Effects of soil properties on urea-N transformation and efficacy of nitrification inhibitor 3, 4-dimethypyrazole phosphate (DMPP)

ABSTRACT

Ammonia (NH₃) and nitrate (NO₃⁻) are two major sources of N loss during the urea-N transformation process in soil. 3, 4-dimethypyrazole phosphate (DMPP) is a nitrification inhibitor that could effectively limit the NO₃ ^– production process but aggravate NH₃ emissions. NH₃ volatilization, NO₃ ^– production and the efficiency of DMPP are all influenced by different soil factors. The objective of this study was to understand the effects of soil properties on N losses after urea application and DMPP efficacy. A laboratory incubation experiment was set up with urea applied to seven different types of soil at a rate of 0.25 g N/kg soil, with or without DMPP, (at a rate of 1% of the N applied in the fertilizer). Urea content, NH₃ losses and mineral N were measured regularly for 50 days after treatment addition. NO₃^–N production rates, inhibition efficiencies of DMPP and urea half-lives were calculated based on the measured NO₃^–N and urea concentrations. Results showed that urea half-life was negatively correlated with soil total organic carbon (TOC) (r = −0.930). NO₃ ^– production rate was correlated with soil pH (r = 0.670), sand (r = 0.639) and clay content (r = −0.632). The efficiency of DMPP for inhibiting NO₃ ^– production was positively correlated with NO₃ ^– production rate (r = 0.721). NH₃ volatilization was positively correlated with soil pH (r = 0.657) and sand content (r = 0.687) and negatively correlated with urea half-life (r = −0.586). Greater inhibition efficiency of DMPP led to higher NH₃ losses. Urea degradation, NH₃ loss, NO₃ ^– production rate and the inhibition efficiency of DMPP were significantly affected by soil properties and the risk of NH₃ volatilization was increased by DMPP application.

KEY WORDS:

• Urea hydrolysis
• ammonia volatilization
• nitrification
• inhibitor efficiency
• soil properties

1. Introduction

Application of N fertilizer has become the most extensive and effective method of supplying N for crop growth. Urea, a solid N fertilizer, has dominated the N fertilizer market due to its beneficial properties of high N concentration (46%), chemical stability and low price.

Urea is hydrolyzed by the urease enzyme and converted into ammonium (NH₄⁺), which is the most effective N nutrient for plant growth (Zaman and Blennerhassett 2010). However, hydroxyl ions (OH⁻) are also produced during urea hydrolysis. The accumulation of NH₄⁺ and OH ^– promotes ammonia (NH₃) formation resulting in N losses through NH₃ volatilization (Zaman and Blennerhassett 2010). Therefore, the rate of urea hydrolysis not only affects the availability of NH₄⁺ from urea fertilizer but also the rate of N loss in the NH₃ form. Meanwhile, NH₄⁺ is also converted to nitrate (NO₃⁻) in soil through the nitrification process, which is carried out by soil microorganisms using the enzymes ammonium monooxygenase (AMO), hydroxylamine oxidoreductase (HAO) and nitric oxide reductase (NOR) (Ruser and Schulz 2015). NH₃ volatilization and NO₃ ^– leaching are two processes largely responsible for N loss from the soil and also trigger environmental issues such as water eutrophication and soil acidification (Philippot, Hallin, and Schloter 2007).

In order to reduce NO₃ ^– production from nitrification, nitrification inhibitors (NIs) have been widely used in agriculture. NIs can inhibit the first step of the nitrification process (i.e., the conversion of NH₃ to NO₂⁻) by reducing the activity of AMO (Subbarao et al. 2006).

Currently, the most widely used NIs around the world are 3,4-dimethypyrazole phosphate (DMPP), dicyandiamide (DCD) and nitrapyrin. In contrast with DCD and nitrapyrin, DMPP has several advantages, such as lower application rate (1% of the N applied in the fertilizer, about 10 times less than the recommended DCD application rate), less mobility in the soil, less volatility than nitrapyrin and it is easier to blend with fertilizer (Kong et al. 2017, 2016; Wu et al. 2007; Zerulla et al. 2001).

It has been well established that urea hydrolyzation, ammonia volatilization, the nitrification process and the effectiveness of DMPP are strongly related to temperature, moisture, pH, soil organic matter content and soil texture (Barth et al, 2001; Xue et al. 2012; Aijânio et al. 2017).

As well, Zhang, Wu, and Zhou (2004) found that peat humus could effectively adsorb DCD and prolong the nitrification inhibition effect by protecting it from biodegradation and excessive leaching. Thus, it is likely that soil humus could also affect the efficacy of DCD. However, DMPP may also be affected by soil humus. Soil humus accounts for 60–90% of soil organic matter and can be divided into humic acid (HA), fulvic acid (FA) and humin (HM). HA and FA are the two most active parts of humus and mainly account for its physical and chemical characteristics (Dou et al. 2008). To date, few studies have focused on the effects of different components of soil humus on DMPP effectiveness.

In addition, the effects of single soil factors on urea hydrolyzation, NH₃ volatilization, DMPP efficacy and the nitrification process have been examined (Aijânio et al. 2017; Barth et al, 2001), but there has been no comprehensive study into the effects of multiple soil physical and chemical properties on urea transformation, N losses or DMPP efficiency.

Thus, in this study, soil humus components (e.g., HA and FA) were extracted and measured and an experiment was conducted to comprehensively evaluate the correlations between soil properties, urea hydrolyzation, ammonia volatilization, nitrification rate and the inhibitory effect of DMPP. Furthermore, an understanding of soil factors influencing urea transformation, N losses and DMPP effectiveness may help in the development of a new fertilizer additive to reduce N losses and/or enhance DMPP efficacy.

2. Materials and methods

2.1. Soil sampling and preparation

Soil samples, which included seven soil types according to Chinese soil classification: Red soil (RS), Black soil (BS), Brown soil (BrS), Yellow soil (YS), Lime concretion black soil (LS), Chernozem (Che) and Desert soil (DS), were collected from six provinces in China during March to early May 2017 (Figure 1). All sites had not received any N inputs since the end of 2012. Ten sampling plots were randomly set up in each sampling location and a total of 100 kg of bulk soil was collected from the upper 20 cm of these sampling plots (about 10 kg per plot). The samples were mixed thoroughly, air dried and then passed through a 2 mm sieve to remove large stones, plant roots and residues. Sampling locations and soil physical and chemical properties are shown in Table 1 and Figure 1.

Table 1. Characteristics of the soils used in the study

Soil type	pH	Soil Texture (%)			EC (μS cm^−1)	TOC (g kg^−1)	TN (g kg^−1)	HA (g kg^−1)	FA (g kg^−1)	Catalase activity (0.1 N KMnO4 ml g^−1 soil)
		Sand	Silt	Clay
Black soil (BS)	6.15	7.3	59.7	33	61	16	1.6	3.5	4.5	18.2
Brown soil (BrS)	6.4	4.7	63.3	32	34	19	1.5	3.1	5.8	16.9
Red soil (RS)	5.94	5.3	51	44.7	30	6.6	0.7	2.0	1.4	18.5
Desert soil (DS)	8.25	24.3	55	20.7	147	13	1.7	2.1	3.1	19.3
Yellow soil (YS)	7.69	23	44.3	32.7	48	9.2	1.1	1.9	2.5	18.6
Chernozem (Che)	8.39	62	20	18	78	12	1.5	2.1	3.7	18.3
Lime concretion black soil (LS)	8.18	15.3	73.7	11	92	11	1.4	2.1	3.5	18.0

EC represents electrical conductivity. TOC represents soil total organic carbon. TN represents soil total nitrogen. HA represents humic acid. FA represents Fulvic acid.

Figure 1. Sampling locations in China. Che: chernozem (site location: N 44°51′10″; E 125°11′3″); BS: black soil (N 43°49′3″ N; E 125°24′5″); BrS: brown soil (N 41°16′17″ N; E 123°26′18″); LS: lime concretion black soil (N 35°5′16″ N; E 115°48′41″); YS: yellow soil (N 31°42′11″; E 116°57′13″); RS: red soil (N 29°48′30″; E 114°18′15″); DS: desert soil (N38°56′11″; E 100°27′38″).

2.2. Experimental design

An aerobic soil incubation was conducted over a 50-day period at a constant temperature of 25°C. Air-dried soil (equivalent to 400 g oven-dry soil) was placed into clean cylindrical polystyrene containers (10 cm in diameter and 10 cm in height), adjusted to approximately 40% water-filled pore space (WFPS) with deionized water and then pre-incubated at 25°C for 1 week to recover soil microorganism activity. After pre-incubation, each soil sample was divided equally into four parts (approximately 100 g each). 10 ml of a solution containing 0.1 g N (0.213 g urea) (U) or a mixture of 0.1 g N and 0.001 g DMPP (UD) was homogeneously sprayed onto each of these portions and then mixed and adjusted to 60% WFPS, which is the optimum WFPS for nitrification to occur (McGeough et al. 2016). Each treatment had six replicates, seven soils resulting in a total of 84 containers.

Three containers for each treatment were prepared for NH₃ collection by being placed into volatilization chambers (15 cm in height and diameter) where a small beaker containing 10 ml of 20 g L⁻¹ boric acid solution and pH indicators (0.044 g L⁻¹ methyl red with 0.066 g L⁻¹ bromocresol green) was used to trap NH₃ (Katukurunda et al. 2013; Soares, Cantarella, and Menegale 2012). The boric acid solution was removed daily and analyzed for NH₃ content until NH₃ could no longer be detected. The other three containers were used to periodically measure urea, NH₄⁺ and NO₃ ^– concentrations in the soil. All of the containers were randomly placed in the incubator. The soil in each container was checked periodically for water loss and, if required, adjusted with deionized water to maintain a WFPS of 60% over the 50-day incubation period. There were 12 sampling occasions (1, 3, 5, 10, 15, 20, 25, 30, 35, 40, 45, and 50 days after treatment application) during the incubation period and 2 g soil was sampled at each sampling occasion for texting the concentrations of NH₄⁺-N and NO₃^–N. Urea concentration was determined initially and then until it could no longer be detected. After each sampling, the incubation containers were placed back into the incubator randomly.

2.3. Analytical methods

Soil total organic carbon (TOC) and total nitrogen (TN) were determined using Total Organic Carbon and elemental analyzer ((TOC-V _CPH; Elementary Vario RL III) respectively.

Soil catalase activity was measured based on the method of Guan (1987); briefly, 2 g air-dried soil (< 2 mm) was shaken with 40 ml distilled water and 5 ml 0.3% hydrogen peroxide (H₂O₂) for 20 mins. 5 ml of 3 N H₂SO₄ was then added and mixed. The sample was then filtered through qualitative filter paper and 25 ml of the filtrate was titrated with 0.1 N KMnO₄.

Soil textures were determined based on the method of Kettler, Doran, and Gilbert (2001), 15 g soil (< 2 mm) was shaken with 45 ml of 3% sodium hexametaphosphate (HMP) for 2 hours, then passed through a 0.053 mm sieve to collect sand (on the sieve). The sieved solution (silt and clay) was stirred thoroughly, then settled undisturbed for a sedimentation period of 6 hours. After the sedimentation period, the suspended clay fraction was discarded. The settled silt fraction and the sand on the sieve were then dried at 105°C to constant weight. The sand and silt contents were calculated, and the clay content was determined by subtracting the sum of the sand and silt weights from the original sample weight. Soil pH and conductivities were determined using a pH-meter (S210 SevenCompact^TM, Mettler, Germany) with a soil:water ratio of 1:2.5 (Bao et al. 1999).

Soil HA and FA were extracted according to a method modified from Brunetti et al. (2007). Briefly, a solution of 0.1 M Na₄P₂O₇ and 0.1 M NaOH was mixed with soil at a ratio of 10:1 (solution to soil). The mixture was shaken horizontally in a water-bath at 72°C for 2 hours at a rate of 145 r min⁻¹, centrifuged at 9600 g for 30 minutes, and then filtered. 5 ml of the filtrate was acidified with 2 M H₂SO₄ to pH 1 (to separate HA and FA). After chilling for 12 hours the mixture was filtered again. The FA passed through and HA was retained on the filter paper. The HA was dissolved in 0.5 M NaOH and the concentrations of HA and FA were determined using a Total Organic Carbon Analyzer (TOC-V _CPH).

Soil NH₄⁺-N and NO₃^–N were extracted from the soil with 2 M KCl and determined by the MgO distillation method (Bao et al. 1999). Urea concentrations were determined by the diacetyl monoxime method after extraction from the soil samples with 2 M KCl (Seaton and Ali 1984). Briefly: 1 g of soil was shaken with 10 ml of 2 M KCl for 60 minutes at a shaking rate of 120 r min⁻¹ and filtered. The Filtrate (0.4 ml) was diluted to 10 ml with deionized water then a 0.4 ml aliquot was mixed with 6 ml of color reagent (diacetyl monoxime combined with thiosemicarbazide. Urea reacts with diacetyl monoxime under acid conditions to give a yellow condensation product; thiosemicarbazide intensifies this reaction), and placed in a boiling waterbath for 15 minutes. After removal from the waterbath, the samples were rapidly cooled in iced water to room temperature. Absorbance was measured on a spectrophotometer at 535 nm wavelength.

The amount of NH₃ volatilized was determined by a titrating the boric acid trap solutions with 0.005 M H₂SO₄ (Katukurunda et al. 2013)

2.4. Calculations and statistical analysis

Traditionally, nitrification rate has been calculated using the following formula: NO₃^–N content/(NH₄⁺-N + NO₃^–N). However, previous research has indicated that not all NH₄⁺-N can be extracted by KCl solution. The non-extractable NH₄⁺ can also be a source of available N to nitrifiers under certain conditions (Feigenbaum et al. 1994). Additionally, NH₃ volatilizations significantly affect soil NH₄⁺ content as well. The conventional formula NO₃^–N content/(NH₄⁺-N + NO₃^–N) may cannot exhibit nitrification rate accurately. Thus, in this study, the average NO₃ ^– production rate was used to represent the nitrification rate. NO₃ ^– production rates are calculated as: (final NO₃^–N content – initial NO₃^–N content)/incubation time. The inhibitory rates are calculated as: (NO₃ ^– production rates in without DMPP treatment – NO₃ ^– production rates in with DMPP treatment)/(NO₃ ^– production rates in without DMPP treatment).

The hydrolyzation of urea in the soils was quantified using regression analysis including 95% confidence limits (P < 0.05). A first-order degradation model (C_t = [k*ln(t) + A)] was applied to the urea concentration data over time, where the constant k is the first order rate constant for urea decomposition. The time taken for the urea concentration to decline to half its initial concentration (i.e., the concentration determined on the first day of incubation) is termed the half-life (t_1/2).

An ANOVA analysis was used to determine the significance of the effects of the different treatments on urea hydrolyzation, NH₃ volatilization and soil NH₄⁺ and NO₃ ^– concentrations in the different soil types. A significance level of 0.05 was used unless otherwise stated. Estimates of means, standard errors of means, standard errors of differences in means and Fisher’s least significant difference (LSD) were calculated using SPSS 19.0. Pearson correlation analysis using Sigmaplot 12 was performed with urea half-life, NH₃ volatilization, NO₃ ^– production rate, inhibitory rate of DMPP and all soil physical and chemical properties.

3. Results

3.1. Urea hydrolyzation

Urea was completely hydrolyzed after day 4 with U and UD treatments in BS and BrS, which exhibited faster hydrolyzation rates than the other soil types (Figure 2). The lowest hydrolyzation rate of urea was found in RS (urea was no longer found after day 8). DMPP application did not show a significant effect on the hydrolyzation rate in any soil type.

Figure 2. Urea concentration in the soil during the incubation period. Vertical bars are standard errors of the mean (n = 3). BS: black soil; BrS: brown soil; RS: red soil; DS: desert soil; YS: yellow soil; Che: chernozem; LS: lime concretion black soil. U:ureaonly;UDureawithDMPP.

The half-lives of urea are shown in Table 2. For both treatments (U and UD), the shortest half-life of urea was found in BrS and the longest half-life was found in RS. DMPP application significantly prolong the half-life urea in BrS by 0.08 day (P < 0.05), while this effectiveness did not observed in other soil types.

Table 2. Average half-lives (days) of urea with different treatments in seven soil types

Soil types	Treatments
	U	UD
BS	1.92 (0.07)	1.91 (0.06)
BrS	1.74 (0.03)	1.82 (0.03)
RS	2.82 (0.10)	2.82 (0.14)
DS	2.12 (0.01)	2.10 (0.05)
YS	2.53 (0.12)	2.56 (0.06)
Che	2.24 (0.08)	2.25 (0.11)
LS	2.09 (0.05)	2.11 (0.08)

U: urea only; UD: urea with DMPP; BS: black soil; BrS: brown soil; RS: red soil; DS: desert soil; YS: yellow soil; Che: chernozem; LS: lime concretion black soil; Values in brackets are standard errors of the mean (n = 3).

3.2. Dynamics of NH₄⁺-N and NO₃^–N concentrations in different soil types

The concentration changes in NH₄⁺ and NO₃ ^– after urea application with or without DMPP in the test soils are shown in Figure 3. Results showed that N remained in the NH₄⁺ form for longer periods in RS than in the other soil types.

Figure 3. Changes in soil mineral N concentrations. Vertical bars are standard errors of the mean (n = 3). BS: black soil; BrS: brown soil; RS: red soil; DS: desert soil; YS: yellow soil; Che: chernozem; LS: lime concretion black soil. U: urea only; UD urea with DMPP.

Application of DMPP significantly increased NH₄⁺ concentrations and prolonged the retention of N in the NH₄⁺ form, in DS, Che and LS, but not in the other soil types (Figure 3). Changes in NO₃^–N over time showed that less NO₃ ^– was produced when DMPP was applied in all soil types except in RS. The NO₃ ^– production rate and the inhibitory rate of DMPP also varied among the different soil types (Table 3), the highest and lowest NO₃ ^– production rates were 3.8 and 0.8 mg kg⁻¹ day⁻¹ in LS and RS, respectively. The strongest inhibition effect of DMPP on NO₃^–N production rates was observed in DS achieving 53.6%, while the weakest was only 2.6% in RS.

Table 3. Average NO₃^–N production rates (mg kg⁻¹ d⁻¹) in each soil type under the urea and DMPP treatments

Soil type	U	UD	Inhibition by DMPP (%)
BS	1.3 ± 0.07	1.1 ± 0.04	14.5 ± 6.8
BrS	2.3 ± 0.13	1.8 ± 0.03	20.3 ± 3.2
RS	0.8 ± 0.06	0.7 ± 0.07	2.6 ± 2.2
DS	2.0 ± 0.06	0.9 ± 0.06	53.6 ± 2.0
YS	1.3 ± 0.05	0.9 ± 0.10	31.2 ± 7.5
Che	3.1 ± 0.26	1.6 ± 0.17	49.0 ± 10.0
LS	3.8 ± 0.08	2.1 ± 0.11	43.0 ± 3.9

U: urea only; UD: urea with DMPP; BS: black soil; BrS: brown soil; RS: red soil; DS: desert soil; YS: yellow soil; Che: chernozem; LS: lime concretion black soil; Values are reported as mean ± standard deviation (n = 3).

3.3. The NH₃ volatilization rate and total amount volatilized

NH₃ volatilization rate exhibited similar trends to NH₄⁺ dynamics in all soil types, rapidly increasing at initially and then decreasing (Figure 4). The timing of the occurrence of volatilization peaks and the cumulative amounts of NH₃ emitted exhibited differences in the different soil types. The earliest NH₃ volatilization peak was found in LS on day 3, and the latest was on day 7 in RS. DMPP application (UD) did not affect the occurrence time of peaks.

Figure 4. Daily NH3 flux from different treatments in seven soil types. Vertical bars are standard errors of the mean (n = 3). BS: black soil; BrS: brown soil; RS: red soil; DS: desert soil; YS: yellow soil; Che: chernozem; LS: lime concretion black soil. U:ureaonly;UDureawithDMPP.

The cumulative NH₃ losses also varied with the different soil types. The order from high to low of cumulative NH₃ volatilization is Che > LS > BrS > DS > BS > YS > RS (Table 4). DMPP application significantly increased NH₃ emissions. In contrast with the U treatment, the UD treatment increased NH₃ losses by 24.8% in DS, which was the greatest effect on NH₃ losses. The smallest effect was found in YS (6.3%) (Table 4).

Table 4. Cumulative NH₃ volatilization (mg N kg⁻¹ soil)

Soil types	Treatments		Aggravated by DMPP (%)
	U	UD
BS	15.6 (0.17)	18.1 (0.35)	19.4 (0.99)
BrS	16.4 (0.04)	19.0 (0.28)	15.9 (1.46)
RS	12.5 (0.11)	14.4 (0.17)	15.2 (2.25)
DS	16.1 (0.18)	20.1 (0.12)	24.8 (1.03)
YS	14.4 (0.24)	15.3 (0.16)	6.3 (0.73)
Che	18.7 (0.13)	22.2 (0.27)	18.7 (2.21)
LS	18.3 (0.20)	22.6 (0.11)	23.5 (1.04)

U: urea only; UD: urea with DMPP; BS: black soil; BrS: brown soil; RS: red soil; DS: desert soil; YS: yellow soil; Che: chernozem; LS: lime concretion black soil; Values in brackets are standard errors of the mean (n = 3).

3.4. Relationships between soil characteristics and urea half-life, NH₃ volatilization, nitrification rate and inhibitory rate of DMPP

The Pearson correlation coefficients, r, and associated probabilities, P, between soil characteristics and urea half-life, NH₃ volatilization, nitrification rate and inhibitory rate of DMPP, are summarized in Table 5: Of the total 9 soil variables assessed, 4 of them had a statistically significant correlation with the NO₃ ^– production rate. These variables included pH (r = 0.670, P < 0.01), sand content (r = 0.639, P < 0.01), clay content (r = −0.632, P < 0.01) and EC (r = 0.774, P < 0.01). The nitrification inhibition rate of DMPP was positively correlated with soil pH (r = 0.931, P < 0.01), sand content (r = 0.611, P < 0.01), EC (r = 0.666, P < 0.01) and catalase activity (r = 0.670, P < 0.01) and negatively correlated with clay content (r = – 0.920, P < 0.01). Urea half-life was negatively affected by TOC (r = −0.930, P < 0.01), HA content (r = −0.718, P < 0.01), FA (r = −0.924, P < 0.01) and catalase activity (r = −0.529, P < 0.05). The amount of NH₃ volatilized was positively correlated with pH (r = 0.657, P < 0.01), sand content (r = 0.687, P < 0.01), FA content (r = 0.504, P < 0.05), EC (r = 0.639, P < 0.01), TOC (r = 0.439, P < 0.05) and negatively correlated with clay content (r = −0.711, P < 0.01).

Table 5. Correlation coefficients (r) between average NO₃^–N production rate, % inhibition rate of NO₃ ^– production by DMPP, half-life of urea, cumulative NH₃ volatilization, % NH₃ losses aggravated by DMPP and soil properties

Soil variable	Rate of net NO3 ^– production	% Inhibition of NO3 ^– production by DMPP	Half-life of Urea	Cumulative NH3 volatilization	% NH3 losses aggravated by DMPP
pH	0.670**	0.931**	−0.029	0.657**	–
Sand	0.639**	0.611**	−0.006	0.687**	–
Silt	−0.238	−0.075	−0.395	−0.229	–
Clay	−0.632**	−0.920**	0.393	−0.711**	–
HA	−0.129	−0.396	−0.718**	−0.076	–
FA	0.317	0.110	−0.924**	0.504*	–
EC	0.774**	0.666**	−0.249	0.639**	–
TOC	0.229	0.068	−0.930**	0.439*	–
Catalase Activity	0.237	0.670**	−0.529*	0.420	–
Cumulative NH3 volatilization	0.931**	0.721**	−0.586**	–	–
% Inhibition of NO3 ^– production by DMPP	0.679**	–	−0.232	0.721**	0.497*
Rate of net NO3 ^– production	–	0.679**	−0.452*	0.931**	–

* P < 0.05; ** P < 0.01; *** P < 0.001.

As well as the effects of soil factors on nitrification rate, urea hydrolyzation and NH₃ volatilization, strong relationships were also found between other variables. The DMPP inhibitory rate was positively correlated with NO₃ ^– production rate (r = 0.679, P < 0.01). The cumulative NH₃ losses was positively correlated with the inhibitory rate of DMPP (r = 0.721, P < 0.01) and negatively correlated with urea half-life (r = 0.679, P < 0.01). No significant relationship was found between urea half-life and DMPP inhibition efficiency, whereas NO₃^–N production rate was negatively correlated with urea half-life (r = −0.452, P < 0.05) (Table 5).

4. Discussion

In this experiment, temperatures and moisture contents were controlled, thus, differences in NO₃ ^– production rate, DMPP inhibitory rate and urea degradation rate (half-life) between different soil types were mainly caused by the different soil physical and chemical characteristics.

4.1. Effects of soil characteristics on urea hydrolyzation

In this study, urea half-life was calculated and used to describe the urea hydrolyzation rate. Results showed that TOC and catalase activity were the only two soil factors (HA and FA are two components of TOC) which significantly reduced urea half-life (Table 3). Urease is the only enzyme that can hydrolyze urea in soil, thus, the urea hydrolyzation rate is determined by the activity of urease in soil (Zaman and Blennerhassett 2010). The urease enzyme is produced by soil microbes (Kot, Zaborska, and Orlinska 2001), higher TOC could lead to higher soil microbial activity and catalase is an oxidoreductase commonly found in microorganisms (Zhang et al. 2005). Thus, Higher TOC content and catalase activity could increase urease activity, and lead to a shorter urea half-life.

4.2. Effects of soil characteristics on Nitrification rate

Microbial oxidation of ammonia is a rate limiting step in the nitrification process and dominates the NO₃ ^– production process (Jia and Conrad 2009; McGeough et al. 2016). Soil and other environmental factors could affect the NO₃ ^– production rate by influencing microbial ammonia oxidation.

Ammonia-oxidizing bacteria (AOB) and ammonia oxidizing archaea (AOA) are two primary groups of microorganisms responsible for the ammonia oxidation process (Purkhold et al. 2000; Wakelin et al. 2014)). Phylogenetic analysis has shown that most operational taxonomic units (OTUs) of AOB are affiliated with the Nitrosospira cluster 3 and AOA are affiliated with Nitrosomonas cluster 7 (Ouyang et al. 2016). Nitrosospira cluster 3 often outcompetes other Nitrosospira (Tourna, Freitag, and Prosser 2010). Thus, the oxidizing ability of AOB on ammonia is stronger than that of AOA and dominates the nitrification process. Many other studies have also verified this relationship (Jia and Conrad 2009; Shen et al. 2008). In the current study, the lowest NO₃ ^– production rate was found in RS, which had the lowest pH value and soils with higher pH values (LS and Che) exhibited higher NO₃ ^– production rates (Table 2). Correlation analysis showed that the NO₃ ^– production rate was positively correlated with soil pH (r = 0.821, P < 0.05). These results align with the effect of soil pH on the activity of AOB, suggesting that the dominant nitrification microorganisms in the test soils were AOB. Previous research has indicated that pH could strongly affect the compositions and activities of AOB and AOA; AOB are most suited to alkaline conditions, and conversely, AOA are more suited to acidic conditions (De Boer and Kowalchuk 2001; Hu, Xu, and He 2014; Hu et al. 2013; Jia and Conrad 2009; Shen et al. 2008; Zhang et al. 2012).

In addition, the substrate for the AMO enzyme is ammonia rather than ammonium (McCarty 1999). The high H⁺ concentration under acidic conditions could restrict ammonium conversion to ammonia, this restriction could directly reduce the substrate concentration of AMO (Hu, Xu, and He 2014; Li et al. 2018). This could be another reason for a lower NO₃ ^– production rate and longer NH₄⁺ storage time under low pH conditions, as found in RS in this study (Figure 2).

The results of this study suggest that soil textural factors such as sand and clay contents can significantly affect NO₃ ^– production rate. Higher NO₃ ^– production rates were observed in Che and DS, which are two soils with lower clay content. This is mainly due to the fact that the nitrification process is an aerobic biological process, which needs a high oxygen concentration. The sand: clay ratio affects soil aerobic – anaerobic processes by changing the soil aeration status and redox potential. Low oxygen concentration and redox potential would occur when there is high clay content, retarding the nitrification process (Benckiser, Schartel, and Weiske 2015; Yu et al. 2007). Furthermore, high clay content could result in high cation exchange capacity (CEC) and strong NH₄⁺ sorption ability (McGeough et al. 2016; Yu et al. 2007), thus decreasing the amount of substrate for nitrifiers.

There was a positive relationship between soil EC and NO₃ ^– production rate (Table 3). Soil EC indicates the ability for electron transfer and the concentration of soluble salts in the soil. The ammonia oxidation process is a microbial oxidation – reduction process which needs electrons. A highly conductive environment could ensure a sufficient supply of electrons to support the ammonia oxidation process (Benckiser, Schartel, and Weiske 2015). In addition, in this study, soil EC exhibited a significant positive relationship with soil pH (Table 4). This result was consistent with previous research and may indicate a possible indirect effect of soil EC on NO₃ ^– production rate by influencing soil pH.

4.3. Impact of soil properties on inhibitory rate of DMPP

A significantly positive correlation was found between the inhibitory rate of DMPP and soil pH (r = 0.881, P < 0.05). The major reason for this relationship could be summarized as follows: Firstly, DMPP has a specific target on AOB (Benckiser, Schartel, and Weiske 2015; Kong et al. 2016; Müller et al. 2002), which dominate the nitrification process under alkaline conditions, and poor inhibitory function on AOA, which play a leading role in nitrification under acidic conditions (Benckiser, Schartel, and Weiske 2015; Kleineidam et al. 2011; Shi et al. 2016). So, DMPP could be more effective in high soil pH conditions than in low. Secondly, DMPP is a weak alkaline salt, thus the high hydroxyl (OH⁻) concentration in soil with high pH could neutralize the phosphate in DMPP and accelerate the hydrolysis of DMPP to 3, 4-dimethypyrazole (DMP), which is the actual functional chemical group that inhibits the ammonia oxidizing process (McCarty 1999).

Results shown in Table 5 suggest that soil texture could play an important role in affecting DMPP inhibitory rate. There would be a strong nitrification potential in coarse soils, and the inhibitory function of DMPP would be amplified by this strong nitrification potential. However, the binding behavior of DMPP to clay would reduce the amount of active DMPP in the soil environment, and thus decrease the contact opportunity between DMPP and soil nitrifiers (Barth et al. 2001). Accordingly, the higher content of clay negatively affected the inhibition rate of DMPP on NO₃ ^– production.

As discussed above, strong soil EC could facilitate the nitrification process, resulting in a faster NO₃ ^– production rate, which in turn would amplify the inhibitory function of DMPP. The positive correlation between soil EC and DMPP inhibitory rate found in this study (Table 5) supports this suggestion.

Barth et al. (2001) observed that catalase and DMPP bound at similar sites on soil particle surfaces. Thus, a competitive relationship may exist between catalase and DMPP. For example, a higher catalase activity may result in greater competition for soil binding sites leaving more DMPP free in the soil environment, thus weakening the physical protection from soil particles and increasing the contact opportunity between DMPP and soil nitrifiers.

4.4. Effects of soil characteristics and DMPP application on cumulative NH₃ volatilization

The higher soil pH and NH₄⁺ accumulation are two important factors for NH₃ volatilization (Zaman and Blennerhassett 2010). NH₄⁺ and OH ^– ions are co-produced during urea hydrolysis and result in an increase in soil pH, faster urea degradation (shorter half-life) will accelerate the accumulation of OH ^– and NH₄⁺ and stimulate NH₃ emissions. Thus, in this study, the cumulative NH₃ volatilization was negatively correlated with urea half-life and positively correlated with soil pH and TOC. Clay could restrict NH₃ emissions by physical protection of NH₄⁺.

DMPP application significantly increased NH₃ volatilization, the percentage of NH₃ lost was positively correlated with the inhibition rate of DMPP (Table 3, Table 4). This was mainly due to DMPP increasing soil NH₄⁺ concentration and pH by inhibiting the nitrification process, which is an acidifying process (Zaman et al. 2009).

4.5. Correlation between urea hydrolyzation, nitrification rate and DMPP inhibition efficiency

In this study, significant inhibition rates of DMPP were found in DS, Che, LS. Significantly increased effects on DMPP on NH₄⁺ were also found in DS, Che, LS, as well (Figure 3), this result mainly caused by the inhibitory effect of DMPP on nitrification process.

Nitrification rate can be affected by increased soil NH₄⁺ concentration in suitable range, and shorter urea half-life means a faster NH₄⁺ accumulation rate, thus, the nitrification rate was negatively affected by urea half-life (Table 5). A greater nitrification potential will lead to greater nitrification inhibition efficiency (McGeough et al. 2016), therefore, a positive relationship was found between nitrification rate and DMPP inhibition efficiency (Table 5).

5. Conclusion

Results in this study indicate that urea hydrolyzation rate is strong positively correlated with soil organic carbon and catalase activity. NH₃ losses from urea fertilizer were positively correlated with soil pH, sand content, organic carbon content and urea hydrolyzation rate. NO₃ ^– production was significantly affected by soil pH, EC and sand and clay content. DMPP effectively inhibited the nitrification process but showed different inhibitory rates in different soils. DMPP efficacy was generally high in soils with high pH, sand content and EC, and low in soils with high clay content. DMPP application inhibited the nitrification process but aggravated NH₃ losses, and a greater inhibition rate resulted in higher NH₃ emissions. Using this knowledge, further study could lead to the development of measures to reduce the NH₃ losses caused by DMPP and enhance its efficacy.

Acknowledgments

This project was funded by the Development and Industrialization of New Stabilized Compound Fertilizer project (2017YFD0200708-2). The authors would like to express thanks to Hanwen Liu, Guohui Wu and Fang Ding from the Institute of Applied Ecology, Chinese Academy of Science for assistance during soil collection and to Professor Song Guan from Jilin Agricultural University for assistance with technical support for measuring soil humic substances.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This work was supported by the Chinese Academy of Sciences [2017YFD0200708-2].