Ammonia Losses, Wheat Biomass, and N Dynamics as Influenced by Organo-Mineral Fertilizer

Abstract

Organo-mineral fertilizers can slow N release to plants, reducing N losses to the environment and enhancing N use efficiency (NUE). Yet, this greater NUE is not always coupled to greater crop yields, which warrants further investigation. Here, we assessed the relationship between N-NH₃ losses from volatilization and wheat (Triticum aestivum L.) biomass and N status. The following treatments were tested: conventional urea (U, 45% N), urea treated with NBPT (N-(n-butyl) thiophosphoric triamide) (U + NBPT, 45.6% N), S-coated urea (U + S; 37% N), Se-coated urea (U + Se; 45% N), organo-mineral fertilizer Azoslow 29 (OMF, 29% N + 50% Azogel^®). The above treatments and non-fertilized control were tested in two soils (LVd and LVAd, 71 and 25% clay, respectively). Semi-open static collectors were used to determine N-NH₃ volatilization 1, 2, 4, 8, 11, 15, 18, 23, 29, and 36 days after application of treatments. Wheat was cultivated for 35 days, and shoot dry mass and total leaf N were determined after harvest. Cumulative N-NH₃ losses from OMF (27 and 32% of N applied in the LVd and LVAd soils, respectively) did not differ from U and (26–32%) and U + Se (24–31%), likely due to organic matter inputs enhancing urease activity in soils. Nevertheless, OMF resulted in 2–4 times greater wheat dry matter than U, U + Se, and U + S, with similar dry mass of U + NBPT for LVAd soils. OMF application enhanced total N removal in wheat leaves relative to the unfertilized control and most N sources. N-NH₃ losses did not reduce biomass yield, but were negatively linked to N accumulation in wheat. The OMF enhanced wheat biomass and nutrition while sustaining environmental quality and promoting circularity in agroecosystems.

1. Introduction

Tropical agriculture is growing in response to the increasing demand for food and feed and economic development of developing countries [1]. This expansion occurs in a scenario of increased synthetic fertilizer use and degradation of natural resources, including soil acidification and loss of biodiversity, water quality issues, and increasing greenhouse gas (GHG) emissions. The recycling of industrial by-products allows for alternative, eco-friendly nutrient sources while mitigating the environmental impacts of landfill waste accumulation. Both efforts (alternative nutrient sources and reduction of environmental waste) are aligned with global initiatives to foster sustainable production patterns [2], enhance circularity in agricultural systems [3], and promote regional food security. This can be particularly relevant for communities with limited access to agricultural inputs and subsidies, oftentimes more susceptible to disruptions in the global fertilizer market, as well as for fertilizer companies seeking to develop more sustainable solutions for plant nutrition.

Raw or treated leather wastes are major by-products of the leather industry (i.e., tanneries) and contain up to 11% protein, which can be hydrolyzed to obtain collagen as a raw material for several industries [4,5,6]. Owing to high N and micronutrient contents, hydrolyzed collagen can be used in organic fertilizers or soil improvers, allowing for the partial displacement of synthetic, fossil fuel-based N fertilizers. The development of alternative N sources is a pressing issue in the agricultural sector, as N is one of the most demanded macronutrients by most crops, with synthetic N-fertilizers accounting for ⅓ of farm operational costs, and causing 2% of global GHG emissions [7]. As such, the use of organo-mineral fertilizers derived from hydrolyzed collagen can enhance agricultural sustainability, potentially reducing production costs and alleviating environmental impacts from synthetic N-fertilizers.

Nitrogen (N) losses due to ammonia (NH₃) volatilization are a major challenge for N use efficiency in agriculture. When surface-applied, conventional urea can have N losses of up to 60% under field conditions, depending on soil type and management practices [8]. Therefore, several fertilizer technologies have been employed to mitigate N-NH₃ losses, including stabilized fertilizers, slow- or controlled-release fertilizers coated with polymers, and urease inhibitors, such as NBPT (N(n-butyl) thiophosphoric amide) [9,10]. Organo-mineral fertilizers based on hydrolyzed collagen have the potential to reduce N-NH₃ losses, as the organic matrix can contribute to a slower and more gradual N release, thus increasing plant uptake and N use efficiency. Yet, knowledge on the impacts of these organo-mineral fertilizers on N-NH₃ losses is scarce, and comparative studies are needed to support their use as an emerging N-fertilizer.

Lastly, the relationship between N-NH₃ losses from various N-fertilizers and crop productivity and nutrition warrants further investigation. Several studies have shown that technologies for fertilizers, including polymer-coated urea or urea treated with urease inhibitors, effectively reduce N-NH₃ losses. However, these lower losses are not necessarily followed by increasing N uptake by plants or higher crop production compared to conventional N sources, such as urea [CO(NH₂)₂] [11,12,13,14]. For instance, Cantarella et al. [15] showcased that urea treated with NBPT reduced N-NH₃ losses by 53%, with productivity increases of 5–12% according to cropping system and soil type. Recently, Santos et al. [9] demonstrated organo-mineral-coated urea reduced N-NH₃ losses by 70% compared to conventional urea, but no gains in foliar N concentration or corn (Zea mays L.) yields were observed. The N-NH₃ volatilization negative impacts on crop yields can be linked to other soil and environmental factors (e.g., light, moisture, and temperature conditions), with N application rates and microbial activity often having an overarching effect on N-NH₃ losses, highlighting the complexity of N dynamics in agroecosystems.

Considering the potential benefits from organo-mineral (OMF) fertilizers, the following hypotheses will be tested in this study: (1) N-NH₃ losses from organo-mineral fertilizers will be lower than synthetic N-fertilizers (e.g., conventional urea) and will vary according to soil type, and (2) lower N-NH₃ losses will contribute to enhanced wheat (Triticum aestivum L.) biomass and N uptake in the OMF-amended plants. The overall goal of this study is to assess the impacts of an organo-mineral fertilizer derived from hydrolyzed collagen [Azoslow 29 (29% N, 50% Azogel^®, hereafter referred to as organo-mineral fertilizer (OMF)], on N-NH₃ losses by volatilization, wheat biomass, and foliar N concentration to further support the use of more sustainable N sources and promote circularity in tropical agriculture. The following N fertilizers were compared: OMF, conventional urea (U, 45% N), urea coated with NBPT (N(n-butyl) thiophosphoric amide) (U + NBPT; 45.6% N), S-coated urea (U + S; 37% N), and Se-coated urea (U + Se; 45% N).

2. Materials and Methods

2.1. Study Location and Soil Characterization

The N-NH₃ volatilization experiment was performed between November and December, 2021, in a greenhouse at the Soil Science Department at Universidade Federal de Lavras, Lavras, Minas Gerais, Brazil (21°13′35″ S, 44°58′43″ W). Temperature (25–32 °C) and humidity conditions were controlled throughout the duration of the experiment. Greenhouse lighting conditions were natural throughout the experiment. The soil samples used as substrate for the N-NH₃ loss experiment were collected from the 0–20 cm depth layer from an area under native Cerrado vegetation (Lavras, Minas Gerais, Brazil). According to the Brazilian Soil Taxonomy [16], the soils were classified as typical dystrophic Red Yellow Latosol (LVAd), with medium texture (25% clay), and typical dystrophic Red Latosol (LVd), with very clayey texture (71% clay). Both soils can be classified as a Typic Haplustox according to the USDA NRCS Soil Taxonomy [17].

After sampling, the soils were air-dried, crushed, and passed through a 4-mm sieve. Then, soil samples were homogenized and stored in plastic bags for further physiochemical analyses. Soil chemical characterization (

Table 1. Chemical characterization of the typical dystrophic Red Yellow Latosol (LVAd) and typical dystrophic Red Latosol (LVd) soils used in the N-NH₃ losses experiment.

Soil †	Chemical Characterization ‡
	pH	SOM	P	K	Ca	Mg	S
	1:2.5 (H2O)	%	mg dm−3
LVd	5.85	<0.2	2.51	19.81	0.46	<0.1	15.27
LVAd	4.81	<0.2	3.15	52.68	0.14	<0.1	2.70
	SB	t	T	V	m	Al	H + Al
	cmolc dm−3			%		cmolc dm−3
LVd	0.61	0.71	1.19	51.00	0	<0.1	0.58
LVAd	0.37	0.47	2.35	16.00	0	<0.1	1.98

† LVd, typical dystrophic Red Latosol (LVd); LVAd, typical dystrophic Red Yellow Latosol; ‡ SOM, soil organic matter; SB, sum of basic cations; t, cation exchange capacity; T, cation exchange capacity at pH 7.0; V, base saturation (%); m, Al saturation (%).

) was performed on the air-dried fine soil, i.e., in the fraction < 2 mm. Soil pH was measured in a soil: water suspension of 1:2.5 (v/v). Soil organic matter (SOM) was determined by Na₂Cr₂O₇ 4 N + H₂SO₄ 10 N oxidation. Exchangeable cations (Ca²⁺, Mg²⁺, and Al³⁺) were extracted using KCl 1 M. Soil P, Na, and K were extracted using the Mehlich-1 solution (0.05 M HCl + 0.0125 M H₂SO₄) in a 1:10 soil: solution (v/v). Soil sulfate was extracted by a solution of Ca(H₂PO₄)₂H₂O 0.01 M and activated carbon, shaking for 5 min followed by 16 h of standing [soil: solution ratio of 1:2.5 (v/v)]. Exchangeable Al was determined by NaOH 0.025 M titration; P and S were determined colorimetrically using a UV/VIS spectrophotometer (B582 model from Micronal, São Paulo, Brazil); K by flame emission photometry (DM-62 model from Digimed, São Paulo, Brazil); and Ca and Mg were determined by atomic absorption spectrometry (AAnalyst 400 model from PerkinElmer, São Paulo, Brazil) [18].

The attributes related to soil fertility (SB, sum of basic cations; t, cation exchange capacity; T, cation exchange capacity at pH 7.0; V, base saturation index; and m, aluminum saturation index) were calculated according to Teixeira et al. [19].

Base saturation was increased to 60% by applying a mixture of Ca and Mg carbonates (p.a.) in a 3:1 molar ratio. Additionally, the soils received basic fertilization, prepared following the fertilization recommendations for pot experiments [20]. A stock solution was prepared using reagent-grade (p.a.) chemicals, including K- and NH₄ phosphates, Cu, Fe, Mn, and Zn sulfates, ammonium molybdate [(NH₄)₆Mo₇O₂₄], and boric acid (H₃BO₃), designed to supply 300 mg dm⁻³ of P, 200 mg dm⁻³ of K, 50 mg dm⁻³ of S, 0.8 mg dm⁻³ of B, 1.5 mg dm⁻³ of Cu, 5 mg dm⁻³ of Fe, 4 mg dm⁻³ of Mn, 5 mg dm⁻³ of Zn, and 0.15 mg dm⁻³ of Mo. Nitrogen was not included in the basic fertilization, as it was provided through the experimental treatments. The stock solution was applied to the dry soil inside plastic bags, which were shaken until the mixture was fully homogenized. Subsequently, approximately 0.5 kg of soil was transferred to pots, which were taken to the greenhouse and maintained until the experiment installation.

2.2. Treatments and N-NH₃ Volatilization Experiment

The greenhouse experiment was set up in a 2 × (5 + 1) factorial with 3 replicates (36 experimental units), corresponding to two soils (LVd and LVAd) receiving five N-fertilizers [conventional urea (U, 45% N), urea coated with NBPT (U + NBPT; 45.6% N); urea coated with S (U + S; 37% N), urea coated with Se (U + Se; 45% N), and the organo-mineral fertilizer (OMF) Azoslow 29 (29% N, 50% Azogel^®)], besides the unamended control (no fertilizer application), in a completely randomized design. The OMF organic matrix utilizes hydrolyzed collagen derived from by-products of the tannery industry (e.g., hides and skins; [18]). The OMF consists of 18% C, 24% inorganic-N, and 5% organic-N.

Soil moisture in the soil pots was kept at 60% of field capacity before the application of the treatments. Fertilizers were hand-applied homogeneously to the soil surface to provide 2 g N per pot. Ammonia was collected using semi-open collection chambers made from plastic bottles (Figure S1). The bottom of each bottle was removed and put on top of the bottles using galvanized baling wire. Inside each collector, a polyurethane foam soaked into 10 mL of acid solution (H₂SO₄ 1 M) was suspended vertically using a rigid wire. The foam served as a trap to capture the volatilized N-NH₃. An 80 mL plastic container was placed at the base of the rigid wire to prevent acid solution leakage and losses (Figure S1). Further details on the construction and assembly of the chambers are described in Araújo et al. [21] and Martins et al. [22]. The chambers were positioned on the soil surface and buried approximately 1 cm into the soil for improved sealing, which allows for consistent N-NH₃ collection under controlled conditions [22]. Wooden sticks and rubber bands were used to improve the fixation of the chambers in the soil. Martins et al. [22] showed that the values of N-NH₃ trapped by the chamber and the N-NH₃ losses measured in the field are linearly correlated (y = 1.66x, R² = 0.99), indicating that the semi-static chambers are a reliable method for N-NH₃ volatilization measurements.

The samples were collected at 1, 2, 4, 8, 11, 15, 18, 23, 29, and 36 days after treatment application. At each collection, the chambers were carefully removed from the soil, the foam was collected and stored in the 80-mL vial, and a new foam soaked in acid solution was placed inside the chamber. Then, the chambers were returned to their respective soil pots, and this procedure was repeated until the end of the evaluation period (36th day). After each collection, the 80 mL vials containing foams soaked in acid solution were taken to the laboratory. The volume was then completed to 50 mL. Aliquots from this extract were transferred to test tubes for distillation by the Kjeldahl method [23], followed by titration with HCl 0.07413 mol L⁻¹ to assess the total N concentration.

2.3. Determination of N-NH₃ Losses

The N values obtained in the laboratory were converted into total N-NH₃ losses according to the following Equation (1):

$$N-NH3={\displaystyle \frac{HCl volume\times 14\times 50\times F\times 0.07143}{aliquot volume}}$$

(1)

where, N-NH₃ is the amount of N-NH₃ volatilized (mg); HCl volume is the volume of HCl 0.07413 mol L⁻¹ consumed per sample during titration (mL); 14 is the N molar mass (g mol⁻¹); 50 is the total volume of the H₂SO₄ solution that collected the N-NH₃ in the chamber (mL); F is the correction factor of the acid used in the titration; 0.07413 is the HCl concentration used in the titration (mol L⁻¹); and aliquot volume is the volume of the H₂SO₄ solution aliquot used in the distillation (mL).

Thus, the daily N-NH₃ losses by volatilization (mg) were obtained per fertilizer treatment and converted into percentage based on the N amount applied via fertilizers (2 g N per pot). To determine the cumulative losses, the daily N-NH₃ losses from the first and second day of collection were summed; these were added to the losses from the fourth day of collection, and so on, until the last sampling date (36th day) [24]. The N-NH₃ values obtained from the semi-static chambers were not calibrated with an external method; as such, it is possible that the measured N-NH₃ values were underestimated for all treatments. Nevertheless, the relative differences among treatments remain valid and should not be affected by this limitation.

2.4. Wheat Cultivation and Harvest

In February 2022, wheat (Triticum aestivum L.), cultivar TBIO Aton, was sown in the soil pots that received the N-fertilizers (November–December, 2021). The pots were randomly displaced on a bench in a greenhouse, following the same experimental design described in Section 2.2 [2 × (5 + 1) factorial with 3 replicates]. Initially, each soil pot received five wheat seeds. Fifteen days after sowing, wheat seedlings were thinned, leaving only three wheat plants per pot.

In the same week, wheat was newly sown in the LVd soil pots. For these soils, seed emergence and seedling growth was initially disrupted due to the formation of a clayey surface crust. As such, the soil surface was gently tilled using a steel trowel to improve soil structure, thus allowing for proper seed germination and radicle development. After 15 days, wheat seedlings were thinned in the LVd soils, leaving only three wheat plants per pot. Daily irrigations kept soil moisture at 60% of field capacity. After 35 days of seedling emergence in the LVd soils, wheat plants were harvested using a garden scissor and split into below- and above-ground biomass [i.e., root and shoot (tillers and leaves)].

Wheat leaves and tillers were triple-washed (distilled water, HCl 10% v/v solution, then distilled water), stored in paper bags, and oven-dried at 65 °C until constant weight (after ±72 h). Then, the dried material was weighed, and a subsample was finely ground in an electric crusher for total N analysis.

2.5. Total N Concentration in Wheat Plants

The total N concentration in shoot dry matter was determined by Kjeldahl distillation [23], followed by titration with HCl 0.07143 mol L⁻¹. The HCl volume consumed during the titration was recorded for each sample and converted into total N concentration according to the Equation (2):

$$Total N=\left[{\displaystyle \frac{\left(0.07143\times 14\times HCl volume\right)}{m}}\right]\times 1000$$

(2)

where, N is the total N concentration in shoot dry matter (g kg⁻¹); 0.07413 is the concentration of HCl used in the titration (mol L⁻¹); 14 is the N molar mass (g mol⁻¹); HCl volume is the volume of HCl 0.07143 mol L⁻¹ consumed per sample during the titration (mL); and m is the mass of the wheat shoot subsample (mg).

2.6. Statistical Analyses

Statistical analyses were performed in the R software v. 4.2.2 [25] using the emmeans package. The assumptions of residue normality and homogeneity of variance were verified by the Shapiro-Wilk and Levene tests (p > 0.05), respectively. Analysis of variance was used to assess the effect of N-fertilizers on daily N-NH₃ volatilization, cumulative N-NH₃ losses, shoot dry mass, and total N concentration in wheat plants. When significant effects were identified, treatments were compared using the Tukey HSD test (p < 0.05). In addition, linear regressions were performed between N-NH₃ losses, wheat shoot dry mass and N concentration to assess potential relationships between N-NH₃ volatilization and plant productivity and nutrition as affected by N-fertilizers.

3. Results

3.1. Daily N-NH₃ Losses by Volatilization

Daily N-NH₃ losses remained mostly similar in the first two days of evaluation (Figure 1a; Table S1). On day 4, the OMF had a 2.22% N-NH₃ loss, higher than the other treatments, and on day 8, it showed a volatilization peak (9.87%; Figure 1a). On day 11, the OMF daily N-NH₃ emission reduced to 4.57%, remaining higher than the losses from the other treatments. On day 15, N-NH₃ loss of conventional urea (5.82%) exceeded those of OMF, which then had losses of less than 2% until the end of the experiment. From day 15 onwards, conventional urea and U + Se had the highest daily N-NH₃ losses, reaching approximately 5% on days 23 and 29 after application of the treatments (Figure 1a; Table S1). U + NBPT had increasing daily losses after 18 days of evaluation, peaking at 5.24% on day 29 but not differing from conventional urea and U + Se. In general, U + S had the lowest daily N-NH₃ losses in the LVd soil, differing from the control only on the last day of evaluation (Table S1; p < 0.05).

As for the LVAd soil, peaks of N-NH₃ losses started 4 days after treatment application (Figure 1b). On day 2, OMF had the highest N-NH₃ loss among N-fertilizers (3.59%; Table S2), and on day 4, it showed a volatilization peak of 10.87%. On day 8, the U and U + Se treatments had peaks of N-NH₃ losses of 9%, surpassing the OMF N-NH₃ loss, which had a sharp reduction in N-NH₃ losses after 8 days (Figure 1b). Between days 8 and 15, U and U + Se had the highest N-NH₃ losses. The peak daily loss of U + NBPT occurred on day 15, not differing from N-NH₃ losses of conventional urea and U + Se (p > 0.05; Table S2). Then, U + NBPT remained as the treatment with highest N-NH₃ losses (1.24–4.2%). Similar to the LVd soil, U + S had the lowest N-NH₃ losses throughout the experiment, differing from the control only on the last day of evaluation.

3.2. Cumulative N-NH₃ Losses as Affected by N-Fertilizers

Due to the increased daily N-NH₃ emissions, OMF had the highest cumulative N-NH₃ losses in the LVd soil from day 4 to the end of the experiment (Figure 2a; Table S3), reaching a cumulative N-NH₃ loss of 27.31%. Conventional urea also had increased N-NH₃ losses, reaching 26.26% at the end of the evaluation period, not differing from OMF (Table S3). Urea coated with S had the lowest cumulative losses at the end of the experiment (2.52%), being similar to the control.

As for LVAd, OMF had the highest cumulative N-NH₃ losses from day 2 until day 15 (Figure 2b; Table S4). On day 18, U + Se reached a cumulative N-NH₃ loss of 28.62%, not differing from OMF (30.04%; p > 0.05). After 23 days, the cumulative N-NH₃ losses from OMF, U, and U + Se reached similar values (nearly 30%). On day 36, U + NBPT reached a cumulative N-NH₃ loss of 30.23%, not differing from OMF, U, and U + Se. Again, U + S was the treatment with the lowest accumulated losses, differing from the control only on the last day of the experiment.

In general, cumulative N-NH₃ losses were greater in the LVAd than in LVd (Figure 3). At the end of the experiment (36th day), conventional urea, OMF, U + Se, and U + NBPT applied to LVAd had the highest cumulative N-NH₃ losses, and the U + NBPT losses in the LVAd soil (30.23%) did not differ from OMF N-NH₃ losses in LVd (26.26%; p > 0.05). Regardless of soil type, U + S application resulted in the lowest cumulative N-NH₃.

3.3. Shoot Drymass and N Concentration in Wheat

An interactive effect of soil type × N-fertilizers was observed for shoot dry mass of wheat plants (p < 0.05). In the LVd soil, wheat plants had lower shoot dry mass than those grown in the LVAd, except for the U and U + S treatments (Figure 4). Shoot dry mass ranged from 0.22 to 0.72 g for LVd, while in LVAd, the mean values ranged from 0.35 to 1.75 g (Figure 4). The lower dry mass observed in LVd was likely a consequence of the delayed seedling emergence caused by crusting in this clayey soil (71% clay content).

In LVd, the highest mean dry mass value was observed for the control treatment (0.72 g), while the lowest value was obtained for the U + S treatment (0.22 g). As for LVAd, OMF led to the highest shoot dry mass (1.75 g), although not differing from U + NBPT (1.34 g; p > 0.05). The application of OMF to LVAd soils increased shoot dry mass by 54% relative to the control, being 2-times higher than conventional urea and U + Se, and 4-times higher than U + S (Figure 4; Figure S2). U + S had the lowest shoot dry mass value for LVAd, with plants being notably smaller than the other treatments (Figure S3).

The soil type × N-fertilizers interactive effect was also significant for the total N concentration (p < 0.05; Figure 5). In general, the mean total N concentration did not differ between plants grown in LVd and LVAd soils, except for the control treatment and for plants that received U + S (Figure 5). In the control treatment, the mean total N concentration in plants grown in LVd (3.8%) was higher than in LVAd (3.0%). In the U + S treatment, the mean total N concentration in plants grown in LVAd (5.9%) was higher than in LVd (4.5%). In both soils, the mean total N concentration in plants that received the U + S treatment was higher than in the control treatment, and the other treatments showed intermediate values. Specifically, the application of OMF increased shoot N concentration in wheat plants by nearly 30% compared to the control in the LVAd soil, not differing from other N sources (p > 0.05).

As a result of increased biomass production, OMF had the highest total N removal across treatments (

Table 2. Total N removal (mg), calculated as total N concentration (%) × shoot dry mass (g), from wheat plants grown in a typical dystrophic Red Latosol (LVd) and a typical dystrophic Red Yellow Latosol (LVAd) receiving the N-fertilizers: Control (without fertilizer application); conventional urea (U; 45% N); NBPT-coated urea (U + NBPT; 45% N); S-coated urea (U + S; 37% N); Se-coated urea (U + Se; 45% N); and the organo-mineral fertilizer (OMF; 29% N).

Treatments	Soils
	LVd	LVAd
Control	27.0 ± 5.7 b †	32.2 ± 5.7 b
U	24.8 ± 5.7 bc	32.0 ± 5.7 b
U + NBPT	10.4 ± 5.7 d	39.9 ± 5.7 b
U + S	9.7 ± 5.7 d	20.8 ± 7.1 bc
U + Se	13.5 ± 5.7 cd	36.7 ± 5.7 ab
OMF	21.5 ± 5.7 bc	68.6 ± 5.7 a

† Means followed by the same letter do not differ by the Tukey test (p > 0.05).

; p < 0.05), although not differing from U + Se in the LVAd. In general, total N removal was greater for LVAd soil than LVd across N sources (Table 2). In the LVd, total N removal was mostly similar among treatments (9.7–24.8 mg), with U + NBPT and U + S showing 1–2-times lower N uptake than OMF (p < 0.05). Similarly, in the LVAd soil, OMF increased N removal by 2-times compared to the control, conventional urea, and U + NBPT, and nearly 3-times relative to the U + S treatment.

3.4. Relationship Between N-NH₃ Losses, Aboveground Biomass, and N Concentration

The shoot dry mass of wheat plants increased with increasing cumulative N-NH₃ losses (Figure 6a), when analyzing data from both soils and excluding the losses from the control treatment (no fertilizer). This becomes more evident when the N-NH₃ losses are analyzed separately by soils (Figure 6b), with improved coefficients of determination (R²) of 0.48for LVd. This trend is reflective of the greater cumulative N-NH₃ losses in the treatments that received the application of OMF and U + NBPT, which also provided a greater shoot dry mass for wheat plants grown in LVAd (Figure 4).

In contrast, the mean total N concentration in the shoots of wheat plants decreased with increasing cumulative N-NH₃ losses (Figure 7). This reduction was observed in the models that included data on N-NH₃ losses and mean total N concentration collected in both soils (Figure 7a), and in the models that included data collected separately in the LVd and LVAd soils (Figure 7b and Figure 7c, respectively). The LVd regression model had a low R² (0.26), possibly due to the low variability of the mean total N concentrations in this soil. The LVAd regression model presented an R² of 0.94, indicating an excellent fit of the model. This fit was probably influenced by the high mean total N concentration in the plants that received U + S application, which also showed the lowest cumulative N-NH₃ losses (2.5% in LVd and 2.9% in LVAd, on average) after 35 days of application.

4. Discussion

Since humidity and temperature in the greenhouse were controlled throughout the experiment, N losses by ammonia (NH₃) volatilization were due to fertilizer and soil properties. In general, the highest peaks of N-NH₃ losses from OMF application (9.87 and 10.87% in LVd and LVAd, respectively) may be related to the organic matrix in its composition. Adding an organic component in soils with low organic matter concentrations (<0.2%) may have stimulated soil urease activity [26]. During hydrolysis, the urea in the OMF decomposes into CO₂ and NH₃, which is then lost through volatilization [27]. Urease activity is highly correlated with organic C [26], which comprises 18% of the OMF used in the present study. Management practices that enhance organic matter levels in soils (e.g., no-till, manure inputs, organo-mineral fertilization) can favor urease activity by providing C as a nutrient source, further enhancing overall microbial activity and N-NH₃ losses [28]. The highest volatilization peaks on the first days of evaluation (day 8 and day 4 in LVd and LVAd, respectively) contributed to the highest cumulative N-NH₃ losses from OMF compared to the other treatments. These volatilization peaks were critical for the cumulative OMF N-NH₃ losses after 36 days, corresponding to nearly 35% of the accumulated losses during the evaluation period. Other aspects of the OMF formulation may have affected N-NH₃ losses, including the larger particle-size compared to the other fertilizers, or the presence of organic N (5%).

The N-NH₃ losses from the other N sources are comparable to those observed in other studies under controlled conditions, highlighting the use of semi-static chambers as a reliable method for N-NH₃ volatilization measurements. Cumulative N-NH₃ losses after conventional urea application in a clayey Oxisol reached 26%, with an 8% loss peak after 4 days [29]. In that study, the SOM concentration was much higher (2.4%) than the soils used in the present study, which may have favored greater urease activity and enhanced N-NH₃ loss peaks after 4 days. Here, the N-NH₃ volatilization peak for urea was 9% and occurred 8 days after application of the treatments in the LVAd. Faria et al. [30] observed a peak N-NH₃ loss of 16% after surface application of conventional urea on the second day of evaluation, and a cumulative loss of 50% after 27 days. However, in that study, the authors used a sandy soil (80% sand) and covered the surface with organic residues of palisade grass [Urochloa brizantha (Hochst. ex A.Rich.) R.Webster]. The addition of organic residues prevents fertilizer incorporation into the soil and provides C as a nutrient source, thus enhancing urease activity [26] and N-NH₃ losses from N-amide sources, such as conventional urea.

Urease activity inhibitors, such as NBPT or Se, can reduce N-NH₃ losses by volatilization in conventional urea [9,30]. However, this was not the case for the present study. In general, N-NH₃ losses for U + Se followed a similar trend to those from conventional urea, with no differences in terms of cumulative N-NH₃ losses. In the study conducted by Faria et al. [30], daily N-NH₃ losses for the U + Se treatment reached 8% after 2 days, with a cumulative loss of 39% at the end of the evaluation period. Here, a peak loss of 9.75% was observed on the 8th day after the application of U + Se in LVAd. This relative delay in the N-NH₃ loss peak was probably due to the lower SOM concentration and lack of organic residues in the soil used in the current study, which may have contributed to maintaining a low urease activity in the soil in the first week of evaluation.

Urea + NBPT delayed the N-NH₃ volatilization peak for LVd, which occurred only 29 days after application of the treatments, resulting in lower cumulative N-NH₃ losses (20.19%) compared to conventional urea (26.26%) at the end of the experiment. Delays in the peaks of N-NH₃ losses are beneficial, as they allow for enhanced N use at the initial stages of crop development. This positive impact, however, was lesser for LVAd, which showed a N-NH₃ volatilization peak 15 days after U + NBPT application. Moreover, cumulative N-NH₃ losses from U + NBPT equaled those of conventional urea at the end of the experiment (Figure 3). Some authors report that the NBPT efficiency as a urease inhibitor can vary between 3 and 14 days, and after this period NBPT loses its inhibitory effect [31], which could explain the reduced efficiency of U + NBPT in the LVAd soil.

Among all treatments, U + S stood out as the one with the lowest N-NH₃ losses. These findings agree with those obtained by Sousa et al. [29], who demonstrated cumulative N-NH₃ losses of 0.4% 16 days after application of U + S. Fertilizers containing water-soluble polymers, such as U + S, gradually release the N-amide contained inside the granules, substantially reducing N losses through NH₃ volatilization. Thus, the N-NH₃ losses from urea coated with S are related to the coating’s dissolution rate. With gradual release, the hydrolysis of urea and the subsequent accumulation of NH₄⁺ and NH₃ on the soil are reduced [32]. It has been demonstrated that even under high temperature (32 °C) and low soil moisture, the urea coated with S substantially minimized N losses [32].

Our findings indicate that N-NH₃ losses do not reduce biomass production in wheat plants under the investigated conditions (i.e., tropical soils and until 36 days). Similar results were observed by Cancellier et al. [11]. The authors demonstrated that different technologies for N-fertilizers, including urea coated with S and polymers, reduced N-NH₃ losses by 37% compared to conventional urea, but did not increase productivity or N concentration in corn (Zea mays L.) plants, which was attributed to soils having increased organic matter levels and N supply, thus preventing a significant impact from the N-fertilizers. Here, the increased dry mass production and N removal observed in the OMF treatment may be attributed to other beneficial properties of this product, not directly measured in this study, such as increased soil nutrient availability, increased soil organic matter content, and greater soil biological activity [5,33]. The lower loss of N-NH₃ from the U + S treatment had no effect on the increase in biomass. The benefits of S-coated urea can influence biomass and productivity in S deficient soils. Shivay et al. (2019) [34] demonstrated that sulfur-coated urea increased rice productivity by up to 25% and improved N use efficiency.

Another potential explanation for the increased biomass after OMF applications includes the N release timing from the fertilizer, which may have better synchronized with plant uptake dynamics, thus leading to greater biomass and N removal. Root dry mass for OMF (0.56 ± 0.08 g) was 2–3 times greater than U + NBPT (0.28 ± 0.08 g), U + S (0.19 ± 0.08 g), and U + Se (0.26 ± 0.08 g), not differing from the control (0.69 ± 0.08 g) or conventional urea (0.39 ± 0.08 g; p = 0.007) when averaged across soils. As such, enhanced root development may have contributed to greater nutrient uptake and above-ground wheat biomass. It is worth mentioning that the biomass results reported in the current study reflect a pot experiment under greenhouse conditions. Pot experiments can limit root development as well as plant growth and nutrient uptake, and results may be different under field conditions.

Other studies have shown lower N-NH₃ losses followed by and/or associated with greater N accumulation in the plant and, therefore, higher N use efficiency. Santos et al. [35] evaluated N-NH₃ losses and the nutritional status of corn plants that received N-fertilizers and showed a trend of reduced foliar N concentration with increased net N-NH₃ volatilization. In a meta-analysis, Abalos et al. [36] showed that nitrification and urease inhibitors increased N use efficiency by 12.9%, but this effect varied according to soil and environmental factors. Some authors, however, found no correlation between lower N-NH₃ losses after the application of N-fertilizer technologies and increased N concentration in the plant [11,13,14]. Instead, the increase in N concentration in corn leaves and grains seems to be better correlated with the increase in N rates applied to the soil, underscoring the relevance of best nutrient management practices for enhanced agronomic efficiency and environmental quality [37].

5. Conclusions

This study assessed the impacts of various N-fertilizers on ammonia (NH₃) losses, wheat biomass production, and N uptake. The cumulative N-NH₃ losses by volatilization resulting from OMF application did not differ from conventional urea and the U + NBPT and U + Se treatments. These higher losses after OMF application are probably linked to the presence of an organic matrix, which may have enhanced urease activity in the soil. Lower N-NH₃ losses following the application of N-fertilizers do not increase dry mass production, but can lead to increases in foliar N concentration in wheat. The U + S treatment, which had the lowest N-NH₃ losses, had the lowest dry mass production and the highest total N concentration. On the other hand, the OMF treatment, which had cumulative N-NH₃ losses comparable to those of conventional urea, U + NBPT, and U + Se, promoted the highest dry mass production when applied in LVAd, not differing from the U + NBPT treatment. In addition, OMF increased the total N concentration in wheat plants compared to the control treatment, showing promise as a sustainable N-fertilizer for highly weathered soils and under the experimental conditions investigated. While these results are encouraging, it should be noted that the effectiveness of OMF may vary depending on soil type, climate, and management practices. Nevertheless, the use of OMF may represent an interesting alternative N- source for wheat production, potentially reducing operational costs and promoting circularity in agricultural systems.