1. INTRODUCTION

Soil contamination by heavy metals is a persistent environmental issue, particularly in areas affected by mining activities. Abandoned mines, in particular, are recognized as major sources of metal pollution, releasing elements such as cadmium (Cd), lead (Pb), and zinc (Zn) into surrounding soils through weathering of waste rock and tailings (Kim et al. 2024a;Peng et al. 2025). These metals can accumulate at high concentrations, surpassing ecological safety thresholds, and exert long-lasting impacts on soil ecosystems due to their non-degradable nature (Wu et al. 2024).

In particular, heavy metal contamination has been known to influence microbial communities (Pang et al. 2023;Wu et al. 2024). Soil microorganisms play essential roles in organic matter decomposition, nutrient cycling, and ecosystem resilience, making them highly sensitive indicators of soil health (Banerjee and van der Heijden 2023). Numerous studies have shown that heavy metals can suppress microbial biomass, reduce enzymatic activity, and disrupt metabolic pathways, thereby compromising soil function (Peng et al. 2025).

Community-level physiological profiling (CLPP) using Biolog EcoPlate^TM has been widely employed as a functional approach to assess microbial metabolic potential in contaminated soils (Eo et al. 2024). By measuring average well color development (AWCD) across a range of carbon substrates, CLPP provides insights into both overall microbial activity and group-specific substrate utilization patterns. AWCD has been shown to serve as a reliable proxy for both microbial diversity and metabolic activity, thereby providing an integrative measure of community functional potential (Costa et al. 2007). Previous research has shown that metal stress can selectively inhibit the utilization of specific substrate groups, such as amino acids and their derivates, thereby shifting the functional structure of microbial communities (Xie et al. 2016). Such shifts often involve a decline in functional diversity, with sensitive metabolic pathways being suppressed and more tolerant microbial taxa dominating carbon utilization. However, the extent to which different heavy metals influence specific metabolic pathways remains poorly understood, particularly in heavily contaminated mine-affected soils.

The Deoksan Mine in Korea, an abandoned Pb-Zn mining site, presents a valuable case for investigating the effects of heavy metal contamination on soil microbial function. The site exhibits spatial heterogeneity in contamination levels, with extremely high Zn and Pb concentrations near mine adits and waste dumps, declining downslope. Pb-Zn mining areas are often characterized by severe contamination with multiple heavy metals, including As, Cd, Cu, Pb, and Zn, extending from adits and mine tailings to adjacent agricultural lands (Rodríguez et al. 2009). Numerous studies have assessed such contamination through risk assessment approaches (Cheng et al. 2018;Wei et al. 2024). However, investigations that employ microbial utilization of specific substrate groups as functional indicators remain limited, highlighting the need for further research in this area.

In this study, we aimed to (1) evaluate spatial differences in soil physicochemical properties and heavy metal concentrations at the Deoksan Mine, (2) examine temporal changes in microbial substrate utilization through AWCD patterns, (3) determine how substrate utilization profiles differ among carbon substrate groups, and (4) identify key heavy metals that regulate microbial functional potential. By combining CLPP data with soil chemistry, this study aims to provide a clearer understanding of how heavy metal stress shapes microbial carbon metabolism and to evaluate the potential of specific substrate groups as functional indicators for ecological risk assessment.

2. MATERIALS AND METHODS

2.1. Study site and selection of sampling locations

This study was conducted at the Deoksan Mine in Susan-myeon, Jecheon-si, Chungcheongbuk-do, Korea. The mine, formerly used for Pb and Zn extraction, is now abandoned (https://miregis.komir.or.kr/). A 30 m×30 m grid was established across the site to guide sampling. Soil samples from each grid cell were analyzed for physicochemical properties (pH, electrical conductivity [EC], organic matter content, cation exchange capacity [CEC], and soil texture) and heavy metal concentrations (Ni, Cu, Zn, As, Cd, Pb).

The site was primarily contaminated with Zn, Cd, and Pb, with concentrations exceeding the worrisome level of soil contamination for forest land defined under the Korean Soil Environmental Conservation Act (600, 10, and 400 mg kg^-1, respectively) in most locations. Metal concentrations were highest near mine adits and waste rock dumps, decreasing downslope.

From this grid (30 m×30 m), three contaminated plots (5 m×5 m) were chosen based on the preliminary survey to ensure relatively homogeneous heavy metal concentrations while representing distinct environmental conditions: the mine adit (S1), the lower section of the waste rock dump (S2), and a shrubland area (S3). A control site (Control) was established in a nearby uncontaminated area with similar physicochemical characteristics and well-developed vegetation, dominated by Quercus aliena and Larix kaempferi.

2.2. Soil sampling and analysis

At each plot (5 m×5 m), three random surface soil samples (0-10 cm depth) were collected using a 10 cm diameter corer. Samples were used for both physicochemical analysis and microbial substrate utilization profiling.

In the laboratory, soil samples were air-dried, sieved (2 mm), and analyzed. Soil pH and EC were measured in distilled water (1 : 5 w/v). Organic matter content was determined by the Walkley-Black wet oxidation method, and CEC was measured using 1 M ammonium acetate (pH 7.0) leaching following Brown’s method (Kulte 1986;Sparks et al. 2020).

Total Ni, Cu, Zn, As, Cd, and Pb were determined after aqua regia digestion (HCl : HNO₃=3 : 1, v/v) and quantified using inductively coupled plasma-optical emission spectrometry (ICP-OES). All measurements were made in triplicate. Soil physicochemical and heavy metal data are presented in Table 1.

2.3. Microbial community substrate utilization profiles

Community-level physiological profiles (CLPP) were assessed using Biolog EcoPlate^TM (Biolog Inc., Hayward, CA, USA) following Grządziel and Gałązka (2018) with minor modifications. Five grams of soil were suspended in 45 mL of 0.85% sterile NaCl, shaken for 30 min, and allowed to settle for 30 min. A 500 μL aliquot of the supernatant was diluted in 49.5 mL sterile saline, vortexed for 30 s, and left to stand for 10 min. Then, 150 μL was inoculated into each EcoPlate^TM well and incubated at 30°C in the dark. Absorbance at 590 nm was recorded every 24 h for 7 days using a microplate spectrophotometer (PowerWave XS; Biotek, Winooski, VT, USA).

Average well color development (AWCD) was calculated as the mean absorbance across all 31 carbon substrates. Substrates were categorized into six groups: (1) Carbohydrates (D-cellobiose, α-D-lactose, β-methyl- D-glucoside, D-xylose, i-erythritol, D-mannitol, Nacetyl- D-glucosamine), (2) Carboxylic acids (pyruvic acid methyl ester, D-glucosaminic acid, D-galactonic acid γ-lactone, D-galacturonic acid, 2-hydroxybenzoic acid, 4-hydroxybenzoic acid, itaconic acid, α-keto butyric acid, D-malic acid), (3) Amines (phenylethylamine, putrescine), (4) Amino acids (γ-aminobutyric acid, L-arginine, L-asparagine, L-phenylalanine, L-serine, L-threonine, glycyl-L-glutamic acid), (5) Phosphorylated chemicals (glucose-1-phosphate, D,L-α- glycerol phosphate), (6) Polymers (Tween 40, Tween 80, α-cyclodextrin, glycogen).

2.4. Statistical analysis

Differences among sites in soil physicochemical properties and heavy metal concentrations were tested using one-way analysis of variance (ANOVA), followed by Tukey’s HSD for post hoc comparisons. The same approach was applied to compare AWCD values for each substrate group at each time point.

Multiple regression models were used to evaluate the influence of heavy metals on total and group-specific AWCD, using data collected at 144 h, at which point site differences were most pronounced. Relative variable importance was calculated using the relimpo package in R. Simple linear regression assessed the relationships between Cd, Pb, Zn concentrations and AWCD values. All analyses were performed in RStudio (RStudio Inc., Boston, MA, USA), with significance at p<0.05.

3. RESULTS

3.1. Soil physicochemical properties and heavy metal concentrations

Soil properties and metal concentrations differed significantly among sites (Table 1). pH ranged from 7.32± 0.04 at S3 to 7.81±0.02 at S1. EC was highest at S1 (0.16±0.00 dS m^-1) and lowest at S3 (0.06±0.00 dS m^-1). TOC was greater at the Control and S1, with S3 lowest. CEC was highest at the Control (44.36±0.40 cmol kg^-1) and lowest at S3 (33.36±0.61 cmol kg^-1). Soil texture was loam at Control and S1, and clay loam at S2 and S3.

Heavy metal concentrations showed marked sitespecific differences. S1 had extremely high Zn (19,064.53 ±323.08 mg kg^-1) and Pb (8,250.49±220.72 mg kg^-1), exceeding levels at all other sites (p<0.05). Ni was highest at S3 (38.98±0.10 mg kg^-1), while the Control consistently had the lowest metal levels.

3.2. Temporal changes in substrate utilization

AWCD increased over time at all sites (Fig. 1), indicating progressive microbial activity. Site effects were significant at most time points (Table 2). From 24 h onward, the Control exhibited the highest AWCD, whereas S2 maintained the lowest. After 120 h, AWCD showed a tendency to stabilize, while S1 continued to increase.

Heatmap analysis revealed that, across all substrate groups, the Control consistently showed the highest levels of microbial activity and substrate utilization (Fig. 2). This pattern was not uniform throughout the incubation but became increasingly evident after 48 h, with the contrast among sites growing more pronounced as the experiment progressed toward 144 h. Although differences among sites were less distinct at 96 h and not statistically significant at that time point (Table 2), the overall trajectory indicated a clear divergence between the Control and the contaminated plots as incubation advanced.

Among the substrate groups, the contrasts were especially apparent for carboxylic acids, amines, and phosphorylated chemicals. Carboxylic acids and amines displayed consistently higher utilization in the Control compared with S1-S3. The most striking pattern was observed for phosphorylated chemicals. Utilization was consistently strong at the Control but remained nearly absent at S1 and S2 throughout the entire incubation period.

3.3. Site- and time-dependent variation in substrate group utilization

One-way ANOVA showed strong site effects for total AWCD at 24 h (F=103.34, p<0.001), with significant differences also at 48 h and 72 h. No site effect was observed at 96 h (p=0.12), but differences reappeared at 120 h and 144 h (p<0.001).

Carbohydrates differed significantly among sites at 24-72 h, with reduced variation thereafter. Carboxylic acids showed significant differences during 24-72 h and again at 120-144 h, with the largest F-value at 24 h. Amines differed only at 24 h and 144 h, whereas amino acids differed at 24-72 h but not later. Phosphorylated chemicals varied among sites from 48 h to 120 h, peaking at 96 h (F=24.47, p<0.001). Polymers differed among sites only between 48 h and 120 h.

3.4. Influence of heavy metal concentrations on substrate utilization

Relative importance analysis at 144 h (Fig. 3) identified Cd and Pb as primary predictors of total AWCD (R²=0.94), while Cd, Pb, and Zn dominated for phosphorylated chemicals (R²=0.97). Carboxylic acids and amines were also strongly influenced by Cd and Pb (R²= 0.87 and R²=0.88, respectively), whereas amino acids (R²=0.62) and polymers (R²=0.59) had lower explanatory power.

Simple linear regression showed strong negative relationships between Cd, Pb, and Zn concentrations and phosphorylated chemical utilization (Fig. 4). For total AWCD, only Cd showed a borderline significant negative relationship (p=0.05), while Pb and Zn were not significant. These results indicate that Cd, Pb, and Zn strongly inhibit utilization of specific substrate groups, particularly phosphorylated chemicals.

4. DISCUSSION

This study demonstrates that heavy metal contamination exerts significant effects on soil microbial functional potential, as reflected in substrate utilization profiles. Our findings indicate that substrate utilization patterns shifted over time, and specific substrate groups, particularly phosphorylated chemicals, were strongly influenced by Cd, Pb, and Zn contamination. These results suggest that phosphorylated chemical utilization can serve as a sensitive indicator for assessing the ecological impacts of heavy metal contamination in soils.

4.1. Influence of soil properties and heavy metals on microbial functional diversity

Marked site-specific differences in pH, EC, organic matter, and CEC were observed, with the most contaminated site (S1) also exhibiting the highest Zn and Pb concentrations. Such physicochemical variation can directly and indirectly influence microbial community function by altering nutrient availability, osmotic stress, and toxicity thresholds (Wang et al. 2021). Although our study did not detect clear relationships between soil physicochemical properties and microbial substrate utilization, further research is needed to elucidate this relationship, considering its critical role in determining microbial activity and ecosystem functioning. Previous studies have reported that high Zn and Pb concentrations can inhibit microbial enzymatic activity and suppress carbon substrate utilization (Klimek et al. 2023). Our results support these findings, particularly the negative associations between Cd, Pb, Zn and specific substrate groups. The observed inhibitory effects on phosphorylated chemical utilization may be related to the disruption of phosphorus metabolism enzymes, which are known to be sensitive to heavy metal stress (Costa et al. 2007).

4.2. Temporal dynamics of substrate utilization

Across all sites, AWCD increased over time, indicating that viable microbial populations were actively metabolizing available carbon substrates. However, the rate and extent of AWCD development varied by site, with S1 showing prolonged increases and S3 maintaining the lowest values throughout incubation. Such differences may reflect both initial microbial community composition and adaptive capacity under metal stress (Niklińska et al. 2006). The plateauing of AWCD at the Control and S2 after 96 h suggests resource depletion or a shift toward slower-growing taxa with lower substrate turnover. In contrast, sustained AWCD increases at S1 could be due to metal-tolerant taxa exploiting residual carbon sources, albeit possibly at the expense of reduced functional diversity. Nevertheless, our data does not allow a definitive confirmation of this mechanism, and further analyses aimed at elucidating the changes in community composition would be required.

4.3. Shifts in substrate group utilization under metal stress

Heatmap analysis revealed distinct substrate use patterns among sites and over time. Carbohydrates and carboxylic acids were consistently the most utilized groups, likely reflecting their central role in microbial energy metabolism (Banerjee and van der Heijden 2023). The significant site effects for carboxylic acids even at late incubation stages suggest that these substrates may serve as a key functional trait differentiating metal-tolerant communities from less tolerant ones (Niklińska et al. 2006).

Phosphorylated chemicals exhibited the strongest negative associations with heavy metal concentrations. This may be because organophosphate utilization is particularly susceptible to inhibition by Pb and Cd, which can competitively bind to enzyme active sites (Gao et al. 2025). The strong R² values (>0.95) in our regression models highlight the sensitivity of this functional group as a potential bioindicator for metal toxicity in soils. Amines and amino acids showed more transient site effects, suggesting that nitrogenous substrates may be less reliable for long-term contamination assessment. However, despite these substrate-specific responses, overall AWCD did not show a consistent inverse relationship with heavy metal concentrations, suggesting that community-level metabolic activity may be maintained through functional redundancy. Other factors, such as site-specific differences in microbial community composition (e.g., the presence or absence of taxa particularly sensitive to certain metals), variation in soil physicochemical properties, and the actual bioavailable fractions of heavy metals, may also have contributed to the observed patterns.

4.4. Methodological considerations and temporal variability

The interpretation of Biolog EcoPlate-based CLPP must be made with caution due to well-recognized methodological limitations. CLPP reflects the metabolic potential of the culturable fraction of soil microbial communities and does not fully represent in situ diversity and activity (Klimek et al. 2023). Consequently, AWCD alone should not be considered a comprehensive measure of microbial function. In our study, AWCD trajectories differed across sites, but statistically significant relationships with heavy metal concentrations were limited. These limitations suggest that CLPP is best viewed as a complementary rather than a standalone tool for assessing microbial responses.

Another important aspect concerns the temporal variability of AWCD patterns. Our data showed that site differences were not statistically significant at 96 h but re-emerged and became pronounced at 120 h and 144 h. This temporal inconsistency was also evident in the Tukey test for phosphorylated chemicals (Fig. 1b), where the Control consistently showed higher values but statistical significance shifted across incubation times. Such a pattern indicates that even when contamination levels are clearly differentiated among sites, the expression of functional differences can fluctuate with time, underscoring the role of temporal variability in microbial responses under heavy metal stress. For this reason, we emphasized the 144 h data, where site contrasts were clearest and associations with heavy metal concentrations were most evident. We acknowledge, however, that the lack of consistent differences across all time points complicates interpretation. At the same time, such temporal fluctuations may themselves reflect instability in microbial activity under heavy metal stress. Future work should therefore integrate multiple incubation times and combine CLPP with molecular approaches to better capture both dynamic responses and underlying mechanisms.

4.5. Implications for ecological risk assessment and soil health monitoring

Our results highlight the utility of Biolog EcoPlatebased CLPP for detecting functional shifts in microbial communities under heavy metal stress. The consistent inhibitory effects of Cd, Pb, and Zn on phosphorylated chemical utilization suggest that this functional group could serve as an early-warning indicator in contaminated soils (Costa et al. 2007). Furthermore, integrating CLPP data with soil physicochemical parameters could improve the sensitivity of ecological risk assessments, particularly within frameworks such as the ISO TRIAD approach (Kim et al. 2024b). However, it should be noted that CLPP primarily captures the metabolic potential of the culturable fraction of soil microbial communities, and may not fully represent in situ activity. Future studies should combine CLPP with high-throughput sequencing and enzyme assays to better understand functional resilience and redundancy in metal-contaminated soils.

5. CONCLUSION

This study revealed that soil microbial functional potential is strongly shaped by site-specific differences in physicochemical properties and heavy metal contamination at the Deoksan Mine. Although overall microbial activity, reflected by increasing AWCD, progressed over time at all sites, significant spatial differences were detected. Substrate utilization patterns varied with incubation time, with carbohydrates and carboxylic acids utilized early, while phosphorylated chemicals became more prominent in the late phase. Among the functional groups, phosphorylated chemical utilization was particularly sensitive to Cd, Pb, and Zn concentrations, showing strong negative correlations. These findings highlight that heavy metal contamination not only reduces overall microbial activity but also alters specific functional pathways, suggesting that phosphorylated chemical utilization may serve as a sensitive bioindicator for assessing ecological risks in contaminated soils.