3.1. Soil characterisation pre-treatment
The gold mine waste, grey sands, was sampled from Greater Bendigo near residential housing. Soil characteristics for pre-treatment were identified (Table 1). Arsenic concentrations were 801 ± 83.4 mg/kg, exceeding the national soil guidance values for residential soils (100 mg/kg). Previous studies have reported As concentrations in grey sands collected from regional Victoria to range between 414 - 2310 mg/kg [5,6,59]. Therefore, As concentrations of ca. 1000 mg/kg was expected and common for grey sands. An exceedance of the national soil guidance values requires further investigation and remediation to minimise human health and ecological risk [11].
The grey sand contained 0.72 ± 0.07% carbon and 0.27 ± 0.11% nitrogen. Soil organic carbon and nitrogen are essential plant macronutrients and indicators for soil productivity/fertility in the agricultural sector [77,78]. In this study, carbon concentrations are low when compared to fertile agricultural soil. For example, a study by Debska et al. [77] aimed to improve soil fertility for a tilled monoculture wheat crop. Carbon and nitrogen concentrations were 2.23 ± 0.06% and 0.24 ± 0.01%, respectively [77]. As stated earlier, carbon is an essential nutrient and an indicator of soil fertility. The grey sands carbon concentrations were depleted, indicating a lack of nutrients, and phytoremediation is needed to improve soil conditions. Other soil characteristics included sand, silt, and clay which were 41.40 ± 2.60%, 37.20 ± 1.30%, and 21.40 ± 1.67%, respectively, and soil texture was a clay loam. Moisture content was 10 ± 0.3% and the pH was 7.78 ± 0.05.
3.2. Phytostabilisation of contaminated soil
3.2.1. Plant growth
Three Australian natives, J. usitatus, P. labillardieri, and T. triandra, were assessed for the ability to phytoremediate As in contaminated gold mine waste, specifically grey sands. Three-month-old seedlings with 5-6 shoots were transplanted into 1 kg pots of grey sands and grown for 100 days. There was no significant difference in plant growth measurements (root/shoot length and shoot count) between seedlings selected for transplant. All three plant species survived and were capable of growth, however, P. labillardieri surpassed T. triandra and J. usitatus.
Figure 1a displays the results of shoot biomass, shoot length and individual shoot count; Figure 1b shows root biomass and root length. Poa labillardieri surpassed the other two species in terms of plant growth measurements; P. labillardieri showed significantly more shoot biomass than J. usitatus (p = 0.011) and significantly more individual shoot counts than T. triandra (p = 0.045) (Fig. 1a). P. labillardieri also produced significantly longer roots (p = 0.029) and root biomass (p = 0.049) than J. usitatus (Fig. 1b). High levels of plant biomass are essential for phytostabilisation. These results are supported by the work of Hayes et al. [42] who conducted a physicochemical and biological investigation of a metal mine at Sunny Corner, New South Wales, Australia. The study investigated Cd, Cu, Pb and Zn contaminants and identified plant species richness. Poa labillardieri and Baeckea utilis (mountain baeckea) were the most abundant plant species and dominated the region closest to the contamination source (<50 m) [42]. As a result, P. labillardieri was identified as a root Pb hyperaccumulator; roots bioaccumulated 974 ± 12 and 975 ± 0 mg/kg of Pb, in soil with Pb concentrations of 19600 ± 7 and 31000 ± 0 mg/kg, respectively [42]. Here the potential of P. labillardieri for use in phytostabilisation has been significantly enhanced through this work, particularly for its ability to grow. To the best of the author’s knowledge, this is the first time that P. labillardieri has been assessed in terms of its efficacy to remediate As-contaminated grey sands; the findings confirm that P. labillardieri has the potential for As phytostabilisation.
3.2.2. Arsenic bioaccumulation
Bioaccumulated elements for Juncus usitatus, Poa labillardieri and Themeda triandra for roots and shoots, post mesocosm experiment, are displayed in Table 2. Arsenic bioaccumulation was primarily in the roots; for example, J. usitatus, P. labillardieri and T. triandra had root As concentrations of 549 ± 210, 399 ± 98 and 356 ± 32.6 mg/kg (mean ± sd; n=3), respectively. There was no significant difference in terms of root As bioaccumulation. Poa labillardieri showed the highest potential for phytostabilisation because it bioaccumulated the most As per plant in the roots (0.098 ± 0.026 mg/plant) and it displayed healthy plant characteristics which are essential for the success of phytostabilisation. It was also visually observed that P. labillardieri grew healthy green shoots with no discolouring or curling of shoots ends which was observed for J. usitatus and T. triandra (Fig. S1-S4). All plant species showed minimal As concentrations (<55.67 ± 9.60 mg/kg) in the shoots and there was no significant difference identified for shoot length.
Previous studies have shown similar results for non-Australian grass species bioaccumulating As primarily in the roots. For example, Lolium perenne (ryegrass) has reported to bioaccumulate approximately 1500 mg/kg and 250 mg/kg of As in the roots and shoots, respectively [31]. Lolium perenne was also grown in mine waste soil collected from the field with elevated As concentrations (5104 mg/kg) and As bioaccumulation was localised in the roots [31]. In Portugal, Agrostis castellana and Agrostis delicatula were assessed for As bioaccumulation from gold mine tailings [79]. Agrostis castellana bioaccumulated 1000 mg/kg in the roots and 170 mg/kg in the shoots [79]. Agrostis delicatula bioaccumulated 1800 mg/kg in the roots and 300 m/kg in the shoots [79]. Lolium perenne, A. castellana and A. delicatula are root hyperaccumulators (>1000 mg/kg) therefore, they are more efficient than plant species investigated in this study. However, L. perenne, A. castellana and A. delicatula shoot As concentrations are elevated (>50 mg/kg) which poses concern for domestic animals and wildlife. For example, it is preferred that shoot concentrations are approximately <50 mg/kg to minimise toxicity entering the food chain i.e. consumption by livestock [80]. In this study, shoot As concentrations are ca. 50 mg/kg, meaning phytostabilisation is possible with low risk to livestock and wildlife.
Of note, J. usitatus, P. labillardieri and T. triandra were also able to bioaccumulate Fe in the roots, 5394 ± 1782, 2091 ± 663 and 3858 ± 681 mg/kg respectively, exceeding normal Fe concentration ranges in plants (30-300 mg/kg d/w), suggesting high bioaccumulation and tolerance for Fe [80]. A previous study, investigating Fe hyperaccumulators, Setaria parviflora and Paspalum urvillei, reported root Fe concentrations at a maximum of ~1750 mg/kg, after 18 days of growth [81]. Other grass species that hyperaccumulate Fe have recorded concentrations of >10 000 mg/kg; for example, Cynodon dactylon has been reported to accumulate 10,712 ± 3844 mg/kg of Fe in the roots after 60 days [82]. Further; as well as Imperata cylindrica reported 23 450 mg/kg in rhizomes and 10 663 mg/kg in the leaves [83]. Root Fe concentrations in this study exceeded Fe concentrations of root hyperaccumulators S. parviflora and P. urvillei, therefore showing potential for root hyperaccumulation. Additionally, all species in this study are preventing bioaccumulation to the shoots, with shoot Fe concentrations below the minimum tolerance for grazing animals (500 mg/kg Fe²⁺) (Table 2) [80]. Iron is an essential micronutrient and effects various plant cellular functions, such as photosynthesis and respiration, and therefore is required in abundance [84]. However, Fe is not readily translocatable, as it is often present in soils as the highly insoluble ferric Fe³⁺ [84,85]. Therefore, plants have evolved to control Fe uptake with reducing and chelating mechanisms [84,85]. In this study, J. usitatus, P. labillardieri and T. triandra are efficiently bioaccumulating Fe which is essential for plant functions and growth [86,87].
Table 2: Bioaccumulated elements for Juncus usitatus, Poa labillardieri and Themeda triandra for roots and shoots, post mesocosm experiment. Elements given in mg per kg of plant dry weight. Values represent mean and ± standard deviation (n=3). Asterisk indicates level of significance between plant species roots and shoots (* = p≤0.05; ** = p≤0.01; *** = p≤0.001).
|
J. usitatus
|
P. labillardieri
|
T. triandra
|
|
Roots
|
Shoots
|
Roots
|
Shoots
|
Roots
|
Shoots
|
As (mg/kg)
|
548.80 ± 210.47
|
21.72 ± 5.35
|
399.10 ± 97.93
|
55.67 ± 9.60 **
|
356.20 ± 32.55
|
21.59 ± 5.87
|
Pb (mg/kg)
|
15.8 ± 6.19
|
<LOD
|
8.68 ± 3.67
|
1.57 ± 0.13
|
10.01 ± 0.97
|
1.40 ± 0.14
|
Sb (mg/kg)
|
0.44 ± 0.16
|
<LOD
|
0.90 ± 0.18
|
<LOD
|
0.38 ± 0.09
|
<LOD
|
Cr (mg/kg)
|
<LOD
|
<LOD
|
0.71 ± 0.49
|
<LOD
|
1.67 ± 0.43
|
<LOD
|
Cd (mg/kg)
|
1.29 ± 0.33
|
1.35 ± 0.76
|
1.72 ± 0.77
|
0.23 ± 0.18
|
0.50 ± 0.12
|
<LOD
|
Cu (mg/kg)
|
115.76 ± 157.87
|
17.51 ± 4.21
|
28.34 ± 5.32
|
9.71 ± 0.97
|
25.30 ± 1.58
|
17.62 ± 0.44
|
Ca (mg/kg)
|
6026 ± 1439.64
|
6448 ± 657.94
|
5689 ± 1589.38
|
9187 ± 2169
|
4402 ± 623.89
|
7605 ± 92.77
|
Zn (mg/kg)
|
133.91 ± 85.06
|
69.70 ± 15.09
|
89.23 ± 15.17
|
37.82 ± 3.23
|
63.44 ± 1.50
|
31.15 ± 5.77
|
Fe (mg/kg)
|
5394 ± 1782.07
|
144 ± 45.52
|
2091 ± 662.90
|
413.2 ± 18.67
|
3858 ± 680.52
|
218.80 ± 60.98
|
Less than limit of detection (<LOD), detection limit for Sb, Cr, Cd, and Pb are <0.012, <0.043, <0.016 and <0.030, respectively.
3.3. Soil-to-plant bioaccumulation
To identify phytostabilising or phytoextracting characteristics, the soil-to-plant bioaccumulation factors were assessed by calculating the bioaccumulation coefficient bioconcentration factor (BCF; soil-to-root), (BAC; soil-to-shoot), and translocation factor (TF; root-to-shoot).
Plants with BCF and BAC values of <1 are considered unsuitable for phytoextraction [31]. Translocation factor values of >1 are indicative of phytoextracting species and values <1 are phytostabilising species [88]. BCF values for J. usitatus, P. labillardieri and T. triandra were 0.69 ± 0.23, 0.50 ± 0.11 and 0.44 ± 0.06, respectively, and BAC/TF values were <0.2, indicating that As concentrations were localised in the roots, which is essential for phytostabilisation. The results confirm that all species are unsuitable for phytoextraction (Table 3) [17,31,70,75]. Previous phytoremediation studies involving grasses have reported similar BCF, BAC and TF values. Lolium perenne grown in mine waste soil (5104 mg/kg As) had BCF, BAC and TF values of 0.7, 0.1 and 0.2, respectively [31]. The higher BCF and <1 BAC/TF values concluded that L. perenne was well suited for phytostabilisation rather than phytoextraction [31]. Additionally, a BCF value above 1 can enhance phytostabilisation. For example, Mensah et al. [31] reported an increase in BCF values (0.7 to >0.9) for L. perenne after the addition of manure, compost and iron oxide soil amendments (Table 3). Lebrun et al. [75] reported similar bioaccumulation values for Salix viminalis grown in mine waste soil (539.06 ± 0.01 mg/kg As) with a BCF value of 0.76 ± 0.11 and BAC/TF values below 0.01. Lebrun et al. [75] was also able to increase the BCF value while keeping the BAC/TF values low (Table 3), hence optimising phytostabilisation. In this study, J. usitatus and P. labillardieri show potential for improved BCF values with the use of soil amendments, as they reported similar BCF values to L. perenne and Salix viminalis (Table 3) [31]. Future studies should assess the use of soil amendments to improve the BCF value for P. labillardieri to enhance phytostabilisation, as it also displayed healthy plant growth measurements mentioned above.
Table 3. Quantified arsenic bioaccumulation using the bioconcentration factor (BCF) for soil-to-root and bioaccumulation coefficient (BAC) for soil-to-shoot. Root-to-shoot arsenic bioaccumulation determined by calculating the translocation factor (TF). Values represent mean and ± standard deviation (n=3) for this study. Additionally, BCF, BAC and TF values from previous studies.
Plant species
|
Soil
|
BCF
|
BAC
|
TF
|
Reference
|
J. usitatus
|
Mine waste
|
0.69 ± 0.23
|
0.02 ± 0.01
|
0.04 ± 0.01
|
This study
|
P. labillardieri
|
Mine waste
|
0.50 ± 0.11
|
0.07 ± 0.01
|
0.14 ± 0.05
|
This study
|
T. triandra
|
Mine waste
|
0.44 ± 0.06
|
0.02 ± 0.01
|
0.06 ± 0.02
|
This study
|
L. perenne
|
Mine waste
|
0.7
|
0.1
|
0.2
|
Mensah et al. [31]
|
L. perenne
|
Mine waste + MC
|
1.2
|
0.1
|
0.1
|
Mensah et al. [31]
|
L. perenne
|
Mine waste + MI
|
0.9
|
0.1
|
0.1
|
Mensah et al. [31]
|
Salix viminalis
|
Mine waste
|
0.76 ± 0.11
|
<0.01
|
<0.01
|
Lebrun et al. [75]
|
Salix viminalis
|
Mine waste + BI
|
0.99 ± 0.09
|
<0.01
|
<0.01
|
Lebrun et al. [75]
|
Salix viminalis
|
Mine waste + C
|
0.82 ±0.09
|
<0.01
|
<0.01
|
Lebrun et al. [75]
|
M = manure; MC = manure + compost; MI = manure + iron; BI = biochar + iron; C = compost.
3.4. Soil characterisation post-treatment
Bulk soil samples were analysed for organic matter, pH, EC, salt, As and possible co-contaminants. Other elements identified included Pb, Sb, Cr, Cd, Cu, Ca, Zn and Fe. Soil characteristics were identified to assess the most suitable candidate for phytostabilisation and for improving soil conditions. Table 4 displays the results of the soil characteristics. Soil amended with J. usitatus, P. labillardieri and T. triandra had As concentrations of 780.80 ± 34.13, 782.50 ± 22.04 and 796.90 ± 44.57 mg/kg, respectively, and there were no statistically significant differences between treatments and control. No other elements exceeded the HIL A, which confirms that co-contamination is not an issue at this site.
Studies by Ollson et al. [5] and Kastury et al. [89] reported similar metal concentrations for grey sands, with the exception of Ca and Fe (1038 - 5089 mg/kg As, n=16; 28.8 - 104 mg/kg Pb, n=16; <LOD - 5.39 mg/kg Sb, n=11; <LOD - 2.41 mg/kg Cd, n=16; 25.7 - 208 mg/kg Cu, n=5; 4.21 - 12.4 g/kg Ca, n=11; 100 - 279 mg/kg Zn, n=5; 23.5 - 33.9 mg/kg Fe, n=16). The results confirm that these metal concentrations are a typical soil elemental composition for grey sands; for example, with elevated As concentrations (800-5089 mg/kg) and low Pb and Sb concentrations of 16-104 and <LOD-5.39 mg/kg, respectively.
Soil characteristics, organic matter, pH, EC salt, and moisture content remained relatively consistent between pre- and post-treatment. The control had a significant increase in soil pH (8.21 ± 0.10) compared to all treatments and significantly decrease in EC compared to P. labillardieri. Salt ranged between 1-2% for all treatments and P. labillardieri had a significantly higher moisture content compared to the control.
Table 4: Soil characteristics post-treatment for organic matter (OM), pH, electrical conductivity (EC), salt, elements, and moisture content (MC). Elements given in mg per kg of soil dry weight. Values represent mean and ± standard deviation (n=3). The national soil guidance values (health investigation levels A (HIL A)) are provided. Asterisk indicates level of significance (* = p≤0.05; ** = p≤0.01; *** = p≤0.001).
|
Soil amended with
J. usitatus (100 d)
|
Soil amended with
P. labillardieri (100 d)
|
Soil amended with
T. triandra (100 d)
|
Unamended soil control (100 d)
|
HIL A (mg/kg)
|
OM (%)
|
1.54 ± 0.13
|
1.46 ± 0.04
|
1.38 ± 0.02
|
1.38 ± 0.06
|
|
pH
|
7.18 ± 0.07***
|
7.89 ± 0.18*
|
7.55 ± 0.03***
|
8.21 ± 0.10
|
|
EC (ms/cm)
|
3.09 ± 0.34
|
3.32 ± 0.89*
|
2.58 ± 0.41
|
1.83 ± 0.27
|
|
Salt (%)
|
1.66 ± 0.57
|
1.66 ± 0.57
|
1 ± 0
|
1 ± 0
|
|
As (mg/kg)
|
780.80 ± 34.13
|
782.50 ± 22.04
|
796.90 ± 44.57
|
743.90 ± 14.22
|
100
|
Pb (mg/kg)
|
24.29 ± 1.24
|
24.90 ± 3.85
|
24.37 ± 1.60
|
16.41 ± 5.92
|
300
|
Sb (mg/kg)
|
<LOD
|
<LOD
|
<LOD
|
<LOD
|
NA
|
Cr (mg/kg)
|
<LOD
|
<LOD
|
<LOD
|
<LOD
|
100
|
Cd (mg/kg)
|
<LOD
|
<LOD
|
<LOD
|
<LOD
|
20
|
Cu (mg/kg)
|
22.61 ± 5.68
|
11.76 ± 2.71
|
39.87 ± 52.69
|
14.55 ± 4.67
|
6000
|
Ca (mg/kg)
|
9676 ± 878.5
|
9645 ± 204.95
|
9960 ± 312.94
|
7475 ± 2618.53
|
|
Zn (mg/kg)
|
98.70 ± 10.41
|
98.43 ± 9.31
|
102.71 ± 5.14
|
84.42 ± 42.76
|
7400
|
Fe (mg/kg)
|
18247 ± 1295.64
|
17974 ± 129.17
|
18012 ± 725.45
|
13279 ± 5106.68
|
NA
|
MC (%)
|
11.91 ± 0.65
|
14.36 ± 2.55*
|
12.85 ± 0.93
|
10.03 ± 1.04
|
|
Less than limit of detection (<LOD), detection limit for Sb, Cr, and Cd are <0.012, <0.043, and <0.016, respectively. Not available (NA).
3.5. Soil microbial analysis - Identification of bacterial 16s rRNA
3.5.1. Bacterial community diversity
A diverse soil microbiome is an indicator of soil fertility and plant health [90–93]. Soil microbes supply plants with nutrients by enhancing nutrient availability and uptake, protect against pathogens and aboveground threats such as insects [91–93]. Assessment of the impact of phytoremediation on the diversity of the microbial community is important in terms of assessing richness, diversity and evenness pre- and post-phytoremediation [94].
Results showed low richness and diversity numbers in the treatments, ranging from 1500-3000 OTUs and a Shannon index of 6.45-6.95. In comparison, healthy soil is reported to contain between 5291-6611 OTUs, and a Shannon index between 7.38-7.96 [95]. The results in this study are consistent with previously reported levels in gold mine tailings and As contaminated soil [90,96].
Kruskal-Wallis analysis showed a significant difference in richness (p = 0.014) and evenness (p = 0.031). The Dunn test was performed to identify the statistically significant difference between pre- and post-treatments; soil amended with T. triandra showed a significant decrease in bacterial richness (p = 0.025). There was no significant difference identified after performing the Dunn test for evenness with the p adjusted method Benjamini-Hochberg.
However, the observed trend of richness and evenness showed that P. labillardieri and T. triandra rhizosphere soil decreased in richness and increased in evenness (Fig. 2). This trend could be explained by a plant’s ability to stabilise a soil microbial community during the establishment of the rhizosphere microbiota [92,97,98]. In this study, the rhizosphere microbiota appeared to have stabilised after colonising P. labillardieri and T. triandra roots, as indicated by the decline in richness and increase in evenness. This is also supported by the PCA plot where P. labillardieri and T. triandra have diverted from the control (Fig. S5). This is promising results because P. labillardieri and T. triandra have stabilised the rhizosphere soil microbial community under toxic conditions of the mine waste soil.
3.5.2. The relative abundance of Phyla
The health and function of a soil microbiome may be determined by the presence of certain microorganisms. For example, the microbial community influences soil As speciation (oxidation states) which in turn impacts a plant’s ability to bioaccumulate As [32,33]. Plants bioaccumulate arsenate, whilst arsenate-respiring microbes can transform arsenate into arsenite, which is unfavourable for plant uptake and also increases the risk of As leaching [32,34]. To understand the composition and function of the rhizosphere in soils amended with plants the relative abundance of the dominant phyla was assessed [90]; further Analysis of Compositions of Microbiomes with Bias Correction 2 (ANCOM-BC2), was used to identify As tolerant microbes in the rhizosphere of J. usitatus, P. labillardieri, and T. triandra.
Collectively, the microbiome of all samples was dominated by five phyla: Actinobacteriota, Proteobacteria, Chloroflexi, Acidobacteriota, and Myxococcota, respectively (Fig. S6). This is consistent with previous studies, which have shown Actinobacteriota, Proteobacteria, and Chloroflex to be the dominant phyla in mine waste [90,99–104]. According to previous studies, Proteobacteria and Actinobacteriota are likely to be higher in the rhizosphere, compared to the bulk soil [97,104]. The rhizosphere is enriched in organic compounds released from plant roots, which stimulates microbial growth and promotes the presence of organisms commonly found in carbon-rich environments (copiotroph) such as Proteobacteria. However, this was not observed in this study. Both Proteobacteria and Actinobacteriota harbour multiple As transforming and detoxifying bacterial strains, and have been identified for their heavy metal resistance and removal [105–108]. For example, Proteobacteria genera Pseudomonas, Hydrogenophaga, Thiobacillus and Acidovorax species have arsenite-oxidising bacteria, transforming arsenite to the less toxic arsenate [102,105,106,109,110]. These genera have been detected amongst the taxa for all samples pre- and post-treatment. Six Pseudomonas species were detected with four identified to species level, P. gradensis, P. taeanensis, P. cuatrocienegasensis, and P. peli. Also detected were three Hydrogenophaga species, with one identified to species level, H. defluvii, one Thiobacillus species, T. thiophilus and one unidentified Acidovorax species. Actinobacteriota genera Bifidobacterium angulatum may remove As from the environment [108], whilst Streptomyces spp. have been shown to detoxify As via methylation [111], suggesting an eco-friendly mitigation strategy for remediating contaminated soil. Bifidobacterium angulatum was not detected in this study; however, three Streptomyces species were detected, with one to species level, S. smyreus.
Metal resistance is supported by this study, as shown by the ANCOM-BC2 waterfall plot for microbial phyla (Fig. 3), with As appearing to have a minimum effect on Proteobacteria and Actinobacteriota. The ANCOM-BC2 waterfall plot for microbial genera (Fig. S7), did not identify a positive or negative LFC for the Proteobacteria and Actinobacteriota genera discussed above, also confirming that Proteobacteria and Actinobacteriota genera are not greatly influenced by As. The presence of eleven Proteobacteria species is encouraging because they have been correlated with available As and reported to have potential for bioremediation, as mentioned above [100,112]. Additionally, Deng et al. [100] reported that Thiobacillus species (β-Proteobacteria) had high relative abundance in soils enriched with metal(loid)s, specifically As, Sb, Pb, Cd, and mercury (Hg). Moreover, Pseudomonas species can reduce various metal(loid)s simultaneously and are metabolically diverse, such as utilising electron acceptors of As⁵⁺ and Sb⁵⁺ [100].
The ANCOM-BC2 analysis also revealed a positive LFC in the phyla Nanoarchaeota and Latescibacterota as one unit (1 mg/kg) of As increased (Fig. 3). Previous studies have identified Nanoarchaeota in water impacted by mine drainage and As-rich sediments and soils [113,114]. However, research investigating archaea and soil arsenic contamination is limited. Su et al. [115] identified Latescibacterota as a keystone phyla in mining impacted soils after reporting positive correlations with As and Sb contamination [115], stating that Latescibacterota can thrive under As-contaminated conditions, which is also supported by this study after a positive LFC was observed in the ANCOM-BC2 analysis. Wang et al. [116] also identified Latescibacterota in mine tailings and discussed the importance of Latescibacterota as a rare species that has a crucial role in nutrient cycling. The addition of soil amendments to the mine tailings enhanced Latescibacterota proliferation and function [116]. Thus, Wang et al. [116] concluded their importance in remediation of mine tailings and potential to influence ecosystem restoration. The fact that Latescibacterota was observed in this study is promising, and suggests that the rhizosphere contains keystone species known to be associated with As-contaminated mine tailings. Future studies should investigate the role of soil amendments to enhance the presence of Latescibacterota in As-rich mine waste soils.
The results did not identify any plant species-specific microbiomes, as indicated by the bacterial relative abundances, with all treatments exhibiting phyla with similar relative abundance (Fig. S6). However, previous studies with larger sample sizes reported rhizosphere microbial compositional shifts. For example, Ling et al. [97] compared bulk soil and rhizosphere soil microbial communities and reported a decline in rhizosphere soil microbial diversity. This decline in diversity was supported by an enrichment in Proteobacteria and Actinobacteriota and a reduction in Chloroflexi, Acidobacteria, and Nitrospirae [97], thus, reiterating the importance of Proteobacteria and Actinobacteriota in the rhizosphere and its influence on As.
Overall, this study identified the dominant phyla found in Bendigo mine waste soil, which were similar to those reported at waste sites on other continents [97]. Importantly, two rhizosphere associated phyla which have previously shown potential for phytoremediation, were the two most dominant phyla in the tested soil, Proteobacteria and Actinobacteriota. Additionally, the positive LFC of the keystone phylum Latescibacterota and the presence of As associated genera was observed in rhizosphere soil. This suggests that the Bendigo mine waste soil potentially harbours the required microorganisms to assist in the selected plant species phytoremediation.