Conservation agriculture affects soil organic matter distribution, microbial metabolic capacity and nitrogen turnover under Danish field conditions

Conservation agriculture (CA) has been reported to affect nutrient cycling. This study aims to investigate how CA induced soil organic matter stratification affects carbon and nitrogen turnover. A case farm study was established on two Danish farms with conventional ploughed tillage (P) and CA practises. Here, we studied how organic matter stratification patterns to 50 cm soil depth differed between the two systems. Further we investigated differences in carbon and nitrogen mineralization patterns in lab incubation experiments. Average stratification ratio, the ratio between soil C and N content in the upper 5 cm and at 20 – 30 cm, the depth of the plough layer in the ploughed system, was 1.86 and 1.61 under CA and 1.04 and 1.06 under P. Carbon respiration from intact soil core incubation was affected by soil total carbon content, and showed stronger stratification in CA than in P. Nitrogen mineralization rates from intact soil core incubation was largest in CA top-layer compared to CA 13.5 – 16.5 cm layer and both P soil layers, with initial mineralization followed by immobilization during the second half of the a four-week incubation. Net change in mineral nitrogen after incubation was only apparent in the 13.5 – 16.5 cm layer in P, with an average N mineralization rate of 0.08 mg N kg (cid:0) 1 soil d (cid:0) 1 . Sieving to 2 mm did not affect N mineralization dynamics. Field-based ammonium to nitrate ratio was higher in CA than in P soils, across varying soil depths and time-points over the entire year. Soil acidity was reduced by one pH unit in CA compared to P. Microbial metabolic capacity was significantly larger in the top 5 cm of CA from the deeper depths, and from the P soils. In conclusion, carbon and nitrogen mineralization, as well as microbial metabolic capacity were strongly affected by an increased stratification ratio of organic matter in CA.


Introduction
Conservation agriculture (CA) farming has been proposed as a strategy to increase both nitrogen use efficiency and soil organic matter (SOM) build-up in cropping systems (Habbib et al., 2017(Habbib et al., , 2016;;Hobbs et al., 2008;Lal, 2009).Conservation agriculture alters SOM pools by vertical stratification through no-till and by residue inputs from crop residues and cover crops.Through this, CA potentially changes the microbial biomass and community structure.It is, however, unclear how these changes affect microbial activity and carbon (C) and nitrogen (N) turnover under temperate field conditions.Increasing our understanding of how tillage practices impact N and C dynamics in a temperate climate can help us design more sustainable cropping systems that stabilize SOM contents, and supply crops more effectively with nutrients.
No-till practices induce stratification of SOM in distinctive depthlayers (Büchi et al., 2017;Franzluebbers, 2002).This differs from conventional tilling practices where the soil is mixed in the plough horizon annually.With time, no-till practices result in depth gradients in soil organic C, total N, potential mineralisable N (PMN), and soil microbial abundance and activity (Kandeler et al., 1999;Sun et al., 2011Sun et al., , 2018;;Büchi et al., 2017;Schmidt et al., 2019).Often, studies on the effect of no-till on SOM dynamics limit their investigation to the uppermost soil layer (Van Den Bossche et al., 2009;Mbuthia et al., 2015;Kraut-Cohen et al., 2020;Sekaran et al., 2020).This focus on the upper layer of soil has earlier caused an overestimation of the effect of no-till on SOM sequestration (Baker et al., 2007;Angers and Eriksen-Hamel, 2008).While earlier studies found no-till management to increase soil organic carbon (SOC) content compared to ploughed soils (Alvarez, 2005;Angers and Eriksen-Hamel, 2008;Franzluebbers, 2010;Johnson et al., 2005), recent studies on potential SOC sequestration by no-till seems to agree that there is little to no effect in temperate climates, when the entire profile and changes in soil bulk density are taken into account (Meurer et al., 2018;Ogle et al., 2019;Sun et al., 2020).Since stratification is even more pronounced in the labile SOM pools subject to fast turnover (Kandeler et al., 1999;Franzluebbers, 2002), studying the depth dimension seems crucial for our understanding of the impact of no-till soils on SOM and residue dynamics.
Despite SOM stratification accounting for most of the increased topsoil SOC under no-till in temperate climates, C turnover might still decrease under no-till due to physio-chemical protection of organic matter.The higher degree of soil aggregate stability in no-till soils may limit microbial access to SOM and thereby prevent turnover (Six et al., 2000).Moreover, structural changes in no-till may limit oxygen availability (Brewer et al., 2018) or increase surface adsorption (Von Lützow et al., 2008) and thereby limit SOM decomposition.
In addition to reduced tillage, CA farming affects SOM by the increased use of cover crops (FAO, 2017), which alters the amount, time-period and the quality of residue input.The increase in residue input may drive increase in SOM content, and in the case of leguminous species residue quality may increase SOM stability (McDaniel et al., 2014;Poeplau and Don, 2015;Tiemann et al., 2015).In fact, McDaniel et al. (2014) found that cover crops increased SOM contents independently from tillage system.The authors argue that while reduced tillage increases physically protected SOM, cover crops and rotational diversity may increase biologically stabilized SOM.Increasing understanding of SOM and the role of microbial biomass in stabilizing SOM sheds light on the importance of N availability for SOM dynamics and stabilization (Coonan et al., 2020;Kirkby et al., 2014;Lavallee et al., 2020).The increase in crop diversity, the organic N input via leguminous cover crops and the extended period of rhizodeposition via cover crops seem to drive microbial activity and therefore increase stable, microbial derived SOM (Cotrufo et al., 2019;De Notaris et al., 2020;Kopittke et al., 2020;Tiemann et al., 2015).
The CA farming induced changes in soil physical conditions and residue input are causing microbial community changes that may impact SOM turnover dynamics (Mathew et al., 2012;Mbuthia et al., 2015;Kim et al., 2020;Li et al., 2020).However, little is known about the functional implications of microbial community changes on C or N mineralization (Nannipieri et al., 2020) and how these changes affect microbial capacity, measured as the ability to decompose C-substrates with varying structural complexity (Campbell et al., 2003).Nonetheless, earlier studies found indications of an impact of microbial community change on C and N turnover due to an increase in C and N cycling enzymes under no-till and cover crops, (Kandeler et al., 1999;Mangalassery et al., 2015;Mbuthia et al., 2015;Van Den Bossche et al., 2009).
Previously, changes in N turnover resulting from CA management practices have been reported: Net N mineralization and nitrification potential was increased in no-till topsoil (Kandeler et al., 1999), showing a clear stratification pattern (Franzluebbers, 2002).The effect of no-till on N mineralization dynamics differs between studies.Some studies report net N immobilization effects of no-till, due to high microbial activity in the residue-rich topsoil layer; hence driving demand for N assimilation especially post-harvest (Laine et al., 2017;Li et al., 2016;Mangalassery et al., 2015;Van Den Bossche et al., 2009).Other studies on net N mineralization report on effects ranging from no net changes (Gómez-Rey et al., 2012) to increased net N mineralization under no-till (Liu et al., 2017).How tillage system affects N turnover in periods with less active primary crop residue decomposition is less studied.In these periods, soil structural differences might be of greater importance compared to residue input parameters.More clear seems to be the effect of cover crops on N turnover dynamics: In their meta-analysis Mahal et al. (2018) found 17 times larger effect on the increase in PMN when leguminous cover crops were included compared to the change from ploughing to no-till.
Often the effect of CA farming on SOM dynamics has been studied on soil samples that have been significantly disturbed during sampling (Gómez-Rey et al., 2012;Kandeler et al., 1999;Liu et al., 2017;Mangalassery et al., 2015).Any soil sampling related disturbance may affect no-till soils to a higher degree than regularly disturbed ploughed soils, since soil structure and soil biology are adapted to the more natural, lesser degree of disturbance.Hence, soil disturbance might reduce in-field generalizability of the analysis results.
Apart from the technical details of the soil analysis, the soil management in itself is crucial to consider for assessing the effect of tillage on SOM turnover dynamics.It takes several years for the soil to adapt to management changes (Cooper et al., 2021;Six et al., 2004).Furthermore, tillage system changes often include many more agronomic activities than would be able to include in a truly replicated small-scale field study.In fact, Cooper et al. (2017) argues that there is a lack of field and farm based studies of tillage systems.Here, we studied a CA cropping system as close to practical farming as possible.We used fields that have been managed according to the respective tillage system for more than 11 years prior the experimental phase, to allow structural changes to have taken place already.
The overall objective of the present study was to investigate how the vertical distribution of SOM in soils from ploughed (P) and CA fields differ and how this affects C and N mineralization.More specifically, our research hypothesis was: (1) CA soil increases SOM stratification compared to P soil.Furthermore, (2) microbial basal respiration and potential catabolic capacity and (3) nitrogen mineralization rate will follow this increased stratification gradient.

Sampling site and study design
A case farm study was performed on two neighbouring farms.The study sites were located near Sorø, Denmark (55 24 ′ 58 N 11 38 ′ 12 E), where fields of two case farms were investigated.The two farms were in close proximity to each other, with intermingled fields.One of the farms, applies a plant production system using conventional mouldboard ploughing (P), with soil inverting tillage every year to a depth of 25 cm.In this system, straw is removed every second or third year, organic fertilization in the form of pig slurry (15-20 Mg ha − 1 , 81-85 kg DM Mg − 1 , 5.0-6.1 kg N Mg − 1 , 1.1-1.3kg P Mg − 1 , 2.5-2.8kg K Mg − 1 , after Sørensen et al., 2020) is applied almost every year and has been so for the last 25 years, and this is supplemented with mineral fertilizers to meet crop requirements.The other farm applies a plant production system following the principles of conservation agriculture (CA) (FAO,

Table 1
Soil texture and organic matter content from the top 50 cm soil in the studied fields: CA exp and P exp being the experimental plots and CA 1 , CA 2 , P 1 , P 2 the additional fields.Numbers in parenthesis represent standard error (n block = 3).In 2017, a field trial was set up on one field of each farm.These two fields (CA exp , P exp ) were located in close proximity (within 1000 m) and had similar topsoil texture (Table 1).A cereal crop rotation (alternating winter wheat and spring barley) was established on the experimental fields during the experimental period.Crops were managed as typical for the farms (Tab.A.1).
Within the experimental fields, a field trial design was installed for a cover crop experiment.Cover crop treatments (with and without cover crop) were established after winter wheat harvest in 2017 following a randomized complete block design with three replicates and a plot size of 6 × 10 m.A mixture of oilseed radish and common vetch (8 kg ha − 1 + 25 kg ha − 1 ) was used as the cover crop treatment, or the soil was left bare in the no cover crop treatment.Cover crops were sown post-harvest using a Weaving GD drill.Straw removal occurred after the farms typical practice.The cover crop treatment was repeated after every cereal harvest in 2017-2018 (Tab. A.1) (Tab. A.1).
Additional to the experimental fields CA exp and P exp , two extra fields of each case farm were included in some of the soil analysis (CA 1 , CA 2 , P 1 and P 2 ).This was done to test, whether findings from the experimental field trial plots applied also at farm level.These four additional fields had the same field history, similar soil texture (Table 1) and are within a radius of 1.2 km to the experimental fields.
Temperature and precipitation measurements were provided by the Danish meteorological institute station (DMI) at Flakkebjerg, which is ca 18 km south-west of the experimental fields.

Bulk samples
Soil samples were collected in March 2019 in a standing winter wheat crop (cv Benchmark) before spring N fertilization.Soil samples were taken from 50 cm deep trenches (Table 2).At the profile wall of each trench, a soil column with the surface area of 10 × 20 cm was sampled to 50 cm soil depth.Each sample was divided in five depth intervals with 0-5 cm, 5-10 cm, 10-20 cm, 20-30 cm and 30-50 cm increments.Soil from each bulk sample was mixed.Stones and plant residues were removed from the samples, which were then kept at 4-5 • C until further analysis.In CA exp and P exp , samples were taken in all three replicate plots of the two cover crop treatments.In CA 1 , CA 2 , P 1 and P 2 samples were taken at three random places in the field (Table 2).

Intact soil cores
In CA exp and P exp , intact soil cores were taken at 3.5-6.5 cm and 13.5-16.5cm depth from the soil trenches (Table 2).Soil cores were taken with a metal ring of 6 cm inner diameter and 3 cm height.The rings were inserted horizontally into the soil profile.In each trench, five cores were taken in each of the two depths.Intact soil cores were kept at 4-5 • C until further analysis.

Field N min samples
Further, soil samples have been taken for N min analysis in March 2019.With soil augers (3 cm inner diameter) 30 cm deep soil cores were sampled, and divided in 10 cm depth intervals.In CA exp and P exp , samples were taken in the cover crop plots with 10-15 soil cores per plot.In CA 1 , CA 2 , P 1 and P 2 15-20 soil cores were taken randomly from the field.Samples from the soil cores were mixed thoroughly for each plot (CA exp , P exp ) or field (CA 1 , CA 2 , P 1 , P 2 ), and kept at 4-5 • C and analysed for N min within 72 h.
Additionally, soil samples have been taken in the field experiment (CA exp , P exp ) at five different time points (Dec 2017, Apr 2018, Sept 2018, Dec 2018and Mar 2019) for N min analysis, prior to the above mentioned experiments.Soil cores were taken with an auger (3 cm inner diameter) to 50 cm depth at intervals of three to five months.In each cover crop treatment plot, 6-10 auger samples were mixed into a bulk sample.These field samples were kept frozen prior to analysis.

Soil profiles
Bulk soil samples from the profiles were sieved to 0.2 cm, stones and organic residues removed, oven dried (105 • C for 24 h) and milled with a ball mill (Retsch 20.741.0004).Soil pH CaCl2 was measured in 0.01 M CaCl 2 (soil:CaCl 2 1:2.5) (Schofield and Taylor, 1955).Soil total C and N concentrations were measured by dry combustion at 1200 • C and gas-chromatographic separation and detection in an elemental analyser (Macro-cube CNS, Elementar, Germany) with acetanilide standards for calibration.
For each intact soil cores, water filled pore space (WFPS) was calculated based on the soil dry weight (bulk density) and the soil water content at the beginning of the incubation, assuming equal particle density of 2.65 g cm − 3 .

Multiple substrate induced respiration
Microbial metabolic capacity along the upper soil profile was measured with a multiple substrate induced respiration (MSIR) method, MicroResp (Campbell et al., 2003;Chapman et al., 2007).Sieved soil (0.2 cm) was partly air dried to obtain a water content of around 17 %, and 38 mg soil was filled into each well of 96-deep-well plates to pre-incubate for seven days at room temperature.A subsample was analysed for gravimetric water content.
CO 2 colorimetric gel traps (1:2 purified agar: indicator, consisting of 12.5 µg ml − 1 Cresol Red, 150 mM KCl and 2.5 mM NaHCO 3 ) were prepared, and calibrated by incubating the traps in air-tight glass containers with specific CO 2 concentrations.

Table 2
Overview of samples and sample depths from different field sites used for the experiments.Bulk samples, incubation rings and soil augers were taken from the two experimental sites (CA exp and P exp ), while from the additional fields (CA 1 , CA 2 , P 1 , P 2 ) only the bulk samples and soil augers were collected.trehalose dehydrate), sugar alcohols (glycerol, myo-inositol), amino acids (L-proline, glycine, L-alanine, L-serine, arginine), and others (urea, titron x-100).A control amended with water only was used to measure basal respiration.Solutions were prepared with 30 mg substrate per gram of soil water.
Dilution of carbon solutions were adjusted after test-incubating extra soil plates with glucose at different dilution levels.The dilution level with the highest respiration was chosen as optimal carbon solution concentration.
Carbon substrates were added to the soil samples, and plates incubated at 25 • C. Plates were tightly sealed with colorimetric gel traps, and their absorbance at 590 nm was measured with a spectrophometeter (Micro-titer plate reader, EON, BioTek, Winooski, Vermont, USA) at 0 and 4 h of incubation.Based on the calibration curve, the absorbance was converted to amount of CO 2 trapped by the colorimetric gel.

Respiration and N mineralization from intact soil cores
Respiration from undisturbed soil samples was measured over a course of four weeks by back-titration of a NaOH solution similar to Müller et al. (2003).Intact soil cores (two replicates per depth, n = 2) were placed into air-tight glass containers with a CO 2 trap with a volume of 1 M NaOH.Bulk densities of the soil cores were slightly increased in CA exp compared to P exp , but this difference was not significant (Table 3).A small beaker of water was added to keep soil water content of the soil samples (top-layer 63.6-65.9% WFPS; bottom layer 59.0-60.2% WFPS, Table 3).Samples were incubated for 28 days at 15 • C. At day 7, 14, 21 and 28, CO 2 trap were taken out and CO 2 respiration was then measured by addition of BaCl 2 and back-titration of each NaOH trap with 0.5 M HCl until the trap reached the inflection point.After each titration, NaOH traps were replaced with fresh ones.The amount of CO 2 respired during each incubation week was calculated from the volume of HCl used for back-titration.Relative CO 2 respiration was calculated as the ratio of cumulatively respired C and C tot content in each soil.
To quantify N mineralization, soil samples were measured for mineral N over the course of the incubation experiment.At the start of the experiment, one soil core of each plot and depth was used for N min analysis.Two soil cores per plot and depth were used for N min analysis, at day 14 and day 28.For N min analysis, soil was removed from the rings and sieved to 0.2 cm.To test, the effect of soil structure disturbance on N mineralization, soil was removed from the soil cores, after day 28.A subsample was sieved with 4 mm and incubated another 14 days, until it was reanalyzed for N min .

Mineral N analysis
Mineral N content was analysed for soil samples form the incubation experiment and from soil augers taken in the field.Samples were analysed for ammonium and nitrate by extraction with 1 M KCL (40 ml KCL: 10 g fresh soil), shaken for 1 h, filtered on 2.5-3 µm filter paperand extracts analysed by flow injection analysis on a FIAstar 500 injection analyser (Foss Analytical, Denmark) after Růžička and Hansen (1988).

Statistical analysis
Soil C and N characteristics and responses were analysed by ANOVA in linear mixed models (Bates et al., 2020).Fixed effects were tillage system and soil depth.All models included a stratification effect, which was defined as the statistical interaction effect between tillage and depth.Stratification ratio was defined as the ratio between average parameter value from the top layer divided by the average parameter value from the deepest layer within the upper 30 cm soil depth (Franzluebbers, 2002).Stratification ratio for time-series were expressed as averages over the entire time-period.Stratification ratio for MSIR was expressed as average over all C sources.
In a subset of CA exp and P exp data, cover crop treatment effect was included as an additive fixed effect in the models.Further, models describing the incubation experiment, had a fixed effect of day of incubation, and a random effect of layer nested in plot.The total concentrations of ammonium as well as ratio between ammonium and nitrate was tested for each field soil mineral N measurement date.
CO 2 respiration rates from the MISR experiment was analysed with the interaction of C-source, tillage system and depth, and included the random effect of plot and well-plate.
Response variables were log-transformed when implied necessary by the distribution of residuals.Post-hoc tests were applied within each soil layer with multi-comparisons at significance level α = 0.05.Linear discriminant analysis (LDA) was performed on the watercorrected CO 2 respiration rate from the MSIR experiment (Venables and Ripley, 2002).Grouping variables for LDA were tillage system and soil depth interactions.LDA was carried out for all carbon substrates, and for substrate groups, based on their chemical structure.
Analysis was executed with R version 4.0.0 (R Core Team, 2020) with the lme4 package (Bates et al., 2020).

Table 3
Average (standard error) bulk density and water filled pore space (WFPS) in the intact soil cores used for incubation from different soil layers of the two tillage systems.
Non-overlapping letters represent significant differences between tillage system and depth combinations with α < 0.05.

Weather conditions
April 2018 to March 2019 was an exceptional dry and warm year compared to the ten-year average (2006)(2007)(2008)(2009)(2010)(2011)(2012)(2013)(2014)(2015) of the same region.The spring barley growing period from April 2018 to July 2018 was on average 2.3 • C warmer and had 176 mm less precipitation than the tenyear average.Comparing the entire year before sampling, from April 2018 to March 2019, the temperature was 1.5 • C higher and precipitation was 500 mm less compared to the 10-year average.Soil samples for incubations were taken in March 2019, which was during a period with relatively high precipitation (Fig. 1).

SOM stratification effect
Long-term CA resulted in a clear stratification of soil organic matter, decreasing with depth, while in regularly ploughed fields a more even distribution of soil carbon and nitrogen was found (Fig. 2, Table 4).The P system showed a decrease in C tot and N tot below 30 cm.The CA system the N tot concentration decreased below 30 cm but not the C tot concentration compared to the 20-30 cm soil layer (Fig. 2).Further, in the CA fields, C-to-N ratio decreased from on average 10.2 in the top 5 cm to 8.8 in the layer 20-30 cm, resulting in an average stratification ratio of 1.16.In the P fields, the C-to-N ratio was more or less constant across the layers, resulting in an average stratification ratio of 0.97 (Table 4).
Significant differences in the extent of stratification was found in all measurements, except in pH CaCl2 (Table 4).Stratification ratio was on average 1.98 in the CA fields, showing a decline with depth.Less stratification was found in the P fields, where stratification ratio was on average 1.25.Stratification ratio of measurements from incubations differed more between systems than stratification ratio of in-field measurements.
Soil acidity was on average about one pH unit lower in the CA fields compared to P fields (Fig. 2), without showing a stratification effect (Table 4).Soil pH CaCl2 showed a field effect in the CA system, where pH CaCl2 in the top 30 cm soil was 5.6, 6.4 and 6.8 in CA exp , CA 1 and CA 2 , respectively (Fig. A.1). Top 30 cm soil pH CaCl2 in the P fields ranged from 6.7 to 7.2.

Basal respiration from intact soil cores
The average daily soil respiration ranged between 2.5 and 7 mg CO 2 -C kg soil − 1 d − 1 (Fig. 3b).There was a time-effect, as soil respiration was highest in the first week of incubation and decreased in the following weeks (Fig. 3).There was no difference in average daily respiration rate between P exp and CA exp in the upper layer (3.5-6.5 cm), but soil respiration was significantly lower in the layer between 13.5 and 16.5 cm in CA exp than in P exp (Fig. 3b).The cover crop treatment decreased average daily C respiration by 22 % compared to no cover crops (Table 5).Soil respiration relative to C org differed with soil depth (Fig. 3c) but not with tillage or cover crop treatment (Table 5).Tillage systems differed slightly in their average WFPS with 61.9 % in CA exp and 62.5 % in P exp , though these differences were not significant.

Multiple substrate induced respiration
The MSIR method showed that all 17 carbon sources resulted in significantly increased respiration, compared to the basal respiration (control, Fig. 4).The MSIR response varied significantly (p < 0.001) between tillage systems and depth but not with cover crop treatment (Tables 4, 5).The soil layer 0-5 cm in CA showed the highest respiration

Table 4
Average stratification ratio (ratio of value for top vs. deepest layer) of depth measurements in CA and P within the upper 30 cm.Stratification effect is expressed as the p-value of the ANOVA interaction term between tillage and soil depth.Measurements from different experiments in the field: a: soil profile (only including the upper 30 cm), b: mineral N profile in the field; and from incubations: c: intact soil core incubation (averaged over day 0 to day 28), d: Multiple Substrate Induced Respiration (MSIR) (water corrected respiration, averaged over all C sources).rates of all soil samples, though the magnitude varied with carbon substrate (Fig. 4).This was also evident in the linear discriminant analysis (Fig. 5), where the 0-5 cm CA topsoil separated strongly from the other soils.The lowest MSIR rates were found in the 5-10 cm layer of the P soil, with the 10-20 cm layer showing slightly higher respiration activity.At 20 cm soil depth were also more or less intact straw residues observed, which had been incorporated in the autumn six months earlier.
There was a highly significant interaction between tillage system and C-source (p < 0.001).Loadings of the LDA showed stronger impact of some C-sources than others on substrate-induced respiration (Fig. 5): Myo-inositol, trehalose and raffinose showed a higher impact for 0-5 cm layer in CA; xylose, maltose, glycin and L-arabinose showed a stronger effect on the respiration from the ploughed soil.However, these differences in loadings could not be ascribed to chemical groups, categorized as mono-, di-, poly-saccharides, sugar alcohols or amino acids.When performing LDA analysis for each chemical group separately, linear discriminant 1 and 2 did not differ compared to the complete dataset (not shown).

Net N mineralization from intact soil cores
Soil net N mineralization in the intact soil core experiment (Fig. 6, Tab.A.2) showed a similar pattern to the CO 2 respiration patterns (Fig. 3).In all samples but the 13.5-16.5cm layer of CA, nitrate and total mineral nitrogen concentrations increased strongly in the first two weeks, indicating soil net N mineralization activity.While at the start of incubation, nitrate concentrations were not different between tillage systems and soil depths, the CA 3.5-6.5 cm layer showed increased ammonium concentrations, which stayed elevated (significantly above the CA 13.5-16.5cm layer, and the two P layers) two weeks into the incubation (Fig. 6).Net N mineralization measured in the first two weeks of the incubation experiment was larger than in the second half of the experiment.Average daily mineralization rates in the 3.5-6.5 cm soil layer of the first two weeks of incubation were 0.44 mg N kg − 1 soil d − 1 and 0.16 mg N kg − 1 soil d − 1 in CA and P, respectively.This corresponded to a mineralization of 0.30 % and 0.10 % of total soil N within the first two weeks.In the 13.5-16.5cm layer, CA showed no changes in the first two weeks (− 0.01 mg N kg − 1 soil d − 1 ) while P showed twice as high net N mineralization compared to the 3.5-6.5 cm layer (0.30 mg N kg soil − 1 d − 1 ), which corresponded to 0.19 % of total soil N in that layer.In the second two weeks of the incubation experiment, both tillage systems and both soil layers showed decreased N concentrations, indicating net N immobilization: average daily N mineralization rates were − 0.48, − 0.02, − 0.19 and − 0.22 mg N kg − 1 soil d − 1 in CA 3.5-6.5 and 13.5-16.5cm layer, and P 3.5-6.5 and 13.5-16.5cm layer respectively.The total change of mineral N across the four week incubation was negative in 3 cases, only in the P 13-16.5 layer the immobilization rate in the second week did not exceed the mineralization rate in the first week.
Net N mineralization was affected by cover crop treatment, as N immobilization was significantly (p < 0.05) lower in the cover crop treatment in the second half of the incubation experiment (Table 5).Across the total length of the experiment, cover crops resulted in total net N mineralization, while the absence of cover crops resulted in total net N immobilization.Tillage system affected the time-dynamics of N mineralization: While the ammonium content stayed high in the CA 3.5-6.5 layer after two weeks incubation, ammonium content was under the detection limit in the P soil (Fig. 6).After four weeks, ammonium was also in CA soil not detectable anymore (Fig. 6).Whether nitrate was lost by gaseous emissions or immobilized was not measured.After four weeks of incubation, there was no difference in total mineral nitrogen between tillage systems (Fig. 6).
The disturbance of soil structure by sieving through a 4 mm sieve at day 28 increased N mineralization in all soils, with the same effect for the two tillage systems (Fig. 6).The increase was only detectable in the nitrate fraction of mineral nitrogen, indicating fast nitrification of any mineralized N.

Mineral N dynamics in the field experiments
The relative proportions of mineral N constituents (ammonium and nitrate) sampled in the field over time differed between the two tillage systems.Independently from time-point, ammonium-N concentrations  were significantly (p < 0.001) higher in CA exp soil compared to P exp soil (Fig. 7).In addition, the NH 4 : NO 3 ratio was significantly higher in CA exp soil compared to P exp soil at all time-points but December 2018 (Fig. 7).Across all time-points, NH 4 :NO 3 ratio was 1.7 in CA exp and 0.5 in P exp .Significantly higher NH 4 concentrations were also predominantly found in March 2019 for all three CA fields in the 0-10 cm soil layer, while in 10-30 layers the differences to P were not significant; however the NH 4 : NO 3 ratio was significantly higher in CA for all three depths (Fig. 8).There was no effect of cover crop treatment on NH 4 or NO 3 concentration in March 2019 (Table 5).

No-till increases stratification
In this study we found that vertical distribution of SOM differed between P and CA fields, with CA showing higher degrees of SOM stratification.In CA fields, residue input and the absence of soil disturbance resulted in an increased average stratification ratio of 1.74 in CA compared to 1.05 in P in the C and N pools.Increased stratification ratios in CA corresponds well to Franzluebbers (2002) who found average SOM pool stratification to be 2.5 in no-till and 1.4 when ploughed.Similar ranges were found in other no-till studies, where SOM pool stratification ranged from 1.41 to 2.31 in no-till and 1.01-1.48 in the ploughed comparison (Frey et al., 1999;Mathew et al., 2012;Sun et al., 2018).These findings support our first hypothesis, which seem to be a consistent characteristic to describe differences in tillage systems.
Tillage system differences in stratification ratios were higher in the incubation measurements, compared to in-field measurements (Table 4).The incubation measurements were more closely linked to SOM turnover dynamics, than the in-field measurements, which show slower changing pools of organic matter.Franzluebers (2002) argues, that measures of turnover dynamics are more sensitive to soil management than measures of total SOM pools.This is supported by the higher metabolic capacity we found in the topsoil of CA compared to P (Figs. 4  and 5).This finding supports the second part of our hypothesis.

Methodology
We measured C mineralization in two different incubation experiments (intact soil core and MSIR), which varied in C application and in  measured C-respiration rates.The differences between incubation experiments may have resulted from varying degrees of soil disturbance, substrate input and incubation time and temperature.However, compared to the first week of the intact soil core incubation, the MSIR basal respiration rate (control) was 1.6-2.4times higher.Based on 10 ℃ temperature differences between core incubation and MSIR experiments, this corresponds to a Q 10 of around 2, which is within expectable ranges (Meyer et al., 2018).
High CO 2 respiration in the first week of the intact soil core incubation indicate that additional oxygen supply at the core surface and disturbance from sampling resulted in mineralization of labile SOM compounds.Gómez-Rey et al. ( 2012) relate the higher turnover rates in the first interval of their incubation to a higher availability of labile SOM substrates.Although we aimed to test C mineralization in undisturbed soil samples, it is likely that our sampling method did in fact disturb the soil environment, altering the chemical and physical exposure of SOM to the soil microbial community.These considerations indicate the methodological difficulties of studying no-till soils with commonly used laboratory analysis, which is often inducing soil disturbance.Since soil structure is one of the major differences between CA and P tillage systems, we need to find new ways to analysis soil processes.In-situ analysis and/or stable isotope studies as for example shown by (Laine et al., 2017) seem a promising strategy.

CA increased microbial metabolic capacity
The microbial community in CA topsoil showed a higher metabolic capacity, shown by the increased MSIR across many C-sources (Figs. 4  and 5).The increased microbial metabolic capacity can be related to higher microbial abundance, as shown by a study by Sun et al. (2016), who found that no-till increased both microbial abundance and microbial activity in the top 5 cm soil.This finding is similar to Kandeler et al. (1999), who found a significant difference in microbial respiration between tillage systems and an increased stratification under reduced tillage practises.The increase in microbial activity is likely related to the higher concentration of residue inputs to the soil surface, rather than residues being distributed into the entire plough-horizon (Sun et al., 2016).
There are indications that the tillage system affected microbial functional diversity.Although C-source did not show significant interactions with tillage on MSIR, the linear discriminant analysis showed a separation of MSIR based on tillage and depth (Fig. 5).This indicates, that the microbial community is differently adapted to respire certain carbon sources.Higher microbial functional diversity could imply that the microbial community in CA is well adapted to decompose the topsoil applied residues.This applies further to the microbes ability to degrade more complex residues that were not mechanically distributed and disturbed due to the no-till element of CA.The changes in abiotic conditions caused by no-till, especially in the soil surface, are likely to not only affect the microbial function (Domeignoz-Horta et al., 2020), but more generally also to allow for more diverse range of microbes to exist in the soil (Holland and Coleman, 1987;Mangalassery et al., 2015;Sun et al., 2016).

CA practices effects on C mineralization
Carbon mineralization from incubation experiments showed strong stratification effects in CA, but not in P (Figs. 3a,4 and 5).Overall, the C mineralization rates from incubated intact soil cores are in the same range as other basal respiration observations (e.g.Grandy et al., 2013).The high C mineralization in CA topsoil seems to have been affected by the higher amount of soil C in that layer, since soil C pools were also stratified with depth (Fig. 3c).Basal C respiration has previously been related to soil microbial biomass and SOM content (Alvarez et al., 1995;Kainiemi et al., 2015;McDaniel et al., 2014).Opposite to Alvarez et al. (1995), we did not see a difference in relative C mineralization between the tillage systems in the top-layer.
Although cover crops have been discussed as crucial for sustainable CA farming practises in temperate climates (Büchi et al., 2018), the effect of cover crops half a year after termination on measured SOM turnover dynamics was low (Table 5).The low cover crop biomass and the short time since the initiation of the cover crop treatment (2 years) may have resulted in the small cover crop effect, compared to the longer term system difference.It should, however, be noted that cover crop effects may play a crucial role in distinguishing C and N turnover effects in CA systems compared to P systems.

Nitrogen mineralization
Nitrogen mineralization rates were increased in the top layer under CA management (Fig. 6, Table 4) supporting the third part of our hypothesis.Similar results were found by Van Den Bossche et al. (2009) who found higher N mineralization from SOM in no-till compared to ploughed soil.The authors relate the higher N mineralization to the younger organic matter input at the CA top-layer (Van Den Bossche et al., 2009), which has been reported to increase initial N mineralization (Oorts et al., 2006).Further, the high N mineralization in the first half of the incubation could be due to soil sampling related soil disturbance which might have increased O2 availability and hence increased microbial activity.This effect was also measured in C respiration (Fig. 3).We did, however, not find an effect of tillage system on N mineralization after soil disturbance with 2 mm sieve (Fig. 6), similar to Oorts et al. (2006).This could indicate that the physical protected labile SOM has been turned over in the initial phase of the experiment.In the study by (Oorts et al., 2006), soil disturbance only had an effect on the deeper layer (5-20 cm) of CA soil, when sieved with 50 µm.This suggests that physical protection of SOM is limited in CA, and does not affect top-layer dynamics.
The initial high N mineralization rates were negated in the second half of the incubation, with almost equally high immobilization rates in the CA top-layer (Fig. 6).Similar results have been observed before (Laine et al., 2017;Liu et al., 2017) and is likely due to the wider C:N ratio of fresh residues in top-layer (Laine et al., 2017, Fig. 2).It is also plausible that some mineralized N was lost via denitrification during the incubation.
While C-mineralization did not show differences relative to soil C, relative soil N mineralization was three times as high in CA top-layer compared to P top-layer in the first two weeks of the experiment.These different patterns in C and N mineralization between tillage systems have also been found by Kandeler et al. (1999).In their study, N mineralization showed larger differences between tillage systems than C mineralization, especially in the 0-10 cm top-layer.One of the reasons for the different turnover rates could be differences in carbon use efficiency in the CA top-layer caused by microbial functional diversity (Domeignoz-Horta et al., 2020;Mangalassery et al., 2015).Changes towards a more fungal dominated microbiome have earlier been reported under no-till management (Frey et al., 1999;Helgason et al., 2009;Sun et al., 2016).Findings from our MSIR incubation support this hypothesis, since C-sources associated to fungal metabolome have separated the CA top-layer from the other soils (Fig. 5).These changes in the microbiome towards more fungi could support a higher C use efficiency (Six et al., 2006).
In general, N mineralization patterns seem to be largely affected by environment, season, crop rotation and incubation time.Seasonal changes in microbial activity are driven by residue input, which are more or less equal between tillage systems in the main growing season.Tillage systems are, however, likely to cause temporal and spatial differences in N mineralization patterns due to differences in cover crop growing period and cover crop termination method.Further, the permanent residue cover in no-till causes not only abiotic changes to the soil environment, but also segregate the residues from the soil and thereby alter many aspects of the decomposition processes.Therefore, a generalization of experimental findings of the impact of CA practices on N mineralization should be taken with care, taking into account these different parameters.
Apart from the generally higher mineralization rates, CA also showed indications of reduced nitrification, since the NH 4 :NO 3 ratio was significantly increased in CA (Figs. 6-8).These differences could be caused by soil structural difference between the tillage systems: While Van Den Bossche et al. ( 2009) did not find an effect of no-till on NH 4 : NO 3 ratio in disturbed soil samples, Laine et al. (2017) found decreased ammonium oxidation in no-till soils in an in-situ labelling experiment.Increased NH 4 :NO 3 ratio could be due to differences in pH, temperature and oxygen between the tillage systems: pH CaCl2 was lower in CA than in P, in particular CA exp was lower (Figs.2, A.1), which seems to have been due to previous field management, since the pH differences are apparent also at depth below 30 cm (Fig. 2).Soil acidity reduces nitrification significantly (Anthonisen et al., 1976).Therefore, it seems likely that at least under CA exp , nitrification was somewhat inhibited by pH.
However, differences in pH can only partly explain the increased NH 4 :NO 3 ratio.Averaged over the top 50 cm soil, pH CaCl2 was at 5.7 in CA exp , while pH CaCl2 was on average 0.9 pH CaCl2 higher in CA 1 and CA 2 , causing the large standard error of pH CaCl2 (Fig. 2).Although CA 1 and CA 2 showed a pH CaCl2 , similar to the P fields, they still showed increased NH 4 :NO 3 ratio (Figs. 8, A.1).
Temperature differences could further have caused nitrification inhibition, as the insolating effect of the residue layer will lower temperature in spring (Oorts et al., 2007;Soane et al., 2012).At low soil temperatures (below 5 ℃), nitrification is reduced (Russell et al., 2007;Taylor et al., 2021), while mineralization is still active at 3.5 ℃ (Dalias et al., 2002).However, the difference in NH 4 :NO 3 ratio was apparent across the entire season, indicating that the temperature effect cannot be the dominating explanation.
The third reason for nitrification inhibition may be lower oxygen availability in CA top-layer due to higher oxygen demand and less airfilled porosity.The increased amount of SOM and microbial functional diversity in the top-layer of CA can lead to higher decomposition rates.In CA, residues are "incorporated" biomechanically in a heterogeneous pattern.Microbes consume high amounts of oxygen in localized "hotspots", which in combination with more compacted topsoil (Blanco--Canqui and Ruis, 2018; Soane et al., 2012) may create local anoxic conditions.The increased water holding capacity and lower pore space in CA decrease air-filled porosity and lower oxygen availability in CA soil.In fact, soil cores from the CA system showed a tendency for higher WFPS compared to P, although that difference was not significant (Table 3).Since nitrification is dependent on oxygen availability, low oxygen or partly anoxic conditions will reduce nitrification rate.In fact, Brewer et al. (2018) found increased anoxic conditions in CA topsoil, which caused decreased nitrification.Using intact soil core samples, Brewer et al. (2018) found strong indications for local anoxic conditions, altering the chemical environment for microbial activity.As nitrification is strongly oxygen (O 2 ) affected (Norton and Stark, 2011), mineral N dynamics in a disturbed soil sample could be significantly altered (Laine et al., 2017).
Reduced nitrification rates under CA may imply that CA is less prone to N 2 O emissions and NO 3 leaching than argued earlier (Laine et al., 2017;Liu et al., 2017).

Perspectives
It is difficult to study long-term tillage effects under realistic field conditions.Relevant comparisons are difficult to manage in plot experiments, and they need to run for several years before reliable results on the long-term effect on SOM are visible.Therefore case-farm studies appear as an appropriate tool to integrate and test knowledge from more controlled research environments.As tillage effects in our study were not investigated in a true comparative field experiment, the total amount of SOM was affected by the long-term field history, especially below the plough-layer (Fig. 2).Further, we cannot exclude the possibility that part of the difference in microbial activity and metabolic capacity might have been caused by previous differences in cover crop application of the two tillage systems.In this context, we deem applying stratification ratios more reliable for comparing tillage effects on SOM pools (Franzluebbers, 2002).Additionally, a higher resolution of soil depth effects in CA, especially in the topsoil and residue layer, is important for better understanding of stratification effects on SOM turnover.
Isotopic tracer studies seem appropriate to study short term effects of management tools and could offer a better understanding of the altered dynamics.

Conclusion
Average stratification ratios of SOM pools and turnover dynamics were 60 % higher in CA compared to P. We advocate for integrating stratification ratios in the analysis of tillage effects on SOM dynamics, both in science, and in practise.
We did not find differences in C respiration activity between the tillage systems, but N mineralization patterns were significantly altered under CA topsoil.Specifically, ammonium to nitrate ratio was increased under CA across seasons, specifically in the upper 10 cm.This implies abiotic or biotic changes in the soil environment, causing nitrification inhibition in CA.
Two years of cover crops did not affect C and N mineralization patterns significantly in either tillage system.
While SOM pools differed with depth, we did not find indications for alternated SOM turnover dynamics that could lead to increased SOM sequestration under CA farming in temperate climates.We did, however, find increased microbial metabolic capacity in CA topsoil.

Declaration of Competing Interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Fig. 1 .
Fig. 1.Total daily precipitation (mm) and daily average temperature ( • C) from March 2018 to March 2019 at the Flakkebjerg field meteorological station, located 18 km south-west of the experimental fields.

Fig. 3 .Fig. 4 .
Fig. 3. Carbon mineralization as weekly CO 2 respiration during incubation of intact soil cores depending on tillage system and soil depth (a).Average CO 2 respiration over 28 days (n = 48), with absolute respiration of soil core (b) and relative CO 2 -respiration to C tot in soil core (c).Non-overlapping letters represent significant differences between tillage system and depth combinations with α < 0.05.Note: Tillage systems are Conservation agriculture (CA) and Ploughed (P) systems.

Fig. 5 .
Fig. 5. Linear discriminant analysis (LDA) of the first two discriminant functions clustering the community responses to multiple substrate induced respiration.Grouped by tillage system and depth.Respiration corrected for basal respiration.Ellipses mark the 95 % confident intervals for each group (n = 12) (left).Arrows mark loadings of parameters (right).Note: Tillage systems are Conservation agriculture (CA) and Ploughed (P) systems.

Fig. 8 .
Fig. 8. Average soil mineral nitrogen in the topsoil by 10 cm depth intervals from three fields per tillage system.Samples were taken in March 2019, in standing winter wheat crop before spring fertilization.Asterisks represent level of significance for difference between tillage system within each layer (p < 0.05(*), p < 0.01 (**), p < 0.001 (***)).Note: Tillage systems are Conservation agriculture (CA) and Ploughed (P) systems.

Fig. A. 1 .
Fig. A.1.Field effect on pH CaCl2 (a) and ammonium to nitrate ratio (b) depth distribution.Note: Tillage systems are Conservation agriculture (CA) and Ploughed (P) systems.
Agronomic practices and experimental activities on the experimental fields from start of the experiment to soil sampling.Anova test results of the development of N-pools during mineralization experiment.
Note: Tillage systems are Conservation agriculture (CA) and Ploughed (P) systems.