Effects of the Fertilizer Added with DMPP on Soil Nitrous Oxide Emissions and Microbial Functional Diversity

Agricultural sites contribute extensively to atmospheric emissions of climate-altering gases such as nitrous oxide. Several strategies have been considered to mitigate the impact of agriculture on climate, among these the utilization of fertilizers added with nitrification inhibitors such as DMPP (3,4-dimethylpyrazole phosphate) may represent a suitable solution. DMPP inhibits the growth and activity of ammonia-oxidizing microorganisms, particularly the ammonia-oxidizing bacteria, which are involved in N2O production. At present, little information is available on the effects of DMPP on the catabolic diversity of soil microbial community. In this study, the N2O emission by soil was performed by using the static chamber technique. The biological determinations of the microbial biomass carbon and the catabolic profile were assessed by measuring the substrate-induced respiration during the entire growing season of a potato crop under two nitrogen treatments: fertilization with and without DMPP. Our results did not show a clear mitigation of N2O emission by DMPP, even if a tendency to lower N2O fluxes in DMPP plots occurred when soil temperatures were lower than 20 °C. Conversely, DMPP deeply affected the microbial biomass and the catabolism of soil microorganisms, exerting a negative effect when it accumulated in excessive doses in the soil, limiting the growth and the capacity of soil microorganism communities to use different substrates.


Introduction
Nitrogen (N) is an essential nutrient for plants and it is often the most limiting factor for crop production. Generally, nitrogen is applied to agricultural soils in a mineral or organic source, frequently in excess of the crop nutritional needs. Nitrogen is used not only by plants but also by soil microorganisms. Microbes are intimately involved in the biological and biochemical transformations occurring in the soil. They are responsible for the mineralization of organic matter and the release of nutrients absorbed by plants. Thus, microbes play an essential role in soil processes, significantly affecting the agroecosystem health through organic matter decomposition, nutrient cycling, and interaction with plants.
Some of N applied to the soil is lost as nitrate (NO 3 − ) by leaching or gas emission, mainly nitrous oxide (N 2 O), making agriculture a sector with a great impact on climate Agriculture 2021, 11, 12 2 of 11 and global environment [1]. The N loss as N 2 O is due to nitrification and denitrification processes, which are controlled by soil physical-chemical conditions such as humidity, temperature, nitrogen and oxygen availability, and pH [2][3][4][5][6]. During nitrification occurring under aerobic conditions, ammonium nitrogen is oxidized to nitrite and nitrate, then reduced in the denitrification process under anaerobic conditions. The supply of ammonium-based fertilizers or organic fertilizers to soil determines an enrichment of ammonium in the soil, leading to the production of NO 3 − and N 2 O [4,5]. Different strategies have been utilized in order to mitigate the emissions of climatealtering greenhouse gases (GHGs) from agricultural sites [7]. Since 1950, many molecules have been identified as possible nitrification inhibitors (NI), i.e., substances constraining N loss from agricultural soils. These compounds added to the fertilizer act to mitigate N loss as NO 3 − and N 2 O by inhibiting biological NH 4 + oxidation, and at the same time, increasing the nitrogen use efficiency (NUE) and yield of crops [8,9]. Among the NIs, 3,4dimethylpyrazole phosphate (DMPP) has been proposed as an advantageous and practical tool for having several characteristics such as low application rate, high efficiency, and low ecotoxicity [10][11][12].
Most studies on DMPP focused attention on its efficiency in reducing soil N 2 O emission [2,3,9,13] and its effects on soil microbial biomass and activity [14][15][16]. These researchers examined the abundance and structure of soil microbial communities in response to DMPP addition, particularly the ammonia-oxidizing microorganisms [15][16][17][18][19][20], but little information is available on the effects of DMPP on functional diversity of soil microbial communities. The microbial functional diversity relates to the capacity to perform different ecological processes in a soil and to utilize a wide collection of substrates. Direct measurements of functional diversity of soil microorganisms could provide more relevant information on soil functioning than measurements of species diversity or community structure, since microorganisms may be present in the soil, but often under rest or dormant phases and not functionally active [21].
In this study, we investigated (1) the effectiveness of DMPP in mitigating N 2 O emission and (2) the outcomes of DMPP application to soil on microbial biomass, and more specifically, the development of the functional responses by soil microbial community after nitrogen fertilization.

Experimental Site
The study was carried out at the experimental farm of the Agricultural Technical Institute "De Cillis" in Ponticelli (Naples, Italy) during spring-summer 2013. The area is characterized by a typical Mediterranean climate with cold, rainy winters and hot, dry summers ( Figure 1). Agriculture 2020, 10, x FOR PEER REVIEW 3 of 12 resulting amount of DMPP applied was 13.9 μg/g dry soil and 4.2 μg/g dry soil, corresponding to the amount of 100 kg N/ha and 30 kg N/ha, respectively. The crop was well watered by means of a sprinkler to avoid water stress during growth.  The soil has a coarse texture (loamy-sand according the USDA textural soil classification) being of volcanic origin. The main soil chemical-physical properties, determined before starting the experiment, are reported in Table S1. Solanum tuberosum L. (cv Agria) tubers were sown on 2013 April 08 on 3 m rows spaced by 0.30 m (3.0 × 3.0 m plot). The experiment was a randomized complete block design with two nitrogen treatments and three replicates. Two treatments were set-up: nitrogen fertilization without nitrification inhibitor (Nitrophoska perfect © , 15:5:20 N:K:P, N as 8% NH 4 + and 7% NO 3 − , and 20% S as SO 3 ) (NH 4 NO 3 plots) and nitrogen fertilization with nitrification inhibitor, the 3,5dimethylpyrazole phosphate (DMPP) (Entec perfect © , 14:7:17 N:K:P, N as 7.9% NH 4 + and 6.1% NO 3 − , and 22% S as SO 3 ; DMPP 0.8% of NH 4 + ) (DMPP plots). Fertilizer was supplied on three events: when sown (100 kg N/ha), and on 04 June 2013 (30 kg N/ha) and 21 June 2013 (30 kg N/ha) by applying the fertilizer near hills that were made on 29 April 2013 by using a disc-type hiller. The amount of N per plot was distributed on the rows per plot. The resulting amount of DMPP applied was 13.9 µg/g dry soil and 4.2 µg/g dry soil, corresponding to the amount of 100 kg N/ha and 30 kg N/ha, respectively. The crop was well watered by means of a sprinkler to avoid water stress during growth.

Biometrical and Physiological Determinations
Harvesting occurred on 09 July. All plants inside each plot were harvested. Tubers were collected, whereas shoot biomass was dried in oven at 75 • C up to constant weight.
The chlorophyll content was used as a proxy to evaluate the plant greenness and was determined by a SPAD (SPAD 502 Plus Chlorophyll Meter, Spectrum Technologies Inc.) at three different times during the plant growing season. Two plants per plot were used and the measurements were performed on three leaves per plant (top, middle, and bottom leaves). Data for each plant were averaged.

Soil Temperature, Humidity, and Nitrogen Content
Soil temperature and moisture were measured during the whole growing season by means of a 107 probe (Campbell Scientific, Inc., Logan, UT, USA) and the CS616 reflectometer (Campbell Scientific, Inc., Logan, UT, USA), respectively. Both probes were introduced into the soil at a 0.0-0.05 m depth inside the chamber and connected to a datalogger (CR1000, Campbell Scientific, Inc., Logan, UT, USA). The water filled pore space (WFPS) was estimated as: where VSWC represents the volumetric soil water content, 2.65 is the average density of the solid matrix calculated on the basis of the relative content of the different mineral constituents, and BD is the bulk density calculated as the dry weight of soil divided by its volume.
Total nitrogen (N tot ) content in the soil was determined by collecting samples from the 0.0-0.10 m soil layer. An integrated soil sample per plot was obtained by collecting and mixing together three different soil samples near each autochamber. Soil was air-dried and sieved (2 mm). Total nitrogen and carbon content was determined by gas chromatograph (Thermo Finnigan, CNS Analyzer).

Soil N 2 O Emissions and Treatments with Acetylene (C 2 H 2 )
Soil N 2 O emission measurements were performed four times during the day and for the whole growing season by automatic chambers (0.30 m diameter and 0.1 m height) linked through a Teflon tube to a gas chromatograph (SRI 8610C, SRI Instruments, Torrance, CA, USA). One chamber per plot was positioned between rows and inserted at 0.03 m depth in the soil. Details on the protocol for gas and flux measurements have been reported in Tedeschi et al. [6].
The contribution of denitrification to total N 2 O fluxes was evaluated by the acetylene inhibition technique (AIT). Soil samples were collected near autochambers at 0.0-0.10 m depth by a PVC cylinder and put into a homemade PVC incubator [5]. The head-space of the incubator was filled with C 2 H 2 to inhibit the nitrification and samples incubated for 6 h at ambient temperature [5]. Untreated samples were also prepared. The air space inside the incubator was frequently mixed by syringes. Air samples were collected at 3 h and 6 h from the head-space with a syringe and analyzed by an SRI 8610C gas chromatograph.

Soil Biological Analisys
Biological analysis were conducted on the same soil subsamples used for the determination of soil N content.
The microbial biomass carbon (C mic ) was evaluated by the substrate-induced respiration (SIR) method as CO 2 evolution from soil following the addition of 2 mL of a D-glucose solution (75 mM) and after the incubation in closed vials (30 mL) for 4 h at 25 • C in the darkness [22]. CO 2 efflux from each sample was measured by an infrared gas analyzer (Model LI6262, LI-COR, Lincoln, NE, USA).
The basal microbial respiration (R esp ) was measured through the LI6262 analyzer and determined after the incubation in vials for 4 h at 25 • C in darkness, following the addition of 2 mL of distilled water to the samples.
The soil metabolic quotient (qCO 2 ), which is the amount of CO 2 -carbon produced per unit of microbial biomass carbon, was calculated as the ratio between basal respiration (R esp ) and C mic . The soil metabolic quotient (qCO 2 ) is an index to evaluate the stress degree of the microbial biomass.
The catabolic response for different substrate types was also determined by measuring the short-term respiration [22] after incubation of the samples for 4 h at 25 • C in darkness following the addition of 19 substrates, namely 7 amino acids (AA), 2 carbohydrates (CH), and 10 carboxylic acids (CA).
The catabolic evenness (E) was determined as E = 1/∑p i 2 , where p i is the ratio of CO 2 evolved for a particular substrate to the sum of all substrates respired.

Statistical Analysis
Statistical analysis of data was performed by Sigma-Plot package (Sigma-Plot 12.2, Systat Software, Inc., San Jose, CA, USA). Differences in biometrical and physiological parameters between fertilization treatments were cheeked by paired t-test (p ≤ 0.05), whereas differences between fertilization treatments or over time inside each treatment were cheeked by one-way ANOVA repeated measurements followed by Duncan's test (p ≤ 0.05).

Plant Greenness, Biomass, and Yield
The plant greenness was unaffected by the type of fertilizer applied (Table 1). A slight, but significant decrease of SPAD values was detected on 09 July, at the harvesting, compared to the previous records. The type of fertilizer did not influence the biomass production and the tuber yield (Table 1).

Soil-Related Measurements and N 2 O Emission
Soil pH showed variations during the studied period and ranged from 6 to 8 in both treatments ( Figure 2a). The wider variation in soil pH was observed in DMPP plots, which dropped to 6 soon after the second fertilization event ( Figure 2a); at that time, soil pH was lower in DMPP plots compared to NH 4 NO 3 plots. Total nitrogen content was higher in DMPP plots compared to NH 4 NO 3 plots after the first fertilization ( Figure 2b). In response to the two additional fertilizations, the DMPP plots showed a lower N content compared to NH 4 NO 3 plots, particularly at the end of the June (Figure 2b).

Soil-Related Measurements and N2O Emission
Soil pH showed variations during the studied period and ranged from 6 to 8 in both treatments (Figure 2a). The wider variation in soil pH was observed in DMPP plots, which dropped to 6 soon after the second fertilization event ( Figure 2a); at that time, soil pH was lower in DMPP plots compared to NH4NO3 plots. Total nitrogen content was higher in DMPP plots compared to NH4NO3 plots after the first fertilization (Figure 2b). In response to the two additional fertilizations, the DMPP plots showed a lower N content compared to NH4NO3 plots, particularly at the end of the June (Figure 2b). Soil temperature increased with the progression of the experimental season ( Figure  3a). The soil temperatures remained below 25 °C until significant volumes of irrigation water were supplied. Peaks higher than 30 °C occurred during the growth season, especially at the end of June (Figure 3a). The soil humidity, evaluated as water filled pore space (WFPS), changed during the season, following the irrigation events ( Figure 3a). WFPS peaked after irrigation, ranging from 80% at the beginning of the experiment to 20% at the end of the plant growing season. In the remaining part of the study period, WFPS values were lower than 70% (Figure 3a).
Overall, the N2O fluxes did not differ between the fertilization treatments ( Figure  3b,c), even if a tendency to lower values in DMPP plots compared to NH4NO3 plots was observed after the first fertilization, when soil temperatures were about 20 °C. In particular, a peak of 80 μg m −2 h −1 was measured in the NH4NO3 plots (Figure 3b), whereas it was about 56 μg m −2 h −1 in DMPP plots (Figure 3c). Thereafter, in concomitance with soil temperature higher than 20 °C, N2O fluxes measured in DMPP plots were higher-although not significant-compared to those measured in NH4NO3 plots (Figure 3b,c). For both treatments, peaks of variable amplitude in N2O emission occurred in response to the irrigation that increased the soil humidity and, in turn, the WFPS values.
The acetylene inhibition technique was used to evaluate the denitrification contribution to total N2O flux. Based on this method, the addition of C2H2, as a nitrification inhibitor, to soil cores collected in the field and stored for a short-term in PVC incubators, gave values approaching to zero (Figure 3b,c). Figure 2. Soil pH (a) and total nitrogen (N tot ) content (b) in NH 4 NO 3 and DMPP plots. Data are means (n = 3) ± standard error. Arrows show the fertilization events; asterisks indicate significant differences between treatments (p ≤ 0.05). Data were analyzed by one-way ANOVA repeated measurements followed by Duncan's test.
Soil temperature increased with the progression of the experimental season (Figure 3a). The soil temperatures remained below 25 • C until significant volumes of irrigation water were supplied. Peaks higher than 30 • C occurred during the growth season, especially at the end of June (Figure 3a). The soil humidity, evaluated as water filled pore space (WFPS), changed during the season, following the irrigation events (Figure 3a). WFPS peaked after irrigation, ranging from 80% at the beginning of the experiment to 20% at the end of the plant growing season. In the remaining part of the study period, WFPS values were lower than 70% (Figure 3a).
Overall, the N 2 O fluxes did not differ between the fertilization treatments ( Figure 3b,c), even if a tendency to lower values in DMPP plots compared to NH 4 NO 3 plots was observed after the first fertilization, when soil temperatures were about 20 • C. In particular, a peak of 80 µg m −2 h −1 was measured in the NH 4 NO 3 plots (Figure 3b), whereas it was about 56 µg m −2 h −1 in DMPP plots (Figure 3c). Thereafter, in concomitance with soil temperature higher than 20 • C, N 2 O fluxes measured in DMPP plots were higheralthough not significant-compared to those measured in NH 4 NO 3 plots (Figure 3b,c). For both treatments, peaks of variable amplitude in N 2 O emission occurred in response to the irrigation that increased the soil humidity and, in turn, the WFPS values.
The acetylene inhibition technique was used to evaluate the denitrification contribution to total N 2 O flux. Based on this method, the addition of C 2 H 2 , as a nitrification inhibitor, to soil cores collected in the field and stored for a short-term in PVC incubators, gave values approaching to zero (Figure 3b,c).

Soil Microbial Activity
The microbial biomass carbon (Cmic) was negatively affected by the first fertilization event occurring when sown in response to high N rates (100 Kg ha −1 ). The Cmic reduction in NH3NO3 plots was lower compared to that observed in DMPP plots (Figure 4a). During the potato vegetative growth, corresponding to the middle experimental period, an increase of microbial biomass in DMPP plots was detected. Conversely, Cmic measured in NH4NO3 plots decreased over time until further fertilizations. Following the two fertilization applications, provided at a rate of 30 kg N ha −1 , Cmic showed a growing trend in both

Soil Microbial Activity
The microbial biomass carbon (C mic ) was negatively affected by the first fertilization event occurring when sown in response to high N rates (100 Kg ha −1 ). The C mic reduction in NH 3 NO 3 plots was lower compared to that observed in DMPP plots (Figure 4a). During the potato vegetative growth, corresponding to the middle experimental period, an increase of microbial biomass in DMPP plots was detected. Conversely, C mic measured in NH 4 NO 3 plots decreased over time until further fertilizations. Following the two fertilization appli- cations, provided at a rate of 30 kg N ha −1 , C mic showed a growing trend in both nitrogen treatments. At harvest and at the beginning of the experiment, C mic showed comparable values (Figure 4a). nitrogen treatments. At harvest and at the beginning of the experiment, Cmic showed comparable values (Figure 4a).
The basal microbial respiration (Resp) increased in NH3NO3 plots after the first fertilization (Figure 4b) and showed a peak in DMPP plots on the 20th day. Afterward, compared to the initial values, Resp increased in NH3NO3 plots, reaching the highest value at the end of the experiment, while it decreased in DMPP plots. In response to the two following fertilizations, NH3NO3 plots showed higher Resp values compared to DMPP plots.
The soil metabolic quotient qCO2 mirrored the trend of microbial respiration ( Figure  4c) in both plots, showing the highest value of about 60 μg C-CO2 mg −1 Cmic h −1 in DMPP plots in combination with the highest basal respiration and the lowest microbial biomass carbon. The catabolic response profile was also deeply affected by the first fertilization event (Figure 4d-f). The substrates-induced respiration decrease in both N treatments; however, the reduction was more marked on the third sampling date in DMPP plots compared to NH4NO3 plots, especially for the catabolic response to AA and CH substrates. A similar trend was also observed for the catabolic response to CA substrates. With the progression of the vegetative growing season, the catabolic response reached the initial values, with no particular differences in trends between the two treatments, except for the CH substrates, which showed higher CO2 production in DMPP plots compared to NH4NO3 plots.
The catabolic evenness did not show appreciable differences either over time or between treatments ( Figure 5). A significant reduction in catabolic evenness on the third sampling date in DMPP plots compared to NH4NO3 plots was evident, the marked alterations occurring together in microbial biomass and respiration. The basal microbial respiration (R esp ) increased in NH 3 NO 3 plots after the first fertilization (Figure 4b) and showed a peak in DMPP plots on the 20th day. Afterward, compared to the initial values, R esp increased in NH 3 NO 3 plots, reaching the highest value at the end of the experiment, while it decreased in DMPP plots. In response to the two following fertilizations, NH 3 NO 3 plots showed higher R esp values compared to DMPP plots.
The soil metabolic quotient qCO 2 mirrored the trend of microbial respiration (Figure 4c) in both plots, showing the highest value of about 60 µg C-CO 2 mg −1 C mic h −1 in DMPP plots in combination with the highest basal respiration and the lowest microbial biomass carbon.
The catabolic response profile was also deeply affected by the first fertilization event (Figure 4d-f). The substrates-induced respiration decrease in both N treatments; however, the reduction was more marked on the third sampling date in DMPP plots compared to NH 4 NO 3 plots, especially for the catabolic response to AA and CH substrates. A similar trend was also observed for the catabolic response to CA substrates. With the progression of the vegetative growing season, the catabolic response reached the initial values, with no particular differences in trends between the two treatments, except for the CH substrates, which showed higher CO 2 production in DMPP plots compared to NH 4 NO 3 plots.
The catabolic evenness did not show appreciable differences either over time or between treatments ( Figure 5). A significant reduction in catabolic evenness on the third sampling date in DMPP plots compared to NH 4 NO 3 plots was evident, the marked alterations occurring together in microbial biomass and respiration. Arrows show the fertilization events, asterisks indicate significant difference between treatments (p ≤ 0.05). Data were analyzed by one-way ANOVA repeated measurements followed by Duncan's test.

Discussion
Agriculture is one of the main sectors contributing to environmental alteration, particularly by climate change through the farm management practices. In order to reduce the negative impact of agriculture on climate and environment, different strategies have been proposed and applied in the last few decades. Among these, enhanced fertilizer use efficiency by crop is a promising tool for mitigating climate-altering gas emissions [24]. Better fertilizer use efficiency may be achieved through the development of fertilizers added with specific inhibitors of biological transformations occurring in the soil, particularly nitrification [5].
In this study, we used 3,4-dimethylpyrazole phosphate (DMPP) as a nitrification inhibitor. Many experimental evidences demonstrated how this specific compound is effective in improving fertilizer use efficiency by crop and in reducing N2O emissions from agricultural soils [15][16][17]. Our data clearly showed that the application of DMPP to the fertilizer improved the soil nitrogen content. Indeed, it is likely that DMPP inhibited ammonia oxidation [17], and WFPS values did not favor nitrate leaching. However, DMPP application did not cause a marked mitigation of N2O emission compared to the NH4NO3 plots, even if a slight decrease of N2O fluxes was observed immediately after the first fertilization application. In these circumstances, the WFPS values were high and the soil temperature was lower than 20 °C, which might represent the threshold above which the effectiveness of DMPP was progressively reduced [4,25]. From May and beyond, greater N2O fluxes were measured in DMPP plots compared to NH4NO3 plots, and no improvement in physiological traits (SPAD and aboveground biomass) or in agronomical traits (tuber yield) was found, confirming the loss of DMPP effectiveness with increasing soil temperatures.
The nitrous oxide emissions were mainly produced by nitrification, as demonstrated by flows close to zero in correspondence with the treatments with acetylene (which completely inhibited nitrification). Moreover, the soil water filled pore space was below 70%a watershed value between nitrification and denitrification-for most of the period studied [26], even if the peaks in N2O emission after the first fertilization event were due to a rise in WFPS above 70%. Many evidences have confirmed the functional importance of ammonia-oxidizing archaea (AOA) and ammonia-oxidizing bacteria (AOB) in N2O production during the nitrification [27][28][29]. We suppose that under our experimental conditions AOB were likely the main contributors to N2O production, since AOA gain advantage from growth on acidic and NH4 + -poor soils, while our experimental site showed Arrows show the fertilization events, asterisks indicate significant difference between treatments (p ≤ 0.05). Data were analyzed by one-way ANOVA repeated measurements followed by Duncan's test.

Discussion
Agriculture is one of the main sectors contributing to environmental alteration, particularly by climate change through the farm management practices. In order to reduce the negative impact of agriculture on climate and environment, different strategies have been proposed and applied in the last few decades. Among these, enhanced fertilizer use efficiency by crop is a promising tool for mitigating climate-altering gas emissions [23]. Better fertilizer use efficiency may be achieved through the development of fertilizers added with specific inhibitors of biological transformations occurring in the soil, particularly nitrification [4].
In this study, we used 3,4-dimethylpyrazole phosphate (DMPP) as a nitrification inhibitor. Many experimental evidences demonstrated how this specific compound is effective in improving fertilizer use efficiency by crop and in reducing N 2 O emissions from agricultural soils [14][15][16]. Our data clearly showed that the application of DMPP to the fertilizer improved the soil nitrogen content. Indeed, it is likely that DMPP inhibited ammonia oxidation [16], and WFPS values did not favor nitrate leaching. However, DMPP application did not cause a marked mitigation of N 2 O emission compared to the NH 4 NO 3 plots, even if a slight decrease of N 2 O fluxes was observed immediately after the first fertilization application. In these circumstances, the WFPS values were high and the soil temperature was lower than 20 • C, which might represent the threshold above which the effectiveness of DMPP was progressively reduced [3,24]. From May and beyond, greater N 2 O fluxes were measured in DMPP plots compared to NH 4 NO 3 plots, and no improvement in physiological traits (SPAD and aboveground biomass) or in agronomical traits (tuber yield) was found, confirming the loss of DMPP effectiveness with increasing soil temperatures.
The nitrous oxide emissions were mainly produced by nitrification, as demonstrated by flows close to zero in correspondence with the treatments with acetylene (which completely inhibited nitrification). Moreover, the soil water filled pore space was below 70%-a watershed value between nitrification and denitrification-for most of the period studied [25], even if the peaks in N 2 O emission after the first fertilization event were due to a rise in WFPS above 70%. Many evidences have confirmed the functional importance of ammonia-oxidizing archaea (AOA) and ammonia-oxidizing bacteria (AOB) in N 2 O production during the nitrification [26][27][28]. We suppose that under our experimental conditions AOB were likely the main contributors to N 2 O production, since AOA gain advantage from Agriculture 2021, 11, 12 9 of 11 growth on acidic and NH 4 + -poor soils, while our experimental site showed an alkaline or subalkaline pH, for most of the study phase. However, the contribution of the heterotrophic denitrification in producing N 2 O during the experimental period, especially for WFPS between 50-70%, cannot be excluded, even if it seems to have a negligible role compared to nitrification, which is unexpected for WFPS values greater than 70%. Barrera et al. [29] found that DMPP induced the abundance of nosZ gene-codifying for the nitrous oxide reductase-diminishing N 2 O losses at high water content below 80%. This hypothesis could also explain the lower N 2 O peak observed in DMPP plots compared to NH 4 NO 3 plots for WFPS values higher than 70%.
Most of studies on the effects of DMPP on soil microorganisms have been aimed at understanding the impacts of DMPP on nitrification and dynamics of ammonia oxidizers [17][18][19][20], whereas little information is available on the impact of DMPP on the functional diversity of the soil microbial communities. It is well known that microbial functional diversity represents the capacity to perform different ecological processes and to utilize a wide assortment of substrates. In this study, we used the substrate-induced respiration (SIR) technique to assess the microbial functional diversity, and the catabolic evenness (E) as an indicator to evaluate the variability in substrate utilization by soil microbial communities. The N fertilization per se negatively affected, at least at higher fertilizer application rates, the physiological responses of microbial communities, determining a reduction in the microbial biomass carbon (C mic ) and community functional diversity. Our data agree with some studies reporting a decrease in microbial biomass with N addition [30,31]. Moreover, C mic showed a negative response to high fertilizer application rates, likely because the microbial communities become C limited, resulting from nitrogen accumulation in the soil, due to the absence of plants in this phase and, therefore, in nutrient uptake. This effect was more evident in DMPP plots where the reduction in C mic was higher than in NH 4 NO 3 plots, followed by an opposite trend in N content. However, it is improbable that only N accumulation may induce stress conditions for the soil microbial community. A negative effect of the 3,4-dimethylpyrazole phosphate on microbial communities cannot be excluded on the basis of the drastic reduction in microbial biomass followed by the increase of microbial respiration and qCO 2 on the third sampling data, even if previous reports showed no negative effect of DMPP on nontarget microorganisms in soils when NI was applied at agronomic doses [32,33]. The increase in microbial respiration and in qCO 2 found in DMPP plots indicated a stress condition for microbial communities, probably related to DMPP in the soil. The method of distributing fertilizer adopted in this study resulted in large DMPP doses in the soil, about 14 µg/g dry soil, after the first fertilization, which may have inhibited the growth and activity of microbes. Tindaon et al. [32] reported inhibitory effects of DMPP on nontarget microorganisms at doses greater than 5 µg/g dry soil in light-textured sandy soil. We hypothesize that the local accumulation of large DMPP doses in the soil may have also constrained the activity of nontarget microorganisms. On the other hand, the application of fertilizer at lower rates (30 kg N ha −1 ), corresponding to about 4.2 µg/g dry soil. DMPP, occurred later during the growing season and did not involve a negative response by microorganisms.
DMPP would seem to have adverse effects on ammonia oxidizer and nontarget soil microorganism populations, as well as on their functions. The functional diversity of soil microbial community was deeply affected by the high DMPP doses, as indicated by the catabolic response profile. The soil amended with DMPP showed a reduced biological activity compared to soil amended with fertilizer without nitrification inhibitor, the former being characterized by a reduced evenness (E). In these soils, microbial communities exhibited a reduced capacity to utilize a wide spectrum of C sources and seemed more adapted to consume only some simple carbon compounds, such as carboxylic acids. It is likely to suppose that high doses of DMPP caused alteration in the composition of soil microbial communities, highlighted by a reduction in the catabolic response, favoring some microbial populations more than others. The loss of DMPP effects with the time and under soil temperature higher than 20 • C enhanced the carbon microbial biomass and improved the functional diversity of soil microbial communities, emphasizing the inhibitory effect of 3,4-dimethylpyrazole phosphate on microbial communities when this NI is accumulated in the soil at high doses.

Conclusions
In the present study, the application of fertilizer added with DMPP did not clearly promote mitigation of N 2 O emission, even if a tendency to lower N 2 O fluxes in DMPP plots occurred when soil temperatures were lower than 20 • C. DMPP would seem to inhibit the growth and metabolism of soil microorganisms. The microbial functional diversity is deeply and negatively affected by DMPP when it is accumulated in excessive doses in the soil, thus limiting the growth and the capacity of soil microorganism communities to utilize different substrates. Our study may provide useful information about the correct application of fertilizer with added DMPP nitrification inhibitor. However, data refer to a single agronomical year and further studies are needed to confirm these promising results.