Salinity Temporal Evolution Assessment under Mediterranean Conditions

Soil salinity is recognized as an important indicator of soil quality (Brady and Will 2008; Litalien and Zeep, 2020). It largely dominates land use strategies, directly affecting decision-making for effective land management (Eswaran et al., 2019; Taghizadeh-Mehrjardi et al., 2021). Salinization is a common pedogenetic process in Algerian soils (Daoud, 1993). Moreover, soil salinization is an important process, especially in arid and semi-arid areas (Saidi et al., 2004), which leads to the degradation of physical, chemical and biological properties of soils. The consequences of this degradation are the reduction of soil fertility, which leads to a reduction in crop yields, and sometimes the disappearance of plant cover.
       
Moreover, according to Abbas et al., (2011) found that globally, soil salinization is spreading at a rate of up to 2 million hectares per year. Vast in area, approximately 77 million hectares of land are salinized due to human activity, of which 58% is irrigated (Metternicht and Zinck, 2003).
       
However, the Solonchaks of Rélizane have a relatively higher degree of belonging with the Calcisols than with the Gypsisols whatever the season and the year considered (Hadj-Mloud and Djili, 2021; Hadj-Miloud et al., 2018).
       
Algeria with more than 20% of irrigated soils affected by salinization (Douaoui and Hartani, 2007), is among the countries most threatened by this problem. This requires careful monitoring of the state and temporal variation of soil salinity to curb the trends of this problem and ensure sustainable land management.
       
This work focuses on the study of the temporal evolution of the salinity state of the Solonchaks located in the Rélizane region. It is therefore:
- Characterize the soil solution, define the saline profiles and determine the chemical facies of the soil solutions.
- Determine temporal changes in salinity between 2010 and 2014, comparing saline profiles to determine the direction and intensity of soil salinity variation.

Study area
 
The soils studied are located in the Relizane region, is known for its two large irrigated perimeters, the perimeter of La Mina and that of Bas-Chéliff.
       
It is in this context that we have studied three soil profiles (P1 and P2) which are present in the perimeter of the Mina and another profile (P3) which is located in the perimeter of Bas-Chélif, more precisely in the research station at Hmadna (Fig 1).
 

       
The geographical coordinates of the studied profiles are presented in the Table 1.
 

       
The climate is semi-arid (253 mm rain/year), temperate in winter with very strong potential evapotranspiration (1500 mm/year).
       
The choice of profiles is based on the diagnostic criteria of Solonchaks according to IUSS Working Group WRB (2014).
       
This study was conducted at the Ecole Nationale Supérieure Agronomique, El-Harrach, Algeria. January, 2021.
 
Methodology
 
The first step is to characterize the soil solution, to define the salt distribution profiles in the soil, and to determine the chemical facies of the soil solutions. Then, it will be a question of studying the temporal evolution of salinity between the wet season of 2010 and 2014 in the irrigated perimeters of Bas-Chéliff and Mina. This monitoring is carried out on three reference profiles of Solonchaks identified by HadjMiloud (2010). To do this, we calculated the rate of salinization by profile. The average EC of the profile is calculated according to the following formula:
          
The difference in the average salinity of the profiles is obtained by the difference between the values   of the EC for the wet season of the 2014 campaign and that of the EC for the 2010 campaign [EC (2014)-EC (2010)]. The salinization rate is obtained by dividing the difference by time (4 years).
       
In order to better analyze the data, we used PCA (principal component analysis). This analysis allows us to determine the anions and cations of the soil solution that have the most influence on the variation in salinity.
       
Subsequently, we performed non-parametric tests (Wilcoxon test). This test reveals whether the differences between the EC values of the 2014 campaign and that of 2010 are significant or not. 

The soil solution (campaign 2010)
 
The descriptive statistics of the examined parameters of profile 1 are presented in Table 2.
 

 
Electrical conductivity
 
Table 2 and Fig 2 show that the surface horizon (Anz) is characterized by high salinity (43.8 dS m^-1) (USSL, 1954). However, this EC is relatively weak compared to the other horizons of the profile. This phenomenon may be due to rain-induced salt leaching, which was procured in the subsurface horizon, leading to an increase in EC (1.6 dS m^-1). On the other hand, we note an increase in the salinity of the two horizons BCnz and Cnz1, reaching values   of 61.6 and 57.5 dS m^-1 respectively.
 

 
pH
 
Table 2 indicates a pH value greater than 8 with an average value which is around 8.2. These values show a relatively alkaline soil reaction, the pH is homogeneous in the profile.
 
Cations
 
Among all the cations in the soil solution (Fig 3), the Na⁺ cation is the most represented with an average rate of 48% and a concentration which varies between 212.7 meq l^-1 and 370.78 meq l^-1.
 

 
Anions
 
The main ion in the soil solution is Cl^- (Fig 4). It represents on average 93% of the total amount of anions with values   that vary between 405 meq l^-1 and 630 meq l^-1.
 

 
Profile 2
 
The descriptive statistics of the studied parameters of profile 2 are indicated in Table 3.
 

 
Electrical conductivity
 
The extreme values of salinity in profile 2 vary between 10.7 dS m^-1 and 56.8 dS m^-1 with an average of 32.46 dS m^-1.                        
These values demonstrate that profile 2 is highly salty. Salinity increases steadily from the surface horizon to the depth horizon, reaching values of 10.7 and 56.5 dS m^-1 respectively (Fig 5). The result indicates that the salt profile is descending.
 

 
pH
 
The pH values   vary between 7.13 and 7.8 with an average of 7.43 (Table 3). These values correspond to a relatively alkaline soil reaction.
 
Cations
 
With concentrations varying between 250 and 950 meq l^-1, Na⁺ occupies on average 46% of the cations in the soil solution (Fig 6).
 

 
Anions
 
The predominant ion in the soil solution is Cl^-. It represents on an average 58% of the total amount of anions with values   that vary between 60 meq l^-1 and 540 meq l^-1  (Fig 7).
 

 
Profile 3
 
The descriptive statistics of the examined parameters of profile 1 are presented in Table 4.
 

 
Electrical conductivity
 
Table 4 and Fig 8 show that the surface horizon has a low salinity compared to the other horizons in the profile (EC= 7.21 dS m^-1). The salinity then increases at the level of the Bgnz1, Bgnz2 and Bgnz3 horizons to reach values   which are respectively 22.4 dS m^-1, 25.8 dS m^-1 and 35 dS m^-1 (Fig 9). The extreme salinity values in profile 3 vary between 7.21 dS m^-1 and 35 dS m^-1 with an average of 24.22 dS m^-1 (Table 4).
 



 
pH
 
The pH values  vary between 7.06 and 7.75 with an average of 7.27 (Table 4). These values indicate a relatively alkaline soil reaction. The pH variation in the profile remains low with regard to the coefficient of variation (CV=3.83%).
 
Cations
 
Among all the cations in the soil solution (Table 4 and Fig 9), the Na+ cation is the most abundant with an average rate of 74%. Ca++comes in second position with a rate of 17% and contents that vary between 25.67 meq l^-1 and 48.27 meq l^-1.
 
Anions
 
The predominant ion in the soil solution is Cl^- (Table 5 and Fig 10). It represents on average 82% of the total amount of anions with values varying between 60 meq l^-1 and 320 meq l^-1.
 

 

 
The chemical facies
 
The Piper diagram (Fig 11) showed that the majority of the chemical facies of the compound soil solutions is of the sodium chloride type. However, we note the presence of the following chemical facies: the magnesian chloride type, chloride without dominance of one of the cations and the sodium hyperchloride facies.
 

       
These chemical facies favor that the soils consumed evolve according to the neutral saline route with a dominant chloride content (Marlet and Job, 2006).
 
Temporal evolution of salinity
 
The study of the temporal evolution of the salinity studied between two companions consists in comparing the saline profiles of the year 2010 and 2014.
 
Comparison of saline profiles between the two companions
 
The comparative study of salt distribution profiles between the two companions focused on 3 profiles (P1, P2, P3). These profiles present complete data for all horizons concerning the 2010 campaign. The comparison of data from 2010 and that of 2014 will make it possible to identify the temporal evolution of salinity over a period of 4 years. Comparison of salt profiles reveals that salinity decreased in all profiles in 2014 compared to 2010 in the three profiles 3 (Fig 12, 13 and 14). Only the surface horizon of profile 3 underwent a slight increase in surface salinity during the year 2014 (EC=3.96 dS m^-1) (Fig 14).
 

 

 

 
Evolution gradient of salinity
 
The comparison of the salinity levels between the two companies will make it possible to determine the direction of the evolution of the salinity of each horizon of the soil.
 
Evolution gradient of salinity per horizon
 
Table 5 shows that globally all the horizons underwent desalination in 2014, with a speed between 1.27 and 33.07 dS m^-1 per year. Only the surface horizon (H1) of profile 3 which recorded salinization with a rate of 0.99 dS m^-1 per year.
       
Furthermore, the surface horizons of profiles 1 and 2 show desalination which occurred at a rate of 9.14 and 33.07 dS m^-1 per year respectively.
       
The subsurface horizons (H2) showed desalination at a rate varying between 1.52 and 12.42 dS m^-1 per year for the three profiles.
       
The depth horizons (H3, H4 and H5) show desalination with a rate varying from 1.27 to 10.87 dS m^-1 per year for the three profiles.
 
Evolution gradient of salinity by profile
 
Table 6 shows that the three profiles underwent desalination between 2010 and 2014. This desalination occurred at a rate between 3.67 and 10.74 dS m^-1 per year.
 

       
In addition, we performed a non-parametric test (Wilcoxon test). This test reveals whether the differences between the EC values of the 2014 campaign and that of 2010 are significant or not. The results of the test (Table 7) reveal that at the significance level α=0.05 (p-value < 0.05), we can reject the null hypothesis, of the absence of significant difference between EC of the samples studied. So the difference between the EC of the 2010 campaign and that of 2014 is significant.
 

 
Multidimensional analysis of the chemical composition of the soil solution
 
We applied the principal component analysis (PCA), in order to know the elements of the soil solution which have the most weight on the variation of the EC.
 
EC PCA and soil solution ions
 
From Fig 15, the F1 axis extracts 59.58% of the inertia from the point cloud. On the F2 axis, the residual inertia that it extracts is 20.43%. Most of the information is represented by the F1 and F2 axis.
 

       
The first axis is formed by the contribution of the following variables: Cl- (r = 0.98), SO4-- (r = 0.79), EC (r = 0.99), Na⁺ (r = 0.96), Mg++ (r = 0.96) and Ca++ (r = 0.26) (Table 8). However, the F2 axis is formed by the contribution of the variables: Ca⁺⁺ (r =0.79) and K+ (r = -0.65). However, the elements of the solution which contribute the most in the formation of the axes are Cl^-, Na⁺, Mg⁺⁺ and SO₄^--. This is explained by the fact that these elements come from more soluble species compared to the other elements (Hadj-Miloud, 2019).
 

       
The evaporated climate of this region (ETP 1500 mm/year), associated with a water deficit, combined with a relatively fine texture, reduces the capillary rise of water from the salty aquifer. He accentuated the salinity of Solonchaks of the Relizane region.
       
However, the chemical composition of the soil solution is similar to the two campaigns, for the cations a classification order of Na+> Ca⁺⁺> Mg⁺⁺>K⁺ type and for the chloride-dominated anions (Cl^-> SO₄^{- -}> HCO₃^-). This Mg⁺⁺ is well established for its profiles 1 and 2 of the 2010 campaign, since the results come from the same report as Hadj Miloud (2010). Hadj Miloud (2019) and Hadj-Miloud (2023).
       
The saline profile is of the convex type for profile 3 (companion 2010 and 2014). The accumulation of salts at the level of the middle part is due to leaching of salts from the surface horizon.
       
The study of the temporal evolution of salinity showed that the majority of horizons underwent desalination between 2010 and 2014 campaign. This desalination occurred at a speed between 3.67 and 10.74 dS m^-1 per year. This would be due to a partial leaching of salts from the surface horizons during the rainy period as shown by Servant (1975). The maximum salinity is observed at the level of middle part of profiles 1 and 3. As a result, the saline profile is of the convex type, for profile 2 the salts have accumulated in the lower part of the profile and consequently the profile saline is descending type.
       
Furthermore, the difference in the variation of salinization rate between the profiles is controlled by the grain size, the morphology (shrinkage cracks) and the presence or absence of a shallow and highly mineralized groundwater table (Hadj-Miloud, 2019).
       
Principal component analysis (PCA) showed that the chemical elements of the soil solution that had the most influence on the variability of salinity were Na⁺, Mg⁺⁺, Cl^- and SO₄^--. We can explain the results of this analysis by the fact that Na⁺ does not enter into the phenomenon of mineral precipitation at very high EC values (Droubi et al., 1980). The evolution of magnesium can be controlled by the precipitation of magnesite. In addition, the results obtained are important for the evaluation of agricultural land (Hadj-Miloud, 2022).

The main results obtained show that the three profiles presented are mainly characterized by very high salinity levels [3.5 < EC (dS m^-1) < 143]. High sodium concentrations in the soil solution have a direct effect on high ESP values   (8.1< ESP (%) < 36.6). Only the surface horizon of profile 3 which is weakly saline (EC = 3.65 dS m^-1) and non-sodium (ESP = 12.2%) (Hadj-Miloud et al., 2022).
       
The extreme salinity value for profile 2 (companion 2010) is visible at the level of the surface horizon (143 dS m^-1), the distribution of salts for companion 2014 is of convex type for profile 1 and 3. On the other hand, the salinity gradient is descending for profile 2.
       
The cationic and anionic composition of the soil solution is strongly dominated by sodium and chlorides for all the profiles. This explains the dominance of the chloride-sodium chemical facies.
       
Finally, the study of the water balance at ground level can further explain the dynamics of salts in the soil.