Phytoremediation of calcareous saline-sodic soils with mesquite (Prosopis glandulosa)

ABSTRACT

We evaluated phytoremediation potential of a calcareous saline-sodic soil with Prosopis glandulosa relying on either N₂ fixation (NFIX) or inorganic N (IN) and gypsum. Straw and root dry matter of NFIX were 60% and 67% higher than that of IN, respectively. While both plants relying on IN or NFIX retained nearly similar Ca²⁺, Na⁺ level in IN lysimeters was 1.8 folds that of NFIX, whereas levels of Na⁺ and K⁺ in roots of NFIX were 13% and 24% higher than that of IN plants. P. glandulosa relying on NFIX and IN released 39% and 50% Na⁺ of that released by the gypsum. P. glandulosa produced more than the gypsum treatment and induced 24% of Ca²⁺ solubilization found in the gypsum treatment. Translocation factor of Na⁺ and the K⁺/Na⁺ ratio in NFIX and IN were 0.58, 0.99 and 13.2, 7.0, respectively. The practical implication of this study conclude that P. glandulosa grown with NFIX or IN is an alternative practice for amelioration of calcareous salt-affected soils.

KEYWORDS:

• Amelioration
• drylands
• land degradation
• N-fixation
• solubilization
• translocation

Introduction

Soils with salts to levels that can adversely affect several of their properties as well as growth of most crops are termed salt-affected soils. The accumulation of salts can both adversely affect plants’ performance (Aslam et al. 2011) and the physico-chemical characteristics of the soil and thereby limiting its potential use (Qadir & Schubert 2002). The rate of salinization of soil and groundwater can be further accelerated with climate change and increased irrigation with brackish water because it leads to increased evaporation and promotes capillary rise of saline ground water (Jesus et al. 2015). In arid and semi-arid regions, spatial and temporal (1999–2000) changes in electrical conductivity (EC) and CI⁻¹ were 16.8 µS cm⁻¹ and 0.18 meq I⁻¹ Y⁻¹, respectively (Yazdanpanah 2016). In more than 100 countries worldwide, occurrence of salt-affected soils is estimated at 1–10 billion ha with a potential of 10–16% increase/year (Qadir & Oster 2002; Yensen & Biel 2006; Aydemir & Sünger 2011). In irrigated dry lands, processes such as salinization and sodication are major driving forces for desertification and thereby threatening the productive capacity of agricultural lands, forestlands and rangelands (Dregne et al. 1991; Mustafa 2007).

The native plants of Prosopis sp are adapted to the conditions of arid and semi-arid regions and are very important in controlling desertification through stabilization of sand dunes, improving soil fertility, reducing soil salinity, providing fuel energy resources, supplying feed and forage for grazing animals, furnishing construction timber, supplementing food for humans and promoting honey production (Manzano & Navar 2000). Prosopis was therefore introduced in Sudan in the year 1917 with the aim of combating desertification by stopping sheet sand encroachment on sandy soils (e.g. Salih 1998; Elfadl & Luukkanen 2003). Prosopis juliflora grows in saline soils with EC from 8 to 51 dS m⁻¹ (Joshi & Hinglajia 2000) and in sodic soils at pH of 10 and exchangeable sodium percentage (ESP) of 60 in the rooting zone (Singh et al. 1998). However, some Prosopis sp such as Prosopis koelziana can tolerate salinity up to 185 dS m⁻¹ (Emtahani & Elmi 2006).

The genus Prosopis contains many N₂-fixing species throughout the world’s semi-arid regions. Previous work with Prosopis glandulosa found that small young trees obtained most of their N from N₂-fixation (NFIX), while mature trees that had accumulated 1.3 Mg N ha⁻¹ in the soil beneath their canopy derived a much smaller percentage of their N₂ from NFIX. Prosopis was earlier said to be an active N₂ fixing tree where N derived from fixation (%Ndfa) determined in wood trunk and leaves were 75% and 25%, respectively (Geesing et al. 2000). Due to the limited infiltration of arid and semi-arid soils, the need for organic amendment is necessary to improve soil hydraulic conductivity (De Boever et al. 2016), which is controlled by particle size distribution (Mazaheri & Mahmoodabadi 2012). This clearly showed that salinity not only negatively affects soil physico-chemical conditions but also microbial functional properties (Min et al. 2016). Although gypsum is superior in reclamation of sodic soils (Ahmed et al. 2015), the ability of organic amendments in depleting deleterious salts from soil have reported positive effects. For example, combined application of organic amendments (Pistacia vera) with gypsum enhanced solubility of Na⁺ (Mahmoodabadi et al. 2013), improving the availability of macronutrients such as N, P and K⁺ (Yazdanpanah et al. 2013) and protect the soil by favoring the appearance of spontaneous vegetation (Tejada et al. 2006). Amelioration strategies mainly rely on leaching of soluble salts with consequent decrease in EC. During reclamation, the final EC and ESP in the leachate may drop to 18% and 11.6% of their respective initial values (Yazdanpanah & Mahmoodabadi 2013). Amendments may increase the soil cation exchange capacity (CEC) and have greater adsorption of divalent cations than Na⁺ and thereby leachate contains high concentrations of Na⁺ (Jalali & Ranjbar 2009). Later, Yazdanpanah et al. (2011) stated that leachate characteristics depend not only on the soil properties but also on the characteristics of the amendment.

The physical reclamation of salt-affected soils through leaching excess salts might be constrained by availability and high cost of high-quality irrigation water (Zhang et al. 2005). Amelioration through proton release from salt-tolerant legumes is increasingly becoming more widely used. Qadir et al. (2003a, 2003b) reported that Na⁺ leaching was equivalent to that of gypsum. Mubarak and Nortcliff (2010) concluded that some legumes that are known to fix N₂ in calcareous salt-affected soils might be an alternative ameliorant to the extremely expensive gypsum through calcite solubilization and a consequent release of Ca²⁺. However, they tested cultivated legumes (mainly: hyacinth bean Dolichos lablab, cowpea Vigna unguiculata L.). Therefore, the aim of this work was to study the phytoremediation potential of P. glandulosa grown in calcareous saline-sodic soil. We hypothesized that bioremediation of such soils could be through (a) solubilization of native Ca²⁺ by the H⁺ released during NFIX and (b) Na⁺ exclusion.

Materials and methods

Soil

The soil, known as Jahgaru in Japan, was collected from the Okinawa Prefectural Kyushu Agricultural Research Center, Okinawa, Itoman city, Makabe 820, Japan. Jahgaru is a heavy-textured smectitic soil and shows problematic physical properties such as severe hardening after air-drying. The soil is classified as loamy, siliceous, thermic Typic Udorthents (CSCC 1994).

Addition of salts

The soil was air-dried, crushed and sieved through 2.0 mm sieve before additions of salts. The soil was made saline-sodic by dissolving NaCl (99.9% purity) in distilled water and sprayed on the soil surface to 80% of saturation and air-dried at 30°C followed by adding only distilled water to 80% saturation and left overnight then after oven dried at 80°C for three days. Addition of distilled water to 80% saturation and drying was repeated for three cycles before the soil was made ready for the next stage. The content of Na⁺ needed to raise the sodium adsorption ratio (SAR) to 19.9 and EC of the extract to 16.0 dS m⁻¹ was calculated according to the following equation:

SAR = Na/[(Ca + Mg)/2]^0.5.

However, the exchangeable sodium percent (ESP) was calculated from SAR values according to the following equation:

ESP (%) = 1.95 + 1.23 SAR (Rashidi & Seilsepour 2008).

The gypsum required (GR) to replace 60% of the exchangeable Na⁺ was calculated according to USDA (1954) as follows:

GR (g/kg soil) = 60% × ESP (33.4%) × 0.86 = 3.75

This was followed by adjustment of CaCO₃ from 5% to 8.3% using analytical powder CaCO₃. Table 1 shows some selected properties of the soil used in the study.

Table 1. Selected physico-chemical properties of the calcareous saline-sodic soil prepared from Jahgaru soil added to the lysimeters (mean ± standard deviation).

Property	Value
pH(paste)	07.59 ± 0.08
ECe (dS m^−1)	16.99 ± 0.48
CEC (cmolc kg^−1)0	16.26 ± 0.58
ESP (%)	22.30 ± 1.63
SAR	19.90 ± 1.60
TN (μg mg^−1)	1.16 ± 0.01
TC (μg mg^−1)	13.77 ± 0.63
C/N	11.87 ± 0.46
CaCO3 (g kg^−1)	80.30 ± 2.43
SP (g kg^−1)	678.8 ± 8.5
Sand (g kg^−1)	158 ± 3.9
Silt (g kg^−1)	427 ± 3.6
Clay (g kg^−1)	415 ± 0.6

Notes: data are represented as means of four replications. Values after means are standard deviations. ECe: electrical conductivity of the extract; CEC: cation exchange capacity; ESP: exchangeable sodium percentage; SAR: sodium adsorption ratio; TN: total nitrogen; TC: total carbon; SP: saturation percentage.

Remediation by P. glandulosa studied in soil lysimeters

The experiment involved the following treatments:

Control (no plant-no gypsum)

Gypsum added according to GR

Prosopis relying on inorganic N (IN)

Prosopis relying on NFIX

Sixteen lysimeters made from polyvinyl chloride (11.7 cm internal diameter and 30 cm height) were prepared. A hole was drilled (1 cm diameter) in the bottom of each lysimeter and a 25 cm leaching hose was tightly fixed inside the hole. A filter paper (No. 5A 110 circle mm, 0.09 mg/circle) was placed at the bottom of each lysimeter. About 2 cm height of acid/distilled-water-washed sand obtained from Japan Sea was then placed on top of the filter paper. Exactly 3.0 kg of the developed saline-sodic soil was then added to the lysimeters. In order to adjust the bulk density to 1.2 g cm⁻³ (almost similar to filed bulk density), the soil height inside the lysimeters was reduced to 25.3 cm by gently hand packing against a wooden table. About 4 cm was then left on top of each lysimeter for irrigation. The amount of distilled water equivalent to 70% of the soil water-holding capacity (WHC) was added in increments to all lysimeters before sowing. Prosopis seeds were immersed in boiled water (100°C) for 10 min, incubated for germination in moistened Petri-dishes at 35°C for 24 h before sowing (Majd et al. 2013). The inoculum (EN 15) supplied by the National Center for Research, Khartoum, Sudan was added at the rate of 10 g/kg soil and thoroughly mixed in lysimeters designated with NFIX and immediately 70% of the soil WHC was added. At sowing (9/12/2015), four germinated seeds were sown (1 cm deep) in the treatments with Prosopis (both relying on NFIX and IN) and soil moisture content was kept at 70% of WHC by regular weighing of an extra lysimeter filled with similar soil and packed to the same bulk density and then kept at 70% WHC. The lysimeters were arranged in a completely randomized design with four replications and placed in biohazard greenhouse. The temperature and light (500 W) during the growth period were maintained at 25°C ± 0.5 and 12 h, respectively.

After 150 (07/05/2016) days from sowing, plants were harvested at 2 cm height from the soil surface using a sharp blade. Plant height was measured (cm), oven dried (80°C for two days), weighed (mg plant⁻¹), finely blended and analyzed for total carbon (TC) and total nitrogen (TN) by dry combustion (JM-MACRO CORDER 1000CN, J-SCIENCE CO, LTD, JAPAN) and content of Na⁺, Ca²⁺, Mg²⁺ and K⁺ (Chapman & Pratt 1982) after wet-acid digestion in a mixture of HNO₃ and 30% H₂O₂ (Jones 2001) using auto-analyzer AA-6800, SHIMADZU Corporation, Japan). The soil in each lysimeter was removed carefully and divided into 0–15 and 15–30 cm, air-dried, crushed and sieved (2.0 mm). Soil paste was made and the pH_paste and EC of the paste extract were measured with DKK-TOA HM-25R, Japan and DKK-TOA CM-30R, Japan, respectively. From each extract, soluble cations (Na⁺, K⁺, Ca²⁺ and Mg²⁺) were determined with AA analyzer,  (Bower & Wilcox 1965), Cl⁻, , and were determined using ion Chromatography (Model L-7610 DEGASSER, Hitachi, Japan). The translocation factor (TF) or mobilization ratio of the element is used to determine the relative translocation of metals from the growing medium to other parts (root and shoot) and was calculated as the ratio of the element content in shoots to that in roots (Padmavathiamma & Li 2007). For salinity stress tolerance, the K⁺/Na⁺ ratio in the straw and root was calculated (Flowers 2004). The SAS (1999) package was used to separate means between treatments, whereas differences between N₂ fixing and non-fixing were determined using the T-Test.

Results

Plant parameters

Straw and root dry matter content and plant height of Prosopis relying on NFIX were higher (P ≤ .02,.01 and .02 for straw, root and plant height, respectively) than those relying on IN by 60%, 67% and 31%, respectively (Table 2). Although straw N content was higher (P ≤ .02) in plants relying on IN (by 29%), the TN in the root of NFIX treatment was higher (P ≤ .02) by 29%. Both Prosopis relying on NFIX or IN contained similar Ca²⁺ level in either the shoot or root. But Ca²⁺ level in the shoot was 50% (1.5 fold) than that in the root (P ≤ .01).

Table 2. Effects of NFIX and IN on dry matter content (DM, mg Plant⁻¹), Plant height (cm), shoot and root content of C, N, Na⁺, Mg²⁺, Ca²⁺ (mg g⁻¹) and C/N ratio of P. glandulosa grown in a calcareous saline-sodic soil. (mean ± standard deviation).

Treatments			C	TN	Na^+	K^+	Mg^2+	Ca^2+
	DM mg Plant^−1	Plant Height cm	g kg^−1		(mg g^−1)				C/N
Shoot
IN	446b ± 103	15.9b ± 2.1	450a ± 2	19.2a ± 1.6	1.41a ± 0.1	9.77a ± 0.6	1.05a ± 0.1	3.33a ± 0.6	23.6a ± 1.9
NFIX	714a ± 148	20.9a ± 1.7	444a ± 5	13.7b ± 2.6	0.77b ± 0.1	9.96a ± 0.7	0.92a ± 0.1	3.30a ± 0.8	18.9a ± 2.1
P ≤ .05	0.02	0.02	0.17	0.01	0.03	0.76	0.12	0.95	0.09
SE	59.89	1.08	3.19	2.06	0.57	0.59	0.06	0.64	1.92
Root
IN	195b ± 38		426.1b ± 5.8	14.7b ± 0.9	1.35b ± 0.2	5.41b ± 0.7	0.86a ± 0.1	2.14a ± 0.3	25.1a ± 1.5
NFIX	326a ± 88		435.4a ± 0.8	17.2a ± 1.0	1.53a ± 0.4	6.70a ± 1.1	0.87a ± 0.1	2.21a ± 0.4	29.1a ± 1.5
P ≤ .05	0.01		0.03	0.04	0.04	0.02	0.93	0.67	0.07
SE	56.92		2.66	0.79	0.19	0.83	0.07	0.15	1.35

Notes: data are represented as means of four replications. Values after means are standard deviations. Means within columns in the Shoot or Root followed by similar lowercase letters are significantly different at P ≤ .05 using Least Significant Difference. SE: standard error; IN: P. glandulosa in lysimeters relying on IN; NFIX: P. glandulosa in lysimeters relying on nitrogen fixation.

The content of Na⁺ in the above ground plant material relying on IN was 1.8 folds (P ≤ .03) than that relying on NFIX. However, the content of Na⁺ in the roots of plants relying on NFIX was almost 13% higher (P ≤ .04) than that relying on IN. The K⁺ level in the shoot was not different (P ≤ .67) between N₂ fixing and non-fixing plants, however, roots of Prosopis relying on NFIX accumulated higher (P ≤ .02) K⁺ (by 24%) level than that relying on IN. In both plants, the ratio of shoot to root K⁺ was 1.5:1–1.8:1. The content of Mg²⁺ in the shoot (P ≤ .12) and roots (P ≤ .93) of both Prosopis was almost similar and varied from 0.86 to 1.05 mg g⁻¹. Accordingly, the K⁺/Na⁺ ratio in the shoot of N₂ fixing Prosopis was almost twofolds (P ≤ .005) that of non-fixing N₂, whereas the ratio was similar (P ≤ .48) in the roots (Figure 1). Only the translocation values (TF) of Na⁺ were higher (P ≤ .05) in non-fixing than N fixing Prosopis by 71% while the TF of K⁺ in the former was 20% higher (P ≤ .37) than in the latter by 20%, though not significant (Table 3).

Figure 1. Effects of NFIX and IN on K⁺/Na⁺ ratio in the straw and root of Prosopis glandulosa grown in a calcareous saline-sodic soil. Note: data are represented as means of four replications. Vertical bars are standard deviations. K+/Na+ ratio values of Straw IN and Straw NFIX or Root IN and Root NFIX with P ≤  .05 are significantly different using T-Test. Straw IN: straw in lysimeters with P. glandulosa relying on IN; Straw NFIX: straw in lysimeters with P. glandulosa relying on IN; Root IN: root in lysimeters with P. glandulosa relying on IN; Root NFIX: straw in lysimeters with P. glandulosa relying on nitrogen fixation.

Table 3. Effects of NFIX and IN on TF (concentration in shoot/concentration in root) of Na⁺, K⁺, Ca²⁺ and Mg²⁺ of P. glandulosa grown in a calcareous saline-sodic soil (mean ± standard deviation).

Treatment	Na^+	K^+	Ca^2+	Mg^2+
IN	0.99 ± 0.3	1.83 ± 0.3	1.22 ± 0.1	1.57 ± 0.4
NFIX	0.58 ± 0.2	1.52 ± 0.3	1.08 ± 0.2	1.55 ± 0.6
P ≤ .05	0.05	0.37	0.96	0.25
SE	0.14	0.29	0.31	0.10

Notes: data are represented as means of four replications. Values after means are standard deviations. SE: standard error; IN: P. glandulosa in lysimeters relying on IN; NFIX: P. glandulosa in lysimeters relying on nitrogen fixation.

Soil parameters

Treatments showed different effects on parameters determined in the 0–15 and 15–30 cm depths (Table 4). Incorporation of gypsum or planting Prosopis relying on NFIX has reduced (P ≤ .02) pH by 0.1–0.2 folds when compared with the control. The EC in the 0–15 cm depth was similarly reduced (P ≤ .001) with both gypsum application and planting Prosopis by 7–14% while in the lower depth, the gypsum treatment reduced (P ≤ .0006) salinity by 27%. In the 0–15 cm depth, the gypsum treatment showed the highest total salts’ concentration. The highest (P ≤ .0001) Na⁺ level was obtained in the gypsum lysimeter (more than twofolds of the control), whereas Na⁺ level in the NFIX lysimeter was higher (P ≤ .0001) than that obtained in the control (1.6 folds). Accordingly, the Na⁺ released in the gypsum, IN and N₂ fixing lysimeters were 75.8, 29.6 and 37.4 meq L⁻¹, respectively (P ≤ .0001). Therefore, the N₂ fixing and non-fixing P. glandulosa released 39% and 50% Na⁺ of that released by the gypsum. At both depths, soluble Ca²⁺ level in Prosopis lysimeters was higher (P ≤ .0001–.01) than that obtained in the respective control lysimeter by 92–100% and 8–15%, respectively. The content of Mg²⁺ in both depths followed similar trend to Ca²⁺ where the levels in lysimeters with Prosopis were higher (P ≤ .002–.0001) than that obtained in control lysimeters by 100% and 17%, respectively. The levels of K⁺ in the soil solution of the gypsum and Prosopis treatments were higher (P ≤ .0001) than that of the control. However, K⁺ level in Prosopis relying on IN obtained in the 0–15 cm depth was exceptionally high (P ≤ .01). The SAR values of the 15–30 cm depth were lower (P ≤ .0001) in gypsum and lysimeters with Prosopis. The level of in the upper depth of lysimeters with Prosopis was higher (P ≤ .02) than in other treatments by 51–57%. However, in the lower depth of NFIX treatment, the level of was higher (P ≤ .05) than other treatments by almost 22%. The level of Cl⁻ in both depths was erratic where it was higher (P ≤ .0008) in the gypsum treatment of the upper depth and lower (P ≤ .0002) than the same treatment in the lower depth. At the 0–15 cm depth, the level of was not different (P ≤ .18) among treatments, whereas in the 15–30 cm depth, the level in N₂ fixing lysimeters was almost 10 folds (P ≤ .005) that found in the inorganic lysimeter. The in the upper depth of the inorganic treatment was 3.5 folds (P ≤ .02) of other treatments. The level of soluble in both depths was high (P ≤ .0001) and showed 1.8–6 folds of other treatments.

Table 4. Effects of NFIX, IN, Gypsum and Control on soil chemical properties (pH_paste, EC, ECe; soluble cations, Na⁺, K⁺, Ca²⁺, Mg²⁺ and anions, HCO₃⁻, CI⁻, NO₂⁻, NO₃⁻, SO₄²⁻ and sodium adsorption ratio, SAR) in the 0–15 and 15–30 cm depths after harvest of P. glandulosa grown in a calcareous saline-sodic soil (mean ± standard deviation).

Treatment	pH(paste)	ECe dS m^−1	(Soluble cations meq L^−1)					Soluble anions (μg^−1)
			Na^+	K^+	Ca^2+	Mg^2+	SAR	HCO3^-	Cl^-	NO2^-	NO3^-	SO4^2 -
0–15 cm
C	7.2a ± 0.2	15.7a ± 1.5	59.2c ± 12.4	0.62c ± 0.1	19.1c ± 6.3	2.2c ± 0.8	18.3a ± 1.5	16.97b ± 3.0	2.7c ± 0.7	0.47a ± 0.5	0.35b ± 0.2	0.32b ± 0.1
G	7.0b ± 0.1	14.2b ± 0.6	135a ± 5.7	1.16b ± 0.1	96.2a ± 3.6	9.3a ± 0.5	18.6a ± 0.9	16.95b ± 1.7	6.7a ± 0.9	0.05a ± 0.0	0.21b ± 0.0	2.08a ± 0.3
IN	7.1ab ± 0.1	14.6b ± 1.4	88.8b ± 20.6	2.6a ± 0.4	38.1b ± 4.3	4.3b ± 0.5	19.4a ± 4.9	26.52a ± 7.6	4.6b ± 1.4	0.52a ± 0.5	1.02a ± 0.6	0.36b ± 0.1
FIX	7.1b ± 0.1	13.5b ± 0.9	96.6b ± 8.8	0.9cb ± 0.1	36.7b ± 6.9	4.5b ± 1.2	21.4a ± 1.2	25.20a ± 3.9	4.9b ± 0.7	0.10a ± 0.1	0.32b ± 0.1	0.38b ± 0.1
CV (%)	1.5	8.1	13.8	15.1	11.54	15.89	13.77	21.63	20.89	122.13	71.01	19.22
P ≤ .05	0.02	0.001	0.0001	0.0001	0.0001	0.0001	0.39	0.02	0.0008	0.18	0.02	0.0001
SE	0.52	0.95	0.85	0.95	0.97	0.93	0.21	0.56	0.74	0.32	0.54	0.97
LSD	0.17	1.81	20.2	0.31	8.45	1.24	NS	7.13	1.51	NS	0.52	0.23
15–30 cm
C	7.0a ± 0.1	17.2a ± 1.3	94.2a ± 5.5	1.2b ± 0.1	88.6bg ± 10.2	10.3b ± 0.9	13.4a ± 0.4	21.36b ± 3.6	6.3a ± 0.8	0.09b ± 0.0	0.18b ± 0.0	1.11b ± 0.1
G	6.9a ± 0.1	12.6b ± 0.5	61.9b ± 2.1	1.2b ± 0.0	78.0c ± 3.6	9.2b ± 0.3	09.4c ± 0.2	21.13b ± 5.9	4.1b ± 0.9	0.11b ± 0.0	0.24b ± 0.0	2.21a ± 0.4
IN	6.9a ± 0.1	15.7a ± 1.8	90.7a ± 6.9	1.3a ± 0.0	102.2a ± 10.6	12.2a ± 1.2	11.9b ± 0.5	21.53b ± 9.1	7.1a ± 0.3	0.06b ± 0.0	1.29a ± 0.3	1.29b ± 0.1
FIX	6.9a ± 0.1	15.8a ± 1.4	89.1a ± 5.2	1.3a ± 0.0	95.9ab ± 8.8	11.9a ± 1.2	12.2b ± 0.5	26.12a ± 4.5	6.1a ± 0.5	0.59a ± 0.1	0.82a ± 0.5	1.14b ± 0.1
CV (%)	1.0	8.2	6.26	4.11	9.57	8.88	3.65	27.32	11.64	88.73	52.28	15.42
P ≤ .05	0.84	0.0006	0.0001	0.01	0.01	0.68	0.0001	0.05	0.0003	0.005	0.001	0.0001
SE	0.6	0.75	0.89	0.59	0.59	0.002	0.94	0.13	0.77	0.64	0.71	0.85
LSD	NS	2.05	8.09	0.08	13.44	1.49	0.64	4.71	1.06	0.29	0.51	0.34

Notes: data are represented as means of four replications. Values after means are standard deviations. Means within columns in the 0–15 or 15–30 cm depths followed by similar lowercase letters are significantly different at P ≤ .05 using Least Significant Difference (LSD). SE: standard error. CV: coefficient of variability. C: lysimeters in the control without gypsum and P. glandulosa; G: lysimeters with gypsum and without P. glandulosa; IN: P. glandulosa in lysimeters relying on IN; NFIX: P. glandulosa in lysimeters relying on nitrogen fixation.

Discussion

Plant parameters

Our data presented here aimed at answering a question if P. glandulosa could be used in reclaiming calcareous saline-sodic soils. The capacity of the plant to fix N₂ and release H⁺ ion involved in solubilization of inherent CaCO₃ is an important principle. Our data illustrated in Figure 1 showed that straw and root dry matter and root N content of P. glandulosa relying on NFIX are higher than that found in lysimeters with P. glandulosa relying on IN. These could possibly be attributed to tolerance of the shrub to salinity stress. These results corroborated findings of previous workers who reported high salinity tolerance for various Prosopis sp. (e.g. Ahmad et al. 1994; Baker et al. 1995; Cony & Trione 1998). Our results suggest that salinity and sodicity stresses could be augmented by supplementing the plants with N₂-fixing bacteria. Accordingly, these findings may indicate that P. glandulosa tree could be a (1) strong alternative for rehabilitation of saline-sodic soils, (2) stabilizing active sands and (3) providing both fuel wood and browse for livestock in drylands ecosystems. Our results verified that of Geesing et al. (2000), who indicated that the tree is an actively fixing N₂ and this suggests that more attention should be given to the inclusion of tree legumes in arid ecosystems. However, it should be carefully noted that some Prosopis species are sensitive to Na₂SO₄ compared to NaCl when EC was more than 5 dS m⁻¹ (Gómez et al. 2008). In this work, NaCl was the main source of salt used to develop salt stress. The question how Prosopis continued to fix N₂ under salt-stress conditions may be explained by the work of Kurdali et al. (2003), who reported 18% higher NFIX by dhaincha (Sesbania aculeate Pers.) under saline than under non-saline soil conditions, and attributed the plant’s fixing ability to the presence of effective salt-tolerant indigenous rhizobial strains and a lower N₂ content of the soil.

The results presented in Table 2 showed that the content of Na⁺ in the shoot of plants relying on NFIX was almost half that in lysimeters with plants relying solely on IN. This low Na⁺ accumulation indicates that Prosopis under saline-sodic conditions has an effective mechanism of excluding the element from the metabolically sensitive parts such as leaves. In contrast, the roots in the non-fixing N₂ Prosopis transferred the Na⁺ to the shoot. Therefore, when plants were forced to fix N₂ under salt-affected soil, the element tend to be retained in the roots of the N₂ fixing plants. Moreover, results of our work also shown in Table 2 indicated that the plants have accumulated Na⁺ in the roots and were able to regulate its transfer to the shoot. That our findings supported other results showed that the Na⁺ in leaves of Prosopis cineraria grown under saline conditions (EC of up to 13.3 dS m⁻¹) exhibited a gradual increase while rapidly increased in roots in response to salinization of soil (Ramoliya et al. 2006). Because high concentration of Na⁺ may be detrimental to root tissues, exclusion of the element might happen during the turnover of fine roots or stem tissue may act as a barrier for translocation. It would be worthwhile to state that these results may indicate the priming effects of the rhizobium in increasing salt tolerance through element exclusion. Our results supported the earlier findings by Misra (2000), who reported that Na⁺ in P. juliflora was confined to older storage components such as stem and branches, suggesting suitability for sodic soil amelioration and that the tree root systems contribute significantly to sodicity alleviation. The use of N₂-fixing legumes in amelioration of calcareous (12.8% CaCO₃) salt-affected (EC of 9.72 dS m⁻¹) soils through exclusion of Na⁺ have been recently assessed by Cao et al. (2012), who found 28.6% reduction of the element in the shoot of N₂-fixing alfalfa. The exclusion of Na⁺ shown in Table 2 of this study is consistent with the fact that ion-specific toxicity eventuates when individual salt ions’ concentration in the soil is in high quantity and plants store them beyond their holding capacity (Ferguson & Grattan 2005). For example, high levels of Na⁺ ions tend to disrupt the uptake of K⁺ ion transporters and thereby resulting in ion-specific toxicity damages photosynthesis (Dubery 2005). Although, in this study, K⁺ content in the shoot of P. glandulosa (Table 2) was similar in N₂-fixing and IN treatments, roots of the latter contained less K⁺ concentration. This might indicate that if the salinity stress continued in the absence of plant/microbial mutual benefits, disruption of K⁺ transfer to shoot would be expected. As presented in Table 2, our study revealed that the accumulation of K⁺ in the shoot and root was greater than that of Na⁺. This trend has been reflected in Figure 1, which showed that the ratio between K⁺ and Na⁺ in both the shoot and root were more than 1. In the roots of both N₂ fixing and non-fixing, this ratio was similar, whereas in the shoot of the former, the ratio was almost double the non-N₂-fixing (Figure 1). Additionally, the content of K⁺ in the straw of P. glandulosa presented in Table 2 showed no significant difference between plants in lysimeters relying on either NFIX or IN. This explained that there was no decline in straw K⁺, suggesting that salt stress is not limiting the growth performance of Prosopis (Erdei & Teleisnik 1993). Despite the high concentration of Na⁺ in the soil solution that might compete with K⁺, our results revealed that the plant is tolerant to salinity stress at the EC level of the study with further positive priming effects of N₂-fixing bacteria. These were recently proved by Bhuiyan et al. (2015), who classified some grasses as salt tolerant when the K⁺/Na⁺ ratio was reported to be more than 1. Maintaining high cytosolic K⁺/Na⁺ ratio as a critical indicator of salt tolerance capacity of plants is well documented (Maathuis & Amtmann 1999; Flowers 2004; Akram et al. 2007). The efficiency of the plant in translocating accumulated elements from roots to shoots measured as TF is a direct estimate of the efficiency of phytoremediation. The effects of salt stress on TF values of Na⁺, K⁺, Ca²⁺ and Mg²⁺ observed in Table 3 indicated that supplying N₂-fixing rhizobium has greatly reduced the translocation of Na⁺ ion to the leaves and thereby increasing the K⁺/Na⁺ ratio in the leaves and improving salt tolerance stress. However, in absence of the N₂-fixing bacteria, P. glandulosa tended to accumulate more Na⁺ ions, thereby may relatively reduce stress tolerance. The similar TF values of Ca²⁺ and Mg²⁺ (Table 3) were consistent with similar content of Ca²⁺ and Mg²⁺ in the straw of mycorrhizal and non-mycorrhizal P. glandulosa presented in Table 2. Our study also observed that the content of Ca²⁺ in the straw of both P. glandulosa was almost threefolds that of Mg²⁺. These findings corroborated of Ramyoliya et al. (2006), who observed that under saline conditions, content of Mg²⁺ in the leaves of P. cineraria was lower than that of both Ca²⁺ and Mg²⁺. Similarly, our results supported those obtained by Scambato et al. (2011), who suggested that mycorrhizal colonization turned Prosopis alba more tolerant to salinity by maintaining the net photosynthesis and controlling transpiration rate.

Soil parameters

The effects of NFIX, IN, application of gypsum on soil chemical properties after harvest of the P. glandulosa shown in Table 4 revealed that there is a reduction in pH found in lysimeters with Prosopis relying on NFIX. This may possibly be attributed to the extra H⁺ proton released during NFIX. The CEC of the soil used in this study is only 16.3 cmol_c kg⁻¹ (Table 1), which indicates a low buffering capacity to resist changes in pH. This result was similar to Schubert et al. (1990), who found, in a laboratory experiment, acidification of soils with low CEC (from 7.11 to 20.88 cmol_c kg⁻¹) under N₂-fixing filed bean (Vicia faba). Significant reduction (0.15 units) in pH was also determined under field conditions of 4 years cultivation of N₂ fixing alfalfa in salt-affected soils (Cao et al. 2012). Our findings supported earlier suggestions that that the acidified rhizosphere with legumes resulted in a considerable extrusion of H⁺ to reduce soil pH (Tang et al. 1999). The present study showed reduction in total soluble salts obtained in Prosopis treatments (Table 4). In both lysimeters planted with Prosopis, the reduction in salinity was almost similar to that of the gypsum treatment. This suggests that Prosopis could have considerable potentials for bioremediation of salt-affected soils. The results of the present study supported previous finding that this tree can be used in bioremediation/ phytoremediation of salt-affected soils (e.g. Bhojvaid & Timmer 1998; Arya 2003). The content of soluble Na⁺ in the gypsum and Prosopis treatments presented earlier in Table 4 was greater than that in the control; this shows that Na⁺ adsorbed to the exchange complex was replaced and injected into the soil solution. This could possibly be explained by high available Ca²⁺ and Mg²⁺ released either directly from the gypsum or through solubilization in non-gypsum lysimeters. Normally, reclamation of such soils based on phytoremediation or organic materials involves enhancement of the ability of Ca²⁺ and Mg²⁺ to replace Na⁺ from the exchangeable complex (Benbouali et al. 2013). Under such conditions, microbial respiration is generally high, resulting in better water aggregate stability and improvement of hydraulic conductivity that enhance removal of salts (Yazdanpanah et al. 2016). In this study, high levels of K⁺ in the soil solution and straw (Table 2) found in the IN treatment is consistent with earlier findings by Fox and Guerinot (1998), who reported that high Na⁺ normally depresses the uptake of K⁺. The low SAR values in the 15–30 cm depth found in the gypsum and Prosopis treatments of this work may further indicate the potential capacity of the plant in amelioration of sodic soils. Due to tolerance and adaptability of Prosopis affinis and Prosopis pallid, Goel and Behl (1998) have recommended Prosopis for afforestation of sodic soils. However, less considerable changes in the SAR of the top 0–15 cm depth in Prosopis lysimeters compared to the gypsum treatment suggests that longer time may be needed to induce significant changes. In elsewhere, reduction in Na⁺ toxicity was observed in a field experiment with Prosopis after 5–7 years when there was build up of macro-elements (Bhojvaid & Timmer 1998).

The ionic composition of the saturated soil paste of this work revealed that the CI⁻¹ and were the dominant anions in both depths. The use of gypsum in our study has increased the level of soluble in the soil to extremely high levels. Sulfate ion is expected to be released in the soil solution during incorporation of gypsum (CaSO₄·2H₂O). The use of gypsum in reclamation of saline-sodic soils is well documented (e.g. Abdel-Fattah et al. 2015). As shown in Table 4 of this study, the concentration of in the top 0–15 cm depth of lysimeters with Prosopis was similar. This similarity could be explained in two possible scenarios: (1) few amount of fine roots that have been reflected in production of low concentration of H⁺ ions to induce significant solubilization of CaCO₃ or (2) the produced as the results of solubilization have been leached down to the 15–30 cm depth. We speculate that the latter scenario was the main reason as the level of in the lower depth in lysimeters with NFIX was 21% higher (P ≤ .05) than that in lysimeters with IN (Table 4). Therefore, the present work revealed that roots of both Prosopis might have similar CO₂ released during respiration, enhancing the partial pressure of CO₂ (PCO₂) that could be involved in dissolution of calcite and thereby resulted in elevated concentration. These findings are similar to others reported elsewhere (Badía 2000; Qadir et al. 2005). As for the lower depth (15–30 cm) where treatments with Prosopis relying on NFIX showed higher levels of , this extra anion concentration could be due to both the influence of the PCO₂ and reaction of the H⁺ proton released during NFIX with calcite and thereby releasing additional content. The content of Ca²⁺ reported in this study indicates that solubilization by the gypsum was more than fourfolds that obtained in both Prosopis treatments. Our results may suggest that the increase in Ca²⁺ and concentrations in the soil solution of Prosopis treatments shown in Table 4 could be due to either dissolution of the calcite as a source of Ca²⁺ potentially enhanced by releases of H⁺ proton or enhanced root respiration that releases CO₂ in the rhizosphere as a result of exposure of Prosopis to salinity, thereby increasing demand for energy (Lambers et al. 1998). Higher levels of Ca²⁺ and levels during phytoremediation of calcareous saline soils using N₂ fixing legumes were earlier reported (Qadir et al. 2003a, 2003b; Mubarak & Nortcliff 2010; Cao et al. 2012).

In conclusion, this study showed that both plants retained similar Ca²⁺, and the Na⁺ level in the straw of non-fixing plants was 1.8 folds that of N² fixing plants, whereas levels of Na⁺ and K⁺ in roots of N² fixing were 13% and 24% higher than that of non-fixing plants. Based on the findings reported here, it can be inferred that (1) N₂ fixing and non-fixing plants released 39% and 50% Na⁺ of that released by the gypsum and (2) both Prosopis produced higher than the gypsum, which induced extra Ca²⁺ solubilization of the gypsum, growing P. glandulosa on calcareous salt-affected soils could be an alternative management practice to reduce salinity and sodicity effects.

Notes on contributors

Mubarak Abdelrahman Abdalla is a Professor of Soil Science at the Department of Soil and Environment Sciences, Faculty of Agriculture, University of Khartoum, Shambat, Sudan. He has written more than 50 articles in the field of land degradation and development.

Abdelkarim Hassan Awad Elkarim is an Associate Professor of Soil Science at the Department of Soil and Environment Sciences, Faculty of Agriculture, University of Khartoum, Shambat, Sudan. He has written more than 20 articles in the field of Soil Chemistry.

Takeshi Taniguchi is an Associate Professor at the Arid Land Research Center, Tottori University. He published many articles in the science of ecology of plant symbiotic microorganisms in drylands. He concentrated in search for effective plants symbiotic microorganisms (mycorrhizal fungi, endophytic fungi, and rhizosphere bacteria) to enhance drought and salt stress tolerance of plants, analysis of the ecological and functional features of symbiotic microorganisms and use of plant symbiotic microorganisms for ecosystem restoration and agricultural production in drylands.

Tsuneyoshi Endo is a Professor in the Faculty of Agriculture, Tottori University, Tottori, Japan. He has excellent achievements in land use effects on soil quality.

Norikazu Yamanaka is a Professor in the Arid Land Research Center, Tottori University. He is specialized in revegetation science: plant ecology and ecosystem restoration in arid lands. Currently he is the Director of the center. He has ample publications and text books in ecological studies of plant communities in arid areas, ecophysiological studies on drought and salt tolerance of woody plants and studies on the ecosystem restoration in arid areas.

Acknowledgements

This study was supported by the Arid Land Research Center, Tottori University, Japan. We are grateful to members of the Division of Environmental Conservation, ALRC, Tottori University for helping us during preparation of soil samples and University of Khartoum, Sudan for providing sabbatical leave.

Disclosure statement

No potential conflict of interest was reported by the authors.

Additional information

Funding

This work was supported by Arid Land Research Center, Tottori University, Japan.