Heavy Metal Pollution Assessment and Survey of Rhizosphere Bacterial Communities from Saccharum spontaneum L. in a Rehabilitated Nickel-Laterite Mine in the Philippines

Abstract

In this study, we assessed soil pollutants and surveyed the bacterial communities using 16S rRNA sequencing to better understand how to improve rehabilitation strategies for nickel-laterite mines in the Philippines. Representative soil samples and rhizospheres from Saccharum spontaneum L. in three post-mining sites rehabilitated in 2015, 2017, and 2019 were collected and analyzed. X-ray diffraction (XRD) identified iron oxyhydroxides, silicates, and clays as major soil components. Based on the pollution load index and contamination degree, the 2015A and 2015B sites were classified as “pristine” and had a “low degree of pollution”, while the remaining sites (2017A, 2017B, 2019A, and 2019B) were considered “moderately contaminated” with nickel, chromium, cobalt, lead, zinc, and copper. An analysis of the bacterial community composition revealed that the phyla Proteobacteria and Actinobacteria, along with the genus Ralstonia, were the most abundant groups across both control and rehabilitated sites. Our results showed that the soil pH and organic matter contents were strongly linked to specific bacterial community composition. These taxa have potential for inoculation in nickel-laterite soils to promote the growth of hyperaccumulator plants. Our results also showed a significant correlation between the structure of the bacterial communities and nickel, chromium, and manganese soil contents, but not with rehabilitation time. Furthermore, we identified the genera Diaphorobacter as potential bioindicators because they are sensitive to nickel and chromium. This study provides valuable baseline data on heavy metal pollution and microbial diversity in a rehabilitated Ni-laterite mine site.

1. Introduction

The Philippines was ranked as the world’s second-largest exporter of nickel (Ni)-laterite ores in 2024, shipping 400,000 tonnes (t) of Ni [1]. It also holds an estimated 4.8 million tonnes (Mt) of Ni reserves, the sixth largest globally [1]. As of 2024, the Philippines had at least 20 operating Ni-laterite mines, two hydrometallurgical plants that process low-grade Ni-laterite ores via high-pressure acid leaching (HPAL), and several potential deposits under various stages of mine development [2,3].

Globally, Ni is primarily found in two types of ore deposits: laterites (comprising ~54% of reserves) and magmatic sulfides (~35% of global reserves) [1]. Ni-laterite ores form near the surface through the prolonged weathering of ultramafic rocks, while magmatic Ni-sulfide deposits result from volcanic or hydrothermal activities [4,5]. Although Ni-laterite ores dominate in terms of reserves, historically, Ni production came from the smelting of sulfide ores due to the technical challenges posed by laterite processing, particularly the need to extract Ni bound in iron oxides [6,7,8,9]. However, with the growing demand, the mining and processing of Ni-laterite ores are becoming more prevalent, raising environmental and social concerns for surrounding communities and ecosystems.

The mining and processing of Ni-laterite ores can release hazardous trace metals that accumulate in soils, surface waters, and groundwater [10]. Hazardous heavy metals like Ni, copper (Cu), manganese (Mn), zinc (Zn), lead (Pb), chromium (Cr), cobalt (Co), and copper (Cd)—commonly found in Ni-laterite-impacted areas—are persistent and non-degradable, posing risks to water quality and food safety through bioaccumulation in the food chain [11,12,13]. For example, Heikkinen et al. [14] reported significant groundwater and surface water contamination near the tailings storage facility (TSF) of the Hitura mining area in Sweden. The authors documented the generation of neutral mine drainage (pH of ~7.2) containing high concentrations of sulfate (SO₄²⁻ of ~7400 mg/L), chloride (Cl⁻ of ~970 mg/L), Zn (~760 mg/L), and Ni (~970 mg/L), which have been impacting local groundwater systems since the 1970s. Similarly, Opiso et al. [11] highlighted the problem of the siltation of streams, rivers, and estuaries, which is detrimental to farming and fishing communities living close to Ni-laterite mining areas in the Philippines. In New Caledonia, Germande and co-workers [15] assessed heavy metal bioaccumulation in marbled eels (Anguilla marmorata) and found that eels living close to the mining area had a higher average Ni concentration (5.14 mg/kg) in their livers than those captured away from the mines (1.63 mg/kg). The authors also showed that Ni, introduced into the river via fine particle deposition and leaching, preferentially accumulated in the eels’ livers and kidneys, leading to dysregulation of numerous genes involved in oxidative stress, DNA repair, apoptosis, reproduction, and both lipid and mitochondrial metabolism [15].

These previous reports underscore the critical need for effective management of Ni-laterite mining wastes, such as silts and overburden, and for the rehabilitation of mined-out areas to prevent heavy metal leaching, bioaccumulation, and siltation. In the Philippines, mining companies are mandated by law to rehabilitate mining sites post-closure [16,17]. However, rehabilitation remains challenging due to the inherently poor quality of Ni-laterite soils. These soils are typically deficient in essential nutrients like carbon (C) and nitrogen (N) and are often enriched with hazardous heavy metals like Ni and iron (Fe), which, in excess, hinder plant growth [18]. Herrera et al. [19], for example, monitored the rehabilitation program of a Ni-laterite mining area in New Caledonia and found it only partially effective. Although endemic plants like Gymnostoma webbianum and Serianthes calycina grew in the rehabilitated area, their growth was slow, limiting the natural development of secondary vegetation.

A promising strategy to enhance the rehabilitation of Ni-laterite mined-out areas is the use of suitable hyperaccumulator plants in combination with beneficial microorganisms found in the soil, rhizosphere, and plant roots. Among these microorganisms, bacteria are particularly important because of their high abundance (106–109 viable cells per cm-3 of soil) and large surface-to-volume ratio (i.e., small size), which enhances their interactions with the environment [20]. Moreover, bacteria are predisposed to accumulate dissolved heavy metals from the environment due to the negative net charge of their cell membranes [21], which could improve plant growth by reducing heavy metal bioavailability via adsorption and coprecipitation reactions [22,23,24]. He et al. [25], for example, reported the high Cr removal ability of Pseudochrobactrum saccharolyticum strain W1, a bacterium capable of thriving even at high concentrations of the highly toxic hexavalent Cr (Cr^VI). Similarly, Pseudomonas sp., like Pseudomonas stutzeri, have been shown to efficiently accumulate and remove heavy metal ions like Cu²⁺, Cd²⁺, Co²⁺, Pb²⁺, and Ni²⁺ from solutions via carbonate precipitation [26,27].

Plant growth-promoting rhizobacteria (PGPR) colonize plant roots and support plant development by improving nutrient uptake and enhancing stress tolerance [28]. Documented modes of action of PGPRs include promoting nutrient bioavailability [29,30], producing phytohormones or modulating ethylene synthesis to stimulate root system development [31,32,33], inducing systematic resistance [34], and protecting plants against phytoparasites [35,36]. For example, Enterobacter cloacae CAL2 and UW4 [37], Pseudomonas putida strain WCS358 [38], and Serratia plymuthica strain IC1270 [39] have been shown to improve the uptake of Fe, an essential element for plant growth, by producing low-molecular-weight compounds called siderophores to sequester ferric ion (Fe³⁺) competitively [40]. PGPRs also carry genes that facilitate plant-beneficial properties like nitrogen fixation [41], phloroglucinol synthesis [42], and pyrroloquinoline quinone synthesis [43]. Although numerous studies have investigated PGPRs in various environments like agricultural lands and industrially contaminated soils, many of these works focused on identifying the genera of PGPRs and how they improved plant growth by analyzing their genes [44]. To date, very few works have focused on assessing how the rhizosphere microbiome community influences the growth of hyperaccumulator plants used for rehabilitation, especially in Ni-laterite mined-out areas.

In this study, the effects of Ni-laterite mine site rehabilitation using Saccharum spontaneum, a hyperaccumulator plant belonging to the grass family, on the structure and diversity of the rhizosphere microbiome were investigated. Specifically, the objectives of this study are to (i) identify hazardous contaminants in the rehabilitated site, (ii) analyze the physico-chemical characteristics of control and rehabilitated soils, (iii) characterize the bacterial communities in the rhizosphere of S. spontaneum, and (iv) determine the composition, relative abundance, and diversity of bacterial communities from the rhizosphere of S. spontaneum in both control and rehabilitated sites. The first and second objectives were achieved by detailed characterization of the soils using techniques like X-ray fluorescence spectroscopy (XRF), X-ray powder diffraction (XRD), and determinations of the organic matter (OM), organic carbon (OC), total nitrogen (TN), available phosphorus (AP), and exchangeable potassium (EK) of the soil. Meanwhile, the third and fourth objectives were accomplished using metagenomics techniques. The outcomes of this work are crucial for improving existing rehabilitation strategies and practices for Ni-laterite mined-out areas and offer insights into how native PGPRs can be beneficial as microbial inoculants to enhance vegetation recovery and improve the agricultural potential of degraded soils.

2. Materials and Methods

2.1. Site Description and Soil Collection

Soil samples were collected from three post-mining rehabilitated sites (2015, 2017, and 2019) of a Ni-laterite mining area in Tubay, Agusan del Norte, Philippines, in August 2021 (Figure 1). Seven soil samples were evaluated: (i) one sample of undisturbed soil adjacent to the mining area (control, 90°17′41″ N 125°30′57″ E), (ii) two samples from areas rehabilitated in 2015 at low (2015A, (90°16′44″ N 125°30′28″ E) and high elevations (2015B; 9°17′38″ N 125°31′02″ E), (iii) two samples from areas rehabilitated in 2017 at low (2017A; 9°17′21″ N 125°30′47″ E) and high elevations (2017B; 9°17′35″ N 125°31′13″ E), and (iii) two samples from areas rehabilitated in 2019 at low (9°16′50″ N 125°30′42″ E) and high elevations (2019B, 9°17′34″ N 125°31′14″ E).

All soil samples were collected from the A horizon (depth of 0–10 cm) at different locations randomly established around the site in a 100 m² plot using sterile, stainless-steel shovels. Samples were collected across each sampling point and mixed to create composite samples [45]. These samples were brought back to the laboratory, air-dried, lightly crushed, sieved through a 2 mm aperture stainless steel screen, placed in labeled polypropylene containers, and stored at −40 °C and 4 °C before processing for metagenomics and physico-chemical characterizations, respectively.

2.2. Analyses of Soil Chemical, Mineralogical, and Physicochemical Properties

To determine the chemical and mineralogical compositions of the soil samples, they were analyzed by XRF (NEXCG, Rigaku Corporation, Tokyo, Japan) and XRD (MultiFlex, Rigaku Corporation, Tokyo, Japan), respectively. Before both analyses, the samples were prepared as follows: (i) the samples were air-dried and screened through a 2 mm aperture sieve to remove plant debris and roots, (ii) the screened samples (<2 mm) were ground using an agate mortar and a pestle to less than 50 µm, and (iii) ground samples were pressed into sample holders for the analyses. XRD peaks were identified using the Match!^® software (Version 3.3.0 Build 88) (Crystal Impact, Bonn, Germany).

Soil pH was measured in situ with a calibrated pH meter (Hanna Instruments, Smithfield, RI, USA), while the moisture content was obtained gravimetrically by drying to a constant weight at 105 °C for 24 h [46]. For OM, OC, TN, AP, and EK, soil samples were analyzed at the Soil and Plant Analysis Laboratory of Central Mindanao University (Musuan, Bukidnon, Philippines). In brief, OM was determined by the Heanes method, while OC was quantified using the Walkley–Black wet oxidation method [46]. AP and TN of the soil were extracted using 0.03 M NH₄F + 0.1 M HCl (1:7) and then determined by an ultraviolet/visible light (UV-VIS) spectrophotometer (Hitachi High-Tech, Tokyo, Japan). Finally, the EK was measured by flame atomic absorption spectroscopy (AAS) (Analytik Jena GmBH+Co. KG, Jena, Germany) using a 1N NH₄OAc as an extractant and with the solution adjusted to pH 7.

2.3. Pollution Indicess to Classify the Extent of Soil Pollution

Four pollution indices—geo-accumulation index (I_geo), contamination factor (CF), pollution load index (PLI), and contamination degree (CD)—were calculated to evaluate the extent of pollution of the soils with Ni, Cr, Co, Pb, Zn, and Cu. The I_geo was calculated using the equation below:

$${\mathrm{I}}_{\mathrm{g}\mathrm{e}\mathrm{o}}={\mathrm{l}\mathrm{o}\mathrm{g}}_2{\displaystyle \frac{{\mathrm{C}}_{\mathrm{n}}}{1.5\ast {\mathrm{B}}_{\mathrm{n}}}}$$

(1)

where C_n is the soil content of contaminant n, B_n is the background content of contaminant n, and 1.5 is a constant introduced to normalize the variances in background levels caused by site-specific geological variations [47]. Muller (1969) [48] devised seven categories for the extent of pollution based on the values of I_geo: (1) unpolluted (I_geo < 0), (2) unpolluted to moderately polluted (0 ≤ I_geo < 1), (3) moderately polluted (1 ≤ I_geo < 2), (4) moderately to heavily polluted (2 ≤ I_geo < 3), (5) heavily polluted (3 ≤ I_geo < 4), (6) heavily to extremely polluted (4 ≤ I_geo < 5), and (7) extremely polluted (I_geo ≥ 5).

The CF of a single target contaminant was determined using the following equation:

$$\mathrm{C}\mathrm{F}={\displaystyle \frac{{\mathrm{C}}_{\mathrm{n}}}{{\mathrm{B}}_{\mathrm{n}}}}$$

(2)

Hakanson (1980) proposed four categories to interpret the calculated values of CF: (1) low contamination (CF < 1), (2) moderate contamination (1 ≤ CF < 3), (3) considerable contamination (3 ≤ CF < 6), and (4) very high contamination (CF ≥ 6). When there are multiple contaminants to be considered, CF can be used to measure CD by simple summation of the CFs of each contaminant. Hakanson (1980) [49] also devised four levels of pollution depending on the values of CD: (1) low degree of pollution (CD < 8), (2) moderate degree of pollution (8 ≤ CD < 16), (3) considerable degree of pollution (16 ≤ CD < 32), and (4) very high degree of pollution (CD ≥ 6).

A pollution index based on the values of CF, called PLI, was introduced by Tomlinson et al. [50], which is calculated using the following equation:

$$\mathrm{P}\mathrm{L}\mathrm{I}=\sqrt[\mathrm{n}]{{\mathrm{C}\mathrm{F}}_1\ast {\mathrm{C}\mathrm{F}}_2\ast \dots {\mathrm{C}\mathrm{F}}_{\mathrm{n}}}$$

(3)

The values of PLI can also be categorized to determine the overall degree of pollution of target areas: (1) pristine (PLI < 1), (2) baseline (PLI = 1), (3) polluted (PLI > 1).

2.4. Identification of Plant Sample

All plant samples were collected in accordance with the guidelines set by the mining company. According to their survey, several plant species identified as hyperaccumulators, including Saccharum spontaneum, were planted (Supplementary Figure S1). The plant specimen was initially identified by a Forester from the company and subsequently verified by a Botanist from MSU-Iligan Institute of Technology, Iligan City, Philippines. Identification of plant samples was based on the published books and journals as references, online identification resources like http://identify.plantnet.org, inaturalist, and Co’s Digital Flora of the Philippines (accessed on 10 March 2024).

2.5. DNA Preparation and Analyses

A cleaning solution was also prepared using a combination of sodium hypochlorite at 10% (vol/vol), sodium dodecyl sulfate at 1% (wt/vol);], sodium hydroxide (NaOH) at 1% (wt/vol);], and sodium bicarbonate (NaHCO₃) at 1% (wt/vol) to sanitize the work surfaces [51]. Sterile conditions were maintained by wearing gloves, cleaning and autoclaving glassware, keeping samples separate, and using sterilized instruments.

Total genomic DNA extraction was performed from 1.0 g of soil samples using the HiPurA Soil DNA Purification Kit (HiMedia Laboratories Private Limited, Thane, India) following the manufacturer’s protocol. The DNA samples were sent to Macrogen, Korea for amplicon sequencing analysis.

The sequencing libraries were prepared according to the Illumina 16S Metagenomic Sequencing Library protocols to amplify the V3 and V4 regions. The input gDNA 5 ng/10 ng was PCR amplified with 5x reaction buffer, 1 mM of dNTP mix, 500 nM of the universal F/R PCR primer, and Herculase II fusion DNA polymerase (Agilent Technologies, Santa Clara, CA, USA). The cycle condition for 1st PCR was 3 min at 95 °C for heat activation, and 25 cycles of 30 s at 95 °C, 30 s at 55 °C, and 30 s at 72 °C, followed by a 5 min final extension at 72 °C. The universal primer pair with Illumina adapter overhang sequences used for the first amplifications was as follows: 16S Amplicon PCR (Forward Primer 5′TCGTCGGCAGCGTCAGATGTGTATAAGAGACAGCCTACGGGNGGCWGCAG and 16S Amplicon PCR Reverse Primer 5′ GTCTCGTGGGCTCGGAGATGTGTATAAGAGACAGGACTACHVGGGTATCTAATCC). The 1st PCR product was purified with AMPure beads (Agencourt Bioscience, Beverly, MA, USA). Following purification, the 10 µL of 1st PCR product was PCR-amplified for final library construction containing the index using NexteraXT Indexed Primer. The cycle condition for 2nd PCR was same as the 1st PCR condition except it used 10 cycles. The PCR product was purified with AMPure beads. The final purified product was then quantified using qPCR according to the qPCR Quantification Protocol Guide (KAPA Library Quantification kits for Illumina Sequencing platforms) and qualified using the TapeStation D1000 ScreenTape (Agilent Technologies, Santa Clara, CA, USA). It was then sequenced using the MiSeqTM platform (Illumina, San Diego, CA, USA).

FLASH software (v1.2.11) evaluated the offline reads for pair-end contigs. The barcode and chimeric sequences were removed from pair-end reads using QIIME software (v1.9.1). The Amplicon Sequence Variants (ASVs) were clustered using UPARSE (v7.0), and the most abundant sequence of each ASV was selected as a representative sequence. Mothur software (v1.30.2) was used for alpha (α) diversity indexes including ACE, Chao1, Shannon, and InvSimpson. Briefly, the Ace and Chao1 indexes were used for the estimation of the bacterial community richness, and the Shannon and Simpson indexes were presented for bacterial community diversity.

2.6. Statistical Analyses

All calculations were performed using the R software v3.6.0, including the alpha diversity species richness (Chao) and evenness (ACE) and Shannon and Simpson indices. Meanwhile, the rarefaction curve, community richness containing Chao1 and Ace indices, and community diversity containing Simpson and Shannon indexes were analyzed by the QIIME platform (v.1.30.1). Spearman rank correlation analysis (p < 0.05) was carried out to correlate soil physico-chemical parameters and heavy metals with bacterial relative abundance. All statistical analyses were performed in the R environment (R software v3.6.0) unless otherwise stated. SILVA was used as a reference database.

3. Results and Discussion

3.1. Comparison of Chemical and Mineralogical Compositions of Soils from the Control and Rehabilitated Areas

The chemical compositions of the soil samples from the control and rehabilitated areas are summarized in

Table 1. Chemical compositions of soils from the control and three rehabilitated sites (2015, 2017, 2019).

Site	SiO2	Al2O3	Fe2O3	MnO	MgO	CaO	TiO2	P2O5	Ni	Cr	Co	Pb	Zn	Cu
	(wt%)	(wt%)	(wt%)	(wt%)	(wt%)	(wt%)	(wt%)	(wt%)	(wt%)	(wt%)	(mg/kg)	(mg/kg)	(mg/kg)	(mg/kg)
Control	33.2	3.64	41.7	0.66	18.0	0.22	0.07	0.04	0.83	0.66	2340	15.8	209	120
2015A	63.5	11.1	1.27	−	21.5	1.20	0.36	0.11	0.26	0.32	924	13.0	72.3	80
2015B	70.5	18.6	0.15	0.20	7.07	0.49	0.41	0.39	0.19	0.16	711	13.9	193	240
2017A	6.50	6.48	79.4	0.86	2.10	0.11	0.2	0.07	1.44	1.16	3980	56.6	356	−
2017B	10.1	5.64	74.7	0.99	4.16	0.10	0.11	0.05	1.37	1.00	3770	51.1	353	−
2019A	13.3	4.59	71.5	1.15	4.44	0.36	0.11	0.12	1.29	1.25	3700	−	410	158
2019B	25.5	4.17	53.5	0.87	12.03	0.13	0.08	0.05	1.54	0.75	3130	21.4	273	122

Note: “−” means not detected.

. The soil from the control was considered the reference profile in this study because it originated from an undisturbed site. Soils from the control are rich in Fe (41.7 wt% as Fe₂O₃), silicon (Si) (33.2 wt% as SiO₂), and magnesium (Mg) (18 wt% as MgO), while those from areas rehabilitated in 2015 contain more Si (63%–71% as SiO₂) and aluminum (Al) (11%–19% as Al₂O₃). In comparison, soils in areas rehabilitated in 2017 and 2019 were dominated by Fe (53%–80% as Fe₂O₃), Si (33.2 wt% as SiO₂), and Al (4–7 wt% as Al₂O₃). Among the rehabilitated areas, the 2019B soil samples had the closest chemical properties to those of the control site.

Goethite (FeOOH) and magnetite (Fe₃O₄) were the major minerals identified by XRD in the control site soils, which is consistent with their high Fe contents, while the significant Mg and Si contents of this sample could be attributed to the presence of lizardite (Mg₃Si₂O₅(OH)₄) and quartz (SiO₂) in moderate to minor amounts (Figure 2). These identified minerals confirm the previous works of Fan and Gerson [52], Aquino et al. [53], and Tupaz et al. [54] on Philippine Ni-laterite ores from Davao Oriental, Zambales, and Mindoro, respectively. Soils from areas of the mine rehabilitated in 2017 and 2019 were also rich in goethite and magnetite but compared with the control, lizardite and quartz were less abundant in them. In contrast, soils from areas of the mine rehabilitated in 2015 had low contents of goethite and magnetite. Instead, they were rich in silicate minerals (quartz, anorthite, and lizardite) and clays (montmorillonite) (Figure 2). These results suggest that in 2015, the soils used to rehabilitate selected areas of the mine were sourced from the deeper, saprolite portion of the soil profile, which is rich in silicate minerals, while those used in the rehabilitation campaigns of 2017 and 2019 likely originated from the limonitic overburden and siltation ponds, which are rich in iron oxyhydroxide minerals [52].

In terms of hazardous heavy metals, the concentrations of Mn, Cr, Pb, and Zn were within the global concentration range for unpolluted soils (MnO = 0.0015–0.77%; Cr₂O₃ = up to ~6%; Pb = 10–40 mg/kg; Zn = 10–300 mg/kg); however, those of Ni, Co, and Cu were 18, 58, and 2 times higher than the global average concentration ranges (Ni = 0.00002–0.045%; Co = up to ~40 mg/kg; Cu = 2–50 mg/kg) [55,56,57,58,59,60,61,62]. These elevated concentrations of Ni, Co, and Cu in the control could be attributed to the natural accumulation of these heavy metals in the iron-rich components of lateritic soils induced by the tropical weathering of heavy metal-bearing ultramafic rocks [53].

Using the soil composition from the control site as background, four pollution indices—I_geo, CF, PLI, and CD—were calculated to gain insights into the degree of heavy metal contamination of the soils used for rehabilitation (

Table 2. Calculated pollution indices of soils from the rehabilitated sites.

Site	Igeo						CF						PLI	CD
	Ni	Cr	Co	Pb	Zn	Cu	Ni	Cr	Co	Pb	Zn	Cu
2015A	−2.27	−1.63	−1.93	−0.87	−2.12	−1.07	0.31	0.48	0.39	0.82	0.35	0.71	0.48	3.07
2015B	−2.73	−2.60	−2.31	−0.77	−0.70	0.23	0.23	0.25	0.30	0.88	0.92	1.75	0.54	4.34
2017A	0.20	0.22	0.18	1.26	0.30	− *	1.73	1.74	1.70	3.59	1.85	− *	1.80	10.6
2017B	0.13	0.00	0.10	1.11	0.17	− *	1.64	1.51	1.61	3.24	1.69	− *	1.67	9.68
2019A	0.04	0.32	0.07	− *	0.39	−0.15	1.55	1.88	1.58	−	1.96	1.36	1.52	8.32
2019B	0.30	−0.40	−0.17	−0.15	−0.20	−0.52	1.85	1.13	1.33	1.35	1.31	1.05	1.32	8.02

* Note: I_geo = geo-accumulation index, CF = contamination factor; PLI = pollution load index; CD = contamination degree; “−” means not measured.

). Overall, the soils used to rehabilitate the mine in 2015 were “pristine” and had a “low degree of pollution” based on the PLI (0.47–0.54) and CD (3.1–4.3) results. In comparison, the soils in 2017A, 2017B, 2019A, and 2019B were “polluted” (PLI = 1.3–1.8) and had a “moderate degree of pollution” (CD = 8–10.6). In terms of the individual heavy metals, the I_geo and CF results showed that the soils used in 2015 were “uncontaminated” (I_geo = −2.7 to −0.7) and had “low contamination” (CF = 0.22–0.92) for Ni, Cr, Co, Pb, and Zn, but those calculated for Cu in the soil used to rehabilitate 2015B were two-fold higher than the background and were classified as “moderately contaminated” based on I_geo (0.22) and CF (1.8). In comparison, the soils used to rehabilitate the mine areas in 2017 were “moderately contaminated” with Ni, Cr, Co, Pb, and Zn, except for Cu. For 2019, the soils used for rehabilitation at both lower and higher elevations were also “moderately contaminated” with Ni, Cr, Co, Zn, and Cu, except for Pb (Table 2).

3.2. Physicochemical Properties of Soils from the Control and Rehabilitated Areas

The physico-chemical parameters of the soil samples are summarized in

Table 3. Physico-chemical properties of soils from the control and three rehabilitated sites (2015, 2017, and 2019) at both lower and higher elevations. Note that pH and moisture content were measured on-site (n = 7).

Sites	pH	Moisture Content (%)	Organic Matter (%)	Organic Carbon (%)	Available Phosphorus (ppm)	Exchangeable Potassium (ppm)	Total Nitrogen (%)
Control	6.19	7.08	1.50	0.87	1.17	225	0.15
2015A	6.88	6.10	0.88	0.51	5.11	37.5	0.19
2015B	6.50	8.28	3.98	2.32	11.9	96	0.14
2017A	6.46	9.86	2.87	1.67	2.26	60	0.23
2017B	6.65	7.30	1.00	0.58	1.26	99	0.22
2019A	6.73	6.15	0.25	0.15	2.12	78	0.23
2019B	6.81	7.34	4.45	2.59	1.66	81	0.18

. The soil pH values in the rehabilitated sites were slightly acidic (6.4) to near neutral (6.9), and they were higher than those measured in the control (6.2). Soil pH is well-known to exert a strong influence not only on plant growth, but also on soil microbiota distribution and composition [63,64,65,66]. Most cultivated crops, for example, can only tolerate slightly acidic (pH > 4.5) to slightly alkaline (pH < 9) conditions [67]. Nutrient availability and the risk of ion toxicity are two of the main reasons why many plants do not grow well under strongly acidic or highly alkaline pH conditions [68,69]. For microorganisms, increases in soil pH typically promote soil nitrogen availability, which increases soil bacterial diversity [69]. Meanwhile, decreases in soil pH enhance litter breakdown, organic carbon availability, and bacterial activities [70]. Finally, soil pH affects the mobility of trace heavy metals, impacts biodegradation, and was found to play an important role in determining metal speciation, solubility from mineral surfaces, movement, and eventually bioavailability of essential trace elements [71,72,73].

The moisture contents (MC) of the soil samples were quite low based on the Philippine soil standards of the Bureau of Soils and Water Management, Department of Agriculture (BSWM-DA) [73], ranging from 6.1% to 9.86%, which could be attributed to particle size distribution changes due to extensive mining activities [74,75,76]. Meanwhile, the organic matter (OM) contents of the soils ranged from extremely low (0.25%) to high (4.45%) (BSWM-DA) [73]. The availability and storage of nutrients in soils are significantly impacted by the presence of soil organic matter [77,78,79]. Organic matter improves soil structure, increasing rainwater infiltration and improving groundwater holding capacity [80]. In terms of organic carbon (OC) contents, this parameter ranged from 0.15% to 2.59%, which could be classified as extremely low to high (BSWM-DA) [73]. The lowest OC content was measured in the 2019A soil sample, at 0.15%, while the highest OC was observed in the 2019B soil sample, at 2.59%. Soil OC is a significant indicator of soil health because it directly influences soil properties, including water-holding capacity, aggregate stability, total nitrogen, pH, and cation exchange capacity [72].

The available phosphorus (AP) ranged from very low (1.17 ppm) to low (11.9 ppm) (BSWM-DA) [73]. The highest AP was measured in 2015B (11.9 ppm), the oldest rehabilitated area of the mine site, while the lowest AP was determined in the control (1.17 ppm). Phosphorus plays an important role in maintaining soil fertility and makes the agricultural production system more sustainable [76]. Although most soils have adequate P, the amount of available P, especially in Ni-laterite soil, is low as the mineralization of this nutrient is slow. Very little P moves in the soil, and this essential nutrient does not readily leach, even with large amounts of precipitation [73,74,75,76]. The exchangeable potassium (EK) in the soils ranged from 37.5–225 ppm, which is classified as low to medium (BSWM-DA, 2022) [73]. Soils from the control site had the highest EK (225 ppm), while the values of the soils from the rehabilitated sites were low. The total nitrogen (TN) was also low in all the sampled sites, ranging from 0.14% to 0.23% (BSWM-DA) [73]. Among the rehabilitated sites, the highest value was obtained from the 2019A sample (0.23%), while the lowest was from the 2015B sample (0.14%). As noted previously, TN is also an essential nutrient for soil production and plant growth [75,76]. It is interesting to note that contrary to the results of Li et al. [81], the soil OC contents did not follow an increasing trend with the increase in reclamation years.

3.3. Bacterial Communities in the Rhizosphere of S. spontaneum L.

The revegetation efforts of the mining company included various plant species, ranging from trees to vegetables. Based on a survey commissioned by the mining company, S. spontaneum was identified as a potential hyperaccumulator plant, which was present in both the control and rehabilitated sites (Supplementary Figure S1). Commonly known as wild sugarcane or Kans grass [82], this species belongs to the Poaceae family and is gaining attention for its ability to hyperaccumulate heavy metals, particularly Ni. Previous studies have demonstrated that S. spontaneum can tolerate and accumulate high concentrations of Ni in its biomass, making it a candidate for phytoremediation in contaminated environments [83,84]. Its extensive root system and high biomass production enable the efficient uptake of heavy metals from the soil, thereby reducing soil toxicity [85]. Additionally, recent studies have shown that S. spontaneum collected from one of the mining sites in Luzon, Philippines, is classified as a Ni hemiaccumulator, which is a category of plants that can tolerate Ni concentrations ranging from 100 to 999 mg/g in dry matter [86]. Its ability to thrive in harsh conditions, along with a high tolerance for other heavy metals such as Cd and Pb, highlight its potential in bioremediation strategies aimed at restoring ecosystem integrity [87].

To determine the different bacterial communities present in the rhizospheres of S. spontaneum from both the control and the rehabilitated sites, 16S rRNA sequencing was used. The amplicon sequence variants (ASVs) approach, a widely utilized high-resolution technique in microbiome research, was employed to examine bacterial diversity. Additionally, a rarefaction curve was used to assess the adequacy of the sample’s sequencing depth. In this study, the rarefaction curve plateaued, confirming that the sequencing depth was sufficient (Supplementary Figure S2). Our results identified a total of 1758 ASVs from high-quality reads obtained from seven (7) samples. The control samples had the highest ASVs (308), spanning 18 phyla, 116 orders, 190 families, and 327 genera, as illustrated in the Venn diagram (Supplementary Figure S3). A total of 47 shared ASVs were identified among all the samples (Supplementary Figure S3). Unique ASVs were identified in the following groups: 2015 (262 ASVs), 2017 (489 ASVs), 2019 (259 ASVs), and control (232 ASVs). These results indicate variations in ASVs across the samples, showing distinct bacterial communities in the rhizospheres of plants from rehabilitated sites [87,88].

The comparison between the bacterial communities present in the rhizospheres of S. spontaneum in the control and rehabilitated sites is illustrated using a heatmap (Figure 3). The heatmap uses a red-to-blue gradient, where red indicates high bacterial density, while dark blue represents lower bacterial density. The most abundant bacterial phyla in the control rhizospheres in order of their relative abundances were Proteobacteria (78.7%), Actinobacteriota (7.46%), Cyanobacteria (3.09%), Acidobacteriota (3.0%), and Bacteroidota (2.93%) (Supplementary Figure S4). In the rehabilitated sites, the relative abundance of the dominant phyla varied across different rehabilitation years. In 2015 (A and B), which was considered as the oldest rehabilitation site, the most abundant phyla in order of their relative abundances were Proteobacteria (88.3%; 80.7%), Actinobacteriota (6.57%; 10.8%), Bacteroidota (1.44%; 2.95%), Cyanobacteria (1.42%; 1.88%), and Firmicutes (0.60%; 1.42%) (Figure 3). Meanwhile, in 2017A, the most abundant phyla in order of their relative abundances were Proteobacteria (70.6%), Actinobacteriota (11%), Acidobacteriota (9.95%), Cyanobacteria (2.23%), and Bacteriodota (1.53%) (Figure 4). In 2017B, Proteobacteria (78.7%), Verrucomicrobiota (13.8%), Actinobacteriota (10.6%), Bacteroidota (2.41%), and Firmicutes (2.36%) were present (Figure 3). In 2019A, the most abundant phyla in order of their relative abundances were Proteobacteria (81.5%), Actinobacteriota (8.19%), Cyanobacteria (3.95%), Bacteroidota (2.26%), and Firmicutes (1.18%). In 2019B, the most abundant phyla in order of their relative abundances were Proteobacteria (85.2%), Actinobacteriota (6.96%), Verrucomicrobiota (2.39%), Cyanobacteria (1.69%), and Firmicutes (1.32%).

The findings of this study revealed that the phyla Proteobacteria and Actinobacteria were the most abundant groups present in both control and rehabilitated sites. Proteobacteria showed the highest abundance in the oldest rehabilitation site (2015), followed by the current rehabilitation site (2019), and then in the 2017 site. This pattern may be attributed to the 2017 site’s location in heavily mined-out areas, whereas by 2019, plant colonization was already evident (Figure 1). Proteobacteria are known for their metabolic versatility, contributing to nitrogen fixation, organic matter degradation, and heavy metal detoxification, which enhance soil fertility and support early plant colonization [89,90]. Certain members of the Proteobacteria phylum, such as Pseudomonas and Rhizobium species, are known for their ability to solubilize phosphate, fix atmospheric nitrogen, and produce plant growth-promoting substances, aiding in the re-establishment of vegetation in disturbed ecosystems [89,91]. Actinobacteriota are prolific decomposers of complex organic compounds and are crucial in soil organic matter turnover and pathogen suppression, promoting healthier soil ecosystems [92]. One known species of Actinobacteria is the genus Streptomyces, since it plays major roles in the cycling of carbon trapped in insoluble organic debris and it is the producer of the majority of, and diverse, bioactive secondary metabolites [92].

Cyanobacteria showed low abundance in most of the rehabilitated and control sites, except in rehabilitated site 2017B. This exception may be attributed to the site’s location at a higher elevation and possibly greater water flow, which can prevent Cyanobacteria from settling and forming blooms, thereby reducing their abundance. Another possible factor is the limited vegetation in the area, which may result in lower organic matter content (Table 3). In addition, this may be attributed to the soil contamination, which could favor the survival of other bacterial groups, such as those in phylum Acidobacteriota (Table 1 and Table 2). Despite their low abundance, Cyanobacteria play an important role in soil stabilization through biocrust formation and contribute to soil fertility via nitrogen fixation and photosynthetic activity, thereby accelerating vegetation growth in degraded lands [93]. Acidobacteriota was detected only in the control and 2017A rehabilitated soils. Their absence from the other rehabilitated sites (2015 and 2019) implied that Acidobacteriota may require specific conditions present in these sites. One possibility is that these genera, often associated with acidic and oligotrophic environments, benefitted from the pH or organic matter levels in 2017A, which had relatively high organic carbon. This is the only highly contaminated site where Acidobacteriota appeared, suggesting a complex interaction in which certain niches (e.g., high organic matter) allowed this species to persist despite the presence of heavy metals. Their scarcity in metal-affected soils could indicate sensitivity to some metals or to the nutrient-poor conditions typical of mine substrates.

Interestingly, Firmicutes displayed no abundance in 2017A. Soil characteristics, such as pH, nutrient availability, contaminants, and organic matter content, can influence the abundance and diversity of these two phyla in the soil [71,74,75,77]. Bacteriodota were present in the control and most of the rehabilitated sites, but were nearly absent in 2019B. Verrumicrobiota, although found in low abundance, were detected only in rehabilitated sites 2017B and 2019B, both located at low elevations and near each other (Figure 1). This genus includes species known to endure extreme oligotrophic conditions, so their presence might reflect adaptation to the nutrient limitations or metal stress in the rehabilitated areas. Identifying why this genus flourished slightly more in 2017B/2019B (maybe due to microhabitats with suitable moisture or substrate availability) would require further investigation, but it is considered as a part of the site-specific community in mined-out areas. Although it is less studied, this group is increasingly recognized for its role in carbon cycling and its ability to adapt to nutrient-poor environments, making it important for the restoration of barren soils [94].

The findings of this study aligned with the previous work of Roesch et al. [95], which reported that forest soil was predominantly composed of Proteobacteria. Similarly, Zhang et al. [96] found that in both natural and secondary forests, the major phyla were Proteobacteria (39.8%), Acidobacteria (17.6%), Actinobacteria (10%), and Verrucomicrobia (8.82%), which were also found in this study. These variations could be attributed to differences in vegetative communities between mined-out and unmined sites [97]. Generally, Proteobacteria were the most dominant phyla in all soil samples across both higher and lower elevation ranges, with relative abundances ranging from 70.6% to 88.3%. This suggests that the microbial community structure differed significantly across the rehabilitated sites, potentially due to environmental factors influencing microbial composition.

Furthermore, the work of Herrera [19] showed that the revegetation of Ni-laterite mine soil, silt, and spoils was dominated by Proteobacteria phylum. Similar results were obtained on a coal and bauxite mine revegetated with mixed plant species, wherein the most dominant phyla were Proteobacteria and Actinobacteriota [95,98]. Moreover, recent findings by Liu et al. [99] showed that the dominant bacterial phylum community groups of a restored coal mine were Proteobacteria (20.5–42.1%), Actinobacteriota (21.6–41.3%), Acidobacteriota (4.72–23.7%), and Chloroflexi (2.90–16.3%) [99]. Proteobacteria, Acidobacteria, Bacteroidota, Chloroflexi, and Firmicutes were also found as the most abundant phyla in a variety of soil environments [100]. Interestingly, the phyla Actinobacteriota, Verrucomicrobiota, Cyanobacteria, Bacteroidota, Planctomycetota, and Firmicutes always co-occurred in all the soil samples, although their relative abundance was low (0.47–13.8%). This also includes the phyla Chloroflexi, Cyanobacteria, Firmicutes, and Gemmatimonadetes, which were also found in unmined soils, but their relative abundances were highly variable and typically represented less than 5% of the 16s rRNA reads [100].

At the genus level, the most abundant genera found in the control and the three rehabilitated sites were Ralstonia and Curvibacter (Figure 5). In lower abundance, other genera identified across these sites included Burkholderia–Caballeronia–Paraburkholderia, Diaphorobacter, Pelomonas, hgcl clade, Sphingomonas, and Novosphingobium (Figure 5 and Supplementary Figure S5). The most dominant genus, Ralstonia, had relative abundances ranging from 29.1% to 38.3% and exhibited a positive correlation with soil Cr concentrations. This explains the elevated abundance of Ralstonia in Cr-rich rehabilitated sites such as 2017A and 2019A. High Mn levels also favored Ralstonia, indicating a broad tolerance to multiple heavy metals rich in Ni-laterite soils. As a result, Ralstonia dominated the metal-rich soils, consistent with its known capacity to withstand and even metabolically interact with heavy metals [63,101,102,103]. These results are noteworthy because Ralstonia includes important soilborne plant pathogens that cause bacterial wilt disease in crops, such as tomatoes and potatoes [101]. In the rhizosphere, the presence of this pathogen can significantly impact bacterial community composition and diversity. However, the presence of diverse microbial communities may influence the colonization success of Ralstonia [100,101]. Moreover, Ralstonia eutropha plays a beneficial role in CO₂ fixation and the biosynthesis of polyhydroxyalkanoates (PHAs), which have applications in bioplastics and medicine [100,101]. Curvibacter species, such as Curvibacter lanceolatus, contribute to the carbon cycle by promoting carbonate precipitation, which may enhance soil structure and nutrient cycling [67,104]. Escherichia and Shigella, commonly associated with foodborne illnesses from contaminated fresh produce, were also detected, but only in the 2015A site. Their presence in this study may be attributed to the use of manure as fertilizer [105].

Other notable genera identified include Diaphorobacter and Sphingomonas, both of which have demonstrated the ability to degrade polycyclic aromatic hydrocarbons (PAHs), which are considered as persistent environmental pollutants. This suggests their potential use in bioremediation and sustainable land management [102,106]. Diaphorobacter was highly sensitive to heavy metals and thrived mainly in soils with low heavy metal contents. The Spearman correlation analysis showed a strong positive correlation between Diaphorobacter and silica (p = 0.024), suggesting a negative relationship with Fe and other heavy metals [10]. This genus prospered in the 2015A/B sites (silica-rich, Fe-poor soils) and was much less common in metal-contaminated sites. Thus, this genus can be used as a sentinel of comparatively normal soil conditions in the post-mining ecosystem. Novospingobium has been shown to degrade estrogens from livestock manure, thereby reducing the environmental and public health risks posed by endocrine-disrupting compounds (EDCs) [107,108]. Finally, Pelomonas saccharophila, identified in the roots and rhizosphere of drought-tolerant Lasiurus sindicus (a perennial plant from Thar Desert in India), was shown by Choudhury et al. [26] to possess the nifH gene.

The genera Burkholderia–Caballeronia–Paraburkholderia were present across all sites but were more abundant in soils with higher pH, notably the 2019A and 2019B rhizospheres. The prevalence of this group in the 2019 site, which had moderate metal levels but relatively high pH suggests that these bacteria thrive under conditions of lower acidity and can tolerate the residual metals present. Meanwhile high exchangeable K in the soil was found to inhibit the Burkholderia–Caballeronia–Paraburkholderia group (as well other genera, like Pelomonas and Sphingomonas), highlighting that multiple soil chemistry factors modulate their abundance. Many Burkholderia-lineage bacteria are known for their metabolic versatility and some metal resistance, aligning with their persistence in rehabilitated mine soils [6,19,63,66]. In soils highly contaminated with heavy metals (2017A and 2017B), there was a notable shift in bacterial community composition, characterized by an abundance of tolerant bacterial taxa, while a decline in populations of more sensitive taxa occurred. Consequently, the elevated concentrations of heavy metals, like Ni, Mn, and Cr likely exerted selective pressure favoring the proliferation of resistant bacterial strains. It is likely that the pronounced toxicity associated with elevated heavy metal content reduced the bacterial adaptability in thriving under such adverse environmental conditions. Potential risks associated with bacterial inoculation of phytopathogenic bacteria include the unintended spread of introduced taxa, adverse effects on non-target plant species, disruption of native microbial communities, and the possible emergence of novel, resistant pathogenic strains. Accordingly, it is recommended that comprehensive risk assessments be conducted, accompanied by the careful selection of bacterial strains, implementation of containment protocols, and ongoing monitoring and evaluation to mitigate these risks effectively.

Another important aspect of this study was determining whether the observed microbial patterns were driven by the time since rehabilitation (i.e., site age in years) or by the differences in soil metal contamination. Our findings indicate that the heavy metal concentration, together with related edaphic factors, has a stronger influence on the soil microbiome than the mere passage of time after rehabilitation. In fact, there was no clear pattern of bacterial community change simply due to increasing rehabilitation age in this study. For example, the 2019 sites (only ~2 years post-rehabilitation) in some cases harbored microbial communities as developed or even more diverse than the older 2017 sites, despite being younger. This lack of a linear age-related trend is consistent with other reports, which found no association between rehabilitation time and the structure and function of bacterial communities [109,110,111,112,113,114]. Instead, the degree of heavy metal pollution and soil conditions at each site were the key drivers distinguishing the microbial communities. The 2017 rehabilitated areas, the most contaminated soils with Ni and other heavy metals, showed a distinct community skewed toward metal-tolerant taxa and missing some groups present in less polluted sites. Meanwhile, the 2015 rehabilitated soils, although the oldest, had the lowest metal content, and their microbial community reflected this almost normal environment (supporting taxa such as Diaphorobacter, which are absent from high-metal sites) [10]. The 2019 sites, despite their younger age, had moderate heavy metal levels and relatively favorable pH/OM, resulting in communities that in some respects approached the diversity of the 2015 sites. These observations suggest that remediation age alone does not dictate microbial recovery; rather, it is the quality of the soil environment (i.e., metal toxicity, nutrient availability, and pH) that governs how closely a rehabilitated soil’s microbiome regains a “natural” structure. Thus, reducing heavy metal stress and improving soil conditions can accelerate microbial community development, whereas even a longer rehabilitation period may fail if pollutants remain elevated. Effective rehabilitation should therefore prioritize lowering metal bioavailability through remediation or amendment alongside planting of hyperaccumulator plants, to create conditions conducive to a diverse microbial ecosystem.

4. Conclusions

In this study, the extent of soil pollution and microbial community structure from the rhizosphere of S. spontaneum from a rehabilitated Ni-laterite mine was assessed. The soils used for rehabilitation in 2015 were “pristine” and had a “low degree of pollution”, but those used in 2017 and 2019 were “moderately contaminated” with at least one of the following heavy metals: Ni, Cr, Co, Pb, Zn, and Cu. Rehabilitated soils were generally low in moisture content (MC), available phosphorus (AP), and total nitrogen (TN); exchangeable potassium (EK) levels ranged from low to medium; and organic matter (OM) or organic carbon (OC) content varied from extremely low to high. The most abundant bacterial phyla across both control and rehabilitated sites were Proteobacteria and Actinobacteriota, while Cyanobacteria was present in all sites except 2017B, likely due to its location in a heavily mined-out area. Acidobacteriota was found only in the control and 2017A sites, in contrast to Firmicutes, which was present in all sites except for 2017A. Verrucomicrobiota had low abundance overall but was more prominent in 2017B and 2019B, which are two adjacent sites.

In this study of a rehabilitated Ni-laterite mine, we found a link between the heavy metal concentrations in the soils and the abundance and diversity of the rhizosphere microbial community across different rehabilitation years. Total metal concentrations and specific metals (Ni, Cr, Mn, and others) showed significant correlations with microbial community, often overriding the effects of time since rehabilitation. Soils from 2015 rehabilitated areas with low metal levels supported higher microbial communities and included taxa sensitive to metal stress, whereas the more contaminated 2017 sites had decreased diversity and were dominated by metal-tolerant genera. Key taxa like Ralstonia prospered under Cr- and Mn-rich conditions, in line with their positive heavy metal correlations, while Diaphorobacter and others were retarded by heavy metals and mainly thrived in the control site. Planting of S. spontaneum and natural succession improved microbial diversity over time, in a manner consistent with previous findings in other post-mining sites. Overall, our results indicate that metal concentration was a dominant factor shaping microbial community structure, and that increasing rehabilitation age was insufficient to recover the original diversity if heavy metal levels remained elevated. Rehabilitation strategies should therefore focus on ameliorating heavy metal contamination (e.g., through soil amendments, selection of metal-tolerant plant microbe partners, or other remediation techniques) to create a soil environment that can support a rich and balanced microbiome.