N 2 O and CO 2 Emissions from Bare Soil: E ﬀ ect of Fertilizer Management

: The paper presents the results of a laboratory experiment focused on the assessment of the e ﬀ ect of di ﬀ erent methods of application of ammonium nitrate (TD—top dressing and DP—deep placement) on N 2 O and CO 2 emissions from soil without crop cover. Nitrogen application increased soil N 2 O–N ﬂuxes by 24.3–46.4%, compared to untreated soil (NIL). N 2 O–N emissions from TD treatment were higher by 12.7%, compared to DP treatment. Soil CO 2 –C ﬂuxes from DP treatment were signiﬁcantly higher by 17.2%, compared to those from NIL treatment. Nonetheless, the di ﬀ erences between soil CO 2 –C ﬂuxes from DP and TD treatments, as well as from TD and NIL treatments, were of no statistical signiﬁcance. The cumulative greenhouse gas (GHG) emissions (a sum of cumulative soil emissions of CO 2 –C and N 2 O–N after conversion to the equivalent of CO 2 –C) from both N-fertilized soils were similar, and higher by 20% than from untreated soil. The obtained data show that the e ﬀ ect of reduction of N 2 O–N soil emissions gained by deep placement of nitrogen fertilizer was completely lost through an increase in CO 2 –C emissions from the soil. This suggests that deep placement of nitrogen fertilizers in sandy soil without crop cover might not lead to a mitigation of soil GHG emissions.


Introduction
Compared to conventional fertilization and cultivation, a deep placement of nitrogen fertilizers provides an alternative method of fertilization, allowing for effective plant nutrition in a later grow stage of plants [1]. Deep placement of nitrogen fertilizers seems to be a promising method of mitigation of N 2 O-N soil emission [2,3]. Gaihre et al. [4] showed that N 2 O-N emissions from soil with deep-placed urea were lower by up to 80% than from top-dressed soil. Chatterjee et al. [5] argued that deep placement of nitrogen could reduce N 2 O-N emissions, since a larger fraction of fertilizer nitrogen can be maintained in the soil for a longer period. According to Chapuis-Lardy et al. [6], a reduction of soil N 2 O-N emission results from N 2 O microbial consumption. Rutkowska et al. [7] evaluated the effect of deep nitrogen placement on N 2 O-N soil emissions from C-poor light sandy soils on four experiments conducted in different regions of Poland. In this study, deep fertilizer placement mitigated the soil emission of N 2 O-N, although the reductions of soil N 2 O emission were lower than those described in the literature. Adviento-Borbe and Linquist [8] observed no significant differences in N 2 O-N soil emissions between broadcasting and deep placement of nitrogen in the soil, whereas Linquist et al. [9] reported that nitrogen deep placement promoted N 2 O-N emissions. These inconsistent results may be explained by differences in the source and amount of N fertilizer and interactions between up to 25 cm depth (ploughing soil layer) from five replications of the experimental treatment of a long-term experiment in Skierniewice (51 • 96 60" N, 20 • 16 63" E). In the treatment, mineral fertilizers were applied annually at the following rates: 90 kg N ha −1 (ammonium nitrate); 26 kg P ha −1 (triple superphosphate); 91 kg K ha −1 (potassium chloride 50%). In the long-term experiment, since 1923, plants have been cultivated in a five-crop rotation: potatoes (30 t manure ha −1 ), spring barley, yellow lupine, winter wheat, and rye (the soil was collected in the autumn, after rye). In the long-term experiment, a conventional ploughing soil tillage was used. The soil was loamy sand with the following fractions in the 0-25 cm layer: 87% sand (>0.05 mm), 5% silt (0.002-0.05 mm), and 7% clay (<0.02 mm). The content of soil organic carbon and total nitrogen was as follows: 7.35 g C kg −1 ± 0.15 standard deviation (SD) and 0.71 g N kg −1 ± 0.02 SD, respectively, and soil NO 3 − -N content was 8.50 mg NO 3 − -N kg −1 ± 0.55 SD (the average from five replications). The soil pH in 1 M KCl was 5.8 (with the range between 5.7 and 6.0 pH). After air-drying and sieving through a mesh of 5 mm, 8 kg of the soil was placed in polyvinyl chloride (PVC) pots. The pots used for the experiment were 20 cm in height with an inside top and bottom diameter of 22 cm and 19 cm, respectively. The pot experiment was conducted in three repetitions, in a fully random system. No plants were cultivated. Over the study period, each pot was weighed daily, and any lost moisture replenished with distilled water to maintain soil moisture content at 60% WFPS. Two days after the first irrigation, the non-fertilized treatment (NIL) and N-fertilized treatments were established. Nitrogen was applied in the fertilized treatments at a dose of 1 g N per pot, in the form of ammonium nitrate (34% N). As we decided to test the nitrogen fertilizer that is most prevalent in Poland, we evaluated the effects of the application of ammonium nitrate, which remains the fertilizer used the most in our agrotechnical and soil conditions, irrespective of a noticeable increase in the urea consumption [25]. The fertilizer was scattered on the surface of soil (top dressing-TD treatment) or applied at a depth of 10 cm (deep placement-DP treatment), which can be extrapolated to 26.31 g N m −2 and 263 kg N ha −1 in field conditions. The extrapolation of the experimental dose of nitrogen to the area of 1 m 2 and 1 ha was made assuming that 1 g of N was applied to the pot area of 0.038 m 2 . Such doses of nitrogen are currently applied in the cultivation of maize on the Polish soil and agricultural conditions. Therefore, we decided to use similarly high nitrogen doses in the laboratory experiment.

N 2 O-N and CO 2 -C Emissions
N 2 O-N and CO 2 -C emissions from the soil were measured in situ by means of infrared spectroscopy using a portable FTIR spectrometer, model Alpha (Bruker, Optic GmbH, Ettlingen, Germany). The measurements of N 2 O-N and CO 2 -C emission was taken before, and a day after, soil nitrogen fertilization, and were continued weekly for the next 51 days (nine test dates) in all replications of the experimental treatments. The emission of N 2 O-N and CO 2 -C (F) from the soil was calculated as an increase in the amount of N 2 O-N and CO 2 -C in the measurement chamber (ø = 16 cm, h = 18.5 cm) after 10 min exposure to the soil surface according to the equation presented by Burton et al. [26]: where: ∆C/ ∆t is the rate of change in N 2 O-N or CO 2 -C concentration inside the chamber, A is the surface area of the chamber, Vc is the total volume of the chamber corrected by temperature, Mmol is molar mass of N 2 O-N or CO 2 -C, and Vmol is the molar volume of N 2 O-N or CO 2 -C inside the chamber, corrected by air temperature using the ideal gas law. N 2 O-N and CO 2 -C emissions from soil were expressed in µg N 2 O-N m −2 h −1 and mg CO 2 -C m −2 h −1 , respectively. Cumulative emissions of N 2 O-N and CO 2 -C (i.e., mg N 2 O-N m −2 and g CO 2 -C m −2 ) were calculated by linear interpolation between two close sampling dates and the numerical integration of the function over time, assuming that fluxes changed linearly among sampling days [27]. The correction for background cumulative N 2 O-N emissions from nitrogen fertilized treatments were calculated as the difference between the  [28], according to the formula: CO 2 eq = N 2 O · 298.

Soil Analysis
Soil NO 3 − -N content was measured at the beginning and (separately in the soil layers: 0-10 cm and 10-20 cm) at the end of the experiment, by means of the Skalar San Plus analyzer (Skalar Analytical BV, Breda, Netherlands), after fresh soil extraction in 0.01 mol CaCl 2 dm −3 with a soil/extractant weight ratio of 1:10. The same soil samples were subject to extraction of dissolved organic carbon (DOC) in accordance with the methodologies described by Zsolnay [29]. The soil samples were extracted in 0.01 mol CaCl 2 dm −3 . The weight ratio of soil to extraction solution was 1:2, and the extraction time was 10 min. Total organic carbon (TOC) content in the soil and soil DOC content were measured by means of the Thermo Electron-C TOC-500 (Shimadzu, Kyoto, Japan). Soil total nitrogen (TN) content was measured using a Vapodest VAP 30 analyzer (Gerhardt, Bonn, Germany).

Statistical Analysis
To determine statistically significant differences between treatments (at p < 0.05), one-way analysis of variance (ANOVA) was carried out. Homogeneous groups for the examined treatments were determined by Tukey's (HSD) multiple-comparison test. Data had been previously tested for normality distribution by a Shapiro-Wilk's test, and for homoscedasticity by Levene's test. Statistical analyses were carried out with the application of the Statistica PL 13.3 software (Tulsa, OKLA, USA).

Soil N 2 O-N Fluxes
The measured soil N 2 O-N fluxes over the study period are presented in Figure 1. Soil N 2 O-N fluxes from all treatments before ammonium nitrate application on the first day (D1) of study were similar. Differences in measured fluxes became noticeable on D2. Compared to NIL treatment and DP treatment, scattering of ammonia nitrate on the soil (TD treatment) resulted in a significant (p < 0.05) increase in the measured soil N 2 O-N flux as early as on D2. On D9, the measured N 2 O-N soil flux from TD treatment was more than twofold higher than from NIL treatment, and approximately 76% higher than from DP treatment. A substantial impact of NH 4 NO 3 top dressing on the measured N 2 O-N soil fluxes was also observed on D16, when soil N 2 O-N emissions from TD treatment reached a maximum. Soil N 2 O-N emissions from NIL and DP treatments peaked one week later (D23). That suggests that deep N-fertilizer placement may delay and reduce the N 2 O-N emissions from the soil. After the peak of N 2 O-N emissions, the amount of nitrous oxide emitted from both N-fertilized soils showed a decreasing pattern until D30, followed by an increase, which resulted in second, smaller peaks reached by DP and TD treatments on D37 and D44, respectively. The N 2 O-N emissions from NIL treatment followed a pattern similar to those from TD treatment; however, the measured fluxes from NIL were significantly lower. The observed changes in N 2 O-N emissions from the tested soils could be caused by an increase in the microbial activity of the soil and, also, to some extent, by changes in natural soil compaction during the experiment.
increased N2O-N emissions of the soil [6,[31][32][33]. The significant (p < 0.05) impact of ammonium nitrate application on N2O-N emission evidenced in our study is similar to experimental results obtained by Bouwman et al. [34], van Groenigen et al. [35], and Meng et al. [36], who observed a significant increase in N2O-N emissions from the soil, particularly at high doses of nitrogen (200-300 kg N ha −1 ). We observed a limited transformation of the fertilizer nitrogen into N2O-N. Corrected for background cumulative soil, N2O-N emissions from TD and DP treatments were low, and accounted for 0.22% and 0.14% of applied N, respectively. In several studies, differing in the soil and atmospheric conditions, plant cover, nitrogen fertilizers, or nitrogen leaching, N2O-N emissions from agricultural soils were found to vary between 0.05% and 15.54% of applied nitrogen [37][38][39][40]. Upon the TD and DP treatments, the measured N2O-N losses from the estimated load of 263 kg N ha −1 equaled, respectively, 0.58 kg and 0.37 kg N ha −1 . Our experiment, however, lasted only 51 days, a fraction of the growing period in Central Poland. The emission of N2O-N from TD and DP treatments extrapolated to the full growing period (March to November, i.e., approximately 210-220 days) reached 2.50 kg and 1.59 kg per ha (i.e., 0.6-0.95% of nitrogen applied per ha). Our earlier laboratory experiment showed that N2O-N emission from soil without mineral fertilization was 0.77 kg N2O-N ha −1 [41]. Therefore, it can be expected that the total soil N2O-N emission from the TD and DP treatments during the growing period would reach a maximum of 3.27 kg and 2.36 kg N2O-N ha −1 , respectively.
According to Zhu et al. [13], in soils where nitrifier denitrification is the dominant pathway of N2O-N production, application of ammoniacal fertilizer results in higher N2O-N emissions, compared with acidic-forming fertilizers (e.g., NH4NO3). In the experimental conditions of the soil, i.e., soil texture, low content of soil carbon, and moisture at level of 60%, promoted the nitrifier denitrification, rather than denitrification, as the dominant pathway of N2O-N production, resulting in a preferential use of the ammonium ion and retention of nitrate in the soil. Most of the NO3 − -N remained accumulated in the soil ( Table 2). In comparison to the NIL treatment, the application of ammonium nitrate as part of the TD treatment increased the content of NO3 − -N in the 0-10 cm and 10-20 cm soil layers 23.8-fold and 2.3-fold, respectively. Under the conditions of deep N-fertilizer placement (DP treatment), the nitrate contents in the 0-10 cm and 10-20 cm soil layers were 1.6-fold and 25.6-fold higher, respectively, than on NIL treatments.
The share of N2O-N emitted from NH4NO3 applied in the DP and TD fertilization systems reached merely 21.2% and 30.0% of total cumulative nitrous oxide emissions from the soil (Table 1). This finding is explained by relatively high emissions of N2O-N from the NIL treatment. Prior nitrogen fertilization, as part of the long-term experiment in Skierniewice, which led to an accumulation of various forms of organic and mineral nitrogen in the soil and, possibly, their  4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 Table 1). The determined high variability of soil N 2 O-N emission patterns over the study period seems to be typical of this process [30]. Measured N 2 O-N fluxes from the soil treated with ammonium nitrate significantly (p < 0.05) exceeded (approximately by 24.3 and 46.4% on the DP and TD treatments, respectively) those from NIL ( Table 1). The cumulative soil N 2 O-N emissions from soils under TD and DP treatments were significantly (p < 0.05) higher, by 26.9% and 42.9%, than those from NIL treatment (Table 1). Several authors indicated that N-fertilization increased N 2 O-N emissions of the soil [6,[31][32][33]. The significant (p < 0.05) impact of ammonium nitrate application on N 2 O-N emission evidenced in our study is similar to experimental results obtained by Bouwman et al. [34], van Groenigen et al. [35], and Meng et al. [36], who observed a significant increase in N 2 O-N emissions from the soil, particularly at high doses of nitrogen (200-300 kg N ha −1 ). We observed a limited transformation of the fertilizer nitrogen into N 2 O-N. Corrected for background cumulative soil, N 2 O-N emissions from TD and DP treatments were low, and accounted for 0.22% and 0.14% of applied N, respectively. In several studies, differing in the soil and atmospheric Agriculture 2020, 10, 602 6 of 14 conditions, plant cover, nitrogen fertilizers, or nitrogen leaching, N 2 O-N emissions from agricultural soils were found to vary between 0.05% and 15.54% of applied nitrogen [37][38][39][40]. Upon the TD and DP treatments, the measured N 2 O-N losses from the estimated load of 263 kg N ha −1 equaled, respectively, 0.58 kg and 0.37 kg N ha −1 . Our experiment, however, lasted only 51 days, a fraction of the growing period in Central Poland. The emission of N 2 O-N from TD and DP treatments extrapolated to the full growing period (March to November, i.e., approximately 210-220 days) reached 2.50 kg and 1.59 kg per ha (i.e., 0.6-0.95% of nitrogen applied per ha). Our earlier laboratory experiment showed that N 2 O-N emission from soil without mineral fertilization was 0.77 kg N 2 O-N ha −1 [41]. Therefore, it can be expected that the total soil N 2 O-N emission from the TD and DP treatments during the growing period would reach a maximum of 3.27 kg and 2.36 kg N 2 O-N ha −1 , respectively.
According to Zhu et al. [13], in soils where nitrifier denitrification is the dominant pathway of N 2 O-N production, application of ammoniacal fertilizer results in higher N 2 O-N emissions, compared with acidic-forming fertilizers (e.g., NH 4 NO 3 ). In the experimental conditions of the soil, i.e., soil texture, low content of soil carbon, and moisture at level of 60%, promoted the nitrifier denitrification, rather than denitrification, as the dominant pathway of N 2 O-N production, resulting in a preferential use of the ammonium ion and retention of nitrate in the soil. Most of the NO 3 − -N remained accumulated in the soil (Table 2). In comparison to the NIL treatment, the application of ammonium nitrate as part of the TD treatment increased the content of NO 3 − -N in the 0-10 cm and 10-20 cm soil layers 23.8-fold and 2.3-fold, respectively. Under the conditions of deep N-fertilizer placement (DP treatment), the nitrate contents in the 0-10 cm and 10-20 cm soil layers were 1.6-fold and 25.6-fold higher, respectively, than on NIL treatments. Values followed by the same letters ( a,b ) in the column are not statistically different (p < 0.05).
The share of N 2 O-N emitted from NH 4 NO 3 applied in the DP and TD fertilization systems reached merely 21.2% and 30.0% of total cumulative nitrous oxide emissions from the soil (Table 1). This finding is explained by relatively high emissions of N 2 O-N from the NIL treatment. Prior nitrogen fertilization, as part of the long-term experiment in Skierniewice, which led to an accumulation of various forms of organic and mineral nitrogen in the soil and, possibly, their transformation to N 2 O-N during the study period, can partly explain the small differences in N 2 O-N emissions from experimental (fertilized) treatments (TD and DP) and the NIL treatment.
Further consideration needs to be given to the soil reaction. In different soil conditions (soil texture, low content of soil carbon, and moisture), several studies showed that optimum values of soil pH for N 2 O-N production ranged between pH 3.9 and 7.0 [42][43][44][45]. Although Russow et al. [46] showed that high NH production [47,48]. In acid soils, approximately 80% of the N2O-N flux results from denitrification. Soil pH of the tested soils (pH 5.8) should be favorable for denitrification. However, we maintained non-variable soil moisture at the level of 60% WFPS, which is unfavorable for denitrification. That was probably the cause of the relatively low N 2 O-N emissions from the soil treated with NH 4 NO 3 (DP and TD treatments). Several authors evidenced a reduction of N 2 O-N emissions from soil after deep N-fertilizer placement [9,49,50]. Liu et al. [3] reported mitigation of N 2 O-N emission after an application of liquid urea-ammonium-nitrate fertilizer at soil depths of both 10 cm and 15 cm. However, the authors observed no significant differences between N 2 O-N emissions from soil fertilized on the topsoil and at a depth of 5 cm. Hosen et al. [51] showed that the deep application of urea neither to the 0-10 cm soil layer nor to the 10-20 cm soil layer did reduce the N 2 O-N emissions. In their opinion, N 2 O-N emission is more dependent on soil conditions than on the depth of fertilizer application. The level of mitigation of N 2 O-N emissions recorded in our study was lower than that recorded by Gaihre et al. [4] in field conditions under rice cultivation. Cumulative N 2 O-N emissions from the soil with top dressing of NH 4 NO 3 application were higher by 12.7%, compared to N 2 O-N emission from the soil with deep N-fertilizer placement (Table 1). Our results were similar to data obtained by Rutkowska et al. [7]. According to the authors, in the majority of experimental locations, in a year with lower precipitation, deep placement of nitrogen fertilizer in maize cultivation resulted in lower than a 17% reduction of N 2 O-N soil emissions. However, on numerous measurement dates, N 2 O-N emissions from soils with deep N-fertilizer placement were similar to, or higher than, those from the conventional top-dressing system of maize fertilization [7]. Chapuis-Lardy et al. [6] hypothesized that conditions interfering with N 2 O-N diffusion in the soil enhanced the N 2 O-N consumption and decreased the N 2 O-N soil emissions as an effect of the longer presence of N 2 O-N in the soil. Our data confirm that the diffusion of N 2 O-N between the place of its formation and the soil surface has an important impact on the emissions from the DP treatment.
Among multifactorial impacts, the soil moisture and content of soil organic carbon act as the proximal regulator of N 2 O-N production in the soil. Both soil moisture and the content of organic carbon regulate the oxygen availability to soil microbes. Zhu et al. [13] showed that oxygen availability regulated pathways of N 2 O-N formation in the soil. N 2 O-N production via NH 3 -N oxidation increased with decreasing O 2 concentration from 21% to 0.5% (vol/vol). At low O 2 concentrations, the contribution of nitrifier denitrification and heterotrophic denitrification to total N 2 O-N production ranged between 34% and 66% and between 34% and 50%, respectively. Heterotrophic denitrification was responsible for all N 2 O-N production at 0% O 2 . Bateman and Baggs [52] showed that N 2 O-N was formed during denitrification at 70% WFPS. According to Butterbach-Bahl et al. [12], N 2 as the end product of denitrification is formed in the soil at soil moisture higher than 80-90% WFPS. The available organic carbon is also a critical factor in controlling denitrification [53]. Its availability promotes the microsite anaerobiosis as a result of increased respiratory demand for oxygen. Therefore, an increase in the availability of labile carbon would favor complete denitrification to N 2 . In our study, soil moisture was constantly maintained at 60% WFPS, and the content of dissolved organic carbon (DOC) in the soil at the end of the study period was low (Table 2). Therefore, the possibility of reducing the soil N 2 O-N emissions from DP treatment as a result of denitrification was limited. The difference in N 2 O soil emissions between DP and TD treatments shows that, at soil moisture of 60% WFPS and low DOC soil content, there is a limited possibility of higher N 2 O microbial consumption via denitrification, as an effect of a longer presence of N 2 O in the light sandy soil.

Soil CO 2 -C Fluxes
Measured soil CO 2 -C fluxes are presented in Figure 1. Soil CO 2 -C fluxes from all treatments before ammonium nitrate application on D1 were similar. Differences in measured fluxes became perceptible on D2. A gradual increase in the emission of CO 2 -C from the soil was recorded for all treatments leading to the first (and the least elevated) peak on D9. Subsequently, we observed a drop Agriculture 2020, 10, 602 8 of 14 in the measured CO 2 -C fluxes, with a first nadir noted for all treatments on D16. This was followed by a substantial rise in the measured CO 2 -C fluxes resulting, for all treatments, in the second (and the most elevated) peak on D23. Similar emission patterns, with a peak on D23, may have resulted from the short period between the soil irrigation and measurement date, and the soil compaction change. Before the experiment, the soil was dried and sieved. During the experiment, irrigation could have resulted in a change in the soil density. Hence, the observed changes could, to some extent, be driven by factors not evaluated in the experiment. Successively, an extensive decline in the measured CO 2 -C fluxes occurred under all treatments, with the second nadir seen on D30 for NIL treatment and on D37 for TD and DP treatments. The third peak and third nadir were observed under all treatments on D44 and D51, respectively. Except for D2 and D30, the differences in measured CO 2 -C fluxes from the studied treatments were not significantly different. On D2, the fluxes for both NIL and DP treatments were significantly higher, as compared with TD treatment. On D30, the measured CO 2 -C flux from DP was significantly higher than those from TD and NIL treatments, whereas the flux from TD was significantly higher, as compared with NIL treatment. As a consequence of daily fluxes, significantly (p < 0.05) higher cumulative CO 2 -C soil emissions were recorded from DP treatment than from NIL treatments ( Table 3). Table 3. Daily CO 2 -C soil emissions and total cumulative CO 2 -C soil emissions from the soil treated with NH 4 NO 3 in the top dressing (TD) and deep placement (DP) fertilization systems.

Treatment CO 2 -C Emissions
Daily Cumulative Values followed by the same letters ( a,b ) in the column are not statistically different (p < 0.05).
Measured soil CO 2 -C daily fluxes over the study period showed a high variability with a range of 12.7-109.2 mg C m −2 h −1 from TD treatment, 12.9-102.5 mg C m −2 h −1 from DP treatment, and 14.1-100.5 mg C m −2 h −1 from NIL treatment (Table 3).
Depending on the way of experimental fertilization, soil CO 2 -C fluxes from DP treatment were significantly (p < 0.05) higher by approximately 17.2%, compared to those from NIL treatment. We observed differences between the measured soil CO 2 -C fluxes from DP and TD treatments, as well as from TD and NIL treatments; however, they were of no statistical significance. CO 2 -C fluxes recorded over the study period were relatively low, compared to those occurring on sandy light soils in the climatic conditions of Poland [54][55][56]. The relatively low soil CO 2 -C fluxes recorded in our study with no crop cover were expected. It is generally acknowledged that CO 2 -C is formed in the soil via both autotrophic and heterotrophic respiration. Adviento-Borbe et al. [23] estimated that autotrophic respiration could account for 36% of total soil respiration in high-yielding maize systems. Rochette et al. [57] evidenced that up to 50% of soil respiration resulted from root respiration and decomposition of root exudates. Amos et al. [24] reported that soil flux of CO 2 decreased as plants reached physiological maturity and senescence, due to the absence of root respiration.
The study data suggest that nitrogen fertilization, and the different methods of its application, has an impact on the CO 2 -C emissions from the soil. Several studies indicated that N-fertilization Agriculture 2020, 10, 602 9 of 14 increased CO 2 -C emissions from the soil [19,20]. Song and Zhang [20] reported, however, that nitrogen fertilization with 250 kg N ha −1 suppressed soil respiration. In addition, Ding et al. [58] revealed that nitrogen fertilization resulted in a reduction of soil CO 2 -C flux. The dose of nitrogen applied in our study could result in no differences in CO 2 -C soil emissions, as exemplified by the lack of differences between the fluxes from TD and NIL treatments (Table 3). DeForest et al.
[59] evidenced that CO 2 -C fluxes could decrease when using high N rates, as a consequence of lower activity of enzymes in the soil. In our study, deep placement of ammonium nitrate resulted in a significant increase in CO 2 -C emission from the soil, compared to NIL treatment. Our results are consistent with data presented by Pareja-Sánchez et al. [22]. The authors demonstrated that nitrogen fertilization increased the CO 2 -C soil emissions as a result of different availability of soil nitrogen for decomposers. At the end of our study, the content of NO 3 − -N in the topsoil layer was substantially higher under TD treatment than under DP treatment, which could reduce the soil enzymatic activity. On the other hand, the content of nitrates in the topsoil layer under DP treatment was insignificantly higher than on NIL treatment. Diffusion of NO 3 − -N from deeper soil layers to the topsoil layer on DP treatment probably increased the amount of nitrogen available for decomposers, and resulted in higher CO 2 -C soil emissions. This may be supported by a higher DOC content in the topsoil layer under DP treatment than NIL treatment (Table 2), a difference which, however, did not reach statistical significance. Adviento-Borbe et al. [23] showed that seasonal fluctuations in soil CO 2 -C fluxes depended on soil NO 3 − -N, though to a lesser extent than on the air temperature and soil moisture. Total cumulative soil CO 2 -C emissions recorded over the study period were correlated only with soil DOC content (Table 4).

Cumulative Greenhouse Gas (GHG) Emissions from the Soil
Cumulative greenhouse gas (GHG) emissions from the soil, calculated as a sum of cumulative soil emissions of CO 2 -C and N 2 O-N converted to the equivalent of CO 2 -C, are presented in Figure 2. Irrespective of the method of nitrogen application, the cumulative GHG emissions from TD and DP treated soil (125.82 g CO 2eq -C m −2 and 126.40 g CO 2eq -C m −2 , respectively) were significantly (p < 0.05) higher, by approximately 20%, than from NIL treatment (104.40 CO 2eq -C m −2 ). No significant differences in soil GHG emissions were found between the two systems of N-fertilizer application. Our data demonstrate that an increase in CO 2 -C emissions from the soil without crop cover counterbalances the reduction of N 2 O-N soil emissions gained by deep placement of nitrogen fertilizer, leading to the conclusion that deep placement of nitrogen fertilizer in light sandy soils without crop cover might not mitigate the GHG emissions from the soil. Due to a greater energy input required for deep placement of the fertilizer into the soil, the total soil GHG emissions may exceed the GHG emissions from the traditional fertilization and cultivation systems. without crop cover might not mitigate the GHG emissions from the soil. Due to a greater energy input required for deep placement of the fertilizer into the soil, the total soil GHG emissions may exceed the GHG emissions from the traditional fertilization and cultivation systems.

Conclusions
Nitrogen fertilization increases the N 2 O-N emissions from the soil. High soil emission of N 2 O-N from topsoil dressed with ammonium nitrate occurs right after fertilizer application. Deep placement of ammonium nitrate at a depth of 10 cm decreases the soil nitrous oxide emissions by 12.7% and increases the soil CO 2 -C emissions by 17.2%. The relatively low difference in N 2 O-N soil emissions between DP and TD treatments shows that, at soil moisture of 60% WFPS and low DOC content in the soil, there is a limited possibility of higher N 2 O microbial consumption via denitrification as an effect of a longer presence of N 2 O-N in the light sandy soil.
However, N 2 O-N diffusion from a deeper soil layer delays and decreases the intensity of soil surface nitrous oxide emissions. An increase in CO 2 -C emissions from the soil without crop cover counterbalances the reduction of N 2 O-N soil emissions gained by deep placement of nitrogen. This suggests that deep placement of nitrogen fertilizers in light sandy soils might not lead to the mitigation of GHG emissions from the soil without crop cover. The deep N-fertilizer placement method, however, will be required for better plant nutrition, rather than the possibility of reducing GHG emissions from soil.
This study bridges a notable gap in the published evidence to characterize the impact of different methods of N application on the N 2 O-N and CO 2 -C emissions from the soil without crop cover. Such conditions, albeit currently unusual, are expected to occur more frequently in the future as a result of atypical weather conditions, which affect plant growth and yielding loss.

Abbreviations
NIL non-fertilized treatment TD top dressing DP deep placement