Effects of Environmental Drivers and Agricultural Management on Soil CO 2 and N 2 O Emissions

: Understanding the roles of natural drivers and anthropogenic activities in greenhouse gas (GHG) emissions of arable ﬁelds is crucial for adopting the most appropriate agricultural management. This study investigated the effect of two tillage treatments of mouldboard ploughing (MP) and no-tillage (NT), and the environmental factors (soil water content and temperature, carbon content and nitrogen forms) on soil carbon dioxide (CO 2 ) and nitrous oxide (N 2 O) emissions. The research was conducted on chernozem soil under winter wheat cultivation. Besides ﬁeld moni-toring, several laboratory experiments took place to examine the effects of environmental drivers and fertilization management on soil GHG emissions. We observed no signiﬁcant difference between the CO 2 emission of MP and NT during a full year period. Nevertheless, signiﬁcant differences were found in the sub-periods (more particularly during vegetation and then after harvest). NT had higher CO 2 emission than MP in all laboratory experiments ( p < 0.001) and in the after harvest period of the ﬁeld trial, measured on bare soil ( p < 0.0001). NT had signiﬁcantly higher N 2 O emission both under laboratory ( p < 0.0001) and ﬁeld conditions ( p < 0.0081). Different fertilization showed no distinguishable effect on N 2 O emission in the laboratory. This study conﬁrms that N 2 O emission of the arable ﬁeld depended more on soil water content than soil temperature, and vice-versa for CO 2 emission.


Introduction
Although the importance of carbon dioxide (CO 2 ) in climate change is well known, it remains in the focus of research interest due to its role in the biosphere and the global ecosystem. More recently however the function of nitrous oxide (N 2 O) emission appears as a new challenge despite the fact that N 2 O has minor concentration in the atmosphere compared to CO 2. Nonetheless, it is a more potent greenhouse gas (GHG) having 310 times higher global warming potential than CO 2 [1]. Moreover, N 2 O also contributes to the processes of ozone depletion [2]. CO 2 has many natural (respiration of biosphere, soils and wetlands, volcanic eruptions) and anthropogenic sources (industry, fossil fuel combustion, transportation, agriculture), as well as natural sinks (forests and oceans) [3]. N 2 O emission might originate from natural sources such as different microbial processes of soils and wetlands, lightning in the troposphere or anthropogenic sources like industry and agriculture, fuel and biomass combustion. There are a number of natural sinks of N 2 O, like photochemical processes in the stratosphere [3].
Soil CO 2 emission has several natural drivers, such as soil temperature (Ts), soil water content (SWC), soil carbon content (TC) and biosphere (e.g., root respiration, micro, mezzo and macro fauna) [4]. N 2 O emission of soils is mainly governed by soil climate like SWC, Ts, soil chemistry and microbial processes such as nitrification or denitrification.

Introduction of Field Trial
The study site is a long-term (>18 years) tillage experiment at Szent István University's Józsefmajor Experimental and Training Farm, Heves County, Hungary (47 41 31.7 N 19 36 36.1 E, 110 m a.s.l). The soil is Endocalcic loamic chernozem [44], which is a dominant soil type in the region of the study. The climate is typical continental type with 560 mm average annual precipitation and 10.3 • C mean annual temperature. The experiment was set up in 2002 and consists of six different tillage treatments (mouldboard ploughing-(MP), notillage (NT), disk tillage, loosening, shallow and deep cultivation), in a randomized design. For the purpose of this study, we investigated the MP and NT treatments. Each tillage treatment had an area of 10 × 105 m in four replicates. In the experiment, crop rotation and adaptive fertilization were used annually. After harvest plant residuals were left on the surface as mulch, and straw was mixed with the soil during tillage operations. The sown crop in season 2018/2019 was winter wheat (Triticum aestivum), and the results of emission measurements of the year 2019 are introduced in this article. The management events and the applied N fertilization doses during the last 5 years of the experiment are summarized in Table 1.

Greenhouse Gas (GHG) Emission Measurements-Laboratory Experiments
To investigate the underlying processes of GHG emissions in a more controlled environment, we set up the following three different laboratory experiments. Detailed information is presented in Tables 2 and 3. (i) In the Tillage experiment we examined the effect of the two tillage methods (MP, NT) on GHG emission under constant air temperature. For the investigations, we collected samples from MP and NT treatments and one fertilizer dose of 80 kg ha −1 N (MP80, NT80) was applied besides non-fertilized control (MP0, NT0). Calcium ammonium nitrate (CAN) was used as fertilizer. SWC was constant in the first part of the experiment and then columns were left to air dry, which enabled us to study the SWC-dependency of GHG emission. (ii) In the Fertilizer experiment we examined the effect of different fertilizer doses of 80 kg ha −1 N (MP80, NT80) and 160 kg ha −1 N (MP160, NT160) on the GHG emissions of MP and NT under constant air temperature. In the Fertilizer experiment, the initial field SWC was maintained in the first part of the experiments, then a higher SWC was used in the second part of this investigation. The laboratory experiments were implemented by the following methodology: we collected undisturbed soil cores (d = 10.5 cm, h = 10.0 cm) from the upper 10 cm of MP and NT treatments. We inserted plastic tubes (d = 10.5 cm, h = 20.0 cm) into the soil and sealed them in the bottom after sampling. During the laboratory measurements, the top of these columns was also sealed to create an incubation chamber, which enabled us to attach these sealed chambers to a cavity ring down spectrometer (CRDS) multi-gas analyser for measuring CO 2 and N 2 O emissions. Each experiment started after a 2-day long pre-incubation period. We maintained SWC on the initial field value for the first part of the experiment until a steady state GHG emission occurred. We did not change the initial SWC of the samples, therefore SWC was higher in NT than in MP treatment in all laboratory experiments imitating field conditions ( Table 2).
The air temperature was constant (21 or 27 • C depending on the experiment) during the investigations (Table 2) except for the Ts dependency experiment, where investigations were performed at 4, 10, 20 and 30 • C (Table 3).
In all laboratory experiments, we measured GHG emissions daily during the first week, then two times weekly during the rest of the experimental duration ( Figure 1).

GHG Emission Measurements-Field Experiment
Regular CO2 emission measurements were performed in seven replicates from 2013 in both MP and NT treatments [46]. In addition, we started N2O measurements in 2019. For the present study we selected the whole year dataset for CO2 including vegetation and

GHG Emission Measurements-Field Experiment
Regular CO 2 emission measurements were performed in seven replicates from 2013 in both MP and NT treatments [46]. In addition, we started N 2 O measurements in 2019. For the present study we selected the whole year dataset for CO 2 including vegetation and after-harvest period and a half-year dataset for N 2 O emission including the after-harvest period (Table 4). We implemented CO 2 measurements weekly and N 2 O measurements on a weekly to monthly basis with 7 (CO 2 ) and 3 (N 2 O) replicates per treatment. To determine GHG emissions, we applied two in situ gas analysers, an infrared gas analyser (IR) for CO 2 emission and a CRDS multi gas analyser for N 2 O emission. The measurements by the different instruments were taken at the same time, in close vicinity. The GHG measurements were based on the dynamic chamber method [47][48][49], where chambers were not automatic but were placed manually. We set the incubation time of the chambers to 2 and 15 min for CO 2 and N 2 O, respectively. GHG fluxes were calculated by quadratic (CO 2 ) and linear (N 2 O) estimations. To investigate the environmental drivers of GHG emissions, we monitored SWC, Ts at every GHG measurement points. Sampling strategy and basic environmental parameters are presented in Table 4.

Instrumentation
We used an EGM-5 (PPSYSTEMS, U.S.) in situ IR gas-analyser to determine CO 2 emissions and a PICARRO G2508 CRDS multi-gas analyser (PICARRO, U.S.) to determine N 2 O emission of the field trial. We also used the PICARRO G2508 to determine CO 2 and N 2 O emissions during all laboratory experiments.
The measurement accuracy of EGM-5 was <1% of reference gas and the measurement range was 0-5000 ppm. In the case of the PICARRO G2508, the accuracy for raw nonsmoothed data was ±600 ppb for CO 2 and ±25 ppb for N 2 O. The guaranteed range was 380-5000 ppm for CO 2 and 0.3-200 ppm for N 2 O.
In the field experiment we also measured the volumetric SWC of the upper 10 cm of each GHG measurement points by CAMPBELL SCIENTIFIC HYDROSENSE II SWC probe with an accuracy of 3% and a resolution of <0.05%. We determined Ts with PPSYSTEMS STP-2 soil temperature probe, which was attached to the EGM5 analyser with an accuracy of ±0.3 • C at 25 • C. In the laboratory experiments, we calculated volumetric SWC from gravimetric SWC of each sample at the end of the experiments.
The leaf area index (LAI) of winter wheat were measured by an ACCUPAR LP80 ceptometer (METERGROUP, U.S.) during ear emergence and flowering phenological phases, which refer to GS59 and GS61 in the Zadoks decimal code system, respectively [50,51].

Soil Chemical and Physical Analyses
We collected soil samples from the upper 10 cm at all N 2 O emission measurement times to determine the total nitrogen (N tot ), total carbon (TC), NH 4 + -N, and NO 3 − -N contents of the samples by wet chemical analyses [52]. We determined the same chemical parameters from the soil columns used in the laboratory experiments by the same methods at the end of the experiments. Sample numbers of the field experiment were 19 and 26 in total (over the after-harvest period) for MP and NT, respectively. Sample numbers of laboratory experiments' chemical parameters can be seen in Tables 2 and 3. For the physical analysis, we collected three undisturbed soil cores (100 cm 3 ) from the upper 5-10 cm layer of MP and NT. We dried the soil cores at 105 • C for 24 h, and we measured the dry weight of the samples to calculate bulk density [53].

Data Handling and Statistical Analyses
Negative CO 2 flux data of field measurements provided by the EGM analyser were omitted because we considered these occasions to be caused by chamber failures. Negative fluxes of CO 2 are a matter of debate in the scientific discussion. Negative CO 2 fluxes are more probable in specific areas like deserts [54], non-vegetated arid lands [55], highly alkaline soils under dry conditions [56], coastal peatlands [57], or estuary wetlands [58] than in the case of continental type arable field during summer and early fall. We used the 5 min smoothed dataset of N 2 O concentrations provided by the PICARRO analyser in order to eliminate the high variance of the raw concentration curve. Negative N 2 O fluxes were considered as zero fluxes. Arable fields can be a sink of N 2 O from the atmosphere under certain circumstances [59], but in our case these negative values converged to zero well, and they were below measurement uncertainty.
For the basic statistical analysis, we used GraphPad Prism version 8.4.3 for Windows, (GraphPad Software, San Diego, California U.S.). We determined the distribution of the datasets with a D'Agostino and Pearson test. We applied all data from the replicates instead of daily mean values of GHG emission, SWC and chemical comparison of laboratory and field investigations. Based on the dataset distribution we used different statistical analyses, which are presented in Table 5. We used Pearson's correlation to find relationships of GHG emission and environmental parameters. We interpreted correlations according to Schober et al. [60].

CO 2 Emission
In general, under laboratory conditions, NT had higher mean CO 2 emissions than MP in all experiments (Table 6). In the Tillage experiment the difference between CO 2 Agronomy 2021, 11, 54 8 of 20 emissions of MP0 and NT0 was significant (p < 0.0001) and it was significant between MP80 and NT80 (p < 0.0001) as well. Fertilization significantly reduced CO 2 emission both in the MP80 (p = 0.0245) and NT80 treatments (p < 0.0067) compared to non-fertilized controls (MP0 and NT0, respectively).
Similar to the findings of the Tillage experiments, NT0, NT80, NT160 had significantly higher mean CO 2 emissions than MP0, MP80, MP160 (p < 0.0001) in the Fertilizer experiment. However, none of the fertilizer doses had significant effect on the CO 2 emission, in MP80 (p > 0.9999); MP 160 (p = 0.1747) compared to MP0; or in NT80 (p = 0.1083), NT160 (p > 0.9999) compared to NT0, but there was a significant difference between NT80 and NT160 (p = 0.0009) Mean CO 2 emissions of NT varied in all laboratory experiments, while the mean CO 2 emissions of MP had only minor variability, thus the emission of MP was more stable ( Table 6).
Treatments had a decreasing trend of CO 2 emission over time in the laboratory experiments, which was steeper in the case of NT compared to MP, especially in the Fertilizer experiment ( Figure 1).

N 2 O Emission
We found similar tendencies in the case of N 2 O emission as for the CO 2 data, the mean N 2 O emission of NT was higher than of MP in all laboratory experiments (Table 7). In the Tillage experiment, N 2 O emission was significantly higher (p < 0.0001) in NT compared to MP treatments regardless of the fertilization doses ( Table 7). The 80 kg ha −1 fertilizer dose had a significant effect (p = 0.0028) on N 2 O emission of MP80 compared to MP0, however, it had a non-significant effect (p > 0.9999) in NT80 compared to NT0 (Table 7).
In the Fertilizer experiment NT0, NT80, NT160 had a significantly higher (p < 0.0001) N 2 O emission than MP0, MP80, and MP160 regardless of fertilization doses. The N 2 O emissions had a nonlinear response to the different fertilizer rates in NT, where NT160 showed the highest mean N 2 O emission. There were significant differences between the mean N 2 O emission of NT0 and NT80 (p = 0.0065) and between NT0 and NT160 (p < 0.0001). Nevertheless, the difference was not significant between NT80 and NT160 (p > 0.9999) ( Table 7). There was no significant effect of fertilizer additions on N 2 O emission for MP treatments (p > 0.0523) ( Table 7).
The temporal course of the N 2 O emissions showed similarities both in the Tillage and Fertilizer experiments (Figure 2). After fertilization, N 2 O emissions peaked in the first week, then there were declining trends in all experiments while the SWC and Ts were constant. The N 2 O emissions reached a steady-state after one week (Figure 2), then decreased during the drying out period (Figure 2a). After N2O emission got constant, we increased the SWC of the treatments from 33.5 ± 1.0 to 40.9 ± 2.0 (avg. MP) and from 40.3 ± 0.6 to 47.7 ± 0.9 (avg. NT) during the second part of the Fertilizer experiment (Figure 2b). The N2O emission was significantly higher under elevated SWC compared to initial SWC in all treatments (p < 0.0103). Still, elevated N2O emissions of MP did not reach the highest emission of any measured N2O values in the NT treatments under initial SWC, even when the SWC of MP was on the same level (Figure 2b).
CO2 emissions of MP and NT treatments highly depended on Ts in laboratory experiments, although the correlations were not significant in all cases (Table 8). Correlation between CO2 emission and Ts was strong in all cases but not always significant.
The N2O emissions showed moderate to strong dependency on Ts with significance After N 2 O emission got constant, we increased the SWC of the treatments from 33.5 ± 1.0 to 40.9 ± 2.0 (avg. MP) and from 40.3 ± 0.6 to 47.7 ± 0.9 (avg. NT) during the second part of the Fertilizer experiment (Figure 2b). The N 2 O emission was significantly higher under elevated SWC compared to initial SWC in all treatments (p < 0.0103). Still, elevated N 2 O emissions of MP did not reach the highest emission of any measured N 2 O values in the NT treatments under initial SWC, even when the SWC of MP was on the same level (Figure 2b).

Environmental Governing Factors of GHG Emissions in Laboratory Experiments
We set a T dependency experiment under initial (field) SWC (33.9 ± 0.4 % in MP0, 39.6 ± 0.3 % in NT0), high SWC (43.3 ± 0.4 % in MP0 and 46.7 ± 0.4 in NT0), and high SWC combined with fertilization (44.5 ± 0.5 % in MP80 and 47.6 ± 0.6 % in NT80; Table 3). CO 2 emissions of MP and NT treatments highly depended on Ts in laboratory experiments, although the correlations were not significant in all cases (Table 8). Correlation between CO 2 emission and Ts was strong in all cases but not always significant. Table 8. Correlation and significance * of Ts and greenhouse gas (GHG) emissions; r 2 and p-values are indicated (Sample numbers are presented in Table 3).

Treatment
Ts The N 2 O emissions showed moderate to strong dependency on Ts with significance only in two cases ( Table 8).
The SWC of the MP and NT treatments were also different during all laboratory experiments (Tables 2 and 3), due to the initial field conditions of the treatments. In the Tillage experiment NT had a significantly higher SWC than MP regardless of fertilization (p < 0.0001 both in 0 and 80 kg ha −1 ).
We could observe similar trends in the Fertilizer experiment as NT treatments (NT0, NT80, NT160) had significantly higher SWC than MP treatments (MP0, MP80, MP160) regardless of fertilization (p < 0.0001 in all treatments). In contrast with tillage, fertilization had no significant effect on SWC neither in MP nor in NT in any experiments (p > 0.9999) ( Table 2).
We examined the SWC dependency of GHG emissions in the second, shorter drying out period, after reaching the steady-state GHG emission phase in the Tillage experiments. The SWC dependency of CO 2 emission was strong in MP0 (r 2 = 0.89, p = 0.0589) and very strong and significant in MP80 (r 2 = 0.99, p = 0.0020). Both control and fertilized samples of NT had moderate correlation with SWC (r 2 = 0.60, p = 0.2277 in NT0; r 2 = 0.58, p = 0.2360 in NT80).
Investigating the SWC dependency of N 2 O emission in the Tillage experiment, MP80 and NT0 samples showed very strong and significant SWC dependency (r 2 = 0.93, p = 0.0344 and r 2 = 0.95, p = 0.0278, respectively). The N 2 O emission of NT80 had a strong, but not significant correlation with SWC (r 2 = 0.74, p = 0.1418) and MP0 had weak N 2 O emission during the whole measurement period, thus SWC dependency could not be interpreted (r 2 = 0.03, p = 0.8354).

Soil Chemical Parameters of Laboratory Experiments
The main soil chemical parameters are provided in Table 9 for all laboratory experiments. Regardless of fertilization rate, all NT treatments had significantly higher N tot content as compared to all MP treatments in the Tillage (p < 0.0137), Fertilizer (p < 0.0001) and Temperature dependency (p < 0.0079) experiments. In the Tillage experiment NT0 and NT80 had significantly higher NH 4 + contents than MP0 (p = 0.0004 and p = 0.0047, respectively), or MP80 (p = 0.0031 and p = 0.0374, respectively). Fertilization had no effect on NH 4 + contents in MP (p = 0.7332) or NT (p = 0.6120) treatments. In the Fertilizer experiment NT0, NT80, NT160 and MP160 treatments had significantly higher NH 4 + contents compared to MP0 (p < 0.0250) and MP80 (p < 0.0026), and there was no significant difference between the NH 4 + content of MP0 and MP80 (p = 0.9979). NH 4 + content of NT treatments was not significant as compared to each other (p > 0.8478). Significantly higher NH 4 + of NT80 compared to MP80 was observed in the Temperature dependency experiment (p = 0.0079; Table 9).
In the Tillage experiment there were no significant differences between the NO 3 − content of MP0 and MP80 (p = 0.7595) and between NT0 and NT80 (p = 0.7590), while the differences were significant between MP0 and NT0 (p = 0.0277) and between MP80 and NT80 (p = 0.0276). In the Fertilizer experiment there were significant differences (p < 0.0233) between the NO 3 − content of all of the fertilizer doses in MP. In NT there was no significant difference (p > 0.9999) in the NO 3 − content of NT0 and NT80, although these fertilizer treatments significantly differed from NT160 (p = 0.0183 and p = 0.0095, respectively). There was a significant difference between the NO 3 − content of MP and NT treatments (p = 0.0211) only under 160 kg ha −1 fertilization. In the Temperature dependency experiment there were no significant differences between the NO 3 − content of the two (MP80 and NT80) treatments (p = 0.1508; Table 9).

GHG Emissions of Field Experiment
The annual course of CO 2 emission showed an increasing trend in the first part of the vegetation period until the end of May, then decreased during ripening till harvest in both MP and NT treatments. A secondary emission peak occurred in the after-harvest period with a decline in fall and winter (Figure 3a).
The whole year investigation period was divided into two sub-periods: vegetation period, and after harvest period with no vegetation (Figure 3, Table 4). In the NT treatment, the after harvest period had a significantly higher mean CO 2 emission than the vegetation period (p < 0.0001) or the whole year period (p = 0.0080). Regarding MP none of the three investigated periods had significant differences as compared to each other (p = 0.4778), thus MP showed a more balanced CO 2 emission (Table 10).

GHG Emissions of Field Experiment
The annual course of CO2 emission showed an increasing trend in the first part of the vegetation period until the end of May, then decreased during ripening till harvest in both MP and NT treatments. A secondary emission peak occurred in the after-harvest period with a decline in fall and winter (Figure 3a).    Table 4).

Treatment
Mean CO 2 Emission (mg m −2 s −1 ) whole year period vegetation period after-harvest period MP 0.063 ± 0.035 a 0.068 ± 0.037 a 0.059 ± 0.031 a NT 0.072 ± 0.032 b 0.051 ± 0.022 c 0.095 ± 0.041 a NT had a slightly higher mean CO 2 emission than MP during the whole year period, but the difference was not significant (p = 0.3933). During the vegetation period (from 18/02/19 to 01/08/19) MP had significantly higher mean CO 2 emission than NT (p = 0.0028). In the after harvest period, where plant and root activity did not prevail, NT had a significantly higher CO 2 emission than MP (p < 0.0001) (Table 11). Small letters indicate significant differences between the treatments within a period. Tillage treat-ments: MP: mouldboard ploughing, NT: no-tillage. (Sample numbers can be seen in Table 4). N 2 O emission data were collected from the after harvest period only, where there was no vegetation on the selected parcels. Our data showed that N 2 O emission had a non-specific temporal trend on the investigated arable field (Figure 3b). Both the temporal variability of N 2 O emission and the deviation of the measurement points were higher in NT than in MP (Figure 3b). Mean N 2 O emission was significantly higher (p = 0.0081) in NT than in MP treatment (Table 11).

Environmental Governing Factors of GHG Emissions in Field Experiment
Regarding the whole year period, the correlation between Ts and CO 2 emission was weak but significant (r 2 = 0.27, p = 0.0043 in MP and r 2 = 0.25 p = 0.0067 in NT (Table 12)), while during the vegetation period, the correlations were weak and nonsignificant (r 2 = 0.002, p = 0.6418 in MP and r 2 = 0.006, p = 0.4277 in NT). CO 2 emission significantly correlated with Ts (r 2 = 0.89, p < 0.0001 in MP and r 2 = 0.45, p = 0.0018 in NT) in each treatment after harvest. N 2 O emission of MP slightly depended on Ts during the after harvest period. There was weak and not significant correlation between Ts and N 2 O emission in MP (r 2 = 0.30; p = 0.1995) and no correlation in NT (r 2 = 0.03; p = 0.6984). Table 12. Environmental drivers of GHG emission and its significance*. r 2 values are indicated. Tillage treatments: MP: mouldboard ploughing, NT: no-tillage. (Sample numbers are presented in Table 3). According to the collected data of the field experiment SWC of NT was significantly higher than in MP (p = 0.0020; Table 4). Bulk density of NT was also slightly higher than MP with 1.6 ± 0.0 and 1.5 ± 0.0 g cm −3 , respectively. We could not observe any correlation between the SWC and the CO 2 emission during the after harvest period (r 2 = 0.00, p = 0.9044) in MP and (r 2 = 0.02, p = 0.7134) in NT. N 2 O emission showed high dependency on SWC of the field trial. Correlation was strong and significant in both tillage treatments (r 2 = 0.83, p = 0.0347 in MP and r 2 = 0.82, p = 0.0399 in NT; Table 12).

Soil Chemical Parameters of the Field Experiment
During the after-harvest period of the field experiment, we collected soil samples every N 2 O measurement day from all sampling points. The mean soil chemical parameters of the treatments are summarized in Table 13. In the field experiment NT had significantly higher mean N tot (p < 0.0001), NH 4 + -N (p = 0.0330), NO 3 − -N (p = 0.0030), and TC (p < 0.0001) content than MP treatment.

Plant Parameters-Field Experiment
According to the background data of the crop, winter wheat had a higher LAI in MP, than in NT during ear emergence (205 days after sowing). The maximum green LAI occurred 236 days after sowing. Winter wheat also had a higher LAI in MP than in NT. Yield, straw, and root data collected at harvest showed higher vegetation biomass in MP as compared to NT (Table 14).

GHG Emissions Affected by Tillage and Fertilization
NT had significantly higher CO 2 emission than MP, not only in all laboratory experiments involving bare soil, but also in field conditions in the after harvest period. The higher TC content in the upper layer of NT than MP might resulted the difference in CO 2 emissions [61][62][63]. We can presume the same reason in the case of laboratory experiments, as within a tillage treatment TC content varies on the scale of a decade rather than annually [64,65].
As a contrast, MP had significantly higher CO 2 emission as compared to NT in the vegetation period. Based on the literature, root respiration is one of the major factors of soil CO 2 emission [66], but its contribution to total soil CO 2 emission has a diurnal and seasonal variability. The root respiration/total CO 2 emission ratio is the highest in vegetation season and strongly correlates with LAI [67]. In our experiment, MP had higher LAI than NT in the vegetation period, thus the presence of winter wheat can be a partial explanation of the significantly higher CO 2 emission of MP than NT over this period. Root biomass data from our field trial also support this assumption.
Our finding regarding the non-significant difference in CO 2 emission of MP and NT during the whole year period coincides with an experiment conducted on a Stagnosol, which also reported similar results of these two tillage methods under winter wheat cultivation [68]. In the arable field investigated in this article, the presence of winter wheat caused higher CO 2 emission of MP than in NT during the vegetation period. In the after harvest period, the higher TC content of NT than MP might explain higher CO 2 emission, while the effect of vegetation and soil chemistry balanced out each other during the whole year. Similar to our finding, in a continuous maize cropping system, in the long term notillage was found to have significantly lower CO 2 emission as compared to conventional tillage during the vegetation period [69]. Although several studies report that conventional practices like ploughing promote soil CO 2 emission as compared to no-tillage or reduced tillage techniques during one or a number of seasons [24][25][26][27], other reports present opposite findings [28][29][30]. The contradiction might be due to differences in the experimental conditions such as soil types, vegetation, tillage practices, or weather conditions [70].
Similar to CO 2 emission, NT had significantly higher mean N 2 O emission than MP both in the laboratory experiments and in the after harvest period of the field trial. The significantly higher N tot , NH 4 + and NO 3 − content of NT than MP can be an explanatory factor of higher N 2 O emission. Bhatia  Differences in tillage treatments change several soil conditions, thus the amount of GHG emitted from the soils also varies. Based on a field trial conducted by Bessou et al. [80] compacted and less aerated soils enhance N 2 O emission. Generally, there are differences between the bulk density of soils in tillage and reduced or no-tillage treatments [81,82]. In the NT treatment of our field trial, SWC was significantly higher than in MP, moreover the bulk density of this treatment was also slightly higher than in MP. Available historical data from this field trial also show that NT has usually higher bulk density than MP [46]. In a synthetic work of Rochette [83] based on 25 field studies, no-tillage was found to have higher N 2 O emission as compared to tilled treatment in the case of poorly-aerated soils, and it had a neutral effect on N 2 O emission in medium and well-aerated soils. Thus it seems that soil compaction and SWC are key factors regarding N 2 O emission in different soil types and tillage methods. NT soils from our laboratory experiments are expected to be less aerated due to significantly higher SWC. Besides nitrogen contents this higher SWC of NT could cause an increase in N 2 O emission.
According to our laboratory results MP had significantly lower N 2 O emission than NT even under a similar level of SWC. On the one hand, the presence of nitrifiers and denitrifiers and their ratio [84] in the investigated tillage treatments might be different, resulting in the measured variances of N 2 O emissions. On the other hand, NT is more compacted even under the same SWC, thus it is less aerated than MP inducing higher N 2 O emissions [85,86].
We could not identify clear patterns regarding the connection between the effect of fertilization on MP and NT treatments of laboratory experiment. Fertilizer doses had no effect in many cases. Similar to our findings, van Groenigen et al. [87] reported non linearity of soil N 2 O emission regarding different N application on sandy and clay soils under maize cultivation, indicating that N 2 O emissions are strongly dependent on soil and fertilizer types. The authors found that there were no significant differences between N 2 O emissions of the treatments in the case of different mineral fertilizer dose applications on clay soil but they were significant on sandy soil. According to another model estimation of Grant et al. [88], N 2 O emission of soils had also a non-linear relationship with fertilizer application in temperate, humid climate, and former fertilizations over previous years were also contributing factors of soil nitrogen stocks and N 2 O emission.
It is our conclusion that the differences between the GHG emissions of the two tillage methods can be altered by soil types, weather and soil conditions, crop cultivation, the propagation of the vegetation, and the combination of these factors.

GHG Emissions Affected by Soil Temperature (Ts) and Soil Water Content (SWC)
Our study confirms-concurrently with other works [89,90]-that Ts is among the main governing factors of soil CO 2 emission. CO 2 emission of MP and NT showed very strong correlation with Ts in our laboratory experiment. This correlation was strong in MP and it was moderate in NT during the after harvest period of the field experiment. We have found that the CO 2 emission of MP was more dependent on Ts than the emission of NT. Similar to our results, Reth et al. [91] report stronger correlation of CO 2 emission with Ts in the laboratory than in the field environment.
In the analysis of the vegetation and the whole year period, the CO 2 dependence on Ts was reduced. Horák et al. [92] also reported different Ts dependencies of CO 2 emission in sub-periods, although they found higher correlations during the beginning of the vegetation period than in the post-harvest period. The presence of vegetation might have an overwriting effect on environmental drivers in our study. We found a similar correlation for the whole year in the same field experiment in 2015 when also winter wheat was cultivated [46]. SWC dependency of CO 2 emissions was moderate to strong in the laboratory and it was minor in field conditions. Reth et al. [91] reported that the effect of SWC on CO 2 emissions of soils largely depended on land use, although, it was less pronounced than the effect of Ts. According to Schaufler et al. [89], SWC dependency of CO 2 emission is not linear or exponential, but there is an optimal interval. The authors found that the highest CO 2 emission occurred at medium SWC. Schaufler et al. [89] also claim that N 2 O emission can be better described by SWC than Ts under field circumstances, which confirms the results obtained in our experiment.
The dependency of N 2 O emission on Ts was moderate to strong in laboratory and minor under field conditions. In the field experiment the main driver of N 2 O emission was rather SWC than Ts, which was very strong in both MP and NT.
Based on our results, the correlation of environmental parameters and GHG emission can be more precisely examined under laboratory circumstances.

Conclusions
Tillage (NT and MP treatments) have significant implications on SWC and soil chemistry. The presence of vegetation (winter wheat) substantially elevated CO 2 emission on MP as compared to NT during the vegetation period. After the harvest, CO 2 emission became significantly higher in NT in comparison with MP, very likely due to the initially higher soil carbon content of NT.
The significantly higher N 2 O emission in NT as compared to MP is most probably the result of the different SWC, aeration and different initial chemical characteristics of the tillage treatments after performing 16 years of the same tillage methods. Tillage proved to have a higher effect on GHG emissions than fertilization practices.
Considering our overall data, NT emitted more CO 2 on the chernozem than MP, but the presence of vegetation had a stronger effect on CO 2 emissions than tillage practices. Higher N 2 O emission can be expected in the case of extremely high SWC (heavy rainfalls and thaws) in this arable field, especially for the NT treatment.