Topolnitsa River and its tributaries (Medetska, Zlatishka, Pirdopska and Bunovska Rivers) were influenced by the ore industry, agriculture activities, and domestic wastewater discharge, which led to the deterioration of its ecological status according to the River Basin Management Plan (2016–2021) (East Aegean River Basin Directorate. MoEW). Three large metal mining and metal processing enterprises are located in this region- “Elatsite Med”, “Dundee Precious Metals” and “Aurubis”. Additionally, in the area, the open-pit “Medet” is located, and although the mine was closed in 1992, it is considered the main source of pollution in this area.

The study area is located in the Topolnitsa-Pirdop valley (Fig. 1). Topsoil samples (0–20 cm) with different Cu concentrations (ranging from 28.05 to 198.9 mg kg^-1) were collected in May 2021. Five sampling sites have been selected from a monitoring map (Kancheva et al. 2018), considering the sources of Cu pollution as follows: S1 – located in the valley of Medetska River at 710 m from the “Medet” mine tailing (42°39'12.9"N, 24°09'04.2"E), S2 – located in the valley of Topolnitsa River (after Zlatishka River) at 250 m from the “Medet” mine tailing (42°39'31.8"N, 24°08'26.6"E), S3 – located in the valley of Pirdopska River at 750 m from the “Medet” mine tailing (42°39'48.2"N, 24°08'11.3E), S4–located in the valley of Bunovska River at 2200 m from the mine tailings of “Elatsite-Med” A.D. (42°36'47.5"N, 24°00'44.8"E), and S5–located in the valley of Topolnitsa River-next to Muhovo village at 25–30 km straight from S4 and S1 (42°24'45.5"N, 23°59'54.4"E). Samples were taken in three replicates per site. All sampled soils were classified as Fluvisols (FAO 2015).

Soil pH was potentiometrically measured in 0.1 M CaCl₂ soil suspension according to ISO 10390:2005. The soil organic carbon (SOC) was determined by sulfochromic oxidation according to BSS ISO 14235:2002. Nitrate (NO₃-N) and ammonium (NH₄-N) nitrogen were determined according to Keeney and Nelson (1982). Inorganic phosphate (P₂O₅) was determined by the method of Olsen (1982). Water content (WtC) was calculated after oven drying (105 °C). The concentration of Cu was measured by inductively coupled plasma-atomic emission spectrometry (ICP-AES) according to ISO 22036:2008.

The genomic DNA was extracted from 0.5 g soil, using the E.Z.N.A soil DNA kit (Omega Bio-tek, USA) according to the manufacturer`s recommended protocol. The quality and quantity of the extracted DNA were subsequently assessed by Qubit 4 fluorometer (Invitrogen) and 1% agarose gel electrophoresis.

Q-PCR analyses were performed using iTaqTM Universal SYBRGreen Supermix (BioRad) on Rotor Gene 6000 (Corbett Life Science) according to the manufacturer’s protocols. Bacterial abundance was estimated using 0.4 μM of primers Eub338f (5’-ACTCCTACGGGAGGCAGCAG-3’) and Eub518r (5’-ATTACCGCGGCTGCTGG-3’) (Aleksova et al. 2020). The three-step thermal program was used at 95 °C for 5 min,45 cycles of denaturing at 95 °C for 10 s, annealing at 61 °C for 30 s, extension at 72 °C for 30 s, and final extension at 72 °C for 10 min. The plasmid DNA from the bacterial standard of Uncultured α – proteobacterium clone BuhD-104 (FM877535.1) was used to generate a linear calibration curve of the threshold cycle versus the gene copy numbers of four-point serial decimal dilution – 10³, 10⁵, 10⁶ and 10⁸. All measurements were run in triplicates. Data were expressed as gene copy numbers per gram of dry soil, and the specificity of amplicons was confirmed by melting curve and agarose gel electrophoresis. Amplification efficiency was estimated to be 109% (R²=0.99). The results were processed via ROTOR-GENE Q SERIES, Software version 2.3.1.

The fungal abundance was estimated using 0.4 μM primers ITS1 (5’-TCCGTAGGTGAACCTGCGG-3’) and ITS4 (5’-TCCTCCGCCTTATTGATATGC-3’) (White et al. 1990). The following thermal program was used: 95 °C for 5 min, 45 cycles of 95 °C for 30 s, annealing 55 °C for 30 s, 72 °C for 60 s, and final extension at 72 °C for 10 min.

A four-point serial decimal dilution of a plasmid DNA of Uncultured Basidiomycota clone LS_Az0_D1_31 (MT785782.1) was used as a standard to generate a linear calibration curve of the threshold cycle versus a number of gene copies ranging from 10² to 10⁵. All measurements were run in triplicates. Data were expressed as gene copy numbers per gram of dry soil. Amplification efficiency was estimated to be 92% (R²=0.99). The results were analysed via ROTOR-GENE Q SERIES, Software version 2.3.1.

The relative ratio of fungi/bacteria was calculated as the ratio of the fungal gene copy numbers to bacterial gene copy numbers (Fierer et al. 2005).

Pearson correlation analysis was used to find the relationships between soil metrics. Hierarchical clustering (Algorithm: UPGMA; Similarity index: Bray-Curtis) was conducted to evaluate the distance between soil physical environments. Non-metric multidimensional scale (nMDS) ordination was applied to assess the similarity between abundance-weighted microbial communities. Statistical analyses were performed using the software PAST (p<0.05).

The values of studied soil properties are shown in Table 1. The soil pH was moderately acidic – pH 5.57 (S1 and S2) to 5.76 (S4). SOC ranged from 1.22 g kg^-1 (S3) to 2.67 g kg^-1 (S2). Concentrations of inorganic nitrate (except S3) and ammonium (except S2 and S4) ions and inorganic phosphates were relatively equally distributed among the soils.

Cu concentrations varied among the soils as follows: the concentrations in S1 and S3 exceeded the maximum permissible concentration (MPC) of 80 mg kg^-1 under Bulgarian Regulation 3/2008 (2008) (http://eea.governement.bg/bg/legislation/soil). The most contaminated soils are located near the tailing of mine “Medet” in the valleys of the Medetska and Pirdopska Rivers. In the other soils, S2, S5 and S4, the concentrations of Cu were under or around MPC. In this regard, S5 can be assumed as a reference site.

Pearson correlation analysis indicated a linear relationship (-0.61; p=0.015) between the level of soil Cu contamination and the distance to the source of pollution.

A cluster analysis of soil abiotic factors was conducted to visualize the similarity among the tested soils (Fig. 2).

Two main clusters were demonstrated: cluster I containing soil S5 with the lowest Cu concentration and cluster II, consisting of other soils sub-clustered according to the Cu threshold concentration of 100 mg kg^-1. This pattern of soil clustering identified the major role of Cu in structuring soil environments.

Bacterial abundance varied in a relatively narrow range. The highest number was observed in S4 (3.24 × 10¹¹), and the lowest one in S5 (1.68 × 10¹¹) (Fig. 3). Pearson correlation analysis did not indicate a significant correlation (0.31; p=0.26) between bacterial abundance and soil Cu concentrations.

The highest fungal abundance was detected in uncontaminated soil S5 (6.71 × 10⁸) and the lowest one in S3 (1.95 × 10⁸) (Fig. 4). The fungal gene copy number variance among the sampling soils was higher (Coef. variance – 40.48) than that of bacterial gene copy number (Coef. variance – 25.07), indicating possible resistance of bacteria to Cu toxicity.

Pearson correlation analysis showed a significant negative correlation (-0.76; p=0.001) between fungal abundance and soil Cu contamination.

The highest relative fungal to bacterial abundance was observed in the uncontaminated soil S5 (3.98 × 10^-3), and the lowest ratio was recorded in the highly contaminated soil S3 (6.80 × 10^-4) (Fig. 5). Fungal to bacterial ratio related negatively with soil Cu (-0.66; p=0.007), probably as a result of the higher Cu impact on the fungal abundance.

Statistical non-metric multidimensional scaling (nMDS) was performed on patterns of fungal and bacterial abundances to assess soil similarities with respect to their microbiology (Fig. 6).

Plots separated the soils into four distinct abundance-weighted microbial communities The soils S2 and S4 grouped closely, indicating similar microbial abundances. The other soils were placed individually both away from the other and from the group S2 and S4. This distribution was probably due to differences in microbial abundance-weighted communities. Our results showed that the important factors influencing soil microbial abundance were soil Cu concentration and water content in S1, and soil organic carbon and phosphates concentrations in S2 and S4 (Fig. 6).