Boron Characterization , Distribution in Particle-Size Fractions , and Its Adsorption-Desorption Process in a Semiarid Tunisian Soil

Boron is essential for crop growth but needed in very small amounts.(e range between boron deficiency and toxicity for plants is quite narrow. (ese stress conditions gravely reduce yield and quality of many crop species. (erefore, understanding the factors and the reactions affecting boron availability in soil is necessary. Against this framework, our research aims to determine the available boron status in a semiarid soil of Dour Ismail irrigated perimeter (North Tunisia). (e objectives are also to investigate boron distribution in different particle-size fractions throughout the soil profile and to understand boron adsorption-desorption mechanisms according to some soil properties. For this purpose, one soil profile was dug in the field plot that had not received any previous boron fertilization. Soil samples were gathered from the different horizons of the profile and analyzed for the main physicochemical properties. Our results showed that the studied soil is Stagnic Fluvisol (clayic). (e highest boron amounts were recorded in deep horizons and were greatly affected by soil salinity, organic matter, and clay contents. However, the increase in the pH level and the high percentage of total lime significantly diminished the available boron amounts in surface layers. (e investigation of depth boron distribution in the different particle-size fractions indicated a considerable contribution of the silt (2–50 μm) fraction (52% of the soil total available boron), while the coarse (>50 μm) and clay (<2 μm) fractions seem to play a less important role. (e adsorption data were fitted to Freundlich adsorption isotherm. It revealed that adsorption of boron increased with the increase of boron concentrations in soil solution. Desorption isotherm denotes that the accumulated boron in soil was not easily released. Adsorption and desorption of boron in soil were greatly affected by soil properties, such as pH, salinity, sand content, clay content, total organic carbon, total nitrogen, and cation exchange capacity.


Introduction
Boron (B) is a necessary micronutrient for the growth of all crops [1].It plays a key role in a diverse range of plant functions including cell division and sugar transport.B is also required for the production of nucleic acids and the development of reproductive structures [2].In soil, B content evolves and must be monitored frequently because of the narrow range between the levels of B deficiency and toxicity as compared to other nutrients [3].erefore, the management of this nutrient is critical.Indeed, appropriate B supply is necessary for obtaining better yields.
In soil, B exists in four major forms: (a) in rocks and minerals; (b) on clay surfaces and iron and aluminium oxides; (c) combined with organic matter (OM); and (d) as boric acid and B in the soil solution [4].Nevertheless, only a few forms are available to plants and their determination is important for delineating the deficient soils.B deficiency is prevalent in coarse-textured acidic soils developed mainly in high rainfall zones.Moreover, soils with low OM content and high adsorption capacity (e.g., soils with high soil pH level and rich in clay minerals) are generally B deficient [5].On the contrary, B toxicity symptoms are often observed in arid and semiarid regions owing to the use of irrigation waters high in B.
Recently, several adverse environmental effects related to B have been traced back to poor manure management practices and inappropriate B fertilizer applications.For example, when humans and animals consume large amounts of B-containing food or drink water, the B concentrations in their bodies may rise to levels that can cause health problems.For plants, B toxicity symptoms are different and they depend on plant species.Stunted growth is common, and fruit trees may be less productive.erefore, in order to recommend the correct rate of B application and to assess the agricultural and environmental B impacts, it is necessary to investigate B distribution in soil and to understand its adsorption-desorption behavior.
In soil, the mobility, transport, and partitioning of B depend on various soil chemical properties, namely, pH level, salinity, and the contents of clay minerals, sesquioxides, carbonates, and OM [5,6].B is retained by adsorption onto soil mineral and humic particles and by forming insoluble precipitates.B adsorption was found to increase gradually in the soil with increasing pH, to reach a maximum around pH 9.0 and to decrease with further increase in pH.Above pH 9, B adsorption was increased from the solution of higher electrical conductivity (EC) [7].Mahler [8] claimed that coarse-textured soils with limited OM content are often low in B available for plants.Subsequently, soils rich in OM and clay contents adsorb a greater B amount [9].e release of sorbed B is of immense significance, as plants respond primarily to the B activity in the soil solution.erefore, adsorption and desorption reactions are among the major factors that regulate B concentration in soil solution and ultimately availability to plants.
Currently, the study of B occurrence and distribution in soils through a multitude of approaches (B characterization, B fractionation, extraction techniques and methodologies, etc.) has been the major focus of agronomists and soil scientists.Consequently, innumerable researchers have studied B dynamic in soil for a long time [10][11][12].Most of these studies have focused on B in soil surface horizons with little attention given to B distribution according to depth.In reality, few researchers investigated this issue in Mediterranean semiarid regions like Tunisia [13,14].us, there is a pressing need for fundamental understanding of B distribution in such areas.
Against this backdrop, the present research was undertaken to better determine the available B content in the Tunisian soil of Dour Ismail irrigated perimeter (North Tunisia), as well as to investigate its depth distribution in the different particle-size fractions.e current study also aims to understand B adsorption-desorption mechanisms with different soil properties.

Soil Sampling and Physicochemical Analysis.
To achieve the objectives of the present study, one pit was dug in Dour Ismail irrigated perimeter situated in the Northwest of Tunisia (Béja Governorate-Goubellat delegation) (Figure 1).
e main geomorphological, bioclimatic, and pedological characteristics of the studied area are detailed in [13].ree soil samples were gathered from each horizons of the profile (the four horizons were considered, see Table 1), air-dried, sieved through a 2 mm diameter mesh, and stored in sealed containers at 5 °C.en, 12 soil samples were analyzed for the main physicochemical properties according to international standards.e soil pH level was measured with a pH meter in water (pH in H 2 O) and in 0.01 M calcium chloride solution (pH in CaCl 2 ) (1 : 5 soil : reagent ratio) (ISO 10390).e salinity was measured as electrical conductivity (EC) with an EC meter in the soil aqueous extract (1 : 5 soil : water ratio) (ISO 11265).Soil texture was determined using the Robinson pipette method after oxidation of the OM with H 2 O 2 and clay dispersion by hexametaphosphate solution (NF X31-107).Total calcium carbonate (TCaCO 3 ) was measured using the volumetric method (NF ISO 10693).Soil samples were also characterized for total organic carbon (TOC) by the sulfochromic oxidation method (ISO 14235) and total nitrogen (TN) by the modified Kjeldahl method (NF ISO 11261).e cation exchange capacity (CEC) was obtained with ammonium acetate following the Metson method (NF X31-130).Finally, the available B was extracted in boiling water and measured by the ICP-AES analytical method (inductively coupled plasma atomic emission spectroscopy) (NF X31-122).
ree replications were performed for each executed analysis and were expressed as an average of the three values.Moreover, we submitted each soil samples to a particle-size fractionation to isolate coarse fraction (>50 µm), silt fraction (2-50 µm), and clay fraction (<2 µm).All separated soil fractions were quantified for their available B contents.B adsorption-desorption reactions were also studied in each horizon.

Particle-Size
Fractionation.Forty grams of soil was dispersed in 200 mL of distilled water with 0.5 g of sodium hexametaphosphate and shaken for 16 h with five agate balls.
e suspension was then filtered through a 50 µm filter under a distilled water jet.e fraction >50 µm was recovered by sieving.e suspended material passing the finer sieve (fraction <50 µm) was further submitted to 10 min ultrasonic treatment in the continuous mode to ensure soil particles dispersion.
e used sonicator is a heat system ultrasonic model W225R (20 kHz frequency, 100 W power). e fraction 20-50 µm was also obtained by sieving (20 µm filter).However, the fractions 2-20 µm and <2 µm were obtained by four sedimentation/decantation and centrifugation cycles.Each fraction was later dried in an oven under 40 °C to reach a constant weight.e different fractions are finely weighed and crushed to be analyzed for available B. In order to simplify the comparison, our results were expressed for the total sand fraction (>50 µm), total silt fraction (2-50 µm), and clay fraction (<2 µm).

B Adsorption-Desorption Studies.
For B adsorption studies, 10 g of soil samples were placed in 50 mL polypropylene centrifuge tubes with 20 mL of 0.01 M CaCl 2 solution and contained di erent quantities of B (5, 10, 20, 40, and 80 mg of B/L as H 3 BO 3 ).e contents were shaken for 24 hours.
e suspension was further centrifuged, and a 10 mL aliquot of the clear supernatant was removed and ltered.Subsequently, B was determined in the ltrate.Soil B adsorption was obtained by subtracting B in solution, after centrifugation and ltration from B added in the di erent concentrations.e following adsorption equation was used to de ne B adsorption in the soils: where x/m the amount of adsorbed B (mg/kg), c the equilibrium B concentration (mg/L), K Freundlich constant, estimates the adsorption capacity (mg/kg), and 1/n Freundlich constant, estimates the adsorption intensity.B desorption was initiated with a removal of 10 mL of aliquot from equilibrated solution and an addition of 10 mL of B-free 0.01 M CaCl 2 solution.
e mixture was resuspended by agitation and equilibrated for 24 hours.e suspensions were centrifuged, and 10 mL of the supernatant was removed for B determination. e amount of desorbed B was calculated as the di erence between equilibrium B concentration at each desorption step and the equilibrium B concentration before each desorption step divided by 2 (dilution factor). is procedure was determined following a method that is described in [15].

Relationship between B Distribution and Soil
Characteristics.Table 1 shows the main physicochemical properties of the studied soil.e pH level was basic (≈8) and slightly dropped with depth.Similarly, the EC values increased as we moved from the surface (1.36 dS/m) to deep layers (2.08 dS/m).e soil texture was dominated by claysilt fractions (≈60%) throughout the pro le. e TCaCO 3 percents decreased with the increase in depth.e A p surface horizon of the studied soil exhibits the highest OM contents with 1.23% of TOC and 1.12% of TN. ese values decreased with depth and reached 0.61% and 0.60%, respectively.e CEC has the same behavior as the TOC and TN. e values varied from 30.67 meq/100 g top to 17.33 meq/100 g bottom.
e greater CEC values recorded at the surface might be due to the higher OM content.Mahajan et al. [16] and Kumari et al. [17] reported a similar relationship.e examination of B distribution in the studied soil showed an increase in the available B content as we moved from surface horizon A p (0.64 mg/kg) to the deep horizon C2 (0.74 mg/kg).e low B values recorded in surface layers may be explained by a slight  increase in the pH level (8.30) and a relatively high TCaCO 3 (13.16%)amount.Indeed, B adsorption in soil depends on the soil pH level.It rises between pH 6.0-8.5 and reaches the highest values around pH 9. B adsorption decreases with further pH increases [3].Furthermore, soils containing important amounts of lime are usually low in available B [18]. e calcium carbonate increases B adsorption by soils since it raises the soil solution pH.Several researchers observed a similar result and failed to get a significant correlation between TCaCO 3 and the available B content of the soil [18].e great B amounts recorded in deep layers correspond to a slight enrichment in TOC (0.61%) and mainly to a great percentage of clay-silt fractions (≈62%).Our results are in line with some earlier studies, which have shown that OM [4] and clay contents [19] affect the B bioavailability.Indeed, Raza et al. [4] and Chaudhary et al. [20] showed a positive correlation between soil OM content and available B. Later, Turan and Horuz [21] claimed that available B is freed as a result of the mineralization of OM caused by microorganisms.Moreover, our results clearly reveal that the clay-silt content (≈62%) in the studied soil seems to be among the determinative factors, which influence B bioavailability.Similar results have been reported in [19,22,23].e authors showed that fine-textured soils usually contain more available B than coarse-textured soils because of their greater content of clay minerals.e C 2 deep horizon of the studied soil showed high amount of available B (0.74 mg/kg) and was slightly saline (2.08 dS/m) as compared to the A p surface horizon (1.36 dS/ m).In fact, some investigators advocated that B adsorption was greater for the high EC solution [7,24].e authors pointed out that B adsorption, as a function of a B concentration solution is independent of salinity in the pH range 4 to 9 up to EC 7.9 dS/m.A statistical analysis was performed to find out possible correlations between the available B contents and the studied soil properties (pH, EC, clay, silt, sand, TCaCO 3 , TOC, TN, CEC, and C/N).ree values of each parameter were used for the correlation matrix.e results reflected a significant and positive-correlated relationship between the available B and the clay fraction (r 2 � 0.64 at p < 0.05).Kumari et al. [17] found a similar correlation.e authors revealed that finetextured soils usually contain more available B than coarsetextured soils.Besides, the EC showed a significant and positive correlation with the available B (r 2 � 0.53 at p < 0.05).Our result agrees with that of [25], which stated that the onset of salinization is accompanied by an increase in the available B content.However, our findings showed a negative relationship between the available B content and the pH level (r 2 � −0.79 at p < 0.05). is appears to correspond to the common belief that an increase in pH would increase B adsorption and consequently cause a reduction in B availability [3].
e same negative relationship was revealed between the available B and the TCaCO 3 content (r 2 � −0.90 at p < 0.05).Indeed, the calcium carbonate acts as an important B adsorbing surface in soil [18].

Depth B Distribution in Particle-Size Fractions.
e investigation of depth B distribution in the different particle-size fractions indicated a considerable contribution of the silt (2-50 µm) fraction to B adsorption in soil especially in depth.
is fraction contains 52% of the total available B (B 2-50μm / TB � 52%).However, the sandy fraction (B >50μm /TB � 26%) and the clay fraction (B <2μm /TB � 15%) seem to play a less important role (Figure 2).Our results do not offer support for the common theory that the finer-sized material and particularly clay minerals should contain more B than the coarser fractions.Indeed, the majority of previous studies [26] attested that the clay fraction has more contribution to the B content of soils.Kumari et al. [17] and Anitha et al. [27] revealed that all B fractions showed a positive correlation with clay content, namely, the readily soluble one (r 2 � 0.88 at p < 0.01 and r 2 � 0.474 at p < 0.05).However, few studies are in line with what we found [17].e authors showed that the silt fraction had a positive but nonsignificant relationship with the available B (r 2 � 0.21).
B partition in sand fraction decreased as we moved from the surface (B >50µm /TB � 32%) to deep horizons (B >50µm /TB � 26%) (Figure 2).In fact, sand content in the studied soil dropped with depth (Table 1).en, the great sand content recorded mainly in surface layers enhanced more content of extractable B. B partition in the clay fraction (B <2μm /TB) has the same behavior as the sand fraction.
e B values dropped from 25% top to 15% bottom although clay content increased in depth.However, some previous studies [28] revealed that the increase in clay content reduced the extractable B content because of the intensity of the adsorption force.Moreover, the lowest B values recorded in the clay fractions especially in depth correspond to significant OM levels (1.05%).Indeed, B adsorption in clay surfaces decreases as OM content increases.Soil OM occupies the reactive adsorption sites on clay fractions [29].

Adsorption of B.
e quantity of B adsorbed in each horizon was traced against the equilibrium concentrations in order to acquire the adsorption isotherm.We noticed that as the equilibrium of B concentration increased, the quantity of the adsorbed B increased in all horizons of the studied soil.However, at high B equilibrium concentration, soil samples show a remarkable difference in B adsorbed than at low B equilibrium concentration (Figure 3).Many researchers have also reported that the adsorption of B increased with the increase of B concentration in soil solution [30,31].e highest amount of B adsorption was recorded in the A/C horizon.Nevertheless, the lowest amount of B adsorption was recorded in A p surface horizon.Indeed, the A/C horizon had the highest pH (CaCl 2 ) value (7.75) of soil solution, while the A p horizon showed the lowest pH (CaCl 2 ) value (7.67).In addition, our results indicated that the B adsorption amount was positively correlated with the soil pH (CaCl 2 ) level (r 2 � 0.95 at p < 0.05) and clay content (r 2 � 0.76 at p < 0.05).Based on these findings, it could be concluded that B adsorption was mainly controlled by the soil pH level.Moreover, it greatly depends on the content of clay fraction of the soil.In general, soils with low clay content will adsorb less B than those with higher clay content [11].
B adsorption behavior in soil has been described using the Freundlich equation.is equation has been successfully veri ed in several soil types [11,19,23].
e adsorption parameters from the Freundlich equation are presented in Table 2. e results showed a wide dispersion of the adsorption capacity values (K) of the di erent horizons of the studied soil.Indeed, the A/C surface horizon showed the highest value (K 4.75 mg/kg), while the A p surface horizon recorded the lowest value (K 0.88 mg/kg).is trend was expected based on the di erent proprieties of the horizons, in particular, the increasing in clay contents and pH values.In this context, Alleoni and Camargo [10] have found that Freundlich constant K follows the tendency of high values in ne-textured soils.In fact, the maximum K value of the A/C horizon could be attributed to the great clay-silt amount (>62%), the high pH (CaCl 2 ) level, and the slightly enrichment in TOC (0.81%) and TN (1.40%).However, the lowest adsorption capacity of A p surface horizon was related to the great amount of sand content (41.46%) and the low pH (CaCl 2 ) level.In addition, the result of the statistical analysis showed a positive correlation between adsorption capacity (K) values and (1) the soil pH (CaCl 2 ) level (r 2 0.93 at p < 0.05), (2) the TN content (r 2 0.60 at p < 0.05), and (3) the clay-silt amount (r 2 0.48 at p < 0.05) and a negative relationship with the sand fraction (r 2 −0.48 at p < 0.05).In reality, it has often been shown that ne-textured soils adsorb more B than coarse-textured soils [22].Besides, our ndings showed that the quantity of OM, particularly TN, is considered as one of the main soil components that a ects B adsorption.Similar results have also been reported in [18,32].e authors showed a positive correlation between the content of soil OM and the amount of the adsorbed B.
e strong relationship between adsorption capacities and the increase in soil pH values (r 2 0.93 at p < 0.05) proves that the pH is one of the most important factors that enhance B adsorption by soil [32].In fact, B adsorption in soil was found to be increased from pH 3 to 8, showed a peak at pH 8 to 10, and declined with further increasing [31].
is positive relationship between soil pH level and B adsorption could possibly be elucidated as the abundance of borate ions to boric acid or by the rising number of adsorption sites at the high pH value [12].Based on our adsorption results, we can conclude that soil pH level, OM, and clay contents are the main chemical properties of soil that a ect the adsorption of B [32][33][34].

Desorption of B.
e knowledge of desorption reaction is interesting for determining if the adsorbed B of soil can be easily released.e amount of the desorbed B was expressed as percentage of the adsorbed B (Figure 4).e A p surface horizon had desorbed high percentage of B (62.2%), whereas the A/C horizon desorbed less B percentage (49.71%).Indeed, the A p surface horizon, which has low a nity for B adsorption, had released a great B amount.erefore, B adsorption and desorption process were oppositely associated.Our results are in line with [11,23], which have reported that soil having low B adsorption desorbs more B amounts.In addition, B desorption was higher in A p surface horizon with a great sand content.In this context, Arora and Chahal [35] observed that coarse-textured soils aimed to desorb large B amounts in comparison with ne-textured soils.e behavior of B desorption was described by the same Freundlich equation, as it has been reported by [15].e comparison of adsorbed and desorbed isotherm values of each horizon indicates two reaction types of B desorption.In fact, when isotherms of B adsorption are similar to desorption isotherms, it means that B adsorption is reversible.For instance, in the case of the A p surface horizon, isotherms of B adsorption (K 0.88) and desorption (K′ 0.84) presented very similar results.is horizon exhibited therefore a reversible B desorption reaction.Nevertheless, the nonconformity between adsorption and desorption isotherms indicates that B adsorption is hysteric and not easily reversible.In reality, all the studied horizons, except the A p surface horizon, exhibited a hysteric adsorption process.For the C 1 horizon, the B desorption isotherm (K′ 0.37) widely di ers from the adsorption isotherm (K 1.61).Desorption process in C 1 deep layer was not easily reversible.us, it contributed to a great retention of B (not released) [36].
Desorption characteristics of the soil were identi ed with a desorption index (DI) calculated as the ratio of Freundlich adsorption and desorption slopes.Indeed, if the n des ′ /n ads value is near one (≈1), it shows reversible desorption.On the contrary, if DI is lower than one (<1), the desorption process is hardly reversible.e data reported in Table 2 show the DI values of the di erent studied horizons.e value of DI of the A p surface horizon (DI 1.13), showing reversible adsorptiondesorption process.e other horizons displayed DI values <1, tending towards hysteric desorption (Figure 5).Previous studies also showed that the content of adsorbed B might not be released in a completely reversible way [15,36].An attempt was made to relate DI values to di erent soil properties.Signi cant correlations were observed with sand fraction (r 2 0.94 at p < 0.05), silt fraction (r 2 0.53 at p < 0.05), TOC content (r 2 0.79 at p < 0.05), and CEC values (r 2 0.59 at p < 0.05).Similarly, Arora and Chahal [11] have shown a positive correlation between the desorbed B amount and the sand content.In addition, our results revealed that OM and silt contents, unlike clays, contribute signi cantly to the release of B in soil and thus to its bioavailability for plants.Indeed, DI values showed a negative correlation with clay content (r 2 −0.76 at p < 0.05), pH (CaCl 2 ) level (r 2 −0.86 at p < 0.05), and EC values (r 2 −0.83 at p < 0.05).ese results could be ascribed to the fact that the adsorption of B increases at high pH level, EC value, and clay content.

Conclusion
Based on our ndings, a variety of soil properties (OM content, particle-size fractions, percentage of calcium carbonate, soil pH level, and salinity) have been identi ed as a ecting the dynamic of B and its distribution in the soil of Dour Ismail irrigated perimeter (North Tunisia).Our results showed that the increase in the pH level and the great percentage of TCaCO 3 signi cantly diminished the available B contents mainly in surface horizons.However, the accumulation of B in depth corresponds to an enrichment in OM (1.05%), to a high EC solution (2.08 dS/m), and to a high percentage of clay-silt fractions (≈62%).
e investigation of depth B distribution in the di erent particle-size fractions showed a considerable contribution of the silt (2-50 µm) fraction (52% of the total available B) to B adsorption especially in depth, while the clay (<2 µm) and the coarse (>50 µm) fractions seem to play a less important role.
e lowest B values recorded in the clay fractions mainly in depth correspond to signi cant OM levels (1.05%).Soil OM occupies the reactive adsorption sites on clay surfaces, therefore decreasing B adsorption.
Indeed, a deeper understanding of the factors a ecting B bioavailability in soils is essential for the assessment of B de ciency or toxicity under di erent conditions.It is worth mentioning that B availability is monitored by the equilibrium B concentration in soil.It described B adsorption and desorption reactions by Freundlich adsorption isotherm.e adsorption of B by soil increased with increasing B concentration in soil solution.
e most important soil properties a ecting the B adsorption capacities (K values) in    Journal of Chemistry the studied soil were: pH (CaCl 2 ) level (r 2 0.93), TN content (r 2 0.60), and clay-silt amount (r 2 0.48).Desorption isotherm denotes that accumulated B in soil cannot be easily released.It exhibits a hysteretic adsorption process in all horizons, except the A p surface horizon, which revealed a reversible adsorption process.Indeed, soils which have a higher a nity for B adsorption tended to desorb less B amounts, namely, A/C horizon.B desorption was positively correlated with sand (r 2 0.94), silt (r 2 0.53), TOC (r 2 0.79), and CEC values (r 2 0.59) and negatively with pH (CaCl 2 ) (r 2 −0.86), salinity (r 2 −0.83), and clay (r 2 −0.76).
Based on our results, it appears possible to declare that OM and silt content can play an important role in B adsorption-desorption process in soil.In reality, the study of B distribution in di erent particle-size fractions, as well as a thorough understanding of the factors that a ect its bioavailability for plants, is essential to better manage B deciency or toxicity in soil.

Figure 1 :
Figure 1: (a) Geographic situation of the study area.(b) Position of the studied soil pro le.

Figure 3 :Figure 2 :
Figure 3: Amounts of the adsorbed B of the di erent horizons of the studied soil.

Figure 4 :
Figure 4: Amounts of desorbed B expressed as percentage of the adsorbed B.

Figure 5 :
Figure 5: Comparison between adsorbed and desorbed B amounts of the di erent horizons of the studied soil.

Table 1 :
Selected physicochemical properties of the studied soil.

Table 2 :
Freundlich adsorption and desorption coe cients and desorption index of the studied soil.