Health Risks Due to Metal Concentrations in Soil and Vegetables from the Six Municipalities of the Island Province in the Philippines

This paper investigated the health risks due to metal concentrations in soil and vegetables from the island province in the Philippines and the potential ecological risks. The concentrations of Cd, Cr, Cu, Fe, Mn, Ni, Pb, and Zn in vegetables and soil in six municipalities of the province were analyzed using the Inductively Coupled Plasma Optical Emission Spectrometry (ICP-OES) Perkin Elmer Optima 8000. It was recorded that all metal concentrations in the soil, except for Cd, exceeded the soil quality standard (SQS). The concentration of Fe and Mn was highest among other metals. The Nemerow synthetical pollution index (Pn) in all soil samples was under Class V which means severe pollution level. Likewise, the risk index (RI) of soil ranged from high to very high pollution risk. Most of the metal concentrations in the vegetables analyzed also exceeded the maximum permissible limit (MPL). All health hazard indices (HHIs) were less than 1, which means potential low non-carcinogenic risk to human population by vegetable consumption. However, it was found that concentration of Cr and Ni in vegetables is a potential health hazard having concentrations exceeding the maximum threshold limit. A 75% temporary consumption reduction of bitter melon, eggplant, sweet potato tops, and string beans produced from two municipalities may be helpful in reducing exposure to target metals. Additional studies are needed to confirm this recommendation. Spatial correlation analysis showed that six out of target metals had datasets that were more spatially clustered than would be expected. The recorded data are useful for creation of research direction, and aid in developing strategies for remediation, tools, and programs for improving environmental and vegetable quality monitoring.


Introduction
The island province of Marinduque in the Philippines, with farming as the major source of income of the population, had been recognized as one of the country's largest copper reserves. Hence, copper mining operations were carried out in Marinduque since Marinduque is an island province, with about 53,344 households in the Republic of the Philippines with coordinates of 13.4767° N, 121.9032° E. This island province has a tropical climate with annual mean temperature of 27 °C. The island does not have a distinct wet and dry season as precipitation occurs all year.
The vegetable and soil samples were collected from six municipalities of Marinduque, namely Boac (B), Buenavista (BV), Gasan (G), Mogpog (M), Sta. Cruz (S), and Torrijos (T). Sampling locations were distributed spatially over the six municipalities of the island province. Coordinates of every sampling point were recorded and plotted using Geographic Information Systems as shown in Figure 1. Generally, the land use of the sampling points from the island province is mainly agricultural [39,40].
Four types of vegetables samples were collected in the individual towns' public markets and household farmyards. A total of 96 samples were assessed, analyzed, and evaluated in triplicates. The vegetables were chosen from the top produced and consumed vegetables in Marinduque, namely string beans (Vigna unguiculata), sweet potato tops (Ipomoea batatas), bitter melon (Momordica charantia), and eggplant (Solanum melongena). Likewise, composite soil samples were collected from each municipality across the province with a stainless-steel auger at 0-30 cm depths. The selection of sites was undertaken on the basis of where vegetables farms are located. Coordinates were recorded using a handheld Garmin global positioning system (GPS) model Montana 650. All vegetable and soil samples were put in a zipper bag, labeled, and transported to the laboratory for treatment and analysis. Sample collection started in 2018 and ended in 2019 completing one year season. Four types of vegetables samples were collected in the individual towns' public markets and household farmyards. A total of 96 samples were assessed, analyzed, and evaluated in triplicates. The vegetables were chosen from the top produced and consumed vegetables in Marinduque, namely string beans (Vigna unguiculata), sweet potato tops (Ipomoea batatas), bitter melon (Momordica charantia), and eggplant (Solanum melongena). Likewise, composite soil samples were collected from each municipality across the province with a stainless-steel auger at 0-30 cm depths. The selection of sites was undertaken on the basis of where vegetables farms are located. Coordinates were recorded using a handheld Garmin global positioning system (GPS) model Montana 650. All vegetable and soil samples were put in a zipper bag, labeled, and transported to the laboratory for treatment and analysis. Sample collection started in 2018 and ended in 2019 completing one year season.

Samples Preparation and Analysis
Vegetable samples were washed, thinly cut/sliced, and dried in a dehydrator at 68 • C for 7-8 h at the MSC-Mapua joint research laboratory in Boac, Marinduque. Dried samples were pulverized into fine particles and stored at room temperature in a properly labeled disposable Petri dish until digestion. Utensils and laboratory equipment were cleaned with soap and water, then rinsed with 1.5% v/v nitric acid and distilled water every use per sample. Additionally, the soil samples were dried at 68 • C for 3 h, ground to a fine powder using an agate mortar and pestle, homogenized, and passed through a 10-mesh sieve before analysis. The concentration of metals such as cadmium (Cd), chromium (Cr), copper (Cu), iron (Fe), lead (Pb), manganese (Mn), nickel (Ni), and zinc (Zn) was analyzed in triplicates using an Inductively Coupled Plasma Optical Emission Spectrometer (ICP-OES) Perkin Elmer Optima 8000. The ICP-OES used in this analysis has detection limits of 0.05, 0.1, 0.1, 0.5, 0.07, 0.5, 0.1, 0.04, and 2.0 for Cd, Cr, Cu, Fe, Hg, Mn, Ni, Pb, and Zn, respectively. The EPA Method 3050B [41], 200.3 [42], and 6010c [43] were followed in the digestion and analysis of samples. The digestion and analysis were conducted at Mapua University's Yuchengco Innovation Center's Wet Laboratory in Manila, Philippines.

Evaluation of Metal Pollution in Soil
The level of soil pollution was assessed by the single-factor (P i ) and the Nemerow synthetical pollution index (P n ) [44]. The P i is a useful tool for determining the pollution degree of a certain heavy metal in soil using Equation (1).
where C i is the concentration of metal i, and S i is the soil quality standard value [45,46]. A higher value of P i indicates more serious pollution, and soil is considered as polluted when P i is greater than 1 (Table 1). To better describe the level of soil pollution when P i is less than 1, the P n can evaluate a range of heavy metal toxicity levels, as shown in Equation (2).
where P imax is the maximum value of the P i and P iave is the average value of the P i . The grading standard for P i is shown in Table 1.

Potential Ecological Risk Index (pERI)
To measure the adverse effects of soil contamination, the risk index (RI) approach established by Hakanson [47] was used. The approach requires four hypotheses elaborated by the work of Agarin et al. [3]. The potential ecological risk index (pERI) considers the concentrations of metals, their environmental impact, and biotoxicity. The RI is computed by following Equation (3).
where T i is the toxic response factor for metal i [44]. The grading standard for pERI is also categorized in Table 1.

Potential Human Health Risk Assessment
The potential human health risk from the consumption of vegetables was estimated by calculating the estimated daily intake (EDI) of metals, health hazard index (HHI), and target cancer risk (TCR) [3]. The EDI in Equation (4) was used to calculate the level of exposure to a particular metal in vegetables via ingestion/oral route [48].
where, E f is exposure frequency (365 days/year); E D is the exposure duration (years); C m is the metal concentration (mg kg −1 ); F IR is the daily average vegetable consumption (g person −1 day −1 ) [49,50]; C f is the conversion factor for fresh vegetable weight to dry weight (0.085) [51,52]; B w is the reference body weight for an adult (kg); and T A is the average exposure time. The exposure duration considered was 70 years [53][54][55]. This is based on the average life expectancy of the Philippine population [53,54]. The Philippine population F IR for bitter melon, eggplant, string beans, and sweet potato tops are 7.42, 14, 6.99, and 7.78 g person −1 day −1 [49,50], respectively. The target hazard quotient (THQ) provides an estimate of the amount of risk associated with pollutant exposure [56]. It was used to evaluate the potential health risks of metal consumption through vegetables [19]. Equation (5) was used to estimate the THQ values of the population as a result of consuming contaminated vegetables as described by Chen et al. [57] and Ezemonye et al. [58].
where, R f D is the oral reference dose (mg kg −1 day −1 ) [56,59]. The R f D is an estimate with uncertainty spanning an order of magnitude of a daily oral exposure to the human population that is likely to be without an appreciable risk of deleterious effects during a lifetime [60]. This number was based on the no-observed adverse effect level (NOAEL) with the consideration of uncertainty factor. If the estimated daily intake of contaminated vegetables of a population is greater than the R f D, then the occurrence of associated health risks increases. Should the THQ be less than one, it is considered safe or has noncarcinogenic consequences. Should THQ be larger than one, the probability of danger increases [19,56]. The HHI [3] was determined by adding the individual metal target hazard quotient values [61] using Equation (6).
Carcinogenic risk assessment, on the other hand, is an estimation of the cumulative probability of developing cancer throughout a lifetime by exposure to a unit dose of a probable carcinogen [34]. The possibility of cancer risks in the studied vegetables through intake of carcinogenic heavy metals was estimated using the cancer risk (CR) [62] as shown in Equation (7). Then, the TCR from heavy metals intake, which can have a carcinogenic effect depending on the exposure level [63], was calculated using Equation (8).
where, CSF is the cancer slope factor and defined as the risk generated by a lifetime average amount of 1 mg kg −1 day −1 of carcinogen chemical and is contaminant specific [63][64][65][66]. Therefore, cancer risk was expressed in terms of incremental lifetime cancer risk, i.e., the probability of developing cancer over a 70-year lifetime period due to a 24 h exposure to a potential carcinogen [55]. The values of CSF for Cd, Cr, Ni, and Pb are 0.38, 0.5, 1.7, and 0.0085 mg kg −1 day −1 [63], respectively. In contrast, if the TCR exceeds the maximum threshold value of 1.00 × 10 −4 , there is a high risk of carcinogen exposure [63,[65][66][67].

Spatial Correlation Analysis
ArcGis Desktop 10.8.1 ArcPro 2.8 [68] was employed to create spatial distribution maps and spatial correlation analysis. Moran's I (index) was used to determine the autocorrelation of datasets to produce spatial distribution maps of metal concentrations in soil and vegetables. Moran's I is classified as positive, negative, and no-spatial autocorrelation. Positive autocorrelation occurs when values clustered together or appear near to each other. Negative spatial autocorrelation occurs when datasets are dispersed, i.e., dissimilar values occur next or near to each other. Should similar and dissimilar values appear randomly, then no spatial autocorrelation occurred.

Statistical Analysis
Pearson's correlation and hierarchical cluster analysis (HCA) was used using IBM SPSS Statistics Version 26.0 to look at the specific relationships between the various metals found in soil and vegetables as well as how certain metals influence their concentration.

Results and Discussion
Subsequent sections elaborate and discuss the results and the potential health implications of recorded concentrations of metals. Data presented below are helpful in determining research directions, creating strategies in soil remediation, soil quality monitoring programs for vegetable quality monitoring, creation of in situ detection tools for metals specific for vegetables and crops, and among other related techniques helpful for reducing health risks. The non-detection of Cd in soil in this study has been attributed to the soil acidity and uptake of some plants [71]. The study of Usman [71] specifically mentioned that plants remediate Cd by phytoextraction and it accumulates in the shoot. Phytoremediation activities in Marinduque were recorded during the period of 2006-2017 [25][26][27][28][29][30]. Other studies in Vietnam [72], Malaysia [73], and Indonesia [74] where there were mining activities recorded an elevated concentration of Cr in soil with no record of Cd concentration.

Concentration of Metals in Soil
It shows that Fe and Mn had the highest concentration of metals in soil across the six municipalities. The high concentration of Fe in the soil can be attributed to the mine tailings from the previous milling of sulfide ores, i.e., copper and zinc, which may have high levels of pyrite (iron sulfide) [75,76]. Likewise, continuous subsurface flow of metals within the island are associated to the two abandoned open mine pits that are located at the higher elevation of the island province [3]. Metal concentration varies spatially as illustrated in Appendix A. Table 2 shows the mean concentrations of metals in the soil samples highlighting the SQS [45,77]. The trend of metal concentrations across the island province is shown in Table 3.

Evaluation of Metal Pollution in Soil
In order to obtain the status of soil pollution from each municipality, the values of P i and P n were calculated using the Nemerow pollution index (NPI). The NPI illustrates the degree of probable pollution a metal contributes, and/or how several metals probably pollute a target environmental medium. As illustrated in Figure 2, all mean P i values of soil from each municipality were greater than five except for Cd which was not detected. This implies that the soil from all municipalities belongs to Class V with severe pollution level (Table 1) and is quite alarming [78]. Heavy metal pollution of soil is a serious environmental hazard, particularly in locations where soils intended for agricultural practices are adjacent to sources of pollution such as mining [44].
As illustrated in Figure 3a, the P n values in all soil from each municipality ranged from 61.35 to 1055.83, which are greater than three indicating that all soil samples were under Class V with severe pollution level ( Table 1). The P n of soil in all municipalities had the following order: BV > S > M > T > G > B. In Figure 3b, the higher pERI was found in the soil of Boac and Mogpog in comparison with Buenavista, Gasan, Sta. Cruz, and Torrijos. More specifically, in Mogpog, the single ecological risk index (E i r ) of Cu was greater than 1200, (E i r = 1968.03), indicating a higher contribution of Cu to RI. The RI of soil in Boac, Buenavista, and Mogpog exceeded 1200; hence, there is very high pollution risk. On the other hand, the RIs of Gasan, Sta. Cruz, and Torrijos were between 600 and 1200 indicating a high pollution risk too. The high pollution risk of the soil due to metal pollution can adversely impact plant growth [79], microbial diversity [80], biota [81], and humans whose exposure includes incidental ingestion, inhalation, and dermal contact [82]. humans whose exposure includes incidental ingestion, inhalation, and dermal contact [82].

Concentration of Metals in Vegetables
Firstly, it must be emphasized that metals such as Fe and Zn are essential micronutrients for human health and other living organisms. The degree of concentrations of metals in vegetables presented below are useful in making strategies, epidemiological studies, further research on edible agricultural yields, and creating more tools for vegetable quality monitoring to protect human health. It will also aid comprehensive soil and vegetable quality monitoring. The

Concentration of Metals in Vegetables
Firstly, it must be emphasized that metals such as Fe and Zn are essential micronutrients for human health and other living organisms. The degree of concentrations of metals in vegetables presented below are useful in making strategies, epidemiological studies, further research on edible agricultural yields, and creating more tools for vegetable quality monitoring to protect human health. It will also aid comprehensive soil and vegetable quality monitoring.
The mean concentrations (mg kg −1 ) of metals in the vegetable samples collected from Marinduque are shown in Figure 4a  It is useful to note that the leafy vegetable (sweet potato tops) had accumulated highest metal concentrations. Similar findings have been reported by Luo et al. [86], cating that leafy vegetables accumulate metals more than non-leafy vegetables. The tribution trend of metal concentration in vegetables is shown in Table 4. It shows tha had the highest concentration, among target metals in all vegetables across the six mu ipalities.  It is useful to note that the leafy vegetable (sweet potato tops) had accumulated the highest metal concentrations. Similar findings have been reported by Luo et al. [86], indicating that leafy vegetables accumulate metals more than non-leafy vegetables. The distribution trend of metal concentration in vegetables is shown in Table 4. It shows that Zn had the highest concentration, among target metals in all vegetables across the six municipalities.

Potential Human Health Risk of Metals by Ingestion
The EDI of metals through consumption/oral intake of vegetables by the population in Marinduque is presented in Table 5. The eggplant contributed most to the population EDI given that it has the highest F IR among the vegetables tested. It was followed by sweet potato tops then string beans and lastly bitter melon as illustrated in Figure 5. The metals of concern in vegetables were Cd, Cr, Fe, Mn, Pb, Zn, and Cu. The EDI of individual metals as a result of vegetable consumption was in the order of Zn > Fe > Mn > Cu > Ni > Cr > Pb > Cd. Bar plots are illustrated in Figures 5 and 6 to visualize the concentration of metals in each vegetable type and the potential contribution of specific vegetables to the HHI of each municipality.   The THQ and HHI values were used to interpret the metals' possible health hazards. The computed potential HHI of metals present in vegetables is shown in Table 7 which evaluated the cumulative effect of ingesting a variety of potentially harmful metals from several vegetables [63]. Ingestion of the vegetable samples recorded to have THQ values ranged from 3.85 × 10 5 to 9.86 × 10 2 Cd, 1.89 × 10 6 -4.86 × 10 Cr, 9.65 × 10 -1.04 × 10 Fe, 2.80 × 10 -4.43 × 10 Mn, 3.00 × 10 -7.36 × 10 Ni, 1.44 × 10 -3.45 × 10 Pb, 6.18 × 10 -1.05 × 10 Zn, and 2.22 × 10 -9.40 × 10 Cu. All of the HHI values were less than 1, hence, there is a potential low non-carcinogenic human health risk to the human population from vegetable consumption [87]. Further, it was     It can be seen from Table 6 that the TCR of Ni due to vegetable consumption in Buenavista (6.08 × 10 4 ), Gasan (8.51 × 10 4 ), and Sta. Cruz (1.26 × 10 4 ) exceeded the maximum threshold value of 1.00 × 10 4 [66,67,83,84]. This indicates potential risk of developing cancer if the population has a daily intake of vegetables shown in Table 5. The corre- The TCR associated with exposure to heavy metals such as Cr, Cd, Ni, and Pb from vegetable consumption was calculated using EDI data (Table 5), CSF [63,78,79], and TCR data shown in Table 6. The computed EDI ranges for various vegetables were 3.85 × 10 −8 -9.86 × 10 −5 Cd, 5.67 × 10 −9 -1.46 × 10 −4 Cr, 6.76 × 10 −5 -7.31 × 10 −3 Fe, 3.92 × 10 −5 -6.20 × 10 −4 Mn, 6.00 × 10 −9 -1.47 × 10 −4 Ni, 1.28 × 10 −7 -1.38 × 10 −4 Pb, 1.85 × 10 −8 -3.16 × 10 −3 Zn, and 8.89 × 10 −5 -3.76 × 10 −4 Cu.
The THQ and HHI values were used to interpret the metals' possible health hazards. The computed potential HHI of metals present in vegetables is shown in Table 7 which evaluated the cumulative effect of ingesting a variety of potentially harmful metals from several vegetables [63]. Ingestion of the vegetable samples recorded to have THQ values ranged from 3.85 × 10 −5 to 9.86 × 10 −2 Cd, 1.89 × 10 −6 -4.86 × 10 −2 Cr, 9.65 × 10 −5 -1.04 × 10 −2 Fe, 2.80 × 10 −4 -4.43 × 10 −3 Mn, 3.00 × 10 −7 -7.36 × 10 −3 Ni, 1.44 × 10 −7 -3.45 × 10 −2 Pb, 6.18 × 10 −8 -1.05 × 10 −2 Zn, and 2.22 × 10 −3 -9.40 × 10 −3 Cu. All of the HHI values were less than 1, hence, there is a potential low non-carcinogenic human health risk to the human population from vegetable consumption [87]. Further, it was found that the highest HHI values in vegetables were in Buenavista and Gasan. It is being emphasized that THQ and HHI were significantly affected by the EDI and R f D values. These values are estimated values with uncertainty. It can be seen from Table 6 that the TCR of Ni due to vegetable consumption in Buenavista (6.08 × 10 −4 ), Gasan (8.51 × 10 −4 ), and Sta. Cruz (1.26 × 10 −4 ) exceeded the maximum threshold value of 1.00×10 −4 [66,67,83,84]. This indicates potential risk of developing cancer if the population has a daily intake of vegetables shown in Table 5. The corresponding TCR of Cr in Buenavista (1.89 × 10 −4 ) and Gasan (2.03 × 10 −4 ) is also higher than the maximum threshold value. It was observed that ingestion of Cr and Ni in vegetables may pose a cancer risk to the population of Marinduque. This recorded result is similar to the findings of Gebeyehu et al. [63]. Chronic exposure to high amounts of Cr causes malignancies of the gastrointestinal tract, respiratory and central nervous systems [74,76]. However, it is considered that the ingested dose of vegetable with elevated concentration of metals is not always equal to the absorbed metal concentration dose as a fraction could potentially be excreted from the body [88]. Hence, regular excretion is important. Additionally, it is noted that the TCR was based on EDI shown in Table 5, CSF, body weight, and 365 days a year of exposure frequency. Therefore, these variables significantly affect the potential to acquire carcinogens and the likelihood of cancer risk during a lifetime exposure. These elevated metal concentrations in the vegetables that posed HHR to the population were associated to the existence of two abandoned open mine pits which are located at the higher elevation of Marinduque [69]. Groundwater flow [21] through porous media, and runoff overflow in the island province potentially caused the elevated concentration of metals in soil that the vegetables may have absorbed from the soil. Given the above information, it is advised to exercise a 75% temporary consumption reduction of bitter melon, eggplant, string beans, and sweet potato tops produced in the municipalities of Buenavista and Gasan. This is to avoid Cr and Ni TCR. Therefore, the proposed monthly average consumption of bitter melon, eggplant, string beans, and sweet potato tops is 6.9, 7.8, 7.4, and 14 g person −1 , respectively. This is to reduce exposure and the probability of cancer occurrence by vegetable ingestion. The TCR at 75% reduction is shown in Table 7. The TCR of Ni in Buenavista and Gasan still exceeds the threshold value despite a 75% consumption reduction. Hence, a comprehensive study covering quality of other vegetables produced in the island will be helpful to identify the location and types of vegetables that must be consumed by the public without posing TCR.
Based on the calculated TCR estimation, a 75% reduction of vegetable consumption would leave Ni alone as concern in the municipalities of Buenavista and Gasan. The TCR of Ni in Buenavista and Gasan at 75% reduction of vegetables consumption is 1.52 × 10 −4 and 2.13 × 10 −4 , respectively. The recommended 75% consumption reduction is temporary as mentioned above. This recommendation, apart from being temporary, is only for bitter melon, eggplant, string beans, and sweet potato tops produced specifically from the municipalities of Buenavista and Gasan. Hence, the effects of the 75% reduction (referred above) do not violate the "food security (FS)" goal defined by the United Nations' Committee on World Food Security. The FS definition states that "all people, at all times, have physical social, and economic access to sufficient, safe, and nutritious food that meets their food preferences and dietary needs for an active and healthy life." The vegetables produced by the above-mentioned municipalities were not safe based on the detected metal concentrations that were beyond the maximum permissible limit [62,66,[82][83][84].

Relationships of Metals in Soil and Vegetables
Pearson correlation analysis was carried out to determine the degree of interrelation and association of the metals in soil and vegetables [89]. As shown in Table 8, the correlation coefficients of Ni-Cr, Zn-Pb, and Cu-Zn were 0.917, 0.940, and 0.921 (p < 0.01), respectively. The value of (p < 0.01) expresses statistical significance of the relationship and association of metals in soil and vegetables. HCA was further used to determine the affinity and behavior of metal in the soil across the province. Figure 7 shows the dendrogram results generated by HCA for metals. The following two primary clusters were identified: (1) Cr-Ni-Cd-Pb-Zn-Cu-Mn and (2) Fe. It produced results that were similar to those of the Pearson correlation analysis.  As shown in Table 9, the correlation coefficients of Cr-Cd, Mn-Fe, Ni-Fe, Ni-Mn, Pb-Cd, and Pb-Cr were 0.986, 0.629, 0.605, 0.852, 0.977, and 0.977 (p < 0.01), respectively. These positive significant correlations between these various metals in soil and vegetables demonstrated that they commonly interact, cooperate with each other to promote potential uptake of vegetables from soil, and are a common source of pollution [41,74,75]. Furthermore, spatial correlation technique was employed to analyze further the relationship of metal concentrations in vegetables that were collected in proximity to where the soil samples originated. Table 10 elaborates the generated Moran's Index that measures the spatial autocorrelation of soil and vegetables. This table illustrates that the concentration of Cr, Ni, Pb, and Zn in the soil is spatially correlated with concentration in vegetables. In addition, though Cd was not detected in soil, the Cd concentration in vegetables shows that it is spatially correlated to nearby environmental media having a Moran's I and p-value of 0.668301 and 0.004269, respectively. This is attributed to the groundwater [21,65] and runoff quality [65].  As shown in Table 9, the correlation coefficients of Cr-Cd, Mn-Fe, Ni-Fe, Ni-Mn, Pb-Cd, and Pb-Cr were 0.986, 0.629, 0.605, 0.852, 0.977, and 0.977 (p < 0.01), respectively. These positive significant correlations between these various metals in soil and vegetables demonstrated that they commonly interact, cooperate with each other to promote potential uptake of vegetables from soil, and are a common source of pollution [41,74,75]. Furthermore, spatial correlation technique was employed to analyze further the relationship of metal concentrations in vegetables that were collected in proximity to where the soil samples originated. Table 10 elaborates the generated Moran's Index that measures the spatial autocorrelation of soil and vegetables. This table illustrates that the concentration of Cr, Ni, Pb, and Zn in the soil is spatially correlated with concentration in vegetables. In addition, though Cd was not detected in soil, the Cd concentration in vegetables shows that it is spatially correlated to nearby environmental media having a Moran's I and p-value of 0.668301 and 0.004269, respectively. This is attributed to the groundwater [21,65] and runoff quality [65].

Conclusions
In this study, the trends of metals in soil and vegetables across the island province in the Philippines were investigated. Soil and vegetable samples were collected from the six municipalities of Marinduque island province and vegetables chosen were among the top produced and consumed in the province, namely string beans, sweet potato tops, bitter melon, and eggplant. All samples were analyzed for the presence of Cd, Cr, Cu, Fe, Mn, Ni, Pb, and Zn and assessed for potential ecological and human health risks. It was recorded that all metal concentrations in the soil, except for Cd, from the six municipalities exceeded the SQS. The concentrations of Fe and Mn in soil were the highest among other target metals. The P n values in all soil samples from each municipality were under Class V which means severe pollution level. Likewise, the pERI of soil in Boac, Buenavista, and Mogpog was considered to have a very high pollution risk while for Gasan, Sta. Cruz, and Torrijos, a high pollution risk was found.
Most of the metal concentrations in the vegetables exceeded the MPL. The sweet potato tops, a leafy vegetable, had accumulated the most metals among the vegetables. Eggplant more likely contributed the greatest TCR to the population. Moreover, all HHI values were less than 1 which means that there is a potential low non-carcinogenic human health risk to the population from the consumption of vegetables. The concentrations of Cr and Ni in target vegetables have the potential risk of cancer occurrence to the population. The TCR of Cr and Ni exceeded the maximum threshold value of 1.00 × 10 −4 . It is further stressed that values of TCR were based on estimated EDI and RfD with uncertainty though addressed by engineering models. Hence, more focused and deeper study is useful.
Based on this record, it is advised to exercise a 75% temporary consumption reduction of bitter melon, eggplant, string beans, sweet potato tops produced from the municipalities of Buenavista and Gasan. Therefore, the proposed temporary monthly average consumption of bitter melon, eggplant, string beans, and sweet potato tops is 6.9, 7.8, 7.4 and 14 g person −1 , respectively. However, more studies are necessary to identify more vegetables and from which area they should be produced in order to sustain the dietary and vegetable preference needs of the population.
The datasets were subjected to an autocorrelation analysis using Moran's Index of the GIS. Records showed that the datasets were more spatially correlated and clustered than would be expected. This means that metal concentrations in soil and nearby vegetables are correlated with each other. Further, the Pearson correlation and hierarchical cluster analyses, on the other hand, revealed that the positive significant correlation between the grouped metals in soil and vegetables demonstrated affinity, promoted potential uptake of vegetables from soil, and a common source of pollution. Hence, expanding the current research to include determining which part of a vegetable stores Ni, contaminant transport analysis, biomonitoring of Cr and Ni, different routes of metal exposure such as eating other foods (root crops, fruits, and cereals), drinking water, and air contact, uncertainty and variability analyses, is crucial to further the understanding of the risks associated with the elevated metal concentrations in the soil and its produce on human health, especially targeting carcinogenic risk. Further, the result of this study is useful for the development of more focused studies for more specific contaminants and edible agricultural yields that are consumed by the public. In addition, epidemiological studies specific to cancer risks are useful to help health authorities and relevant government agencies to create necessary programs to protect public health.