The higher relative concentration of K + to Na + in saline water improves soil hydraulic conductivity, salt-leaching efﬁciency and structural stability

. Soil salinity and sodicity caused by saline water irrigation are widely observed globally. Clay dispersion and swelling are inﬂuenced by sodium (Na + ) concentration and electrical conductivity (EC) of soil solution. Speciﬁcally, soil potassium (K + ) also signiﬁcantly affects soil structural stability, but for which con-cern was rarely addressed in previous studies or irrigation practices. A soil column experiment was carried out to examine the effects of saline water with different relative concentrations of K + to Na + (K + / Na + ), including K + / Na + of 0 : 1 (K0Na1), 1 : 1 (K1Na1) and 1 : 0 (K1Na0) at a constant EC (4 dS m − 1 ), and deionized water as the control (CK), on soil physicochemical properties. The results indicated that at the constant EC of 4 dS m − 1 , the inﬁltration rate and water content were signiﬁcantly ( P < 0 . 05) affected by K + / Na + values, and K0Na1, K1Na1 and K1Na0 signiﬁcantly ( P < 0 . 05) reduced saturated hydraulic conductivity by 43.62 %, 29.04 % and 18.06 %, respectively, compared with CK. The volumetric water content was signiﬁcantly ( P < 0 . 05) higher in K0Na1 than CK at both 15 and 30 cm soil depths. K1Na1 and K1Na0 signiﬁcantly ( P < 0 . 05) reduced the desalination time and required leaching volume. K0Na1 and K1Na1 reached the desalination standard after the ﬁfth and second inﬁltration, respectively, as K1Na0 did not exceed the bulk electrical conductivity required for the desalination prerequisite throughout the whole inﬁltration cycle at 15 cm soil layer. Furthermore, due to the transformation of macropores into micropores spurred by clay dispersion, soil total porosity in K0Na1 dramat-ically decreased compared with CK, and K1Na0 even increased the proportion of soil macropores. The higher relative concentration of K + to Na + in saline water was more conducive to soil aggregate stability, alleviating the risk of macropores reduction caused by sodicity.


Introduction
Freshwater shortage resulting from elevated demand for water resources as well as the irrational exploitation and use after economic and population growth (Zhang and Xie, 2019;Prajapati et al., 2021) constrains the sustainability of agricultural production (Aparicio et al., 2019).Alternative water resources with variable water quality (such as saline water) are being considered for agricultural irrigation in several desert and saline areas (Singh et al., 2021;Liu et al., 2022a).Utilizing saline water could partly alleviate the undersupply of freshwater for agricultural production (Yang et al., 2020).However, the other side of the coin is that saline water irrigation could result in soil salinization and/or sodicity.Once the soil is salinized and/or alkalized, soil hydraulic properties, like infiltration rate, saturated hydraulic conductivity and permeability, will change inevitably (Scudiero et al., 2017).And cations in the soil solution change the soil structural characteristics through the soil clay-particle dispersion and flocculation (Bouksila et al., 2013;Hack-ten Broeke et al., 2016;Zhang et al., 2018).Therefore, in order to optimize saline water utilization, the effects of saline water quality on the soil hydraulic properties and pore structure characteristics should be paid more attention.
Saline water irrigation can increase the monovalent ions concentration in soil solution and affect soil structure (Qadir et al., 2007(Qadir et al., , 2021)).Excess sodium (Na + ) from saline irrigation water is adsorbed onto the clay surface in salt-affected soils where sodium compounds predominate contributing to the disintegration of soil structure (Marchuk and Rengasamy, 2011;Belkheiri and Mula, 2013;Awedat et al., 2021).As percolation progresses, the thickness of the diffusion doubleelectron layers increases due to the relatively larger hydrated radius of Na + , and the repulsive force between adjacent diffusion double-electron layers appears to increase, resulting in the dispersion and swelling of soil particles (Alva et al., 1991;Reading et al., 2015).
Soil calcium (Ca 2+ ) and magnesium (Mg 2+ ) can alleviate soil dispersibility by replacing Na + in soil colloids, the outer layers of the Ca 2+ and Mg 2+ containing colloidal particles do not adsorb water molecules, turning Na + qualitative hydrophilic colloid into Ca 2+ and Mg 2+ hydrophobic colloids (Marchuk and Rengasamy, 2011;Tsai et al., 2012).Colloidal particles move closer to each other, promoting soil particles to form water-stable aggregates, thus improving soil structural stability (Gharaibeh et al., 2009;McKenna et al., 2019).Therefore, the concentration of Na + in relation to Mg 2+ and Ca 2+ (sodium adsorption ratio, SAR) (US Salinity Laboratory Staff 1954) is a crucial criterion for soil structural stability and hydraulic conductivity (Rengasamy and Marchuk, 2011).Although SAR can be used to predict soil clay dispersion effect caused by cations, the controlling mechanism of dispersion in SAR is presumed to be exchangeable Na + .However, Na + does not alone cause soil dispersion, since the chemical component of clay structure integrity is mainly a function of ionic valence and hydration radius (Marchuk et al., 2014).Potassium (K + ) has been overlooked because saltaffected soils typically contain low amounts of K + .However, Li et al. (2022) reported that under the continuous recycling use of underground saline water, water-soluble and exchangeable K + is higher than Ca 2+ and Mg 2+ in the Hetao irrigation district -one of the large irrigation districts in China.It is anticipated that the long-term use of irrigation water with high K + concentrations may therefore create substantial challenges in preserving good soil structure and adequate infiltration rates (Sposito et al., 2016).K + is not as effective as Na + in generating soil particle dispersion and swelling problems, yet Marchuk and Marchuk (2018) pointed out that K + could substitute Na + on exchange sites to encourage Na + leaching and increase water conductivity to some extent.A lower concentration of K + may have positive effects on soil permeability due to the substitution of exchangeable Na + by K + with lower dispersive potential, increasing aggregates stability and soil pore connectivity (Buelow et al., 2015).Traditional SAR ignored the role of K + , and a newly proposed equation, the cation ratio of soil structural stability (CROSS), could integrate the effects of Na + and K + in soil, which is an important indicator for assessing the quality of saline water (Rengasamy and Marchuk, 2011).
Thus, we hypothesized that the amount of K + relative to Na + would certainly have an effect on soil structural stability, which could be evaluated in a column experiment under controlled conditions.The specific objectives of this study were to (1) ascertain the effect of irrigation saline water with different relative concentrations of K + to Na + (K + / Na + ) on transport and distribution of water and salt, (2) determine the effect on soil pore structural characteristics, and (3) predict these effects using a newly proposed index (CROSS) rather than SAR.

Soil sampling location and properties
The study soil was collected from a layer of 0-40 cm field in Yangling ( 108• 04 E, 34 • 20 N), Shaanxi Province, China.After being air-dried, the soil was grounded to pass through a 2 mm sieve.The soil's physical and chemical properties are listed in Table 1.Soil particle size distribution was measured by the Laser Mastersizer 2000 (Malvern Instruments; Malvern, UK), and according to the international classification system, soil texture was classified as silty clay.Soil bulk density was calculated using the soil core method.EC e and pH were measured using a conductivity meter (DDS-307; China) and pH meter (PHS-3C; China), respectively.Total soluble salts refer to the total amount of soluble salts in soilsaturated paste extract.Flame photometry (6400A; China) was used to measure soluble Na + and K + , concentrations of CO 2− 3 and HCO − 3 were tested using the neutral titration method, Cl − was analyzed using the silver nitrate titration method, SO 2− 4 was determined using barium sulfate turbidimetric method, and Mg 2+ and Ca 2+ were specified using ethylene diamine tetraacetic acid (EDTA) titrimetric method (Bao, 2005).

Experimental design
Soil columns were prepared using transparent polyvinyl chloride cylinders, with an internal diameter of 20 cm and a height of 50 cm (Fig. 1).Round and small holes (6 mm diameter) were arranged equally at the bottom of each cylinder for drainage.A 5 cm depth quartz sand was laid at the bottom of the soil column as a filter layer before packing to prevent small soil particles from being washed away.After that, airdried soil was packed at 40 cm height with a bulk density of 1.35 g cm −3 (referring to the original level of the soil).The sieved, dry soil was poured into each soil column in the 5 cm sections for uniform compaction, and the layer's surface was roughened to ensure a tight connection to the next layer.The soil column was then allowed to stand in the laboratory for 24 h before starting the experiments described herein.The constant water head (2 cm, using a Mariotte bottle) infiltration experiment was conducted with three replications for each treatment.
The experiment implemented alternate leaching, as the prolonged leaching process of soil substrates is more helpful in illuminating the function of electrolyte effect and cation exchange (Shaygan et al., 2017).The next infiltration was performed 2 d after the drainage of the previous infiltration was completed.Soil layers were regarded as reaching the desalination prerequisite when the soil salt content came to less than 0.3 %, which meant that bulk electrical conductivity was less than 1.5 dS m −1 (transformation from salt content to bulk electrical conductivity) (Yin et al., 2022).Water application was stopped when the bulk electrical conductivity of all treatments at 15 cm depth reached the prerequisite for desalination.This experiment was planned to fill all the pores in the soil column throughout the infiltration cycle; therefore the water applied at the first infiltration was described by the pore volume equation (Huang and Xu, 2010) as follows: where V p is the pore volume (cm 3 ), V s is the volume of filled soil (cm 3 ), TP is the soil total porosity (%), d s is the soil particle density (2.65 g cm −3 ) (Huang and Xu, 2010) and BD is the bulk density (g cm −3 ).According to Eqs. ( 2) and (3), around 6 L of water was required in the first infiltration.Required water volume for each subsequent leaching was determined by the volume of leachate at the first infiltration; it was 0.5 L each time.

Soil property measurements
During the whole experimental period, soil volumetric water content and bulk electrical conductivity were real-time monitored at 15 and 30 cm soil depths from the soil surface by capacitance sensors (ECH2O 5TE, METER Group; USA) (Fig. 1).Leachate was collected in the leachate catcher below the soil column.Cumulative leachate volume was monitored over time to determine the saturated hydraulic conductivity (K sat ; cm min −1 ) of each treatment by using a derivation of Darcy's approach (Sahin et al., 2011) as follows: where V l is the leachate volume (cm 3 ), H is the length of filled soil (cm), A is the surface area of soil column (cm 2 ), t https://doi.org/10.5194/soil-9-339-2023SOIL, 9, 339-349, 2023  is the leaching time of measurement (min) and h is the height of constant water head (cm).
To determine the amount of salt released, we measured the volume and EC of the leachate.The leachate was collected at 3 h intervals when the leachate started to drain, and leachate was stored in 100 mL wide-mouth polypropylene reagent bottles.The salt accumulated in the soil column was determined by subtracting the salt in the leachate from the applied water, the salination rate (R s , %) indicated the ratio of salt accumulated in the soil column at every time of infiltration to the salt content at the first applied water.Leaching efficiency (L e ; g L −1 ) refers to the amount of desalination per unit of water volume in the desalination process.R s and L e were calculated as follows: where m s is the salt content accumulated in soil column at each infiltration (g), m w is the salt content in the total water used for the first infiltration (g), m 1 is the mass of salts after the first infiltration (g) and w is the total water volume used for leaching (L).Soil samples were collected from each soil column at 5 cm intervals with the 0-40 cm soil layer 3 d after the final infiltration.Soil BD was calculated using the soil core method, and TP was calculated by Eq. ( 3) based on BD.The soil water characteristics' curve was measured with the high-velocity centrifugal method (CR21 Hitachi; Japan) and calibrated by RETC software (PC Progress Inc.; Prague, Czech Republic).Currently, several defining sizes of macropores are proposed, rather than a precise definition and pore size range (Cameira et al., 2003;Kim et al., 2010;Hu et al., 2018;Budhathoki et al., 2022;Aldaz-Lusarreta et al., 2022).In this study, macropores were defined as the pores with diameters larger than 1 mm, whereas micropores were defined as smaller than 1 mm (Luxmoore, 1981;Wilson and Luxmoore, 1988).Based on the capillary pressure data, the relationship between pore diameter (d; mm) and water suction (S; Pa) was described according to the capillary bundle model (Jury et al., 1991):

Statistical analysis
Statistical analysis among all treatments with different K + / Na + was performed in SPSS 22.0 software, using oneway analysis of variance (ANOVA) based on the least significant difference (LSD) test at 95 % significance level (P < 0.05).All figures were created through Origin 2022b.

Results
3.1 Soil-saturated hydraulic conductivity (K sat ) The K0Na1, K1Na1 and K1Na0 significantly (P < 0.05) reduced K sat by 43.62 %, 29.04 % and 18.06 % compared with CK, respectively (Fig. 2).Additionally, K sat was negatively correlated with CROSS of saline water, and increasing the CROSS of the applied saline water generally reduced K sat .

Soil water content
Water content increased immediately after each infiltration in all treatments and then gradually decreased, and the degree of variation tends to stabilize (Fig. 3).Also, water content at deeper soil depths was greater than at shallow soil depths at the same time during the whole infiltration period.The water content ranged from 0.39-0.41and 0.40-0.42cm 3 at 15 and 30 cm soil depths, respectively.K0Na1 had the highest water content at both 15 and 30 cm soil depths.K1Na1 and K1Na0 were greater than CK at 15 cm soil depth and lower than CK at 30 cm soil depth, and the water content of K1Na1 was higher than K1Na0 at both 15 and 30 cm soil layers.

Soil salination rate (R s ) and leaching efficiency (L e )
The R s and L e under CK were not shown in Fig. 4; because deionized water was used as the control and there was almost no salt contained in the study soil, CK was considered negligible for salt accumulation and leaching.R s peaked at the first infiltration, and approximately 70 %-80 % of the salt in the saline water was retained in the soil column, after which the subsequent leaching had lower R s values (Fig. 4).The lower the relative concentration of K + to Na + , the larger soil R s .Among the three saline water treatments, K1Na0 had the lowest R s and highest L e at five infiltrations.

Soil bulk electrical conductivity
Bulk electrical conductivity of K0Na1, K1Na1 and K1Na0 ranged from 1.0-2.0dS m −1 at 15 cm and 1.5-2.5 dS m −1 at 30 cm soil depth (Fig. 5).After the first infiltration, bulk electrical conductivity in 15 cm soil depth peaked and then exhibited a general downward trend in the following infiltrations.However, more salts were leached to deeper layers, where salt began to accumulate instead of desalination occurring, and bulk electrical conductivity at 30 cm soil depth gradually increased following the infiltration events.Overall, K0Na1 had the highest bulk electrical conductivity among all treatments at both 15 and 30 cm, and K1Na1 was higher than K1Na0.
At 15 cm soil depth, K0Na1 reached the soil desalination prerequisite after the fifth infiltration, while K1Na1 https://doi.org/10.5194/soil-9-339-2023SOIL, 9, 339-349, 2023 reached the desalination prerequisite after the second infiltration, and K1Na0 did not exceed the desalination prerequisite during the whole infiltration period.Among all saline water treatments, K1Na0 reduced the desalination time and required leaching volume to reach the standard of desalination.K0Na1, K1Na1 and K1Na0 did not meet the desalination prerequisite at 30 cm soil depth, and the increased volume of infiltration water also increased the bulk electrical conductivity.

Soil bulk density (BD) and total porosity (TP)
Soil BD varied from 1.30-1.40g cm −3 across all treatments, and BD was below 1.35 g cm −3 at 0-10 and 35-40 cm soil layers; however it was over 1.35 g cm −3 at 10-35 cm soil depth (Fig. 6).K0Na1 significantly (P < 0.05) enhanced soil BD throughout the soil column profile compared with CK.TP first diminished with soil depth to reach a minimum at about 30-35 cm and then slightly increased at 35-40 cm.The TP of K1Na1 and K1Na0 slightly improved after five times of infiltration, and only K0Na1 showed a decline compared with CK.Overall, over the whole infiltration period, K1Na0 was most conducive to the formation of soil pore structure and increasing the total pore volume.The saline water with lower CROSS was beneficial for reducing soil BD and increasing TP.

Proportion of micropores and proportion of macropores
Micropores were the most dominant pores for all treatments: the volume of micropores accounted for more than 40 % of the total soil volume; however, the proportion of macropores did not exceed 8 % (Fig. 7).The 0-5 cm soil layer had the

Effects of saline water on soil water movement and redistribution
As a crucial soil hydraulic characteristic, K sat reflects the transportation ability of water and solutes (Braud et al., 2001;Maillard et al., 2011;Albalasmeh et al., 2022).The cation composition and EC of soil solution affect K sat by controlling electrostatic repulsive pressure through surface potential and midpoint potential between adjacent particles and consequently influence water movement (Fares et al., 2000;Liu et al., 2022b).Specifically, the relative concentration of K + to Na + in saline water was related to the swelling and dispersion of soil particles (Yu et al., 2016;Zhu et al., 2019).
Dispersed clay particles clogged soil macropores to subsequently restrict water transport (Awedat et al., 2021).The Na + has a relatively higher ionicity index than K + ; as a result, the low relative concentration of K + to Na + decreased the degree of covalency in clay-cation bonds, which was detrimental to clay particles aggregation (Marchuk and Rengasamy, 2011).Therefore, in our study, the high relative concentration of K + to Na + promoted the flocculation and stabilization of soil clay particles, resulting in an increased water hydraulic conductivity (Fig. 2).After a certain period of irrigation, soil moisture redistributed at different depths of soil column (Fig. 3).Soil water moved further down during the phase of water redistribution soon after each irrigation, reducing the water content in the upper soil layers.As the upper soil layers drained, https://doi.org/10.5194/soil-9-339-2023SOIL, 9, 339-349, 2023 the lower soil layers still had water inflow (Kargas et al., 2021), increasing the water content in the lower soil layers.
The results also implicated that the retention of soil water by Na + was stronger than that by K + ; the cause may be that Na + can increase the thickness of the diffuse double layers around soil colloids theoretically due to its larger hydrated radius and lower charge than K + , and the adjacent double layers overlapped to provide more space between layers (He et al., 2015), where, subsequently, more water can be retained (Fig. 3).Additionally, our study showed that K1Na1 was even more beneficial than deionized water for water downward transport (Fig. 3), which was because deionized water (CK) (below 0.2 dS m −1 ) tended to leach soluble minerals and salts, especially Ca 2+ , from the surface soil layers.This would lead to the reduction of its original, solid soil structural stability.In the absence of salt and Ca 2+ , the dispersed tiny particles filled the smaller pore spaces in soil, reducing even more channels for water flow and exacerbating water retention in deeper soil layers (Ayers and Westcot, 1985).However, a lower concentration of soluble salts could increase colloid flocculation and thereby improve soil aeration and water conductivity (Tang and She, 2016).

Effects of saline water on soil salination and desalination process
Numerous factors influence soil salt-leaching efficiency; for example, increasing EC and reducing SAR definitely improve clay flocculation, which can enhance salt leaching (Ebrahim Yahya et al., 2022).Na + is more likely to trigger soil clay dispersion and swelling than K + ; thus Na + generally inhibits water infiltration, which is detrimental to salt leaching (Smiles and Smith, 2004).Adding K + could promote displacement of the adsorbed Na + and then decrease Na + concentration and salt accumulation in soil solute through leaching.A greater reduction in Na + concentration was associated with a higher rate of cation exchange rate, and the slow rate of solute leaching from aggregates reduced the total leaching efficiency (Shaygan et al., 2017).During the leaching process, water flow preferentially passed through the macropores rather than aggregates.The slow water transportation through aggregates induced the slow removal of solutes from the aggregates, leading to a reduced leaching efficiency.In our study, the alternate leaching was implemented to improve solute leaching (Figs. 4 and 5).The soil solutes diffused into the aggregates surface during the rest period, improving salt leaching due to the water flow in macropores (Al-Sibai et al., 1997).Saline water with more K + could increase the magnitude of cation exchange due to the substitution of Na + on exchange sites by K + with lower dispersive potential (Shaygan et al., 2017); the intensive release of cations from the soil further improved salt's leaching efficiency.In addition, the integrity of soil aggregates created by combining clay particles and the other soil components enhanced by K + can benefit solute transportation (Marchuk and Rengasamy, 2011).

Effects of saline water on soil pore structure characteristics
Due to the long-term continuous irrigation, the upper layer of soil was immersed in water, resulting in the swelling and loosening of soil particles (Vaezi et al., 2017;Håkansson and Lipiec, 2000), which caused the reduction of BD in the surface layer (Fig. 6).The subsoil BD increased with depth under the impact of water pressure and self weight due to the declining pore diameter and pore branching closure (Schjønning et al., 2013).And for soil at the bottom, the loss of soil particles from small holes was responsible for the abrupt reduction in BD (Fig. 6).The value of CROSS in saline water could reflect changes in soil BD and TP, in agreement with the result of Marchuk and Marchuk (2018).The high CROSS implied an increase in the proportion of monovalent exchangeable cations, thickening the double layer at the interface between the clay surface and soil solution.Hence, soil swelling occurred at the expense of water-conducting pores.Additionally, aggregates slaking and subsequent clay dispersion and deposition of clay particles within the pore space contributed to the reduction in TP (Marchuk and Marchuk, 2018).Soil macropores play a crucial role in water and solute transport, accounting for 85 % of the total infiltration volume (Wilson and Luxmoore, 1988;Weiler and Naef, 2003;Kotlar et al., 2020).For saline water with the same EC, a decrease in K + concentration may enhance soil clay dispersion, resulting in the loosening of clay particles from the aggregates.The dispersed clay particles moving with the water could clog the soil macropores, promoting the conversion of macropores to micropores (Cameira et al., 2003), thus reducing the macropore volume (Fig. 7).

Conclusions
We explored the effects of the relative ratio of K + to Na + in saline water on soil hydraulic characteristics and structural stability via a soil column experiment.Irrigation with saline water of K + / Na + of 1 : 0 caused fewer pore blockages due to soil clay particle dispersion than 0 : 1, which increased soil-saturated hydraulic conductivity.The presence of K + accelerated the sustained Na + replacement and leaching, alleviating salt accumulation and enhancing leaching efficiency.K + positively affected the establishment of soil structure due to the transformation of micropores into macropores, and the ever-increasing unobstructed water-conducting channels sped up water and solute transport.The rational use of saline water with adequate K + could help mitigate the structural deterioration caused by Na + .Appropriate adjustment of the relative concentration of K + to Na + in saline water during infiltration could ameliorate soil structural properties.In addi-tion to Ca 2+ and Mg 2+ (primary concerns in earlier studies), the relative concentration of K + to Na + is an essential indicator for assessing the suitability of saline water quality for irrigation and should be considered when using saline water.

Figure 1 .
Figure 1.Illustration of the experiment apparatus (a) and schematic diagram (b).

Figure 2 .
Figure 2. Saturated hydraulic conductivity (K sat ) under different treatments.K0Na1, K1Na1 and K1Na0 indicate the saline water at EC of 4 dS m −1 with K + / Na + of 0 : 1, 1 : 1 and 1 : 0, respectively.CK is deionized water.Different letters after the means of K sat indicate statistical differences (P < 0.05) among treatments based on LSD.Bars indicate standard deviations of means.

Figure 3 .
Figure 3. Variation of water content over time under different treatments at 15 (a) and 30 cm (b) soil depths during the five times of infiltration.

Figure 4 .
Figure 4. Salination rate (R s ) (a) and leaching efficiency (L e ) (b) at five infiltrations under all saline water treatments.K0Na1, K1Na1 and K1Na0 indicate the saline water at EC of 4 dS m −1 with K + / Na + of 0 : 1, 1 : 1 and 1 : 0, respectively.Different lowercase letters followed by means of R s indicate statistical differences (P < 0.05) among treatments based on LSD, and different uppercase letters followed by means of L e indicate statistical differences (P < 0.05) among treatments based on LSD.Bars indicate standard deviations of means.

Figure 5 .
Figure 5. Variation of bulk electrical conductivity over time under treatments with different K + / Na + at constant EC at 15 (a) and 30 cm (b) soil depths in the period of five times of infiltration.K0Na1, K1Na1 and K1Na0 indicate the saline water at EC of 4 dS m −1 with K + / Na + of 0 : 1, 1 : 1 and 1 : 0, respectively.CK is deionized water.

Figure 6 .
Figure 6.Soil bulk density (BD) and total porosity (TP) throughout the soil column profile under different treatments.K0Na1, K1Na1 and K1Na0 indicate the saline water at EC of 4 dS m −1 with K + / Na + of 0 : 1, 1 : 1 and 1 : 0, respectively.CK is deionized water.The blue horizon columns represent BD, while the green horizon columns represent TP.Different lowercase letters followed by means of BD indicate statistical differences (P < 0.05) among treatments based on LSD, and different uppercase letters followed by means of TP indicate statistical differences (P < 0.05) among treatments based on LSD.Bars indicate standard deviations of means.

Figure 7 .
Figure 7. Proportion of micropores and proportion of macropores in total soil volume throughout the soil column profile under different treatments.K0Na1, K1Na1 and K1Na0 indicate the saline water at EC of 4 dS m −1 with K + / Na + of 0 : 1, 1 : 1 and 1 : 0, respectively.CK is deionized water.The blue horizon columns represent proportion of micropores, while the green horizon columns represent proportion of macropores.Different lowercase letters followed by means of proportion of micropores indicate statistical differences (P < 0.05) among treatments based on LSD, and different uppercase letters followed means of proportion of macropores indicate statistical differences (P < 0.05) among treatments based on LSD.Bars indicate standard deviations of means.

Table 1 .
The physicochemical properties of study soil.

Table 2 .
Saline water settings with different K + / Na + at a constant EC.