Impact of Tree Species and Substrates on the Microbial and Physicochemical Properties of Reclaimed Mine Soil in the Novel Ecosystems

: Evaluating how different tree species and substrates affect the microbial and physico-chemical properties of technosols from combustion wastes and reclaimed mine soil (RMS) is vital in species selection to enhance restored ecosystem services. This research aimed to evaluate the impact of pioneer and N-ﬁxing tree species and substrates on the post-mining soil microbial and physicochemical properties. Common birch ( Betula pendula Roth) and Scots pine ( Pinus sylvestris L.), as the commonly introduced species on reclaimed mine soils (RMS) in eastern and central Europe, were selected as pioneer species, whereas black alder ( Alnus glutinosa (L) Gaernt.) and black locust ( Robinia pseudoacacia L.) were selected as N-ﬁxer species. Soil samples were collected from different RMS developed from three substrates (ﬂy ashes, clay, and sand) and measured for the content of total nitrogen (N t ), organic carbon (C org ), exchangeable calcium (Ca 2+ ), exchangeable potassium (K + ), exchangeable magnesium (Mg 2+ ), C to N ratio (C:N), basal respiration rate (RESP), and microbial biomass carbon (C mic ). The research indicated that tested tree species inﬂuenced water holding capacity (WHC), N t , C:N, and RESP value. The highest N t accumulation in soil was observed under N-ﬁxing, but it did not transfer into higher organic carbon content under N-ﬁxers. The soil under pine had a greater C:N ratio than the soil under birch, alder, and locust. The RESP rate was highest under birch. In terms of substrate type, RMS developed on Miocene clays exhibited higher carbon and macronutrient contents followed by ashes, whereas sands exhibited the lowest values of both physicochemical and microbial properties. The study suggested that both tree species and substrates affect microbial activities and physicochemical properties of RMS; however, the substrate effect is stronger.


Introduction
The exploitation of minerals and other geological materials (such as sand and coal) is the primary driver of ecosystem service degradation [1,2]. The post-mining soil usually exhibits disturbed water-air-soil relationships, very low soil microbial biomass, deficiency of available water, variation in texture (from high clay content to dominant coarse fraction), highly acid or alkaline pH, and nutrient deficiency [3][4][5][6]. These features restrict the growth and development of trees in reclaimed mine soils (RMS) [7,8]. Additionally, the planted trees frequently face significant constraints of essential nutrient supply [9,10]. However, if appropriate reclamation and rehabilitation methods are implemented, post-mining sites have a high potential for ecosystem services [11][12][13]. Afforestation is likely one of the most promising methods of mitigating the effects of exploitation and land degradation by

Soil Sampling and Measurements
A total of seventy-two (6 replications × 3 RMS substrates × 4 tree species) 10 × 10 m plots were established randomly. Soil samples were collected in August and September 2019. Tree stands on plots ranged in age from 18 to 44 years (Table 1). At each plot, one composite sample was collected from five locations (from four corners and at the center) at the depth of 0-5 cm with 5 cm diameter auger after the removal of the litter layer. Because of the presence of large waste rock fragments below 5 cm, the soil sample was only taken from the top layer. The samples were sieved with 2 mm mesh prior to laboratory analysis. Samples for physicochemical analyses were air-dried, whereas samples for microbial analyses were stored field-moist at 4 • C.
LECO TruMac CNS analyzer was used to analyze the content of organic carbon (C org ) and total nitrogen (N t  Prior to microbial analyses, the samples were adjusted to 50% of maximum water holding capacity (WHC) and pre-incubated at 22 • C for 6 days; WHC was determined gravimetrically according to Schlichting and Blume [41]. For microbial biomass carbon (C mic ) and basal respiration rate (RESP) measurement, samples (50 g d.w.) unamended for RESP measurements and amended with 8 mg glucose for C mic measurements were incubated at 22 • C in gas-tight jars. The incubation time was 24 h for the determination of RESP and 4 h for C mic . The jars contained beakers with 5 mL 0.2 M NaOH to trap the evolved CO 2 . After the jars were opened, 2 mL 0.9 M BaCl 2 was added to the NaOH; the excess of hydroxide was titrated with 0.1 M HCl in the presence of phenolphthalein as an indicator. C mic was calculated from the substrate-induced respiration rate according to the equation given by Anderson and Domsch (1978): C mic [mg g −1 ] = 40.04 y + 0.37, where "y" is mL CO 2 × h −1 × g −1 .

Statistical Analysis
Data were analyzed using Statistica 12.0 Software (StatSoft, Inc., (Tulsa, OK, USA), 2014). Prior to analysis, the data were tested for normality (Shapiro-Wilk test, p < 0.05). Two-way analysis of variance (ANOVA) was used to compare the effect of tree species and substrates on the measured properties. HSD test was run if significant (p < 0.05) effects were found. A linear correlation was used to assess the relationships between macronutrients and microbial activities. All correlation coefficients significant at p < 0.05 are shown.

Soil Texture and Water Holding Capacity
The soil under the pioneer and N-fixing tree species did not show any significant difference in soil texture on the same substrate ( Table 2). The lowest value of sand fraction was observed under alder on clays (22%) and the highest under pine on sands (89%). The lowest silt fraction was under pine on sands (5%) and the highest under black locust on ashes (30%) and for the clay fraction the lowest value was observed under birch on ashes and pine on sands (6%) and the highest under alder on clays (64%). The soils under birch on ashes revealed the highest WHC (104.7%), whereas soils under alder on sands revealed the lowest (29.8%; Table 2). Explanations: 71 ± 12-mean ± SD; P = pine; B = birch; A = alder; Bl = black locust; As = ashes; Sa = sands; Cl = clays; WHC = water holding capacity. Means in each column followed by different letters are significant at p < 0.05 level.

Soil Chemical Properties and Microbial Activities
Tree species (TS) had only effect on N t , C:N, and RESP, while substrate (SU) significantly affected all studied chemical and microbial activities. No interaction effect between tree species and soil substrate (TS × SU) was found (Table 3). Table 3. Two-way ANOVA results for tree species (TS) and substrate (SU) effect on soil chemical properties and microbial activities: N.S.-not significant, C org -organic carbon, N t -total nitrogen, Ca 2+ -calcium, K + -exchangeable potassium, Mg 2+ -exchangeable magnesium, C:N-C to N ratio, RESP-basal respiration rate, and C mic -microbial biomass carbon. The N t content was higher under N-fixing species (alder and locust) than under pine. Soils under pine exhibited significantly higher C:N ratio than under other tree species. Soils under birch exhibited significantly higher RESP compared to soils under pine and black locust (Table 4). Table 4. Tree species effect on soil chemical properties and microbial activities: C org -organic carbon, N t -total nitrogen, Ca 2+ -calcium, K + -exchangeable potassium, Mg 2+ -exchangeable magnesium, C:N-C to N ratio, RESP-basal respiration rate, and C mic -microbial biomass carbon. Means in each row followed by different letters are significant at p < 0.05 level.  (72.86 g kg −1 ) and N t (3.06 g kg −1 ), whereas sands exhibited the lowest values. Ca 2+ was significantly higher in ashes (4.21 g kg −1 ) and lower in sands (0.77 g kg −1 ). Sands also exhibited significantly lower values of K + , Mg 2+ and C:N (0.04 g kg −1 , 0.06 g kg −1 and 16.14, respectively). The higher RESP value was observed in ashes (2.40 µM CO 2 g −1 24 h −1 ) and the lower RESP value was observed in sands (1.67 µM CO 2 g −1 24 h −1 ). The highest value of C mic was measured in clays substrate (600.79 µg g −1 ) than ashes and sands (Figure 1).

Relationships between Macronutrients and Microbial Activities
Basal respiration rate and microbial biomass carbon were positively correlated with C org , N t , Ca 2+ , Mg 2+ , and K + . C org was positively correlated with N t , Ca 2+ , Mg 2+ , and K + . N t was also positively correlated with C org , Ca 2+ , Mg 2+ , and K + ( Table 5). The highest pHH2O and pHKCl values were observed in ashes (8.02 and 7.64, respectively), while the lowest values were in sands (5.68 and 4.84, respectively). Clays exhibited the highest values of Corg (72.86 g kg −1 ) and Nt (3.06 g kg −1 ), whereas sands exhibited the lowest values. Ca 2+ was significantly higher in ashes (4.21 g kg −1 ) and lower in sands (0.77 g kg −1 ). Sands also exhibited significantly lower values of K + , Mg 2+ and C:N (0.04 g kg −1 , 0.06 g kg −1 and 16.14, respectively). The higher RESP value was observed in ashes (2.40 μM CO2 g −1 24 h −1 ) and the lower RESP value was observed in sands (1.67 μM CO2 g −1 24 h −1 ). The highest value of Cmic was measured in clays substrate (600.79 μg g −1 ) than ashes and sands ( Figure 1).  Substrate effect on soil chemical properties and microbial activities: Corg-organic carbon, Nt-total nitrogen, Ca 2+ -calcium, K + -exchangeable potassium, Mg 2+ -exchangeable magnesium, C:N-C to N ratio, RESP-basal respiration rate, and Cmic-microbial biomass carbon. Means in each bar chart followed by different letters are significant at p < 0.05 level.

Relationships between Macronutrients and Microbial Activities
Basal respiration rate and microbial biomass carbon were positively correlated with Corg, Nt, Ca 2+ , Mg 2+ , and K + . Corg was positively correlated with Nt, Ca 2+ , Mg 2+ , and K + . Nt was also positively correlated with Corg, Ca 2+ , Mg 2+ , and K + ( Table 5). Substrate effect on soil chemical properties and microbial activities: C org -organic carbon, N t -total nitrogen, Ca 2+ -calcium, K + -exchangeable potassium, Mg 2+ -exchangeable magnesium, C:N-C to N ratio, RESP-basal respiration rate, and C mic -microbial biomass carbon. Means in each bar chart followed by different letters are significant at p < 0.05 level.

Tree Species Effect on Soil Properties and Microbial Activities
The impact of tree species on soil WHC differed significantly between birch and pine.
Cejpek et al. [42] also measured the lowest soil moisture in the reclaimed pine sites, whereas the highest soil moisture was in the reclaimed alder sites which may be related to soil development. The influence of soil substrate on studied chemical and biological properties was greater than that of tree species at the early stage of novel ecosystem development. Similarly, spoil heap after lignite mining in Poland had a greater impact on the content of C, N, P, C:N ratio, and microbial properties (microbial biomass carbon and respiration, enzyme activity) than tree species [30]. In contrast to our findings, Šnajdr et al. [43] found that tree species is the most important determinant of the chemical and biological properties of soils in the Sokolov brown-coal mining district's post-mining sites (Czech Republic). This phenomenon, however, could be attributed to the low diversity of substrates on the spoil heap in the Sokolov brown-coal mining district, which was formed of Miocene clays of the Cypris formation [43]. In contrast to Józefowska et al. [44], our study found no interaction effect between tree species and soil substrate on soil microbial biomass and respiration. According to these authors, the specific mix of tree species and the substrate has a large impact on faunal activity and microbial community composition.
The N t content was significantly higher under N-fixing species (alder and locust) than under pine ( Table 4). The higher N t value observed in soils under alder and black locust indicates the effectiveness of N-fixing tree species in enriching this deficient element in RMS. Similarly, Wang et al. [45] revealed that N-fixing forests had 20%-50% higher N t in the 0-5 cm soil layer than non-N-fixing forests. The lack of differences in N t content between birch and N-fixers could be attributed to the stimulation of non-symbiotic N-fixing bacteria growth [46,47]. The non-significant differences in C org content observed among tree species are opposite to the findings of previous studies. The soils under N-fixing species usually contain more carbon than soils under non-N-fixers [48]. For example, Ussiri et al. [49] observed a 42% increase in soil carbon under black locust in 10 years afforestation on reclaimed mine soils, while only 11% increase under pine, and Chodak and Niklińska [30] observed higher organic carbon accumulation under alder than under pine. Increased soil organic matter input and reduced decomposition of older soil carbon are associated with higher C org content in N-fixing tree species [48]. However, this phenomenon was not observed in our soils.
Soils under pine exhibited significantly higher C:N ratio than under other tree species. The lowest C:N values were observed under alder and black locust. This may be due to the lower carbon to nitrogen ratio in the litter of nitrogen fixing species [27,50,51]. Similarly, Józefowska et al. [44] observed lowest (14) C:N under alder compared to non-N-fixing species growing on different substrates in RMS.
There was no difference observed for microbial biomass among tree species. However, the highest RESP was exhibited under birch ( Table 4). The previous studies by Chodak and Niklińska [37] and Józefowska et al. [44] also revealed higher respiration rate and microbial biomass under birch than pine on post-mining sites. Differences in microbial biomass and basal respiration rate between tree species and substrates may have implications for tree nutrient availability. High microbial biomass and basal respiration rate frequently result in high nutrient availability to trees [52,53], by increasing both microbial biomass turnover and non-microbial organic material degradation. This result showed that a strong and positive correlation exists between soil microbial biomass and basal respiration rate with N t content (Table 5). This confirms the findings of Józefowska et al. [44], who indicated that N t content was positively and strongly correlated with soil microbial activities in mine soil. Frouz et al. [27] also reported a positive correlation between soil microbial biomass and N t content. Soil microorganisms play a vital role in the biological transformations and develop most of the stable N and C nutrients and other vital nutrients [54].

Substrate Effect on Soil Properties and Microbial Activities
All of the investigated physicochemical and microbial properties were significantly influenced by the substrates. The lowest WHC values were observed in sands. Ashes exhibited the highest values of WHC (86.7%-104.7%) followed by clays (62.1%-75.9%). The storage of water on spoil heaps is hampered by the soil's incomplete development [55].
The pH values varied significantly across all substrates. According to Soil Science Division Staff [56], sands were strongly acidic to very strongly acidic, clays were moderately acidic to slightly acidic, and ashes were moderately alkaline to slightly alkaline, which is a characteristic of combustion wastes [40,57].
C org content varied significantly across substrates, with clays having significantly higher C org content than ashes and sands. In line with this, Chodak and Niklińska [37] revealed that substrate texture had a greater effect on C org of RMS than tree species. Woś and Pietrzykowski [39] also observed the highest organic carbon stock in soils developed from Neogene (Miocene) clays, which is linked to the element's in-situ accumulation and geogenic (fossil) carbon content. The differences found in our study could also be attributed to geogenic carbon content [58]. In contrast to sands, Miocene clays contain some geogenic carbon [59]. The higher values found in clay-based soils revealed that soils with more silt and clay content have more organic carbon [39]. This discrepancy is caused by clay particles binding SOC in the form of organic-mineral complexes. These complexes are resistant to leaching and microorganism breakdown [60,61]. Because of the incomplete combustion of lignite in the power plant, combustion wastes contain some unburned carbon [62,63]. In the case of ashes from the Bełchatów power plant, the amount of unburned material was estimated to be 2% [64]. Despite this, the carbon content of ashes-derived soils was lower than that of clay and comparable to that of sands. Clays, like C org , had a higher N t value than ashes and sands. C org accumulation in soils is generally highly correlated with N t accumulation [65,66]. Similarly, Treschevskaya et al. [67]'s study on soil development processes on post-mining sites found that clay substrates had higher N t content than chalk and marl, sand, and sandy loam.
Significantly higher Ca 2+ was observed in ashes than in sands and clays and it was mainly due to the combustion of lignite coal [68]. Ashes and clays, on the other hand, had significantly higher K + and Mg 2+ values than sands. Except for Ca 2+ , the observed higher macronutrient content in clay may be due to its high capacity for holding macronutrients [69]. Mining waste sites with sandy substrates are characterized by unfavorable soil structure and nutrient deficiencies [70].
The study revealed that substrates have a greater influence on microbial activities than tree species. Clays, in particular, had greater microbial biomass than ashes and sands. In line with this, Chodak and Nikliska [37] found that soil texture, rather than vegetation properties, influenced microbial activity on reforested mine spoils in Poland. The type of soil is a major factor in determining microbial activity [71,72]. Carletti et al. [73] found that geologic parent material had a significant impact on the microbial community in northern Italian montane spruce forests. According to Józefowska et al. [25] and Reich et al. [74], the substrate influences soil biological properties in both natural and post-mining soils.

Conclusions
The substrate influences physicochemical and microbial properties more than tree species. Except for N t , there is no significant difference in physicochemical and microbial properties between pioneer and N-fixing tree species. The highest carbon and macronutrient content was found in Miocene clays, followed by combustion wastes, and the lowest physicochemical and microbial properties were found in sands. The study found that tree species influenced WHC, N t , C:N, and RESP value. The highest N t in soils was observed under N-fixers, but it did not transfer into higher organic carbon content under N-fixers. The C:N ratio of pine soils was higher than that of birch, alder, and locust soils. Birch had the highest basal respiration rate. As a result of the stronger effect of substrate on RMS than tree species, reclamation of the substrate improves physicochemical and microbial properties and facilitates ecosystem restoration in mine sites.

Data Availability Statement:
Supplementary data to this article can be found online at https://data. mendeley.com//datasets/86wywbx3ps/2, accessed on 18 October 2022.

Conflicts of Interest:
The authors declare that there is no known competing financial interests nor personal relationships that could have appeared to influence the work reported in this paper.