1 Introduction

Soil microorganisms are a key component of terrestrial ecosystem functioning, supporting energy flow and organic matter cycling, ecosystem biodiversity, and stability (van der Heijden et al. 2008; Eisenhauer et al. 2017; He et al. 2022). Therefore, soil microorganisms are indicators of ecosystem health and proper functioning (Gałązka et al. 2022). Soil microbial characteristics can serve as biological indicators on recultivated post-industrial sites, including respiration rate (Chodak and Niklińska 2010), biomass (Chodak et al. 2015), enzymatic activity (Chodak et al. 2021), and, finally, functional and/or genetic diversity and community structure of the community (Markowicz et al. 2015; Moradi et al. 2020). The use of soil microorganisms as indicators of site restoration success may be hampered by the fact that soil parameters are influenced by several environmental factors. Both vegetation characteristics and soil physical and chemical properties are among the most important factors that most strongly shape soil microbial communities (Chodak et al. 2016; Klimek et al. 2020a; He et al. 2022). To study soil bacteria, we used a standard medium, the Biolog® ECO plates, which allow quick assessment of the fulfilment of ecological functions that can be performed by different microbial species (Preston-Mafham et al. 2002).

Reclamation of post-industrial wastelands faces several challenges (Frouz et al. 2008; Chodak et al. 2021; Likus-Cieślik et al. 2023). The most important role of well-established vegetation in reclamation efforts is to ground stabilization, prevent soil erosion, promote water infiltration and retention, and restore landscape connectivity (Vidal-Macua et al. 2020). Reclamation success depends on the properties of dumped or excavated material, including the presence and concentration of toxic substances, but also on local water availability and vegetation arrangement (Chodak and Niklińska 2010; Likus-Cieślik et al. 2023). The effectiveness of reclamation processes and the permanence of vegetation cover are critical in densely populated areas, mainly because of dust exposure and the associated risk to human health. More than half (57%) of the world’s population lives in cities (The World Bank, 2021). Accordingly, some mine tailings that were previously located outside the cities and on unused land may become integral parts of growing urban areas (Spórna and Krzysztofik 2020; Yan et al. 2022). Vegetation on reclaimed mine tailings located in cities suffers from intensified environmental challenges, such as increased air temperature, water shortage, air and soil pollution, compaction, and alkalisation (Godefroid and Koedam 2003; McKinney 2006).

The aim of the study was to analyse the relationship between vegetation and soil bacteria features on reclaimed heaps, as knowledge of the ecological linkages between above- and below-ground biota is crucial for improving understanding of the maintenance and stability of ecosystem processes (Zappelini et al. 2015). It is worth noting, that most efforts towards investigating the effect of vegetation on soil biota on reclaimed wastelands have focused on assessing the effect of the dominant tree species used for reclamation. For example, many studies have revealed the effectiveness of remediation using silver birch Betula pendula (Chodak et al. 2015) or black alder Alnus glutinosa L. (Sroka et al. 2018) and confirming their boosting effects on soil microorganisms. However, vegetation diversity in afforested sites is most strongly related to undergrowth plant species (Rawlik et al. 2020; Kondratenko et al. 2022), the same applies to temperate forest sites (Klimek et al. 2015, 2016). It has been shown that the relationships between the diversity of plants and soil bacteria were different in the early and late stages of temperate forests (Li et al. 2016). Afforested heaps can therefore serve as a model for studying these relationships at an early stage of succession, especially on heaps formed with unpolluted or moderately polluted excavated material.

We expect, that vegetation diversity composition will favour functional (metabolic) soil bacteria diversity on a recultivated ‘Solvay’ process tailing ponds located in expanding residential areas in the city of Krakow in southern Poland. Tailing ponds resulted from the production of soda ash (sodium carbonate, Na2CO3) (Steinhauser 2008) and contain highly alkaline (pH = 11–12) and saline wastes, prone to intensive dusting (Krzak 2005; Sroczyński 2009). Our results may also be useful for the upcoming reclamation of similar areas in Poland and across the world (Gołub and Piekutin 2020; Wang et al. 2020), as many branches of industry continuously produce large amounts of alkaline residues (Gomes et al. 2016).

2 Materials and methods

2.1 Study site

The study was conducted in the area of recultivated soda waste piles and settlements of the former “Solvay” plant, located in southern Poland, in the city of Krakow, by the Wilga River (Fig. 1). The factory operated between 1906 and 1990s and in the past, its location was outside of the city boundaries. Soda ash (sodium carbonate, Na2CO3), the factory's main product is the raw product used in many industrial applications, including the production of glass, detergents, and paper (Steinhauser 2008). Nearly 5 million tonnes of waste (NaCl, CaCl2, Ca(OH)2, CaCO3, CaSO4, P2O5, and SiO2) was generated from factory, which were deposited in oval-shaped sedimentation tanks near the plant, across the area of about 100 ha. Deposited ashes formed white dunes, so-called “White Seas” (in Polish “Białe Morza”) and had a strong negative impact on the environment. Highly alkaline (pH = 11–12), salty, and prone to intensive dusting wastes dunes were finally recultivated in the 1990s by covering them with a cap layer composed of clay, slag, and soil, enrichment with sewage sludge of thickness 20–60 cm (locally 1 m) (Sroczyński 2009). This created a physical barrier that prevented direct exposure to waste materials and allowed for vegetation introduction. Nowadays the area is mostly covered by vegetation; the former sedimentation tanks (currently heaps) are mostly overgrown by a variety of vegetation types: cryptogamic, ruderal, meadow-like assemblages with shrubs and, partially, with trees and shrubberies located in a local land depression, on heaps sides and closer to the Wilga river.

Fig. 1
figure 1

Maps presenting study sites distribution: a Cracow city location on the map of Europe, b Cracow city agglomeration scheme with main roads and rectangular marked region denoting Białe Morza area, c city region of Białe Morza with four former settlements—heaps (A, B, C, D)

Full size image

2.2 Sampling design

Four former sedimentation tanks (A, B, C, and D) (Fig. 1) were studied and nine plots (10 × 10 m size) were investigated within each heap: five on the flat top and four at the sides (North, South, East, and West). The latter were located in the middle of the heaps’ height, approximately 10 to 15 m above the bottom of the Wilga River valley. The plots within the top area were selected in the middle of the heap plane and directly above the four side plots. Differences in waste composition and environmental effects (such as rain runoff) were expected to cause variations in vegetation composition between plots located on the tops and sides of the settlements. Northern parts of heap A were excluded from the sampling due to the ongoing construction of a new public road and only 7 samples were collected there. The final number of plots included in the study was 34 (19 plots on the tops and 15 plots on the slopes).

2.3 Vegetation survey and soil sampling

The vegetation survey and soil sampling was carried out once during one working day. On each site, vegetation, that is, vascular plants including trees, shrubs and the floor species, was characterized on a 100-m2 study plot using the Braun-Blanquet method (1964). The collected data included the number of vascular plant species per plot (N plant), the number of tree species (N tree), the number of shrub species (N shrub), the number of floor species (N floor), and plant coverage in three layers of vegetation (note that summed plot coverage by trees, shrubs and floor layer may exceed 100%). The data on plant cover in the relevés were transformed from the Braun-Blanquet scale into a 1 – 9 ordinal scale (van der Maarel 1979), and the H’plant was calculated according to the equation:

$${H}^{\prime}_{plant} = -\sum_{i=1}^{N\;plant}{p}_{i}\left({log}_{10}{p}_{i}\right)$$
(1)

where pi denotes the frequency for the i-th species of plant, and N plant is the number of vascular plant species at a particular site. Within the plant species found, invasive species were separated according to (Tokarska-Guzik et al. 2012).

Soil samples were collected in the central part of the plot, where a vegetation survey was performed. The upper 10 cm depth of the soil was collected with a spade. After the transport of soil to the laboratory, the soils were sieved with a 1 cm wide sieve to remove plant residuals, stones, and soil animals, and stored at 4 °C field moist until further analyses.

2.4 Soil physical and chemical analysis

The dry weight (DW) of the soil samples was determined by measuring the mass loss (water) after the soil samples had been dried at 105 °C for 24 h. Subsequently, the samples were subjected to burning in a muffle oven at 550 °C for 24 h for determining organic matter (OM) content. The water holding capacity (WHC), which was the amount of water that a given soil can hold without leaking, was measured using a standard gravimetric method after the soil was soaked for 24 h in net-ended plastic pipes immersed in water. The soil pH was measured in air-dried subsamples (2 g) shaken in deionised water (1:10 w:v) for 1 h at 200 rpm. Organic carbon (C) and total nitrogen (N) were analyzed by dry combustion of approximately 5 mg milled soil samples with an elemental analyzer (Vario El III, Elementar Analysen Systeme GmbH, Germany). The flow-injection analyzer (FIA compact, MLE) was used to analyze the total P concentration, after wet mineralization of 0.5 g DW of subsamples in suprapure 65% HNO3 (Merck). To assess the accuracy of the mineralization process, four blank samples and three replicates of standard certified material (CRM025-050, Sandy Loam 8, RT Corp.) were analyzed with the soil samples. Each analysis was performed in three subsamples from each soil sample, and the data was averaged.

2.5 Biolog® ECO plates analysis of soil bacteria

Bacterial activity and functional diversity were analysed using Biolog® ECO plates. The Biolog® ECO plates are 96 well microplates, that contain 3 sets of 31 common carbon sources and employ a tetrazolium redox dye as an indicator of microbial metabolism (http://www.biolog.com). The substrates were six compound guilds: carbohydrates, carboxylic acids, amines and amides, amino acids, polymers, and miscellaneous (Campbell et al. 1997). All soil samples were replicated three times.

The soil samples were shaken in 0.9% NaCl for 30 min at 200 rpm. The supernatants containing microbes (100 μl) were diluted in 9.9 ml of 0.9% NaCl. Solutions of 100 μl per well were inoculated into the Biolog® ECO plates. To prevent contamination, all tools used were sterile and the preparation of Biolog® ECO plates was done under a laminar flow chamber. The absorbance in particular wells was measured as light absorbance at 590 nm using a Tecan with i-control software (Tecan Group Ltd., Männedorf, Switzerland). The first measurement was carried out just after inoculation and then was measured daily for 7 days. The average well-colour development (AWCD) values of the substrates were assigned to guilds and used to indicate the results of bacterial functional diversity for each plate and sample (Preston-Mafham et al. 2002). The overall rate of substrate utilization by microorganisms was expressed by the AUC (Area Under the Curve), which was calculated as follows:

$$AUC = \sum\limits_{i = 1}^{N} {\sum\limits_{t = 1}^{n - 1} {} \frac{{(A_{n} + A_{n + 1} )}}{2}\times (t_{n + 1} - t{}_{n}} )$$
(2)

where An and An+1 are the absorbance of each well (substrate, n) at two consecutive measurements at two different measurement times for tn and tn+1. The calculated AUC values were summed for each soil sample and the proportions of individual substrates from this total area were expressed as area %. The area % values for all substrates were then used for characterization of community-level physiological profiles (CLPP). statistical analyses. Functional diversity of soil bacteria was also expressed as the number of substrates decayed (R, richness) and by the Shannon diversity index (H’bact), which was calculated as:

$${H^{\prime}}_{bact}=\;-\sum_{i=1}^{R}{p}_{i}\left({log}_{10}{p}_{i}\right)$$
(3)

where pi is the ratio of the activity on each substrate to the sum of activities on all substrates.

2.6 Statistical analysis

Before statistical analyses, right- or left-skewed data were transformed to fulfil the normality criterion. The Shapiro–Wilk test was used to confirm the normality of the data distribution within groups.

Differences in vegetation characteristics (coverage, N plant, N tree, N shrubs, N floor, H’plant, N invasive, share of invasive species) between plots from two locations (heap tops and heap slopes) were compared with the one-way ANOVA test. Comparisons of the means for the groups were made using the Tukey test. One-way analysis of similarities (ANOSIM) was used to test for significant differences in the vegetation composition between plots on heap tops and slopes (permutation N = 9999, with Bray–Curtis similarity index). The similarity percentage (SIMPER) procedure was applied to identify which plant species made the largest contribution to the average dissimilarity between plots on heap tops and slopes.

Differences in soil physical–chemical properties (OM, WHC, pH, C, N, and P content, and C:N and C:P ratios) between soils from two locations (heap tops and heap slopes) were compared with the one-way ANOVA with the Tukey test.

Differences in soil bacteria characteristics (AUC, R, H’bact, substrate guilds % use on plates) between soils from heap tops and heap slopes were compared with the one-way ANOVA test. Comparisons of the means for the groups were made using the Tukey test. One-way analysis of similarities (ANOSIM) was used to test for significant differences in the CLPP between plots on heap tops and slopes (permutation N = 9999, Bray–Curtis similarity index).

Multiple-variable analysis was performed to indicate correlations between soil physical and chemical properties, vegetation and bacterial indices.

Non-parametric multidimensional scaling analysis (NMDS) with the Bray–Curtis similarity measure was used to compare bacterial CLPP profiles for heaps tops and slopes, with selected soil physical and chemical properties, and vegetation characteristics.

The Mantel test was used to identify the relation between vegetation diversity and composition and CLPP pattern. This test confirmed the relationship between the two matrices after permuting the rows (stands) and columns (individual plant occurrence vs. single substrate use, %) of the matrix. Random permutations were run 5000 times, with the correlation being recalculated after each permutation and Bray–Curtis matrices similarity measure was used to assess the one-tailed p-value and R-value for the model.

ANOVA analyses and correlations were performed using Statgraphics Centurion 18 software (StatPoint Technologies Inc., Warrenton VA, U.S.A.) and multivariate analyses using PAST 2.17c software (Natural History Museum, University of Oslo, Norway).

3 Results

3.1 Vegetation characteristics

On all plots altogether, 110 different vascular plant species were found. The number of plant species ranged from 5 to 19 per plot, and the most common species, distantly often, were the tree Betula pendula (growing on 21 plots) and the perennial invasive plant, Solidago canadensis (on 20 plots). These species generated also the highest plot coverage comparing other plant species. 34 plant species were found on only one single plot, and the next 25 were found on two plots. Of 11 invasive plant species found, the most common was the abovementioned Solidago canadensis, but also Solidago gigantea, Parthenocissus inserta, and Robinia pseudoacacia.

Heaps tops and slopes differed in some vegetation characteristics (Table 1). There were significant differences between heap tops and slopes in plant species number in vegetation layers; sites located on heap tops were characterized by a lower number of tree species (p = 0.0037) and shrubs species (p = 0.0113), but the higher number of floor species (p = 0.0156) than heap slopes (Table 1). However, the vegetation on the heaps' tops and slopes was characterized by similar coverage, and composed of a similar entire number of species, H’plant as well as the number of invasive species and their share in total number of species (Table 1).

Table 1 Differences in vegetation characteristics: total coverage (%), total number of vascular plant species (N plant), number of tree species (N trees), shrub species (N shrubs) and floor species (N floor), number of invasive species (N invasive), share of invasive species in total number of species (share inv., %) between flat tops and slopes of heaps (mean ± SD; n = 19 (tops) or n = 15 (slopes)). ANOVA test results were given in a separate column entitled P value. Significant differences between tops and slopes are in bold (P < 0.05)
Full size table

ANOSIM analysis showed a significant difference in vegetation composition on tops versus slopes (R = 0.2277; p = 0.0003). Overall Bray–Curtis dissimilarity in SIMPER analysis was 83.53 and this analysis revealed, that the first 10 plant species were responsible for 36.5% cumulative contribution in dissimilarity between tops and slopes. The most influential species was Solidago canadensis, an invasive perennial herbaceous plant, which was more abundant on tops than on slopes (Table 2). Two subsequent most influential species were native tree species, that is Betula pendula and Populus tremula, which were more abundant on slopes that on tops (Table 2).

Table 2 Results of SIMPER analysis of dissimilarity in vegetation composition between flat tops and slopes of heaps (mean ± SD; n = 19 (tops) or n = 15 (slopes)). Only ten the most influential species were given. Invasive species were indicated by bolded font
Full size table

3.2 Soil physical and chemical properties

Soils were characterized by a low OM content (9.7% on average), WHC (44.3% on average), and biogenic element content. Top and slope soils did not differ in OM content, WHC, or soil C and N content (Table 3). There was a significant difference in P content between the heap top and slope (p < 0.05), indicating that soils on heap slopes have higher P content than on tops (Table 3). Soil pH ranged from 6.5 to 8.8, and more than 65% of soil samples have pH above 8.0. Higher pH occurred at sites with a thicker cap layer or excavated dump material, as soil pH was negatively correlated with soil WHC and N and P content (see Supplementary Material).

Table 3 Differences in soil physical and chemical properties: organic matter content (OM, %), maximum water holding capacity (WHC, %), soil pH measured in water, content of C, N, and P (% of soil dry mass, DW) and C:N and C:P ratios between flat tops and slopes of heaps (mean ± SD; n = 19 (tops) or n = 15 (slopes)). ANOVA test results were given in a separate column entitled P value. Significant differences between tops and slopes are in bold (P < 0.05)
Full size table

3.3 Soil bacteria activity and functional diversity

Most of the substrates on Biolog® ECO plates were utilised by bacteria (26 from 31 on average for all soil samples). Sites on heap tops and slopes did not differ in soil bacteria activity and functional diversity indices (Table 4). Carbohydrates, carboxylic acids, and amino acids were among the most used substrate guilds (above 20% each), but utilisation of particular substrate guilds did not differ between site locations. ANOSIM analysis indicated a lack of difference in bacterial CLPP between heap tops and slopes (R = -0.04139; p = 0.8532).

Table 4 Differences in soil bacteria indices based on Biolog® ECO plate test: bacteria activity (AUC) and functional diversity (R, H’bact) indices, as well as the use of substrates from six chemical guilds (% of use) between flat tops and slopes of heaps (mean ± SD; n = 19 (tops) or n = 15 (slopes)). ANOVA test results were given in a separate column entitled P value
Full size table

3.4 Relations between vegetation, soil physical and chemical properties, and soil bacteria indices

Correlation analysis indicated that many vegetation indices were highly correlated with each other (Table 5). For instance, the number of tree species was positively correlated with the number of shrub species (R = 0.75; p < 0.0001), but negatively with the number of floor species (R = -0.59; p = 0.0006). The most numerous floor species have the highest involvement in the plant diversity index H’plant (R = 0.983; p < 0.0001). The higher number of invasive species per plot, the higher was their share in total number of plant species (R = 0.86; p < 0.0001), nevertheless invasive species number do not participate in H’plant. The only correlation found between vegetation indices and soil physical and chemical properties was a negative correlation between the share of invasive species in the total number of species and soil pH; however, that correlation was not particularly strong (R = -0.36; p = 0.0376) (Table 5).

Table 5 Heatmap with Pearson correlations for ecological parameters about vegetation, and soil bacteria properties at the recultivated post-industrial Solvay heaps in Kraków (for units see text). Correlations are displayed in red (positive) and green (negative); colour saturation denotes the strength of the relationship (statistically significant at p<0.05 are indicated in bold)
Full size table

The general soil bacterial indices (AUC, R, H'bact) were highly correlated, which is a commonly observed phenomenon (Table 5). However, these general bacterial indices did not reflect any of the vegetation data (Table 5). Bacterial activity AUC was positively correlated with WHC and P content (Table 5). Both bacterial functional diversity indices (R, H’bact) were positively correlated with OM, WHC, and N content, but negatively with soil pH (Table 5). Number of plant species, number of floor species and H’plant were positively correlated with amino acids use on Biolog® ECO plates (Table 5).

CLPP pattern did not differ between heap tops and slopes, as CLPP profiles between heap tops and slopes were strongly overlapping as shown in Fig. 1. Two NMDS dimensions allowed to obtain a stress value of 0.1842. Soil WHC, the share of invasive species and next soil P content were the most important factors shaping CLPP, surpassing other factors and environmental variables (Fig. 2).

Fig. 2
figure 2

Plot of the two dimensions of NMDS analysis of CLPP profiles according to site properties including soil characteristics: WHC, pH, and contents of phosphorus (P) and vegetation characteristics: coverage, plant diversity index H’plant, and share of invasive species in the total number of species (share inv.). Analysis results were presented for stand locations which were denoted with different symbols; the red triangles symbol denotes heap tops, and the blue triangles denote heap sides. Points representing stands of extreme values for each environment separately were connected by a line forming convect hulls; their overlapping indicates similarities in bacterial community CLPP

Full size image

The Mantel test showed that bacterial CLPP was not related to vegetation diversity and composition. The Bray–Curtis similarity measure indicated a relationship between CLPP and vegetation is not significant (R = 0.0399; p = 0.3041).

4 Discussion

Sustaining biological diversity is an extremely important challenge in a world rapidly changing under human activity. Recultivated post-industrial areas are an important element in maintaining the biological functions of soils in highly urbanized areas, including soil functions. This is also true for alkaline landfills. We showed that in this specific environment, the effect of vegetation on soil bacterial properties was moderate and the most important factor maintaining soil bacteria activity and functional diversity was simply OM content, which was correlated with other basic soil properties, such as WHC and nutrients content (mostly C and N). Although vegetation is an important source of soil organic fraction through the production of litter and root exudates (Rodríguez-Loinaz et al. 2008; Klimek et al. 2016), these soil properties were are apparently mostly related to the cap properties, created during the initial phase of recultivation processes.

The studied soils were characterised by relatively high activity and functional diversity indices, comparable to forest soils in Poland. Klimek et al. (2016) found that in various forest types, the averaged values for AUC, R, and H’bact were 39.0, 24.4, and 1.10, respectively. These values were found for soil A horizon, which corresponds to soils in the current study. Wasak et al. (2019), in their study conducted in a forested mountainous area near Kraków city, found averaged values for AUC, R, and H’bact were 31.1, 19.5, and 1.21, respectively. High soil bacteria activity has also been found in other recultivated landfills in Poland. For example, studies on reclaimed and afforested sand mine pits near Szczakowa in Upper Silesia in Poland revealed that AUC ranged up to 70.0 in soil A horizon (Chodak et al. 2015; Sroka et al. 2018). These authors also determined H’bact, but their calculations engaged another formula (using ln instead log10) and therefore their results cannot be directly compared with ours. In general, comparisons between studies using the common Biolog® method should be made with caution, as many procedural and computational issues can affect the final results obtained, such as the dilution of the microbial suspension (Rutgers et al. 2016).

Plot vegetation coverage on Białe Morza heaps was similar to the recultivated Kamieńsk (brown coal) heaps in central Poland (106% and 123%, respectively) (Rawlik et al. 2020). Białe Morza and Kamieńsk heaps, although originating from the activity of different industries, were similar because of the similar duration of the recultivation process (about 30 years) and the uniquely high substrate pH, which was alcaic on both heap caps (8.1 and 7.2, respectively). The average species richness of vascular plants on the Białe Morza heaps was lower than that obtained by Woś et al. (2018) on the former Piaseczno open-pit sulphur mine and Szczakowa sand pit mine excavation (Poland), which averaged 18.5. It is worth noting, that recultivation processes on the latter two sites started 60 years ago, resulting in a later successional chronosequence than on the Białe Morza site. For comparison, the number of vascular species in deciduous temperate forest stands in Poland ranges from 10 to 35 (Chodak et al. 2016). Meanwhile, some studies indicate that plant diversity is increasing in urban areas after four decades of unassisted vegetation development in a post-industrial area (Salisbury et al. 2021). On the other hand, species identity may be more important than species richness, at least because of the possibility that some species have an extraordinary effect on the functioning of the community (ecosystem) (so-called keystone species) (Giller and O’Donovan 2002). For example, silver birch, which occurred on more than 60% of plots in our study, has been confirmed to favour higher plant floor species diversity on post-industrial sites compared to other pioneer tree species (Woś et al. 2018).

In general, the recovery of ecosystem multifunctionality on post-industrial technosols is greater under higher biomass-producing species (Singh et al. 2023). Unfortunately, some high biomass-producing plant species are invasive (Czarniecka-Wiera et al. 2020). In disturbed ecosystems, such as post-industrial areas, there is a serious risk of invasion by alien species, which tend to dominate native species (Olszewski 2021). At the Białe Morza site, the proportion of invasive plant species in the total number of species was very high (18 ± 10%), while the overall proportion of invasive plant species on a national scale is 1.3%, and the combined proportion of invasive and potentially invasive species is approximately 3.5% (Tokarska-Guzik et al. 2012). For comparison, in a study of intraurban railway flora in East-Central European cities (Lublin and Lviv), invasive species accounted for 8.7% and 15.4%, respectively (Denisow et al. 2017). Two goldenrod species, Solidago canadensis and S. gigantea have especially strong negative effect on potential species diversity (Herr et al. 2007; Klimek et al. 2020b). Among the trees, an invasive species, Robinia pseudoacacia, which was a common tree species in the study area and the fourth numerous invasive species, is reported to grow in a wide range of habitats and therefore can be used to green habitats seriously disturbed by industry (Środek and Rahmonov 2022). Therefore, the planned introduction of high biomass producing native plant species during recultivation is particularly important for soil bacteria functioning in alkaline degraded soils.

The lack of relationships between soil bacterial CLPP and vegetation diversity and composition (Mantel test) could be due to the uniform influence of high pH on Solvay landfills, which is rather unique in the background of Poland, where most of the country is covered by the areas of acid soils with pH below 5.5 (Latawiec et al. 2017). So high soil pH on recultivated Solvay heaps continues for more than 100 years after recultivation, as shown in the Jaworzno area, also located in southern Poland (Sutkowska et al. 2015). Soil pH is extremely important for soil biota, especially bacteria, which are characterised by an enormously high ratio of surface area to volume (Portillo et al. 2013). Many studies have shown the importance of soil pH for soil bacteria in different soils, for example, Fierer and Jackson (2006) showed that soil pH was the best predictor of bacterial richness and diversity across various ecosystem types from North and South America. Similar results using analogue sampling were obtained by Rousk et al. (2010). Furthermore, these second authors assumed that the apparent direct effect of pH on bacterial community composition was probably due to the narrow soil pH ranges for optimal bacteria growth compared to fungi. Studies on the effect of soil pH on soil organisms are biased by other environmental factors that change simultaneously with soil pH, such as nutrients availability (Wan et al. 2021) or the composition of plant root exudates (Lauber et al. 2009; Massenssini et al. 2015), but it is generally accepted, that pH close to neutral is associated with the highest diversity of soil bacteria (Shen et al. 2013; Catania et al. 2022). This may be related to the fact, that many bacteria have intracellular pH values close to neutral (Madigan and Martinko 2006), and therefore pH values far from neutral may impose a significant stress that certain bacterial taxa. In our study, we found that bacterial functional diversity indices were negatively correlated with soil pH. These may be explained by the fact, that higher pH apparently occurred at plots with a thicker cap layer or excavated dump material. Surprisingly, all these factors did not differ between heap tops and slopes. Heap tops and slopes differed in plant species composition, meaning that a meadow-like vegetation assemblage occurred on tops, whereas assemblages on heap slopes were more forest-like. Nevertheless, this was not reflected in vegetation coverage or H’plant.

Carboxylic acids, carbohydrates and amino acids were among the most used carbon compounds, which can be associated with the fact that they are numerously represented on Biolog® ECO plates (9 carboxylic acids, 7 carbohydrates, and 6 amino acids in all 31 substrates). These chemical guilds of the substrates are particularly essential components of root exudates, which are known to support the growth and development of a bacterial community, hence differentiation between the experimental treatments has often been reported (Furtak et al. 2020). We found that soil bacteria from plots with a higher total number of plant species, number of floor species and H’plant utilised a higher proportion of amino acids, but not of any other chemical guilds of substrates on Biolog® ECO plates. Zhang et al. (2021) reported higher utilisation of amino acids on Biolog® ECO plates after the application of organic fertilizer to the soil. In fact, during the recultivation process in the Białe Morza area, the cap layer material was enriched with sewage sludge from municipal wastewater treatment plant (Okrutniak 2010), which may favour vegetation in some locations. On the other hand, it is questionable, whether the effect of a single fertilizer application effect can last so long. Moreover, amino acid use by bacteria on Biolog® ECO plates was negatively correlated with bacterial AUC and soil WHC, which may indicate the occurrence of stressful conditions for soil bacteria studied here. Chai and Schachtman (2022) reported, that the root exudation profiles of N-stressed plants often show a marked decrease in amino acids. Another possible explanation is that amino acids were exuded by small-sized N-fixing floor plants, which were more abundant in plots with higher vegetation diversity.

5 Conclusions

Soil is a globally universal environment, and soil microorganisms are characterised by high dispersal rates, which favour general parallels in soil microbial communities in many ecosystems. In our study, soil bacterial activity and functional diversity were shaped rather by soil properties than by vegetation properties, but improvement of soil OM content or WHC could be achieved by accurately managing vegetation diversity and composition. It is worth to pointing out again that unarranged urban greenings, also recultivated post-industrial areas may support biological diversity for both plants and soil microorganisms, the most important components of terrestrial ecosystems, and can provide important ecosystem services for citizens. There is a need to prevent biodiversity loss in disturbed and semi-natural landscapes, in which urban areas play an increasingly important role (e.g. Alvey 2006; Bonthoux et al. 2019).