Introduction

Modern agriculture is considered one of the main sources of heavy metal pollution (Jayakumar et al., 2021; Soleimani et al., 2023). The term "heavy metal" refers to a group of metallic and metalloid elements associated with environmental pollution, such as arsenic (As), cadmium (Cd), chromium (Cr), lead (Pb), mercury (Hg), nickel (Ni), copper (Cu) and zinc (Zn) (Alloway, 2022). Its presence in the environment is worrying due to its inability to biodegrade, its ability to accumulate in biological tissues, and its harmful effects on human and animal health, including chronic toxicity, mutagenicity, teratogenicity, and carcinogenicity (Dasharathy et al., 2022; U.S. Environmental Protection Agency [EPA], 1986). In particular, As, Cd and Pb have been classified as carcinogenic to humans (Flora & Agrawal, 2017; Rahman & Singh, 2019; Rezaei et al., 2019).

The accumulation of these elements in agricultural soils can be promoted by the continuous application of agrochemicals and organic byproducts (Alengebawy et al., 2021). This phenomenon is also observed in winegrowing systems -whether conventional, integrated, or organic- where improper input management can alter the distribution and accumulation of heavy metals in the soil (Alloway, 2022).

Grape derived products -such as juices, macerated drinks, spirits, pisco, and wines- are widely consumed worldwide for their nutritional and functional value, as they provide vitamins, minerals, antioxidants, fiber, phytochemicals, and carotenoids, vital components for protecting the body against chronic diseases. The raw material for these products is must, which is the liquid obtained from pressing grape bunches (Figure 1), and is processed according to the desired product type.

Despite the nutritional benefits of grape products and their importance in the international market, there is concern about their potential role as a source of impurities and contaminants, such as heavy metals. To our knowledge, this is the first study to analyze heavy metal content in winegrowing areas of the Arequipa region of Peru. Therefore, the objective of this study was to evaluate the potential ecological risk associated with the accumulation of heavy metals in agricultural soils of four wineries in the Arequipa region (La Joya, Majes Tradición, CIEPA-Majes, and San Isidro), as well as the health risks associated with the level of contamination of Moscatel grape must.

Materials and methods

Study area

The study was carried out in four winegrowing areas located in the Arequipa region, Peru: La Joya (16° 27’ 27.06’’ S and 71° 48’ 16.62’’ W, at 1 626 m a. s. l.), Majes Tradición (16° 13’ 33.24’’ S and 72° 26’ 36.42’’ W, at 464 m a. s. l.), Agricultural Research, Education and Production Center (CIEPA) - Majes of the National University of San Agustín (UNSA) (16° 19’ 51.48’’ S and 72° 13’ 22.32’’ W, at 1 440 m a. s. l.) and San Isidro (16° 34’ 13.86’’ S and 71° 55' 45.84'' W, at 1 660 m a. s. l.) (Figure 2). The variety grown in these sites is Moscatel, which is one of the most important and oldest in the region, and is recognized for its high aromatic capacity.

In addition to sampling, surveys were conducted among technical staff at the production centers to gather information related to agronomic management and physicochemical conditions.

Soil and must sampling

From a bibliographic review on the metals with the greatest presence in agricultural soils and their potential as a contaminant (Aytop et al., 2023; Mendoza et al., 2020; Soleimani et al., 2023), eight elements were selected for monitoring: As, B, Cd, Cu, Fe, Hg, Pb and Zn.

Soil and must sampling was carried out during the 2023 harvest: February 16 in La Joya, February 15 in Majes Tradición, March 24 in CIEPA-Majes and March 31 in San Isidro. Three composite samples were taken at each site, consisting of three soil subsamples (1 kg each, at a sampling depth of 0-20 cm) and must (1 L each, obtained from pressing 30 kg of grapes) (Figure 1).

Laboratory analysis

The concentration of heavy metals was determined by inductively coupled plasma mass spectrometry (ICP-MS) (ICP-MS 7900, Agilent Technologies, USA). The analyses were performed by Bhios Laboratorios S.R.L., accredited by the National Institute of Quality (INACAL). The results were expressed in milligrams per kilogram of fresh weight of the composite samples (mg∙kg^-1). The analyses were performed by triplicate.

Ecological risk assessment due to soil contamination

Based on the analysis results, the level of ecological risk derived from the presence of heavy metals was estimated using three indicators: contamination factor (CF), pollution load index (PLI), and potential ecological risk (PER). The procedure for calculating each index based on the results of heavy metal concentrations in the samples and international reference values is detailed below.

Contamination factor

The CF was used to measure the relationship between the content of each heavy metal in the soil and the reference geographic value (value in uncontaminated soil), which is an indicator of contamination from anthropogenic activities associated with a single heavy metal. The calculation was performed using the following equation (Hakanson, 1980):

where C _soil is the concentration of each metal in the soil samples and C _ref is the reference concentration of each metal in uncontaminated soils. In Peru, there are no studies on reference values for metals in soils; therefore, the values reported by Taylor and McLennan (1995) for As, B, Cd, Cu, Fe, Pb, and Zn, corresponding to 1.5, 15, 0.098, 25, 35,000, 20, and 71 mg∙kg^-1, respectively, and by McKeague and Wolynetz (1980) for Hg (0.06 mg∙kg^-1), were used. According to Hakanson (1980), contamination levels can be classified as low (CF < 1), moderate (1 ≤ CF < 3), considerable (3 ≤ CF < 6), and very high (CF ≥ 6).

Pollution load index

The PLI allows soil quality to be assessed based on the degree of heavy metal contamination. The PLI is calculated using the nth root of the product of the contamination factors for each metal:

where CF is the contamination factor and n is the number of heavy metals considered. According to Tomlinson et al. (1980), the interpretation of the result of this equation can be: presence of contamination (PLI > 1) or no metal contamination (PLI < 1).

Potential ecological risk

The PER was determined using the method proposed by Hakanson (1980), which allows the evaluation of the potential toxic effect of heavy metals in the soil using the following equations:

Where PERI _i is the potential ecological risk index of a metal i, Tr _i represents the toxicological response factor of metal i and CF _i is the contamination factor of metal i. The toxicological response factor considered for As, B, Cd, Cu, Fe, Hg, Pb and Zn were 10, 2, 30, 5, 1, 40, 5 and 1, respectively (Hakanson, 1980; Mirzaei-Aminiyan et al., 2018; Xu et al., 2020). The PERI was classified as: low (PERI < 40), moderate (40 ≤ PERI < 80), considerable (80 ≤ PERI < 160), high (160 ≤ PERI < 320) and very high (PERI ≥ 320). For its part, the PER value was interpreted as: low (PER < 95), moderate (95 ≤ PER < 190), high (190 ≤ PER < 380) and very high (PER ≥ 380) (Hakanson, 1980).

Health risk assessment

To estimate potential health risks, per capita wine consumption values in Peru (6.03 g·day^-1) were used (Promotora dels aliments catalans [PRODECA], 2023), since the must is mainly used for winemaking and its consumption is well documented in various countries. Non-cancer risks were assessed for all metals considered, and carcinogenic risks for As, Cd, and Pb. The assessment was conducted using the model proposed by the United States Environmental Protection Agency (EPA, 1989), which considers three mechanisms of contaminant translocation: ingestion, dermal absorption, and inhalation. This study focused on the ingestion mechanism of toxic elements associated with grape must consumption.

Non-carcinogenic risks

The health index (HI) method, widely used due to its simplicity and usefulness in identifying potential risks (Quispe et al., 2021), was applied. The HI is calculated by summing the risk quotients (HQ) of the chemical elements to which a person is exposed, according to the following equations:

where ED is the exposure duration (30 years for adults), EC is the element concentration (μg·g^-1), EF is the exposure frequency (365 days∙year^-1), IR is the ingestion rate, RfD is the oral reference dose of each element (μg·g^-1∙day^-1), BW is the body mass (70 kg for adults) and Atn is the average exposure time for the non-carcinogenic element (10 950 days). The RfD values for chronic exposure used in this study were (μg·g^-1∙day^-1): As = 0.0003, B = 0.2, Cd = 0.001, Cu = 0.04, Fe = 0.7, Hg = 0.0003, Pb = 0.0036, and Zn = 0.3 (de Miguel et al., 2007; EPA, 1986; Taylor et al., 2023). HI values are categorized as follows: negligible (HI < 0.1), low impact (0.1 ≤ HI < 1.0), moderate impact (1.0 ≤ HI < 4.0), and very high risk (HI ≥ 4.0) (EPA, 1989).

Carcinogenic risks

Carcinogenic risk (CR) estimates the cumulative lifetime probability of developing cancer from exposure to a given dose of the carcinogen. This parameter was derived from the following equation:

where ED is the exposure duration (70 years for adults), Atn is the average time of exposure to carcinogens (25 550 = 365 days·year^-1 × ED), and SF is the slope factor (mg·kg^-1∙day^-1: As = 1.5, Cd = 0.038, and Pb = 0.0085) (EPA, 1986; Mohammadi et al., 2024). The interpretation of the CR values is: no significant risk (CR < 1×10^-6), low risk (1×10^-6 ≤ CR < 1×10^-5), moderate risk (1×10^-5 ≤ CR < 1×10^-4), high risk (1×10^-4 ≤ CR < 1×10^-3) and very high risk (CR ≥ 1×10^-3) (EPA, 1989).

Statistical analysis

The results for heavy metal content in soil and must were subjected to analysis of variance and a Tukey means comparison (p < 0.05). Additionally, Pearson correlation analyses (p < 0.05) were performed to explore associations between variables. All statistical analyses were performed using RStudio version 4.4.1 (R Core Team, 2024). The reported values correspond to the average of three replicates per sample.

Results and discussion

Heavy metals in the soil

The management and fertilization characteristics in the four production areas were similar, with the exception of La Joya, where an additional 15 t∙ha^-1 of cattle manure was applied (Table 1). Several studies have shown that soil physicochemical properties, such as cation exchange capacity (CEC), organic matter (OM) content, electrical conductivity (EC), and pH, directly influence the mobility and bioavailability of metals (Antoniadis et al., 2017). In the case of pH, it has been reported that variations in this parameter -induced by the accumulation of metals such as Pb, Cu, and Zn- can alter soil biological activity by inhibiting the activity of mycorrhizal fungi and soil enzymes (Angon et al., 2024).

Table 1. Agronomic management and principal physicochemical characteristics of the soil from four winegrowing zones in the Arequipa region of Peru.

The physicochemical changes associated with heavy metals range from modifications in soil structure to OM degradation (Angon et al., 2024; Antoniadis et al., 2017; Eckhard et al., 2011). Although not the aim of this study, it should be mentioned that the effects on microorganisms are of concern, as they play a fundamental role in long-term soil stability and structure (Wang et al., 2022).

The correlation analysis between the concentration of metals in the soil samples and the physicochemical characteristics showed strong positive associations between EC and the concentrations of Cd (r = 0.82; p = 0.0011) and Fe (r = 0.94; p < 0.0001), and moderate negative associations with Pb (r = -0.66; p = 0.0202) and Cu (r = -0.62; p = 0.0330). The OM was positively associated with Pb (r = 0.79; p = 0.0024) and Cu (r = 0.77; p = 0.0034), and negatively with As (r = -0.58; p = 0.0469), Cd (r = -0.64; p = 0.0252) and Fe (r = -0.86; p = 0.0004) (Figure 3), while pH presented weak or moderately negative associations, with the exception of Pb, Cu and Zn, where the correlations were slightly positive (Figure 3). The diversity of results reflects the complexity of chemical interactions in soil. Regarding potential synergies between metals, strong correlations were observed between Cu and Pb (r = 0.89; p = 0.0001), as well as between Fe and Cd (r = 0.85; p = 0.0004).

Heavy metal concentrations in soils showed significant differences among the study areas for Cd, Cu, Fe, Pb, and Zn (Table 2). The highest values for Hg, Cu, and Pb were reported in La Joya, possibly associated with the higher OM content and soil pH conditions. In contrast, the highest concentrations of Fe, Cd, and As were observed in Majes Tradición.

Table 2. Analysis of variance and comparison of means of heavy metal concentration (mg∙kg^-1) in soils from four winegrowing areas in the Arequipa region, Peru.

In agricultural soils, the permissible limits for Pb are the highest, followed by Zn and Cu, while Cd has the lowest limits (Alengebawy et al., 2021). This means that the accumulation of Cd in the soil, even at lower concentrations, is more toxic than that of Cu, Zn, and Pb. According to the results obtained, the limits established by the Ministerio del Ambiente del Perú (MINAM, 2017) for As (50 mg∙kg^-1), Cd (1.4 mg∙kg^-1), Hg (6.6 mg∙kg^-1), and Pb (70 mg∙kg^-1) were not exceeded in any of the evaluated areas. However, compliance with these limits does not guarantee the absence of ecological risk, since the interaction and cumulative effect of multiple metals in the same site can pose a significant risk to the ecosystem.

Statistical analysis revealed significant differences only in the levels of Cd, Cu, and Fe, with the soils of the Majes Tradición winery showing the greatest accumulation. In general, the concentration of heavy metals in the soils was observed to be distributed as follows: Fe > Zn > B > Cu > As > Pb > Cd > Hg (Table 2). Except in Majes Tradición, the presence of the metals classified as the most dangerous to the environment (Hg, Pb, and As) was detected (EPA, 1989).

Potential ecological risk

The CF revealed that all four production areas are contaminated by As and B, with higher concentrations in Majes Tradición and CIEPA-Majes (Table 3). These areas share common characteristics, such as a drip irrigation system and low OM contents (Table 1). According to Antoniadis et al. (2017), these conditions may favor the accumulation of As and B, since in soils with a slightly alkaline pH (average pH 7.43) these elements are not absorbed by plants because they tend to form insoluble borates and, therefore, are not available to plants. Furthermore, the efficiency of drip irrigation can reduce the leaching of these metals into subterranean layers, which contributes to their accumulation in the first centimeters of the soil.

Table 3. Assessment of potential ecological risks due to the accumulation of heavy metals in soils of four winegrowing areas in the Arequipa region, Peru.

In studies conducted in vineyards in France, Brazil, Croatia, and Spain, accumulation of Cu has been reported due to the continuous application of copper fungicides (Zołnowski et al., 2013). This accumulation can compromise the microbiological quality of the soil (Alengebawy et al., 2021; Angon et al., 2024) and decrease the productive potential of vineyards (Jayakumar et al., 2021). Although in the present study Cu was the fourth most abundant heavy metal (Table 2), its concentration does not represent a significant ecological risk in any of the areas evaluated (Table 3).

The PLI revealed that the four wine production centers have a moderate level of contamination, while the PERI identified a constant risk associated with B in all four zones, as well as a considerable and moderate specific risk from cadmium (Cd) in La Joya and CIEPA-Majes, respectively (Table 3). The PER showed that none of the zones is at low ecological risk, with a high level of risk observed in La Joya and a moderate level in the other wine production centers (Table 3).

Various studies indicate that, in addition to pesticides, organic waste such as compost, animal manure, and humus are also sources of pollution, as they tend to accumulate heavy metals such as Pb, Ni, Cd, Cr, Cu, and Zn (Canet et al., 1998). However, the use of organic amendments improves the biological, physical, and chemical quality of soils (Eckhard et al., 2011; Wang et al., 2022); in addition, it exerts a medium- and long-term buffering effect on the concentrations of some metals such as B (Jayakumar et al., 2021; Xu et al., 2020).

In the Arequipa region, the use of organic waste has been promoted by the technical and academic community due to its benefits on the physicochemical and biological properties of soil in arid and semi-arid areas. Therefore, it is controversial to consider them as the main source of contamination. However, fertilizers used in agricultural production are likely the most significant source of heavy metals in the area. The results of this study are expected to serve as a basis for future research to confirm this hypothesis and establish sustainable agronomic practices.

Heavy metals in the must

The metal content of viticulture products, especially wine, is regulated by national laws and trade organizations to preserve organoleptic properties and food safety (Agostini & Daudt, 1997; Orescanin et al., 2003; Tariba, 2011). The presence of heavy metals in grape must can come from both natural and anthropogenic sources. Among natural sources, soil constitutes the main reservoir of metals, which are absorbed by the vine roots and transported to the fruit; this provides the largest amount of metal ions present in the grape (Agostini & Daudt, 1997; Prabagar et al., 2021). On the other hand, anthropogenic sources include agronomic management practices (fertilization, phytosanitary treatments, etc.) and environmental pollution (Mahlungulu et al., 2023; Orescanin et al., 2003; Tariba, 2011).

Prabagar et al. (2021) reported the presence of Zn, Ni, Cu, As, and Pb in grape berries grown in Sri Lanka. This partially coincides with the findings of the present study, where the presence of at least three heavy metals was found in the must from the four winegrowing areas in the Arequipa region (Table 4). However, only the concentration of Cd in the must from La Joya exceeded the permissible limits established by the International Organization of Vine and Wine (OIV, 2024). In the case of B, although there are no explicit limits for this element in wine, the concentrations in La Joya and Majes Tradición exceeded the reference value established by the World Health Organization (WHO) (2.4 mg∙L^-1) (Health Canada, 2020).

Table 4. Analysis of variance and comparison of means of heavy metal concentration (mg∙kg^-1) in must samples from four winegrowing areas in the Arequipa region, Perú.

Some heavy metals, such as Pb and Cd, have the ability to mimic essential nutrients and compete with them for root uptake sites (Tariba, 2011). However, it is important to study each production area and evaluate the specific dynamics of soil-plant interactions, since, for example, in samples from San Isidro -the area with the highest soil Cd content (0.356 mg∙kg^-1)- no Cd was detected in the must (Tables 3 and 4).

The concentration of B, Cd, Pb, and Zn in the must varied significantly between the study areas. In general, the order of abundance was: B > Fe > Zn > Cu > As > Cd > Pb > Hg (Table 4), which mostly coincides with the distribution observed in the soils (Table 2). The higher concentration of B, Zn, and Fe, both in the soil and in the must, could be explained by their affinity with clay particles and soil OM (Eckhard et al., 2011), so they have a greater probability of being absorbed and transported by the roots to the fruits. The absorption of heavy metals by plants represents a potential health risk when they enter the food chain through the consumption of contaminated agricultural products (Jayakumar et al., 2021; Tariba, 2011).

Health risk assessment

Health problems caused by exposure to heavy metals are increasing due to their bioaccumulation capacity and their entry into the food chain (Orescanin et al., 2003; Shaheen et al., 2016; Tariba, 2011). Although Cu, Zn, and Fe are necessary for various physiological processes, their intake at high concentrations can cause acute or chronic toxicity (EPA, 1986; Nag & Cummins, 2022; Rezaei et al., 2019).

The non-carcinogenic HQs calculated for the musts revealed that the main contributors were As, B, Fe, Hg, Pb, and Zn (Table 5). The SI indicated a very high health risk in the samples from La Joya, Majes Tradición, and CIEPA-Majes, while the samples from San Isidro were classified as medium risk. These results are consistent with those reported for wines from different regions of the world (Tariba, 2011).

Table 5. Assessment of potential health risks from the presence of heavy metals in Moscatel grape must from four winegrowing regions in the Arequipa region of Peru.

The CR values associated with As showed high levels in the musts from La Joya, Majes Tradición, and CIEPA-Majes (Table 5). Similar results have been reported for vegetables such as leek, mint, and cilantro grown in the Arequipa region (Quispe et al., 2021), so it is important to evaluate the magnitude of the risks associated with this metal in other agricultural products. Regarding Cd and Pb, the results indicate that La Joya presents high and medium risks, respectively, while in Majes Tradición both were classified as low (Table 5). Although only Cd exceeded the MRLs in La Joya (Table 4), the CR, IS and RC values suggest significant risks to consumer health (Table 5).

While these findings are worthy of attention, it is important to interpret them with caution, as the values obtained correspond to the raw material (must), not the finished product. Furthermore, it should be noted that per capita wine consumption rates in Peru are low compared to other countries. However, the identification of both carcinogenic and non-carcinogenic risks justifies the need to conduct heavy metal traceability studies in finished products to accurately determine the risks to consumers' health.

Conclusions

The results showed the presence of potential ecological and health risks associated with the accumulation of heavy metals in soils and Moscatel grape musts in the four winegrowing regions of the Arequipa region. Although risk levels ranged from low to very high, San Isidro had the lowest heavy metal contamination.

The findings of this work are expected to motivate research to determine the sources of contamination in the region and the implications for the microbiological quality of the soil and for direct consumption products.