Introduction

Land degradation due to increasing salinity or sodicity is a worldwide problem that needs attention for food security to the increasing population1. According to the recent estimates, the total area of salt‑affected soils (SAS) of the world amounts to 1 381 million ha (Mha), or 10.7% of the total global land area befall in many parts of sub-humid, semi -arid, and arid regions of the world2. Out of the 329 million hectares geographical area of India, 175 million hectare is degraded and 6.74 million hectares is salt affected land3 with excess amounts of soluble (saline) and/or sodic salts, which adversely affect crop/vegetation growth and yield4,5. These soils are having high sodium (Na) content and high exchangeable sodium percentage (ESP) which causes dispersion of clay resulting drastic reduction in soil permeability6. Due to high dispersion and swelling of clay these soils are unsuitable for crop production7. Reclamation and proper management of SAS has been encouraged in developing as well as developed countries8,9,10,11 and numerous efforts have been made to develop these soils for productive land use systems12.

In India, salt affected soils account for about 6.74 million hectares and the future projections indicate an increase in the area of SAS to the extent of about 16 million hectares by 2050 due to inappropriate irrigation and changing climate3. Of the total SAS, sodic lands accounts for about 56.3% (3.8 million hectare) in India13,14 and as per recent assessment, in the Indo-Gangetic plains, slightly sodic areas increased by 51.8% between 1996 and 202315.

Crop growth and productivity are severely affected due to this problem, resulting in an annual loss of about 4 billion US$. This loss is expected to rise manifold with probable increase of SAS3,16. Indo-Gangetic plain zone in India covering five major provinces is highly infested with sodicity problem covering about 2.5 million hectare sodic lands (Fig. 1). Most of the sodic lands in this region are geogenic origin and seepage from canal network subsequently increase salts in the soil profile contributing in augmenting this problem. Salt affected soils in Indo-Gangetic plains is a critical issue for proper land use under the rainfed and irrigated regions15,17. During past two decades various efforts have been made to reclaim these sodic soils through chemical amendments using gypsum (CaSO4⋅2H2O) but it did not gain momentum due to its limited availability in open markets and higher cost18. Due to poor soil health conditions, farmers usually allocate share of this land parcel to grow crops, fuel, fodder, and timber for their livelihood security. Among the crops, rice and wheat are the predominant crops cultivated in reclaimed as well as partially reclaimed sodic lands. In certain conditions where irrigation water is not available in abundant quantity, the mono crop of rice in rainy season is practiced and fields are left uncultivated during winter season resulting salts coming up on the surface again. However, biological approaches, such as phytoremediation and plantation of multipurpose tree species and grasses to meet out the fuel, fodder and timber requirement have shown encouraging results in amelioration of sodic soils19,20. Changes in land uses lead to a change in the physico-chemical and biological properties of soil. These changes attributed to the anthropogenic activities in the soil surface and sub-surface layers. Soil organic carbon (SOC) is a major constituent to assess the fertility status of sodic soils. The loss of SOC due to different land uses under normal soils is well documented however, in sodic soils it is not scientifically studied because rice-wheat cropping system is predominant in sodic soils21,22,23,24. On the other hand, the impact of land use changes on soil physico-chemical and biological properties of sodic lands is hardly noticed by the land managers to initiate ameliorative measures25. Most of the studies on effect of land use changes on salt affected soils are confined to their ecological restoration however, the present study was planned with the objectives (1) to assess the influence of different land use-land covers on soil physico -chemical and biological properties and (2) find out the changes in nutrient status of sodic soil under different LULCs. For future sustainable land use planning of degraded sodic soils, a long-term study was conducted in Indo-Gangetic plains at the Research Farm of ICAR -Central Soil Salinity Research Institute (CSSRI), Regional Research Station (RRS), Lucknow, India. The hypothesis of the present study was that the different land uses have potential for amelioration of sodic soil through changes in soil properties for which systematic assessment was done.

Fig. 1
figure 1

Extent of salt affected soils in Indo-Gangetic plains.

Full size image

Materials and methods

Study site

Present study was initiated at the research farm of ICAR—Central Soil Salinity Research Institute, Regional Research Station, Lucknow, U.P., India covering an area of 24 ha sodic land and situated at an elevation of 120 m AMSL representing sodic soils of Central Indo-Gangetic plains. It is located between 26° 47’ 45”− 26° 48’ 13” N and 80° 46’ 7”− 80° 46’ 32” E (Fig. 2). Study site represents a semi-arid sub-tropical climate characterized by hot summer and cold winter. The topography of the research farm resides in the incurvation of moderately sloping plains between 119 and 125 m contours.

Fig. 2
figure 2

Location map of ICAR-CSSRI, research farm, Shivri, Lucknow, India.

Full size image

The mean annual rainfall (2000–2020) of the study site is 817 mm that is generally monsoonal occurring between June to October months. The average annual evaporation was 1580 mm. The average maximum and minimum temperatures recorded in the months of May and January were 39 °C and 7.1 °C, respectively (Fig. 3). Thus, temperature regime was hyperthermic and the moisture regime was mainly ustic.

Fig. 3
figure 3

The climatic features (1995–2015) of Shivri Research Farm.

Full size image

The soil was highly sodic (pH1:2 10.5 and ESP 89) fine loamy, mixed, hyperthermic and classified as Aquic Natrustalfs26. The site was lying abandoned and has very scanty vegetation with some grass species like Sporobolus marginatus, Cynodon dactylon, Sacharum spontanium, Aristida obscendens, Leptochloa fusca. Prosopis juliflora, Zizyphus numularia, Asparagus racemosus, Coculus pendulus (Fig. 4).

Fig. 4
figure 4

Initial Soil status of experimental site.

Full size image

Experimental design

During 1995 the sodic farmland was divided in to different land parcels. One parcel of the land (1 hectare) was reclaimed for cultivation of rice and wheat using chemical amendments like gypsum (CaSO4.2H2O) and the remaining 6 ha land was equally divided in to six land parcels for bio-amelioration under different LULCs and put under observation for 20 years. Long-term field experiment consisting of seven different land use land covers viz. (L1) (barren grassland), (L2) Rice-Wheat, (L3) Pastoral (Panicum maximum), (L4) Silvi-Agriculture (Jatropha + seasonal crops), (L5) Silvipastoral (Prosopis + Laptochloa fusca followed by Trifolium alexandrinum), (L6) Silviculture (multipurpose tree species), (L7) Horticulture (fruit species) was conducted under Randomized Block Design (Table 1) with five replicate plots.

Five plots of 10 × 10 m size were selected under each LULC for soil sampling. Five sampling points representing whole plot were selected in each plot and samples were collected from three depths (0–15, 15–30 and 30–60 cm) at each sampling point from corresponding LULCs in the month of May. These five samples mixed thoroughly to form a homogenized sample. Accordingly, 15 soil samples collected from each plot and a total 105 (7 × 3 × 5) soil samples were collected from seven LULCs.

Table 1 Description of the selected land use-land covers.
Full size table

The water samples from different irrigation sources viz. tube wells and hand pumps available at the farm were collected and analysed for quality appraisal. Three water samples were collected from tube wells, two of them from a depth of 55 m and one from 110 m. Two samples were collected from the hand pumps of 20 m depth installed at the farm. The irrigation water used for different LULCs was having low electrolyte concentration (0.56–0.72 dS m− 1) and pH (7.50–8.29) which were quite safe for irrigation in all types of crops. Among the cations, Na+ dominates over Ca++ and Mg++ followed by K+ whereas, HCO3− plus CO3− anions dominates over Ca++, and SO42− are absent. The residual alkalinity in these waters ranged from 1.3 to 1.7 me l− 1 and SAR from 2.9 to 3.3 me l− 1 (Table 2). Prolonged irrigation with such waters may cause slow alkalization of soils.

Table 2 Water quality of study site.
Full size table

Initial soil properties of the study site

Soil mineralogy of the study site

Semi quantitative estimates of sand, silt and clay minerals in sodic soils of experimental site were conducted before initiating the LULC studies. From the analysis it was found that the Quartz (0.426 nm) mineral dominates in sand followed by plagioclase feldspars (0.321 to 0.328 nm), mica (1.0 nm), chlorite (1.4 nm), orthoclase feldspars (0.312 to 0.323 nm), amphibole (0.845 nm) and mixed layer minerals (1.21 nm). Calcite (0.303 nm) was absent in the sand fraction of A horizon but was present in minute quantities in B horizon. The silt fraction mineralogy was similar to that of the sand fraction with an exception that the mica increased considerably in silt. General order of mineral preponderance was mica, plagioclase, quartz, orthoclase, chlorite, kaolinite, mixed layer, calcite and amphibole. According to semi-quantitative XRD for clay minerals, the as cendency of illite, followed by chlorite, kaolinite, mixed layer and smectite minerals were estimated26 (Table 3).

Table 3 Semiquantitative estimates of clay minerals (%) in sodic soils of Shivri experimental farm.
Full size table

Methodology to assess soil properties

To monitor the initial soil properties of the study site, eight soil profiles were dugout up to a depth of 120 cm at the farm in 1995, collected soil samples from each horizon and analysed for soil physico-chemical and biological properties. Soil bulk density (BD) was computed from intact soil samples by weighing a known volume of a 5.3 cm internal diameter core sampler27. Water holding capacity (WHC) and soil porosity (SP) were determined following the procedure recommended by keens box28. Double concentric infiltrometer cylinder with 60 cm outer and 30 cm inner diameters was used to measure Infiltration rate (IR)29,30. The observation of first water fall was recorded after 5 min and the subsequent observations were taken after 10, 20, 30, 40, 50, 60, 90 and 120 min of ponding water. The homogenized samples were dried and passed through a 2.0 mm sieve and used for analyzing soil properties. Soil pH and electrical conductivity (EC) were determined in 1:2 soil: water ratio. Digital pH and conductivity meters were used to measure soil pH and EC using 1:2 soil water suspension31. Exchangeable sodium percentage (ESP) was computed following the method32. Organic carbon (OC) content was analyzed through rapid titration method described by33. Available nitrogen was assessed by soil distillation with KMnO4 and NaOH34, available P and K by the Olsen’s sodium bicarbonate extraction35 and sodium acetate extraction methods, respectively. The Flame photometer was used to measure Na+ and K+ concentrations in soil saturation extract and the concentration of Ca2+ and Mg2+ in soil extract was estimated by the Versenate titration method36. Titration method with 0.01 N H2SO4 using phenolphthalein and methyl orange indicators was used to determine carbonate (CO3-) and bi-carbonate (HCO3- ) contents in soil extract. Soil MBC was determined using fumigation extraction method37 and MBN by CHCl3 fumigation extraction technique38. The soil MBP was measured using chloroform fumigation extraction method39 and the triphenyl tetrazolium chloride (TTC) method40 was used to determine soil dehydrogenase activity (DHA). The average initial physico-chemical and biological properties of the study site are given in Table 4.

Table 4 Initial properties (year 1995) of the study site (N = 8).
Full size table

Data analysis

The statistical analysis of experimental data was carried out using SPSS 18.0 to test the effects of LULCs on soil physio-chemical and biological properties using the two-way ANOVA. The Pearson’s multiple correlation analysis was carried out to find out the relationship between soil parameters under different LULCs. Least significance difference (LSD) test at p < 0.05 was used to analyze the comparison between the treatment means. All data are an average of five replicates ± SE of the composite soil samples. The principal component analysis (PCA) which limits the variables and extracts smaller number of independent factors (principal components) for deducing the association among observed variables was performed by varimax rotation with Kaiser Normalization. Varimax rotation was employed because orthogonal rotation minimizes the number of variables with a high loading on each component and therefore facilitates the interpretation of PCA results. The number of significant principal components were selected in view of the Kaiser criterion with eigenvalue greater than 141. The analysis will help in observing the trend of changes in soil properties under different LULCs uncovering the relationships between observations and variables and among the variables.

Results

Effect of different land use-land covers on soil physical properties

The soil physical properties including sand, silt, clay content, BD, SP, WHC, and IR were significantly (p ≤ 0.05) influenced by the different LULCs (Table 5). The sand and silt contents ranges from 60.10% to 68.50% and 19.50% to 24.40%. The soil under rice-wheat land use was considerably higher in sand content and lowest in clay content. It was followed by the silvi -agriculture, pastoral and horticulture LULCs. The reduction in clay content under different land uses ranges 2.8% to 36.68%. The highest reduction (36.68%) in this soil textural fraction was recorded under continuous use of land parcel under rice-wheat land use. This was followed by silvi-agriculture (20.78%) and horticultural (20.22%) LULCs. The silt content was higher in silvi-pastoral followed by silviculture and pastoral land uses. The silt content after 20 years of silvi -pastoral and silvicultural land uses increased to the level of 23.23%, and 8.1% over the initial values and 20.60% and 5.8% over traditional rice -wheat land use, respectively. The results indicated that the mean BD (g cm-3) of sodic soils was significantly (p ≤ 0.05) influenced by the different LULCs and over the period of study. Among the land uses studied, the BD in agroforestry LULCs was significantly (p = 0.005) different from BGL and rice –wheat land uses. The highest reduction from initial value of bulk density 1.60 g cm-3 in surface soil was recorded in silvi-pastoral land use (1.44 ± 0.002 g cm-3) followed by silviculture, horticulture, pastoral, and silvi-agricultural LULCs; whereas, the lowest (1.56 ± 0.002 g cm-3) in rice-wheat land use. From the data given in Table 9 it was revealed that there was a negative correlation with soil organic carbon and positively correlated with clay content. The soil porosity ranged from 42.05% to 56.34% across all LULCs compared to initial value of 42.4%. The highest (56.34 ± 1.13%) porosity was recorded under silvi-pastoral followed by silviculture (52.20 ± 1.36%), pastoral (45.44 ± 0.85%), and horticulture (44.22 ± 0.65%) LULC; while the lowest (43.63 ± 1.12%) under rice–wheat (Table 5). Soil infiltration rate (IR) also affected significantly (p ≤ 0.05) under different LULCs. In our study infiltration rate in different land uses ranged from 11.80 mm to 24.52 mm day-1 after 20 years compared to initial value of 4.0 mm day-1. Maximum improvement in this parameter was documented under silvi-pastoral land use followed by silviculture, horticulture and pastoral and the minimum under BGL and rice-wheat land uses. The IR of sodic soils under agroforestry LULCs increased to more than 20% over the traditional rice-wheat land use (Table 5).

Table 5 Effect of land use-land covers on soil physical properties (n = 10).
Full size table

Effect of different land use-land covers on soil chemical properties

Soil pH1:2, electrical conductivity (EC1:2), ESP, and SOC were significantly (p ≤ 0.05) affected by different LULCs. The pH1:2 of all the land parcels used for different LULCs for 20 years was still within the sodicity level (8.74–10.08) although reduced from the initial value of 10.50 in surface soil. The pH1:2 of surface soil (0–15 cm) of land parcel put under continuous cultivation of rice-wheat for 20 years was significantly lower than the land parcels covered under agroforestry land use and BGL. However, under sub-surface soils (15–30 and 30–60 cm) highest reduction in soil pH1:2 was recorded under silvipastoral land use. Soil EC1:2 of surface soil (0–15 cm) under different land uses ranges from 0.45 to 1.43 dSm-1 and it increased with the increasing soil depth. Maximum reduction in soil EC1:2 in surface (76%) soil was recorded in rice-wheat land use and in sub-surface (63%) under silvipastoral LULC followed by silviculture and pastoral land uses. The soil ESP increased with increasing soil depth but there was significant reduction in ESP irrespective of soil depth in all the land uses over the BGL. The ESP of surface soil before start of experiment was 89. Among the land uses, silvipastoral LULC showed 32%, 48%, and 54% reduction in ESP at 0–15, 15–30, and 30–60 cm soil depths, respectively, which was significantly lower than rest of the land uses (Fig. 5). The SOC content in all three soil depths (0–15, 15–30 and 30–60 cm) varied significantly under different LULCs. It decreased with increasing soil depth across the land uses. The highest improvement in SOC content at all three soil depths was recorded under silvipastoral (0.44, 0.36. 0.28%) followed by silvicultural (0.37, 0.31, 0.24%) LULCs and the lowest in BGL and rice-wheat land uses. Among the LULCs, the accumulation of SOC under agroforestry land uses was more pronounced up to 30 cm depth and beyond that, the difference between the LULCs was not statistically significant.

Fig. 5
figure 5figure 5

Effect of land use-land covers on soil pH, EC, ESP and SOC (n = 10) Bars showing SE.

Full size image

Effect of different land use-land covers on soil microbial properties

Different LULCs significantly influenced the microbial properties of sodic soil. The mean MBC, MBN and MBP varied from 42.3 to 170.4, 12.2 to 45.3, and 7.3 to 21.6 µg g− 1, respectively in 0–15 cm soil layer across the land uses (Fig. 6). The highest concentrations of MBC, MBN and MBP were recorded in silvipastoral followed by silviculture and pastoral land uses. The MBC in agroforestry-based LULCs exhibited a significant difference (p ≤ 0.05) over the BGL and rice-wheat land uses. The soil of silvipastoral land use had the highest mean value (170.4 µg g− 1) of MBC, followed by silviculture (152.5 µg g− 1) and pastoral (132.0 µg g− 1) and the lowest (42.3 µg g− 1) in BGL and rice-wheat (63.4 µg g− 1) compared to initial value of 24.34 µg g− 1. In our study, DHA varied significantly among the land uses. The highest values of DHA (6.1 µg TPF g− 1 h− 1) was observed in silvipastoral followed by pastoral (5.5 µg TPF g− 1 h− 1) and silviculture land uses in comparison to DHA of 1.38 TPF g− 1 h− 1 in the initial composite soil sample. Minimum DHA content was recorded in BGL (2.13 TPF g− 1 h− 1) and rice-wheat (4.13 µg TPF g− 1 h− 1) land uses.

Fig. 6
figure 6

Effect of different land use-land covers on soil microbial properties.

Full size image

Effect of different land use land covers on soil fertility

Available nitrogen content in soil under different LULCs ranged from 94.56 to 183.64, 92.20 to 143.20, and 65.40 to 83.40 kg ha-1 at 0–15, 15–30, and 30–50 cm soil depths, respectively. It was found to be higher in all the LULC compared to the intial value of 94 kg ha-1. The highest mean available nitrogen contents irrespective of soil depths were recorded under silvipastoral followed by silviculture LULCs, which was significantly higher over the BGL, and rice-wheat land uses (Table 6). The available N content in surface soil was found to be 95.36% higher in silvipastoral land use compared to initial soil content. Available phosphorus and available potassium in the study site were also significantly affected with different LULCs over the initial content of 25 and 388.8 kg ha-1, respectively. Available P and K in agroforestry based LULCs were significantly higher over BGL and rice-wheat land uses. The data given in Table 6 indicated that among the LULCs, the highest amounts of available N and P irrespective of soil depths were recorded in silvipastoral LULC. However, the level of available P in 0–15 (38.30 kg ha-1) and 15–30 cm (31.50 kg ha-1) soil depths in rice-wheat land use was at par with pastoral because of continuous application of phosphatic fertilizers in rice-wheat land use. It was computed that in surface soil, available P content was build up to the tune of 6, 53.2, 45.6, 66.0, 125.60, 72.8 and 29.8 per cent over the initial content in control, rice-wheat, pastoral, silvi-agriculture, silvipastoral, silviculture and horticulture land use, respectively. Similarly, available K content in surface soil was 57% higher than the initial value in silvipastoral land use followed by 38.9% in silviculture, 34.36% in silvi-agriculture, 30.97% in pastoral, 24.10% in horticulture, 11.24% in rice-wheat and 0.93% in BGL land use.

Table 6 Effect of land use-land covers on soil fertility (n = 10).
Full size table

Effect of different land use land covers on ionic content in soil

There were significant (p ≤ 0.05) variations in ionic contents under different LULCs (Table 7). The level of Ca + Mg in 0–15, 15–30 and 30–60 cm soil depths was highest under L5 (Silvipastoral) followed by L6 (silviculture) LULCs whereas, the lowest in BGL and rice-wheat land uses. The K and Na contents under different LULCs ranges from 0.11 to 0.43 and 2.15 to 8.55 meq l-1, respectively. The maximum K content in 0–15, 15–30 and 30–60 cm soil depths was recorded under treatment L5 followed by L2. The Na content at all the soil depths was lowest in L5 followed by L6 LULCs whereas, the highest in BGL and rice-wheat land uses. Similar trend was observed in Cl contents. After 20 years of different LULCs, the levels of CO3 and HCO3 contents varied from 1.33 to 6.33 and 1.00 to 16.30 meq l-1, respectively. The average maximum CO3 content was recorded under BGL and rice-wheat land uses whereas, the minimum under L5 and L4 LULCs. However, the HCO3 content in surface soil (0–15 cm) was lowest under L3 followed by L2 and L1 whereas, under 15–30 and 30–60 cm soil depths it was lowest under L3 and L2 land uses (Table 7).

Table 7 Effect of different land use-land covers on ionic content.
Full size table

Principal component analysis

The principal component analysis limits the variables and extracts smaller number of independent factors for deducing the association among observed variables (soil properties). Three principal components (PC1, PC2 and PC3) with eigenvalues greater than 1 were extracted (Table 8). The PCA led to reduction of the initial dimension of the dataset to three components, which explains 65.9% of the total dataset variance. The first component (PC 1) of the rotated matrix showed strong positive correlation with available nitrogen (> 0.90) while as good positive correlation (> 0.75, < 0.90) was observed with BD, WHC, IR, available K, Na and OC (Table 9). The first component showed moderate correlation (> 0.60, < 0.75) with Ca + Mg and CO3. The second component showed strong correlation with clay and pH and good correlation with K, HCO3, EC and ESP. Similarly, the third component (PC 3) showed moderate correlation with silt, porosity, available P, Ca + Mg and Cl, and good correlation with sand and available K (Fig. 7).

Fig. 7
figure 7

Principal component analysis (PCA) of soil properties under different land uses in salt affected soils.

Full size image
Table 8 Principal component analysis of soil parameters.
Full size table
Table 9 Pearson’s correlation matrix among soil properties.
Full size table

Discussion

Present study revealed that the afforestation of sodium salt laden degraded sodic soils for more than 20 years showed a substantial reduction in salt content and improvement in physical, chemical, and biological properties, and favours in increasing SOC in surface and sub-surface layers and nutrient status of sodic soils compare to BGL and rice-wheat42,43,44. This is attributed to the enrichment of soil with the addition of organic matter, and thereby recycling of important nutrients20,45. In our study, soil physical properties like BD, porosity, and WHC influenced significantly with changing in LULCs. Among the different LULCs, highest improvement in soil physical properties was recorded under agroforestry based LULCs over the BGL and rice-wheat land uses. The highest improvement in these parameters after twenty years of study was recorded under P. juliflora based silvipastoral system. This may be attributed to the higher level of SOC in agroforestry LULCs. This result is in conformity with the results reported by workers14,46,47,48,49 in their earlier studies. The highest BD recorded in rice-wheat land use may be attributed to the exhaustive tillage operations in rice-wheat land use which may create a compact soil layer that increase BD50. Soil infiltration rate that is an important indicator of improvement in sodic soils has also influenced with LULCs. The improvement in this parameter ranges from 11.80 mm to 24.52 mm day− 1. This variation in infiltration rate may be attributed to differences in soil texture, BD and organic matter contents51. Lesser the organic matter, higher bulk density and declined porosity in rice-wheat and BGL results in poor soil structure and, thereby decreased infiltration rate52.

Soil chemical properties like pH, EC and ESP are the important parameters that decides the severity of the sodicity in the soil. In our study, the pH1:2 of surface soil (0–15 cm) in rice-wheat land use was significantly lower than that of land parcels covered under different agroforestry LULCs. This is due to reclamation of these parcels with chemical amendments like gypsum (CaSO4⋅2H2O) and continuous cultivation of rice-wheat crops53,54. However, highest reduction in this parameter in sub-surface (25–30 and 30–60 cm) soil layers recorded under silvipastoral land use52,55. This is because of replacement of exchangeable sodium by calcium due to the root exudates, and consequent leaching of sodium salts. Similarly, reduction in EC and ESP in surface soil (0–15 cm) was more pronounced in rice-wheat land use because of reclamation of this land parcel with gypsum (CaSO4⋅2H2O) and continuous cultivation of rice-wheat crops56,57. However, in sub-surface layers the reduction in these parameters were more in agroforestry LULCs than that of BGL and rice-wheat land uses. This may be due to high organic matter contents through decomposition of leaf litter, which mobilize the cations to neutralize the sodicity by reducing pH and evolution of CO2 that helps to mobilize the inherent Ca2+. This released Ca2+ can accelerate the reclamation by replacing the exchangeable Na+ from the soil58. This was more pronounced in silvipastoral land use, which proved more efficient in reducing soil pH, EC and ESP. Prosopis based land use significantly decreased exchangeable sodium percentage59. The organic carbon content is another important chemical characteristic of sodic soils. These soils are very poor (< 1.0 g kg− 1) in organic carbon content. The organic matter plays an important role in reduction of salt level60,61. After 20 years of study, it was found that the percent increase in organic carbon content in all the LULCs was higher in the surface (0–15 cm) layer and it declined with increasing depth. The highly significant increase in organic carbon was recorded in silvipastoral land use followed by silviculture and the lowest in BGL and rice-wheat land uses. The lower organic matter in rice-wheat and BGL compared to agroforestry land uses was most likely because of the less vegetation coverage. The trees continuously add organic matter through litter in the upper soil and increases root turnover52,62,63 which further increased SOC due to positive priming64. Our results are in conformity with the earlier findings reported65,66. Farooqi et al.67 links reclamation of salt-affected soils with increased carbon sequestration in soils. Lowest SOC in rice-wheat land use system in the present study may be due to removal of crop residues, intensive cultivation practices, and taking more microbial carbon from the soil system68,69,70. Declining in organic matter content may reduce soil quality as well as productivity of salt affected soils. Increasing area under rice-wheat land use may reduce the amount of organic matter and availability of essential nutrients in the soil resulting poor soil health and crop productivity. Organic matter content in soil plays an important role in the conversion and mineralization of nutrients in the soil. The highest mean available N content irrespective of soil depths was recorded under agroforestry LULCs, which was significantly higher over the BGL, and rice-wheat land uses (Table 6). This can be ascribed to the accumulation of organic residues in the form of leaf litter on the soil where BGL and rice-wheat land uses had lesser soil organic matter content than silvi-pastoral and silvicultural land uses. It has also been reported that the mineralization of the accrued soil organic matter to ammonia is the major sources of nitrogen in the soil71. Changes in land uses resulted in reduction of total nitrogen may affect the fertility and productivity of sodic soils because nitrogen is an essential nutrient for plant growth. In silvipastoral systems, trees help in maintaining soil fertility, cycle nutrients, improve microclimate and improve overall system productivity72. Improved soil fertility in agroforestry LULC having tree canopies may result from litter fall or dung inputs from sheltering animals59. Similarly, the increment in available P and K contents were recorded under agroforestry LULCs. This may be due to decreasing in soil pH and ESP and increasing soil organic matter contents, which may increase the availability of P and K to the plants. Additionally, the organic acid production during breakdown of organic materials increases P release. Soil quality improvement under P. juliflora in degraded lands including sodic lands has been reported from Indo-Gangetic plain region73.

Soil microbial properties like MBC, MBN and MBP, which are an indicator of microbial activities in the soil changes due to different LULCs. The highest improvement in these parameters was recorded under agroforestry LULCs. This may be attributed to the more organic matter in the top humus soil that promotes microbial activities57,74. MBC is a complex sign of changes in SOC because it has a much faster rate of turnover. Previous land use studies also reported significant reduction in MBC using conventional practices for rice-wheat land use as compared to forestland use75,76,77,78. Dehydrogenase activities (DHA) of the soil described the oxidative activity of soil microorganisms also influenced with LULCs. The higher DHA values recorded under silvipastoral and pastoral LULCs may be attributed to the assortment in vegetation, improvement in soil properties and environmental factors affecting soil properties. However, low DHA values in BGL and rice–wheat land uses were due to low SOC and repressing effects of mineral fertilizers as observed in present study. The organic matter added through leaf litter could also be a contributing factor to motivate DHA by increasing availability of substrate for the soil microbial community in agroforestry land uses79,80.

Among the ionic contents analyzed in this study, Na+ content, which was highly dominant in surface soils, decreased in all the LULCs. After twenty years of different LULCs, it decreased in surface layers (0–15 cm) and accumulated in the lower levels due to leaching of salts and dissolution of CaCO3 by the plant roots in the presence of CO2, causing replacement of adsorbed Na+. The deep-rooted trees in silvipasture and silviculture land uses could improve the nutrient status of degraded soil by redistributing nutrients from the deeper soil layers.

Conclusions

This study evaluated the impact of different land use–land cover (LULC) systems on the physico-chemical and biological properties of degraded sodic soils. Results showed that long-term agroforestry practices significantly improved soil health by increasing SOC, available N, P, K, microbial biomass (C, N, P), and dehydrogenase activity, while reducing soil pH and ESP. Among all systems, the silvipastoral system—P. juliflora with salt-tolerant grasses (L. fusca, T. alexandrinum)—was most effective in enhancing soil fertility and microbial activity. Conversely, traditional rice–wheat systems exhibited poor soil health due to low organic inputs and intensive cultivation. Overall, the findings highlight agroforestry, particularly silvipastoral land use, as a nature-based solution for restoring and sustaining sodic soils. Although, the present study is confined to represent sodic soils dominant in the Indo-Gangetic plains but may be useful for the other geographical areas having similar degraded lands. The study will be useful in strategizing nature based solution for rejuvenation of sodic lands through agroforestry land uses and management of soil fertility and ecological sustainability.