The Response of the Soil Microbiome to Contamination with Cadmium, Cobalt and Nickel in Soil Sown with Brassica napus

: Soil fertility is determined by biological diversity at all levels of life, from genes to entire biocenoses. The aim of this study was to evaluate bacterial diversity in soil contaminated with Cd 2+ , Co 2+ and Ni 2+ and sown with Brassica napus . This is an important consideration because soil-dwelling microorganisms support phytoremediation and minimize the adverse effects of heavy metals on the environment. Microbial counts, the inﬂuence (IF HM ) of Cd 2+ , Co 2+ and Ni 2+ on microorganisms, the colony development (CD) index, the ecophysiological diversity (EP) index and genetic diversity of bacteria were determined under controlled conditions. Soil contamination with Cd 2+ , Co 2+ and Ni 2+ signiﬁcantly inﬂuenced microbial diversity and increased the values of CD and EP indices. The tested heavy metals decreased the genetic diversity of bacteria, in particular in the phyla Actinobacteria and Proteobacteria . Bacteria of the genera Arthrobacter, Devosia, Kaistobacter, Paenibacillus, Phycicoccus, Rhodoplanes and Thermomonas were identiﬁed in both contaminated and non-contaminated soil. These bacteria are highly resistant to soil contamination with Cd 2+ , Co 2+ and Ni 2+ .


Introduction
Environmental stress caused by various pollutants, including heavy metals, has a negative effect on the biological activity of soil. Any changes in environmental conditions can modify the composition of soil-dwelling microorganisms and their activity [1]. Biological diversity should be preserved and changes that take place at all levels of life, including microorganisms that occupy the lowest level of the trophic system, have to be monitored to maintain ecosystem health [2]. Microorganisms constitute the largest group of soil-dwelling organisms. They participate in the circulation of elements in nature, including carbon, nitrogen, sulfur and phosphorus, the decomposition of organic matter, energy flow and they influence the physical parameters of soil and soil structure [3][4][5]. Unlike organic pollutants, heavy metals do not undergo microbial decomposition [6] and their total concentrations in soil persist for a long time after deposition [7]. Heavy metals lead to the biological and chemical degradation of soil [8]. They exert a negative effect on microbial populations and the activity of soil enzymes. Heavy metals disrupt not only microbial diversity, but also biochemical processes [9,10]. Natural and anthropogenic changes in the environment elicit quick responses from microorganisms which, due to their sensitivity, can rapidly adapt to new conditions. Microorganisms have developed defense mechanisms that minimize the toxic effects of heavy metals [11]. These mechanisms involve extracellular and intracellular binding of trace elements. The most widely encountered extracellular mechanisms which eliminate heavy metals include ATP-binding cassettes (ABC), inner membrane transporters of the resistance-nodulation-division (RND) family, small multidrug resistance (SMR) proteins, major facilitator superfamily (MFS) proteins and multidrug and toxic compound extrusion (MATE) proteins [12,13]. Intracellular binding mechanisms involve metallothionein proteins with high affinity for divalent cations

Experimental Design
A pot experiment was established in a greenhouse (Poland) with four replications. In the first stage, soil samples of 3.5 kg each were passed through a sieve with 1 cm mesh size. In the second stage, soil samples were contaminated with nickel, cobalt and cadmium chlorides at Ni 2+ -400 mg kg −1 , Co 2+ -80 mg kg −1 and Cd 2+ -8 mg kg −1 . Each treatment had 4 replicates (n = 16). Heavy metal doses were established based on the Regulation of the Minister of the Environment [38] and soil contamination levels in Poland [39][40][41][42] and other European Union countries and the world [43,44]. In Polish soils, the average content of Ni 2+ has been estimated at 20 mg kg −1 [39]. Soil contamination with Ni 2+ does not occur frequently around the world, but in some areas, human activity has led to excessive Ni 2+ accumulation in soil. High levels of soil contamination with Ni 2+ have been reported near the Sudbury steel mill in Canada (1600-2150 mg kg −1 ) [45,46], in the vicinity of the non-ferrous metal factory in Plovdiv, Bulgaria (303 mg kg −1 ) [47], Selebi Phikwe copper-nickel mine in Botswana (267 mg kg −1 ) [48], the former sewage sludge landfill in Denmark (212 mg kg −1 ) [49] and a defunct waste incineration plant in the Czech Republic (122 mg kg −1 ) [50]. The geochemical background value of cobalt in soils has been determined at 6.5 mg kg −1 DM in Poland and at 4.5-12 mg kg −1 DM around the world [39]. Point-source pollution with cobalt has also been reported in Japan (23 mg kg −1 soil) [51], Egypt (26.8 mg kg −1 ) [52], Sweden (14 mg kg −1 ) [53], in the vicinity of nonferrous metal ore mines in southern Sardinia (30-495 kg −1 DM) [54] and non-ferrous metal factories in the USA (>154 mg kg −1 ) [39]. Natural cadmium levels range from 0.06 to 1.1 mg kg −1 in Polish soils [39] and from 0.2 to 1.3 mg kg −1 in European soils [55]. In locally contaminated soils in the USA (Arizona), cadmium deposition was determined at 394 mg kg −1 to 1777 mg kg −1 [56]. In urban soils, cadmium concentrations were determined at 6.7 mg kg −1 in Pakistan, 2.6 mg kg −1 in Iran and 1.2 mg kg −1 in Denmark. In agricultural soils fertilized with sewage sludge, cadmium levels reached 19.3 mg kg −1 in India, around 3.5 mg kg −1 in Pakistan, 2.5 mg kg −1 in Zimbabwe and around 2.0 mg kg −1 in Serbia, Egypt and Mexico [57]. Brassica napus plants were fertilized with the following macronutrients: N-70 mg kg −1 soil DM, P-30 mg kg −1 soil DM, K-56 mg kg −1 soil DM and Mg-10 mg kg −1 soil DM. Fertilizers were applied in the following form: nitrogen-CO(NH 2 ) 2 , phosphorus-KH 2 PO 4 , potassium-KCl and magnesium-MgSO 4 · 7H 2 O. CO(NH 2 ) 2 contained 46.6% N, KH 2 PO 4 -22.8% P; KCl-52.4% K, and MgSO 4 · 7H 2 O-9.9% Mg. Soil samples were combined with fertilizers, placed in polyethylene pots and brought to capillary water capacity of 50%. Soil not contaminated with heavy metals but fertilized with macronutrients was the control. The following soil treatments were analyzed: (1) uncontaminated soil (control), (2) soil contaminated with Ni 2+ , (3) soil contaminated with Co 2+ and (4) soil contaminated with Cd 2+ . The average ambient temperature was 16.5 • C, air humidity was 77.5% and daytime length ranged from 14 h 4 min to 16 h 30 min. The following soil treatments were analyzed: uncontaminated soil and soil contaminated with cadmium, cobalt and nickel. The emerged seedlings were thinned and six Brassica napus plants were left per pot. The plants were harvested during flowering (BBCH 60-Biologische Bundesanstalt Bundessortenamt und Chemical Scale). The experiment lasted 50 days. Soil samples for analyses of microbiological properties were collected in the same period.

Method of Microbiological Analysis
The counts of organotrophic bacteria, actinobacteria and fungi were determined by the serial dilution method. Soil samples (10 g) were placed in sterile isotonic saline solution (90 cm 3 0.85% NaCl) and shaken for 30 min (130 rpm). One cm 3 of each serial solution and 18 cm 3 of a selective medium were placed in sterile Petri dishes and incubated in a thermostat at a temperature of 28 • C for 10 days. Colony forming units (cfu) were counted daily for 10 days. Microbiological analyses were carried out according to the procedures described by Boros-Lajszner et al. [58]. The growth rate of various microbial groups (organotrophic bacteria, actinobacteria and fungi) and the number of emerged microbial colonies were determined to calculate the colony development (CD) index and the ecophysiological diversity (EP) index in uncontaminated soil and in soil contaminated with Ni 2+ , Co 2+ and Cd 2+ . The microbial CD index was calculated based on the formula developed by De Leija et al. [59] and the EP index, based on the formula proposed by Sarathchandra et al. [60].
The microbial CD index was calculated with the use of the following formula: where N 1 , N 2 , N 3 . . . N 10 is total number of microbial colonies on experimental days 1, 2, 3 10 divided by the total number of colonies identified during the entire experimental period. The microbial EP index was determined with the following formula: where pi is the number of microbial colonies on a given day divided by the total number of colonies.
The influence of heavy metals (IF Hm ) on microbial counts and Brassica napus was determined with the use of the following formula: where IF Hm -heavy metal pollution index, A Hm -microbial counts/number of plants in soil contaminated with heavy metals, A 0 -microbial counts/number of plants in uncontaminated soil.
DNA was isolated from the rhizosphere of Brassica napus with the Genomic Mini AX Bacteria + kit. The polymerase chain reaction was carried out with the Q5 Hot Start Hight-Fidelity 2X Master Mix according to the manufacturer's instructions. A metagenomic analysis of the 16S rRNA coding gene was performed based on the sequence of the V3-V4 hypervariable region of 16S rRNA. The selected region was amplified and a library was developed with the use of specific polymer sequences 341F and 785R. Data were collected and standardized in the last stage of library development. The 16S rRNA gene was sequenced in the MiSeq Reporter (MSR) v 2.6 program with 2 × 250 bp paired-end runs in Illumina v 2. The reads were identified to the genus level in the QIIME program based on the GreenGenes v 13.8 reference database. Sequencing was performed by Genomed SA (Warsaw, Poland).

Statistical Analysis
Data were processed statistically in Statistica 13.1 (Palo Alto, CA, USA) [61]. The results were compared in ANOVA and Tukey's HSD test. Homogeneous groups were identified at a significance level of p = 0.05. The analyzed data had normal distribution and similar variance. Metagenomic profiles were analyzed in STAMP 2.1.3. [62] and Circos 0.68 software [63]. The number of operational taxonomic units (OTUs) associated with each class and order was directly proportional to each bandwidth combining bacterial taxa with the corresponding soil samples. Each bacterial class and order were represented by a unique color. The external ring represents the total percentage of 16S sequences and the internal ring denotes the number of 16S amplicon sequences allocated to a given taxon.

Results
The microbiological properties of soil were modified under the influence of Ni 2+ , Co 2+ and Cd 2+ (Table S1). Heavy metals significantly decreased the abundance of organotrophic bacteria and fungi, but their influence on actinobacteria was less pronounced. These observations were confirmed by the values of the heavy metal pollution index (IF HM ) ( Figure 1). Ni 2+ followed by Co 2+ and Cd 2+ exerted the greatest negative effect on the soil microbiome, in particular, organotrophic bacteria and fungi. Actinobacteria were least sensitive to heavy metal contamination.
tified at a significance level of p = 0.05. The analyzed data had normal distribution and similar variance. Metagenomic profiles were analyzed in STAMP 2.1.3. [62] and Circos 0.68 software [63]. The number of operational taxonomic units (OTUs) associated with each class and order was directly proportional to each bandwidth combining bacterial taxa with the corresponding soil samples. Each bacterial class and order were represented by a unique color. The external ring represents the total percentage of 16S sequences and the internal ring denotes the number of 16S amplicon sequences allocated to a given taxon.

Results
The microbiological properties of soil were modified under the influence of Ni 2+ , Co 2+ and Cd 2+ (Table S1). Heavy metals significantly decreased the abundance of organotrophic bacteria and fungi, but their influence on actinobacteria was less pronounced. These observations were confirmed by the values of the heavy metal pollution index (IFHM) ( Figure  1). Ni 2+ followed by Co 2+ and Cd 2+ exerted the greatest negative effect on the soil microbiome, in particular, organotrophic bacteria and fungi. Actinobacteria were least sensitive to heavy metal contamination.
In soil contaminated with nickel, the colony development (CD) index of organotrophic bacteria (CD = 52.8) was higher and the CD index of actinobacteria (CD = 26.3) and fungi (CD = 38.2) (Figure 2) was lower than in the control treatment. The ecophysiological diversity (EP) index of organotrophic bacteria, actinobacteria and fungi was significantly higher in soil contaminated with heavy metals (Figure 2). The EP index was highest in actinobacteria, followed by organotrophic bacteria and it was lowest in fungi (EP = 0.42-0.47).   (Figure 2) was lower than in the control treatment. The ecophysiological diversity (EP) index of organotrophic bacteria, actinobacteria and fungi was significantly higher in soil contaminated with heavy metals (Figure 2). The EP index was highest in actinobacteria, followed by organotrophic bacteria and it was lowest in fungi (EP = 0.42-0.47). Bacterial phyla Actinobacteria and Proteobacteria were predominant in soil samples contaminated with heavy metals (Figure 3). Actinobacteria was the dominant phylum in all treatments and it accounted for 35.6% of the microorganisms in soil contaminated with nickel, 57.3%-in soil contaminated with cadmium, 61.5%-in soil contaminated with cobalt and 62.5%-in uncontaminated soil ( Figure 3). Proteobacteria was the second most abundant phylum, which was predominant in nickel-contaminated soil (49.1%). The phylum Actinobacteria was characterized by the greatest variations in the number of OTUs. The number of OTUs specific to Actinobacteria was 25.9% higher in soil contaminated with cobalt than nickel and it was 21.7% higher in soil exposed to cadmium than nickel.
(a) Bacterial phyla Actinobacteria and Proteobacteria were predominant in soil samples contaminated with heavy metals (Figure 3). Actinobacteria was the dominant phylum in all treatments and it accounted for 35.6% of the microorganisms in soil contaminated with nickel, 57.3%-in soil contaminated with cadmium, 61.5%-in soil contaminated with cobalt and 62.5%-in uncontaminated soil (Figure 3). Proteobacteria was the second most abundant phylum, which was predominant in nickel-contaminated soil (49.1%). The phylum Actinobacteria was characterized by the greatest variations in the number of OTUs. The number of OTUs specific to Actinobacteria was 25.9% higher in soil contaminated with cobalt than nickel and it was 21.7% higher in soil exposed to cadmium than nickel. Bacterial phyla Actinobacteria and Proteobacteria were predominant in soil samples contaminated with heavy metals (Figure 3). Actinobacteria was the dominant phylum in all treatments and it accounted for 35.6% of the microorganisms in soil contaminated with nickel, 57.3%-in soil contaminated with cadmium, 61.5%-in soil contaminated with cobalt and 62.5%-in uncontaminated soil (Figure 3). Proteobacteria was the second most abundant phylum, which was predominant in nickel-contaminated soil (49.1%). The phylum Actinobacteria was characterized by the greatest variations in the number of OTUs. The number of OTUs specific to Actinobacteria was 25.9% higher in soil contaminated with cobalt than nickel and it was 21.7% higher in soil exposed to cadmium than nickel. Contamination with Cd 2+ , Co 2+ and Ni 2+ modified the soil microbiome also at the class level ( Figure S1). In uncontaminated soil and in soil contaminated with cadmium, cobalt and nickel, the highest number of OTUs were associated with the classes Actinobacteria (55.7%, 47.7%, 52.4%, 18.2%), Alphaproteobacteria (15.2%, 16.1%, 15.3%, 29.3%), Betaproteobacteria (5.1%, 7.5%, 5.3%, 4.0%), Thermoleophilia (5.1%, 6.8%, 6.9%, 15.4%) and Gammaproteobacteria (4.6%, 4.0%, 4.1%, 15.7%), respectively. At the level of microbial classes, heavy metals (in particular, nickel relative to cadmium and cobalt) exerted the most varied effects on Actinobacteria and Gammaproteobacteria, reflected in the differences in the number of their OTUs.
Bacteria of the family Micrococcaceae (phylum Actinobacteria) were predominant in control soil and in treatments contaminated with Cd 2+ and Co 2+ ( Figure 5). Under exposure to Ni 2+ , the number of Micrococcaceae OTUs decreased nearly 20-fold, whereas a significant increase was observed in the abundance of bacterial families Gaiellaceae, Nocardioidacea and Intrasporangiaceae of the phylum Actinobacteria, family Xanthomonadaceae of the phylum Proteobacteria and families Sphingobacteriaceae and Hyphomicrobiaceae of the phylum Bacteroidetes. The structure of bacterial communities was also modified in treatments contaminated with Cd 2+ and Co 2+ relative to control, but these changes were not as spectacular as those observed in Ni 2+ treatments. The number of OTUs associated with bacterial families Gaiellaceae and Hyphomicrobiaceae was significantly higher in soil contaminated with Cd 2+ and Co 2+ than in uncontaminated soil. In turn, bacteria of the family Nocardioidaceae were more abundant in soil contaminated with cobalt than in the control treatment and in soil exposed to cadmium. Minerals 2021, 11, x FOR PEER REVIEW 9 of 16 Bacteria of the family Micrococcaceae (phylum Actinobacteria) were predominant in control soil and in treatments contaminated with Cd 2+ and Co 2+ ( Figure 5). Under exposure to Ni 2+ , the number of Micrococcaceae OTUs decreased nearly 20-fold, whereas a significant increase was observed in the abundance of bacterial families Gaiellaceae, Nocardioidacea and Intrasporangiaceae of the phylum Actinobacteria, family Xanthomonadaceae of the phylum Proteobacteria and families Sphingobacteriaceae and Hyphomicrobiaceae of the phylum Bacteroidetes. The structure of bacterial communities was also modified in treatments contaminated with Cd 2+ and Co 2+ relative to control, but these changes were not as spectacular as those observed in Ni 2+ treatments. The number of OTUs associated with bacterial families Gaiellaceae and Hyphomicrobiaceae was significantly higher in soil contaminated with Cd 2+ and Co 2+ than in uncontaminated soil. In turn, bacteria of the family Nocardioidaceae were more abundant in soil contaminated with cobalt than in the control treatment and in soil exposed to cadmium. The number of OTUs associated with specific bacterial genera exceeded 1% in eight genera (Devosia, Kaistobacter, Rhodoplanes, Thermomonas, Mesorhizobium, Lysobacter, Rhodobacter, Sphingomonas) of the phylum Proteobacteria, in two genera (Arthrobacter, Phycicoccus) of the phylum Actinobacteria, in two genera (Bacillus, Paenibacillus) of the phylum Firmicutes and in one genus (Deinococcus) of the phylum Thermi. Bacterial genera Arthrobacter, Devosia, Kaistobacter, Paenibacillus, Phycicoccus, Rhodoplanes and Thermomonas were identified in uncontaminated soil and in soil contaminated with cadmium, cobalt and nickel ( Figure 6). Bacillus and Mesorhizobium were dominant in soil contaminated with Cd 2+ , Co 2+ and Ni 2+ . Genera Lysobacter and Rhodobacter were most abundant in soil exposed to nickel. The number of OTUs specific to Deinococcus was highest in control soil and in treatments contaminated with cadmium and cobalt, whereas the abundance of Sphingomonas was highest in the control treatment and in soil contaminated with Ni 2+ . The correlation coefficients presented in Table S2 confirmed the presence of significant correlations between the microbial community and Brassica napus yields. Plant yields were positively correlated with the counts of all microbial groups and the colony development (CD) index of actinobacteria and fungi and they were negatively correlated with the CD index of organotrophic bacteria, the ecophysiological diversity (EP) index of actinobacteria and the Shannon-Wiener index describing the diversity of bacterial communities at the phylum, class, order and family level. The number of OTUs associated with specific bacterial genera exceeded 1% in eight genera (Devosia, Kaistobacter, Rhodoplanes, Thermomonas, Mesorhizobium, Lysobacter, Rhodobacter, Sphingomonas) of the phylum Proteobacteria, in two genera (Arthrobacter, Phycicoccus) of the phylum Actinobacteria, in two genera (Bacillus, Paenibacillus) of the phylum Firmicutes and in one genus (Deinococcus) of the phylum Thermi. Bacterial genera Arthrobacter, Devosia, Kaistobacter, Paenibacillus, Phycicoccus, Rhodoplanes and Thermomonas were identified in uncontaminated soil and in soil contaminated with cadmium, cobalt and nickel ( Figure 6). Bacillus and Mesorhizobium were dominant in soil contaminated with Cd 2+ , Co 2+ and Ni 2+ . Genera Lysobacter and Rhodobacter were most abundant in soil exposed to nickel. The number of OTUs specific to Deinococcus was highest in control soil and in treatments contaminated with cadmium and cobalt, whereas the abundance of Sphingomonas was highest in the control treatment and in soil contaminated with Ni 2+ . The correlation coefficients presented in Table S2 confirmed the presence of significant correlations between the microbial community and Brassica napus yields. Plant yields were positively correlated with the counts of all microbial groups and the colony development (CD) index of actinobacteria and fungi and they were negatively correlated with the CD index of organotrophic bacteria, the ecophysiological diversity (EP) index of actinobacteria and the Shannon-Wiener index describing the diversity of bacterial communities at the phylum, class, order and family level.

Discussion
Soil has a complex structure and it is a reservoir of various chemical compounds and trace elements that can disrupt the homeostasis of the soil environment. Heavy metals, including cadmium, cobalt and nickel, are toxic at high concentrations and can induce

Discussion
Soil has a complex structure and it is a reservoir of various chemical compounds and trace elements that can disrupt the homeostasis of the soil environment. Heavy metals, including cadmium, cobalt and nickel, are toxic at high concentrations and can induce changes in the microbiological properties of soil [64]. In the present study, the calculated values of heavy metal pollution index (IF HM ) confirmed that heavy metals compromised the activity of the examined microbial groups, in particular, organotrophic bacteria and fungi. Actinobacteria were least sensitive to the toxic effects of Cd +2 , followed by Co +2 and Ni +2 . The survival of this microbial group can probably be attributed to the acquisition of resistance to adverse environmental conditions. The physicochemical properties of soil, including pH, sorption capacity and organic matter content, could also play a role in microbial survival [65]. Microorganisms have evolved resistance to harsh environmental conditions. Prolonged exposure to trace elements can induce changes in cell structure and lead to microbial death [66][67][68]. According to Sang-Hwan et al. [69], pollutants affect the rate of soil processes and can inhibit soil respiration. Microbial imbalances compromise the diversity of the soil microbiome.
Soil-dwelling microorganisms are effective indicators of soil quality [70]. Exposure to heavy metals induces changes in the proportion of r-/K-strategists in the bacterial community. According to Sarathchandra et al. [60], an increase in the value of the CD index implies that r-strategists (rapidly growing microorganisms) are more abundant than K-strategists (slowly growing microorganisms). A decrease in the value of the EP index suggests that sensitive microorganisms are replaced by microorganisms that are more resistant to soil contamination with heavy metals [59]. In the current study, the CD index for organotrophic bacteria was higher in treatments contaminated with Ni 2+ and Co 2+ ; the CD index for actinobacteria was higher in soil exposed to Cd 2+ and Co 2+ ; and the CD index for fungi was higher in treatments contaminated with Co 2+ than in uncontaminated soil. The value of the EP index was highest in Actinobacteria, followed by organotrophic bacteria and fungi. In polluted regions, microbial activity also contributes to immobilization of metals in the soil environment [71,72]. Microorganisms can neutralize pollutants through their accumulation, binding and storage in the form of phosphates, oxalates and carbonates in cell structures and organelles [73,74].
Heavy metal pollution changes microbial diversity in many terrestrial ecosystems [75,76]. In this study, soil contamination with Cd +2 , Co +2 and Ni +2 also induced changes in microbial diversity at all taxonomic levels. Bacterial phyla Actinobacteria and Proteobacteria were predominant and similar observations were by Franke-Whittle et al. [77], Sun et al. [78] and Greening et al. [79]. Proteobacteria occur not only in soils contaminated with heavy metals [80], but also in polluted waters [81] and sediments [75]. In a study by Epelde et al. [82], the proportions of microbial taxa with high metabolic plasticity, including Chloroflexi and Ktedonobacteria, increased in the bacterial community in the presence of metals. In contrast, the proportions of Actinobacteria and Acidobacteria remained stable. In the present experiment, in addition to Proteobacteria and Actinobacteria, bacterial genera of the phyla Firmicutes and Thermi were also resistant to soil contamination with Cd 2+ , Co 2+ and Ni 2+ . Similar results were reported by Pradhan et al. [83] and Yang et al. [84] who analyzed bacterial communities in soils exposed to heavy metals. According to Emenike et al. [85], native microorganisms can effectively remove heavy metals from soil. Microorganisms develop tolerance to heavy metals by binding these contaminants on the surface of cells [86,87]. Peptidoglycan carboxyl groups are the main cationic binding sites in Gram-positive bacteria and phosphate groups-in Gram-negative bacteria [21]. In the current study, both Gram-positive and Gram-negative bacteria were identified in contaminated soil. The predominant Gram-positive bacteria were Arthrobacter, Phycicoccus, Bacillus, Paenibacillus and Deinococcus, whereas the most abundant Gram-negative bacteria were Devosia, Kaistobacter, Rhodoplanes, Thermomonas, Mesorhizobium, Lysobacter, Rhodobacter and Sphingomonas. These bacteria can be used in the bioaugmentation of soils contaminated with Cd 2+ , Ni 2+ and Co 2+ to restore soil homeostasis.
Microbial tolerance to heavy metals is an individual trait. There is evidence to indicate that microorganisms survive in contaminated environments by reducing metals to a lower oxidation (less toxic) state or by binding metals with the involvement of specific and non-specific compounds that are metabolic products of microorganisms [19,22,88,89]. Specific compounds include metallothioneins and non-specific compounds include lowmolecular-weight organic acids, alcohols and high-molecular-weight polysaccharides [22]. De Jaysankar et al. [90] identified bacteria tolerant to Hg, Cd and Pb. In other studies, bacteria resistant to Cr, Zn, Cd, Pb, Cu, Ni and Co were isolated from various environments contaminated with heavy metals [91,92]. However, microorganisms resistant to changing environmental conditions can also disrupt soil homeostasis by eliminating sensitive microorganisms. Bacteria characteristic of heavy metal-contaminated soils were also isolated in the present study. These were: Bacillus and Mesorhizobium which were predominant in treatments contaminated with Cd 2+ , Co 2+ and Ni 2+ , as well as Lysobacter and Rhodobacter that were most abundant in soil exposed to Ni 2+ . The evolution of resistant bacteria decreases biodiversity in a given environment [93].

Conclusions
Contamination with Ni 2+ , Co 2+ and Cd 2+ disrupts the microbiological balance of soil. In the group of the analyzed heavy metals, Ni 2+ induced the greatest and Cd +2 induced the smallest changes in the soil microbiome. Heavy metals contributed to changes in the values of the colony development (CD) index and the ecophysiological diversity (EP) index. Exposure to heavy metals modified the proportions of r-and K-strategists in soil. The microbial EP index was higher in treatments contaminated with heavy metals than in uncontaminated soil. The metagenomic analysis revealed that Actinobacteria and Proteobacteria were the dominant bacterial phyla. Bacteria of the genera Arthrobacter, Devosia, Kaistobacter, Paenibacillus, Phycicoccus, Rhodoplanes and Thermomonas were identified in uncontaminated soil as well as in treatments contaminated with heavy metals. These bacteria are characterized by high resistance to soil contamination with Ni 2+ , Co 2+ and Cd 2+ .