Risk assessment and soil heavy metal contamination near marble processing plants (MPPs) in district Malakand, Pakistan

Abstract

Soil heavy metals (HMs) pollution is a growing global concern, mainly in regions with rapid industrial growth. This study assessed the concentrations, potential sources, and health risks of HMs in agricultural soils near marble processing plants in Malakand, Pakistan. A total of 21 soil samples were analyzed for essential and toxic HMs via inductively coupled plasma‒optical emission spectrometry (ICP‒OES), and probabilistic health risks were evaluated via Monte Carlo simulation. The concentrations (mg/kg) of Ca (29,250), P (805.5) and Cd (4.5) exceeded the average shale limits of 22,100, 700, and 3.0 mg/kg, respectively, and indices such as Nemerow’s synthetic contamination index (NSCI) and the geoaccumulation index (I_geo) categorized the soil sites as moderately polluted. The potential ecological risk index (PERI) indicated considerable to high ecological risk for As and Cd. The deterministic analysis indicated non-carcinogenic risks for children (HI > 1), whereas the probabilistic analysis suggested no significant risk (HI < 1) for both adults and children. Both methods indicated that the total cancer risk for Cr, Ni, Cd, and As exceeded the USEPA safety limits of 1.0E-06 and 1.0E-04. Sensitivity analysis identified heavy metal concentration, exposure duration, and frequency as key risk factors. The study suggested that HM contamination is mainly anthropogenic, poses a threat to soil and human health, and highlights the need for management strategies and surveillance programs to mitigate these risks.

Introduction

Soil is an indispensable natural resource, fundamental to food production and biodiversity maintenance, and plays an important role in national security and global ecological stability^1,2. However, the integrity of this resource is increasingly threatened by the pervasive issue of soil pollution, which has become a pressing global environmental concern³. The contamination of soil with hazardous substances such as radioactive elements, nanomaterials, and heavy metals significantly undermines soil resilience and disrupts ecosystems^4,5. These pollutants primarily originate from anthropogenic activities, including industrialization, urbanization, mining, and the excessive use of agrochemicals, as well as natural events such as volcanic eruptions and erosion of mineral deposits^6,7. The global significance of soil pollution lies in its far-reaching impacts on agricultural productivity, human health, and environmental sustainability^8,9. Contaminated soils can lead to the accumulation of toxic elements in the food chain, posing severe risks to food security and public health on a global scale^10,11. Moreover, soil degradation due to pollution exacerbates the loss of arable land, which is already under pressure from a growing global population^12,13. Addressing soil pollution is therefore critical to ensuring the long-term viability of global food systems, the health of populations, and the preservation of biodiversity^14,15.

Millions of tons of marble waste, consisting of fine powder and fragmented rock, are generated worldwide in marble quarries and marble processing plants (MPPs)^16,17. During the quarrying and refining processes, nearly 70% of the valuable mineral resources of marble are lost, with approximately 40% of marble waste being released into the surrounding atmosphere, leading to widespread environmental pollution^18,19. MPPs, particularly in Khyber Pakhtunkhwa, Pakistan, contribute significantly to wastewater and environmental pollution, as large amounts of waste are released as sludge and sludge containing harmful marble particles^20–24. This waste, which is often disposed of improperly in water bodies and land, poses severe ecological risks, including damage to aquatic ecosystems, soil, and human health^20,25,26. Despite the introduction of pollution control programs by the government, poor enforcement has hindered their effectiveness^27,28. Although the marble industry plays a crucial role in socioeconomic development worldwide, growing concern about its impact on the environment has led to an increasing focus on recycling waste and byproducts to align with sustainable development goals^19,29. Recently, modified biochar composites (Mar-BC800 and Sep-BC800) were developed from marble waste, calcium-rich sepiolite, and agricultural waste, and have been shown to be highly effective, eco-friendly, and cost-efficient for phosphate removal from waste streams³⁰. Moreover, their potential for repurposing as fertilizers presents a sustainable solution for enhancing soil fertility and improving soil properties³¹.

In developing countries without prior treatment, farmers use marble wastewater as an alternative to irrigating crops because it is a low-cost and readily available resource^32–34. In the short term, marble wastewater can have a positive effect on plant growth as it contains high quality nutrients such as calcium, potassium, iron and copper^24,25. However, farmers may not be fully aware of the potential long-term consequences of this practice, as marble waste also contains heavy metals^35–37. Heavy metals such as Al, As, Cd, Cr, Co, Ni, Pb, and Hg are toxic to plants even at low concentrations^38–40. The accumulation of these elements at relatively high concentrations in agroecosystems poses a major threat to soil microbes and disrupts the distribution and structure of the microbial community^4,41,42. Similarly, their toxic behavior in soil disrupts the morphological and physiological processes of crops by reducing their growth rate, nutrient imbalance, stomatal movement, and inhibition of photosynthesis⁴³. In human and animal body tissues, soil heavy metals accumulate through the food chain^44,45 and cause serious effects such as nephrotoxicity, neurotoxicity, hypertension, infertility, and carcinogenicity^20,46,47. Moreover, prolonged exposure to marble waste through inhalation and dermal contact poses serious health risks to local populations and workers, including conditions such as urolithiasis and abnormal lung function^{16,17,22,35,48}. Therefore, it is crucial to assess the risk and extent of soil heavy metal contamination near MPPs⁴⁹. Such assessments are essential for developing effective mitigation strategies to protect both the environment and public health^6,50.

Geochemical factors in soil, water and sediments contain valuable information on element ratios and should be treated as compositional data⁵¹. Using classical statistical tools directly on the input concentrations can lead to spurious results because different geometric rules govern compositional data⁵². Traditional methods such as Euclidean distance and covariance are unsuitable for geochemical concentrations, resulting in spurious elements^53,54. Geochemical data are positive vectors, represented by a constant sum, that are distinct from real Euclidean space⁵⁵. Log-ratio transformations such as the centered log-ratio (CLR), isometric log-ratio (ILR) and additive log-ratio (ALR) can handle compositional element concentration data^56–58. CLR provides an alternative to raw data but introduces collinearity, whereas ILR avoids collinearity^57,59. The application of CLR and CoDa principles, such as variation matrices, biplots, and dendrograms, allows the establishment of linear relationships between CLR variables and the differentiation of geochemical sources^42,60. Identifying unusual patterns related to heavy metal exposure in the area of interest is crucial²⁶.

Similarly, the evaluation of soil pollution status, identification of potential sources of contamination and appraisal of the prospective risks to human populations are prerequisites for the successful implementation of any action plan⁶¹. Soils near the marble industry are most likely polluted by marble effluents, marble slurry dumping and the production of marble powder from marble processing plants²⁵. Numerous studies have been conducted in Pakistan on heavy metal water pollution related to the marble industry^20,37. These studies have shown that marble industry effluents strongly affect water and sediment quality and cause the bioaccumulation of heavy metals in aquatic fauna and flora. Similarly, with respect to health risk assessment, most studies have typically focused on deterministic risk quantification of exposure to soil contaminated with heavy metals, which could either underestimate or overestimate the risk^62,63 In addition, owing to the uncertainty of heavy metal concentrations and the specific variability between individuals, identifying the most hazardous element for the population via a deterministic method is difficult^64,65. Fortunately, as the most common method of probabilistic risk analysis, Monte Carlo simulation can determine the likelihood of a risk being exceeded and help identify the priority metal for risk control^66–68. Therefore, a comprehensive investigation that considers the metal enrichment status and probabilistic risk assessment of heavy metals at the national level is valuable for understanding the contamination traits and contributions of various sources prior to taking efficient measurements to improve soil quality and human health^61,69. In addition, for the production of healthy crops and ensuring the health security of farmers, researchers, soil managers, livestock and common people, it is essential to study agricultural soils contaminated with marble waste^25,70–72. Therefore, the current study aimed to (i) evaluate the extent of heavy metals (HMs) contamination in agricultural soils withassociated human health risks using both deterministic and probabilistic health risk assessment modelsand (ii) identify potential sources of heavy metal contamination in soil via a compositional multivariate technique.

Materials and methods

Background of the study area

The present study was conducted between September and December 2019 on agricultural soil near MPPs in the district of Malakand (Fig. 1).

Fig. 1.

The study area map showing sampling sites was generated via ArcGIS software version 10.2.2.

Geographically, District Malakand is located between 34.3718° N latitude and 72.2420° E longitude and is the gateway to Dir, Chitral, Swat, Shangla, and Buner Districts and the currently merged, federally administered, tribal area of Mohmand and Bajaur. Malakand District covers an area of 952 km² with a population density of 475 inhabitants per km²⁷³. The land of Malakand District is fertile and surrounded by mountains rich in biodiversity. A large part of the agricultural land in the study area is irrigated by the Swat River (Fig. 1). Winters are very cold, and the temperature can decrease to – 2 °C in January. Summers are hot, and the temperature reaches 41 °C from June to July⁷⁴. The average annual rainfall is between 600 and 650 mm⁷⁵. Rice is an important cash crop with a total irrigated area of 4991 hectares⁷⁶. The valuable mineral reserves reported from several locations in the Malakand district are mica, quartz, marble, Malakand granite, and Dargai chromite⁷⁷. The coordinates of the individual sampling points were recorded by a global positioning system (Garmin eTrex 10). The study area map (Fig. 1) was generated via ArcGIS version 10.2.2.

Soil sampling and analytical methods

The agricultural soils in the study area are sandy loam. Soil samples were collected from 21 sites (Fig. 1) in triplicate at a depth of 0–20 cm via a hand auger following a quincunx sampling pattern⁷⁸. Finally, one kilogram composite sample for each site was prepared by thoroughly mixing the five sub soil samples and sealed in clean, labeled polyethylene bags^79,80. The soil samples were air dried, mechanically ground, sieved through 2 mm sieve and stored for further analysis⁸¹. 0.5 g of the dried soil sample was added in triplicate to a 50 mL Erlenmeyer flask with 26 mL of a triacid mixture of HNO₃, HCl_, and HF at a ratio of 9:3:1 and stored overnight⁸². The samples were completely digested on a hot plate by gently increasing the temperature to 155 °C until white dense fumes emerged⁸³. After cooling, the digested sample was diluted to 50 mL with distilled water, filtered through Whatman filter paper (0.45 μm) in a clean volumetric flask and stored at room temperature⁸⁴. The concentrations of essential elements and heavy metals in the extracted soil solutions were analyzed using inductively coupled plasma optical emission spectrometry (ICP‒OES; Thermo Scientific, iCAP 6000, UK).

Physicochemical analysis of the soil samples

A soil suspension was prepared using a soil to water ratio of 1:2.5 in double deionized water, then stirred and centrifuged at 3000 rpm for 4 h. The pH, electrical conductivity (EC), and total dissolved solids (TDS) of the soil samples were determined via a portable meter (Model PH-2012, WT01 China)⁸⁵.

Quality assurance and quality control

For quality assurance and quality control, special attention was given to sampling, preservation, transportation, storage and each experimental technique used in laboratory analysis⁸⁶. Analytical-grade acids and reagents provided by Merck (Germany) were used in this study. The standard reference material (SRM) used to construct calibration curves of the tested elements was IAEA-Soil-7. The reference soil material (IAEA-Soil-7) was prepared and analyzed according to the certified analytical standard quality control procedure⁸⁷. Eight different linear concentration standards were prepared. Each element with a good linear plot and a correlation coefficient greater than 0.999 was observed in the construction of standard curves. The traceability of the standards was verified via high-performance ICP‒OES against the corresponding SRM (ISO 2009). Each sample was analyzed in triplicate, and the mean value was used to interpret the results. The value of the relative standard deviation (RSD) for the analyzed elements was ≤ 2.0% (Table S1). The limit of detection (LOD) for each element was determined by digesting five blanks (Table S1). The recoveries were within the EPA limits of 92–123% (Table S1).

Assessment methods for soil heavy metal pollution

In the present study, the soil heavy metal pollution status was assessed through pollution indices developed by Muller^87,88; Hakanson⁹⁰ and Tomlinson et al.⁹¹.

Pollution Index and Nemerow Synthetic Contamination Index

The pollution index (Pi) reflects the pollutant level and primary exposure factors, serving as a tool and is used to assess heavy metal pollution in soil⁹². Pi can be calculated via the following equation.

$$\text{Pi}=\frac{{\text{C}}_{\text{i}}}{{\text{B}}_{\text{i}}}$$ 1

where C_i is the measured concentration of the ith element and B_i is the background value of the ith element in the world average shale⁹³. The Nemerow synthetic contamination index (NSCI) is used to represent the overall level of soil contamination by heavy metals. Soil HM contamination is a complicated process triggered by the synchronized action of numerous heavy metals⁹⁴. The NSCI can be calculated via Eq. 2.

$$\text{NSCI}=\sqrt{\frac{{\text{P}}_{\text{iave}}^2+{\text{P}}_{\text{imax}}^2}{2}}$$ 2

where P_iave is the mean value of P_i for the analyzed HM and P_imax is the maximum P_i value of the HM in a sample. The soil risk categories based on the P_i and NSCI are listed in Table S2.

Index of geo-accumulation (I_geo)

The index of geo-accumulation (I_geo), originally defined by Muller⁸⁸ is used to estimate the natural and anthropogenic factors of soil contamination⁹⁵. This method assesses the environmental impact while considering preindustrial and existing metal concentrations. The I_geo is calculated using the following equation, adapted from El Azhari et al.⁹⁶.

$${\text{I}}_{\text{geo}}={\text{log}}_2\left[\frac{{\text{C}}_{\text{n}}}{1.5\text{Bn}}\right]$$ 3

where C_n signifies the measured concentration (mg/kg) of metals in the soil sample and Bn is the geochemical background (mg/kg) of the tested metal. A factor of 1.5 compensates for background values of the respective metal owing to minor environmental and anthropogenic influences. On the basis of I_geo, Muller⁸⁸ proposed seven risk categories, as shown in Table S2.

Ecological risk index (Eri)

Eri measures the potential ecological risk of a single element in soils⁹⁰ and can be calculated as follows:

$${\text{E}}_{\text{r}}^{\text{i}}={\text{T}}_{\text{r}}^{\text{i}}\times {\text{P}}_{\text{i}}^{\text{i}}$$ 4

where ${\text{E}}_{\text{r}}^{\text{i}}$ is the potential ecological risk factor for the ith trace element and where $T_r^i$ is the coefficient of the ith element, which is 10 for As; 30 for Cd; 05 for Co, Cu and Ni; 01 for Mn and Zn; and 02 for Cr and V^90,97–100. ${\text{P}}_{\text{i}}^{\text{i}}$ is the pollution index of trace elements taken from Eq. (1). The evaluations of E_rⁱ are listed in Table S2.

Potential ecological risk index (PERI)

The comprehensive description of the ecological risk from a wide range of potentially toxic metals was determined through the application of the potential ecological risk index (PERI)^90,101. The PERI results were calculated via the following equation.

$$\text{PERI}=\sum _{\text{i}=1}^{\text{n}}{\text{E}}_{\text{r}}^{\text{i}}=\sum _{\text{i}=1}^{\text{n}}{\text{T}}_{\text{r}}^{\text{i}}\times {\text{P}}_{\text{i}}^{\text{i}}$$ 5

where PERI designates the cumulative ecological risk of trace elements, ${\text{E}}_{\text{r}}^{\text{i}}$ is the potential ecological risk factor for the ith trace element and ni is the number of trace elements, which is 09 in this study. $T_r^i$ is the coefficient of the ith element, and ${\text{P}}_{\text{i}}^{\text{i}}$ is the pollution index of the trace elements taken from Eq. 1. The evaluations of PERI are listed in Table S2.

Assessment of health risk

Exposure assessment (noncarcinogenic)

Human exposure to soil heavy metals (HMs) typically occurs through accidental ingestion, inhalation, and dermal contact with soil particles from water and air^102,103. In the present study, the noncarcinogenic health risks from individual heavy metal across the three exposures routes were calculated via the following equations ¹⁰⁴.

$${\text{ADD}}_{\text{ing}}=\frac{{\text{C}}_{\text{soil}\times {\text{R}}_{\text{ing}}\times \text{EF}\times \text{ED}}}{\text{BW}\times \text{AT}}\times \text{CF}$$ 6

$${\text{ADD}}_{\text{inh}}=\frac{{\text{C}}_{\text{soil}\times \text{I}{\text{R}}_{\text{inh}}\times \text{EF}\times \text{ED}}}{\text{BW}\times \text{AT}\times \text{PEF}}\times \text{CF}$$ 7

$${\text{ADD}}_{\text{dermal}}=\frac{{\text{C}}_{\text{soil}\times \text{SA}\times \text{AF}\times \text{ABF}\times \text{EF}\times \text{ED}}}{\text{BW}\times \text{AT}}\times \text{CF}$$ 8

where ADD_ing, ADD_inh, and ADD_dermal are the average daily doses (mg/kg/day) of HM through ingestion, inhalation, and dermal contact, respectively. All other factors and values used in Eqs. 6–8 to estimate the exposure level and risks are listed in Table S3.

HQ is the ratio of the average daily dose (ADD; mg/kg/day) and the reference dose (RfD; mg/kg/day) for a given toxicant. HQ values > 1 signify the occurrence of a noncarcinogenic risk¹⁰⁵ and can be calculated via Eq. (9).

$$\text{HQ}={\text{HQ}}_{\text{ing}}+{\text{HQ}}_{\text{inh}}+{\text{HQ}}_{\text{dermal}}=\frac{{\text{ADD}}_{\text{ing}}}{{\text{RFD}}_{\text{ing}}}+\frac{{\text{ADD}}_{\text{inh}}}{{\text{RFD}}_{\text{inh}}}+\frac{{\text{ADD}}_{\text{dermal}}}{{\text{RFD}}_{\text{dermal}}}$$ 9

where HQ_ing, HQ_inh, and HQ_dermal are hazard quotients (HQs) for the exposure routes, and RFD_ing, RFD_inh, and RFD_dermal represent the corresponding reference doses listed in Table S4.

Hazard index

The hazard index (HI) estimates the overall noncarcinogenic health risks from exposure to multiple potentially toxic HMs. HI is the sum of all HQs computed for individual HMs for a given route of exposure¹⁰⁶. The HI can be calculated via the following equation.

$$\text{HI}=\sum _{\text{K}=1}^{\text{n}}\text{HQ}={\text{HQ}}_1+{\text{HQ}}_2+\cdots ....+{\text{HQ}}_{\text{n}}$$ 10

where 1, 2, …, n is the HQ calculated for individual toxic HMs. The population is considered safe if HI < 1 and vulnerable if HI > 1¹⁰⁷.

Carcinogenic risk assessment

The carcinogenic risk (CR) assessment estimates the likelihood of an increased risk of cancer for a person due to lifetime exposure to potentially toxic elements. The cancer risk from lifetime exposure to As, Cr, Cd, and Ni was calculated across the three exposure routes via the following Eqs. ¹⁰⁵.

$$\text{CR}={\text{ADD}}_{(\text{ingestion},\text{inhalation},\text{dermal})}\times \text{CSF}$$ 11

$$\text{TCR}=\sum \text{Cancer risk}={\text{CR}}_{\text{ing}}+{\text{CR}}_{\text{inh}}+{\text{CR}}_{\text{dermal}}$$ 12

where CR is the cancer risk from the individual exposure pathway, and TCR is a cumulative cancer risk. CSF is the cancer susceptibility factor (mg/kg/day) of the HMs listed in Table S4. According to the USEPA¹⁰⁸, cancer risk values below 1 × 10⁻⁶ are considered insignificant, while those higher than 1 × 10⁻⁴ are likely to be detrimental. According to the classification of the WHO¹⁰⁹, As, Ni, Cd, and Cr (VI) compounds are being studied as Group 1 carcinogens; Pb compounds as Group 2 A; and Co and its compounds as Group 2B. In contrast, Cu, Zn, and Mn are not classified as carcinogenic, and their cancer risk has not been studied.

Probabilistic risk assessment

Uncertainty is inherent in risk assessments because of limited precise information, variability in environmental systems, and differences in individual human characteristics^110,111. To address these uncertainties, Monte Carlo simulation, a random sampling method based on probability theory, is commonly used to quantify variability and uncertainty in risk assessment¹¹². To achieve this, the probability distributions of the variables listed in Table S3 were used as input parameters to evaluate the probability functions for the health risks¹¹³. After performing 10,000 iterations, stable exposure distribution results were derived, and the values at quantiles 5th and 95th of the exposure distribution were used to assess the probabilistic risk¹¹⁴. A sensitivity analysis was conducted to evaluate the degrees of contribution of the exposure factors to the results. A positive value indicates a positive correlation between the exposure factor and health risks, while a negative value indicates a negative correlation^115,116.

Statistical analysis

Descriptive statistical characteristics, including the mean, standard deviation (SD), standard error (SE), skewness, kurtosis, and coefficient of variation (CV), of the 19 elements were calculated via SPSS software (version 25 for Windows). Monte Carlo simulation, sensitivity analysis and the best-fitting distribution for each factor were performed via Oracle Crystal Ball version 11.1.3.0. Compositional data analysis approaches such as centered log-ratio (clr) transformation, variation matrix, clr-biplot, coda-dendrogram, and factor analysis were applied to establish a linear relationship between the variables and differentiate the source of contamination in agricultural soils^42,59,60,117. The variation matrix (Table 4), clr-biplots (Fig. 14a,b), and Coda-dendrogram (Fig. 15) were generated via Coda-Pack software¹¹⁸ and the R package ‘‘compositions’’¹¹⁹. The factor analysis (Table 5) was conducted using the statistical software JMP version 16.0¹²⁰.

Table 4.

Normalized variation matrix for the chemical elements of Table 1: Bold values show linear association.

	pH	EC	TDS	Al	As	Ca	Cd	Co	Cr	Cu	Fe	K	Mg	Mn	Ni	P	Si	Sr	Ti	V	Zn	Zr
pH	0
EC	0.06	0.00
TDS	0.31	0.41	0
Al	0.12	0.17	0.27	0
As	0.14	0.15	0.25	0.89	0
Ca	0.04	0.12	0.24	0.45	0.33	0
Cd	2.19	2.17	1.65	1.8	0.89	1.29	0
Co	4.21	4.41	3.03	0.28	0.82	0.45	1.29	0
Cr	0.10	0.07	0.22	0.18	0.43	0.22	1.54	0.35	0
Cu	0.29	0.23	0.36	0.66	0.72	0.51	2.22	0.72	0.3	0
Fe	0.27	0.32	0.22	0.54	1.31	0.83	1.56	0.86	0.79	1.58	0
K	0.13	0.15	0.51	0.07	0.89	0.46	2.12	0.43	0.17	0.52	0.79	0
Mg	0.07	0.09	0.25	0.36	0.59	0.24	1.3	0.31	0.32	0.48	0.69	0.5	0
Mn	0.18	0.28	0.21	0.12	1.06	0.58	1.72	0.47	0.39	0.98	0.25	0.28	0.38	0
Ni	0.04	0.11	0.15	0.17	0.51	0.17	1.3	0.2	0.16	0.45	0.55	0.25	0.08	0.24	0
P	0.49	0.38	0.92	0.65	1.51	1.33	2.53	1.19	0.83	1.23	0.92	0.58	1.1	0.68	0.87	0
Si	0.59	0.95	0.85	1.02	1.77	1.16	1.96	0.82	1.41	2.09	1.12	1.14	1.11	0.93	0.93	1.64	0
Sr	0.12	0.14	0.32	0.23	0.61	0.23	1.69	0.26	0.14	0.38	1.09	0.18	0.34	0.55	0.2	1.06	1.17	0
Ti	0.33	0.38	0.56	0.42	1.57	1.07	1.93	0.8	0.74	1.54	0.25	0.68	0.89	0.33	0.68	0.93	1.59	1.01	0
V	0.10	0.09	0.16	0.3	0.45	0.27	1.47	0.28	0.09	0.42	1.07	0.34	0.39	0.63	0.25	1.17	1.57	0.12	0.9	0
Zn	1.04	0.97	1.09	3.11	1.85	2.31	1.48	3.19	2.82	3.83	2.27	3.17	2.74	2.7	2.51	2.93	2.6	3.10	3.33	3.11	0
Zr	0.27	0.26	0.24	0.88	0.32	0.44	0.85	0.93	0.61	1.2	0.94	1.0	0.66	0.87	0.52	1.61	1.75	0.83	1.22	0.66	1.44	0

Fig. 14.

clr-biplot (a) PC1 vs PC2 (b) PC1 vs PC3 of the elements in study area.

Fig. 15.

CoDa-dendrogram shows association of the data set of the geochemical composition of the soil generated by the clustering algorithm (Wards method).

Table 5.

Factor analysis (Varimax rotation) of the centered log ratio (Clr) transformed data set.

Variables	Factors
	F1	F2	F3	Communality
Al	0.89	0.27	0.15	0.908
As	− 0.81	0.37	− 0.24	0.860
Ca	− 0.52	0.40	0.33	0.611
Cd	− 0.70	− 0.41	− 0.13	0.844
Co	0.17	0.10	0.71	0.918
Cr	0.19	0.92	− 0.07	0.916
Cu	− 0.01	0.80	0.27	0.821
Fe	0.43	− 0.71	− 0.18	0.891
K	0.75	0.54	− 0.01	0.925
Mg	− 0.13	0.04	0.77	0.870
Mn	0.80	− 0.37	0.06	0.861
Ni	0.12	0.20	0.81	0.879
P	0.61	0.00	− 0.44	0.738
Si	0.14	− 0.55	0.38	0.902
Sr	0.12	0.78	0.38	0.880
Ti	0.68	− 0.31	− 0.19	0.930
V	− 0.06	0.84	0.15	0.945
Zn	− 0.57	− 0.51	− 0.52	0.933
Zr	− 0.70	− 0.08	− 0.33	0.726
Eigen value	5.84	5.15	2.37	–
% of total explained Variance	30.76	27.11	12.45	–
% of cumulative Variance	30.76	57.87	70.32	–

Ethical approval and consent to participate.

The current research study was approved by the Institutional Research Committee of the Department of Botany at Islamia College Peshawar. The article does not include any humans or animals as research objects. All methods were conducted following the appropriate guidelines outlined in the methods section.

Results and discussion

Assessment of physicochemical properties and heavy metal concentrations

The assessment of physicochemical properties and heavy metal concentrations in soils near marble processing plants showed significant variation in the studied parameters (Table 1). The soil pH, ranging from 7.3 to 9.0, with a mean of 8.3, indicates slightly alkaline conditions. Similarly, EC and TDS values ranging from 213.3 to 556.6 µs/cm and 2231 to 1072.67 mg/L, with mean values of 374.60 and 635.80, respectively, indicate variability in soil salinity levels. These variations in soil pH and salinity may be attributed to differences in water quality, soil composition, and the influence of industrial activities, such as marble processing, on the surrounding environment^23,121. High salinity levels could lead to soil degradation, impacting agricultural productivity¹²². Similarly, soil alkalinity impacts both physical and chemical properties, including texture, structure, color, infiltration, porosity, surface crusting, swelling, pH, and nutrient availability¹²³. Managing soil alkalinity and salinity through proper irrigation practices, soil amendments, and crop selection is crucial for maintaining soil health and ensuring successful agricultural productivity^124,125. Similar results were also reported by Eman et al.¹²⁶, which is consistent with our current results.

Table 1.

Descriptive statistics of physicochemical properties and heavy metal (HM) concentration (mg/kg) in agricultural soils near MPPs in District Malakand.

	Statistics								Background concentration
Variables	Mean	Min	Max	SE	SD	CV (%)	Skewness	Kurtosis	Average Shale	FAO
pH	8.3	7.3	9.0	0.18	0.48	5.80	− 1.20	3.10	−	6–8.5
EC (µs/cm)	374.60	213.3	556.6	31.33	84.25	29.20	0.16	0.71	−	400
TDS (mg/L)	635.80	223.0	1072.67	100.50	266.0	41.80	0.15	0.88	−	450
Al	10,914.30	5100	23,250	1067.8	4893.3	44.8	1.20	1.10	80,000	−
As	5.30	2.0	10.0	0.50	2.30	42.0	0.70	− 0.48	13	20
Ca	19,881.50	12,650	29,250	1163.5	5331.7	26.8	0.4	− 1.0	22,100	−
Cd	2.0	0.5	4.50	0.30	1.40	74.50	0.74	− 0.86	0.3	3
Co	6.0	2.0	11.0	0.50	2.40	41.30	0.46	− 0.21	19	50
Cr	27.0	13.0	40.0	1.60	7.50	27.90	− 0.51	− 0.35	90	100
Cu	20.10	6.50	40.50	2.30	10.50	52.30	0.60	− 0.66	45	100
Fe	4699.71	1075	13,200	555.2	2544.0	54.10	2.0	5.50	47,200	50,000
K	2105.70	734.5	3990.5	207.6	951.6	45.20	1.0	− 0.90	26,600	−
Mg	5420.71	2438.5	8050	365.8	1676.1	30.90	0.5	− 0.80	15,000	−
Mn	149.30	68.50	315.5	16.0	73.40	49.20	0.96	− 0.17	850	2000
Ni	18.40	12.50	27.0	0.90	4.10	22.40	0.40	− 0.56	68	50
P	292.0	65.5	805.5	45.2	207.3	71.10	1.1	0.60	700	−
Si	12.50	1.50	25.50	1.50	7.0	55.90	0.3	− 0.50	7300	−
Sr	35.14	13.50	60.50	2.90	13.40	38.0	0.47	− 0.57	300	−
Ti	600.50	125.50	1471.0	82.70	378.9	63.10	1.16	0.32	4600	−
V	39.21	16.50	59.0	2.70	12.50	32.0	− 0.27	− 0.45	130	−
Zn	5.80	0.50	19.0	1.30	6.10	106.70	1.21	0.08	95	300
Zr	3.0	1.0	4.50	0.20	1.10	37.50	− 0.01	− 1.01	160	−

SE; standard error, SD: standard deviation; CV: coefficient of variance.

The highest mean concentration (mg/kg) was observed for Ca (19881.50), followed by Al (10914.30), Mg (5420.71), Fe (4,699.71) and K (2105.70) (Table 1). Titanium (Ti) had a mean concentration of 600.50 (mg/kg), whereas phosphorous (P) recorded a mean concentration of 292.0 (mg/kg). The maximum concentrations (mg/kg) of Ca (29250) and P (805.5) exceeded the world average shale concentration (Table 1). These findings are consistent with those of Aukour and Al-Qinna³³, Ozcelik¹²⁷, Ahmad et al.¹⁶ and Noreen et al.¹²⁸, who reported elevated levels of these elements in soils and water impacted by industrial activities such as marble processing and mining. The high Ca concentrations likely reflect the presence of calcareous materials, which are common in regions with intensive marble processing¹²⁹. Elevated P levels could be linked to industrial discharge or the accumulation of phosphate-rich dust¹³⁰. Such elevated concentrations may have ecological implications, including potential disruptions to soil nutrient balances and risks of eutrophication in nearby water bodies¹³¹. The findings underscore the need for monitoring and managing these elements to mitigate potential environmental and agricultural impacts.

The mean concentrations (mg/kg) of Mn (149.30), V (39.21), Sr (35.14), Cr (27.0), Cu (20, 10) and Ni (18.40) were found to be within the limits of world average shale (Table 1), indicating relatively lower contamination levels¹⁰⁵. Similarly, the mean concentrations (mg/kg) of Si (12.50) Co (6.0), Zn (5.80), As (5.30), Zr (3.0), and Cd (2.0) were below the FAO permissible limits (Table 1), suggesting a lower risk to soil quality and ecosystem health¹³². However, the maximum concentration of Cd (4.50 mg/kg) exceeded the FAO limit of 3.0 mg/kg (Table 1), raising concerns about potential soil and human health risks¹³³. Similar results were also reported by Khan et al.³⁶ for Cd concentrations in different marble industries of Khyber Pakhtunkhwa, Pakistan, highlighting consistent patterns of cadmium accumulation in soils associated with marble processing activities in the region. Elevated cadmium levels have been associated with industrial emissions and waste, particularly in areas affected by mining and manufacturing activities¹³⁴. Cadmium is a toxic metal that can accumulate in the food chain, posing long-term health hazards¹³⁵.

The coefficient of variation (CV) quantifies the spatial variation in geochemical variables¹³⁶. In the present study, the coefficient of variation (CV) values varied from 22.40 to 106.70% (Table 1). Low variations were obtained for Ni, Ca, Cr, Mg, V, and Zr; moderate variations were obtained for Si, Fe, Cu, Mn, K, Al, As, and Co; and high variations were obtained for Ti, P, Cd and Zn (Table 1). According to Zuo¹³⁷, small CV values are associated with areas that have similar geochemical backgrounds, while large CV values characterize areas that have different geochemical backgrounds. The irregular distribution of elements in the current study area seems to be influenced by anthropogenic rather than natural influences¹³⁸. The statistical skewness and kurtosis indices were used to ascertain the normality or abnormality of the heavy metal distribution¹³⁹. The kurtosis values of Al, Fe, P, Ti and Zn were greater than zero, indicating steeper than normal distributions of these elements (Table 1). The skewness values of Al, K, Fe, P, Ti and Zn were greater than one, indicating that these elements were positively skewed to lower concentrations, highlighting potential environmental concerns^140,141. Understanding these associations helps in identifying the sources of soil contamination and in developing strategies to mitigate their impact on soil health and agricultural productivity¹⁴².

Assessment of soil heavy metal pollution

Pollution indices are among the most effective tools for assessing heavy metal contamination in soils, offering more insight than simply measuring total metal concentrations¹⁴³. In this study, single indices were calculated to evaluate the pollution levels of individual heavy metals in the soil. These indices are based on the total metal concentrations, along with geochemical background levels or preindustrial baselines, to provide a comprehensive assessment of soil contamination. Similarly, the Nemerow synthetic contamination index (NSCI) is employed to assess contamination levels across different sites¹⁴⁴. The results of the pollution index (Pi) (Table S5) show that the Pi values (1 ≤ 2) for As are indicative of slight pollution at sites S1-S3 and S14; larger variations in Pi values (2 ≤ 3) for As indicate moderate pollution at soil sites S6-S9 and S15–S18; and the Pi values > 4 are indicative of heavy pollution at soil sites S4, S11-S14, and S19-S21. Similarly, Pi values between 1 and 2 for Cd indicate slight pollution at soil sites S1, S6, and S8-S10; moderate pollution (2 ≤ Pi < 3) at sites S2–S5, S7, and S17; and heavy pollution (Pi > 4) at soil sites S11–S16 and S18–21. The Nemerow synthetic contamination index (NSCI) values for heavy metals varied from 1.20 to 4.11, with a mean of 2.55 (Table S5). The results indicate that 38% of the soil samples were slightly contaminated, 29% were moderately contaminated, and 33% were heavily contaminated. Overall, the soil exhibited moderate pollution with a mean NSCI value of 2.55 (Table S5). Contamination of the soil sites may be associated with emissions from road traffic, marble processing plants (MPPs), and car repair workshops, which release pollutants into the surrounding atmosphere^25,145,146. The results of the current study are consistent with the findings of Yüksel et al.¹⁴⁷, Weissmannova et al.¹⁴³, Wang et al.¹⁴⁴ and Chen et al.¹⁴⁸ who reported varying levels of soil and sediment contamination linked to industrial activities, traffic emissions, and other anthropogenic sources. These patterns align with established knowledge about the impact of such activities on environmental pollution, particularly in areas with similar sources of contamination.

The index of geoaccumulation (I_geo) originally proposed by Muller⁸⁸ provides valuable insights into the severity, range, and abundance of heavy metal contents in the soils of the studied sites¹⁴⁹. The I_geo results indicated that cadmium (Cd) had the highest value of 1.274, suggesting significant pollution, followed by arsenic (As) with a value of 0.663, indicating substantial contamination. Copper (Cu) and cobalt (Co), with I_geo values of 0.09 and 0.062, respectively, indicate moderate pollution. Vanadium (V), chromium (Cr), and nickel (Ni) exhibited relatively lower I_geo values of 0.061, 0.060, and 0.054, indicating comparatively lower contamination levels. Manganese (Mn), titanium (Ti), and strontium (Sr) had even lower I_geo values of 0.035, 0.026, and 0.024, suggesting minimal pollution. Likewise, Zinc (Zn) and zirconium (Zr) exhibited the lowest I_geo values of 0.012 and 0.004, indicating very low levels of contamination. The I_geo values of As, Cu, Co, V, Cr, Ni, Mn, Ti, Sr, Zn, and Zr categorized the soil sites as slightly to moderately polluted. The I_geo values between 1 and 2 for Cd are indicative of moderate pollution at soil sites S11 and S14–S16, whereas I_geo values between 2 and 3 at soil sites S12–S13 and S18–S21 are indicative of moderate to strong pollution. The I_geo values of Cd between 0 and 1 in the soil samples (S1–S10, S17) indicate unpolluted to moderate pollution. Similarly, the I_geo values of As (1 ≤ I_geo ≤ 2) indicate a moderate level of pollution at soil sites S13 and S19–S21. On the other hand, I_geo values of As between 0 and 1 classify the soil sites (S1–S12 and S14–S18) as uncontaminated to moderately contaminated (Table S6). The pollution pattern indicates that Cd and As are the primary contaminants of concern, with their levels varying from unpolluted to highly contaminated, depending on the site. Other metals, such as Cu, Co, V, Cr, and Ni, contribute to moderate pollution, while Mn, Ti, Sr, Zn, and Zr pose minimal environmental risk. The variation in I_geo values across different sites highlights the spatial heterogeneity in soil contamination, suggesting that targeted remediation efforts may be needed in specific areas with relatively high pollution levels¹⁴⁴.

Evaluation of the potential ecological risk of heavy metals

The potential ecological risk index provides an accurate assessment of heavy metal contamination in agricultural soil by considering metal toxicity relative to their reference concentrations in the Earth’s crust¹⁵⁰. Table S7 presents the results of the ecological risk index (Eri) and the potential ecological risk index (PERI) assessments for heavy metals (HMs), showing substantial variation in the mean Erⁱ values for individual heavy metals across the soil sites. The range of Erⁱ values for Cd was (10–450) > As (2.5–65.5) > Cu (0.72–4.5) > Co (0.53–2.89) > Ni (0.92–1.99) > V (0.25–0.91) > Cr (0.29–0.89) > Mn (0.08–0.37) > Zn (0.01–0.2) (Table S7). The Eri values (< 15) for Cu, Co, Ni, V, Cr, Mn, and Zn at the soil sites indicated a low ecological risk associated with these metals. The Eri values for As, ranging from 15 to 30, indicate a moderate ecological risk at soil sites S2-S3, S6-S9, and S14-S18. Eri values between 30 and 60 signify considerable risk at sites S4-S5, S10-S13, and S19-S20. Site (S21) poses a high ecological risk for As (60 ≤ Eri < 120), whereas site (S1) is associated with a low ecological risk (Eri < 15). Furthermore, considerable ecological risk of Cd was characterized for S6 and S8-S10, while a high ecological risk was characterized for S4–S5, S7, and S17. Very high ecological risk was identified at sites S11–S16 and S18–S21, whereas sites S1–S3 exhibited lower ecological risk for Cd (Table S7). The current findings suggest that cadmium is the most concerning heavy metal, with several sites facing very high ecological risks, followed by arsenic, which presents moderate to high risks at various sites. Other metals such as Cu, Co, Ni, V, Cr, Mn, and Zn, generally pose low ecological risks across the studied areas. These results are strongly supported by the studies of Din et al.¹⁴⁹, Chen et al.¹⁵¹ and Ajah et al.¹⁵², who also highlighted the significant ecological risks associated with cadmium and arsenic, while noting lower risks from other metals.

The PERI values reflect how different biological communities react to toxic substances, highlighting the potential ecological risks associated with hazardous elements¹⁵³. The potential ecological risk values in the analyzed soil samples ranged from 18.52 to 511.35, with a mean value of 217.49 (Table S7). Almost 42.85% of the soil sites were at very high ecological risk, while 23.80% were at considerable ecological risk. In addition, approximately 14.28% of the sites were associated with moderate ecological risk, and 14.28% were classified as low ecological risk sites. The main contributor to the PERI appeared to be Cd (83.8%), followed by As (14.4%) (Fig. 2). In contrast, the cumulative contribution of Co, Mn, V, Cr, Zn, Cu and Ni to the PERI was 3.0%. The contribution rate of heavy metals to the PERI correlates not only with their concentration but also with their toxicity response factors¹³⁸. In the current study, the pattern observed highlights Cd and As as the dominant hazardous elements in the soil, which pose the greatest ecological threats owing to their high toxicity and significant presence¹⁵⁴. This finding is consistent with the results of Lu et al.¹⁵⁵, Long et al.¹⁵⁶ and Kumar et al.¹⁵⁷, who also identified Cd and As as major contributors to potential ecological risks, especially in soils and sediments affected by human activities like mining. This study underscores the need for increased attention to Cd and As levels in agricultural soils because of their substantial ecological impact. The results also align with broader research, such as studies on sediment contamination in large freshwater lakes in China, where Cd was found to pose a moderate national-level risk due to significant emissions from human activities¹⁵⁸.

Fig. 2.

Contribution percentage (%) of heavy metals to potential ecological risk.

Assessment of potential health risk

A major global concern is human exposure to heavy metals due to environmental pollution. Heavy metals such as As, Cd, Cr, Pb, and Ni are highly toxic, classified as priority metals of significant public health concern¹⁵⁹. Even at minimal exposure levels, they cause damage to multiple organs and are classified as human carcinogens¹⁶⁰. Therefore, this study presents a health risk assessment of heavy metals for individuals working in marble processing plants (MPPs) and in surrounding agricultural soils, addressing both potential non-carcinogenic and carcinogenic health risks.

Noncarcinogenic health risk

Table 2 presents the average daily dose (ADD, mg/kg/day), hazard quotient (HQ), and hazard index (HI) for non-carcinogenic risks of heavy metals (HMs) through ingestion, inhalation, and dermal contact for both children and adults. A comparatively greater trend in ADD values (mg/kg/day) was observed for children than for adults for both the ingestion and inhalation routes (Table 2). The ADD_ingestion values (mg/kg/day) for the children decreased in the order of Al (1.40E-01) > Fe (6.01E-02) > Mn (1.91E-03) > V (5.01E-04) > Cr (3.44E-04) > Cu (2.57E-04) > Ni (2.35E-04) > Co (7.49E-05) > Zn (7.37E-05) > As (6.76E-05) > Zr (3.59E-05) > Cd (2.44E-05). Compared to children, adults exhibited similar patterns of average daily dose (ADD) for ingestion and inhalation of the studied heavy metals, but with lower intake levels. Conversely, higher values of ADD_dermal were obtained for adults than for children in the pattern of Al > Fe > As > Mn > V > Cr > Cu > Ni > Zn > Zr > Co > Cd (Table 2). According to this study, children appear to be more susceptible to health risks than adults are through the ingestion and inhalation routes. On the other hand, adults are more susceptible to such risks through the dermal route.

Table 2.

Mean values of average daily dose (ADD), hazard quotient (HQ), total hazard quotient (THQ), and cumulative hazard index (HI) of non-carcinogenic risks for adults and children.

Elements	ADD ingestion	ADD inhalation	ADD dermal		HQ ingestion	HQ inhalation	HQ dermal	HQ total
Adults
Al	1.50E-02	2.20E-06	1.70E-04		1.50E-02	1.83E-01	1.67E-02	2.15E-01
As	7.24E-06	1.06E-09	2.48E-06		2.41E-02	3.55E-06	5.26E-03	2.94E-02
Cd	2.61E-06	3.84E-10	2.97E-08		2.61E-02	3.84E-06	1.04E-03	2.71E-02
Co	8.02E-06	1.18E-09	3.66E-08		4.01E-04	2.07E-04	2.26E-03	2.87E-03
Cr	3.68E-05	5.42E-09	4.20E-07		1.23E-02	1.89E-04	4.77E-02	6.02E-02
Cu	2.75E-05	4.05E-09	3.14E-07		7.42E-04	1.01E-07	1.07E-02	1.15E-02
Fe	6.44E-03	9.47E-07	7.34E-05		9.20E-03	4.30E-03	1.21E-02	2.56E-02
Mn	2.05E-04	3.01E-08	2.33E-06		1.46E-03	2.10E-03	2.92E-03	6.48E-03
Ni	2.52E-05	3.71E-09	5.75E-08		1.26E-03	1.80E-07	1.47E-03	2.73E-03
V	5.37E-05	7.90E-09	6.12E-07		7.67E-03	1.13E-06	5.26E-03	1.29E-02
Zn	7.89E-06	1.16E-09	5.40E-08		2.63E-05	3.87E-09	6.06E-02	6.07E-02
Zr	3.85E-06	5.66E-10	4.39E-08		4.81E-02	7.07E-06	2.53E-04	4.84E-02
Total	2.18E-02	3.20E-06	2.50E-04	HI	1.46E-01	1.90E-01	1.66E-01	5.03E-01
Children
Al	1.40E-01	1.03E-05	1.37E-04		1.40E-01	8.55E-01	1.37E-03	9.96E-01
As	6.76E-05	4.97E-09	1.99E-06		2.25E-01	1.66E-05	1.62E-02	2.41E-01
Cd	2.44E-05	1.79E-09	2.39E-08		2.44E-01	1.79E-05	2.39E-03	2.46E-01
Co	7.49E-05	5.51E-09	2.94E-08		3.74E-03	9.64E-04	1.83E-06	4.71E-03
Cr	3.44E-04	2.53E-08	3.37E-07		1.15E-01	8.84E-04	5.61E-02	1.72E-01
Cu	2.57E-04	1.89E-08	2.52E-07		6.93E-03	4.70E-07	2.10E-05	6.95E-03
Fe	6.01E-02	4.42E-06	5.89E-05		8.58E-02	2.01E-02	8.41E-04	1.07E-01
Mn	1.91E-03	1.40E-07	1.87E-06		1.36E-02	9.81E-03	1.04E-03	2.45E-02
Ni	2.35E-04	1.73E-08	4.61E-08		1.18E-02	8.40E-07	8.54E-06	1.18E-02
V	5.01E-04	3.69E-08	4.91E-07		7.16E-02	5.27E-06	7.02E-03	7.86E-02
Zn	7.37E-05	5.42E-09	4.33E-08		2.46E-04	1.81E-08	7.22E-07	2.46E-04
Zr	3.59E-05	2.64E-09	3.52E-08		4.49E-01	3.30E-05	4.40E-04	4.49E-01
Total	2.03E-01	1.49E-05	2.01E-04	HI	1.37E + 00	8.87E-01	8.54E-02	2.34E + 00

The hazard quotient (HQ) values of Al, As, Cd, Cr, Co, Ni, Cu, V, Fe, Mn, Zr, and Zn for adults and children were < 1.0 (Table 2). Therefore, no apparent health risk from exposure to a single HM is indicated. However, in children, the cumulative HI value from ingestion, inhalation and dermal exposure to heavy metals was > 1.0 (Table 2), indicating that prolonged exposure of children to soils near the marble industry is harmful to their health²⁵. In addition, the probability distributions of total hazard index (HI_total) from ingestion, inhalation, and dermal exposure to heavy metals for adults (Fig. 3a) and children (Fig. 3b) were assessed via a probabilistic approach with Monte Carlo simulation. HI _total scores were found to be less than 1.0 at the mean, 5th, and 95th percentiles for both children and adults, indicating no apparent risk for the target population. The deterministic approach estimated higher HI_total scores for both children and adults (Table 2) compared to the probabilistic approach shown in Fig. 3a,b. Similar results were also recorded by Sanaei et al.¹¹⁴, where the HI values calculated through the deterministic approach were higher than those from probabilistic values. This disparity can be attributed to the overestimation of risk caused by the inherent uncertainties of deterministic analysis¹¹⁵.

Fig. 3.

Probability distribution of total hazard index (HI) for soil heavy metals presented in Tabe 2 for (a) adults and (b) children from ingestion, inhalation and dermal contact route.

Ingestion, inhalation, dermal and total cancer risk (TCR)

Cancer risk represents the probability of developing cancer from prolonged exposure to carcinogens¹⁶¹. The potential cancer risk from ingestion, inhalation, and dermal exposure to soil heavy metals (As, Cd, Co, Cr, and Ni) for adults and children is shown in Table 3. Commonly referenced acceptable risk levels range from 1.0E-06 to 1.0E-04 (i.e., one in 1,000,000 to one in 10,000)¹⁶². This study follows the USEPA¹⁶² guidelines, which define carcinogenic risk levels at 1.0E-04 and 1.0E-06, and considers these levels as potential indicators of adverse health risks for the exposed population. Through deterministic approach, the total carcinogenic risk (TCR) from soil heavy metal exposure ranged from 5.6E-10 to 6.9E-04 for children and from 1.2E-10 to 7.4E-05 for adults (Table 3). The carcinogenic risk from Cr, Ni, Cd and As ingestion was slightly higher in children than in adults and exceeded the USEPA limits of 1.0E-06 and 1.0E-04. The carcinogenic risk from inhalation and dermal exposure to these elements in children and adults was below the USEPA limits (Table 3). However, the TCR for Cr, Ni, Cd, and As was consistently higher in children than in adults across all exposure pathways (Table 3). Furthermore, the contribution of each element to the TCR decreased in the order of Cr > Ni > Cd > As, suggesting greater potential for the carcinogenic risk of Cr. The TCR for Co fell within the safety limits of 1.0E-04, indicating no apparent carcinogenic risk. Table 3 shows that Cr and Ni contribute significantly to the carcinogenic risk in children, and consequently, children are at higher risk than adults.

Table 3.

Estimated carcinogenic risks through ingestion, inhalation, and dermal pathways for children and adults.

	Adult					Children
Carcinogenic	CRingestion	CRinhalation	CRdermal	TCR	CRingestion	CRinhalation	CRdermal	TCR
As	4.8E-06	2.5E-07	6.8E-07	5.8E-06	4.5E-05	1.2E-07	5.4E-07	4.7E-05
Cd	6.9E-06	6.1E-11	–	6.9E-06	6.4E-05	2.8E-10	–	6.4E-05
Co	–	1.2E-10	–	1.2E-10	–	5.6E-10	–	5.6E-10
Cr	7.4E-05	1.3E-10	3.5E-08	7.4E-05	6.9E-04	6.0E-10	2.8E-08	6.9E-04
Ni	1.7E-05	4.4E-09	6.8E-08	1.7E-05	1.6E-04	2.1E-08	5.5E-08	1.6E-04
Total	1.0E-04	2.5E-07	7.8E-07	1.0E-04	9.5E-04	1.2E-06	6.3E-07	9.6E-04

In addition to point estimation of carcinogenic risk, a probabilistic approach was applied to assess the carcinogenic risk in the exposed groups. This approach considers the proper distribution of parameters, including the metal concentration, ingestion rate (IR), exposure duration (ED), exposure frequency (EF), and body weight (BW) (Table S3). The simulated carcinogenic risk results for As, Cr, Cd, Ni, and Co from ingestion, inhalation, and dermal exposure for adults and children are shown in Figs. 4, 5, 6, 7, 8, 9, 10 and 11. In adults, the probabilistic cancer risk values from ingestion for As (Fig. 4a), Cr (Fig. 4b), and Ni (Fig. 4c) at the 5th, mean and 95th percentiles exceeded the minimum acceptable limit of 1.0E-06. The Cd level at the 95th percentile from the ingestion route also exceeded the minimum threshold of 1.0E-06 (Fig. 4d). The probabilistic cancer risk from dermal exposure to As (Fig. 5a) and Cr (Fig. 5b) in adults was slightly higher than from ingestion at the mean, 5th, and 95th percentiles. In adults, the potential cancer risk for As (Fig. 6a), Cd (Fig. 6b), Cr (Fig. 6c), Co (Fig. 6d), and Ni (Fig. 6e) via inhalation route was below the USEPA limits of 1.0E-04 and 1.0E-06. However, the total cancer risk for adults (Fig. 7d) from ingestion (Fig. 7a), inhalation (Fig. 7b), and dermal contact (Fig. 7c) route at the 95th percentile, exceeded the upper limit of 1.0E-04. This finding indicates a potential cancer risk for the adult group.

Fig. 4.

Shows the probability of cancer risk for (a) As (b) Cr (c) Ni (d) Cd in adults via ingestion route.

Fig. 5.

Probability of cancer risk for (a) As (b) Cr and (c) Ni in adults through dermal route.

Fig. 6.

Probability distribution of cancer risk for (a) As (b) Cr (c) Co (d) Cd and (e) Ni in adults through inhalation route.

Fig. 7.

Cumulative probability of cancer risk for adults from (a) ingestion, (b) inhalation (c) dermal and (d) total cancer risk from the three pathways.

Fig. 8.

Cumulative probability of cancer riskk for (a) As (b) Cr (c) Cd (d) Ni in children through ingestion route.

Fig. 9.

Probability distribution of cancer risk for (a) As (b) Cr (c) Cd (d) Co (e) Ni in children through inhalation route.

Fig. 10.

Cumulative probability of cancer risk for (a) As (b) Cr (c) Ni in chidren through dermal route.

Fig. 11.

Cumulative probability of cancer risk for children from (a) ingestion (b) inhalation (c) dermal route and (d) total cancer risk from the three exposure pathways.

For children, at the 95th percentile, the probabilities of cancer risk resulting from soil ingestion exposure to As (Fig. 8a), Cr (Fig. 8b) Cd (Fig. 8c), and Ni (Fig. 8d) were 1.83E-04, 2.52E-04, 1.99E-05, and 5.27E-04, respectively. These values exceeded the safety limit of 1.0E-04 and were relatively higher than the corresponding risks for the target group. Similarly, the probability of cancer risk due to inhalation exposure to soil Cr at the mean and 95th percentiles (Fig. 9b) was greater in children than in adults and exceeded the threshold of 1.0E-04. In addition, the probability of cancer risk in children due to inhalation exposure to soil As (Fig. 9a), Cd (Fig. 9c), Co (Fig. 9d) and Ni (Fig. 9e) at the mean and 95t^h percentiles was within the permissible limits of USEPA. In children, the potential cancer risks of As (Fig. 10a) and Cr (Fig. 10b) through the dermal pathway were lower than in adults but exceeded the lower safety limits of 1.0E-06. For Ni, the dermal exposure risk both in children (Fig. 10c) and adults (Fig. 5c) was within the permissible limits. In addition, the total cancer risk in children (Fig. 11d) from the combined pathways of ingestion (Fig. 11a), inhalation (Fig. 11b) and dermal contact (Fig. 11c) at the mean, 5th, and 95th percentiles were 6.95E-04, 5.49E-04, and 8.62E-04, respectively, and were above the upper safety limit of 1.0E-04, suggesting a greater carcinogenic hazard for children than adults. The slightly elevated risk in children can be attributed to their higher intake rates relative to body weight, increased absorption efficiency, and developmental susceptibility¹⁶³. The exceedance of USEPA thresholds suggests that the levels of these metals pose a significant carcinogenic threat, necessitating urgent measures to mitigate exposure, particularly in vulnerable populations such as children¹⁶⁴. Wu et al.¹⁵³ also reported a high carcinogenic risk for children in the soil around electronics manufacturing sites.

In a previous study, the carcinogenic risk higher than 1.0E-04 through soil ingestion for adults and children was shown to predict health hazards for the target group¹⁶⁵. In the current study, the slightly higher risk value in adults through the dermal pathway aligns with findings from a similar study conducted in the mining areas of central China¹⁶⁵. Similarly, the cancer risk via the inhalation route was negligible for adults (Fig. 7b) and unacceptable at the 95^th percentile for children (Fig. 11b). Cao et al.¹⁶⁶ reported that consuming food grown on contaminated soil and direct skin contact with soil are significant human exposure pathways. Moreover, industrial and agricultural activities have led to the significant discharge of heavy metals into the surrounding soil. In accordance with Wang et al.¹⁶⁸ and Yang et al.⁶⁵, the use of pesticides and fertilizers significantly contributes to the occurrence of As and Cd in soil. Moreover, polluted water, smoke, and metal contaminated dust from smelting and metal processing can migrate into agricultural soils and increase the risk of cancer development⁶⁵.

In the current study, the carcinogenic risks from As, Cr, Cd, and Ni are probably linked to the combustion of fossil fuel, inorganic fertilizers, battery workshops, and the emission of marble dust^17,22,25,167. In addition, excessive accumulation of Cr in human tissues triggers gastric and lung cancer¹⁶⁸. Similarly, exposure to nickel can produce a variety of human health side effects, such as pulmonary fibrosis, nasal and lung cancer, and kidney and cardiovascular diseases¹⁶⁷. Similarly, soil arsenic exposure significantly associated with mortality rates from colon, stomach, kidney, lung and nasopharyngeal cancers¹⁶⁹. According to Suwatvitayakorn et al.¹⁷⁰, Cd has the potential to accumulate in the human body, resulting in acute and chronic effects on various organs and systems, with a particular impact on the kidneys and skeletal system. Therefore, it is equally important to carefully treat and monitor extremely noxious pollutants such as Cd, Cr, As, and Ni, even at very low concentrations in soil, air, and water.

Despite the carcinogenic or noncarcinogenic risk, children are more vulnerable to the potential health risk attributed to the existence of HMs in the soils surrounding MPPs. This suggests that children are more susceptible to adverse health effects from exposure to HMs, as they are more likely to ingest HMs orally from hand to mouth¹⁷¹. Among the different routes of exposure, the ingestion route contributes the most to the overall cancer risk in both children and adults, followed by the dermal route, and the inhalation route contributes very little to the overall cancer risk. Similar results were also reported by Saha et al.¹⁷², Kusin et al.¹⁷¹, Cai et al.¹⁷³ and Hao et al.¹⁶⁵.These findings strongly support our current results. Our study underscores the significant health risks from heavy metal exposure in agricultural soils near industries, emphasizing the need for focused risk control measures, as also suggested by Butt et al.⁴⁸, Khan et al.²⁵, Muhammad et al.¹⁷⁴ and Noreen et al.¹²⁸ in their respective studies.

Sensitivity analysis

A sensitivity analysis was performed to determine the importance of the variables involved in calculating noncarcinogenic and carcinogenic risk¹⁷⁵. The total hazard index (HI) < 1 at the mean, 5th, and 95th percentiles for both children (Fig. 3b) and adults (Fig. 3a) suggests that there is no apparent noncancerous risk for these two groups. However, sensitivity analysis of the total HI in adults indicated that heavy metal concentration (60.4%), exposure duration (30.4%) and exposure frequency (8.4%) were the most predictive noncancer risk factors (Fig. 12b).These results are in agreement with those of the study conducted by Fallahzadeh et al.¹⁷⁵. Similarly, for children, exposure duration (67.3%), heavy metal concentration (26.5%) and exposure frequency (6.4%) were important factors for noncancer risk (Fig. 12a). Sanaei et al.¹¹⁴ reported that the ingestion rate is the main factor influencing health risk, followed by exposure duration and metal concentration, while body weight is a less influential variable, partially supporting our study results. In addition, the assessment of the contributions of HMs to adult cancer risk through ingestion, inhalation, and dermal contact (Fig. 13b) indicated that As had the most significant impact, accounting for 83.6% of the total risk, followed by Cr (12.4%) and Ni (9%). For children, greatest contribution to the total cancer risk was simulated for Ni (59.5%), followed by Cr (21.6%) and As (18.3%) from the ingestion, inhalation and dermal exposure routes (Fig. 13a). The findings of this study align with those of Chen et al.¹⁶⁹, Yang et al.⁶⁵ and Mallongi et al.⁶⁶, who identified As, Cd, Cr, and Ni as the main contributors to total risk variance.

Fig. 12.

Sensitivity analysis of HI total for (a) chidren (b) adults.

Fig. 13.

Sensitivity analysis of total cancer risk for (a) chidren (b) adults.

Looking for associations and heavy metal sources

A normalized variation matrix used to show the linear association for the chemical components⁴². The analysis of the normalized variation matrix (Table 4) for the chemical parts (Table 1) reveals key associations between soil pH, EC, TDS, and the studied elements in agricultural soils near marble industries. These associations provide insights into the geochemical behavior of these elements and their interactions within the soil matrix¹⁷⁶. According to Boente et al.⁴² variations less than 0.2 suggest a proportional or linear relationship, while variations greater than 1.0 indicate a lack of proportionality or linear connection. In the current study, strong linear relationships were observed between soil pH and elements such as Ca (0.04), Ni (0.04), Mg (0.07), Cr (0.10), Al (0.12), V (0.10), Sr (0.12), K (0.13), As (0.14) and Mn (0.18). These values suggest a proportional or linear relationship, indicating that changes in soil pH significantly affect the solubility and availability of these elements¹⁷⁷. Therefore, soil pH plays a crucial role in determining the solubility and mobility of elements¹⁷⁸. The very strong relationship of calcium with pH indicates that calcium carbonate plays a crucial role in buffering the pH of the soil and stabilizing pH levels¹⁷⁹. However, acidic conditions often increase the solubility of these elements which can lead to higher concentrations in water and soil¹⁸⁰. EC is an important indicator of soil salinity and nutrient availability¹⁸¹. The strong linear relationships of EC with Cr (0.07), Mg (0.09), V (0.09), Ni (0.11), Ca (0.12), Sr (0.14), K (0.15), As (0.15) and Al (0.17) (Table 4) suggest the impacts of nutrient availability and soil chemical processes¹⁸². The relationships of EC with Ca, K, Mg, and Sr contribute significantly to the ionic balance, reflecting their role in soil structure and maintaining soil health and fertility^182,183. The presence of these elements in a balanced proportion ensures that the soil remains fertile and capable of supporting healthy plant growth¹⁸⁴. However, an imbalance, often indicated by a high EC, could suggest salinity issues, which can hinder plant growth and lead to soil degradation over time¹⁸⁵. The TDS is a measure of the total concentration of dissolved substances in soil water, which directly impacts soil water quality and its suitability for plant growth¹⁸⁶. The linear proportions of TDS with Ni (0.15) and V (0.16) highlight how increases in TDS can affect soil and water quality¹⁸⁷. Elevated TDS levels can indicate the presence of excessive salts or other dissolved substances, which may reduce the availability of water to plants, potentially leading to water stress and reduced plant growth^188,189. The associations between pH, EC, TDS, and the studied elements provide valuable insights into the potential sources of these elements and their environmental implications^12,26. For example, the strong relationship between pH and Ca suggests that calcium carbonate from marble industries could be a significant source, contributing to the observed soil chemistry²³. The relationships involving EC and TDS with elements such as Ni, Cr, and V may indicate industrial or agricultural inputs, such as fertilizers or industrial emissions, as potential sources¹⁹⁰. Similarly, values below 0.2 for Al versus K (0.07), Mn (0.12), Ni (0.17) and Cr (0.18) (Table 4) indicate that these element pairs are considered proportional and can be easily explained by the fact that where Al is present, it is likely to be accompanied by relatively high concentrations of K, Mn, Ni and Cr. Furthermore, the smaller contribution values for pairs such as Cr to V (0.09), Sr (0.14), Ni (0.16), and K (0.17), along with Sr to V (0.12), Ca to Ni (0.17), and K to Sr (0.18), underscore the existence of a linear binary relationship among these elements. This suggests that the concentration of one element in a pair can strongly influence the concentration of the other. For example, the linear relationship between Cr and V or Sr implies that fluctuations in chromium levels could directly impact vanadium or strontium concentrations. Similarly, the associations of Ca with Ni and K with Sr suggest a strong geochemical linkage between these pairs, likely due to co-deposition or similar environmental factors that influence their distribution. The observed associations may also indicate shared sources or pathways of these elements, which is consistent with the patterns identified in other studies by Boente et al.⁴², Khan et al.²³; and Khorshidi et al.¹⁹¹.

In addition, appropriate measures to find sources of HM contamination in agricultural soils are of paramount importance for developing sustainable management approaches and protecting soil quality¹⁹². Therefore, a clr-biplots (Fig. 14a, b) based on a centered logarithmic transformation (clr) was constructed to interpret the inter-elemental associations and source identification of HMs in agricultural soils^42,193. Figure 14a explains 62% of the variation between PC1 and PC2 and Fig. 14b explains 53% of the variation between PC1 and PC3 of the studied elements, indicating multiple element relationships related to natural geology and anthropogenic activities¹⁹⁴. Additionally, strong element associations between Mn-P-Ti, Al, and K are explained by the vicinity of their vertices and their rays pointing in a similar direction, indicating a combination of geogenic and anthropogenic origins¹⁹⁵. The vertices of Mn, P, and Ti are superimposed, and Ti was found to have the highest vector length, which could be associated with the co-precipitation of TiO₂, MnO, and P₂O₅ found in marble powder and fertilizers²⁵. Similarly, Cu, V, Sr, Cr, Mg, Ni, and Co form another association due to the close proximity of their vertices and the alignment of their rays in the same direction, as illustrated in Fig. 14a,b. This association is likely linked to vehicle emissions and industrial activities¹⁹⁴. A close relationship between Zn, As, Cd, and Zr forms a separate group and could be associated with smelter activities and the regular use of inorganic fertilizers^196,197. Fe and Si form another association because their rays also point in the same direction. However, in Fig. 15, Fe, Mn, Ti, and Ca shifted their positions and entered the elemental associations of Zn–Cd–As–Zr–Fe, Cu–Ti–Cr, Al–K–Mn, and Mg–Ca–Sr–Ni, indicating their affinity for these elements and sharing both natural and anthropogenic sources.

Factor analysis (Varimax rotation) of the centered log-ratio transform (clr) dataset was applied to discriminate potential geochemical source patterns and the associations and interrelationships between variables in the topsoil¹⁹¹. The three-factor model (Eigen > 1) explained 70.32% of the variance overall, with the contributions of each factor being F1 (30.76%), F2 (27.11%), and F3 (12.45%) (Table 5). A variable with factor loadings greater than 0.5 was included to explain the main components of each factor. The three-factor model captured a good correlation between the variables and their potential sources, showing a communality greater than 0.6 (60% variability). Factor one (F1) was dominated by Al, Mn, K, and Ti with strong positive loadings of 0.89, 0.80, 0.75, and 0.68, respectively, which could be attributed mainly to road transport and industrial activities¹⁹⁸. For factor two (F2), the Cr, V, Cu, and Sr loadings were 0.92, 0.84, 0.80, and 0.78, respectively, indicating that the marble industry was the source of origin²⁵. Factor three (F3) showed strong positive loadings, mainly with Ni (0.81), Mg (0.77), and Co (0.71), suggesting that river water could be another important source of pollution¹⁴⁵.

Cluster analysis (CA) is commonly used with PCA to verify the results of associations and HMs sources¹⁹³. For this purpose, the coda dendrogram (Fig. 15) is chosen as a visual method to identify possible relationships, pairwise or involving larger groups¹⁹⁹. The short vertical bars in the coda dendrogram correspond to linear associations between the clustering variables. Therefore, three groups of elements can be identified: (i) Al, K, Mn, P, Mn, Si, Fe, and Ti, which are not highly toxic and are of geogenic origin, and (ii) Cr, Ni, V, Cu, Co, Mg, Ca, and Sr, which become toxic in excess quantity, indicating anthropogenic and lithogenic sources²⁰⁰. The third group consists of As, Cd, Zn, and Zr, which are potentially toxic and derived from the extensive application of chemical fertilizers¹⁹⁷. According to Weissmannová et al.¹⁴³, different elements grouped in the same cluster originate from a common source. The results of the coda dendrogram (Fig. 15) fully agreed with the results of the clr-biplots (Fig. 14a,b), variation matrix analysis (Table 4), and factor analysis (Table 5). These observations reinforce that industrial effluents, intensive fertilizer use, vehicle emissions, and irrigation with contaminated water are key sources of toxic elements, verifying findings from Jehan et al.¹⁴⁵, Khan et al.²⁵, and Sarwar et al.²⁰¹.Therefore, these findings strongly support our current study results (Fig. 16).

Fig. 16.

Pollution sources of heavy metals and potential health risk assessment.

Conclusion

In the current study, the concentration of macro elements and heavy metals in soil near marble processing plants (MPPs) in district Malakand, was assessed to determine the extent of contamination and potential health risks. The concentrations of macro elements and heavy metals in the soil decreased in the following order: Al > Mg > Fe > K > Ti > P > Mn > V > Sr > Cr > Cu > Ni > Si > Co > Zn > As > Zr > Cd. The maximum concentrations of Ca, P and Cd exceeded the acceptable limits of the FAO and the world average shale. The Pollution Index (Pi) and Nemerow Synthetic Contamination Index (NSCI), indicated different levels of contamination from slight to heavy contamination. The Geoaccumulation Index (I_geo) indicated significant contamination of the soil sites by Cd and As, moderate contamination by Cu, Co, V, Cr, and Ni, and minimal contribution from Mn, Ti, Sr, Zn, and Zr. The potential ecological risk assessments revealed that Cd poses the highest ecological risk, followed by As. The pollution indices and compositional exploratory data analysis revealed that the spatial variability of soil contamination is primarily influenced by factors including industrial activities, ro ad traffic, and marble processing plants. The health risk assessments indicated elevated non-carcinogenic risks for children, with the total cancer risk fromthe potential toxic metals above safe levels, particularly for children. The key factors influencing health risks included metal concentration, exposure duration, and frequency, with As and Cr being significant contributors to cancer risk. The long-term monitoring of heavy metal contamination and its health impacts should be implemented, enforcing regulations on emissions from industrial and vehicular sources should be enforced, and soil remediation techniques such as phytoremediation and biochar application should be employed to reduce contamination levels. Additionally, public health programs should be developed to limit children’s exposure to contaminated soils, and alternative agricultural practices and land-use policies should be explored to minimize exposure and risks.

Author contributions

Asghar Khan: Conceptualization, investigation, methodology, data curation, formal analysis, writing—original draft, software. Muhammad Saleem Khan: Conceptualization, Project administration, Resources, Supervision, Validation, Writing—review & editing. Qaisar Khan: Data curation, Writing—review & editing. Kishwar Ali: Formal analysis, Writing—review & editing. Fazal Hadi: Formal analysis, Validation, Writing—review & editing. Ghulam Saddiq: Validation, resources, Writing—review & editing.

Funding

No grant for the current research was awarded by any organization or research institute.

Data Availability

All the data are included in the article.

Ethics approval and consent to participate

We, the authors of this manuscript, hereby declare that the content presented in this research article is original and has not been published elsewhere. All contributions made by coauthors are appropriately acknowledged.

Consent for publication

We declare that this manuscript is original, has not been published before and is not currently being considered for publication elsewhere.

Competing interests

The authors declare no competing interests.

Supplementary Information

The online version contains supplementary material available at 10.1038/s41598-024-72346-7.