequently, WFPS values are affected by the uncertainty that arises from the assumption of constant bulk density values. It should be noted that, as a result of these considerations, the calculated WFPS values during phases of highest soil moisture are sometimes above 100% (Fig. 1, Figs. S1.1–1.5). Nevertheless, we decided to show WFPS values and use them for interpretation of N2O fluxes, because WFPS is assumed more suitable as explanatory variable than gravimetric and volumetric soil water contents.
N2O fluxes were calculated using linear regression, robust linear regression, and Hutchinson-Mosier regression for non-linear fluxes (Pedersen et al. 2010) using the R package gasfluxes. The choice of regression method was based on an algorithm as described by Leiber-Sauheitl et al. (2014). This was conducted with an update of the HMR package to version 0.4.1 (Pedersen 2017) which was based on the calculation of standard errors. This resulted in linear calculation of the majority of N2O fluxes (> 98%). To identify samples that had failed, the distribution of the square roots of the standard errors of the N2O and CO2 fluxes was used.
The flux rates are expressed as means (n = 4) with standard deviation of the replicated field plots. Cumulative annual N2O emissions were calculated from N2O fluxes per plot using linear interpolation between two measurement dates.
To determine the pathways of N2O production pathways and the progress of N2O reduction we used the three-dimensional N2O isotopocule model (3DIM), which is based on three isotopic signatures of N2O (δ15N, δ15NSP, and δ18O) (Lewicka-Szczebak et al. 2017, 2020; Yu et al. 2020). The model enables the determination of the following quantities:
fbD+nD, ffD, fNi = estimated fractions of gross N2O production originating from bacterial denitrification (including nitrifier denitrification), fungal denitrification, and nitrification, respectively.
The 3DIM applies a stable isotope mixing model in the Bayesian framework which allows for integrating three N2O isotopic signatures into a single model in order to identify the most probable solution for the rN2O and the mixing proportions of different N2O production pathways (Lewicka-Szczebak et al. 2020). The detailed isotopic characteristics applied are presented in the Table S11 and follow the most recent review paper (Yu et al. 2020) taking into account the analyzed isotopic signatures of the substrates for this case study. The soil water and mineral N isotopic signatures were determined in samples collected during three one-week field campaigns (November 2015, March 2016, and May/June 2016) with the objective of conducting intensive stable isotope analyses (Lewicka-Szczebak et al. 2020). The water isotopic signatures (δ18OH2O) showed relatively stable values of 6.4 ± 1‰. The δ15N of NO3− ranged from − 9 to 7‰ with a mean of 0.85 ± 1.8‰. No significant differences were observed between the field sites or treatments. The δ15N of NH4+ showed a significantly elevated value of 41.4 ± 8‰. This value was the result of a rapid consumption of NH4+ following its addition to the soil and was accompanied by very low NH4+ contents. Additionally, in this field study the addition of NH4+ was conducted in all treatments (either in mineral or organic form), and its dynamic transformations resulted in rapid decrease in NH4+ content peaks (Fig. 1). Consequently, it can be assumed that this study also involves significant isotopic fractionation of the NH4+ pool. This assumption can be further reinforced by an analysis of the sample points in the δ15NSP-δ15N space (Fig. S3 C, D). The high number of sampling points on the right side or below the reduction line of bacterial denitrification (bD) can be explained only by strongly 15N-enriched endmember values of nitrifier-denitrification (nD) and nitrification (Ni). Two possible cases of N2O mixing and reduction can be considered in the modeling process:
Case 1
N2O produced from bacterial denitrification is first partially reduced to N2, followed by the mixing of the residual N2O with N2O from other pathways,
Case 2
N2O produced by various pathways is first mixed and then reduced.
In this section only the results of calculations based on Case 2 will be presented. This is due to the enhanced performance and reliability of 3DIM when Case 2 is applied in cases where the bD and nD fractions cannot be distinguished with sufficient precision (Lewicka-Szczebak et al. 2020). However, due to the strong dominance of bacterial denitrification, the differences between results for both cases are very low.
Statistical analysis
The statistical analyses were conducted using R 3.3.2 (R Core Team 2020). For each test, an analysis of variance (ANOVA) was conducted to analyze the effects of tillage (conventional tillage, no-tillage and strip-tillage), fertilizer application type (organic and/or mineral fertilizer, and surface application, closed slot injection and surface application followed by incorporation), crop type (maize and winter wheat), and location (Soest and Braunschweig) on cumulative N2O fluxes. A post hoc test was conducted using the Tukey HSD test to perform pair-wise comparisons of all tests where differences were indicated by ANOVA. The variance homogeneity and approximate normality of residuals were evaluated through diagnostic plots as described by Zuur et al. (2010). For all tests, the significance level was set to p ≤ 0.05.
The propagated uncertainty was calculated using the Gaussian error propagation equation, with consideration given to the standard deviations of all individual parameters. In 3DIM, the Markov-chain Monte Carlo method is utilized with the Metropolis condition: Li+1/Li ≥ α, where α is a random variable sampled from a uniform distribution.
To investigate the impact of measured environmental parameters on N2O fluxes, a generalized additive model (GAM) was applied on log-transformed N2O fluxes using the R package mgcv version 1.8–31 (Wood 2006, 2011). The model is defined by the relation of N2O fluxes to a linear combination of predictor variables, that are estimated from parametric or smoother functions of explanatory parameters. The degree of smoothing is estimated by the penalized maximum likelihood approach. An offset of 18 µg N2O-N m− 2 h− 1 was used to ensure that negative fluxes remained within the dataset, resulting in a total dataset of n = 3091. The variance homogeneity and approximate normality of the residuals were checked by diagnostic plots according to Zuur et al. (2010). The parameters used were NO3− content and WFPS by site and treatment, as these have been demonstrated to be primary regulators of N2O fluxes. CO2 fluxes were used as an indicator of microbial and plant respiratory activity (Deppe et al. 2016), due to the fact that the related O2 consumption impacts N2O production from nitrification and/or denitrification processes. The inclusion of the soil NH4+-N content and soil temperature did not result in an improvement of the model. To ascertain whether the impact of environmental parameters differed according to soil types on the two sites and management system (C-, L- or O-system), the site and system were included in the analysis. In addition, a random effect smoother (bs="re”) for plot was included (Pedersen et al. 2019).
Results
Soil temperature, precipitation and soil moisture
The mean annual temperature at Soest was 10.83oC and 11.24oC, respectively, in the first (June 1, 2014 to June 1, 2015) and second experimental year (June 1, 2015 to June 1, 2016). At the Braunschweig site, the corresponding values were 10.35oC and 10.55oC respectively. The mean temperatures were thus higher than the 30-year average (10.38oC at Soest and 9.92oC at Braunschweig, respectively). At Soest, the precipitation was slightly lower in the first experimental year (sum: 736 mm, average: 2.02 mm d− 1) than in the second experimental year (764 mm, average: 2.09 mm d− 1). At Braunschweig, precipitation in the first and second experimental years was 491 mm (average: 1.34 mm d− 1) and 647 mm (average: 1.77 mm d− 1), respectively. This is in agreement with long-term average annual precipitation of Soest and Braunschweig (2.11 and 1.63 mm d− 1), indicating that the Braunschweig climate was drier than the Soest climate in both years. In comparison to the long-term averages, the precipitation levels at the Soest site were lower in both experimental years. In contrast at the Braunschweig site this was only the case during the first experimental year.
The time course of water-filled pore space (WFPS, Fig. 1 and Figs. S1.1 to S1.6) exhibits typical (excluding peak events) high values during the hydrological winter season (here defined as beginning November 1 and ending February 28) with high values up to approx. 100%. While periods of extended high WFPS values were similar in both winter seasons at the Soest site, the duration of wet periods was longer in the second season at the Braunschweig site. In contrast to the Braunschweig site where WFPS values were consistently close to 100% during extended periods in winter, this occurred only during brief periods at the Soest site. During the growing season (defined as beginning in March and ending in October), typical drying of soils was interrupted by high precipitation events, resulting in a short-term increase in WFPS (e.g. in September 2015). In all other cases, WFPS values followed the dynamics of vegetation and precipitation (Fig. 1 and Figs. S1.1 to S1.6).
Time course of regularly monitored data of treatment g-O-M (Trailing hose in growing crop + mineral amendment, O-system, maize) of Braunschweig (left panel) and Soest (right panel) sites. (A) precipitation and soil temperature, (B) soil moisture (water–filled pore space, in %)., (C) N2O flux, (D) process ratios (yellow: fBD+nD= fraction of heterotrophic bacterial denitrification + nitrifier denitrification, purple: fNI, fraction of nitrification, red: fFD, = fraction of fungal denitrification, green: rN2O = fraction of residual N2O), (E) soil respiration index (dark chamber CO2 flux), (F) blue: NO3- -N; red: NH4+ -N content. Dates of management activities are marked with colored bars in panel B (sowing = black, harvest =red, organic fertilization with biogas digestate = green, mineral fertilization = blue)
Mineral N
Mineral N (Nmin = NO3− + NH4+) reflected fertilization (Fig. 1F, Fig. S1.1 to S1.5) where NO3− and NH4+ peaks after fertilization were always evident. In treatments with mineral fertilizer application only (Fig. S1.1, S1.5), total Nmin reached up to approx. 200 kg N ha− 1. In treatments including organic and mineral fertilizer application in maize, Nmin peaked up to > 300 kg N ha− 1 in maize (treatments i-L-M at Braunschweig site, Fig. S.1.4 and d-O-M at Soest site, Fig. S1.5). Smallest peaks of up to 70 kg N ha− 1 occurred in wheat with surface application of digestate in the Soest site (g-O-W, Fig. S1.3). NH4+ peak values were always about one order of magnitude lower than NO3− values. While NO3− fertilizer peaks typically lasted several weeks, NH4+ peaks were more short-lived. Additional mineral N peaks occurred after harvest, presumably due to mineralization of crop residues, and were strongest in maize at Braunschweig with up to > 100 kg N ha− 1 in 2015 (Fig. S.1.2 treatment g-O-M and Fig. S.1.4, treatment i-L-M). Total mineral N typically declined after fertilization during crop growth due to plant uptake.
N2O fluxes
Gas sampling for measuring N2O emissions was done, with very few exceptions, weekly from June 2014 until June 2016. The time course of N2O fluxes in Soest and Braunschweig and the timing of management events is shown in Fig. 1C and Figs. S1.1 to S1.5. At both sites, peak N2O fluxes occurred following some (but not all) fertilizer events, during periods when high moisture coincided with high NO3−, and following tillage or harvest. Maximum fertilizer-induced peaks were highest (> 1000 µg N m− 2 h− 1) in O-system treatments with maize and digestate application in growing crop or mineral fertilization only (Fig. S1.2, treatment g-O-M and Fig. S1.5, treatment s-O-M, respectively). Maximum fertilizer–induced peaks were lowest (i.e. always below 500 µg N m− 2 h− 1) in all winter wheat treatments (Fig. S1.1, treatment s-C-W and S1.3, treatment g-O-W). Post-harvest peaks in fall occurred in all treatment and both years, but were generally lower than fertilizer induced peaks.
Cumulated N2O emissions and yields of crops and management systems were compared for both years. To evaluate differences between treatments, we first compared all treatments with organic fertilization (upper block of Table 2). In the first year, maize in the L-system (treatment i-L-M) at the Soest site showed lower areal emissions than maize in the O-system (g-O-M) of both sites. There were higher areal emissions of N2O in maize in the O-system (g-O-M) at both sites compared to treatments with organic fertilizer of the L-system (i-L-M). Biogas digestate application in the growing crop (O-system) thus resulted always in higher areal fluxes compared to digestate incorporation before sowing (L-system). Comparison between sites showed that g-O-M had higher areal fluxes in Braunschweig in both years, but the other treatments were never different between sites. Comparison between the two crop types showed that maize (g-O-M treatment) exhibited up to 4 times higher fluxes than wheat (g-O-W) in Braunschweig in both years. There was the same trend in Soest, but differences were not significant. Comparison between the years showed no different fluxes for any of the treatments and sites.
Yield-scaled fluxes showed similar differences between treatments, but less pronounced due to the lower yield in wheat. In contrast to areal fluxes, yield scaled flux of g-O-W was not different from g-O-M. This was due to lower maize yields in the second year, which was relatively dry during maize growth. Otherwise, cases of significant differences matched areal fluxes.
Moreover, we compared all O-system treatments with maize in order to test differences between below-root strip-tillage application (d-O-M), mineral fertilizer application (s-O-M) or trailing hose organic fertilizer application (g-O-M) (see d-O-M and s-O-M data in the middle block of Table 2 and g-O-M data in the upper block). This was done only for the Soest site since d-O-M was not established at Braunschweig. There were no significant differences between treatments and trends were different for both years with highest areal and yield-scaled fluxes in s-O-M in the first year and in d-O-M in the second year.
The last comparison was conducted to test differences in winter wheat between mineral fertilization (s-C-W) and trailing hose application of biogas residues in growing crop (g-O-W) on C- and O-systems of both sites (see s-C-W data in lower block of Table 2, g-O-W data in upper block). Areal fluxes of s-C-W and s-O-W treatments were not different at either of the sites, while areal fluxes of the s-C-W treatment were lower at the Soest site (about one third) compared to the Braunschweig site. There were no significant differences between yield-scaled emissions of those treatments.
Yields
Maize yields were between 10.1 and 20.6 Mg ha− 1 a− 1 in Braunschweig and between 14.6 and 20.8 Mg ha− 1 a− 1 in Soest (Table 2). All yields of the second year were lower compared to the first year, where differences were more pronounced in Braunschweig. Winter wheat yields did not exhibit significant differences between years and ranged between 9.1 and 11.1 Mg ha− 1 a− 1. Winter wheat yields were about 10% higher in Soest, but this difference was only significant in the first year.
Environmental controls of N2O fluxes
Generalized additive models (GAM) had been applied to investigate the variance of N2O fluxes in relation to explanatory variables such as WFPS or NO3− content in the topsoil layer. The used GAM fitted WFPS and NO3− content to a combination of the interactions between the cultivated crops, the management systems, the fertilization treatments and the investigated sites (Soest and Braunschweig). The GAM explained 41% of the variance of the log-transformed N2O fluxes (see Supplementary data: Table S10 and Fig. S2). The effect of site and management system was highly significant. Additional parameters, such as CO2 flux (as a proxy for microbial and plant respiratory activity), WFPS, NO3− and plot selected per site, were significant and improved the goodness of the GAM fit. The impact of CO2 was similar for both sites. The interaction between WFPS and NO3− content in soil, separated per treatment, significantly affected the N2O fluxes at each site. For all treatments, the highest N2O fluxes correlated with high WFPS values (> 80%). The impact of NO3− content was different in the treatments, while in the mineral fertilization treatment (s), high N2O fluxes coincided high WFPS values and NO3− contents of 200 to 300 kg N ha− 1 in Soest and > 300 kg N ha− 1 in Braunschweig. This interaction was found to be lower for the trailing hose application treatment, but is more similar between the two sites. The N2O fluxes of the strip-till treatment (d-O-M), which was only applied in Soest, again coincided with high WFPS values and high NO3− contents in soil (> 300 kg N ha− 1).
Isotopic signatures of N2O and processes of N2O production and reduction
The contribution of N2O production pathways and progress of N2O reduction was determined based on the 3DIM approach and using the measured isotopic signatures of N2O (see Materials and Methods, Data treatment). Due to the lack of clear isotopic distinction in δ15NSP and δ18O values between the pathways bD and nD, we cannot separate them and can only present their joint contribution. The 3DIM model can theoretically divide these processes. However, the distinction is inaccurate, as it is mostly based on the differences in δ15N values only, and could not be validated so far (Lewicka-Szczebak et al. 2020). The results provided by the model iterations can be evaluated for the correlations between the modeled pathway fractions, where high correlation coefficients indicate weak distinction between particular fractions (Lewicka-Szczebak et al. 2020). This is definitely the case for bD and nD fractions, where the correlation coefficient reaches 0.80 and 0.71 for Soest and Braunschweig, respectively. Much better distinction was achieved between the pathways fD and Ni, since their isotopic signatures differ in two values: δ18O and δ15N (see Fig. S3). The correlations between modelled ffD and fNi are much lower with 0.46 and 0.40 for Soest and Braunschweig, respectively. Therefore, the 3DIM results of the N2O pathway fractions are presented as fbD+nD, ffD and fNi. (Table 3).
At the Soest site, these values were statistically not different among treatments and exhibited quite similar means. rN2O values were between 0.19 and 0.37, indicating that most of the produced N2O was reduced to N2. fbD+nD values were between 0.76 and 0.95, indicating that most of the N2O production originated from bacterial denitrification and only a small fraction from nitrification or fungal denitrification.
In contrast, Braunschweig results were more variable and showed highest rN2O for the g-O-M and g-O-W treatments and lowest of the s-C-W treatment. The i-L-M treatment exhibited the lowest fbd+nD (0.5) indicating the largest contribution from processes other than denitrification. For Soest, ffD and fNi were similar in magnitude, ranging from 0.06 to 0.15 and from 0.07 to 0.16, respectively, and exhibited no significant differences between treatments. At the Braunschweig site, ffD of the g-O-W and s-C-W treatments were lower compared to g-O-M and i-L-M. The highest ffD with values of > 0.4 were noted when N2O fluxes and WFPS values were high (on July 11 2014 and August 20 2015 of the g-O-M treatment, Fig. S-1.2; on July 3.2014 and July 11 2014 of the i-L-M treatment, Fig. S.1.4). Interestingly, the ffD values exceeding 0.4 all occurred at the Braunschweig site and were always associated with high rN2O (> 0.6, Fig. S4B). fNi of the i-L-M treatment was higher compared to the other treatments, reaching values > 0.3 (Fig. S4H). These high values occurred when N2O fluxes and WFPS values were low (Fig. S.1.4).
NH3 emissions
Ammonia (NH3) emissions were measured at both sites in maize following digestate application (Table 4). The levels of NH3 emissions were similar for both sites within the two seasons. The highest NH3 emissions occurred after application of biogas digestate with trailing hose (g-O-M) with up to 6.5 kg NH3-N ha− 1 at Braunschweig (season 2). The incorporation of biogas digestate (i-L-M) reduced NH3 emissions significantly, except at the Soest site in the first season. The application by strip till (d-O-M) resulted in almost no NH3 emissions in both seasons in Soest. At the Braunschweig site, this treatment was not feasible due to the relatively high amount of clay in the soil. The incorporation by conventional tillage or by strip till can only be carried out before or during sowing, respectively. Therefore, fertilization with those techniques took place earlier in the year compared to the application with trailing hose into the growing crops (Fig. 1, Fig. S1.1-S1.5). Therefore, the higher NH3 emissions of the latter treatments were probably not only due to the omitting of incorporation, but probably also because of the higher soil and air temperatures at the date of fertilization.
Discussion
Nitrous oxide emissions
Emission events that are influenced by climate, soil properties and management practices
In both years and on both sites, the results demonstrated that high emissions of N2O were primarily observed during periods of events such as after heavy rainfall, soil tillage, sowing, or fertilization. In accordance with previous studies, N2O emissions were found to be contingent upon the temporal dynamics of soil moisture (Senbayram et al. 2014), Nmin content (Helfrich et al. 2020) and tillage (Petersen et al. 2008, 2011; Krauss et al. 2017). High N2O fluxes frequently occurred in maize during the late spring and early summer months when soil temperatures increased and soil moisture and mineral N contents were high due to the late uptake of N and water by maize (Senbayram et al. 2014). Despite the higher precipitation observed in the first year, the annual N2O fluxes of the two seasons were not found to be statistically different. However, the differences in precipitation and soil moisture were particularly pronounced during the spring and early summer, with clearly drier conditions in 2015. This resulted in alterations to N2O fluxes as evidenced by the cumulated fluxes of June/July which were higher in 2014 across all maize-based treatments.
Soil processes that cause N2O emissions
It is well established that high moisture content, soil temperatures, and the content of easily degradable soil carbon accelerate denitrification, a major source of N2O (Butterbach-Bahl et al. 2013; Müller and Clough 2014). The dominance of denitrification is clearly supported by the stable isotope results. These results indicate a strong dominance of bacterial denitrification and/or nitrifier denitrification in all treatments (mean fbD+nD values from 0.76 to 0.90, Table 3), with the exception of the i-L-M treatment of Braunschweig (mean fbD+nD value = 0.50), where nitrification showed significantly higher contribution (mean fNi = 0.26) than in other treatments. This was mostly due to exceptionally elevated fNi values on May 18 and 24, 2015, with fNi = 0.68 ± 0.07. The exceptionally dry spring of 2015 in Braunschweig may be a contributing factor. Following the application of CAN as a mineral fertilizer, NH4+was found to be in plentiful supply as a substrate of nitrification, following mineral fertilization with CAN. Subsequently, nitrification was likely intensified by relatively low soil moisture (67% WFPS), whereas conditions were less favorable for N2O production from heterotrophic denitrification. Under these conditions, similar fluxes of N2O from nitrification and bacterial denitrification are typical (Zhu et al. 2013) and the relatively low observed N2O fluxes (4 to 12 g N ha− 1 d− 1) are in agreement with the low N2O yield of nitrification (Hink et al. 2017).
While the soil pH and organic matter content were comparable between the two sites, the Braunschweig site exhibited a higher clay content (25–35%) than the Soest site (14%). The water retention characteristics, pore size distribution, and pore structure of a soil are largely influenced by its texture (Weller et al. 2022). As the clay content increases, the rate of aeration is typically slowed, which in turn increases the rate of denitrification (Jungkunst et al. 2006). However, reduced gas exchange also enhances N2O reduction to N2 (Harter et al. 2016), thus negating any clear net effect of texture on N2O fluxes. This is reflected by our data, which showed that the Braunschweig site with higher clay content exhibited higher areal fluxes in both years only in g-O-M treatments. However, the other treatments were never significantly different between sites. Both sites were limed to maintain a target soil pH close to neutral as indicated by their loamy texture. Therefore, the inhibition of N2O reduction due to low pH (Žurovec et al. 2021) did not apply to either of the sites. Nevertheless, the mean rN2O values of both sites were found to range between 0.22 and 0.55 which is higher than the global mean rN2O for agricultural soils of 0.11 reported in the review by Scheer et al. (2020). It is important to note that isotopic values were only available for high N2O flux events when soil moisture and NO3− content was high. This is due to the limited sensitivity of this approach when N2O fluxes are low (Buchen et al. 2018). Consequently, the isotopic evidence was confined to the typical denitrification events. It is possible that N2O reduction was inhibited to some extent, as it is known that this effect can result from high NO3− levels and has been shown to result in elevated rN2O at high pH (Senbayram et al. 2019, 2022). Isotopic data also indicated that denitrification was the dominant process in the N2O fluxes in almost all of the treatments (Table 3). This is in line with the observed increase in denitrification during periods of elevated N2O flux, which is accompanied by elevated moisture and NO3− contents. However, the isotopic values were only available for a limited number of dates. Nevertheless the observation that peak flux periods account for the majority of the cumulated yearly fluxes indicates that isotopic results represent the majority of the emitted N2O. The differences in fbd+nD between sites were only evident for the i-L-M treatment which exhibited the lowest fbd+nD (0.5) for the Braunschweig site. Approximately half of the N2O flux originated from fungal denitrification and nitrification, respectively (ffD = 0.24 and fNi = 0.26). The high ffD observed for the g-O-M treatment of Braunschweig also indicates that elevated ffD occurred in both maize treatments at this site. The highest ffD values (0.40 to 0.49) were observed in i-L-M and g-O-M during the wet spring 2014. This period was characterized by high moisture levels (86% WFPS) and high N2O fluxes (90 to 1300 g N ha− 1 d− 1) between June 6 and July 7, coinciding with unusually high rN2O (0.67 to 0.73). Because high N2O fluxes and high rN2O are in consistent with the lack of N2O reductase in fungal denitrifiers (Rohe et al. 2017), the observed maximum ffD values at these dates are plausible. There is only limited knowledge regarding the control of fungal N2O due to the limitations of current methods (Rohe et al. 2021). However, there is evidence that the interaction of labile carbon sources, high soil moisture and, high NO3− levels can enhance fungal denitrification and related N2O flux (Senbayram et al. 2018). These observations indicate that fungal denitrification may contribute to peak fluxes of N2O. However, further studies, including analysis of microbial community composition and functional genes (Rohe et al. 2020) are necessary to clarify this.
Management system and treatment effects
The simulated management systems (cash crop farm, livestock farm with biogas plant, climate-optimized farm with biogas plant) of our study differed with respect to tillage, fertilizer type, crop rotations and, fertilization techniques. Each of these aspects affects the proximal and distal controls of N2O processes in a specific manner. This includes water content, bulk density, mineral N, labile organic C, pH, and oxygen in a specific way (Groffman et al. 1988; Butterbach-Bahl et al. 2013; Müller and Clough 2014). These controls are affected by five features of the experimental setup: (1) crop type, (2) tillage, (3) fertilizer type, (4) fertilizer application technique, and (5) soil properties induced by the management system). These features are discussed in the following section. While these features impact N2O dynamics to some extent individually, it is necessary to consider the interaction between them and with site characteristics in order to understand the results and draw conclusions with respect to mitigating N2O and NH3 fluxes while maintaining yield.
Crop rotation
Diverse crop rotations, including the frequent use of legumes, intercrop** and organic fertilizers typical of organic farming are known to enhance organic matter content, soil structure, and biological activity in soils. While these factors are conducive to crop productivity, they also imply the risk of mineral N accumulation after harvest and high N2O fluxes (Hansen et al. 2019). This is in line with the observation that, in the majority of cases, clear post-harvest peaks of NO3− and/or N2O fluxes were evident in O-system treatments at both sites (Figures S 1.2, S.1.3 and S 1.5), while similar peaks were either absent or less pronounced in the L- and C-systems.
Several studies have demonstrated that the cultivation of maize can result in significantly increased N2O emissions in comparison to the cultivation of winter wheat (Drury et al. 2006, 2012; Senbayram et al. 2014). This can be due to the fact that the uptake of N by plants and the uptake of water by maize are still limited during the early summer period, whereas winter wheat is at its most productive growth stage and the uptake of nutrient and water is at its highest during this stage (Senbayram et al. 2014). Higher fluxes under maize are supported by our data. Crop type effects could be evaluated from treatments g-O-M compared to g-O-W that differed in crop type only. Significant differences were observed with fluxes of g-O-M approximately four times higher than those of g-O-W. This was evident at the Braunschweig site during the first year, which was characterized by higher precipitations. Moreover, trends with at least 50% higher N2O fluxes were observed in the second year, and also at Soest in both years. While comparing these treatments (g-O-M to g-O-W) with isotopic data we can note significant difference in the share of N2O originating from fungal denitrification at the Braunschweig site (Table 3, ffD). The treatment g-O-M shows higher ffD than g-O-W. This may indicate that greater contribution of fungal denitrification may be partially responsible for the higher N2O fluxes observed for the g-O-M treatment.
Tillage
The impact of strip-till and no-till on N2O fluxes and in general greenhouse gas emissions is still controversially discussed in present literature. Some studies have indicated reduced emissions (Jacinthe and Dick 1997; Drury et al. 2006, 2012; Bayer et al. 2016) whereas some other studies have found indications for increased emissions using these cultivation systems (Abdalla et al. 2013; Bayer et al. 2016), or even no differences (Johnson et al. 2010). In both years and both sites, the O-system areal N2O emissions were more than twice as high than those of the L-systems, with the difference being statistically significant except for the second year in Soest. For yield-scaled fluxes, the O-system fluxes were always higher than those of the L-system. The results of the O-system, which was managed with zero-tillage and direct seeding, could be explained by higher emissions due to the fact that tillage was restricted beyond mulch tillage. These results are consistent with those of other studies that have observed an increase in emissions under no-till, strip-till or in general conservation tillage. The reported reasons include higher bulk densities, organic matter contents, and soil water contents (Abdalla et al. 2013; Huang et al. 2018). The first factor is supported by our data, which indicates that the top soil bulk density was highest in the O-system with zero tillage and lowest in the C system conventional tillage (Tables S7 and S8). Indication for elevated availability of organic C was given for the 0–10 cm layer of the O-system treatment in Braunschweig, as basal respiration was approximately 50% higher compared to the L- and C-system treatments (2016 data of Table S6). This may provide an explanation for the more pronounced enhancement of N2O fluxes in the O-system (g-O-M vs. i-L-M, Table 2) observed at the Braunschweig site.
Fertilizer type
The application of liquid organic fertilizers on agricultural fields affects nitrification and denitrification dynamics by the simultaneous transport of mineralizable organic N and mineral N, as well as labile organic carbon (C) which causes oxygen (O2) consumption during its respiratory decomposition (Grosz et al. 2022). In the boundary zone between organic fertilizer and bulk soil, this potentially leads to the local co-occurrence of anoxia and high NO3− concentration from nitrification of fertilizer-derived NH4+. The water retention of organic fertilizer results in an increased water content in manure or digestate clumps (Petersen et al. 2003). These factors can result in high denitrification rates (Baral et al. 2016), potentially accompanied by highly variable N2O/(N2 + N2O) ratio of denitrification. Moreover, the reduction of N2O may be inhibited by low pH, which can result from acidification by previous nitrification of fertilizer-derived NH4+ and a locally high NO3− concentration. However, the impact of C inputs on N2O fluxes may be masked in soils with a high content of soil organic matter (Möller 2015). Moreover, field-applied digestates have been demonstrated to result lower N2O fluxes compared to untreated slurries, in part due to their more recalcitrant organic carbon content (Hou et al. 2015).
The present study did not identify any differences in the amount of emitted N2O based on type of fertilizer utilized (mineral or organic). For both crops, the fluxes observed in the mineral fertilized treatments were not (statistically) different from those observed in the organic fertilization treatments (Table 2). It is possible that the supply of labile organic carbon from the applied digestate was insufficient to enhance N2O production, or that this effect was diminished by limitation of nitrate supply for denitrifiers. This is because nitrification of digestate-derived NH4+ following digestate application is typically delayed (Köster et al. 2015). However, while the time course of NO3− content differed between organic and mineral fertilizer treatments (Fig. S1), the mean NO3− content was not (statistically) different. It is also possible that the organic carbon present in the biogas residues may have disproportionately contributed to the reduction of N2O to N2 during denitrification (Müller and Clough 2014). Laboratory incubations have demonstrated that liquid organic fertilization results in high N2 fluxes with rN2O levels that are significantly lower than those observed in mineral fertilizer treatments (Köster et al. 2015). However, this assumption must be rejected based on isotopic data, which actually show that the lowest rN2O values are observed for the mineral fertilization (Table 3). The results indicate that organic fertilization did not result in an increase in either N2O flux or N2 flux.
Biogas digestate application technique
Slurry injection methods improve the efficiency of the N uptake by crops, reduce N leaching and N losses by NH3 volatilization (Webb et al. 2010; Emmerling et al. 2020), yet may potentially enhance N2 and N2O fluxes via nitrification and denitrification (Buchen-Tschiskale et al. 2023). While slurry injection has been demonstrated to effectively inhibit NH3 emission, N2O fluxes are found to be enhanced, but with a variable extent dependent on the interaction with the types and application rates of manure, soil texture, vegetation, and climate (Hou et al. 2015).
A comparison was conducted between the O-system treatments and maize (Table 2) in order to test for differences between below-root digestate application by strip-till-depot fertilization (d-O-M), mineral fertilization (s-O-M) and trailing hose digestate application (g-O-M). This comparison was only conducted for the Soest site, as the d-O-M treatment had not yet been established at Braunschweig. There were no significant differences between the treatments. However, the trends differed between the two years, with the highest areal and yield-scaled N2O fluxes observed in the s-O-M treatment in the first year and in the d-O-M treatment in the second year.
The highest N2O fluxes were observed in O-system treatments with maize (g-O-M), which were associated with trailing hose surface application without incorporation. It is not to be expected that surface application will result in higher N2O fluxes than incorporation, since aeration close to the surface is not in favor of denitrification (Webb et al. 2010). Since the higher N2O fluxes observed in the g-O-M treatments compared to the i-L-M treatments were assumed to be due to zero-tillage (see above), it is not possible to rule out the possibility that these differences would have been even more pronounced, if zero tillage had been compared with surface application and conventional tillage with identical application techniques. However, this comparison was not part of the study.
Ammonia emissions
The data show that high NH3 emissions occurred only in treatments with trailing hose application in maize without incorporation (g-O-M treatments). This is in agreement with the known enhancement of NH3 volatilization when digestate is in contact with the atmosphere for an extended period (Sommer et al. 2003). The fluxes were further enhanced by the high air temperature during the period of application in June and July. In general volatilization is decreased as the depth of slurry application increases (Sommer and Hutchings 2001). In comparison to other studies (e.g., Möller and Stinner 2009; Herr et al. 2019), the general level of NH3 emission was relatively low at both sites. The incorporation of digestate following trailing hose application within four hours of application resulted in significant reduction in the risk of NH3 emissions, with the i-L-M treatment with immediate cultivator incorporation to 12 cm depth demonstrating a reduction of up to 90% when compared to surface application of the g-O-M treatment. This is in line with the findings of a recent meta-analysis by Emmerling et al. (2020), which demonstrated a 63% reduction in NH3 emissions through manure incorporation in comparison to surface application. The application of strip-till below root fertilization (d-O-M treatment) effectively inhibited NH3 fluxes, as evidenced by the absence of detectable emissions following fertilization. This finding aligns with previous observations on slurry strip-till application in maize by Pietzner et al. (2017). In agreement with the i-L-M treatment, this confirms earlier work that incorporation is the most effective measure to lower NH3 fluxes (Webb et al. 2010).
Combined management system effects on N2O and NH3 fluxes
The highest N2O fluxes were obtained for the O-system with maize crop**. Soil conditions with high mineral N and soil moisture content, as well as evidence from isotopic values of emitted N2O, indicated that N2O fluxes of all treatments were dominated by denitrification. Moreover, N2O isotopes suggested that the N2O/(N2 + N2O) ratio of denitrification was comparable among the treatments, with no lower levels observed in the O-system treatments. These results are in line with the third hypothesis which states that the O-system treatments exhibit higher denitrification rates. This is evidenced by the fact that N2O fluxes from denitrification were enhanced in g-O-M treatments. However, the hypothesis that enhanced N2O reduction to N2 by heterotrophic bacterial denitrification would result from the higher bulk density and organic C content of O-system treatments (hypothesis 4) was not confirmed. This was because the stable isotope values of N2O did not show lower N2O/(N2 + N2O ) ratios of denitrification in O-system treatments. The elevated share of fungal denitrification observed in the i-L-M treatment of the Braunschweig site during phases of high soil moisture and NO3− content indicates that the type of N2O production process may be a relevant factor. This may be one of the reasons why our assumption that heterotrophic bacterial denitrification would enhance N2O reduction to N2 in the O-system was not confirmed.
The results indicate that O-system treatments with maize, excluding digestate incorporation, exhibited higher N2O and NH3 fluxes. Furthermore, the data did not indicate any enhanced drought resilience. This raises the question of how the O-system could be further optimized. The utilization of below-root injection techniques, in combination with other techniques for organic fertilization and nitrification inhibitors, has the potential to reduce N2O emissions (Dittert et al. 2001; Ruser and Schulz 2015). Furthermore, the development of injection techniques for growing crops, as demonstrated by the recent success of slot application of slurry in winter wheat, could possibly improve the mitigation of N2O and NH3 emissions (ten Huf et al. 2023; Buchen-Tschiskale et al. 2023).
Conclusions
A comparison was conducted between the N2O and NH3 fluxes and yields of three management systems – with an evaluation of the controlling soil properties. With regard to our hypotheses, we were unable to confirm that organic fertilization leads to higher N2O fluxes in general (hypothesis 1) as only the O-system with maize exhibited the highest N2O fluxes. However, the L-system demonstrated no consistent differences to the C-system. It was suspected that this effect was absent due to the stabilization of organic C during the digestion of pig slurry. This confirmed that the digestion of manures is an effective tool to generate energy and lower N2O fluxes (Möller 2015). The highest N2O emissions in the O-system treatments without digestate incorporation in maize were due to the soil’s increased denitrification potential resulting from reduced tillage, which led to higher bulk density and organic carbon availability. The hypotheses that the incorporation of digestate would result in a reduction in NH3 fluxes (hypothesis 2) was clearly confirmed. The findings demonstrated that the application of digestate to growing crops without incorporation or at a late stage prior to fertilization before seeding resulted in elevated NH3 fluxes. The time course of NH3 fluxes and the subsequent decline after incorporation provides evidence of the potential for mitigating NH3 fluxes, if the lag time after digestate application can be minimized. It was demonstrated that the strip tillage incorporation was an effective method for reducing NH3 fluxes, while it did not result in enhanced N2O fluxes. Enhanced N2O reduction in the O-system (hypothesis 4) was not confirmed since lower N2O/(N2 + N2O) ratios of denitrification were not observed. Overall, the beneficial aspects of the O-system including more stable soil structure and saving of resources are potentially counteracted by enhanced N2O and NH3 emissions. To avoid this, optimized management concepts should include measures to lower NH3 fluxes through acidification and/or advanced injection techniques, and possibly mitigate N2O fluxes through nitrification inhibitors. Given the significant effort required to conduct field-scale studies of management effects on N2O fluxes and processes, and the difficulty in identifying significant effects due to high field-scale variability, future studies should be planned and combined with N-cycle modeling (e.g. Grosz et al. 2023). Models that have been enhanced through the incorporation of suitable field data may prove to be a valuable tool for the development of field management strategies that aim to minimize negative N emission while maintaining or increasing yields.
Data availability
The datasets generated and/or analyzed during the current study are available from the corresponding author upon request.
References
Abdalla M, Osborne B, Lanigan G, Forristal D, Williams M, Smith P, Jones MB (2013) Conservation tillage systems: a review of its consequences for greenhouse gas emissions. Soil Use Manag 29:199–209. https://doi.org/10.1111/sum.12030
Baral KR, Arthur E, Olesen JE, Petersen SO (2016) Predicting nitrous oxide emissions from manure properties and soil moisture: an incubation experiment. Soil Biol Biochem 97:112–120. https://doi.org/10.1016/j.soilbio.2016.03.005
Bayer C, Gomes J, Zanatta JA, Vieira FCB, Dieckow J (2016) Mitigating greenhouse gas emissions from a subtropical Ultisol by using long-term no-tillage in combination with legume cover crops. Soil Tillage Res 161:86–94. https://doi.org/10.1016/j.still.2016.03.011
Buchen C, Well R, Helfrich M, Fuss R, Kayser M, Gensior A, Benke M, Flessa H (2017) Soil mineral N dynamics and N2O emissions following grassland renewal. Agric Ecosyst Environ 246:325–342. https://doi.org/10.1016/j.agee.2017.06.013
Buchen C, Lewicka-Szczebak Dominika F, Heinz W Reinhard (2018) Estimating N2O processes during grassland renewal and grassland conversion to maize crop** using N2O isotopocules. Rapid Commun Mass Spectrom 32:1053–1067. https://doi.org/10.1002/rcm.8132
Buchen-Tschiskale C, Well R, Flessa H (2023) Tracing nitrogen transformations during spring development of winter wheat induced by 15N labeled cattle slurry applied with different techniques. Sci Total Environ 871:162061. https://doi.org/10.1016/j.scitotenv.2023.162061
Bundesministerium der Justiz (2020) Verordnung über die Anwendung von Düngemitteln, Bodenhilfsstoffen, Kultursubstraten und Pflanzenhilfsmitteln nach den Grundsätzen der guten fachlichen Praxis beim Düngen (Düngeverordnung - DüV). https://www.gesetze-im-internet.de/d_v_2017/BJNR130510017.html
Butterbach-Bahl K, Baggs EM, Dannenmann M, Kiese R, Zechmeister-Boltenstern S (2013) Nitrous oxide emissions from soils: how well do we understand the processes and their controls? Philos Trans R Soc B Biol Sci 368:20130122. https://doi.org/10.1098/rstb.2013.0122
Deppe M, Well R, Kücke M, Fuss R, Giesemann A, Flessa H (2016) Impact of CULTAN fertilization with ammonium sulfate on field emissions of nitrous oxide. Agric Ecosyst Environ 219:138–151. https://doi.org/10.1016/j.agee.2015.12.015
Deppe M, Well R, Giesemann A, Spott O, Flessa H (2017) Soil N2O fluxes and related processes in laboratory incubations simulating ammonium fertilizer depots. Soil Biol Biochem 104:68–80. https://doi.org/10.1016/j.soilbio.2016.10.005
Dittert K, Bol R, King R, Chadwick D, Hatch D (2001) Use of a novel nitrification inhibitor to reduce nitrous oxide emission from15N-labelled dairy slurry injected into soil. Rapid Commun Mass Spectrom RCM 15:1291–1296. https://doi.org/10.1002/rcm.335
Drury CF, Reynolds WD, Tan CS, Welacky TW, Calder W, McLaughlin NB (2006) Emissions of Nitrous Oxide and Carbon Dioxide. Soil Sci Soc Am J 70:570–581. https://doi.org/10.2136/sssaj2005.0042
Drury CF, Reynolds WD, Yang XM, McLaughlin NB, Welacky TW, Calder W, Grant CA (2012) Nitrogen Source, Application Time, and Tillage effects on Soil Nitrous Oxide emissions and Corn Grain yields. Soil Sci Soc Am J 76:1268–1279. https://doi.org/10.2136/sssaj2011.0249
Emmerling C, Krein A, Junk J (2020) Meta-analysis of strategies to reduce NH3 emissions from slurries in European Agriculture and Consequences for Greenhouse Gas Emissions. Agronomy 10:1633
Erisman JW, Galloway J, Seitzinger S, Bleeker A, Butterbach-Bahl K (2011) Reactive nitrogen in the environment and its effect on climate change. Curr Opin Environ Sustain 3:281–290. https://doi.org/10.1016/j.cosust.2011.08.012
Flessa H, Beese F (2000) Laboratory Estimates of Trace Gas Emissions following Surface application and injection of cattle slurry. J Environ Qual 29:262–268. https://doi.org/10.2134/jeq2000.00472425002900010033x
Flessa HGJ, Hofmeier M, Dittert K, Ruser R, Osterburg B, Poddey E, Wulf S, Pacholski A (2014) Minderung Von Stickstoff-Emissionen Aus Der Landwirtschaft: Empfehlungen für die Praxis Und Aktuelle Fragen an die Wissenschaft. Senat der Bundesforschungsinstitute des Bundesministeriums für Ernährung und Landwirtschaft, Berlin
Galloway JN, Townsend AR, Erisman JW, Bekunda M, Cai Z, Freney JR, Martinelli LA, Seitzinger SP, Sutton MA (2008) Transformation of the Nitrogen cycle: recent Trends, questions, and potential solutions. Science 320:889–892. https://doi.org/10.1126/science.1136674
Groffman PM, Tiedje JM, Robertson GP, Christensen S (1988) Denitrification at different temporal and geographical scales: proximal and distal controls. In: Wilson JR (ed) Advances in nitrogen cycling in agricultural ecosystems. CAB International, Wallingford, UK, pp 174–192
Grosz B, Kemmann B, Burkart S, Petersen SO, Well R (2022) Understanding the Impact of Liquid Organic Fertilisation and Associated Application techniques on N2, N2O and CO2 fluxes from agricultural soils. Agriculture 12:692. https://doi.org/10.3390/agriculture12050692
Grosz BP, Matson A, Butterbach-Bah K, Clough T, Davidson EA, Dechow R, DelGrosso S, Diamantopoulos E, Dörsch P, Haas E, He H, Henri CV, Hui D, Kleineidam K, Kraus D, Kuhnert M, Léonard J, Mueller C, Petersen SO, Sihi D, Vogeler I, Well R, Yeluripati JB, Zhang J, Scheer C (2023) Modeling denitrification: can we report what we don’t know? AGU advances 4:e2023AV000990. https://doi.org/10.1029/2023AV000990
Hansen S, Berland Frøseth R, Stenberg M, Stalenga J, Olesen JE, Krauss M, Radzikowski P, Doltra J, Nadeem S, Torp T, Pappa V, Watson CA (2019) Reviews and syntheses: review of causes and sources of N2O emissions and NO3 leaching from organic arable crop rotations. Biogeosciences 16:2795–2819. https://doi.org/10.5194/bg-16-2795-2019
Harter J, Guzman-Bustamante I, Kuehfuss S, Ruser R, Well R, Spott O, Kappler A, Behrens S (2016) Gas entrapment and microbial N2O reduction reduce N2O emissions from a biochar-amended sandy clay loam soil. Sci Rep 6. https://doi.org/10.1038/srep39574
Helfrich M, Nicolay G, Well R, Buchen-Tschiskale C, Dechow R, Fuss R, Gensior A, Paulsen HM, Berendonk C, Flessa H (2020) Effect of chemical and mechanical grassland conversion to cropland on soil mineral N dynamics and N2O emission. Agric Ecosyst Environ 298. https://doi.org/10.1016/j.agee.2020.106975
Herr C, Mannheim T, Müller T, 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. https://doi.org/10.1002/jpln.201800376
Hey C, Hertin J, Baron M, Dahms H, Doyle U, Dross M, Elmer C-F, Jung S, Kruse A, Kuhn T, Nabitz L, Salomon M, Schmid E, Steuwer SD, Täuber S, Volkens A, Wiegand S, Wiese J (2015) SRU 2015, Stickstoff: Lösungsstrategien für Ein drängendes Umweltproblem. Sondergutachten. Sachverständigenrat für Umweltfragen (SRU). Geschäftsstelle des Sachverständigenrates für Umweltfragen, Berlin
Hink L, Nicol GW, Prosser JI (2017) Archaea produce lower yields of N2O than bacteria during aerobic ammonia oxidation in soil. Environ Microbiol 19:4829–4837. https://doi.org/10.1111/1462-2920.13282
Hou Y, Velthof GL, Oenema O (2015) Mitigation of ammonia, nitrous oxide and methane emissions from manure management chains: a meta-analysis and integrated assessment. Glob Change Biol 21:1293–1312. https://doi.org/10.1111/gcb.12767
Huang Y, Ren W, Wang L, Grove JH, Yang X, Tao B, Goff B (2018) Greenhouse gas emissions and crop yield in no-tillage systems: a meta-analysis. Agric Ecosyst Environ 268:144–153. https://doi.org/10.1016/j.agee.2018.09.002
Hutchinson GL, Mosier AR (1981) Improved Soil Cover Method for Field Measurement of Nitrous Oxide fluxes. Soil Sci Soc Am J 45:311–316. https://doi.org/10.2136/sssaj1981.03615995004500020017x
Jacinthe P-A, Dick WA (1997) Soil management and nitrous oxide emissions from cultivated fields in southern Ohio. Soil Tillage Res 41:221–235. https://doi.org/10.1016/S0167-1987(96)01094-X
Johnson JMF, Archer D, Barbour N (2010) Greenhouse Gas Emission from contrasting management scenarios in the Northern Corn Belt. Soil Sci Soc Am J 74:396–406. https://doi.org/10.2136/sssaj2009.0008
Jungkunst HF, Freibauer A, Neufeldt H, Bareth G (2006) Nitrous oxide emissions from agricultural land use in Germany – a synthesis of available annual field data. J Plant Nutr Soil Sci 169:341–351. https://doi.org/10.1002/jpln.200521954
König V, Kießling G (2001) Untersuchungsergebnisse und Hinweise Zur Sachgerechten Behandlung Von Nmin -Bodenproben. In: unpublished results by Thüringer Landesanstalt für Landwirtschaft. Jena, Germany
Köster JR, Cárdenas LM, Bol R, Lewicka-Szczebak D, Senbayram M, Well R, Giesemann A, Dittert K (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. https://doi.org/10.1016/j.soilbio.2015.01.021
Kramps-Alpmann D (2017) Klimaoptimierte Anpassungsstrategien in Der Landwirtschaft (optimierter Klimabetrieb) II. Unpubl Proj Rep
Krauss M, Ruser R, Muller T, Hansen S, Mader P, Gattinger A (2017) Impact of reduced tillage on greenhouse gas emissions and soil carbon stocks in an organic grass-clover ley - winter wheat crop** sequence. Agric Ecosyst Environ 239:324–333. https://doi.org/10.1016/j.agee.2017.01.029
Laughlin RJ, Stevens RJ (2002) Evidence for fungal dominance of denitrification and codenitrification in a grassland soil. Soil Sci Soc Am J 66:1540–1548. https://doi.org/10.2136/sssaj2002.1540
Leiber-Sauheitl K, Fuß R, Voigt C, Freibauer A (2014) High CO2 fluxes from grassland on histic gleysol along soil carbon and drainage gradients. Biogeosciences 11:749–761. https://doi.org/10.5194/bg-11-749-2014
Lewicka-Szczebak D, Augustin J, Giesemann A, Well R (2017) Quantifying N2O reduction to N2 based on N2O isotopocules – validation with independent methods (Helium incubation and 15N gas flux method). Biogeosciences 14:711–732. https://doi.org/10.5194/bg-14-711-2017
Lewicka-Szczebak D, Lewicki MP, Well R (2020) N2O isotope approaches for source partitioning of N2O production and estimation of N2O reduction – validation with 15N gas-flux method in laboratory and field studies (2020). Biogeosciences 17:5513–5537. https://doi.org/10.5194/bg-17-5513-2020
Loftfield N, Flessa H, Augustin J, Beese F (1997) Automated gas chromatographic system for rapid analysis of the atmospheric trace gases methane, carbon dioxide, and nitrous oxide. J Environ Qual 26:560–564
Loyon L (2018) Overview of animal Manure Management for Beef, Pig, and Poultry farms in France. Front Sustain Food Syst 2:36. https://doi.org/10.3389/fsufs.2018.00036
Ma ST, Shan J, Yan XY (2017) N2O emissions dominated by fungi in an intensively managed vegetable field converted from wheat-rice rotation. Appl Soil Ecol 116:23–29. https://doi.org/10.1016/j.apsoil.2017.03.021
Mohn J, Wolf B, Toyoda S, Lin CT, Liang MC, Bruggemann N, Wissel H, Steiker AE, Dyckmans J, Szwec L, Ostrom NE, Casciotti KL, Forbes M, Giesemann A, Well R, Doucett RR, Yarnes CT, Ridley AR, Kaiser J, Yoshida N (2014) Interlaboratory assessment of nitrous oxide isotopomer analysis by isotope ratio mass spectrometry and laser spectroscopy: current status and perspectives. Rapid Commun Mass Spectrom 28:1995–2007. https://doi.org/10.1002/rcm.6982
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. https://doi.org/10.1007/s13593-015-0284-3
Möller K, 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. https://doi.org/10.1016/j.eja.2008.06.003
Müller C, Clough TJ (2014) Advances in understanding nitrogen flows and transformations: gaps and research pathways. J Agric Sci 152:S34–S44. https://doi.org/10.1017/s0021859613000610
Müller C, Laughlin RJ, Spott O, Rütting T (2014) Quantification of N2O emission pathways via a 15N tracing model. Soil Biol Biochem 72:44–54. https://doi.org/10.1016/j.soilbio.2014.01.013
Nyameasem JK, Ruser R, Kluß C, Essich C, Zutz M, ten Huf M, Buchen-Tschiskale C, Flessa H, Olfs HW, Taube F, Reinsch T (2023) Effect of slurry application techniques on nitrous oxide emission from temperate grassland under varying soil and climatic conditions. Grass Forage Sci 78:338–358. https://doi.org/10.1111/gfs.12612
Pacholski A (2016) Calibrated Passive Sampling - Multi-plot Field measurements of NH3 emissions with a combination of dynamic tube Method and Passive samplers. J Vis Exp 53273. https://doi.org/10.3791/53273
Pedersen A (2017) Package’HMR’–Flux Estimation with Static Chamber Data. Version 04 2
Pedersen AR, Petersen SO, Schelde K (2010) A comprehensive approach to soil-atmosphere trace-gas flux estimation with static chambers. Eur J Soil Sci 61:888–902. https://doi.org/10.1111/j.1365-2389.2010.01291.x
Pedersen EJ, Miller DL, Simpson GL, Ross N (2019) Hierarchical generalized additive models in ecology: an introduction with mgcv. PeerJ 7:e6876. https://doi.org/10.7717/peerj.6876
Petersen SO, Nissen HH, Lund I, Ambus P (2003) Redistribution of Slurry Components as Influenced by Injection Method, Soil, and Slurry properties. J Environ Qual 32:2399–2409. https://doi.org/10.2134/jeq2003.2399
Petersen J, Hansen B, Sørensen P (2004) Nitrification of 15N-ammonium sulphate and crop recovery of 15N-labelled ammonium nitrates injected in bands. Eur J Agron 21:81–92. https://doi.org/10.1016/S1161-0301(03)00087-X
Petersen S, Schjonning P, Thomsen I, Christensen B (2008) Nitrous oxide evolution from structurally intact soil as influenced by tillage and soil water content. Soil Biol Biochem 40:967–977. https://doi.org/10.1016/j.soilbio.2007.11.017
Petersen SO, Mutegi JK, Hansen EM, Munkholm LJ (2011) Tillage effects on N2O emissions as influenced by a winter cover crop. Soil Biol Biochem 43:1509–1517. https://doi.org/10.1016/j.soilbio.2011.03.028
Petersen SO, Baral KR, Arthur E (2016) Manure distribution as a predictor of N2O emissions from soil. Anim Prod Sci 56:549–556. https://doi.org/10.1071/an15534
Pietzner B, Rücknagel J, Koblenz B, Bednorz D, Tauchnitz N, Bischoff J, Köbke S, Meurer KHE, Meißner R, Christen O (2017) Impact of slurry strip-till and surface slurry incorporation on NH2 and N2O emissions on different plot trials in Central Germany. Soil Tillage Res 169:54–64. https://doi.org/10.1016/j.still.2017.01.011
R Core Team (2020) R: A Language and Environment for Statistical Computing. Vienna, Austria. http://www.R-project.org/
Ravishankara AR, Daniel JS, Portmann RW (2009) Nitrous Oxide (N2O): the Dominant ozone-depleting substance emitted in the 21st Century. Science 326:123–125. https://doi.org/10.1126/science.1176985
Röckmann T, Kaiser J, Brenninkmeijer CAM, Brand WA (2003) Gas chromatography/isotope-ratio mass spectrometry method for high-precision position-dependent 15N and 18O measurements of atmospheric nitrous oxide. Rapid Commun Mass Spectrom 17:1897–1908. https://doi.org/10.1002/Rcm.1132
Rohe L, Well R, Lewicka-Szczebak D (2017) Use of oxygen isotopes to differentiate between nitrous oxide produced by fungi or bacteria during denitrification. Rapid Commun Mass Spectrom 31:1297–1312. https://doi.org/10.1002/rcm.7909
Rohe L, Oppermann T, Well R, Horn MA (2020) Nitrite induced transcription of p450nor during denitrification by Fusarium oxysporum correlates with the production of N2O with a high 15N site preference. Soil Biol Biochem 151:108043. https://doi.org/10.1016/j.soilbio.2020.108043
Rohe L, Anderson TH, Flessa H, Goeske A, Lewicka-Szczebak D, Wrage-Mönnig N, Well R (2021) Comparing modified substrate-induced respiration with selective inhibition (SIRIN) and N2O isotope approaches to estimate fungal contribution to denitrification in three arable soils under anoxic conditions. Biogeosciences 18:4629–4650. https://doi.org/10.5194/bg-18-4629-2021
Ruser R, Schulz R (2015) The effect of nitrification inhibitors on the nitrous oxide (N2O) release from agricultural soils-a review. J Plant Nutr Soil Sci 178:171–188. https://doi.org/10.1002/jpln.201400251
Ruser R, Flessa H, Schilling R, Beese F, Munch JC (2001) Effect of crop-specific field management and N fertilization on N2O emissions from a fine-loamy soil. Nutr Cycl Agroecosystems 59:177–191. https://doi.org/10.1023/a:1017512205888
Scheer C, Fuchs K, Pelster DE, Butterbach-Bahl K (2020) Estimating global terrestrial denitrification from measured N2O:(N2O + N2) product ratios. Curr Opin Environ Sustain 47:72–80. https://doi.org/10.1016/j.cosust.2020.07.005
Sehy U, Ruser R, Munch JC (2003) Nitrous oxide fluxes from maize fields: relationship to yield, site-specific fertilization, and soil conditions. Agric Ecosyst Environ 99:97–111. https://doi.org/10.1016/s0167-8809(03)00139-7
Senbayram M, Chen R, Wienforth B, Herrmann A, Kage H, Mühling KH, Dittert K (2014) Emission of N2O from Biogas Crop Production systems in Northern Germany. BioEnergy Res 7:1223–1236. https://doi.org/10.1007/s12155-014-9456-2
Senbayram M, Well R, Bol R, Chadwick DR, Jones DL, Wu D (2018) Interaction of straw amendment and soil NO3 content controls fungal denitrification and denitrification product stoichiometry in a sandy soil. Soil Biol Biochem 126:204–212. https://doi.org/10.1016/j.soilbio.2018.09.005
Senbayram M, Budai A, Bol R, Chadwick D, Marton L, Gundogan R, Wu D (2019) Soil NO3 level and O2 availability are key factors in controlling N2O reduction to N2 following long-term liming of an acidic sandy soil. Soil Biol Biochem 132:165–173. https://doi.org/10.1016/j.soilbio.2019.02.009
Senbayram M, Well R, Shan J, Bol R, Burkart S, Jones DL, Wu D (2020) Rhizosphere processes in nitrate-rich barley soil tripled both N2O and N2 losses due to enhanced bacterial and fungal denitrification. Plant Soil 448:509–522. https://doi.org/10.1007/s11104-020-04457-9
Senbayram M, Wei Z, Wu D, Shan J, Yan X, Well R (2022) Inhibitory effect of high nitrate on N2O reduction is offset by long moist spells in heavily N loaded arable soils. Biol Fertil Soils 58:77–90. https://doi.org/10.1007/s00374-021-01612-x
Shen W, Xue H, Gao N, Shiratori Y, Kamiya T, Fujiwara T, Isobe K, Senoo K (2020) Effects of copper on nitrous oxide (N2O) reduction in denitrifiers and N2O emissions from agricultural soils. Biol Fertil Soils 56:39–51. https://doi.org/10.1007/s00374-019-01399-y
Shoun H, Fushinobu S (2016) CHAPTER 14. Denitrification in Fungi. In: Moura I, Moura JJG, Pauleta SR, Maia LB (eds) Metallobiology. Royal Society of Chemistry, Cambridge, pp 331–348
Sommer SG, Hutchings NJ (2001) Ammonia emission from field applied manure and its reduction—invited paper. Eur J Agron 15:1–15. https://doi.org/10.1016/S1161-0301(01)00112-5
Sommer SG, Génermont S, Cellier P, Hutchings NJ, Olesen JE, Morvan T (2003) Processes controlling ammonia emission from livestock slurry in the field. Eur J Agron 19:465–486. https://doi.org/10.1016/S1161-0301(03)00037-6
ten Huf M, Reinsch T, Zutz M, Essich C, Ruser R, Buchen-Tschiskale C, Flessa H, Olfs HW (2023) Effects of Liquid Manure Application techniques on Ammonia Emission and Winter Wheat Yield. Agronomy 13:472. https://doi.org/10.3390/agronomy13020472
Toyoda S, Yoshida N (1999) Determination of Nitrogen isotopomers of Nitrous Oxide on a modified isotope ratio Mass Spectrometer. Anal Chem 71:4711–4718. https://doi.org/10.1021/ac9904563
van Kessel C, Venterea R, Six J, Adviento-Borbe MA, Linquist B, van Groenigen KJ (2013) Climate, duration, and N placement determine N2O emissions in reduced tillage systems: a meta-analysis. Glob Change Biol 19:33–44. https://doi.org/10.1111/j.1365-2486.2012.02779.x
Velthof GL, Mosquera J (2011) The impact of slurry application technique on nitrous oxide emission from agricultural soils. Agric Ecosyst Environ 140:298–308. https://doi.org/10.1016/j.agee.2010.12.017
Venterea RT, Clough TJ, Coulter JA, Breuillin-Sessoms F (2015) Ammonium sorption and ammonia inhibition of nitrite-oxidizing bacteria explain contrasting soil N2O production. Sci Rep 5:12153. https://doi.org/10.1038/srep12153
Walter K, Don A, Flessa H (2015) Net N2O and CH4 soil fluxes of annual and perennial bioenergy crops in two central German regions. Biomass Bioenergy 81:556–567. https://doi.org/10.1016/j.biombioe.2015.08.011
Webb J, Pain B, Bittman S, Morgan J (2010) The impacts of manure application methods on emissions of ammonia, nitrous oxide and on crop response—A review. Agric Ecosyst Environ 137:39–46. https://doi.org/10.1016/j.agee.2010.01.001
Weller U, Albrecht L, Schlüter S, Vogel HJ (2022) An open soil structure Library based on X-ray CT data. SOIL 8:507–515. https://doi.org/10.5194/soil-8-507-2022
Westley MB, Popp BN, Rust TM (2007) The calibration of the intramolecular nitrogen isotope distribution in nitrous oxide measured by isotope ratio mass spectrometry. Rapid Commun Mass Spectrom 21:391–405. https://doi.org/10.1002/Rcm.2828
Wood SN (2006) Generalized additive models: an introduction with R. Chapman & Hall/CRC, Boca Raton, FL
Wood SN (2011) Fast stable restricted maximum likelihood and marginal likelihood estimation of Semiparametric Generalized Linear models. J R Stat Soc Ser B Stat Methodol 73:3–36. https://doi.org/10.1111/j.1467-9868.2010.00749.x
Wrage-Monnig N, Horn MA, Well R, Muller C, Velthof G, Oenema O (2018) The role of nitrifier denitrification in the production of nitrous oxide revisited. Soil Biol Biochem 123:A3–A16. https://doi.org/10.1016/j.soilbio.2018.03.020
Yu L, Harris E, Lewicka-Szczebak D, Barthel M, Blomberg MRA, Harris SJ, Johnson MS, Lehmann MF, Liisberg J, Müller C, Ostrom NE, Six J, Toyoda S, Yoshida N, Mohn J (2020) What can we learn from N 2 O isotope data? – analytics, processes and modelling. Rapid Commun Mass Spectrom 34:e8858. https://doi.org/10.1002/rcm.8858
Zhu X, Burger M, Doane TA, Horwath WR (2013) Ammonia oxidation pathways and nitrifier denitrification are significant sources of N2O and NO under low oxygen availability. Proc Natl Acad Sci U S A 110:6328–6333. https://doi.org/10.1073/pnas.1219993110
Žurovec O, Wall DP, Brennan FP, Krol DJ, Forrestal PJ, Richards KG (2021) Increasing soil pH reduces fertiliser derived N2O emissions in intensively managed temperate grassland. Agric Ecosyst Environ 311:107319. https://doi.org/10.1016/j.agee.2021.107319
Zuur AF, Leno EN, Elphick CS (2010) A protocol for data exploration to avoid common statistical problems: data exploration. Methods Ecol Evol 1:3–14. https://doi.org/10.1111/j.2041-210X.2009.00001.x
Acknowledgements
We would like to thank Dorothee Kramps-Alpmann, Karen Korte, Ulrike Ernst, and Karoline Röper for their contributions to the field work and data evaluation related to soil and plant sampling and GHG flux measurement. We would also like to thank Annette Goeske and Martina Heuer for their assistance with the IRMS analysis of gas samples, Daniel Ziehe, Andrea Oehns-Rittgerod, and Kerstin Gilke for GC analysis of gas samples, Roland Fuß for his assistance with the statistical data evaluation, and Ulrike Manthey for language proofreading of the manuscript. We would like to thank the reviewers and editors for their detailed and constructive feedback on the manuscript.
Funding
Open Access funding enabled and organized by Projekt DEAL. This research work was supported by funds from the Federal Ministry of Food and Agriculture (BMEL) through the Federal Office for Agriculture and Food (BLE) and conducted in cooperation with the research unit 2337: “Denitrification in Agricultural Soils: Integrated Control and Modeling at Various Scales (DASIM)” (German Research Foundation, grant WE 1904/10 − 1 to Reinhard Well). Dominika Lewicka-Szczebak was supported by the “Polish Returns” program of the Polish National Agency of Academic Exchange (NAWA) and the grant Opus-516204 of the National Science Centre Poland (NCN). Open Access funding enabled and organized by Projekt DEAL.
Open Access funding enabled and organized by Projekt DEAL.
Author information
Authors and Affiliations
Corresponding author
Ethics declarations
Conflict of interest
The authors have no conflicts of interest that are relevant to the content of this article.
Additional information
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Electronic supplementary material
Below is the link to the electronic supplementary material.
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Well, R., Ruoss, N., Grosz, B. et al. Effect of agricultural management system (“cash crop”, “livestock” and “climate optimized”) on nitrous oxide and ammonia emissions. Biol Fertil Soils (2024). https://doi.org/10.1007/s00374-024-01843-8
Received:
Revised:
Accepted:
Published:
DOI: https://doi.org/10.1007/s00374-024-01843-8