Assessing the Performance of Bio-Based Nitrogen Fertilisers Under Salinity and Drought Stress in Spinach: A Preliminary Trial

Abstract

Recently, the EU approved RENURE-criteria materials to be used as substitutes for synthetic N fertilisers. Several studies have been performed on the agronomic efficacy and potential environmental impacts of different bio-based fertilisers (BBFs) from biomass recovery, including the RENURE-criteria materials. But information is lacking about their effectiveness under abiotic stress conditions like salinity and drought. The predictions for climate change-induced increased drought and soil salinisation for the European soils have also increased, making it inevitable to understand BBF performance in these impending situations. Two RENURE-criteria top-priority materials (ammonium nitrate (AN) and ammonium sulphate (AS) and another commercially used BBF—an evaporator concentrate (CaE)) were evaluated in a pot trial growing spinach under salinity and drought stress with a reference ‘no stress’ condition to examine crop growth, nutrient uptake, and nitrogen fertiliser replacement value (NFRV). Agronomically, BBFs performed at par with the synthetic fertiliser (SF) under unstressed and salt-stressed conditions, whereas, under drought stress, BBFs outperformed the SF treatment. AS exhibited the highest yield and nutrient uptake, displaying an NFRV of 3.1 and 1.8 under no-stress and salt-stress conditions, respectively. Salt stress did not negatively impact the crops grown in this trial, potentially due to the higher potassium content in the system, which alleviated the possible negative impacts of high sodium. This study delves into the agronomic response, without evaluating crop physiological changes, and, hence, should be taken as a preliminary step into further investigation of observed elemental interactions (that could be potentially driving stress mitigation) while also examining the crop physiology during the duration of stress.

1. Introduction

Nitrogen (N) is one of the key nutrients with the ability to limit plant growth, when insufficiently applied. Currently, agricultural production rests largely on the shoulders of synthetically manufactured N fertilisers, obtained from the carbon dioxide (CO₂)-intensive Haber–Bosch process. Consuming 1% of the global energy supply [1], the Haber–Bosch process produces ~180 million tonnes of ammonia (NH₃) annually [2] and generates 1.87 tonnes of CO₂ per tonne of NH₃ produced [1]. Next to these environmental concerns, Europe also faced brutal ramifications of the ongoing war between Russia and Ukraine. The fertiliser price hike in 2022 resulted in affordability as well as availability crunches in Europe, causing farmers to cut back on field applications of fertilisers. The sharp increase in natural gas prices led to decreased production of NH₃ needed to produce synthetic N fertilisers [3]. In the wake of these effects, it is inevitable that feasible substitutes to the current N fertilisers are identified to avoid impedance to the European agricultural sector. Animal manure, sewage sludge, and food and other organic wastes are categorised as the most promising biomass streams to extract agricultural nutrients [4]. The European Union’s (EU) Circular Economy Action Plan urges avoiding nutrient leakage from the agricultural sector and closing the nutrient loop by recycling, recovering, and processing nutrients for use as fertilisers [5]. Supported by improvements in regulations like the EU Fertilising Products Regulation (FPR 2019/1009) and the now-approved introduction of recovered nitrogen from manure (RENURE) material categorisation [6,7], dependency on the Haber–Bosch-derived synthetic N fertilisers could be reduced. These recovered nutrients, referred to as bio-based fertilisers (BBFs) in this study, can be defined as ‘materials or products derived from biomaterials (plant, animal or microbial origin, often wastes, residues or side-streams from agriculture, industry or society) with a content of bioavailable plant nutrients suitable to serve as a fertiliser for crops’ [8]. These fertilisers are derived from using different nutrient recovery technologies (physical/chemical/thermal/biological) in order to up-concentrate nutrients from the initially treated biomaterials, thus improving their nutrient efficiency. Ashes, struvites, ammonium (NH₄⁺) salts (ammonium nitrate/sulphate), evaporator concentrate, etc., are examples of different types of BBFs.

Studies in the past have tested the agronomic efficacy and environmental impacts of various N-based BBFs (including the RENURE-materials), both under controlled and uncontrolled conditions [9,10,11,12,13,14]. Although previous research expounds upon fertiliser performance under regular and unstressed experimental conditions, their behaviour under situations of abiotic stress is currently unexplored. Salinity and drought stress are two known abiotic stresses that can significantly limit crop growth and productivity [15]. Soil salinity, a key constraint to crop production, affects millions of hectares of land globally, and costs billions of euros annually [16,17,18]. Increased temperature and evapotranspiration as a result of global climate change can increase soil salinity, thus negatively impacting crop production [19]. While the present extent of salinisation on the surface or underground European soil profile is uncertain, the common consensus is on the increasing trend of soil salinisation in the last few decades; this can be attributed to the changing climate, land use and management practices [20]. Similarly, drought is another important abiotic stress limiting agricultural production, especially in the arid and semi-arid regions of the world [21], and considered to be the primary destructive crop yield-limiting factor [22]. Adverse impacts on crop production might increase owing to climate change, which is predicted to increase the frequency and severity of crop water stress [23,24]. A study involving future drought predictions for Europe under different CO₂ emission scenarios among other factors predicts more frequent and severe drought conditions for Western Europe even under moderate emission conditions [25].

Under such circumstances, it is vital to test the alternative nutrients that could replace the commonly used synthetic N fertilisers to understand their behaviour, performance, and impacts (if any) under potential conditions of stress. With this goal at hand, the current study tested three N-based BBFs—two of which are the RENURE-criteria high-priority materials- ammonium nitrate (AN) from stripping/scrubbing of liquid fraction (LF) of digestate and ammonium sulphate (AS) from air washing of pig stable. The third BBF tested is a concentrate after evaporation (CaE) of LF of digestate, a product already locally marketed as an N and potassium (K) fertiliser, with traces of phosphorus (P). In this study, these three BBFs were tested to determine their fertiliser performance against the commercially used N synthetic fertiliser (SF) calcium ammonium nitrate (30% N) under two stressors: salinity and drought. A plant growth experiment with spinach (Spinacia oleracea L.) was performed to answer the following research questions:

(i)

Do N-based BBFs, specifically the RENURE-criteria NH4+ salts, exhibit similar performance to SF in terms of crop yield and N uptake under salinity and drought stress?

(ii)

Do N-BBFs change soil parameters like pH and EC under abiotic stressors compared to the SF fertilisation?

(iii)

Can N-based BBFs replace their synthetic counterparts as a better choice of N fertilisation in areas prone to abiotic stresses?

2. Materials and Methods

2.1. BBF Collection and Characterisation

The AN tested in this study was obtained from the stripping/scrubbing unit of a pig farm with an anaerobic digestion (AD) plant located in Gistel, Belgium, with a capacity of 11,000 fattening pigs, and a manure capacity of 60,000 tonnes y⁻¹. The AS was obtained from air washing stable air of a pig farm in Belgium, by capturing the NH₃-rich air using sulphuric acid. The CaE is the residual concentrate after evaporation process in an AD plant in Belgium that processes different types of organic wastes. More details on these BBFs and their origin can be found in previous studies [13,14]. All BBFs were collected in air-tight polyethylene sampling bottles of 1 L each and stored at 4 °C.

All the BBFs were physio-chemically characterised to determine their dry matter (DM), pH, electrical conductivity (EC), total elemental composition, and mineral N (NH₄⁺-N and (nitrate) NO₃⁻-N) as per the protocols mentioned in the earlier study [13] (

Table 1. Physio-chemical characterisation of tested BBFs on fresh weight basis.

	BBFs Tested
Parameters	Ammonium Nitrate	Ammonium Sulphate	Concentrate After Evaporation
pH	6.5	6.2	9.4 *
EC (mS cm−1)	312	211	102
DM (%)	27	21	18
Total N (g kg−1)	100	42	6.9
NH4+-N (g kg−1)	44	42	2
NO3−-N (g kg−1)	56	<0.002	<0.002
Total P (g kg−1)	0.01	0.07	1.4
Total K (g kg−1)	0.03	0.23	14
Total Na (g kg−1)	0.09	2.1	20

*: pH-KCl for concentrate after evaporation. EC: electrical conductivity; and DM: dry matter.

).

2.2. Soil Sampling and Characterisation

The soil used for this study was collected from the top layer (0–30 cm) of an arable field in Wingene, West Flanders, Belgium, in March 2021. Predominantly sandy in nature, soil in this part of Flanders is characterised as the Z.c.h. soil type (soil with sandy texture and a moderately poor drainage class with signs of rust deeper than 60 cm and a post-podzol B-horizon) [26]. Spinach was grown in 2020, harvested in early June and followed with maize, which was harvested in late September. The soil was sampled from the field before fertilisation for an impending field trial and air-dried in a greenhouse, where it was sieved using a 2 mm mesh and mixed thoroughly for homogeneity. The soil was stored in the greenhouse until the experiment commenced. Sub-samples of the soil were taken for measurement of pH-potassium chloride (KCl), EC, total elemental composition and mineral N content [13] (

Table 2. Physio-chemical characterisation of soil before start of experiment.

pH-KCl	EC (µS cm−1)	OM (%)	OC (%)	Total N (g kg−1)	NH4+-N (mg kg−1)	NO3−-N (mg kg−1)	Total P (g kg−1)	Total K (g kg−1)	Total Na (g kg−1)
5.6	122	1.9	0.93	1.3	14	41	1.2	1.0	0.05

EC: electrical conductivity; OM: organic matter; and OC: organic carbon.

).

2.3. Experimental Set-Up

The air-dried soil was moved from the greenhouse to a basement, four days before the start of experiment, and water was added to the soil to reach 50% water holding capacity (WHC). After a thorough mixing of the soil, the containers were covered with pin-holed parafilm and left in the dark to activate microbial activity in the dry soil.

The experiment tested five treatments, each with four replicates under two stressors (salt and drought) and one without any stress (5 × 4 × 3), i.e., a total of 60 pots. Seeds of spinach were used, which were previously tested for their germination potential. The treatments tested in the experiment were as follows: (i) unfertilised control; (ii) synthetic N (synthetic ammonium nitrate; 30% N); (iii) AN; (iv) AS; (v) CaE. To supplement the growth of crops and ensure an equal application of P and K in all treatments, synthetic fertilisation was performed using triple super phosphate (46% P₂O₅) and potassium sulphate (K₂SO₄) (PAT; 30% K₂O, 10% MgO, and 42.5% SO₃). An important point to be considered here is the additional application (23% more than the recommended dosage) of K₂O from the treatment with CaE—a consequence of satisfying the N requirements of the crop. Hence, to ensure that every treatment received equal K application, additional K₂SO₄ was added to all other fertilised treatments, accordingly, using PAT. This led to an over-fertilisation with S in the case of treatment with AS (~115% more than SF and AN, and ~1265% more than CaE). The nutrient application dosage was determined based on the crop requirement for spinach (200 kg N ha⁻¹, 55 kg P₂O₅ ha⁻¹ and 250 kg K₂O ha⁻¹; personal communication with Inagro vzw).

For application in the current pot trial, only a very small quantity of the AN was needed (

Table 3. Application rates of products (g pot⁻¹) and macronutrients (mg pot⁻¹) in the pots as per recommended nutrient dosage.

Treatments	Product Application (g pot−1)						Nutrients Applied (mg pot−1)
	CAN	TSP	PAT	AN	AS	CaE	Ntotal	P2O5	K2O	SO3
Control	0	0	0	0	0	0	0	0	0
SF	0.24	0.10	0.64	-	-	-	73	20	112	158
AN	0.12	0.05	0.32	-	-	-	73	20	112	158
AS	-	0.10	0.64	0.88	-	-	73	20	112	339
CaE	-	0.03	0.16	-	0.12	3.9	73	20	112	25

CAN: Calcium ammonium nitrate; TSP: Triple super phosphate; PAT: Potassium sulphate; SF: Synthetic fertiliser (CAN); AN: Ammonium nitrate; AS: Ammonium sulphate; CaE: Concentrate after evaporation.

) due to its high N content, which necessitated mixing AN with demineralised water (to reach the required concentration), thus enabling a substantial quantity of the product (4.17 mL of product) to be pipetted out and mixed with the soil homogenously. All the synthetic fertilisers were also handled similarly; the granules were first ground into a fine powder, which was mixed with demineralised water and added to the soil, enabling a thorough homogenous mixing of the fertilisers with the soil. At the onset of the experiment, each pot (height = 18 cm; top diameter = 12.6 cm) was filled, firstly, with one kg of the pre-incubated soil. BBFs were added to the remaining 600 g of the soil in a container and thoroughly mixed for homogeneity before adding it to the pots. After water addition, 10 seeds were placed in each pot evenly using a pair of forceps, slightly pushing them under the soil surface. Each pot was covered with a pin-holed plastic bag to avoid loss of moisture and placed in the dark for seed germination. Since none of the pots exhibited any seed germination in the first 10 days, the moisture content was reduced, and five seeds were sown in each pot. Within 4–6 days of resowing, germination was observed in all pots. The pots were watered thrice a week, or as per loss of moisture, and the position of pots was randomised during each watering session throughout the experimental duration.

The two stressors were applied after six weeks of resowing the seeds, to ensure that plants in all pots reached similar growth phase and to avoid loss of plants at their initial vulnerable stage. Salt stress application was performed in stages to avoid sudden shock to the plants. Sodium chloride (NaCl) solution was prepared and applied during watering sessions in increasing concentrations of 25 mM to a final concentration of 200 mM, within a period of 25 days. The concentration of NaCl application was decided as per practice from the literature, where successful salinity impact was created by applying this level of salt [27,28,29]. Moisture in these pots was adjusted considering the addition of NaCl solution. Drought stress was imposed by reducing water application to 30% of actual crop requirement.

2.3.1. Plant and Soil Analyses Post-Harvest

The plants were harvested 75 days after the resowing of seeds by clipping the spinach at the stem above the soil using trimming scissors. Any residual soil on the stem was wiped off using tissues and cleaned well. The number of plants in each pot, along with the fresh weight (FW), was recorded. The DM of plants was estimated by oven-drying the samples at 50 °C for five days, after which the dried plant material was finely ground using a mortar and pestle. These ground samples were then analysed to determine the total elemental composition of the plant matter. Total N, P, K, S, Mg, sodium (Na), and calcium (Ca), along with heavy and trace elements were determined [13].

Soil samples from each pot were transferred to plastic bags and the roots were separated. The soil was thoroughly mixed before a sub-sample was transferred to smaller plastic bags and frozen for mineral N analysis. At this point, the moisture content in the soil samples was determined by drying sub-samples of soil at 105 °C for 24 h. The remaining soil was air-dried and sieved using 2 mm mesh to remove smaller roots. Sub-samples from the sieved soil were transferred to 100 mL plastic containers to perform physio-chemical characterisation of soil (as mentioned in Section 2.2).

2.3.2. Calculations and Statistical Analysis

After obtaining results from plant analyses, yield, N, K, S and Na uptake in plant matter, and parameters of fertiliser efficiency like apparent nitrogen recovery (ANR) and nitrogen fertiliser replacement value (NFRV) were determined. The calculations for ANR and NFRV [30,31,32] were performed as follows:

$$\mathrm{ANR} = {\displaystyle \frac{(N uptake TREATMENT \left({kg ha}^{-1})\right)-(N uptake CONTROL \left({kg ha}^{-1}\right))}{Total N applied TREATMENT \left({kg ha}^{-1}\right)}}$$

(1)

$$\mathrm{NFRV}={\displaystyle \frac{ANR BBF}{ANR synthetic N fertiliser }}$$

(2)

Results were statistically processed using the statistical software Sigmaplot 14. Both yield types (fresh and dry yield) of all three stressor categories (except dry yield of non-stressed category) were evaluated using non-parametric ANOVA (Kruskal–Wallis test) since the conditions for parametric ANOVA were not met. Either the normality test (Shapiro–Wilk) failed (p < 0.05), or the test of Variance (Brown–Forsythe) failed (p < 0.05) for these comparisons. Multiple comparisons using Dunn’s post hoc test followed the Kruskal–Wallis test. In the case of dry yield under the unstressed category, parametric ANOVA was performed after the conditions of normality and variance were met, and this was followed by multiple comparisons by Tukey’s post hoc analysis. For nutrient uptake, similarly, parametric and non-parametric ANOVA was performed depending on the fulfilment of assumptions, particularly the test of normality and variance. A two-way ANOVA was also performed to confirm the effects of NaCl application for salinity stress creation by using fertiliser treatments and stressors as the two fixed factors, and Na uptake was evaluated as the dependent variable. For ANR and NFRV, parametric ANOVA followed by Tukey’s test was performed.

Soil parameters after harvest were analysed using two-way ANOVA (p = 0.05), followed by Bonferroni’s multiple comparison post hoc test to understand how the effect of stressor, fertiliser and their interaction (treatment × Stressor) could have potentially impacted the mineral N, pH, and EC of the soil. Correlations between yield and nutrient uptake, and between elemental interactions were estimated using Pearson’s Correlation (p = 0.05).

3. Results and Discussion

3.1. Yield and Elemental Uptake

Fertiliser effect on crop growth was compared individually under each stressor category. Firstly, fresh and dry yield of crops per pot was determined and differences between treatments were statistically determined. The unfertilised plants (control treatment) demonstrated a fresh yield of 13 ± 7.8, 11 ± 10, and 9.4 ± 0.45 g pot⁻¹ and dry yield of 0.89 ± 0.58, 0.99 ± 0.98, and 0.79 ± 0.16 g pot⁻¹ under unstressed, salt-stressed and drought-stressed conditions, respectively (

Table 4. Mean fresh and dry yield (g pot⁻¹) of spinach presented with their standard deviation (n = 4).

		Treatments
Stressor	Yield Type (g pot−1)	Control	Synthetic Fertiliser	Ammonium Nitrate	Ammonium Sulphate	Concentrate After Evaporation
No stress	Fresh yield	13 ± 7.8 a	22 ± 1.0 ab	25 ± 5.8 ab	34 ± 2.0 b	24 ± 1.1 ab
	Dry yield	0.89 ± 0.58 A	2.1 ± 0.19 B	2.1 ± 0.56 B	2.8 ± 0.19 B	2.1 ± 0.33 B
Salt stress	Fresh yield	11 ± 10 a	26 ± 3.2 ab	28 ± 3.6 ab	33 ± 7.3 b	32 ± 1.0 b
	Dry yield	1.0 ± 0.98 A	2.4 ± 0.37 AB	2.3 ± 0.37 AB	2.8 ± 0.58 B	2.8 ± 0.15 B
Drought stress	Fresh yield	9.4 ± 0.45 a	No biomass at harvest	14 ± 3.8 ab	10 ± 1.4 ab	16 ± 3.6 b
	Dry yield	0.79 ± 0.16 A	No biomass at harvest	1.6 ± 1.0 AB	1.2 ± 0.17 AB	1.8 ± 0.39 B

For both yield types, the treatments under each category of stress are compared amongst each other using the non-parametric ANOVA (Kruskal–Wallis test) due to failure of the normality test (p < 0.050) or failure of the test of Variance (p < 0.050) (except for dry yield under no stress—tested by parametric ANOVA, followed by pairwise comparison by Tukey’s post hoc test). When significant differences between treatments are found after the Kruskal-Wallis test (p < 0.05), post hoc testing with Dunn’s pairwise comparison is performed. Significant differences between treatments of each category of stress are indicated by lower-case letters for fresh yield and upper-case letters for dry yield. Comparisons between treatments are performed within the same stressor category only (e.g., the treatment ‘Control’ is compared with the treatment ‘SF’ under ‘No stress’ only and not with SF under ‘Salt stress’ OR ‘Drought stress’).

).

The control treatment exhibited high variabilities within replicates under both ‘no stress’ and ‘salt stress’ conditions. AS-treated plants had significantly higher fresh yield compared to control under ‘no stress’, whereas, under ‘salt stress’ conditions, AS and CaE-treated plants had significantly higher fresh yield. No differences were observed between the different BBFs or between the SF-treated plants relative to BBF-treated plants’ yield. Although the AS-treated crops showcased higher mean fresh yield (34 ± 2 g pot⁻¹) than the other fertilisations (SF (22 ± 1 g pot⁻¹) and both BBFs (25 ± 5.8 and 24 ± 1.1 g pot⁻¹ for AN and CaE, respectively)), this was not detectable statistically in this trial. Under ‘salt stress’, both AS and CaE-treated plants have significantly higher fresh yield compared to the control.

In the case of drought-stressed plants, it can be seen how the crops suffered under the stress of water shortage, but specifically the pots treated with SF succumbed to the drought stress. In this treatment, all the plants started to wither within the first week of stress application and eventually died in the following weeks. Hence, no plant material from this treatment was available at the time of harvest for further analyses. Under drought stress, CaE exhibited significantly higher fresh yield (16 ± 3.6 g pot⁻¹) in comparison to the control (9.4 ± 0.45 g pot⁻¹). It is already known that drought interferes with many important growth functions like photosynthesis, nutrient and water relations, and ultimately reduces productivity [33,34]. The low water potential in the soil and in the plant reduces developmental activities of the plant cells and tissues [35]. Although not significantly higher, the improved mean yield of plants fertilised with CaE compared to the other treatments could be potentially attributed to the presence of organic matter (OM) in this BBF. As per the product information provided by the producer for an earlier batch of the same product, the CaE contains about ~20% OM, and this product is sold as a commercial (organic) fertiliser. Application of organic amendments has exhibited ameliorative effects on plant growth under drought stress in some studies [36,37] and has also shown to improve soil microbial functions by enhancing the stability of the soil microbiome [38]. These effects of organic fertilisers could be due to the modulation of nutrient mobility and availability under varying soil moisture conditions resulting in enhanced plant growth [39,40]. In this trial, CaE is the only organic BBF, and, compared to the mineral treatments (both SF and mineral BBFs), CaE contributes OM into the soil (~0.6 g OM/1.6 kg soil), potentially benefiting the condition of induced drought stress in this scenario. The results of the current trial demonstrate that fertilisation with BBFs yielded noticeably resilient spinach in terms of withstanding drought stress, relative to spinach fertilised with SF. Since plant responses to drought stress depend on the plant species, intensity and duration of stress and also on the period of plant growth when affected by stress [41,42,43], further research on the tested BBFs and SF observing growth of different crops could give additional insights into the poor performance of SF under drought stress seen in this trial.

Uptake of N, S, Na and K in each pot was calculated from the elemental content detected in the analysed plant matter (

Table 5. Mean N, S, K and Na uptake (g pot⁻¹) of spinach presented with their standard deviation (n = 4).

		Treatments
Stressor	Nutrient Uptake (g pot−1)	Control	Synthetic Fertiliser	Ammonium Nitrate	Ammonium Sulphate	Concentrate After Evaporation
No stress	N	0.05 ± 0.03 c	0.09 ± 0.01 bc	0.11 ± 0.02 ab	0.14 ± 0.02 a	0.07 ± 0.01 bc
	S	0.003 ± 0.002	0.009 ± 0.001	0.005 ± 0.005	0.008 ± 0.006	0.008 ± 0.001
	K	0.09 ± 0.06 α	0.24 ± 0.05 β	0.23 ± 0.08 β	0.28 ± 0.04 β	0.21 ± 0.05 αβ
	Na	0.002 ± 0.001 *	0.004 ± 0.002 *°	0.004 ± 0.001 *°	0.009 ± 0.002 *°	0.015 ± 0.003 °
Salt stress	N	0.03 ± 0.03 a	0.11 ± 0.01 bc	0.10 ± 0.02 c	0.14 ± 0.02 b	0.10 ± 0.01 c
	S	0.003 ± 0.002 B	0.009 ± 0.002 A	0.002 ± 0.000 B	0.013 ± 0.002 A	0.009 ± 0.000 A
	K	0.13 ± 0.14	0.26 ± 0.06	0.26 ± 0.07	0.30 ± 0.06	0.30 ± 0.01
	Na	0.009 ± 0.006 *	0.019 ± 0.007 *°	0.014 ± 0.002 *°	0.021 ± 0.006 *°	0.035 ± 0.014 °
Drought stress	N	0.03 ± 0.00	n.a.	0.09 ± 0.05	0.07 ± 0.01	0.07 ± 0.02
	S	0.003 ± 0.001 C	n.a.	0.001 ± 0.001 B	0.006 ± 0.001 A	0.006 ± 0.001 A
	K	0.10 ± 0.02	n.a.	0.17 ± 0.10	0.13 ± 0.01	0.17 ± 0.04
	Na	0.002 ± 0.00 *	n.a.	0.004 ± 0.00 *°	0.002 ± 0.00 *°	0.013 ± 0.00 °

n.a.: not applicable. All treatments are compared at individual stress categories for differences in each nutrient uptake using the parametric ANOVA test (p < 0.05) followed by Tukey’s post hoc pairwise comparison. When the assumptions for the standard ANOVA are unmet, i.e., the Normality test (p < 0.050) or the test of Variance (p < 0.050) failed, the non-parametric ANOVA (Kruskal–Wallis test) is performed (p < 0.05) followed by Dunn’s post hoc pairwise comparison. Significant differences between treatments of each category of stress are indicated by lower-case letters for N uptake, upper-case letters for S uptake, symbols α and β for K uptake and the symbols * and ° for Na uptake. Comparisons between treatments are performed within the same stressor category only (e.g., treatment ‘Control’ is compared with treatment ‘SF’ under ‘No stress’ only and not with SF under ‘Salt stress’ OR ‘Drought stress’). Drought-stressed SF plants are dead at harvest; hence, no nutrient uptake can be calculated.

). As expected, N uptake was lowest in the control treatment under unstressed and salt-stressed conditions. However, under unstressed conditions, the N uptake in the control was similar to that of SF and CaE. Plants treated with AS exhibited high N uptake in unstressed and salt-stressed plants, similar to AN under unstressed and similar to SF under salt-stressed conditions.

Drought-stressed plants had poor nutrient uptake, in general, and due to high variability within treatments, significant differences were not observed between treatments for N uptake. However, in the case of S uptake, it could be observed that SF, AS and CaE-fertilised drought-stressed plants had comparable S uptake, and significantly higher S uptake compared to control and AN-fertilised plants. The same trend was observed in the case of salt-stressed plants, whereas unstressed crops showed no significant differences between treatments for S uptake.

K uptake only exhibited differences under the unstressed category, where control treatment had significantly lower K uptake compared to SF, AN and AS. Although CaE seemingly exhibits higher mean K uptake compared to control, there was no significant difference between the two treatments for K uptake (p = 0.102 by Tukey’s pairwise comparison).

All sets of stressor categories demonstrate significantly higher Na uptake in CaE-fertilised plants compared to the control. While mean Na uptake in CaE was markedly higher relative to other fertilisations, statistical evaluation did not mirror these findings (p > 0.05 by Dunn’s pairwise comparison). To ensure that the added NaCl produced the intended salinity effect, a two-way ANOVA test was performed (excluding the SF treatment due to missing values under SF-drought-stressed plants, which precluded interaction analysis) to evaluate the effects of BBF treatment and stressor on Na uptake. This indicated that applied salt stress led to higher Na uptake rates in all treatments in the salt-stress category (p < 0.001), and, among the BBF treatments, this effect was most pronounced in CaE, followed by AS and AN.

Induced salinity stress, as mentioned earlier, seemed to have no negative impact on the yield of spinach, as compared to the unstressed set of treatments. However, studies that followed a similar set-up and NaCl application rates were able to see more pronounced effects [27,29,44]. This could be potentially explained by the nutrient uptake interactions that were observed in these plants. High concentrations of NaCl exposure known to adversely affect crop performance due to the salinity-induced nutritional imbalance like the Na⁺ antagonism to K⁺ and Cl⁻ antagonism to NO₃⁻ [45] were the reason to choose this compound as a salt stressor in this experiment. But, under the current experimental conditions, it was observed that plants were not impacted by the applied salt stress, and exhibited yield at par with the unstressed set of treatments. Previous studies have tried to evaluate the impact of different levels of K to alleviate the induced salt toxicity from NaCl and have found some incongruous conclusions [46,47,48,49]. Although the 2018 study [49] concluded that the observed salt tolerance in spinach was a result of specific gene expressions and that spinach is possibly more salt-tolerant than previously thought, the other three studies pointed to a correlation between increasing K⁺ observed to be alleviating the higher Na⁺ found in plant tissues. In this experiment, since K was added in excess to the actual crop requirement as mentioned in the experimental description, all fertilised plants received 23% additional K (than the recommended/optimal K dose).

Generally, in an event of induced salt-stress, the plants could have an inhibition in the translocation of mineral nutrients, especially K⁺. Due to their similar physiochemical properties, Na⁺ can compete with K⁺ for the binding sites in metabolic processes in plants, thus disturbing different metabolic processes [50,51]. Increased saline conditions can interfere with the K functioning in multiple ways: (i) high Na⁺ can inhibit K⁺ activity in soil reducing K uptake; (ii) in low K systems, Na⁺ interferes with K⁺ translocation from root to shoot, and, in general, Na⁺ also competes with K⁺ for uptake sites at the plasma membrane, again reducing K uptake [52]; (iii) salinity stress can cause plasma membrane disruption and favours K⁺ leakage that results in rapid cytosolic K⁺ reduction [53]; (iv) significant membrane depolarisation and K⁺ leaking through depolarisation-activated outward-rectifying K⁺ channels [51]. A solution to dealing with the aforementioned Na⁺ and K⁺ interactions is stated in a study [46], as maintenance of K⁺ level above a threshold and maintaining a high cytosolic K⁺/Na⁺. Upon examining the correlation between yield and K uptake, strong positive correlation was observed in all cases with an r value of 0.838, 0.936 and 0.974 for crops grown under no stress, salt-stress and drought stress conditions, respectively (p < 0.001 in all cases). A higher K supply corresponded to increased K⁺ in the plant tissues, potentially reducing the negative impacts associated with Na⁺ concentration. On the other hand, it has been deduced in studies that, in K-deficient systems, the negative impacts of salinity were significantly increased during photosynthetic activity and were accompanied by increased salt sensitivity, inhibition of N and photosynthetic C assimilation [54,55,56]. In the context of the tested BBFs, these findings highlight the potential value of products like CaE with high K to be used in regions with higher soil salinity. Additionally, this strategy can also be incorporated when fertilising with RENURE-criteria materials like the two NH₄⁺ salts (AN and AS) tested in this study. Incorporating K-rich BBFs into fertilisation plans may provide both essential nutrient support and help mitigate salt stress. It is well established that NH₄⁺ salts can contribute to increased soil salinity, particularly during dry seasons or in saline soils. Therefore, supplementing K under these conditions could be advantageous for salt-sensitive crops.

Although this trial primarily focusses on yield and nutrient uptake to draw conclusions, it is strongly recommended that future research investigate the plant physiological changes, stomatal conductance, chlorophyll index, etc., to monitor the impacts of stress.

3.2. ANR and NFRV

The calculation of ANR was done to understand how efficiently the crop utilises the applied N within the induced stress conditions. Unstressed and salt-stressed plants treated with AS exhibited significantly higher ANR (1.4 and 1.8, respectively) compared to the other fertilisations (Figure 1). This outcome could be attributed to the higher N uptake observed in these plants, which could be a potential consequence of the higher S fertilisation and uptake in AS-fertilised crops. Increasing S application dosage has been observed to positively impact the N uptake in certain crops, where, specifically in the case of spinach, there was an observed considerable increase in the amino acid cysteine in a past study [57]. In the current experiment, the salt-stressed plants demonstrate this N-S interaction more prominently and with statistical significance, where a positive correlation between the N and S uptake was observed for the salt-stressed plants (Pearson’s co-efficient r = 0.72, p < 0.001).

NFRV signifies the amount of synthetic fertiliser saved when replaced with a BBF while attaining the same crop yield, i.e., this indicator demonstrates the suitability of a BBF relative to a synthetic N fertiliser [13]. This indicator is commonly used to assess the agronomic suitability of BBFs compared to conventional synthetic N fertilisers, due to its direct reflectivity on the practical substitution potential of alternative products. Hence, this indicator is very relevant to the RENURE-criteria products tested in this study. NFRV depends on ANR and, hence, the N uptake results, thus showcasing a similar pattern in outcome. Both the unstressed and salt-stressed AS-treated plants exhibited significantly higher NFRV (3.1 and 1.8, respectively), compared to AN (1.9 and 0.91, respectively) and CaE-treated plants (0.59 and 0.80, respectively). Overall, the NH₄⁺ salts show better performance and potential compared to the tested SF based on their NFRV. Since NFRV is a direct comparison of the BBF with SF, and no live plants were obtained for the drought-stressed SF treatment, NFRV could not be calculated for drought-stressed BBF-fertilised plants. And this directly indicates that, for drought-stressed plants, BBF performance can be said to be better compared to the SF treatment under the current experimental conditions.

3.3. Effect of BBF Application on Soil

Soil from each pot was assessed to determine effects, if any, of fertiliser application and induced stress (two-way ANOVA followed by multiple comparisons using the Bonferroni post hoc test).

Table 6. Mean ± standard deviation of mineral N (NH₄⁺-N and NO₃⁻-N) on fresh weight and pH and electrical conductivity (EC) of air-dried soil (n = 4) presented for each stress category.

		Treatments
Stressor	Soil Parameters	Control	Synthetic Fertiliser	Ammonium Nitrate	Ammonium Sulphate	Concentrate After Evaporation
No stress	NH4+-N (mg kg−1)	2.7 ± 0.97	3.0 ± 1.1	1.9 ± 0.41	2.3 ± 1.0	2.2 ± 0.62
	NO3−-N (mg kg−1)	4.6 ± 6.9 a	51 ± 51 b	0.60 ± 0.29 a	11 ± 20 ab	1.7 ± 2.9 a
	pH	5.69 ± 0.06 C	5.34 ± 0.05 A	5.57 ± 0.09 C	5.21 ± 0.03 A	5.84 ± 0.09 B
	EC (µS cm−1)	92 ± 25 α	223 ± 32 β	102 ± 12 α	168 ± 13 αβ	146 ± 13 αβ
Salt stress	NH4+-N (mg kg−1)	2.3 ± 0.28	2.0 ± 0.38	2.3 ± 0.45	2.2 ± 0.76	2.2 ± 0.77
	NO3−-N (mg kg−1)	7.8 ± 9.9	18 ± 16	1.8 ± 2.8	12 ± 21	1.7 ± 2.9
	pH	5.74 ± 0.02 A	5.41 ± 0.08 C	5.64 ± 0.07 AB	5.31 ± 0.04 C	5.69 ± 0.08 B
	EC (µS cm−1)	440 ± 26 αβ	530 ± 68 αβ	427 ± 51 β	483 ± 68 αβ	551 ± 165 α
Drought stress	NH4+-N (mg kg−1)	2.9 ± 0.61	2.8 ± 0.62	3.1 ± 1.1	2.2 ± 0.42	2.6 ± 0.68
	NO3−-N (mg kg−1)	1.2 ± 0.36 c	132 ± 32 a	46 ± 22 b	26 ± 19 bc	2.1 ± 2.2 c
	pH	5.76 ± 0.04	5.41 ± 0.05	5.48 ± 0.06	5.30 ± 0.04	5.79 ± 0.10
	EC (µS cm−1)	91 ± 29 γ	263 ± 32 αβ	151 ± 11 βγ	283 ± 30 α	120 ± 20 γ

All soil parameters are compared based on both the effect of treatment and effect of stress using the two-way ANOVA, followed by multiple comparison using the Bonferroni post hoc stress, where effect of treatment, stress and Treatment × Stress are determined (p < 0.05). In the case of NO₃⁻-N, the normality and variability test failed (p < 0.05) due to high variability between replicates. The ANOVA test for independent factors and factor interactions were significant (p < 0.01). Lower-case letters indicate significant differences for NO₃⁻N, upper-case letters for pH and symbols α, β, γ indicate differences in the case of EC of soil. Significant differences between interactions (Treatment × Stressor) are not indicated in the table.

presents the results of mineral N, pH and EC of soil, the key parameters that were seen to exhibit noticeable differences between treatments. Mineral N (NH₄⁺-N and NO₃⁻-N) analysis of soil post-harvest helped to shed light on the impacts of stress on N use by plants. The two-way ANOVA test was performed on soil mineral N to understand effects of both fertiliser application and stressor used. Most NH₄⁺-N would have converted to NO₃⁻-N, which the plant utilised for its growth; no significant differences under any of the treatment or stressor conditions were observed. However, large variabilities within and differences between treatments were observed in the soil NO₃⁻-N results. Within unstressed and drought-stressed conditions, SF-treated soil exhibited the highest NO₃⁻-N, albeit the very high variability seen under unstressed (also in salt-stressed) SF treatments. Soil samples from this treatment were analysed twice to rule out analytical errors, but both analyses gave similar outcomes. The wide spread of data appears to reflect inherent variability rather than measurement inconsistencies, making it inappropriate to exclude any values as outliers. Under drought-stressed conditions, the high-soil NO₃⁻-N in the SF-treated soil further validates the poor growth and eventual wilting of the crops, which is potentially an aftermath of low N mobility under low moisture-conditions. In general, higher variability is common in the case of soil NO₃⁻-N, but no specific reasons could be pointed to in this trial. Comparison of stressors within individual treatment (treatment × stressor) showed significantly higher NO₃⁻-N in the case of water-stressed soil samples in SF (p < 0.001) and AN treatment (p < 0.05), as expected, due to lower uptake. Salt-stressed soil had similar NO₃⁻-N comparable to their unstressed counterparts, indicating no effect on N mobility under conditions of higher than usual salinity.

From interaction with local farmers, it has been understood that there is concern about the potential of recovered nutrients’ impact on soil pH, especially for RENURE materials like AS that can acidify soils (when the NH₄⁺ is converted to NO₃⁻, an H⁺ is released into the soil, thus increasing its pH). In this trial, it can be observed that, among pots fertilised with different BBFs, AS was seen to have significantly lower pH between the tested treatments but at par with the synthetic N treatment. Using AS on acidic soils could use some caution, especially if liming is not a common practice. Additionally, it is recommended to validate these results in fields, especially under long-term application of BBFs.

EC of soil was monitored to oversee the impacts of the RENURE NH₄⁺ salts like AN and AS tested in this trial, i.e., if these high-salt BBFs change the soil EC. Also, EC analysis was another parameter used to ensure that the inducement of salt stress was performed adequately in the experiment to satisfy the stress condition. Despite the recommended dose of NaCl application for stress inducement, we find that the soil EC values were relatively lower at the time of analysis. This could be attributed to loss of some of the salt solution as a leachate and/or because of the air-drying and long storage of the soil samples before analyses (due to practical limitations). It has been reported that the handling and storage of samples could affect the EC in soil extracts [58,59]. This could lead to redistribution or precipitation of soluble salts, thus resulting in lower EC. Hence, the reported EC values could likely underestimate the actual salinity experienced by the plants during the experimental duration. Nevertheless, the stressed soil samples were found to have higher EC than the unstressed soil samples by a factor of ~5, 2, 4, 3, and 4 times for control, SF, AN, AS and CaE, respectively.

It was observed that, among the fertilised unstressed plants, SF-treated soil had high EC, comparable to both AS and CaE. In the case of SF, this is in tandem with the exceptionally high NO₃⁻-N observed in the soil as mentioned earlier. Nitrate loading in groundwater from agricultural use has been shown to have a direct effect on groundwater EC [60]. Among the RENURE-criteria NH₄⁺ salts, soil treated with AN had significantly lower EC (102 µS cm⁻¹), relative to the SF treatment (223 µS cm⁻¹). The higher EC in the SF (CAN) can be attributed to the additional ionic strength imparted by the Ca ions in CAN.

Among drought-stressed treatments, AS had significantly higher EC compared to the other treatments (283 µS cm⁻¹). This can be attributed to the high salt index of AS, which can increase the concentration of soluble ions in the soil, which accumulates further during lack of moisture due to drought [61]. The AS would dissociate into NH₄⁺ and sulphate (SO₄²⁻) ions, compared to the AN that dissociates into NH₄⁺ and NO₃⁻ ions. Since SO₄²⁻ ions have a higher charge relative to the NO₃⁻ ions, the ionic strength of AS application would be stronger than that of AN application. Comparison between different stressors on the same treatment was performed, and as expected, all salt-stressed soil samples exhibited significantly higher EC compared to their unstressed and drought-stressed counterparts (p < 0.05).

It can be concluded that, for this trial, the treatment with the RENURE-criteria NH₄⁺ salts, or induced salinity by NaCl addition did not negatively affect the soil parameters in a way that could compromise agronomic performance of the fertilisers. However, nitrate content variability requires further study.

4. Conclusions

This study evaluated the effect of N-based BBFs, specifically two high-priority RENURE materials like AN and AS, and another commercial BBF (CaE) in the growth of spinach under induced abiotic stressors like salinity and drought. Overall, the results of this study indicate comparable agronomic performance demonstrated by all three tested BBFs, compared to the SF. Plants fertilised with the BBFs exhibited comparable yield and N, K, S and Na uptake. Under drought stress, all the SF-fertilised plants withered away, leaving no plant biomass for further analyses. On the other hand, plants fertilised with BBFs could withstand the condition of drought, although with a poorer yield than their unstressed and salt-stressed counterparts. The lack of sufficient moisture stunted their growth and gave poor quality and minimal biomass, which were indicative of poor nutrient uptake. The possible reason for the resilience of crops grown in salinity stress is the increased K in plant tissues, which could have alleviated the negative impacts of increased Na in the system. However, we recommend stricter salinity control during the experiment to also ensure that the plant is subjected to sufficient stress to make stronger conclusions. Significantly higher N recovery (measured by ANR) and N replaceability value (measured by NFRV) were observed in AS-treated plants, potentially from the excessive S fertilisation that could have boosted the N uptake, which was observed especially in the salt-stressed plants. The interactions of Na-K and N-S, under salinity stress need to be further delved into, to understand how to alleviate and manage abiotic stress conditions, which can be expected globally in the near future. The aim of this study was to learn about agronomic performance of novel BBFs, specifically of the RENURE products, under stress conditions, which have yielded favourable outcomes for these materials. BBFs such as CaE with high K content warrant further study to better understand their role in reducing abiotic stress and enhancing both crop fertilisation and resilience to soil or climate changes.

We recommend further trials that can also evaluate plant physiological indicators, like chlorophyll index, stomatal conductance, etc. which can give a deeper understanding of stress effects, and identify the absence of this information as a limitation in the current trial.