CHEMICAL FRACTIONATION AND ENVIRONMENTAL RISK ASSESSMENT OF POTENTIALLY TOXIC ELEMENTS IN TOPSOILS OF THE CAATINGA BIOME

Abstract

The distribution and mobility of potentially toxic elements were evaluated in 50 topsoils from the Caatinga biome in Brazil using BCR (Community Bureau of Reference) sequential extraction, the risk assessment code (RAC), and principal component analysis (PCA). The residual fraction was dominant for most elements, with concentrations reaching up to 154 mg kg^-1 for Ni and 82 mg kg^-1 for Cr. However, significant mobility was observed in the acid-soluble fraction, particularly for Ni (0.16-10 mg kg^-1) and Cd (0.12-1.24 mg kg^-1), whereas Pb (0.75-15 mg kg^-1) and Zn (0.34-4.6 mg kg^-1) showed moderate mobility. Consequently, RAC classified Ni (13.3%) and Cd (12.7%) as medium environmental risk, while other metals showed low risk. PCA revealed that soil pH correlated positively with Ni, Cu, and Cr, and negatively with Zn, Cd, and Pb. Organic matter and soil texture were also important in influencing metal mobility. These findings indicate that, despite the geogenic origin of these soils, the high mobility of Cd and Ni in specific pH conditions poses potential ecological risks. This suggests that environmental monitoring in the semi-arid Caatinga ecosystem should focus on bioavailable fractions rather than total content to prevent trophic transfer effectively.

Keywords:
sequential extraction; bioavailability; multivariate analysis; semi-arid region; geochemical mobility.

INTRODUCTION

Soil is a repository for various potentially harmful and toxic substances. Its condition as an interface among the biosphere (land biomass, marine biomass, and humans), the lithosphere (crust, soils, and sediments), the hydrosphere (freshwater and seawater), and the atmosphere, makes it a transit station for contaminants.¹

Evaluating and characterizing soils, especially those influenced by contaminants, is one of the most significant environmental challenges. Considering this, several types of research have been developed to understand the dynamics of pollutants within the soil system. Trace metals occupy a prominent place among these and are being intensively studied.²-⁴

The potential danger of contaminants in soils arises not just from their overall concentration but primarily from their bioavailability. Trace metals can exhibit toxicity to organisms and must be chemically accessible. Consequently, relying solely on the total content is insufficient to accurately gauge the risk involved.⁵ Metals bound loosely to chemical species exhibit high availability, as they can easily dissolve with a slight change in soil conditions, like pH or redox potential. In contrast, metals tightly bound to chemical species demonstrate strong stability.

Significant changes in environmental conditions, which generally do not occur in nature without human influence, are required to become soluble. When metal concentrations are high under these conditions, their bioavailability is low and, consequently, the risk of toxicity is negligible. Thus, chemical speciation studies enable the determination of bioavailability and associated risks of trace metals.⁶,⁷

Operational speciation involves identifying and quantifying different chemical forms of an element within a sample. This is typically done using a simple or sequential extraction process, in which various chemical extractors are applied to the same sample in succession. Each extraction phase has distinct conditions compared to the previous one. This method is commonly employed in soil analysis due to its simplicity and the availability of instrumental procedures in most laboratories.⁵,⁸

Sequential extractions do not provide complete chemical speciation but may yield realistic characterizations that are of great interest for the behavior of an element as a contaminant. Metal contents (which may not always represent the total) can be associated with their mobility and their ability to be transferred to living organisms. In summary, sequential extraction can approximate the distribution of trace elements across soil stages and their relative mobility.¹

Different sequential extraction schemes exist, and all practices follow the same procedure. To harmonize the various techniques used, the European Community BCR 1992 conducted a study involving multiple extractors and their applications.⁹,¹⁰ The BCR (Community Bureau of Reference) method established three steps for extracting metals associated with acid-soluble, reducible, and oxidizable phases. However, due to specific challenges encountered with the BCR scheme, the original protocol was reviewed and optimized, resulting in a revised three-step procedure.

A previous study³ highlighted the relationships between soil chemistry and biodiversity within the Caatinga ecosystem. In this study, the status of the surface soil elements was investigated using the accumulation index, finding some areas with high levels of degradation, probably due to the use of fertilizers and pesticides in agriculture, as well as the presence of industries in the region. However, a deeper and more detailed understanding is needed. Employing sequential extraction methods can provide a more accurate analysis of how trace metals are distributed across different soil fractions, distinguishing between bioavailable and inert components. In regions like the Caatinga, where biodiversity is rich and vulnerable, it is crucial to understand the nature of metal contamination. This knowledge helps assess direct ecological risks and supports the development of effective conservation strategies. By identifying the exact forms and mobility of these metals, researchers can provide actionable data that helps policymakers and environmental managers implement measures to mitigate the adverse effects of such contamination, thereby promoting sustainable land-use and conservation practices in these areas.

For this reason, this work employed a method encompassing four extraction fractions: metal in exchangeable ions and carbonates, metal associated with Fe and Mn oxides, metal bound to organic matter (OM), and residual or lithogenic phase. Excellent speciation methods are essential to accurately determining the total concentration and different chemical species.

In essence, the hypothesis suggests that the total concentration of potentially toxic elements (PTEs) is insufficient to predict their environmental impact, making chemical speciation essential for accurate risk estimation. Therefore, the primary objective of this research is to evaluate the geochemical fractionation and environmental risk of Cd, Cr, Cu, Ni, Pb, and Zn in topsoils of the Caatinga biome. This was achieved by applying the BCR sequential extraction method, calculating the risk assessment code (RAC) to quantify potential mobility, and using multivariate statistical tools (principal component analysis (PCA)) to identify the geo-edaphic factors controlling this availability.

EXPERIMENTAL

Study area

The study area was the state of Pernambuco in the Northeast region of Brazil. It occupies a territorial area of 98,311.6 km² and has a population of 9,674,73.¹¹ The samples were selected in eight different soil types, which were distributed as follows: Planossol (alfisols) (18%), Luvisolo (aridisols) (8%), Regolithic Neosol (entisols Psamments), Litholic Neosol (entisol Lithic Orthents) and Quartzarenic Neosol (entisols Quartzipsamments) (30%), Gleisol (gelisols) (12%), Oxisol (oxisols) (10%), and Argisol (ultisols) (20%).¹² Soil samples were collected from areas under Caatinga vegetation cover, a unique Brazilian biome characterized by drought-resistant, withered scrub trees throughout most of the year.³,¹³Figure 1 details the soil type in the 50 topsoil samples collected.

Figure 1
Study area, sampling sites, and geological details

The tropical climate of Pernambuco is characterized by distinct local variations influenced by its geographical location, limited rainfall, and the interplay of meteorological systems in the area. It has two seasons: a rainy season from March to June and a dry season from July to December.

The climate in the countryside of the state is milder due to the Borborema Plateau. During winter, temperatures in some cities can drop to 10 °C or even 5 °C, while on the Pernambuco coast, the average temperature ranges from 25 to 30 °C.

Sample collection and preparation

Trenches of 0.7 × 0.7 m were opened according to the procedures established by EMBRAPA.¹⁴ The collected samples were conditioned in plastic bags and labelled. They were air-dried and then sieved through a 2-mm sieve to remove stones and other debris. Afterwards, they were macerated in an agate mortar, with granulometry reduced to pass through a 75 μm sieve (ABNT/ASTM 200). The humidity was determined by weighing 0.1 g of each sample and drying it in an oven at 105 °C until constant weight.

The mobility of PTEs is limited, resulting in their accumulation in the uppermost soil layers and minimal leaching to lower horizons.¹⁵ Consequently, when these metals are introduced onto soil surfaces due to human activities, their concentrations diminish with increasing depth, and their availability depends on the specific characteristics of each soil. Considering this, soil samples were collected at depths of 0 to 20 cm.

Soil pH was measured with a combined electrode submerged in a soil-water suspension at a 1:2.5 ratio. Determining soil density involves measuring the sample mass and its volume. The volume was obtained by collecting an undisturbed soil sample through a cylinder of known internal volume.

A dry combustion method with a CHN elemental analyzer (TruSpec CHN LECO^® 2006, St. Joseph, USA) was used to determine the carbon concentration. The organic matter of the soil was estimated from its organic carbon content, with carbon accounting for 58% of the soil’s organic matter. Granulometric analysis (sand, clay, and silt) was conducted using the pipette method, following the EMBRAPA guidelines.¹⁶

The sequential extraction method was used to determine the concentrations of Cd, Cr, Cu, Ni, Pb, and Zn in the available fractions. This method was based on the protocol outlined in the BCR Program, which involved a series of steps designed to isolate and quantify these metals in various forms.¹⁰ The methods used for processing the samples were divided into four stages.

The exchangeable ions and those bound to carbonate were extracted in the initial (first) stage. To achieve this, a combined 1.0 g of each soil sample with 40 mL of a 0.11 mol L^-1 acetic acid solution (HOAc) was placed in polyethylene containers. The mixture was agitated for 16 h at ambient temperature. Subsequently, the contents of the polyethylene containers were transferred to tubes and centrifuged. The resulting supernatant was collected and preserved in the refrigerator at 4 °C until analysis. The remaining residues were rinsed with 20 mL of ultrapure water from a Milli-Q system (with a resistivity of 18.0 μS cm^-1) and then centrifuged for 10 min. The washing solution was discarded to ensure no loss of solid residues from the containers.

In the second step, elements bound to Fe-Mn oxides were extracted. After the first step, each sample residue was mixed with a 40 mL hydroxylamine hydrochloride solution at 0.5 mol L^-1. The solution was acidified with HNO₃ to pH 1.5 before being added to the residue. The mixture was agitated at room temperature for 16 h, and the resulting extract was separated from the residue by centrifugation. The supernatant was preserved at 4 °C for later analysis, while the residue underwent the washing procedure outlined in the first step.

In the third step, elements bound to organic matter were extracted. Considering the sample treatment, 10 mL of H₂O₂ 8.8 mol L^-1 (30 v v^-1) was added to the second-stage residue, which was then digested for 1 h at room temperature and manually shaken to facilitate the reaction. Digestion was continued for one hour in a thermostatic bath at 85 °C, the bath temperature being increased to about 99 99.8 °C to reduce the volume to a few milliliters. Then, 10 mL of H₂O₂ was added to the beaker, the beaker was covered, and it was heated in a thermostatic bath at 85 °C for 1 h.

Afterwards, the vessel was opened, and the bath temperature was increased to evaporate the sample to dryness. Then, 50 mL of 1 mol L^-1 ammonium acetate solution (C₂H₇NO₂) was added to the residue, and the mixture was shaken for 16 h at room temperature. After completing this step, the precipitate was centrifuged as described earlier, and the resulting extract was stored at 4 °C until analysis.

In the fourth step, the residual fraction was extracted. At this stage, the residue prepared in the third step was transferred to a Pyrex vessel to prevent loss, and 7 mL of 12 mol L^-1 HCl and 2.3 mL of 15.8 mol L^-1 HNO₃ were added. The mixture was then digested at room temperature for 16 h. The mixture was heated on a heating plate for 2 h, until it boiled. The residues were filtered and transferred to 25 mL volumetric flasks after the vessel was cooled to room temperature, and the volume was completed with HNO₃ (2 v v^-1).

Measurement system

The metals were determined using flame atomic absorption spectrometry (FAAS) with a VARIAN^® Spectr AA-220FS (California, USA). The background correction was performed using a deuterium hollow-cathode lamp. Acetylene was used as the fuel gas, and air as the oxidant, with gas flows of 2 and 13.5 L min^-1, respectively. For metal quantification, calibration curves were prepared using dilutions of stock solutions at 1000 μg mL^-1 for each metal (Spectrosol from Merck, Darmstadt, Germany). The other instrumental parameters are detailed in Table 1. The concentration ranges of the curves were 0.53-2.0 mg L^-1 for Cd and Zn and 0.11 2.0 mg L^-1 for Cr, Cu, Ni, and Pb.

Table 1
Instrumental operating conditions for metal determination by FAAS

Element	Wavelength / nm	Lamp current / mA	Slit width / nm
Cd	228.8	4.0	0.5
Cr	357.9	7.0	0.2
Cu	324.7	4.0	0.5
Ni	232.0	4.0	0.2
Pb	217.0	5.0	1.0
Zn	213.9	5.0	1.0

The efficiency of the sequential extraction method was evaluated using the CRM BCR-701 (lake sediment), which has certified values for the available fractions and informative values for direct material analysis. Accuracy was evaluated using the normalized error (E_n) score, calculated according to Equation 1. The appropriate range for the reference material results was considered to be between -1 and 1, as recommended by ISO 13528.¹⁷

(1) $${\mathrm{E}}_{\mathrm{n}}=\frac{({\mathrm{x}}_{\mathrm{obt}}-{\mathrm{x}}_{\mathrm{cert}})}{\sqrt{{\left({\mathrm{U}}_{\mathrm{obt}}\right)}^2-{\left({\mathrm{U}}_{\mathrm{cert}}\right)}^2}}$$

where: x_obt and x_cert are the obtained and certified values from the sample analysis; U_obt and U_cert are the expanded analytical uncertainties at the 95% confidence level.

Risk assessment and statistical analysis

To assess the potential environmental risk and mobility of the PTEs, the RAC was calculated. The RAC is defined as the percentage of the metal present in the exchangeable and acid-soluble fraction (F1) relative to the total concentration (sum of all fractions). The calculation is expressed by Equation 2.

(2) $$RAC=\left(\frac{C_{F1}}{\sum _{i=1}^4{C}_{Fi}}\right)\times 100$$

where C_F1 is the metal concentration in the acid-soluble fraction, and ∑C_Fi is the sum of concentrations in all four fractions. The risk classification criteria are: RAC < 1% (no risk); 1% ≤ RAC ≤ 10% (low risk); 11% ≤ RAC ≤ 30% (medium risk); 31% ≤ RAC ≤ 50% (high risk); and RAC > 50% (very high risk).¹⁸

This study employed principal component analysis (PCA), Varimax rotation, and Kaiser-Meyer-Olkin normalization. PCA is a mathematical technique that transforms a set of potentially interrelated variables into a more concise set of uncorrelated variables, known as principal components. The initial principal component effectively captures the most significant variability inherent in the data. Subsequent components account for residual variability beyond that justified by the preceding components. This approach facilitates the examination of interconnections among the observed variables. This method analyzes relationships among the observed variables.³

RESULTS AND DISCUSSION

As shown in Table 2, the content of Cd, Cr, Cu, Ni, Pb, and Zn in the sample CRM BCR-701 associated with the four fractions: metal in the form of exchangeable ions, and carbonates (step 1), a metal associated with oxides of Fe and Mn (step 2), metal bound to organic matter (step 3) and residual or lithogenic phase (step 4), analyzed by the FAAS technique. In each extraction step, the number E_n was calculated. The results obtained are close to the certificates and reflect the efficiency of the method; the number E_n demonstrates this with values between -1 and 1.

Table 2
Values of certified and obtained concentrations of the chemical elements in the BCR-701, analytical uncertainties at the 95% confidence level, and the average number of E_n (normalized error)

Metal		Step 1 / (mg kg-1)	Step 2 / (mg kg-1)	Step 3 / (mg kg-1)	Step 4 / (mg kg-1)
Cd	Cert. value	7.34 ± 0.35	3.77 ± 0.28	0.27 ± 0.06	0.13a ± 0.08
	Obt. value	7.69 ± 0.82	4.24 ± 0.65	< LOD	< LOD
	En	0.39	0.66
Cr	Cert. value	2.26 ± 0.16	45.7 ± 2.0	143 ± 7	62.5a ± 7.4
	Obt. value	2.50 ± 0.80	47.70 ± 0.64	141 ± 5	54.7 ± 3.6
	En	0.30	0.95	-0.15	-0.95
Cu	Cert. value	49.3 ± 1.7	124 ± 3	55.2 ± 4.0	38.5a ± 11.2
	Obt. value	50.2 ± 2.6	129 ± 21	54.8 ± 0.73	29.7 ± 2.9
	En	0.30	0.23	-0.11	-0.76
Ni	Cert. value	15.4 ± 0.9	26.6 ± 1.3	15.3 ± 0.9	41.4a ± 4.0
	Obt. value	17.3 ± 2.7	31.8 ± 6.0	12.9 ± 2.2	42.2 ± 1.5
	En	0.69	0.84	-0.98	0.18
Pb	Cert. value	3.18 ± 0.21	126 ± 3	9.3 ± 2.0	11.0a ± 2.0
	Obt. value	3.00 ± 0.05	128 ± 0.77	10.6 ± 1.3	7.6 ± 2.5
	En	-0.82	0.52	0.53	-0.59
Zn	Cert. value	205 ± 6	114 ± 5	45.7 ± 4.0	95a ± 13
	Obt. value	218 ± 14	116 ± 7	42.82 ± 0.78	99.3 ± 1.8
	En	0.84	0.20	-0.71	0.32

• ^a
  Informative value. LOD: limit of detection = 0.1 mg kg^-1.

Figure 2 presents the percentages of Cd, Cr, Cu, Ni, Pb, and Zn in the acid-soluble (step 1), reducible (step 2), oxidizable (step 3), and residual (step 4) fractions. The figure shows the distribution of the 50 soil samples in each fraction.

Figure 2
Percentage distribution of Cd, Cr, Cu, Ni, Pb, and Zn in the four fractions obtained by the modified BCR sequential extraction procedure. Fractions: F1 (acid-soluble), F2 (reducible), F3 (oxidizable), and F4 (residual)

The acid-soluble fraction indicates the presence of PTEs that could be released into the environment under increased acidity. This fraction poses the highest environmental risk due to its increased mobility.

The acid-soluble fraction levels of Cd, Cr, Cu, Ni, Pb, and Zn ranged from 0.12 to 1.24 mg kg^-1; 0.11 to 2.53 mg kg^-1; 0.12 to 0.68 mg kg^-1; 0.16 to 10 mg kg^-1; 0.75 to 15 mg kg^-1 and 0.34 to 4.6 mg kg^-1, respectively, with Cd, Ni, and Zn in high proportion. The same behavior was observed by Khadhar et al.⁴

The reducible fraction of PTEs is the metal content bound to iron and manganese oxides that would be released if the substrate were subjected to more reducing conditions. The levels of Cd, Cr, Cu, Ni, Pb, and Zn in this fraction ranged from 0.20 to 1.6 mg kg^-1; 0.39 to 37 mg kg^-1; 0.17 to 13 mg kg^-1; 0.11 to 26 mg kg^-1; 2.2 to 44 mg kg^-1 and 0.41 to 9.9 mg kg^-1, respectively. According to Huang et al.,¹⁹ adding Cd to Mn oxide lowered the adsorption of this metal. Cd and Pb are consistently higher in step 2, indicating that both are mobile in the environment. Studies²⁰ in the Vale do Ribeira region indicate that Pb transport is mainly associated with Fe and Mn oxides and hydroxides. Pb, having a large ionic radius, can occupy adsorption spaces with low binding energies. Thus, the adsorption of Pb by the oxides of Fe and Mn is considered the main retention process of this metal in the soil.

The oxidizable fraction indicates the amount of metal bound to organic matter and sulfides, which would be released into the environment if conditions became oxidative. The levels of Cd, Cr, Cu, Ni, Pb, and Zn for this fraction ranged from 0.66 to 1.35 mg kg^-1; 0.77 to 22 mg kg^-1; 0.10 to 17 mg kg^-1; 0.10 to 12 mg kg^-1; 11 to 26 mg kg^-1 and 0.27 to 15 mg kg^-1, respectively. The abundance of Cu observed in step 3 can be attributed to its strong attraction to soluble organic ligands. The formation of these complexes has the potential to enhance the mobility of Cu in soils.⁴,²¹ According to the EPA,²² the predominant form of zinc (Zn) in contaminated soils was associated with iron and manganese oxides. In our study, more Zn content was found in the residual fraction, indicating the influence of the natural substrate of the soil. The Cd fraction was the highest obtained in this step; similar results were found in the research of Chavez et al.²³

The residual metal fraction is bound by the strongest association with the crystalline structures of the minerals, and it is not always easy to separate them from the extracted material. In the modified BCR protocol, the residual fractions were digested with aqua regia. The concentration ranges of Cd, Cr, Cu, Ni, Pb, and Zn within this fraction were as follows: 0.16 to 29 mg kg^-1, 0.87 to 82 mg kg^-1, 0.93 to 54 mg kg^-1, 0.77 to 154 mg kg^-1, 8 to 39 mg kg^-1, and 0.63 to 57 mg kg^-1, respectively. In general, this step revealed the largest metal fraction. This finding implies a reduced risk of pollution associated with these elements, as the residual fraction contains metals that are unlikely to be released under typical environmental conditions.

The calculated RAC values followed the order: Ni (13.3%) > Cd (12.7%) > Zn (9.3%) > Pb (6.4%) > Cr (2.9%) > Cu (2.7%) (Figure 3). According to the classification criteria, Ni and Cd posed a medium risk (RAC between 11 and 30%) to the local environment. This suggests that a significant portion of these elements is weakly bound to the soil matrix and may enter the food chain via plant uptake or leach into groundwater.

Figure 3
Environmental risk assessment (RAC) of topsoils in Pernambuco. The lines indicate the risk thresholds: low risk (1-10%) and medium risk (11-30%)

Zn presented a RAC of 9.33%, classifying it as low risk but approaching the medium-risk threshold (10%). Pb, Cr, and Cu exhibited values below 10%, falling within the low risk category, indicating limited mobility and a strong association with more stable soil fractions (reducible and residual). These results highlight that, despite the geogenic nature of the soils, specific monitoring strategies should focus on Ni and Cd, given their greater geochemical mobility in the Caatinga biome.

The geo-ecological parameters are essential for evaluating soil sensitivity to pollutant presence. When metals are introduced to the soil surface, their downward transport is limited unless the metal retention capacity of the soil is exceeded or the interaction of the metal with the waste matrix increases mobility. Factors such as the breakdown of the organic waste matrix, shifts in pH and redox potential, or variations in soil solution composition resulting from different remediation methods or natural weathering processes can enhance metal mobility. For this reason, the relationship between these, in particular pH, soil density, granulometry, carbon, organic matter and the metal contents found was studied (Table 3).

Table 3
Pearson’s correlation matrix between carbon (C), organic matter (OM), sand, clay, silt, pH, density (Dens.), Cd, Cr, Cu, Ni, Pb, and Zn (metal concentrations represent the sum of the four fractions)

	C	OM	Sand	Clay	Silt	pH	Dens.	Cd	Cr	Cu	Ni	Pb	Zn
C	1	1
OM	1	1
Sand	-0.33	-0.33	1
Clay	0.66	0.66	-0.66	1
Silt	-0.47	-0.47	0.04	-0.45	1
pH	-0.36	-0.36	0.17	-0.36	0.76	1
Dens.	-0.16	-0.16	0.66	-0.62	0.26	0.09	1
Cd	0.29	0.29	-0.27	0.41	-0.32	-0.37	-0.24	1
Cr	0.43	0.43	-0.40	0.61	-0.10	0.10	-0.40	-0.01	1
Cu	0.05	0.05	0.08	0.18	0.06	0.34	-0.24	0.14	0.51	1
Ni	0.15	0.15	-0.28	0.40	0.20	0.39	-0.48	-0.14	0.78	0.71	1
Pb	0.32	0.32	-0.11	0.23	-0.35	-0.58	0.01	0.63	-0.15	-0.25	-0.44	1
Zn	0.19	0.19	-0.01	0.14	0.07	-0.02	0.04	0.44	-0.01	0.20	0.05	0.31	1

Significant positive correlations were observed among Pb, Cd, and Zn, indicating that these elements exhibit similar geochemical behavior and retention mechanisms in the studied soils. This strong association is likely driven by their high affinity for soil colloids, particularly clay minerals and organic matter, which provide negatively charged sites for cation adsorption and complexation. Furthermore, the positive correlation found between Cr, Ni, and Cu and the finer soil fractions (silt and clay) suggests that their distribution is controlled by the specific surface area and cation exchange capacity (CEC) of the soil matrix. Conversely, the negative correlation between most metals and the sand fraction reflects a “dilution effect”, where quartz-rich sandy soils, having fewer binding sites, exhibit naturally lower metal concentrations. These correlations support the hypothesis that, for Caatinga topsoils, metal distribution is primarily governed by geo-edaphic characteristics (texture and OM) rather than diffuse anthropogenic inputs.

Complementing Pearson’s correlation analysis, a multivariate approach was applied to identify the underlying geo-edaphic factors controlling element availability. Thus, the relationship between potentially toxic elements and soil physicochemical properties was evaluated using PCA. The results are visualized in a biplot (Figure 4), where the first two principal components (PC1 and PC2) accounted for 58.35% of the total variance (PC1: 34.17% and PC2: 24.18%).

Figure 4
Principal component analysis (PCA) biplot showing the multivariate relationships between potentially toxic elements and soil physicochemical properties

The biplot reveals distinct geochemical associations based on the length and direction of the vectors. The first component (PC1) was strongly influenced by soil texture, organic matter and C. A cluster formed by clay, OM, and C showed strong positive correlations with most metals (particularly Cr, Cd, Ni, and Zn), as indicated by acute angles between their vectors. In contrast, sand and density displayed a strong negative correlation with these elements, appearing in the opposite quadrant. The second component (PC2) highlighted the influence of pH. The pH vector was oriented in the opposite direction to that of the group formed by Cd, Pb, and Zn, suggesting a negative correlation and reinforcing the conclusion that higher acidity (lower pH) favors the mobility and potential availability of these elements. Meanwhile, Ni, Cu, and Cr formed a distinct cluster positively associated with the finer soil fractions (clay and silt), suggesting their retention in the soil matrix.

Soil pH plays a crucial role in metal extraction from soil samples. In acidic conditions, the high concentration of H⁺ ions competes with metal cations for binding sites on soil colloids, resulting in increased solubility and availability. Conversely, as pH increases (particularly above 7.0), metal retention is significantly enhanced due to two primary mechanisms: specific adsorption and precipitation. At higher pH levels, metal cations undergo hydrolysis to form hydroxylated species, which are preferentially adsorbed onto the increasingly negative surface charges of soil particles. Furthermore, alkaline conditions promote the formation of insoluble precipitates, such as metal hydroxides and carbonates. This precipitation process effectively removes metals from the soil solution, drastically reducing their mobility and bioavailability compared to acidic environments.²²,²⁴

Ni and Pb contents in the F1 fraction exhibited a positive correlation, consistent with Sungur et al.²⁴ findings in agricultural soils. Conversely, a negative correlation was observed between soil pH and Cd in the F2 and F3 fractions, while a positive correlation was evident in the F4 fraction. pH displayed positive correlations, notably with the less mobile F3 and immobile F4 fractions, showing connections with Zn and Cu. Zinc hydrolyses when pH exceeds 7.7, leading to strong adsorption onto soil surfaces. Furthermore, zinc forms complexes with inorganic and organic ligands, influencing its interaction with the soil surface’s adsorption sites, as the EPA noted.²² Ni exhibited a positive correlation with pH in both mobile and immobile phases. This implies that, beyond soil pH and other parameters, organic matter content is essential for understanding how soil properties affect Ni availability.

In soil, chromium has two potential oxidation states: trivalent chromium (Cr³⁺) and hexavalent chromium (Cr⁶⁺). While Cr⁶⁺ is highly mobile and toxic, Cr³⁺ is generally less mobile and tends to bind to the soil solid phase. Soil pH governs the speciation and solubility of these forms. Under acidic conditions, the reduction of Cr⁶⁺ to Cr³⁺ is favored; however, at low pH, Cr³⁺ remains soluble and mobile. As soil pH rises (typically above 5.5), Cr³⁺ tends to precipitate as hydroxides or co-precipitate with iron oxides, thereby significantly reducing its mobility. This mechanism explains the positive correlation established in this research between soil chromium levels and pH: higher pH values promote the precipitation and retention of chromium in the soil matrix, whereas acidic conditions favor its solubility and potential leaching. This behavior is similar to that found by other researchers.²⁵

Beyond soil pH, the presence of OM in soil is another highly significant factor influencing the availability of PTEs. Organic matter interacts with metals, forming exchange complexes or chelates. Adsorption can be so strong that it stabilizes, as with Cu, or it forms very stable chelates, as with Pb and Zn. Organometallic complexes are often incorporated, thereby facilitating metal solubility, availability, and dispersion, as soil organisms can degrade them. This leads to the persistence of toxicity.

The influence of organic matter on the availability of metals has been extensively investigated. Research indicates that as organic matter content in soils decreases, the adsorption of PTEs onto soil components also decreases. Additionally, dissolved organic matter in soils can enhance the movement and absorption of PTEs by plant roots, as reported by Zeng et al.²⁵ Altering soil pH to a higher level could mobilize metals because of the intricate interactions in soils abundant with dissolved organic matter, as noted by the EPA.²²

In general, the relationship between metals present in the soil and organic matter content showed limited significance. Among them, Cr exhibited the most notable positive correlation. Cr³⁺ can form soluble organic complexes when interacting with natural organic matter in the surrounding environment, thereby becoming accessible to biological processes.²⁵ This finding explains this study’s positive correlation between extractable Cr contents and organic matter.

A positive relationship exists between organic matter and carbon content with Zn in the F2 fraction and Cr, Ni, and Pb in the F3 and F4 fractions. A negative relationship with Cu was found in the F2 fraction. The similarities and differences were due to metal affinities against organic carbon and organic matter.

Organic matter is strongly related to soil texture; sandy soils with low clay content present low OM concentration. Fine-textured soils are likely to originate from secondary minerals, which are easily altered and generally represent the primary source of PTEs. Coarse textured soils comprise primary minerals, such as quartz, with low PTE levels. Clay soils retain more metals by adsorption, while sandy soils lack fixation capacity; therefore, contamination in deeper layers is more likely.²⁶,²⁷

Cationic metals such as Cd and Pb are less soluble in calcareous soil with increasing pH.²² This is consistent with the results obtained in the current study for cadmium in fractions 1, 2, and 3. Our research showed a strong correlation between clay and Cd, whereas the relationship with Pb was insignificant.

Clay minerals, carbonates, hydrous iron oxides, and manganese oxides can adsorb cadmium, or cadmium may precipitate as cadmium carbonate, hydroxide, or phosphate. The behavior of cadmium in soil is primarily influenced by pH, similar to other cationic metals. When the soil is acidic, cadmium solubility rises, resulting in minimal adsorption by soil colloids, hydrous oxides, and organic matter. In contrast, at pH values above 6, cadmium binds to the solid soil phase or precipitates, resulting in a significant reduction in cadmium solution concentrations. Lead exhibits a strong affinity for organic ligands, and forming such complexes can notably enhance lead mobility in soil.²⁰

The soil texture also influenced nickel, which correlates well with clay in the studied fractions. Ni tends to attach itself to clays, as well as to iron and manganese oxides and organic matter. As a result, it becomes separated from the soil solution.²⁸,²⁹ Surface soils can be chelated, forming chelates of considerable solubility, and clay can form stable complexes that can even express their distribution in the soil profile. Both inorganic and organic ligands can form complexes with nickel, enhancing its mobility within soils.²²

Research indicates that Cr and Ni frequently interact with clay minerals throughout pedogenic processes. As a result, they exist as structural components of clay minerals rather than as exchangeable ions on the surface of clays.²¹ Cr in soils has little mobility and can form complexes with Fe and Mn oxides, organic matter, and the clay mineral fraction. This element was the one that most correlated with clay in this research.

In this study, heavy-textured soils (sand and silt) correlate negatively with most metals (Figure 4), although there is a strong positive correlation between them. Organic matter and clay content were positively correlated with the metals analyzed, but not with pH for most elements. Silt content correlated only positively with Cu in the F2 and F4 fractions and poorly with other metals.

Soils with greater texture and elevated pH levels exhibited more effective metal attenuation, whereas sandy and low-pH soils exhibited less efficient metal retention. Clay soils containing low-pH oxides exhibited relatively good preservation of anionic metals. Similar to cationic metals, lighter-textured soils were less effective at retaining anions.

In Figure 5, we can see how the different soil types found in the region, alfisols, aridisols, entisols (Lithic Orthents, Psamments, and Quartzipsamments), gelisols, oxisols, and ultisols, behaved with respect to the physical-chemical parameters studied. In the case of gelisols, determining the content of sand, silt, and clay was impossible.

Figure 5
Behavior of soil type with physical-chemical parameters. (a) Stacked column graph representing soil texture (sand, clay, silt), in %. Content of C (g kg^-1), OM (organic matter) (g kg^-1), density (g cm^-3), and pH in the secondary axis; (b) bar chart with the chemical elements studied (values represent the sum of the four fractions). Content expressed in mg kg^-1

As shown in Figure 5, entisol has higher sand content and lower silt and clay content than other soils. This is explained by the absence of well-defined soil horizons, which requires more time for fine materials to accumulate.³⁰ It is also worth noting that aridisols and oxisols have high clay content and low sand content.

The pH value was lower in soils with higher OM and C content (gelisols, oxisols, and ultisols). According to the literature,³⁰ entisols and aridisols have higher pH than alfisols, gelisols, oxisols, and ultisols. In our study, the first ones had a higher pH, with an average value of 6.73 and 6.21 for aridisol and entisol, respectively.

Figure 5b shows that aridisols have high Ni, Cu, and Cr contents, with mean values of 103, 46, and 65 mg kg^-1, respectively. This accumulation pattern can be attributed to the limited leaching processes typical of arid environments, which prevent the downward migration of metals, leading to their retention in surface horizons.³¹ In contrast, highly weathered soils like oxisols and ultisols presented lower concentrations for most elements, likely due to long-term leaching losses.

Studies indicate that the concentrations of Cd, Cr, Cu, Ni, Pb, and Zn across soil types can vary with factors such as soil origin and history, as well as human activity in the area. In areas without significant human influence, the concentration of these metals appears relatively low compared to other types of soil. However, the concentration of these metals can be higher in mining areas or other activities that release PTEs into the soil.

CONCLUSIONS

The application of the BCR sequential extraction method revealed distinct geochemical behaviors among the studied elements. While the residual fraction was the primary host for most metals, indicating a lithogenic origin, the acid-soluble fraction played a critical role in the mobility of specific elements.

The risk assessment code provided quantitative evidence that Ni and Cd present a medium environmental risk in the topsoils of the Caatinga biome. Zinc showed a low risk but bordered the medium threshold, whereas Pb, Cr, and Cu were predominantly associated with stable fractions (reducible and residual), posing negligible immediate risks.

Multivariate statistical analysis confirmed that soil pH, organic matter, and texture are the primary determinants of the availability of these potentially toxic elements. The negative correlation between pH and the mobile fractions of Cd and Zn suggests that any acidification of these soils (natural or anthropogenic) could trigger a significant release of these toxic elements into the soil solution.

These findings highlight that, even in areas without massive industrial activity, the intrinsic geochemical characteristics of semi-arid soils can promote high bioavailability of toxic elements. This is particularly important in the unique and sensitive Caatinga biome, where the balance between soil health and agricultural productivity is critical. This research also sets a precedent for similar environmental studies in other regions, underscoring the need to address soil contamination in the face of increasing industrial and agricultural pressures.

ACKNOWLEDGMENTS

The authors thank the Radioecology and Environmental Control Laboratory of the UFPE Department of Nuclear Energy and the Northeastern Regional Center for Nuclear Sciences (CRCN-NE) for their chemical analyses. To the FACEPE (grant No.: BFP-0009-3.09/17, APQ-0245-3.01/21, BFP-0101-3.01/22 and APQ-1049-3.09/22) and a scientific initiation scholarship (grant No. BIC-0074-4.06/23); the CNPq (grant No.: 304557/2023-4 and 175873/2023-2); the CAPES (grant No. 88882.379364/2019-01); and to the Pro-Rectory of Research and Innovation PROPESQI - UFPE for the assistance in the Public Notice PROPG No. 05/2023 of the Dean of Graduate Studies at the Federal University of Pernambuco (grant No. 23076.055174/2023-86) for funding.

DATA AVAILABILITY STATEMENT

All data is available in the text. A preliminary version of this manuscript was deposited as a preprint on the Research Square server (https://dx.doi.org/10.21203/rs.3.rs-3328679/v2).

REFERENCES

• ¹ Coringa, J. E. S.; Pezza, L.; Coringa, E. A. O.; Weber, O. L. S.; Acta Amazonica 2016, 46, 161. [Crossref]
  » Crossref
• ² Alovisi, A. M. T.; Cassol, C. J.; Nascimento, J. S.; Soares, N. B.; da Silva Junior, I. R.; da Silva, R. S.; da Silva, J. A. M.; Geoderma Regional 2020, 22, e00298. [Crossref]
  » Crossref
• ³ Fernández, Z. H.; dos Santos Júnior, J. A.; Amaral, R. S.; de França, E. J.; Menezes, R. S. C.; Bezerra, M. B. C. F.; Coutinho, A. P.; Rojas, L. A. V.; Santos, J. M. N.; Bezerra, J. D.; de Araújo, E. E. N.; Chem. Ecol. 2022, 38, 108. [Crossref]
  » Crossref
• ⁴ Khadhar, S.; Sdiri, A.; Chekirben, A.; Azouzi, R.; Charef, A.; Environ. Pollut. 2020, 263, 114543. [Crossref]
  » Crossref
• ⁵ Siahcheshm, K.; Orberger, B.; Wagner, C.; Environ. Earth Sci. 2022, 81, 34. [Crossref]
  » Crossref
• ⁶ Hernandez-Soriano, M. C.; Jimenez-Lopez, J. C.; Sci. Total Environ. 2012, 423, 55. [Crossref]
  » Crossref
• ⁷ Kennou, B.; El Meray, M.; Romane, A.; Arjouni, Y.; Environ. Earth Sci. 2015, 74, 5849. [Crossref]
  » Crossref
• ⁸ Emurotu, J. E.; Nigerian Journal of Basic and Applied Sciences 2021, 28, 48. [Crossref]
  » Crossref
• ⁹ Fernández, E.; Jiménez, R.; Lallena, A. M.; Aguilar, J.; Environ. Pollut. 2004, 131, 355. [Crossref]
  » Crossref
• ¹⁰ Ure, A. M.; Quevauviller, P.; Muntau, H.; Griepink, B.; Int. J. Environ. Anal. Chem. 1993, 51, 135. [Crossref]
  » Crossref
• ¹¹ Instituto Brasileiro de Geografia e Estatistica (IBGE); 2021. [Link] accessed in March 2026
  » Link
• ¹² United States Department of Agriculture (USDA); Keys to Soil Taxonomy, 2022. [Link] accessed in March 2026
  » Link
• ¹³ Caatinga; Ministério do Meio Ambiente e Mudança do Clima, 2024. [Link] accessed in March 2026
  » Link
• ¹⁴ Empresa Brasileira de Pesquisa Agropecuária (EMBRAPA); Manual de Análises Químicas de Solos, Plantas e Fertilizantes, 2ª ed.; Embrapa Informação Tecnológica: Brasília, 2009.
• ¹⁵ Wuana, R. A.; Okieimen, F. E.; Int. Scholarly Res. Not. 2011, 2011, 402647. [Crossref]
  » Crossref
• ¹⁶ Empresa Brasileira de Pesquisa Agropecuária (EMBRAPA); Manual de Métodos de Análise de Solo, 3^a ed.; EMBRAPA: Brasília, DF, 2017.
• ¹⁷ International Organization for Standardization (ISO); ISO 13528: Statistical Methods for Use in Proficiency Testing by Interlaboratory Comparisons; ISO: Genebra, 2022. [Link] accessed in March 2026
  » Link
• ¹⁸ Liu, Z.; Wu, Q.; He C, Zhang, P.; Wang, X.; Luo, Z.; Yang, L.; Huang, C.; Frontiers in Environmental Science 2024, 12, 1509277. [Crossref]
  » Crossref
• ¹⁹ Huang, X.; Chen, T.; Zou, X.; Zhu, M.; Chen, D.; Pan, M.; Int. J. Environ. Res. Public Health 2017, 14, 1145. [Crossref]
  » Crossref
• ²⁰ Tramonte, K. M.: Estudo da Disponibilidade de Metais em Sedimentos do Sistema Estuarino-Lagunar de Cananéia-Iguape; Tese de Doutorado, Universidade de São Paulo, São Paulo, Brasil, 2014. [Link] accessed in March 2026
  » Link
• ²¹ Ghayoraneh, M.; Qishlaqi, A.; J. Geochem. Explor. 2017, 180, 1. [Crossref]
  » Crossref
• ²² McLean, J. E.; Bledsoe, B. E.; Ground Water Issue. Behavior of Metals in Soils; United State Environmental Protection Agency (USEPA): Washington, D.C., 1992. [Link] accessed in March 2026
  » Link
• ²³ Chavez, E.; He, Z. L.; Stoffella, P. J.; Mylavarapu, R. S.; Li, Y. C.; Baligar, V. C.; Chemosphere 2016, 150, 57. [Crossref]
  » Crossref
• ²⁴ Sungur, A.; Soylak, M.; Yilmaz, E.; Yilmaz, S.; Ozcan, H.; Soil Sediment Contam. 2014, 24, 1. [Crossref]
  » Crossref
• ²⁵ Zeng, F.; Ali, S.; Zhang, H.; Ouyang, Y.; Qiu, B.; Wu, F.; Zhang, G.; Environ. Pollut. 2011, 159, 84. [Crossref]
  » Crossref
• ²⁶ Gharaibeh, M. A.; Marschner, B.; Heinze, S.; Moos, N.; Geoderma Regional 2020, 20, e00245. [Crossref]
  » Crossref
• ²⁷ Reyes, A.; Cuevas, J.; Fuentes, B.; Fernández, E.; Arce, W.; Guerrero, M.; Letelier, M. V.; J. Geochem. Explor. 2021, 220, 106653. [Crossref]
  » Crossref
• ²⁸ Dai, Z.; Wang, L.; Tang, H.; Sun, Z.; Liu, W.; Sun, Y.; Su, S.; Hu, S.; Wang, Y.; Xu, K.; Liu, L.; Ling, P.; Xiang, J.; Chemosphere 2018, 207, 440. [Crossref]
  » Crossref
• ²⁹ Rasheed, K.; Solomon, A.; Aiyesanmi, A.; Phys. Sci. Int. J. 2019, 21, 1. [Crossref]
  » Crossref
• ³⁰ Empresa Brasileira de Pesquisa Agropecuária (EMBRAPA); Sistema Brasileiro de Classificação de Solos, 5ª ed.; Embrapa Solos: Brasília, DF, 2018.
• ³¹ Kabata-Pendias, A.; Trace Elements in Soils and Plants, 4^th ed.; CRC Press: Boca Raton, 2011.

Edited by

• Associate Editor handled this article:
  Mario Henrique Gonzalez

Publication Dates

• Publication in this collection
  03 July 2026
• Date of issue
  2026

History

• Received
  30 Dec 2025
• Accepted
  16 Mar 2026
• Published
  24 Mar 2026