Synergistic effects of phosphogypsum and wheat varieties on saline–sodic soil properties and wheat productivity in Ethiopia

Abstract

Soil salinity and sodicity are major constraints, limiting nutrient availability, degrading agroecosystems, and threatening agricultural productivity. In Ethiopia, most irrigated lowland areas are affected by salinity and/or sodicity, which sharply reduces crop yields. Addressing this problem requires effective and timely management strategies. Field and pot experiments were conducted on saline–sodic soils at the Metehara Sugar Estate during the 2022/23 and 2023/24 growing seasons to evaluate the synergistic effects of phosphogypsum (PG) application and wheat varieties on selected soil physicochemical properties and wheat performance. Factorial combinations of PG rates and wheat varieties were tested using a completely randomized design for pots and a randomized complete block design for field trials, each with three replicates. Soil samples were collected before soil amendment and after harvest to assess changes in physicochemical properties. Application of PG significantly improved wheat growth, physiology, and yield, as well as soil properties. The most effective rates, 150% and 200% of the gypsum requirement (GR), reduced soil pH from 8.16 to 7.93–7.91 and lowered exchangeable sodium percentage (ESP) from 34.3% to 11.0–7.9%. Applying PG alone at 150% or 200% GR rate, or growing the salt-tolerant wheat variety ETBW-5879, enhanced physiological traits such as relative water content, chlorophyll SPAD content, photosynthesis, and transpiration. Application of PG at 150 or 200% GR, combined with the salt-tolerant variety ETBW-5879, resulted in superior performance in terms of plant height, leaf area, above-ground dry biomass, and grain yield. At PG 200% GR, ETBW-5879 increased grain yield by 107% over untreated Gambo and by 224% over untreated Dande’a, while at PG 150% GR, the increases were 103% and 217%, respectively. Even application of PG at 100% GR with ETBW-5879 produced a grain yield that was 6% higher than PG at 200% GR with Gambo and 21% higher than PG at 200% GR with Dande’a. These results indicate that applying PG at 150% GR in combination with ETBW-5879 optimizes wheat yield on saline–sodic soils at Metehara. However, further validation across other saline–sodic soils and multi-year field trials in Ethiopia is necessary to draw robust conclusions.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-026-55507-8.

Introduction

The global population is projected to increase from 8.5 billion in 2030 to a peak of 10.4 billion by 2080¹, driving greater food demand and intensifying pressure to expand agriculture into lowland irrigated areas. However, the combined effects of irrigated land mismanagement and climate change are aggravating land degradation, resulting in widespread secondary salinization and sodification across arid, semi-arid, and coastal regions. These processes severely constrain agricultural productivity^2,3. In sub-Saharan Africa, salt-affected soils are estimated to cover between 19 million ha⁴ and 69 million ha⁵. In Ethiopia’s Central Rift Valley, salinity and sodicity are well-documented constraints that significantly limit crop production^2,6. At the Metehara Sugar Estate, sugarcane yields have declined markedly, with 6% (779.6 ha) of land classified as severely salinized, 20% (2,474.0 ha) as moderately salinized, and 24% (2,999.3 ha) as slightly salinized⁷. These problems are aggravated by poor irrigation practices and inadequate drainage systems, which raise groundwater levels near the soil surface and intensify secondary salinization and sodification^2,6,7.

Ethiopia is one of the largest wheat producers in sub-Saharan Africa, and wheat plays a critical role in national food security and rural livelihoods⁸. Nevertheless, domestic production has not kept pace with increasing demand. Rapid population growth, urbanization, shifting dietary preferences toward wheat-based foods, climate variability, and rising food prices have widened the demand–supply gap^8,9. As a result, Ethiopia spends more than USD 700 million annually on wheat imports¹⁰, placing significant pressure on foreign exchange reserves and highlighting the urgency of improving local productivity. Although the government aims to achieve wheat self-sufficiency through productivity gains and expansion of irrigated agriculture, the increasing prevalence of salt-affected soils in irrigated lowlands remains a major constraint. This challenge underscores the need for effective reclamation strategies, particularly for sodic and saline–sodic soils, which often require substantial investment in chemical amendments¹¹.

Reclamation of sodic and saline–sodic soils typically involve the application of soluble Ca²⁺ sources to replace exchangeable Na on the soil exchange complex and facilitate its leaching. Gypsum and gypsum-like by-products are widely used for this purpose^6,12. Among these, PG has emerged as a cost-effective and readily available amendment, extensively utilized in countries such as Morocco, Pakistan, Egypt, India, Brazil, and Mexico¹³. Application of PG improves soil chemical and physical properties by reducing ESP, supplying Ca²⁺, P, and SO₄²⁻, and enhancing soil aggregation, hydraulic conductivity, and moisture retention^12,14,15. However, its agronomic effectiveness is highly context-specific and depends on soil characteristics, crop genotype, and management practices.

Although several studies have reported improvements in wheat growth and yield components following PG application in saline-sodic and sodic soils^12,16, most have evaluated PG as a standalone amendment without explicitly considering plant genetic responses. In Ethiopia, research on PG application is still very limited, with only one published study¹² reporting its effects on wheat growth and soil chemical properties under the sodic soils of Alage. Importantly, no study has systematically investigated the interaction between PG application rates and salt-tolerant wheat genotypes in relation to both soil reclamation dynamics and crop performance under saline–sodic environments. This represents a critical knowledge gap, particularly in the context of integrated soil–plant management approaches.

Wheat is moderately salt-tolerant, with threshold salinity levels of 6 dS m⁻¹ for soil and 4 dS m⁻¹ for irrigation water^17,18. However, substantial varietal differences in salt tolerance exist. For example, Nekir et al.¹¹ reported that the ETBW-5879 (Honqolo) variety produced the highest grain and biomass yields under severe salt stress conditions (up to 20 dS m⁻¹). Integrating salt-tolerant varieties with appropriate soil amendments, therefore, represents a promising strategy to enhance both reclamation efficiency and crop productivity.

The novelty of this study lies in its integrative approach, which simultaneously evaluates the combined effects of PG application rates and wheat varietal responses on soil physicochemical reclamation and crop productivity under saline–sodic soil conditions. Unlike previous studies that have treated soil amendment and crop selection independently, this research explicitly examines their interaction as a unified management strategy, providing new insights into the synergistic mechanisms governing soil–plant responses in salt-affected systems. It is hypothesized that: (i) increasing PG application rates will significantly reduce soil sodicity indicators through enhanced Ca²⁺-mediated Na displacement and leaching; (ii) salt-tolerant wheat varieties will exhibit superior growth and yield performance under saline–sodic conditions due to improved physiological and adaptive traits; and (iii) the combined application of optimal PG rates and salt-tolerant varieties will produce a synergistic effect, resulting in greater improvements in soil properties and wheat productivity than either approach alone. Accordingly, this study evaluates the interaction effects of varying PG rates and wheat varieties on soil physicochemical properties and wheat productivity at the Metehara Sugar Estate.

Materials and methods

Experimental site and climate

The experiments were conducted under controlled pot and field conditions using various methodological approaches. Two-season pot experiments (2022/23 and 2023/24) were conducted in a lath house, a simple shelter constructed from wire netting, at Hawassa University (7.0558° N, 38.4717° E) using saline–sodic soils collected from the Metehara Sugar Estate. A one-season field experiment (2023/24) was conducted on the farmland of the Metehara Sugar Estate (8.8805° N, 39.9495° E; Supplementary Fig. 1) at an altitude of 955 m above sea level. The climate of the study area is classified as a warm semiarid (SA2) agroecological zone¹⁹. Based on agro-meteorological data collected from the Metehara sugar estate between 1985 and 2022 (Supplementary Fig. 2), the area receives a mean annual rainfall of 549 mm, with average annual minimum and maximum temperatures of 17 °C and 33 °C, respectively. The annual reference crop evapotranspiration (ETo) is estimated at 2059.6 mm, with mean daily ETo ranging from 5.0 to 6.2 mm day⁻¹. The site is a key irrigated wheat-growing area in the Central Rift Valley of Ethiopia, where soil salinity and sodicity are the main constraints to crop production, as in most irrigated lowland areas of the country.

Properties of soils, irrigation water, and phosphogypsum

Surface soil samples (0–20 cm depth) were randomly collected from an 11.5-ha field (MZ-50) at the Metehara Sugar Estate using a manual shovel, as higher salt concentrations are typically found in surface soils compared to subsurface layers. Ten subsamples taken at 100 m intervals were combined to form a composite sample. The samples were air-dried, crushed, and passed through a 5-mm sieve for the pot experiment, and a portion was analyzed for selected physicochemical properties. The soil had a loam texture and was classified as saline–sodic according to the criteria of Allison and Richards²⁰, with ECe > 4 dS m⁻¹ and ESP > 15% (Supplementary Table 3). Tap water from Hawassa University was used to irrigate the pot experiment, while water from diversion weirs along the Awash River, before contamination from salty Lake Basaka, was used to irrigate the field experiment. Irrigation water samples from both sources were collected and analyzed (Supplementary Table 3). According to the irrigation water quality guidelines of Ayers and Westcot²¹, the water used in both experiments had ECw < 0.7 dS m⁻¹ and SARw < 3, indicating no adverse effects on nutrient or water availability for sensitive crops.

The PG material used in this study was obtained through a donation from Morocco’s Office Chérifien des Phosphates (OCP), whereas pure gypsum was collected from the Mojo Gypsum and Calcium Carbonate factory. Both PG and pure gypsum were characterized by high calcium and sulfate contents (Supplementary Table 4), indicating their suitability as calcium-source amendments for the reclamation of sodic and saline–sodic soils. Compared with pure gypsum, PG exhibited relatively higher acidity and greater available phosphorus content.

Treatment and gypsum requirements

The experiments involved two factors: wheat variety and PG levels. Three wheat varieties: Dande’a (salt-sensitive), Gambo (moderately tolerant), and ETBW-5879 (highly salt stress-tolerant) were used. These varieties were selected from a preliminary screening of 265 genetic materials from the Werer and Kulumsa Agricultural Research Centres and further evaluated for salt tolerance at the Werer Research Centre. All selected varieties are well adapted to the Rift Valley agro-ecological zones of Ethiopia.

The doses of PG applied in the pot experiment were 0% GR (control), 50% GR (2.14 g kg^-1), 100% GR (4.28 g kg^-1), 150% GR (6.42 g kg^-1), and 200% GR (8.55 g kg^-1). A positive control consisting of pure gypsum at 100% GR (2.78 g kg^-1) was also included. In the field experiment, the corresponding PG rates were 0, 9.4, 18.9, 28.4, and 37.8 Mg ha⁻¹, with a positive control of pure gypsum at 12.3 Mg ha⁻¹. Gypsum requirement (GR) rates were calculated using a soil bulk density of 1.23 g cm⁻³ and a soil depth of 15 cm, following a modified method of FAO²² (Eq. 1).

1

where; GR is the gypsum requirement (g kg^-1) of soil; ESP_i is actual (initial) exchangeable sodium percentage (%); ESP_f is the final ESP to be reached after reclamation (%); CEC is the cation exchange capacity (Cmolc kg^-1); 0.86 g is the amount of CaSO₄·2H₂O required to reduce the sodium content of the soil²⁰, F is the purity of CaSO₄ (88%).

The rate of phosphogypsum requirement (phosphogypsum application to the soil) was calculated from the quantity of gypsum requirement, as given in Eq. (2).

2

where; PG_R is the quantity of PG required (Mg ha⁻¹), GR is the quantity of gypsum required (Mg ha⁻¹), and 0.65 is the conversion factor between PG and pure gypsum²³. This factor is used because PG is largely composed of calcium sulfate dihydrate (CaSO₄·2H₂O), the same compound as gypsum, but it contains impurities such as P₂O₅, fluoride, organic matter, and trace elements, which reduce its effective gypsum equivalence.

Experimental design and agronomic management

A factorial combination of 18 treatments (3 wheat varieties and 6 PG rates) was arranged in a completely randomized design with three replications, using 54 packed pots. Pots with a height of 30 cm and a diameter of 25 cm were used in the experiment. In the pot experiment, each pot was filled with 5 kg of soil and had a surface area of 490.9 cm². The amount of soil added to each pot was carefully adjusted to approximate the bulk soil mass under field conditions, thereby ensuring that the experimental setup realistically represented field conditions in terms of soil volume, bulk density, and rooting environment. Phosphogypsum was applied once in both experimental conditions. It was finely ground and thoroughly mixed with the soil 30 days before planting to enhance reactivity and ensure uniform distribution and interaction within the soil matrix. Based on the defined surface area of each pot, a seed rate of 0.49 g per pot, equivalent to approximately ten seeds was sown. Following germination, the seedlings were thinned to seven plants per pot to maintain an appropriate plant population. This seeding rate was not arbitrarily chosen; rather, it was carefully calculated to correspond to the recommended field seed rate (100 kg ha⁻¹) when proportionally scaled down to the pot level, thereby ensuring consistency and comparability between the pot and field experiments.

A field experiment was conducted to evaluate the effectiveness of PG on different wheat varieties under field conditions. Before starting the field experiment, the site was uniformly plowed and pulverized with a tractor. The factorial treatments were arranged in a randomized complete block design with three replications (Supplementary Fig. 3). The plot size of the experimental unit was 20 m × 10 m (200 m²), giving 33 rows of 20 m length with a row spacing of 0.3 m. The PG was uniformly broadcast and thoroughly mixed into the soil by hand 30 days before planting to allow sufficient reaction time and ensure a homogeneous distribution throughout the treated soil depth. A seed rate of 100 kg ha⁻¹ was used for each variety, with nitrogen fertilizer applied as urea (46 kg N ha⁻¹) in two splits at tillering and anthesis. Weeds and insects were controlled with Pallas-45 herbicide and Karate 5% EC insecticide, respectively.

The sowing dates for the pot and field experiments were done during the off-season for irrigated wheat crop production. For the pot experiments, sowing occurred on October 23, 2022, for the first season and on November 2, 2023, for the second season. Conversely, sowing was performed on October 18, 2023, during the cropping season for the field experiment. The experiments included multiple harvests to account for wheat variety differences and treatment effects on maturity (Supplementary Tables 1 and 2).

Crop data collection and measurement

All crop growth, physiological, and yield trait data were collected from five plants in the pot experiment (pot area = 490.9 cm²) and from four representative quadrants per plot in the field experiment (each quadrant measuring 1 m × 1 m = 1 m²). At physiological maturity, plant height was measured in the pot experiment and in the field by sampling five plants per quadrant. Aboveground dry biomass and grain yield were determined at harvest for both experiments. Plants were manually harvested and threshed, with grain moisture standardized to 12.5%. Aboveground biomass was oven-dried at 70 °C for 72 h before weighing.

Leaf area and relative water content were measured on flag leaves during anthesis. Leaf area was measured from 10 randomly selected leaves using a LI-3000C portable area meter. Relative water content (RWC) was measured immediately after cutting the lamina base, with the leaves sealed in plastic bags and transported to the laboratory. Fresh weight (FW) was measured after excision and full turgid weight (TW) after hydrating leaves in 100 ml distilled water for 24 h. After soaking, the leaves were blotted and dried using tissue paper. Dry weight (DW) was measured after oven drying at 70 °C for 72 h. Relative water content was calculated using Eq. (3) by Schonfeld et al.²⁴ as follows:

3

Chlorophyll content and leaf gas exchange parameters were measured from 10 randomly selected leaves at the initiation of the flag leaf stage. Chlorophyll content was quantified using a SPAD-502 chlorophyll analyzer (Soil–Plant Analyses Development, Minolta Camera Co., Ltd., Osaka, Japan), as described by Zhang et al.²⁵. Soil–Plant Analyses Development measurements were conducted between 7:00 and 9:00 am to mitigate the influence of light intensity on chloroplast movement. SPAD readings were obtained at the midpoint of each leaf on one side of the midrib, and the average values were subsequently recorded. Leaf gas exchange parameters were evaluated using a CIRAS-3 portable photosynthesis system (CIRAS-3 PP system Inc.) on the 50^th day after emergence. The photosynthesis rate [μmol (CO₂) m^-2 s^-1] and transpiration rate [mmol (H₂O) m^-2 s^-1] were measured from the youngest fully expanded leaves taken on sunny days between 9:00 and 11:00 a.m. All photosynthetic measurements were conducted under actual field and pot conditions, using ambient light intensity, temperature, and atmospheric CO₂ concentration to ensure optimal plant physiological performance.

Soil sampling and preparation

Soil samples from the pot experiment were collected at the end of the second growing season, whereas samples from the field experiment were collected at the end of the first growing season. The samples were air-dried and crushed to pass through a 2 mm sieve, then analyzed for selected soil properties to assess the effects of PG application on soil conditions across the three wheat varieties.

Laboratory analyses

Phosphogypsum and pure gypsum analyses

Four grams of PG and pure gypsum, on an oven-dry basis (80–85 °C), were sieved to < 2 mm and transferred into 250 mL bottles. The samples were mixed with 200 mL of distilled water and shaken at low speed for 1 h. The suspensions were subsequently filtered through Whatman No. 41 filter paper. The filtrates were analyzed for EC and pH using a digital conductivity meter and a pH meter, respectively²⁰. Available P was determined using the modified Olsen method²⁶, and SO₄^2- was determined by precipitation using the barium sulfate method described by Allison and Richards²⁰. Basic cations (Ca, Mg, Na, and K) were extracted using 1 M ammonium acetate at pH 7²⁷. Ca and Mg were measured using atomic absorption spectrophotometry (AAS), whereas Na and K were determined using a flame photometer.

Soil and water analyses

Particle size distribution was determined using the Bouyoucos hydrometer method²⁸, and the soil textural class was determined using a textural triangle²⁹. Soil bulk density was determined using the core method³⁰. Soil pH, electrical conductivity (ECe), soluble cations (Ca²⁺, Mg²⁺, K⁺, and Na⁺), and anions (CO₃²⁻, HCO₃⁻, Cl⁻ and SO₄²⁻) were determined from saturated paste extract following the methods described by Allison and Richards²⁰. The extracted solution was used to determine the soil pH with a digital pH meter, ECe with a conductivity meter, soluble Ca²⁺ and Mg²⁺ with an AAS, and soluble Na⁺ and K⁺ with a flame photometer. Chlorine was determined by titration against 0.1 N AgNO₃ method, whereas SO₄^2- was analyzed by precipitation using the barium sulfate method described by Allison and Richards²⁰. The CO₃^- and HCO₃^2- ions were determined by titrating with sulfuric acid (H₂SO₄) to phenolphthalein and methyl orange endpoints, respectively²⁰. The sodium adsorption ratio (SAR) of the soil solution was calculated from the concentrations of soluble Na⁺, Ca²⁺, and Mg²⁺, as shown in Eq. (4).

4

Organic carbon was analyzed by the method described in Walkley and Black³¹, and available P was determined using the modified Olsen method²⁶. Cation exchange capacity and exchangeable basic cations (Ca, Mg, Na, and K) were extracted using 1 M ammonium acetate (pH 7) by the method described in van-Reeuwijk²⁷. Exchangeable Ca and Mg were determined using AAS, while a flame photometer was used for Na and K measurements. ESP was calculated using Eq. (5).

5

The irrigation water used in the pot and field experiments was analyzed for electrical conductivity (ECw), pHw, and the concentrations of water-soluble cations (Ca²⁺, Mg²⁺, K⁺, and Na⁺) and anions (CO₃²⁻, HCO₃⁻, Cl⁻ and SO₄²⁻). The sodium adsorption ratio (SAR) was calculated using Eq. 4. The adjusted sodium adsorption ratio (adj-SAR) modifies the traditional SAR by accounting for calcium precipitation or dissolution (Cax²⁺), ionic strength, and HCO₃^-/Ca²⁺ ratio²¹. It is computed from leaching water Cax²⁺, Mg²⁺, and Na⁺ concentrations, as shown in Eq. (6). The chemical characteristics of the irrigation water are presented in Table 1.

6

Table 1.

Changes in particle size distribution and bulk density (BD) as influenced by PG application on saline-sodic soil.

Treatments	Clay (%)		Silt (%)		Sand (%)		Texture		BD (g cm-3) from field
	Pot†	Field¥	Pot†	Field¥	Pot†	Field¥	Pot†	Field¥
Phosphogypsum
Initial value	21.52		28.40		50.08		L		1.23
PG/G 0% GR	20.5c	20.7c	30.1a	30.7a	49.4a	48.6a	L	L	1.26a
PG 50% GR	20.9b	21.0bac	31.2a	30.9a	47.9a	48.1a	L	L	1.23b
PG 100% GR	21.2a	21.2ba	29.7a	31.0a	49.1a	47.8a	L	L	1.20 cd
PG 150% GR	21.4a	21.5a	30.7a	30.5a	47.9a	48.0a	L	L	1.18ed
PG 200% GR	21.5a	21.4a	30.6a	30.8a	47.9a	47.8a	L	L	1.16e
G 100% GR	21.1ba	20.9cb	30.3a	31.1a	48.6a	48.0a	L	L	1.22cb
p value	0.02	0.0003	0.26	0.12	0.12	0.08			0.0001

PG = phosphogypsum; G = pure gypsum; L= loam; Means followed by the same letters within a column do not differ statistically at P < 0.05; † = the pot experiment conducted for two seasons; ¥ = the field experiment conducted for one season.

Net dispersive charges and cation ratio of soil structural stability

The concept of net dispersive charge, defined as the difference between dispersive and flocculating charges, provides a clear understanding of clay behavior in soils. Rengasamy et al.³² introduced a model for the dispersive charge (DC) of soils, which is defined as the total dispersive charge contributed by all exchangeable cations (Na, K, Mg, and Ca) in relation to exchangeable Na (Eq. 7). Conversely, the flocculating charge (FC) of the soil solution is determined by the presence of electrolytes or water-soluble cations (Ca²⁺, Mg²⁺, K⁺ and Na⁺) in relation to water-soluble Ca²⁺ (Eq. 8). The net dispersive charge was calculated based on the amount of clay dispersed, as expressed in Eq. (9).

7

8

9

where: the concentration of soluble cations in the flocculated soil–water suspension is expressed as Cmolc kg^-1 (on soil basis) in Eq. (8), and the concentration of exchangeable cations is expressed Cmolc kg⁻¹ in Eq. (7). The cation ratio of soil structural stability (CROSS) was calculated from water-soluble cations, according to Rengasamy and Marchuk³³, as shown in Eq. (10).

10

where: the unit of each ion concentration is meq 1^–1 in Eq. (10).

Following the harvest of the wheat crop, soil samples were collected and analyzed for ECe, pH, exchangeable Ca, ESP, and available P. Soil bulk density was determined only from the field experiment. The flocculating, dispersive, and net dispersive charges, as well as the cation ratio of soil structural stability, were calculated. In general, selected soil properties were analyzed to assess the effectiveness of PG application on saline–sodic soils across three wheat varieties in both pot and field experiments.

Statistical analysis

A separate analysis of variance was conducted for the pot and field experiment data using the GLM procedure in SAS version 9.4³⁴, with each experiment analyzed according to its respective experimental design. Data normality was evaluated using the Shapiro–Wilk test³⁵. Subsequently, a mixed-effects model was applied, with phosphogypsum (PG), wheat variety, and their interaction PG and wheat variety treated as fixed effects, while replications were considered random effect. Treatment means were compared using Tukey’s honest significant difference (HSD) test at p < 0.05, following the procedures described by Gomez and Gomez³⁵. Principal component analysis (PCA) was performed using the PRINCOMP procedure in SAS. Pearson correlation coefficients were calculated using SAS version 9.4.

Field guidelines and permission

The field experiment was conducted on a Metehara sugar estate farmland after obtaining permission from the sugar estate farm manager and the local agricultural office. The farmer was informed of the study’s objectives, and consent was obtained before establishing the experimental plots.

Results

Soil particle size distribution and bulk density

ANOVA results indicated that soil particle size distribution was not significantly affected by PG rate or wheat variety (Supplementary Table 5), and the soil textural class remained loam, consistent with the initial condition (Table 1). However, the clay fraction showed a small but statistically significant response to PG application in both the field (p < 0.001) and pot (p < 0.05) experiments. In the pot experiment, the clay content increased slightly from 20.0% in the control to 20.9%, 21.2%, 21.4%, and 21.5% at GR levels of 50%, 100%, 150%, and 200%, respectively (Table 1). A similar trend was observed in the field experiment, where clay content rose from 20.7% to 21.0%, 21.2%, 21.5%, and 21.4% at the corresponding application rates (Table 1). Notably, PG applied at 100% GR did not differ significantly from pure gypsum (Table 1).

In the field experiment, soil bulk density was significantly reduced by PG application (p < 0.0001), as shown in Table 1 and Supplementary Table 5. Compared to the control, bulk density decreased by 2.4%, 4.8%, 6.4%, and 8.0% at 50%, 100%, 150%, and 200% GR, respectively (Table 1). Furthermore, PG applied at 100% GR resulted in a 1.6% greater reduction in bulk density compared to pure gypsum applied at the same rate (Table 1). In contrast, neither wheat variety nor its interaction with PG had a significant effect on soil bulk density (Supplementary Table 5).

Selected soil chemical properties

The data presented in Table 2 and Supplementary Table 5 demonstrate that PG application significantly reduced soil pH (p < 0.001) and ESP (p < 0.0001) across all wheat varieties under both pot and field conditions. In the pot experiment, PG applied at 50%, 100%, 150%, and 200% GR decreased soil pH by 0.7%, 2.1%, 2.8%, and 3.1%, respectively, compared to the initial soil (Table 2). Correspondingly, ESP declined from 34.34% to 26.0%, 14.0%, 12.1%, and 7.9% at the respective application rates. A similar trend in the reduction of soil pH and ESP was observed under field conditions. Notably, PG applied at 100% GR produced significantly greater reductions in soil pH (by 0.6% and 0.7%) and ESP (by 6.7% and 8.2%) than pure gypsum at the same rate in the pot and field experiments, respectively (Table 2).

Table 2.

Changes in electrical conductivity, pH, exchangeable Ca, exchangeable sodium percentage, and available P as influenced by PG application on saline-sodic soils in the pot and field experiments.

Treatments	pH		ECe (dS m-1)		Ex. Ca (Cmolc·kg−1)		ESP (%)		Av. P (mg kg-1)
	Pot†	Field¥	Pot†	Field¥	Pot†	Field¥	Pot†	Field¥	Pot†	Field¥
Phosphogypsum
Initial value	8.16		14.20		10.79		34.34		6.95
PG/G 0% GR	8.15a	8.17a	4.49d	6.42c	11.4e	12.9e	32.7a	34.7a	7.5e	7.8e
PG 50% GR	8.10b	8.13b	4.67dc	6.69bc	14.4d	15.1d	26.0b	28.1b	10.7d	9.7d
PG 100% GR	7.99d	8.06c	4.82bc	6.76b	18.0c	18.9c	14.0d	13.9d	14.3c	14.3c
PG 150% GR	7.93e	8.00d	5.01ba	6.97ba	24.4b	25.6b	12.1e	11.0e	15.9b	15.7b
PG 200% GR	7.91e	7.96e	5.19a	7.19a	27.8a	29.7a	7.9f.	9.0f.	17.3a	17.2a
G 100% GR	8.04c	8.11b	4.80bc	6.74bc	16.8dc	18.9c	15.0c	15.5c	8.2e	8.1e
p value	0.0008	0.0004	0.0001	0.0007	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001
Wheat varieties
ETBW-5879	8.00b	8.05b	4.67b	6.64b	18.2b	19.7b	18.0a	18.6a	12.1a	12.0a
Gambo	8.04a	8.07ba	4.91a	6.85ba	19.0ba	20.4a	18.0a	18.6a	12.3a	12.2a
Dande’a	8.02ba	8.09a	4.92a	6.90a	19.1a	20.5a	17.8a	18.9a	12.4a	12.3a
p value	0.047	0.035	0.022	0.037	0.041	0.019	0.84	0.38	0.55	0.32

PG = phosphogypsum; G = pure gypsum; Means followed by the same letters within a column do not differ statistically at P < 0.05; † = the pot experiment conducted for two seasons; ¥ = the field experiment conducted for one season.

Wheat variety had a statistically significant effect on soil pH (p < 0.05) in both pot and field experiments, whereas ESP was not significantly influenced by variety (Table 2; Supplementary Table 5). Notably, the salt-tolerant variety ETBW-5879 resulted in a greater reduction in soil pH of saline–sodic soils compared with the salt-sensitive varieties Gambo and Danda’e (Table 2). In contrast, the interaction between PG application and wheat variety was not statistically significant for either soil pH or ESP (Supplementary Tables 5 and 7), indicating that their combined effects did not exert a synergistic influence on these soil chemical properties.

The application of increasing rates of PG significantly affected soil ECe in both pot (p < 0.0001) and field (p < 0.001) experiments relative to the initial soils (Table 2; Supplementary Table 5). Following wheat cultivation, PG application resulted in a slight increase in soil ECe compared with the control, with increments of 0.18, 0.33, 0.52, and 0.70 dS m⁻¹ in the pot experiment at application rates of 50%, 100%, 150%, and 200% GR, respectively (Table 2). A similar pattern was observed under field conditions, where ECe increased by 0.27, 0.34, 0.55, and 0.77 dS m⁻¹ at the corresponding PG rates. Notably, PG applied at 100% GR did not differ significantly from pure gypsum at the same rate (Table 2).

Application of increasing rates of PG significantly (p < 0.0001) increased exchangeable Ca following wheat cultivation under both pot and field conditions compared with the control (untreated soil), in shown in the Table 2 and Supplementary Table 5. Soil samples from both experiments showed that PG application increased exchangeable Ca relative to the control by 2.2–3.0, 6.0–6.5, 12.7–13.0, and 16.4–16.8 cmolc kg⁻¹ at 50, 100, 150, and 200% GR, respectively (Table 2). However, soils treated with pure gypsum at 100% GR did not show a significant difference in exchangeable Ca compared with those treated with PG at the same rate, indicating that both Ca sources are equally effective when applied at equivalent GR levels (Table 2).

Exchangeable Ca in the saline–sodic soil was significantly (p < 0.05) influenced by the cultivation of different wheat varieties under both pot and field conditions (Table 2; Supplementary Table 5). The salt-tolerant variety ETBW-5879 showed a greater decrease in exchangeable Ca compared with the salt-sensitive varieties Gambo and Danda’e (Table 2), suggesting higher Ca uptake and/or enhanced mobilization from the soil. However, the interaction between PG application and wheat variety had no significant effect on exchangeable Ca (Supplementary Tables 5 and 7), indicating that varietal differences did not modify the response of this soil property to PG application.

As shown in Table 2 and Supplementary Table 5, wheat cultivation under both pot and field conditions significantly increased soil-available P (p < 0.0001) with increasing PG application rates. Application of PG at 50%, 100%, 150%, and 200% GR improved post-harvest soil fertility under both pot and field conditions compared with the untreated control, increasing available P by 42.9% (3.2 mg kg⁻¹) and 24.6% (1.2 mg kg⁻¹); 91.8% (6.8 mg kg⁻¹) and 94.4% (6.6 mg kg⁻¹); 113.2% (8.4 mg kg⁻¹) and 102.1% (7.9 mg kg⁻¹); and 132.4% (9.9 mg kg⁻¹) and 121.6% (9.5 mg kg⁻¹), respectively (Table 2). Across both experimental setups, PG application increased soil-available P by 75.8% compared with pure gypsum, likely due to its residual P content, which is absent in conventional gypsum (Table 2). However, wheat variety and its interaction with PG application did not significantly influence soil P levels (Table 2; Supplementary Table 5). This indicates that their combined effects did not produce a synergistic response in soil phosphorus content.

Net dispersive charge and cation ratio of soil structural stability

Application of PG significantly (p < 0.0001) increased the flocculating charge (FC) while decreasing the dispersive charge (DC), net dispersive charge (NDC), and the cation ratio of soil structural stability (CROSS) in saline–sodic soils (Table 3; Supplementary Table 5). In both pot and field experiments, PG applied at 200% GR reduced clay DC by 56.6% (223.0 Cmolc kg⁻¹) and 55.9% (233.6 Cmolc kg⁻¹), respectively, compared with the control (Table 3). Conversely, the same treatment increased clay FC by 101.7% (327.1 Cmolc kg⁻¹) in the pot experiment and by 98.9% (278.8 Cmolc kg⁻¹) in the field experiment (Table 3). Moreover, PG at 200% GR substantially reduced CROSS by 67.4% (11.9 meq L⁻¹) in the pot experiment and by 73.7% (14.0 meq L⁻¹) in the field experiment (Table 3).

Table 3.

Changes in dispersive charge (DC), flocculating charge (FC), net dispersive charge (NDC), and cation ratio of soil structural stability (CROSS) as influenced by PG application on saline-sodic soils in the pot and field experiments.

Treatments	DC (Cmolc kg-1)		FC1:1 (Cmolc kg-1)		NDC		CROSS1:1 (meq l-1)
	Pot†	Field¥	Pot†	Field¥	Pot†	Field¥	Pot†	Field¥
Phosphogypsum
Initial value	417.67		308.07		109.60		18.24
PG/G 0% GR	394.0a	417.9a	321.5c	281.7d	81.5a	136.1a	17.6a	19.0a
PG 50% GR	343.1b	363.6b	348.6c	303.3d	-5.5b	60.2b	14.1b	16.1b
PG 100% GR	223.5dc	227.9c	531.4b	450.4c	-307.9dc	-222.5c	10.0dc	11.2c
PG 150% GR	209.6d	200.9d	560.4b	498.2b	-350.7d	-297.3d	8.8d	7.7e
PG 200% GR	171.0e	184.3e	648.6a	560.5a	-477.6e	-376.3e	5.8e	5.0f.
G 100% GR	232.3c	238.4c	518.9b	444.3c	-286.6c	-205.9c	10.5c	9.8d
p value	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001

PG = phosphogypsum; G = pure gypsum; Means followed by the same letters within a column do not differ statistically at P < 0.05; † = the pot experiment conducted for two seasons; ¥ = the field experiment conducted for one season.

Increasing PG application rates consistently reduced NDC, thereby enhancing clay flocculation and promoting the formation of more stable soil aggregates (Table 3). The 150% GR treatment produced effects on DC, FC, NDC, and CROSS that were similar but slightly less pronounced than those observed with the 200% GR treatment across both experimental systems (Table 3). In contrast, the application of pure gypsum at 100% GR did not result in significant differences in DC, FC, NDC, or CROSS compared to PG applied at the same GR level (Table 3). Wheat variety, as well as its interaction with PG application, had no significant effect on DC, FC, NDC, or CROSS (Supplementary Tables 5 and 6).

Growth, physiology, and yield traits

As shown in Supplementary Table 8, both PG application rate and wheat variety significantly influenced plant growth, physiological performance, and grain yield across both pot and field experiments. Significant effects were observed on plant height, leaf area, relative water content, chlorophyll SPAD values, photosynthetic rate, transpiration rate, dry biomass, and grain yield under both experimental conditions. However, the interaction between PG application and wheat variety was significant only for plant height, leaf area, aboveground biomass, and grain yield, whereas physiological parameters did not show significant interaction effects.

Plant height and leaf area

Under both experimental conditions, PG application rate, wheat variety, and their interaction significantly affected plant height and leaf area (Fig. 1; Supplementary Table 8). Application rates of 150–200% GR consistently enhanced these growth parameters (Supplementary Table 9). The salt-tolerant variety ETBW-5879 consistently outperformed the salt-sensitive varieties Gambo and Dande’a under both pot and field conditions. A stronger interaction effect was observed at higher PG levels, particularly at 200% GR, where ETBW-5879 recorded maximum plant height and leaf area values of 72.0 cm and 24.7 cm², respectively, in 2022/23, and 73.0 cm and 23.9 cm², respectively, in 2023/24 (Figs. 1A, B, D, and E). Under field conditions, the corresponding values further increased to 83.0 cm and 20.6 cm² in 2023/24 (Figs. 1C and F). In untreated soils, salt-tolerant wheat varieties such as ETBW-5879 showed increases in plant height of 18–35% and in leaf area of 28–47% compared with salt-sensitive varieties such as Gambo and Dande’a (Figs. 1A–F). Although slightly lower than those observed at 200% GR, the application of 150% GR with ETBW-5879 also resulted in substantial improvements in both parameters across experimental systems (Figs. 1A–F).

Fig. 1.

Effect of PG application on plant height (A–C) and leaf area (D–F) of the tested varieties in saline-sodic soils. Means followed by the same letters do not significantly differ at P < 0.05; PG = phosphogypsum; G = pure gypsum.

Relative water and chlorophyll contents

The relative water content (RWC) and chlorophyll SPAD values of wheat leaves were significantly affected by PG application rate and wheat variety under both pot and field conditions (Table 4; Supplementary Table 8). Across both experimental conditions, PG application markedly improved RWC and SPAD values compared with the untreated control, indicating enhanced physiological performance under saline–sodic stress conditions (Table 4). Specifically, PG applied at 50%, 100%, 150%, and 200% GR increased RWC by 13%, 19%, 28%, and 27%, respectively, relative to the control (Table 4). Similarly, SPAD values, which reflect leaf chlorophyll concentration, increased by 22%, 34%, 38%, and 36% at 50%, 100%, 150%, and 200% GR, respectively, compared with the control (Table 4). Among treatments, PG application at 100% GR consistently outperformed equivalent rates of pure gypsum, increasing RWC by 6.9% and SPAD values by 6.8% (Table 4).

Table 4.

Mean values of relative water content and chlorophyll content in SPAD values of wheat as influenced by PG and varieties under pot and field experiments.

Treatments	RWC (%)			Chlorophyll SPAD values
	Pot		Field 2023/24	Pot		Field 2023/24
	2022/23	2023/24		2022/23	2023/24
Phosphogypsum
PG/G 0% GR	62.1d	64.1d	63.0d	32.4c	34.8d	35.0c
PG 50% GR	74.2cb	76.1cb	66.3c	40.5b	42.7c	41.8b
PG 100% GR	77.9b	78.3b	74.3b	46.1a	45.6ba	44.9a
PG 150% GR	80.8a	81.7a	79.8a	47.1a	47.7a	45.9a
PG 200% GR	80.1a	81.7a	79.1a	46.4a	47.1a	45.5a
G 100% GR	70.9c	74.1c	69.4c	40.9b	43.3bc	42.5b
p value	0.0001	0.0005	0.0009	0.0001	0.0001	0.0001
Wheat varieties
ETBW-5879	83.2a	84.4a	79.9a	46.9a	50.2a	48.8a
Gambo	73.0b	76.0b	71.3b	42.6ba	43.5b	42.7b
Dande’a	66.8c	67.6c	64.7c	37.1b	36.8c	36.4c
p value	0.0001	0.001	0.0006	0.003	0.0001	0.0001

Means followed by the same letters do not significantly differ statistically at P < 0.05; PG = phosphogypsum; G = pure gypsum.

Growing wheat varieties significantly influenced RWC and chlorophyll SPAD values under both pot and field conditions (Supplementary Tables 8 and 10). In particular, the salt-tolerant variety ETBW-5879 consistently exhibited higher RWC and SPAD values than the salt-sensitive variety Dande’a (Table 4). Specifically, ETBW-5879 showed an 11–14% increase in RWC and a 9–13% increase in SPAD compared with Gambo, and a 23–25% increase in RWC and a 21–27% increase in SPAD compared with Dande’a under both experimental conditions (Table 4). However, the two-way interaction between PG application and wheat variety was not significant for RWC and SPAD (Supplementary Tables 8 and 10), indicating that varietal differences and PG effects acted independently without a synergistic influence on these physiological traits.

Photosynthesis and transpiration rates

The average rates of photosynthesis and transpiration were significantly affected (p < 0.001) by both wheat variety and PG application level, as shown in Supplementary Tables 8 and 10. Increasing the PG application rate from 50 to 200% GR resulted in a marked enhancement of both physiological processes across the experimental conditions (Table 5). Relative to the control, PG application at 50%, 100%, 150%, and 200% GR increased the average photosynthesis rate by 16.6% (1.6 μmol CO₂ m⁻² s⁻¹), 39.9% (3.9 μmol CO₂ m⁻² s⁻¹), 67.5% (6.6 μmol CO₂ m⁻² s⁻¹), and 63.3% (6.2 μmol CO₂ m⁻² s⁻¹), respectively. Similarly, transpiration rates increased by 23.1% (0.6 mmol H₂O m⁻² s⁻¹), 57.9% (1.5 mmol H₂O m⁻² s⁻¹), 107.7% (2.9 mmol H₂O m⁻² s⁻¹), and 102.6% (2.8 mmol H₂O m⁻² s⁻¹), respectively (Table 5). However, pure gypsum applied at 100% GR did not differ significantly from PG at the same rate for either photosynthesis or transpiration, indicating comparable physiological responses to both amendments at this application level.

Table 5.

Mean values of photosynthesis and transpiration rates of wheat as influenced by PG and varieties under pot and field experiments.

Treatments	Photosynthesis rate (μmolCO2 m−2 s−1)			Transpiration rate (mmolH2O m−2 s−1)
	Pot		Field 2023/24	Pot		Field 2023/24
	2022/23	2023/24		2022/23	2023/24
Phosphogypsum
PG/G 0% GR	9.2c	9.9c	10.2d	2.6c	2.7d	2.8d
PG 50% GR	10.6cb	11.7cb	11.9c	3.2bc	3.4c	3.3c
PG 100% GR	12.5b	14.7ba	13.9b	4.0ba	4.5b	4.2b
PG 150% GR	14.7a	17.1a	17.4a	5.0a	5.9a	5.8a
PG 200% GR	14.6a	16.0a	17.3a	4.8a	5.5a	6.0a
G 100% GR	11.9b	12.5b	13.6b	3.7b	3.8bc	3.9cb
p value	0.0001	0.0001	0.0001	0.0001	0.0001	0.0001
Wheat varieties
ETBW-5879	13.7a	15.0a	15.7a	5.0a	5.7a	5.3a
Gambo	12.3b	13.3b	14.3b	3.8b	4.4b	4.3b
Dande’a	10.9c	11.4c	12.1c	2.9c	2.8c	3.4c
p value	0.0008	0.001	0.0001	0.0001	0.0001	0.0001

Means followed by the same letters do not significantly differ at P < 0.05; PG = phosphogypsum; G = pure gypsum.

The results further showed that salt-tolerant wheat varieties, particularly ETBW-5879, exhibited significantly higher photosynthesis and transpiration rates than the salt-sensitive varieties Gambo and Dande’a under both pot and field conditions (Table 5). Specifically, ETBW-5879 recorded 8.9–11.2% higher photosynthesis and 18.4–24.5% higher transpiration rates than Gambo, and 20.4–23.6% higher photosynthesis and 35.5–51.6% higher transpiration rates than Dande’a (Table 5). These findings indicate that ETBW-5879 has a greater capacity to sustain gas exchange under saline–sodic stress conditions. Nevertheless, the interaction between PG application and wheat variety was not statistically significant for either photosynthesis or transpiration (Supplementary Tables 8 and 10), suggesting that their effects on these physiological traits were independent.

Dry biomass and grain yields

As shown in Supplementary Table 8, PG rate, wheat variety, and their interaction (PG × variety) significantly influenced wheat dry biomass and grain yield under both pot and field conditions. Application of PG at 50%, 100%, 150%, and 200% GR increased dry biomass yield by 15%, 38%, 50%, and 53%, and grain yield by 21%, 67%, 87%, and 92%, respectively, relative to the control across all tested varieties (Supplementary Table 9). Both biomass and grain yield reached their highest values at 150% and 200% GR, indicating a clear dose-dependent response of wheat to PG application under saline–sodic conditions (Fig. 2). Moreover, PG consistently outperformed pure gypsum, resulting in significantly greater improvements in both dry biomass and grain yield (Supplementary Table 9).

Fig. 2.

Effect of PG application on dry biomass (A–C) and grain yield (D–F) of the tested varieties in saline-sodic soils. Means followed by the same letters do not significantly differ statistically at P < 0.05; PG = phosphogypsum; G = pure gypsum.

Varietal differences significantly influenced dry biomass and grain yield under both pot and field conditions (Supplementary Table 8). In the field experiment, untreated saline–sodic soil markedly constrained wheat productivity, with dry biomass and grain yield recorded at 14.3 t ha⁻¹ and 4.2 t ha⁻¹ for ETBW-5879, 13.2 t ha⁻¹ and 3.3 t ha⁻¹ for Gambo, and 11.9 t ha⁻¹ and 2.6 t ha⁻¹ for Dande’a (Supplementary Table 9). Similar trends were observed in the pot experiments across both seasons (Supplementary Table 9). Overall, the salt-tolerant variety ETBW-5879 consistently outperformed the more sensitive varieties, Gambo and Dande’a, producing substantially higher dry biomass and grain yield under saline–sodic conditions.

Applying increasing rates of PG in combination with salt-tolerant wheat varieties significantly enhanced both dry biomass and grain yield under pot and field conditions (Figs. 2A–F). The salt-tolerant variety ETBW-5879 showed markedly greater yield advantages over the more sensitive varieties under untreated saline–sodic conditions. Specifically, the grain yield of ETBW-5879 increased by 107% and 224% at 200% GR, 103% and 217% at 150% GR, and 89% and 194% at 100% GR, compared with Gambo and Dande’a grown without PG application (Figs. 2D–F). Notably, even at 100% GR, ETBW-5879 produced grain yields that were 6% higher than Gambo and 21% higher than Dande’a under the highest PG rate (200% GR), highlighting the strong contribution of varietal tolerance relative to amendment rate (Figs. 2D–F). A similar pattern was observed for dry biomass, where the combined effect of PG application and the salt-tolerant variety consistently resulted in superior biomass production across both experimental setups (Figs. 2A–C).

Principal component analysis and Pearson correlation

Principal component analysis (PCA) was conducted to evaluate the combined effects of PG application and salt-tolerant wheat varieties on saline–sodic soil properties and wheat productivity under pot and field conditions (Supplementary Table 12; Supplementary Figs. 4A and 4B). In the pot experiment, the first two principal components (PC1 and PC2) accounted for 42.7% and 32.1% of the total variance, respectively, with a cumulative contribution of 74.7%. In the field experiment, PC1 and PC2 explained 45.6% and 33.6% of the variance, respectively, giving a cumulative total of 79.3%. These results indicate that a substantial proportion of the variability in the dataset is captured by the first two components in both experimental systems.

In the pot experiment, PC1 was strongly associated with ESP, FC, DC, NDC, and CROSS, reflecting the dominant soil sodicity and structural stability gradient (Supplementary Table 12). PC2 was primarily linked to plant growth and physiological traits, including plant height, grain yield, RWC, photosynthetic rate, and transpiration rate, thereby representing crop performance responses (Supplementary Table 12). A comparable pattern was observed under field conditions, where PC1 was mainly driven by ESP, NDC, and CROSS, indicating soil physicochemical constraints, whereas PC2 was associated with plant height, leaf area, biomass, and photosynthetic attributes, reflecting crop growth and productivity responses (Supplementary Table 12).

Overall, PCA results demonstrate that PG application rate and varietal differences are the main sources of variation influencing both soil properties and wheat performance (Supplementary Table 12). PC1 primarily represents the gradient of soil structural improvement under saline–sodic conditions, whereas PC2 captures the associated physiological and yield responses of wheat. These findings highlight that increasing PG application, particularly when combined with salt-tolerant wheat varieties, substantially improves soil structural quality and enhances crop productivity.

Correlation analysis indicated that both the application of PG and the use of salt-tolerant wheat varieties were consistently associated with soil physicochemical properties, thereby enhancing wheat growth and productivity under the two experimental conditions (Supplementary Table 11). Soil pH and ESP exhibited a strong positive correlation (r = 0.94***), indicating their close association in alkaline sodic soils and their combined influence on the development of sodicity. Both soil pH and ESP showed negative correlations with ECe, exchangeable Ca, and available P, while demonstrating positive correlations with NDC, CROSS, and bulk density. Furthermore, exchangeable Ca and available P showed strong negative correlations with soil pH (r =  − 0.83*** and r =  − 0.67**), ESP (r =  − 0.86*** and r =  − 0.58**), NDC (r =  − 0.88*** and r =  − 0.59**), CROSS (r =  − 0.71** and r =  − 0.67**), and bulk density (r =  − 0.84*** and r =  − 0.41*), indicating that improvements in these soil properties were closely associated with increased Ca and P availability following PG application.

Grain yield showed strong positive correlations with key growth and physiological traits, including plant height (r = 0.72***), leaf area (r = 0.52**), dry biomass (r = 0.86***), RWC (r = 0.77***), chlorophyll SPAD value (r = 0.89***), and photosynthetic rate (r = 0.70**) (Supplementary Table 11). In addition, grain yield was positively correlated with exchangeable Ca (r = 0.31*) and available P (r = 0.29*), while showing negative correlations with ESP (r =  − 0.47*) and CROSS (r =  − 0.28*) (Supplementary Table 11). In general, these relationships indicate that improvements in soil chemical properties and physiological performance enhance wheat productivity. In contrast, increased sodicity and structural instability adversely affect crop growth and yield under saline–sodic conditions.

Discussion

Soil particle size distribution, bulk density, net dispersive charge and cation ratio of soil structural stability

Findings from both pot and field experiments showed that PG application slightly increased soil clay content relative to the control (Table 1; Supplementary Table 5). This increase is primarily attributed to higher ionic strength, which promotes flocculation and suppresses clay dispersion, thereby enhancing the retention of fine particles¹⁵. The observed change reflects improved preservation of existing clay rather than the formation of new clay minerals. Under sodic conditions, Na⁺ promotes clay dispersion and the loss of fine-textured particles, whereas Ca²⁺ supplied by PG enhances flocculation, improves aggregate stability, and reduces clay mobility³⁶. In contrast, sand and silt fractions remained largely unaffected, consistent with their lower sensitivity to flocculation–dispersion processes³². In addition, PG itself may contain 20–25% clay-sized particles, which could directly contribute to the slight increase in the measured clay fraction compared with the control treatment³⁶. Calcium-induced flocculation further promotes the formation of stable aggregates, improves soil structure, and increases porosity, thereby contributing to reduced bulk density. Reported reductions in bulk density range from 3–5% in moderately sodic soils to 15–24% under higher PG application rates (15–60 t ha⁻¹) under favorable conditions^15,37. Similar responses have been documented in Solonetz soils in Russia³⁸ and saline–sodic soils in Morocco¹⁵ and China³⁷.

Application of PG reduced dispersive charge, net dispersive charge, and CROSS values, while increasing flocculating charge, thereby enhancing flocculation and aggregate stability in Na-dominated saline–sodic soils (Table 3). These changes are mainly attributed to Ca²⁺-induced displacement of exchangeable Na and the resulting increase in electrolyte concentration, which compresses the diffuse double layer and suppresses clay dispersion^15,39. The shift toward flocculating charge dominance (i.e., NDC < 0) indicates improved cation balance and greater structural stability of the soil³². In this context, CROSS proved to be a particularly sensitive indicator of PG effectiveness in soil amelioration. Overall, the observed structural improvements were primarily driven by PG-mediated chemical processes, with negligible contributions from varietal effects or their interaction with PG application^12,15.

Selected soil chemical properties

As shown in Table 2 and Supplementary Table 5, PG application decreased both soil pH and ESP, confirming its effectiveness in ameliorating soil alkalinity and sodicity. These results are consistent with earlier reports^12,40–42. The observed reductions are mainly attributed to Ca²⁺ released from PG, which displaces exchangeable Na⁺ from soil exchange sites and enhances the leaching of HCO₃⁻ (69–73%) and CO₃²⁻ (50–60%)^12,41. Additionally, sulfate ions released from PG combine with the displaced Na⁺ in the soil solution to form soluble sodium sulfate, thereby facilitating its leaching^41,42. In addition, the presence of sulfuric acid (0.1–1.0% by weight) and phosphoric acid (0.2–1.5% by weight) in PG contributes to soil acidification, further lowering pH and ESP^40,41. The decline in pH may also reflect increased H⁺ activity following Ca²⁺-induced displacement of Na⁺ from soil colloids⁴². Collectively, these mechanisms represent a salting-out effect that enhances Na⁺ removal and improves soil chemical conditions. Despite these reductions, pH and ESP remained within ranges suitable for crop growth. Compared to pure gypsum, PG resulted in a greater reduction in pH and ESP (Table 2), likely due to its approximately 2.2-fold higher dissolution rate, which enables more rapid release of Ca²⁺ and SO₄²⁻ and accelerates Na exchange and leaching processes⁴³.

Although PG application reduced soil ECe relative to the initial condition, values remained slightly higher than those of the control (Table 2), consistent with previous reports^14,15,40. This transient increase in ECe is primarily driven by PG dissolution, which releases Ca²⁺ and SO₄²⁻ into the soil solution⁴². The released Ca²⁺ displaces exchangeable Na, K, and Mg from soil exchange sites³⁸, thereby increasing the concentration of these cations in the soil solution and causing a temporary rise in electrical conductivity. The extent of this response is governed by PG application rate, soil texture, and leaching intensity⁴¹. Owing to its strong affinity for soil colloids, Ca²⁺ preferentially replaces exchangeable Na, thereby promoting soil structural improvement¹². Concurrently, SO₄²⁻ facilitates Na⁺ removal from the soil profile through the formation of soluble sodium sulfate complexes⁴². Collectively, these processes enhance exchangeable Ca²⁺ saturation on soil exchange surfaces and underpin the progressive reclamation of saline–sodic soils^15,40.

Present results further showed that PG application significantly increased soil available P (Table 2), consistent with previous studies^12,44. This response can be attributed to the additional supply of P from PG, as evidenced by the increased concentration of available P in the soil solution^14,44. Based on a report by Kalinitchenko et al.³⁸, PG is a byproduct of phosphate fertilizer production, contains appreciable amounts of residual phosphorus, typically ranging from 25 to 75 g kg⁻¹. Its use as a P amendment also provides environmental benefits by reducing reliance on phosphate rock mining, thereby mitigating associated environmental degradation, resource depletion, and waste accumulation⁴⁴. Across both experimental setups, PG application increased soil available P by 75.8% relative to pure gypsum (Table 2), likely due to its residual P content, which is absent in conventional gypsum. In sodic soils, PG application also increased the levels of soluble P, calcium-bound P, and iron-bound P fractions compared with both the initial soil condition and soils treated with pure gypsum¹⁴. This improvement is likely associated with the dissolution of PG, which releases Ca²⁺ and sulfate ions that enhance P availability and promote the transformation of P into more stable and plant-accessible forms⁴².

Wheat varieties significantly affected soil pH, ECe, and exchangeable Ca in both pot and field experiments (Table 2; Supplementary Table 5). Salt-tolerant wheat varieties improve soil health through multiple mechanisms, including enhanced ion selectivity at the root membrane, ion accumulation in the cytoplasm, and increased rhizosphere ion exchange, resulting in greater cation than anion uptake⁴⁵. To maintain electrochemical balance, roots release H⁺ ions, thereby lowering soil pH⁴⁶. In addition, higher root exudation, particularly organic acids and CO₂, further contributes to pH reduction⁴⁷.

Salt-tolerant varieties also exhibited lower ECe (Table 2), which was attributed to enhanced Na⁺ (30–70%) and Cl⁻ (20–60%) uptake from the soil solution, together with efficient ion compartmentalization and exclusion mechanisms, improved K⁺ selectivity, and effective maintenance of ion homeostasis⁴⁶. These traits collectively reduce salt accumulation in the soil solution, leading to a decline in ECe over time⁴⁷. Similarly, the greater reduction in exchangeable Ca observed under the salt-tolerant variety ETBW-5879 compared with the salt-sensitive varieties Gambo and Danda’e (Table 2) is likely due to its more extensive root system and stronger rhizosphere activity, which enhance Ca²⁺ mobilization and uptake⁴⁶.

However, the interaction between PG application and wheat variety had no significant effect on soil pH, ECe, exchangeable Ca, ESP, or available P (Supplementary Tables 5 and 7), indicating the absence of synergistic effects on these properties. This suggests that PG-driven chemical amelioration via Ca²⁺-mediated displacement of exchangeable Na and varietal influences through root-mediated ion uptake and rhizosphere modification operated independently, with changes in soil properties primarily reflecting their main effects rather than any interactive mechanism.

Growth, physiology, and yield traits

Application of PG significantly alleviated salinity–sodicity stress across wheat varieties, significantly enhancing growth, physiological performance, and yield (Supplementary Table 8). Increasing PG rates consistently improved plant height and leaf area (Supplementary Table 9), aligning with responses reported in in wheat^12,16 and other crops such as pepper⁴¹ and sunflower⁴². These improvements are primarily driven by coupled chemical and physical soil amelioration: PG supplies Ca²⁺, which displaces exchangeable Na⁺, increases phosphate and sulfate availability, and induces slight acidification that enhances nutrient accessibility³⁸. Concurrently, reductions in sodium-induced dispersion improve aggregation, porosity, permeability, and aeration, thereby promoting root development and more efficient water and nutrient uptake. These enhanced cellular activities, including division and elongation, further support canopy expansion and biomass accumulation⁴⁸. Comparable responses have been reported in rice, where PG application increased leaf area by 78–152%⁴⁹. Notably, PG at 100% gypsum requirement outperformed pure gypsum (Supplementary Table 8), likely due to its higher solubility and additional nutrient contributions, resulting in more effective Na⁺ displacement, improved soil structural stability, and ultimately greater crop productivity⁵⁰.

Relative water content and chlorophyll SPAD values of wheat leaves were significantly affected by PG application in both pot and field experiments (Table 4; Supplementary Table 8). Phosphogypsum reduced sodium hazards and improved soil physical conditions, enhancing root activity, soil aeration, and nutrient uptake. These changes helped maintain cell turgor, reduce membrane damage, and support photosynthesis, resulting in increased RWC^50,51. For example, PG increased quinoa RWC by up to 80% under normal irrigation and 69% under deficit conditions⁵⁰. Applying PG also significantly increased SPAD values by improving nutrient availability (N, P, K, Ca, and Mg), essential for chlorophyll synthesis⁵¹. Similar increases (35–40%) in chlorophyll content have been reported in soybean at 100% GR⁵². Moreover, PG at 100% GR produced higher RWC and SPAD values than pure gypsum (Table 4), likely due to its greater solubility and nutrient supply, which enhance Ca²⁺-mediated Na⁺ displacement and improve soil structure. Consequently, PG-treated wheat maintained better water status and chlorophyll content, leading to improved physiological performance^14,50.

The use of PG improved wheat photosynthesis and transpiration by improving soil structure, reducing sodicity, and increasing the availability of essential nutrients (P, SO₄²⁻, Ca, N, Mg, and K) that support chlorophyll synthesis and stomatal regulation^52,53. Improved soil conditions maintained higher leaf water status and optimised stomatal function, resulting in increased carbon assimilation and regulated transpiration^41,52. In addition, PG significantly increased dry biomass and grain yield in both pot and field experiments (Supplementary Table 9). Adding Ca²⁺ from PG improved soil structure, infiltration, and hydraulic conductivity by displacing Na from exchange sites, reducing ESP, and promoting clay flocculation⁴⁹. These enhancements improved root growth, nutrient uptake, and reduced ionic stress, resulting in increased wheat productivity^52,53. As a result, increasing PG application rates resulted in higher biomass accumulation and grain yield⁵⁰, with PG outperforming pure gypsum due to its increased nutrient supply and better effects on soil physicochemical properties and rhizosphere conditions.

Varietal differences in plant height and leaf area reflected genetic variation in salt tolerance⁵⁴. The salt-tolerant variety ETBW-5879 consistently exhibited greater plant height and leaf area than the salt-sensitive varieties Gambo and Dande’a under both experimental conditions (Supplementary Table 9), consistent with previous report¹¹. Salt-tolerant genotypes maintained higher relative water content under salinity stress through improved osmotic adjustment, reduced water loss, and enhanced membrane stability, thereby supporting plant growth and biomass accumulation⁵⁵. They also exhibited higher SPAD values, likely due to lower reactive oxygen species production, reduced membrane damage, and decreased chlorophyll degradation^47,54. Moreover, salt-tolerant varieties sustained higher photosynthetic and transpiration rates under saline–sodic soil conditions through efficient stomatal regulation, improved ion homeostasis, and stronger antioxidative defence mechanisms⁵⁶, consistent with earlier studies reporting 25–30% higher rates than those of sensitive genotypes⁵⁷. Consequently, ETBW-5879 produced greater dry biomass and grain yield owing to enhanced photosynthetic efficiency and carbon assimilation^54,55. The combined application of PG and salt-tolerant varieties further amplified these positive effects, significantly increasing grain yield and dry biomass production under both pot and field conditions (Figs. 2A–F).

The combined application of PG and salt-tolerant wheat varieties synergistically enhanced plant height and leaf area by improving soil structure, increasing nutrient availability, and strengthening tolerance to salt-stress⁴⁸. Applying PG reduced sodium hazards, promoted root development, and improved soil physical conditions, thereby facilitating the uptake of Ca, P, and sulfate nutrients essential for leaf expansion and vegetative growth^48,51. Salt-tolerant varieties exhibited greater leaf area due to superior photosynthetic efficiency and enhanced CO₂ assimilation capacity⁵⁴,which contributed to improved canopy development and biomass accumulation. However, the interaction between PG and variety was not significant for photosynthesis and transpiration under either pot or field conditions, indicating that all genotypes responded similarly in terms of gas exchange. This suggests that PG exerted a uniform positive effect on physiological performance regardless of varietal differences. Overall, the combined use of PG and salt-tolerant wheat varieties improved soil conditions, nutrient uptake, and plant growth, consistent with previous findings reported for crops cultivated under salt-affected soils^48,54,55.

Principal component analysis and Pearson correlation

Principal component analysis was performed to assess the combined effects of PG application and salt-tolerant wheat varieties on saline–sodic soil properties and wheat productivity under both pot and field conditions (Supplementary Table 12; Supplementary Figs. 4A and B). The application of PG was associated with a reduction in ESP, which in turn likely improved soil physical properties by increasing FC and reducing DC, NDC, and CROSS. These effects are consistent with previous findings indicating that Ca²⁺ supplied by phosphogypsum replaces Na⁺ on exchange sites, thereby enhancing soil aggregation and structural stability^12,15,42. In the other hand, PCA showed that salt-tolerant wheat varieties were positively associated with improved plant height, grain yield, relative water content, and photosynthetic performance, supporting earlier reports on enhanced physiological adaptation under salinity and sodicity stress^54–56. Overall, the PCA identified the key soil and crop variables influenced by PG and varietal selection, providing valuable insights for optimizing soil management strategies and improving crop productivity in saline–sodic environments^48,54,55.

The correlation matrix from pot and field experiments shows how PG rates and salt-tolerant wheat varieties affect saline–sodic soil properties and wheat productivity (Supplementary Table 11). The observed relationships indicate that soils with high ESP and pH are generally associated with unstable soil structure, higher bulk density, and reduced availability of essential nutrients². The strong positive association between exchangeable Ca and available P further demonstrates that PG amendments improve nutrient availability by enhancing soil chemical balance and structural stability¹². Overall, the correlation matrix confirms that PG application plays a crucial role in improving the physicochemical quality of saline–sodic soils, thereby creating more favorable conditions for plant growth^50,58.

Grain yield exhibited strong positive correlations with key growth and physiological traits (Supplementary Table 11). This indicates that improvements in soil conditions due to PG application, together with the use of salt-tolerant wheat varieties, substantially enhance wheat growth, physiological performance, and ultimately grain productivity under saline–sodic conditions^38,49,50. Grain yield showed positive correlations with exchangeable Ca and available P (Supplementary Table 11), indicating that improved soil fertility under PG application enhances nutrient availability and supports higher wheat productivity under saline–sodic conditions. Increased exchangeable Ca likely improves soil aggregation and nutrient retention, while available P supports root growth and key metabolic processes essential for yield formation⁴². In contrast, grain yield was negatively correlated with ESP and CROSS (Supplementary Table 11), reflecting the detrimental effects of sodicity and ionic imbalance on soil structure and plant performance^51,52. High ESP promotes clay dispersion and reduces soil porosity, while elevated CROSS indicates reduced structural stability, ultimately restricting water and nutrient uptake and lowering yield^12,15.

Conclusions

This study confirms that soil salinity and sodicity severely constrain wheat productivity in irrigated lowland areas of Ethiopia by degrading soil chemical balance, limiting nutrient availability, and reducing plant physiological performance. The combined application of PG and salt-tolerant wheat varieties effectively mitigates these constraints, leading to substantial improvements in soil quality and crop yield. Phosphogypsum application, particularly at 150–200% GR, significantly reduced soil pH and ESP, improving soil structure and fertility. These improvements enhanced key physiological traits such as relative water content, chlorophyll content, and photosynthetic activity, resulting in higher biomass and grain yield. Among the tested varieties, ETBW-5879 consistently showed superior tolerance and productivity under saline–sodic conditions, outperforming Gambo and Dande’a across all PG levels. It also maintained high yield even at lower PG rates, indicating strong adaptation and efficient resource use. Overall, integrating PG at 150% GR with the salt-tolerant variety ETBW-5879 is the most effective strategy for improving wheat yield and soil health in saline–sodic soils at the Metehara Sugar Estate. However, further multi-location and multi-year field trials validation is required to confirm the wider applicability of these findings.

Author contributions

Ashenafi Worku undertook this study as part of a PhD program at Hawassa University, supervised by Sheleme Beyene and Kibebew Kibret . The first draft of the manuscript was written by Ashenafi Worku , and the authors were involved in the revision and editing of the draft manuscript and approved the final manuscript.

Funding

This study was conducted in Ethiopia as part of a PhD dissertation entitled “Evaluation of Phosphogypsum as a Reclaiming Material for Salt-Affected Soils.” The authors gratefully acknowledge the financial support provided by OCP Ethiopia through Hawassa University for the implementation of this research. Additional financial support for the purchase of chemicals required for soil properties analysis was provided by the Ethiopian Institute of Agricultural Research (EIAR).

Data availability

The datasets analyzed in the current study are available from the corresponding author upon reasonable request.

Declarations

Competing interests

The authors declare no competing interests.