Replacing Mineral with Organic Fertilisers in Maize Basal Fertilisation: Impacts on GHG Emissions and Yield

Abstract

Portuguese farmers seek evidence that organic fertilisers, particularly manure-based ones, can be safely used as partial replacements for mineral fertilisers (MFs), taking advantage of their nutrient and organic matter (OM) content. This study aimed to clarify the effects of applying organic fertilisers, especially under no-till practices in maize production. The experiment involved replacing basal mineral nitrogen (N) fertilisation with composted animal manure (CM) and pig slurry (PS) across three contrasting fields with varying soil characteristics, tillage techniques, and OM supplementation. Results indicated that site 1, which had the lowest clay and OM content, benefited the most from organic fertilisers, particularly in enhanced nutrient absorption in both maize leaves and grain. In this site, partial nutrient balance of N was significantly higher in the PS treatment (0.56 kg grain N exported kg N applied⁻¹) compared with the treatment with only MF (0.44). The impacts on greenhouse gas (GHG) emissions depended on site characteristics. CM led to higher emissions at site 2 (CH₄) and site 3 (N₂O), while PS did not increase GHG emissions at any site. Under no-till practices at site 3, CM resulted in higher global warming potential (154 kg CO_2-eq ha⁻¹) compared with the PS (128 kg CO_2-eq ha⁻¹) and MF (109 kg CO_2-eq ha⁻¹) treatments. Overall, this study suggests that organic fertilisers, particularly PS, can be a viable strategy for improving soil health and maintaining sustainable maize production in Portuguese agriculture. However, their effectiveness depends on factors such as soil texture, OM content, and tillage practices. In soils with lower OM content or under no-till practices, pig slurry emerges as a suitable alternative, replacing 30% of total mineral N fertilisation without compromising productivity or increasing GHG emissions.

1. Introduction

The specialisation of livestock production and crop production resulted from increased agricultural intensification aimed at meeting higher food demand due to growing global population and changing dietary patterns [1]. However, this specialisation has also created two new problems. Firstly, the fertilisation of agricultural crops relies almost exclusively on mineral fertilisers (MF), with little or no incorporation of organic matter (OM) [2], leading to soil impoverishment [3]. Secondly, specialised and confined animal production generates large amounts of manure, particularly slurry, which can have substantial environmental impacts if not appropriately managed [4,5]. Inadequate manure management can also result in significant ammonia (NH₃) and greenhouse gas (GHG) emissions, namely, nitrous oxide (N₂O), methane (CH₄), and carbon dioxide (CO₂), thereby directly contributing to climate change and decreasing air quality [4,6,7].

These two issues present an opportunity to be addressed simultaneously. Utilising animal manure as fertiliser in agriculture could reduce production costs, especially given the recent rise in MF prices, and enhance soil organic matter (SOM) content, which is particularly important in Mediterranean climates [8,9]. Additionally, it could promote a circular economy by using locally produced forage and crops for animal feed.

The use of animal slurry and manure, whether raw or composted, as organic fertilisers (OF) is once again increasing, albeit slowly. These fertilisers provide a significant nutrient supply for crops across the European Union (EU), though their application is more concentrated in regions with high livestock densities, such as Germany, the Netherlands, and Denmark [10,11]. However, their use is predominantly in organic farming. For instance, in Italy—the European country with the second largest organically managed land area in 2023 [12]—93% of total OF are applied in organic farming [10]. The EU has set a goal to increase organic farming from the current 9% of agricultural land to 25% [13], further encouraging the use of OF while reducing reliance on MF. Similar efforts are underway in China, the world’s largest consumer of MF, where declining nitrogen (N) use efficiency has underscored the need for more sustainable fertilisation practices [14].

Globally, the application of manure N to soils varies by region, with Europe leading at 41.1%, followed by Asia at 34% and the Americas at 20.6% in 2022 [15]. Despite its recognised benefits for soil health and sustainability, the widespread adoption of OF presents several challenges. First, there are logistical challenges related to transportation and the need for specific equipment for application [16,17], often increasing the costs associated with transport and application [11]. Second, there is uncertainty about the amount of nutrients applied, both in terms of quantities and characteristics. This is because, unlike MF, the composition of OF is highly variable, complicating the accurate determination of application rates needed to supply specific amounts of N, phosphorus (P) and potassium (K) [18]. The application of predefined nutrient ratios, such as N:P, is not feasible with OF, leading, in some cases, to the over-application of certain elements [19]. Also, estimating the fraction of plant-available nutrients is challenging, as many nutrients in OF are in organic forms and not readily available to plants [20,21]. Third, there is a risk of soil and water contamination with unknown elements, such as pathogenic microorganisms and residues of veterinary drugs [22,23]. Fourth, the uneven distribution of these fertilisers at both national and European scales often results in limited availability in certain regions, restricting their accessibility where they are needed [11]. Lastly, it has been shown that the application of OF to soil might increase NH₃ and GHG emissions due to the application of ammonium (NH₄⁺) and carbon (C) [24,25]. Nevertheless, such emissions can also occur with MF as a consequence of incorrect and excessive application rates and an energy-intensive production process [24,26].

Given the previously mentioned challenges associated with the use of OF, there is a risk of decreasing yield and crop quality, which can directly impact farm sustainability. Hence, before replacing MF with organic alternatives, it is essential to conduct a comprehensive assessment of these factors and on different site characteristics.

Such assessment was conducted on maize in this study. Maize (Zea mays L.) is among the three most important crops globally, alongside rice and wheat [27]. This includes Portugal, where maize production reached 700 thousand tons in 2022 [28]. High-yielding maize cropping systems historically require significant nutrient inputs, particularly N, traditionally applied in the form of MF [29,30,31]. However, in some regions, there is a growing trend towards using manure. This trend is particularly notable in regions with intensive livestock production, often characterised as a landless farming system [32]. In these areas, there are substantial amounts of manure and/or slurry available, but there is limited land available for its disposal.

The use of animal manure can also be complemented by other strategies to enhance SOM, such as no-till or reduced tillage practices. However, regulations sometimes mandate the incorporation of animal manure into the soil unless mitigation measures are employed [33,34]. It is believed that integrating OF with conservation agriculture practices could yield significant benefits for both farmers and the environment. Therefore, the potential advantages of these combinations need to be thoroughly evaluated.

In this study, we evaluated the effects of substituting MF with poultry manure compost and pig slurry in basal fertilisation of maize production, on maize yield and GHG emissions in three contrasting fields, in terms of agricultural practices (conventional or conservation) and soil characteristics. This experiment was conducted at a farm scale, enabling the assessment of this replacement under real-life conditions. While field-scale studies are common, the inherent variability of these environments highlights the continued need for research into the impacts of OF on crop performance and environmental emissions under different field characteristics. By conducting such evaluations, the effectiveness of OF, such as manures and composts, can be better understood, promoting their use in agriculture. This, in turn, can help reduce reliance on MF while also easing the burden of waste management for livestock farmers.

2. Materials and Methods

2.1. Experimental Sites and Soil Characteristics

A field experiment was conducted during the 2021 campaign on a commercial farm located at Azinhaga (Quinta da Cholda—Portugal) at three different sites, all dedicated to maize production and irrigated using a sprinkler system. Both site 1 and 2 have been managed under conventional agriculture practices, while in site 3, no-till practices have been implemented for the past 15 years (direct seeding). In contrast to site 2 and 3, site 1 has received yearly applications of organic amendments over the last 15 years in addition to the mineral fertilisation. In all sites, maize has historically been continuously cultivated.

At each site, nine soil samples were randomly collected from the 0–20 cm soil layer at the beginning of the experiment, prior to basal fertilisation and seedling, and were analysed using standard methods as described by Fangueiro et al. [35]. This was conducted to characterise soil fertility at each site and assist in designing fertilisation plans. The main characteristics of the soil of each site are presented in

Table 1. Soil physicochemical characteristics at each experimental site. Mean values and standard errors of nine replicates.

	Site 1	Site 2	Site 3
pH (H2O)	7.3 ± 0.1	7.5 ± 0.1	7.0 ± 0.1
Soil organic matter (g kg−1 dry soil)	13.0 ± 0.0	21.0 ± 0.1	15.0 ± 0.1
Olsen-P (mg P2O5 kg−1)	32.7 ± 2.95	29.0 ± 1.7	17.2 ± 3.4
Egner-Riehm-K (mg K2O kg−1)	139.7 ± 7.1	183.3 ± 12.1	144.4 ± 5.6
Electrical conductivity (dS m−1)	0.095 ± 0.0	0.137 ± 0.0	0.092 ± 0.1
NH4+ (mg N kg−1)	1.7 ± 0.3	1.5 ± 0.1	1.3 ± 0.2
NO3− (mg N kg−1)	19.3 ± 1.7	35.7 ± 6.9	27.8 ± 1.6
Texture class	Sandy	Sandy loam	Sandy loam
Coarse sand (g kg−1)	642.6 ± 62.4	240.1 ± 11.3	220.3 ± 34.6
Fine sand (g kg−1)	140.4 ± 44.5	292.8 ± 12.6	336.4 ± 25.7
Silt (g kg−1)	35.7 ± 6.9	19.3 ± 1.7	27.8 ± 1.6
Clay (g kg−1)	101.7 ± 7.6	193.3 ± 7.6	195.6 ± 15.8

.

According to the Köppen climate classification [36], the farm is marked by a Mediterranean temperate climate, characterised by rainy winter and dry and hot summer. Meteorological data were recorded by a weather station located at the farm. Air temperature and precipitation recorded during the experiment are presented in Figure 1.

2.2. Experimental Design

To evaluate the effectiveness of three different fertilisation strategies on maize yield, we employed a single large plot design in each experimental site, rather than multiple smaller plots. As such, in each experimental site, three plots (20 m × 500 m) were established, each receiving different fertilisers at basal fertilisation: pig slurry (PS), composted manure (CM), and standard MF, ammonium sulphate (40:0:0). Since this experiment was conducted on a commercial farm, an unfertilised treatment was not included in the experimental design. The basal fertilisation options were as follows:

Site 1: pig slurry (32 m³ fresh material ha⁻¹), composted manure (10 T fresh material ha⁻¹), and MF (150 kg ha⁻¹).

Site 2: pig slurry (32 m³ fresh material ha⁻¹), composted manure (10 T fresh material ha⁻¹), and MF (150 kg ha⁻¹)

Site 3: pig slurry (32 m³ fresh material ha⁻¹ + 175 kg MF ha^-1), composted manure (12 T fresh material ha⁻¹ + 175 kg MF ha^-1), and MF (400 kg ha⁻¹).

In all sites, total N fertilisation amounted to around 300 kg N ha⁻¹, calculated considering the estimated crop production and soil fertility. Differentiation between treatments was conducted during basal fertilisation, top dressing, and irrigation (

Table 2. Nitrogen fertilisation plan for each treatment and site, detailing each fertilisation event.

	Basal	Seedling	Top Dressing	Irrigation
	Kg N ha−1
Site 1
MF	60.0	22.5	78	140
CM	85.6	22.5	52	140
PS	168.6	22.5	0	109
Site 2
MF	60.0	22.5	78	140
CM	93.4	22.5	44	140
PS	80.5	22.5	50	140
Site 3
MF	160.0	22.5	0	118
CM	163.4	22.5	0	118
PS	166.6	22.5	0	118

MF—mineral fertiliser, CM—composted manure, PS—pig slurry.

). Application doses during basal fertilisation varied between OF due to differences in dry matter and N content, which affected application feasibility and justified adjustments between treatments. These were the minimum quantities required to ensure homogeneous application. Top dressing and irrigation applications were then adjusted so that all treatments received the same total amount of N throughout the crop cycle. At site 3, top dressing was not feasible due to the existing management practices (no-till); therefore, the N that would have been applied during this time (70 kg N ha⁻¹) was applied immediately during basal fertilisation in this site.

Management practices were consistent across the three treatments at each site, including P and K fertilisation and herbicide application. Detailed information on the management practices for each experimental site can be found in the Supplementary Materials (Table S1).

Within each plot, corresponding to each treatment, three sub-plots were used as pseudo-replicates for all the remaining assessments. This approach was chosen due to logistical constraints and field uniformity. The large area of the plots (1 ha) allowed for more practical simulation of real-world fertiliser application on a commercial scale, which enhances the applicability of our findings to large-scale agricultural practices. Field conditions were carefully monitored to ensure uniformity, and rigorous application protocols were followed to maintain consistency across the plot, as well as a robust sampling strategy to ensure representative data collection. Overall, this design provides valuable insights into fertiliser performance under conditions that closely reflect practical agricultural scenarios, offering a meaningful contribution to the field of agronomy.

Pig slurry was sourced from a nearby swine farm and applied using an 8 m³ tanker equipped with a splash plate. The CM was obtained from a commercial brand and applied using a conventional manure spreader. Both the slurry and the compost were incorporated into the soil to a depth of 30 cm by ploughing within 24 h of application in sites 1 and 2. In site 3 (no-till), a 5 mm irrigation was applied to minimise ammonia emissions. To ensure an even distribution, the tractor with the splash plate was equipped with a flowmeter that indicated the amount of slurry being applied. For the application of CM, the tractor maintained a steady speed, and the manure spreader was weighed before and after application to accurately determine the amount applied. The composted manure applied at site 1 was derived from turkey manure, while the compost used at sites 2 and 3 was derived from poultry manure, hence the different application rates of CM in site 3 compared with sites 1 and 2. The OF were analysed using the methodologies described by Fangueiro et al. [35], and the main characteristics are shown in

Table 3. Main characteristics of the OF applied in each site. To be noted that, in site 1, the CM was derived from turkey manure, while in sites 2 and 3, the CM derived from poultry manure, hence the slightly different characteristics. Mean values and standard errors of three replicates.

	pH	EC	DM	Total N	P	K
		mS cm−1	%	g kg−1 (FM)
Site 1
PS	7.68 ± 0.01	8.79 ± 0.02	4.68 ± 0.09	2.68 ± 0.01	0.49 ± 0.03	0.86 ± 0.03
CM	6.61 ± 0.01	10.18 ± 0.01	83.54 ± 0.44	16.86 ± 0.93	40.54 ± 0.32	16.07 ± 0.11
Sites 2 and 3
PS	7.34 ± 0.01	10.26 ± 0.02	4.03 ± 0.03	2.92 ± 0.01	0.51 ± 0.04	0.83 ± 0.02
CM	8.62 ± 0.03	13.70 ± 0.27	32.23 ± 0.76	8.05 ± 0.48	9.27 ± 0.24	4.35 ± 0.15

PS—pig slurry, CM—composted manure, EC—electrical conductivity, DM—dry matter, FM—fresh matter.

.

2.3. Analytical Procedures

2.3.1. Maize Growth and Yield

A chlorophyll meter (SPAD 502, Konica Minolta, Ramsey, NJ, USA) was used to assess the chlorophyll levels in the leaves during the flowering stage, evaluating 20 plants per replicate of each treatment plot. On the same day, 20 leaves per replicate and treatment plot were sampled for N, P, and K monitoring. Maize yield was measured during harvest directly from the harvesting equipment for each treatment plot (each plot harvested at once), therefore not permitting any replicates. During this process, maize grains were also sampled and then dried at 80 °C for 24 h to determine dry matter content and N, P, and K concentration. Analysis of maize plant parts (grains and leaves) was performed following the methods described by Temminghoff and Houba [37].

The nitrogen–partial nutrient balance (N-PNB) was calculated as the ratio between the total N exported by the grain and the total N applied. In the absence of a zero-N treatment, the N-PNB was used as an index for nitrogen use efficiency. N-PNB measures the amount of N output per unit of N input [38], which is practical for field experiments and can be easily utilised by farmers to monitor the efficiency of N fertilisation in each crop campaign.

2.3.2. Gaseous Emission Measurements

GHG emissions from each treatment plot were measured throughout the experiment following the methods described by Fangueiro et al. [39]. In each plot, three stainless-steel chambers were installed on the planting line. The chamber includes (1) a stainless steel base, 29.5 cm long, 20 cm wide, and 14 cm high, which was inserted into the ground to a depth of about 6 cm and remained in the same place until the end of the measurement period, and (2) a cover, 35 cm long, 25 cm wide, and 8.2 cm high, equipped with Teflon tube (Ø = 4 mm; length = 300 mm) to facilitate air sampling. Air sampling was conducted by closing the chamber, followed by immediate sampling (T0) of 30 mL of air from the headspace using a syringe. Additional samples were collected at 30 min (T1) and 60 min (T2) after closure. Gas fluxes were then calculated using linear regression based on the data collected at T0, T1, and T2 (Equation (1)). Plants growing inside the chambers were removed during sampling, and the chambers were temporarily removed and replaced each time machinery passed through the planting lines. Air samples were analysed by gas chromatography with a GC-2014 instrument (Shimadzu, Kyoto, Kyoto, Japan) to determine the concentrations of N₂O, CH₄, and CO₂. The temperature was recorded at each sampling point.

2.4. Calculations and Statistical Analysis

Gas fluxes were calculated using linear regression based on the data collected at T0, T1, and T2. The emission fluxes (G) for N₂O, CH₄, and CO₂ were calculated as described in Fangueiro et al. [40]:

$$G (\mathrm{g} \mathrm{C} \mathrm{or} \mathrm{N} {\mathrm{day}}^{-1} {\mathrm{ha}}^{-1}) =\frac{(C \times 1440 \times 12) \times M}{V \times \frac{273 + T}{273} \times a}$$

(1)

where C (dm³ dm⁻³ min⁻¹) represents gas fluxes obtained by linear regression corrected for time (1440 min) and headspace (12 L), M is the molecular weight of C (12 g mol⁻¹) or N (14 g mol⁻¹), T is the mean temperature of the environment (25 °C), V is the volume of an ideal gas (22.4 m³ mol⁻¹), and a is the area occupied by the chamber (0.06 m²).

The cumulative emissions of CO₂, N₂O, and CH₄ over the entire experiment were estimated by linear interpolation of the respective emission rates between two successive measurements. Yield-scaled emissions were obtained by dividing the cumulative gas emissions by the maize yield in each treatment. The global warming potential (GWP) for each treatment and forage crop was determined using the GWP coefficients for N₂O (273) and CH₄ (27.2) [41].

Statistical analysis was performed using one-way analysis of variance (ANOVA) to identify significant differences between treatments and sampling dates. When ANOVA indicated significant effects, Tukey’s Honest Significant Difference (HSD) test was used as a post hoc test to compare mean differences at a 0.05 probability level. Relationships between variables were evaluated using the Pearson correlation index. Statistical analysis was conducted using Statistix 7.0 (Analytical Software, Tallahassee, FL, USA).

3. Results and Discussion

3.1. Maize Yields and Plant Nutrient Content

The leaf chlorophyll levels did not show significant differences between treatments at sites 2 and 3 (

Table 4. Maize growth and production indicators at the different sites and for each treatment. Mean values of three replicates.

	Maize Growth Indicators				Production Indicators
	Leaf Content (mg kg−1)			Chlorophyll Levels (SPAD Levels)	Grain Yield (kg DM ha−1)	Grain Content (mg kg−1)			N-PNB (kg N Exported kg N Applied−1)
	N	P	K			N	P	K
Site 1
CM	17.70 a	2.67 a	38.48 a	35.27 b	15,667	8.97 b	3.31 a	3.93 a	0.47 b
MF	14.09 b	2.49 b	39.41 a	42.07 ab	17,287	7.62 c	3.25 a	3.80 a	0.44 b
PS	16.22 a	2.50 b	36.69 b	43.72 a	17,261	9.72 a	3.28 a	3.90 a	0.56 a
Site 2
CM	19.77 a	2.83 a	39.99 a	49.40 a	20,224	10.68 a	3.23 a	3.89 a	0.72 a
MF	19.76 a	3.00 a	44.95 a	49.43 a	19,768	9.83 a	3.31 a	4.04 a	0.65 a
PS	19.06 a	2.79 a	40.97 a	48.57 a	20,259	9.04 a	3.26 a	3.97 a	0.61 a
Site 3
CM	21.18 a	2.77 a	41.17 a	56.71 a	15,838	8.66 a	3.17 a	3.89 a	0.46 a
MF	20.92 a	2.73 a	38.37 a	54.77 a	15,934	9.15 a	2.65 a	3.36 a	0.49 a
PS	20.65 a	2.91 a	39.83 a	54.46 a	16,130	8.80 a	3.08 a	3.84 a	0.47 a

For each site and for each column, values followed by a same letter are not statistically different according to the ANOVA analysis. CM: composted manure; MF: mineral fertiliser; PS: pig slurry; DM: dry matter; N-PNB: nitrogen–partial nutrient balance.

), according to the SPAD readings. However, at site 1, the pig slurry (PS) treatment led to the highest levels, followed by the mineral fertiliser (MF) and the composted manure (CM) treatments, reaching 43.72, 42.02, and 35.27, respectively. The relationship between SPAD readings and leaf N content has been widely discussed, showing a positive correlation [42,43]. However, in this experiment, no significant correlation was found between the two variables, though a general trend was observed where higher SPAD readings tended to coincide with increased leaf N content. According to Rocha et al. [44], the results can vary, either decreasing or increasing the correlation between the SPAD readings and leaf N content, depending on the maize hybrid evaluated and with environmental factors [42].

Similarly, no significant differences were observed between treatments at sites 2 and 3 in terms of leaf N, P, and K content. At site 1, the lowest leaf N content was observed in the MF treatment (14.09 mg kg⁻¹), while the highest N content was found in the CM (17.70 mg kg⁻¹) and PS (16.22 mg kg⁻¹) treatments. These results were interesting considering the equalised N application rate across treatments. The significant differences observed at site 1, as opposed to the other two sites, might be attributed to the poorer soil conditions, as indicated by the lower SOM and N content (Table 1). In soils with lower fertility, the application of OF has been shown to have a more pronounced effect on crop performance and soil microbial biomass [45], explaining the higher leaf N content in the treatments with OF (CM and PS). Similarly, Duong et al. [46] found that in soils with less clay content, the application of compost increased wheat shoots N and P content to levels higher than those observed in soils with higher clay content. The authors suggested that this effect might be due to OF promoting the development of finer roots, which have an enhanced ability to access and utilise soil nutrients. The highest leaf P content was observed in the CM treatment (2.67 mg kg⁻¹), compared with PS (2.50 mg kg⁻¹) and MF (2.49 mg kg⁻¹) treatments, whereas K content was highest in the CM (38.48 mg kg⁻¹) and MF (39.41 mg kg⁻¹) treatments and lowest in the PS (36.69 mg kg⁻¹) treatment. Since all treatments received a same amount of K before the N fertiliser application, these differences can be explained by the extra amount of P and K applied through the OF. Indeed, P and K content was higher in the CM compared to PS (Table 3), which lead to a higher application rate of these nutrients in the CM treatment and possibly increased nutrient recovery. This was particularly evident at site 1, where CM, derived from turkey manure, had higher P and K content than the CM applied at sites 2 and 3.

Differences between MF and OF treatments in terms of grain yield were lower than 2.5% and 1.2% in sites 2 and 3, respectively. However, at site 1, while differences between grain yield in MF and PS represent less than 0.2%, a yield decrease of ~9% was observed in CM relative to MF and PS. Additionally, sites with conventional agricultural practices (site 2) showed a trend toward higher yields, while no-till practices (site 3) tended to result in lower yields. The impact of different tillage systems on crop yield has been widely debated in the literature, with outcomes varying depending on soil and climate conditions. For example, Barut et al. [47] found significantly lower silage yields from corn under no-till compared to conventional tillage in a Mediterranean study. In another experiment with grain crops, no-till practices have been associated with higher soil penetration resistance, which can inhibit root growth and negatively impact crop yield [48]. On the other hand, some studies have shown that no-till practices can improve yields under certain conditions. In Croatia, Bogunovic et al. [49] reported higher grain yields with no-till compared to conventional tillage, but only during dry years, likely due to improved soil water retention. A similar outcome was observed in a rainfed maize system in Spain [50]. At site 3, there is a slight tendency for higher yield with PS, though the effect is minimal.

PS (9.72 mg kg⁻¹) resulted in the greatest N recovery in maize grains, followed by CM (8.79 mg kg⁻¹) and MF (7.62 mg kg⁻¹), at site 1. At the other two sites, no significant differences were observed between treatments. Furthermore, in all three sites, P and K recovery by maize grains were statistically similar between treatments. At site 1, the nitrogen–partial nutrient balance (P-PNB) indicated that PS (0.56 kg N exported kg⁻¹ N applied) was more efficient at supplying N to plants compared with CM (0.47 kg kg⁻¹) and MF (0.44 kg kg⁻¹). This could be attributed to the slower mineralisation rate in the CM treatment compared with PS [51,52], which may have affected crop N absorption and its translocation to the grains. In contrast, fresh pig slurry has demonstrated higher microbial activity compared with composted pig slurry, as evidenced by increased soil respiration and enzyme activity, likely due to greater nutrient availability in a Spanish horticultural experiment [52]. Additionally, the impact of different OF like CM and PM can vary across different soil types. For example, Cassity-Duffey et al. [53] observed a slower mineralisation rate of feather meal in clay soil, whereas a pellet mix in the same soil did not yield the same result, likely due to differences in particle size between the two OF. This suggests an interactive effect between soil particle size and the type of OF, similar to how pig slurry proved more effective in sandy soil (site 1) but had no impact on more clay soils (sites 2 and 3).

Furthermore, the basal fertilisation at site 2 was applied later (Supplemental Table S1), after a heavy rainfall that happened after the application of OF in site 1. This, combined with the finer texture and higher SOM levels, might explain the absence of significant differences in nutrient recovery by grains and N-PNB values among the fertilisers at sites 2 and 3. Overall, the results indicated that the partial replacement of MF with PS and CM did not negatively affect the N-PNB or nutrient recovery, since the levels of nutrients recovered from OF were never significantly lower than those from MF.

3.2. Greenhouse Gas Emissions

The highest CH₄ emissions at site 1 occurred within the first day of measurement across all treatments, including in the MF treatment (Figure 2a), reaching 166.82 g C ha⁻¹ d⁻¹ in the latter treatment. However, this was followed by negative fluxes in all treatments, reaching −109.87 g C ha⁻¹ d⁻¹ in the MF treatment. This could be due to unusually warm temperatures at this time, reaching 30 °C, and dryer conditions, which led to soil to act as a sink for CH₄ [54,55]. Additionally, the coarser texture of the soil in this site could have led to better aeration and increased CH₄ oxidation [56].

Subsequently, during the irrigation period (15 May 2021–24 August 2021) at this site, notable CH₄ emissions were observed, mostly in the CM treatment. Irrigation probably increased water-filled pore space (WFPS) which, along with the release of organic compounds from the OF, could have promoted methanogenesis and increased CH₄ emissions during this period. This has been previously observed in an irrigated Mediterranean field, with the application of both manure and slurry [57]. The higher emissions in the CM treatment were, however, surprising, considering the higher dry matter content (Table 3) and the composting process the manure underwent, which results in a much more stable organic fertiliser. In fact, previous researchers have found that composted manure resulted in lower CH₄ emissions when compared with raw slurry [58], which contrasts with the present findings, although this was observed in an experiment with flooded rice. The higher CH₄ emissions in the CM treatment in this experiment could be explained by the higher C content of CM compared with PS, leading to higher C in the soil, and higher CH₄ emissions [24] given the optimal conditions: higher soil moisture. By the end of the experiment, CM (1453 g C ha⁻¹) resulted in higher cumulative CH₄ emissions compared with PS (-276 g C ha⁻¹) but did not significantly differ from the MF (455 g C ha⁻¹) treatment (

Table 5. Cumulative and yield-scaled emissions of methane (CH₄), nitrous oxide (N₂O), carbon dioxide (CO₂), and the global warming potential (GWP) at the three sites and for each treatment considered. Mean values and standard errors of three replicates.

Treatments	Site 1		Site 2		Site 3
	CH4
	g C ha−1	g C t−1 DM	g C ha−1	g C t−1 DM	g C ha−1	g C t−1 DM
CM	1453 a	93 a	3657 a	181 a	1224 a	77 a
MF	455 ab	26 ab	1866 b	94 b	−393 a	−25 a
PS	−276 b	−16 b	2287 ab	113 b	830 a	51 a
	N2O
	g N ha−1	g N t−1 DM	g N ha−1	g N t−1 DM	g N ha−1	g N t−1 DM
CM	3212 a	205 a	8175 a	404 a	5570 a	352 a
MF	2446 a	142 a	7767 a	393 a	4098 b	257 b
PS	2872 a	166 a	7470 a	369 a	4736 ab	294 ab
	N2O (% of total N applied)
CM	1.071 a	-	2.725 a	-	1.857 a	-
MF	0.815 a	-	2.589 a	-	1.366 b	-
PS	0.957 a	-	2.490 a	-	1.579 ab	-
	CO2
	kg C ha−1	kg C t−1 DM	kg C ha−1	kg C t−1 DM	kg C ha−1	kg C t−1 DM
CM	5972 a	381 a	4204 a	208 a	3389 a	214 a
MF	4883 a	282 a	4099 a	207 a	3282 a	206 a
PS	5010 a	290 a	4368 a	182 a	2657 a	165 a
	GWP
	kg CO2 eq ha−1	kg CO2 eq t−1 DM	kg CO2 eq ha−1	kg CO2 eq t−1 DM	kg CO2 eq ha−1	kg CO2 eq t−1 DM
CM	1430 a	91 a	3640 a	180 a	2434 a	154 a
MF	1066 a	62 a	3400 a	172 a	1744 b	109 b
PS	1222 a	71 a	3287 a	162 a	2062 ab	128 ab

For each site and for each column, values followed by a same letter are not statistically different according to the ANOVA analysis. CM: composted manure; MF: mineral fertiliser; PS: pig slurry; DM: dry matter.

). PS was the only treatment that acted as a CH₄ sink.

In site 2, significant CH₄ emissions were observed in all treatments, mainly during the initial 6 days of measurement (Figure 2d), with emissions peaking at 259.50 g C ha⁻¹ d⁻¹ in the PS and MF treatments and at 245.35 g C ha⁻¹ d⁻¹ in the CM treatment. The CH₄ peak in the PS treatment could be due to CH₄ produced during PS storage that was subsequently released after soil application [35]. In the CM treatment, this could be explained by the higher C that was provided, giving substrate for methanogens [24], while in MF treatment, the quickly available N applied might have prevented CH₄ oxidation [55,57].

In contrast, site 3 showed no significant CH₄ emissions during the first four sampling dates (28 days) (Figure 3a). However, small emissions did occur after June, during the irrigation period, in all treatments, but higher in the CM treatment, reaching 33.00 g C ha⁻¹ d⁻¹ in late June. During this time, CH₄ fluxes were somewhat erratic, with small peaks, probably due to irrigation and higher C availability, mixed with negative fluxes, also small. As opposed to what was previously observed at site 1, these negative fluxes were observed during the irrigation period. Although CH₄ oxidation is favoured during warm and dry conditions, the negative fluxes during this period could be explained by the lack of top-dressing fertilisation, resulting in low N availability and consequently higher CH₄ oxidation rates [55], due to low competition for oxidising bacteria [59]. Regardless, emissions, either positive or negative, were rather small at site 3 when compared with the other two sites, indicating that no-till practices under this study’s conditions led to lower CH₄ emissions. This was slightly unexpected given the typically higher soil WFPS associated with no-till practices. There are conflicting results in the literature regarding this outcome. For instance, Struck et al. [60] found that conventional agricultural practices in maize increased CH₄ uptake compared with no-till, which was observed in this study, whereas a review by Six et al. [61] suggested the opposite effect. A more recent review by Maucieri et al. [62] showed that the impact of no-till practices on CH₄ emissions was negligible in dry climates, , and observed a slight, though not significant, tendency for higher emissions in maize under no-till, which contrasts with this study’s findings. The authors emphasised the need for further research focused specifically on the impact of no-till practices on CH₄ emissions, as most studies have predominantly concentrated on CO₂ and N₂O.

Regarding the cumulative CH₄ emissions, different trends were observed in the three different sites, with notably higher values of CH₄ emissions at site 2 (Table 5). The differences between the sites could be explained by variations in soil textures but also agricultural practices and history of OF application. The coarser texture at site 1 could have led to better aeration and increased CH₄ oxidation [56], while the no-till practices implemented at site 3 under the present study’s conditions led to lower emissions. Kravchenko et al. [56] also explained that the different microbial communities in each soil could also play a significant role in impacting CH₄ emissions and oxidation, which is something that should be considered in future studies.

N₂O emissions at site 1 remained consistently below 70 g N ha⁻¹ day⁻¹, averaging approximately 15 g N ha⁻¹ day⁻¹ across all treatments (Figure 2b). While fluxes were highly variable, N₂O emissions tended to increase in all treatments after irrigation began, likely due to higher WFPS [55]. The highest peaks were observed in the PS and CM treatments, reaching 58.73 g N ha⁻¹ d⁻¹ on 19 June and 47.11 g N ha⁻¹ d⁻¹ on 20 May, respectively. Although smaller peaks were also noted in the MF treatment. N₂O emissions at this site appeared lower compared with the other two sites, probably due to differences in soil texture or the timing of fertiliser application. For instance, Chantigny et al. [63] found higher emissions in a clay soil compared with a soil with courser texture in a field experiment with corn, again, likely due to higher WFPS in the clay soil, which can create anaerobic conditions and favour denitrification. In fact, it has been shown that sandy soils have a lower denitrification potential than clay soils [64], further explaining the lower emissions observed in site 1. Nitrification reactions are typically more dominant with MF, while denitrification tends to prevail under organic fertilisation [55,65]. However, in this site, no significant differences in cumulative N₂O emissions were observed between the different types of fertilisers (Table 5). This suggests that, regardless of the dominant N₂O-producing process, the partial replacement of MF with OF did not increase emissions at this sandy soil.

At site 2, the average N₂O emission rate was approximately 42 g N ha⁻¹ day⁻¹ across all treatments throughout the experiment (Figure 2e). The highest peaks were observed in the MF and CM treatments 9 and 17 days after fertiliser application (8 April and 16 April), reaching 192.30 g N ha⁻¹ d⁻¹ and 187.04 g N ha⁻¹ d⁻¹, respectively. This coincided with a rainfall event, which, as previously observed, could have promoted N₂O emissions. However, considering the cumulative N₂O emissions, there were no significant differences between fertiliser treatments (Table 5). Following these peaks, N₂O emissions stabilised at around 50 g N ha⁻¹ day⁻¹ until harvest.

In site 3, the average N₂O emission rate was approximately 30 g N ha⁻¹ day⁻¹ in the MF and PS treatments throughout the whole experiment, while it averaged around 48 g N ha⁻¹ day⁻¹ in the CM treatment (Figure 3b). The highest emissions occurred at the beginning of the experiment, following fertiliser application, especially in the CM treatment (132.83 g N ha⁻¹ d⁻¹). The higher emissions in the CM treatment were somewhat surprising, given the higher stability of the material as a result of the composting process [66], which leads to a slower mineralisation rate and gradual nutrient release. However, the higher C content in this material could have promoted microbial activity and reduced oxygen (O₂) availability, creating optimal conditions for denitrification and increased N₂O emissions [63]. In contrast, the lower N₂O emissions in the PS treatment could be due to improved plant N absorption (thus lower N availability for nitrification and/or denitrification reactions), as PS has a higher proportion of NH₄⁺ that is easily available and consumed by the plants, unlike solid manures or composted manures [57]. Emissions at this site decreased after the initial peaks but increased again around the beginning of June, likely due to precipitation observed during this time. Higher cumulative N₂O emissions were observed in the CM (5570 g N ha⁻¹) treatment compared with MF (4098 g N ha⁻¹); however, cumulative emissions in the PS (4736 g N ha⁻¹) treatment did not significantly differ from those in the MF treatment (Table 5).

It is also noteworthy that the emission factor (EF) for N₂O emissions was numerically highest at site 2, followed by sites 3 and 1. Although sites 2 and 3 are relatively similar in soil texture and neither has a history of OM application, the EF was lower at site 3 compared to site 2. This aligns with findings by Six et al. [61], who reported that after 20 years of no-till practices, N₂O emissions are lower compared to conventional ploughing practices, regardless of climate type. Similarly, Struck et al. [60] observed a decrease in N₂O emissions with the adoption of no-till in a field experiment in Germany with maize, despite an increase in WFPS. The lower EF values at site 1 might be attributed to its coarser soil texture, as previously noted. Additionally, higher N leaching at site 1, particularly given the timing of fertiliser application, could also contribute to the reduced EF observed there.

CO₂ emissions at site 1 were very similar between fertiliser treatments (Figure 2c), with emissions peaking on several dates but increasing from the beginning of the experiment up until mid-June, after which emissions decreased. The observed peaks in June were higher in the CM treatment, reaching 65.77 g C ha⁻¹ d⁻¹, although not significantly higher compared with PS (42.15 g C ha⁻¹ d⁻¹) and MF (45.43 g C ha⁻¹ d⁻¹). At site 2, CO₂ emissions increased from less than 20 kg C ha⁻¹ day⁻¹ initially to over 30 kg C ha⁻¹ day⁻¹ in mid-April, mirroring trends seen at other sites, probably attributed to higher soil moisture content, considering irrigation and rain events, which promotes CO₂ emissions, especially in Mediterranean climates [25]. Values fluctuated then between 20 and 30 kg C ha⁻¹ day⁻¹, except for a peak exceeding 60 kg C ha⁻¹ day⁻¹ observed in all treatments during the first 20 days of June, likely related to the start of irrigation. At site 3, an initial peak in CO₂ emissions was observed immediately after application across all treatments (Figure 3c), with the highest values in the CM treatment (reaching 120.00 g C ha⁻¹ d⁻¹), likely due to the higher C availability that promoted soil respiration [67]. This trend mirrors the pattern observed in our N₂O daily fluxes, further supporting the earlier observation that high soil respiration due to microbial activity results in low O₂ availability which leads to higher N₂O emissions. There was a subsequent second peak at the beginning of June in all treatments, with higher emissions in the CM and MF treatments, reaching 134.90 g C ha⁻¹ d⁻¹ and 111.92 g C ha⁻¹ d⁻¹, respectively. From mid-July to harvest in mid-October, CO₂ emissions stabilised, showing similar values across all three treatments. Nevertheless, in neither site were observed differences between treatments in terms of cumulative CO₂ emissions (Table 5).

In terms of GWP (Table 5), the results followed similar trends to the cumulative N₂O and CO₂ emissions, showing only significant differences in site 3. In this site, which was under no-till practices, partial replacement of MF with CM (2434 kg CO₂ eq ha⁻¹) led to a higher GWP compared with only MF (1744 kg CO₂ eq ha⁻¹) but did not differ from PS (2062 kg CO₂ eq ha⁻¹). PS also did not differ from MF. This was also observed when emissions were adjusted to grain yield. This highlights the potential of PS as a crop fertiliser, capable of sustaining crop production under no-till practices while maintaining similar emissions, when compared with an exclusively mineral fertilisation. Additionally, it reduces the need for MF production, along with the associated raw material requirements and emissions from production.

4. Conclusions

Replacing 30% of mineral N fertilisers with OF, such as pig slurry and composted animal manure, showed promising benefits, particularly in soils with lower clay and OM content. At site 1, this substitution significantly enhanced nutrient absorption, while at site 3, under no-till practices, yields were lower, but the use of pig slurry helped counteract this effect. Therefore, the use of OF is highly recommended, especially in poorer soils and under conservation management practices, suggesting that OF can support productivity in more challenging conditions, improving soil fertility and plant performance. However, soil texture, SOM content, and tillage practices should always be considered. While composted manure increased CH₄ and N₂O emissions at certain sites, pig slurry did not contribute to higher emissions, reinforcing its potential as a sustainable alternative.

There is still the need for further research on optimising OF application under different soil and climatic conditions. Long-term studies should assess whether adjustments in application timing, considering rainfall events, or different combinations of organic and mineral fertilisers can maximise yields while minimising emissions. Additionally, future research should explore how these practices influence overall farm profitability and resilience to climate variability.