The effects of soil micro-environment on the adaptation of Cynodon dactylon (L.) Persoon in a copper smelter’s slag dump

Abstract

Background and Aims

Cynodon dactylon (L.) Persoon is a clonal plant often used in slope restoration and greening projects. Two different types of C. dactylon were found in a copper smelter’s slag dump. The relationship between the two C. dactylon phenotypes and the micro-environment in the slag was investigated, aiming to identify the most suitable form of C. dactylon for restoring such slag dumps.

Methods

Copper, lead and zinc were extracted using the modified European Community Bureau of Reference sequential extraction procedure, and their morphologies were analysed. A risk screening procedure was used to assess the risk posed by soil contamination with Cu, Pb and Zn. Correlation analysis, redundancy analysis, principal component analysis and significance of difference were used to analyse the relationships between plant growth and the physicochemical properties of the soil and its Cu, Pb and Zn content. The structure of the microbial community was analysed using high-throughput sequencing.

Key Results

With only mild Cu contamination, the Cu enrichment coefficient of upright C. dactylon (type B) reached 0.71. The Cu transport mode was rhizomatous. The Cu content of the roots was significantly higher (P ≤ 0.05) than that in creeping C. dactylon (type A), and the relative abundance of the rhizosphere microbial community was significantly greater (P ≤ 0.01) than that of the A-type. The function of the genes is mainly metabolic, followed by replication and repair functions.

Conclusions

Upright C. dactylon (B) exhibits a capacity for Cu enrichment that makes it highly promising for the ecological restoration of copper smelting slag dumps.

INTRODUCTION

Mining and the use of mineral resources is the basis of much economic development and social well-being, but large piles of extracted residues not only affect soil quality, biodiversity, water resources and landscapes, but also make the land uninhabitable and visually unacceptable. They can also migrate into the soil under the effects of wind erosion and leaching (Mendez and Maier, 2008; Manhart et al., 2019), contaminating the surrounding soil and jeopardizing health (Zhang, 2020). It is estimated that more than ten billion tons of tailings are produced every year worldwide, and the large amount of tailings has resulted in tremendous waste of time and space, in addition to soil pollution, especially heavy metal pollution (Adiansyah et al., 2015).

Modern remediation techniques for soil heavy metals are mainly physicochemical and bioremediation (Li et al., 2024). Among the bioremediation techniques, phytoremediation and plant–microbial joint remediation have been studied extensively in recent years owing to their low cost, simple operation and wide practicability (Sarwar et al., 2017). A group led by Liu evaluated the effectiveness of Arundo donax L. for in situ remediation of Cd and Pb contamination in non-ferrous metal mining and smelting zones (Liu et al., 2019). Shi et al. (2023) reported using a rhizobium–legume symbiont to remediate heavy metal-contaminated soil. Yang et al. (2013) reported that Cynodon dactylon could be used as a pioneer species for the phytoremediation of manganese mine tailing sands. Song et al. (2022) used C. dactylon to remediate polycyclic aromatic hydrocarbons and Cd in contaminated soil.

Cynodon dactylon is an extremely heat-resistant, drought-tolerant and trampling-resistant grass. It is a perennial of the Poaceae Cynodon dactylon genus (grass family, Teff subfamily) with ample biomass, a fast growth rate and strong adaptability (Chen, 2000). It can spontaneously use roots, stolons, transverse rhizomes and other organs to produce potentially independent individuals in natural conditions (Dong, 1996). Phenotypic plasticity is one of the key characteristics of C. dactylon that enable it to adapt to complex and variable environments by coordinating physiological and morphological responses (Sheehy and Laskowski, 2023). There have been many studies of the variation in phenotypic traits of C. dactylon from cloned plants. Hutchings and De Kroon (1994) attributed the clonal phenotypic plasticity of the plant, primarily morphological plasticity, to its foraging behaviour. Dong and Pierdominici (1995) found that C. dactylon stolons show greater morphological plasticity in response to light than the rhizomes. Chen et al. (2021) have demonstrated that temperature can affect the height of the plant and its photosynthesis. Xie et al. (2016) did not find noticeable phenotypic changes in C. dactylon, but they showed that the heavy metal tolerance of the plant was significantly related to the soil microorganisms present, and that populations from heavily metal-polluted areas had greater genetic diversity than those from less polluted areas. Many studies have shown that the morphological variability of wild C. dactylon is strongly influenced by latitude, soil and climate (e.g. Zhang et al., 2018; Wang et al., 2020). Iqbal et al. (2022) showed that the 15 different ecotypes of C. dactylon they collected might have been a result of environmental variation.

In this study, two different growth traits of C. dactylon were found in a copper smelting slag storage area undergoing ecological restoration. Considering the phenotypeic plasticity of C. dactylon, it is speculated that there are important relationships between different growth traits of C. dactylon and several heavy metal ions, soil microorganisms and essential nutrients in the soil. These phenotypic differences might affect the enrichment capacity of heavy metals in soils.

To determine which C. dactylon growth type is most suitable for copper slag remediation and to explore the relationships between differences in C. dactylon growth traits and the soil micro-environment, the growth traits of erect and creeping C. dactylon, their ability to accumulate Cu, Pb and Zn, the physical and chemical properties of the rhizosphere soil and the microbial community structure were investigated at a copper slag storage site in Xichang, China. The purpose of this study was to provide theoretical support for the bioremediation of heavy metal-contaminated soil and to determine the type of C. dactylon most suitable for practical ecological restoration of such sites.

MATERIALS AND METHODS

Sampling

The samples in this study were collected from a copper smelting slag disposal site in Xichang City, Liangshan Yi Autonomous Prefecture, Sichuan Province, China (Fig. 1A and 1B) (102°19929′ E, 27°98028′ N, elevation 1599 m a.s.l.). A satellite image of the area is shown in Fig. 1C. Over time, pollution control and ecological restoration measures have been implemented at the study site (Fig. 1D). Plants have settled in the slag dump. The plant species are largely homogeneous, with C. dactylon, Paspalum paspaloides, Bidens pilosa, Eupatorium adenophora, Ricinus communis, Equisetum ramosissimum, Imperata cylindrica and Juncus ochraceus Buchenau dominating. Among them, C. dactylon has the greatest biomass, is the dominant species in the study area and shows two growth traits: upright and creeping. The study area has a subtropical highland monsoon climate, with annual sunshine hours ranging from ~1600 to 1800 h and an average annual temperature of 17.2 °C.

Fig. 1.

(A) A geographical location map (China, Sichuan). (B) A geographical location map (China, Sichuan Liangshan Prefecture). (C) A satellite image of the sampling points. (D) A real scene image of the sampling points. The red arrows and shapes indicate plot A; the blue arrows and shapes indicate plot B. (E) A photograph of plant A1. (F) A photograph of plant b1.

In China, no permit is required for collecting C. dactylon. Specimen selection was based on the growth traits of the stems in the study area. Five 1 m × 1 m sampling squares each were established for plants with upright and creeping stems. The creeping C. dactylon sampling points were labelled A1–A5; those of the upright C. dactylon were labelled B1–B5. A small shovel was used to collect into sampling bags intact roots of plants growing naturally in representative areas with large biomass of a single growth type. The C. dactylon samples were transported back to the laboratory and cleaned with deionized water for analysis of their physiology and biochemistry. The shovel was cleaned with deionized water before collecting the next sample. The surface soil on the root system of the C. dactylon down to 5 cm was collected and shaken loose on the spot. Some of the soil was sealed in a sampling bag and brought back to the laboratory to be dried naturally for physical and chemical analyses. The other portion of the soil was placed in sterile sampling bags, packed with ice in a foam box, and transported to the laboratory for storage at −80 °C (de Celis et al., 2024). Samples packed in ice in a foam box were sent to Biomarker Technologies Co., Ltd. in Beijing to analyse the structure of their microbial communities.

All the sampling, preparation and long-term storage were conducted in accordance with ISO technical specification 10381-2:2002, Soil quality sampling part 2: Guidance on sampling techniques.

Analytical methods

Plant growth and Cu, Pb and Zn analysis.

A ruler was used to determine the height of C. dactylon individuals, and Vernier callipers were used to determine the internode length, leaf length and leaf width. According to China standard DB62/T 4508-2022 for crop sampling and preservation, three to five samples were randomly selected from each sampling point, mixed to form a mixed sample and analysed centrally. The roots, stems and leaves were separated and cleaned twice with deionized water. The samples were then killed by heating at 105 °C for 30 min, dried at 70 °C, cut into pieces and placed in labelled sample bags. Their Cu, Pb and Zn contents were determined by microwave digestion with flame atomic absorption (Supplementary Data Method S1).

Physicochemical indices and determination of the forms of the heavy metals.

A portion of the naturally air-dried rhizosphere soil samples was sieved through a 2 mm sieve and used to determine the pH, available phosphorus and quick-acting potassium. Another portion was sieved through 100 mesh and used to determine the state and content of heavy metals.

The Cu, Zn and Pb contents of the soil samples were determined via the European Community Bureau of Reference continuous extraction method (Chakraborty et al., 2014). The ionic state (C1), reducible state (C2), oxidizable state (C3) and residual state (C4) were extracted sequentially with a mixture of 9 mL of nitric acid + 3 mL of hydrochloric acid + 1 mL of hydrofluoric acid, with microwave solubilization (a Shanghai Xin Yi, TANK Pro instrument). The specifics of the extraction process are shown in Supplementary Data Method S2. The same method was used with the plant samples. C1, C2 and C3 are termed the bioavailable states more easily absorbed and used by plants (Rodríguez et al., 2009). Chinese national standard samples and blanks were used for quality control during the Cu, Zn and Pb analyses (see Chinese standards GSB04-1725-2004, GSB04-1761-2004 and GSB04-1742-2004). When the sample was determined, a blank control was set for each batch, and the determination was repeated three times.

The pH of the soil was determined according to ‘Soil Testing Part 2: Determination of Soil pH’ (standard NY/T1121.2-2006). The soil-to-water ratio was 1:2.5. The available phosphorus concentration in the soil was determined using an SFP-3 comprehensive analysis device for soil composites (FHK, Japan). The quick-acting potassium (SAK) concentration was determined according to ‘Determination of quick-acting and slow-acting potassium content in soil’ (NY/T889-2004) with microwave digestion (Shanghai Xinyi, TANK Pro). There too, the determination method was the same as that used for the plant samples. Neutral ammonium acetate solution leaching was used with flame photometry to determine SAK levels according to ‘Determination of soil fast-acting potassium and slow-acting potassium content’ (NY/T889-2004).

Soil microbiology.

The five samples of each growth type collected were combined into a mixed sample. The bacterial community structure in the rhizosphere soils of the two growth types was analysed by 16S sequencing using the Illumina NovaSeq 6000 high-throughput sequencing platform with the 16s v3 + v4_b amplicon and the primers F: ACTCCTACGGGGAGGCAGCA and R: GGACTACHVGGGTWTCTAAT. Quality filtering was performed using v.3.0 of the Trimmomatic software and Cutadapt software v.1.9.1. DADA2 denoising was performed using QIIME2 2020.6, followed by the delineation of operational taxonomic units (OTUs) and amplicon sequence variants. Analysis of diversity, ANOVA, correlation analysis and functional prediction were also performed.

Data processing and statistical analysis

The means and standard deviations of the raw data were calculated, and the data were subjected to one-way ANOVA, bivariate correlation analysis and analysis of the significance of differences using SPSS Statistics v.23 software (IDM Corporation, USA). The graphing used Origin 2024 software (OriginLab, USA). Principal component analysis and redundancy analysis were performed using Past v.4.15 software (University of Oslo, Norway).

Bioconcentration and translocation

The ratio of root-to-soil heavy metal content was used as the bioconcentration factor (BCF_root) (Song et al., 2015; Buscaroli, 2017; Oliveira et al., 2023). The translocation factor (TF) characterizes heavy metal transport among plant organs. Values close to or greater than one indicate that a plant has a strong ability to transport heavy metals between organs (Yang et al., 1996). TF1 is the root–stem transfer coefficient, and TF2 is the stem–leaf transfer coefficient. If TF1 is greater than TF2, root–stem transport is the dominant mode of heavy metal transport, and vice versa. The formulas are as follows:

$$\begin{array}{l}\mathrm{B}\mathrm{C}{\mathrm{F}}_{\mathrm{r}\mathrm{o}\mathrm{o}\mathrm{t}}=\frac{C_{\mathrm{r}\mathrm{o}\mathrm{o}\mathrm{t}}}{C_{\mathrm{s}\mathrm{o}\mathrm{i}\mathrm{l}}}\end{array}$$ (1)

$$\begin{array}{l}\mathrm{T}\mathrm{F}1=\frac{C_{\mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m}}}{C_{\mathrm{r}\mathrm{o}\mathrm{o}\mathrm{t}}}\end{array}$$ (2)

$$\begin{array}{l}\mathrm{T}\mathrm{F}2=\frac{C_{\mathrm{l}\mathrm{e}\mathrm{a}\mathrm{f}}}{C_{\mathrm{s}\mathrm{t}\mathrm{e}\mathrm{m}}}\end{array}$$ (3)

where C_root is the concentration of heavy metals in the roots (in milligrams per kilogram), C_soil is the concentration of heavy metals in the soil of the rhizosphere (in milligrams per kilogram), and C_stem and C_leaf are the heavy metal concentrations in the stems and leaves (in milligrams per kilogram).

Heavy metal risk assessment

Based on the heavy metal accumulation pattern, the secondary phase concentration ratio risk screening procedure (RSP) can be used to quantify contamination risk (Zhao et al., 2013; Chen et al., 2017). The formula is as follows:

$$\begin{array}{l}\mathrm{R}\mathrm{S}\mathrm{P}=\frac{M_{\mathrm{s}\mathrm{e}\mathrm{c}}}{M_{\mathrm{p}\mathrm{r}\mathrm{i}\mathrm{m}}}=\frac{\mathrm{C}1+\mathrm{C}2+\mathrm{C}3}{\mathrm{C}4}\end{array}$$ (4)

where C4 (in milligrams per kilogram) is considered the primary phase (M_prim) and the sum of C1 + C2 + C3 (in milligrams per kilogram) is considered the secondary phase (M_sec). An RSP value ≤ 1 indicates no pollution risk; 1–2 indicates mild pollution risk; 2–3 indicates moderate pollution risk; and an RSP value ≥ 3 indicates severe pollution risk (Deng et al., 2024).

RESULTS

Cynodon dactylon growth

As shown in Fig. 2, C. dactylon (B) tended to be significantly taller than creeping C. dactylon (A), with significantly longer leaves (P ≤ 0.01). Cynodon dactylon (A) had significantly longer stem nodes (P ≤ 0.05). There was no significant difference in leaf width between the two growth types of C. dactylon.

Fig. 2.

Analysis of the growth traits of two forms (A and B) of Cynodon dactylon. Lowercase letters indicate significant differences between groups, and error bars represent standard deviations.

Heavy metal content and physicochemical properties of the soil

The physical and chemical properties of the soil are summarized in Supplementary Data Table S1. The pH of the soil around the rhisoomes of type A ranged from 7.78 to 7.98, whereas around type B it ranged from 7.42 to 7.49, thus for both types of C. dactylon the soil was slightly alkaline.

With type A, the available phosphorus content ranged from 10.00 to 49.00 mg kg⁻¹, and for type B it ranged from 22.00 to 47.33 mg kg⁻¹. The available potassium for type A ranged from 189.58 to 231.46 mg kg⁻¹, and for type B from 173.04 to 434.35 mg kg⁻¹. The differences in pH, available phosphorus and available potassium were not, therefore, significant.

Fig. 3 shows that the soil Zn, Pb and Cu at both sample sites was dominated by the residual state (C4) material. The relative content of C4-state Cu at plot A was 56.3–76.8 %, whereas at plot B it was 31.4–55.2 %, a significant difference (P ≤ 0.05). For C4-state Pb, the corresponding values were 80.7–87.0 and 77.0–84.6 %, a difference that is not significant. For C4-state Zn, the values were 80.6–86.6 and 78.0–84.9 %, again a difference that is not significant. The lowest relative percentages were for the ionic state (C1), with values of 1.3–9.2 and 4.3–10.2 % for Cu at plots A and B, respectively. The comparable values for Pb were as low as 1.7 and 1.3%, and for Zn as low as 0.3 and 0.5%. The residual state is generally considered to be more stable and less bioavailable. The relative contents of the bioavailable state of Cu were 23.2–43.7 and 39.6–49.7 % at plots A and B, respectively, a significant difference (P ≤ 0.05). The relative contents of the bioavailable state of Pb (13.0–19.3 and 15.4–23.0 %) and of Zn (13.5–19.4 and 15.2–22.0 %) were not significant. The difference in bioavailable Cu might have led to differences in the growth between C. dactylon types A and B. The differences in bioavailable Pb and Zn are unlikely to have been influential.

Fig. 3.

Percentage content of soil heavy metal states: Cu (A), Pb (B) and Zn (C). C1 represents the ionic state, C2 the reducible state, C3 the oxidizable state and C4 the residual state.

Heavy metal risks

The ionic form of heavy metals is the most migratory and toxic of the four forms, and it is also the most bioavailable form, hence the ionic forms of copper, lead and zinc were used to evaluate their environmental risks. The average content of ionic state Cu observed in the type A samples was 232.25 mg kg⁻¹. For Pb it was 484.62 mg kg⁻¹ and for Zn 151.49 mg kg⁻¹. With the type B samples, the corresponding averages were 316.44, 322.50 and 262.62 mg kg⁻¹. According to Chinese standard GB 15618-2018, the ionic Cu exceeded the screening value (100 mg kg⁻¹) by 2.32 and 3.16 times with the two types of samples. The ionic Pb exceeded the screening value (170 mg kg⁻¹) by 2.85 and 1.90 times. Ionic zinc, however, did not exceed the screening value (300 mg kg⁻¹). According to standard GB 15618-2018, Cu and Zn have no control values, but that for Pb is 1000 mg kg⁻¹.

The RSP method was used to evaluate the risks of Zn, Pb and Cu, with the results shown in Table 1. The RSP values of Pb and Zn in plots A and B were <0.3, indicating no pollution risk. The RSP value of Cu in plot A was the largest at 0.777, again with no risk. In B2, there was mild contamination, and in B3 there was a medium risk, although samples B1, B4 and B5 had no pollution risk. However, their RSP values were 0.961, 0.915 and 0.811, respectively, which are very close to one, indicating a mild risk potential.

Table 1.

Rhizosphere soil heavy metal risk evaluation table for Cynodon dactylon.

RSP	A1	A2	A3	A4	A5	B1	B2	B3	B4	B5
Cu	0.777	0.582	0.645	0.600	0.303	0.961	1.019#	2.186##	0.915	0.811
Pb	0.164	0.239	0.188	0.149	0.220	0.237	0.299	0.229	0.207	0.182
Zn	0.210	0.180	0.241	0.155	0.239	0.246	0.224	0.283	0.179	0.227

^#Light pollution; ^##moderate pollution.

To summarize, there was no environmental pollution risk from Pb or Zn in the rhizosphere soil of sample plots A and B. There was only a mild risk from Cu in sample plot B, which might have had some impact on the growth of upright C. dactylon.

Heavy metal bioconcentration and translocation

Fig. 4 shows that Zn, Pb and Cu levels in the roots and stems of type A C. dactylon were lower than those in type B, although the levels in the leaves were higher. In creeping C. dactylon (A), both Pb and Cu were highest in the roots, at 2008.29 and 1236.49 mg kg⁻¹, respectively. Zn was highest in the stems, at 3701.47 mg kg⁻¹. In general, the Zn, Pb and Cu levels were lower in the leaves than in the roots and stems. Upright C. dactylon (B) had the highest content of Zn, Pb and Cu in its roots (6334.11, 3252.23 and 2561.01 mg kg⁻¹, respectively). There was little in the stems and leaves. The Zn, Pb and Cu levels in the roots of type B C. dactylon were greater than those in the roots of type A, with significant differences in Cu (P ≤ 0.05) but no significant differences in Zn and Pb. The differences in Zn, Pb and Cu in the stems were not significant; but in the leaves they were (P ≤ 0.05). Indeed, the differences in Cu content were highly significant (P ≤ 0.01).

Fig. 4.

Cu, Pb and Zn contents in the roots, stems and leaves of Cynodon dactylon. White circles represent the overall data mean; grey shaded areas indicate the 25–75 % confidence interval.

Fig. 5A shows that the bioconcentration factors of Cu, Pb and Zn in type B C. dactylon were higher than those in type A, and that the bioconcentration capacities of Zn, Pb and Cu were BCF_root-Cu > BCF_root-Zn > BCF_root-Pb. The bioconcentration factor of Cu in type B C. dactylon ranged from 0.36 to 0.71, which was significantly higher than that in type A (P ≤ 0.01) The bioconcentration factor of Zn in type B C. dactylon ranged from 0.13 to 0.31, which was again greater than that in type A, although the difference was not statistically significant. Likewise, the BCF of Pb in type B C. dactylon ranged from 0.08 to 0.32, showing no significant difference from that in type A. The bioconcentration factor of Pb in type B C. dactylon ranged from 0.08 to 0.32, with no significant difference from that in type A.

Fig. 5.

(A) A radar plot of the enrichment coefficient of Cynodon dactylon. (B) A radar plot of the translocation coefficient of C. dactylon.

Fig. 5B shows that the translocation factors TF1 and TF2 of creeping C. dactylon (A) for Zn, Cu and Pb were higher than those of upright C. dactylon (B). The two types of C. dactylon exhibited rhisomes enriched in Zn and Cu. In upright C. dactylon (B), the rhisomes were enriched in Pb, whereas creeping C. dactylon (A) exhibited stem–leaf transport of Pb. Generally, the Cu, Pb and Zn transport capacities of creeping C. dactylon (A) were greater than those of upright C. dactylon (B).

The root and stem translocation factor (TF1) values for Cu, Pb and Zn in creeping C. dactylon (A) ranged from 0.02 to 2.59, from 0.04 to 1.21 and from 0.06 to 3.32, respectively, with TF1_Zn > TF1_Cu > TF1_Pb. The stem–leaf transfer coefficient (TF2) values of Cu, Pb and Zn ranged from 0.09 to 3.32, from 0.1 to 2.6 and from 0.04 to 1.00, respectively, with TF2_Cu > TF2_Pb > TF2_Zn. Creeping C. dactylon (A) accumulated Zn and Cu primarily in its roots and stems, and Pb tended to be distributed more in its leaves. Upright C. dactylon (B) also accumulated Cu, Pb and Zn in its roots and stems, but its stem–leaf transport coefficients for all three were smaller than those of type A C. dactylon.

Soil contamination and growth of C. dactylon

The Zn, Pb and Cu levels found in the roots and leaves of types A and B were correlated with the levels of bioavailable Zn, Pb and Cu in the soil. The Cu in the leaves of the C. dactylon in sample plot B exhibited a strong positive correlation with the bioavailable Cu content of the soil (R² = 0.961, P ≤ 0.01). The Cu content of C. dactylon roots in sample plot B exhibited a strong positive correlation with the bioavailable Pb concentration (R² = 0.972, P ≤ 0.01). The Pb content of C. dactylon leaves in sample B was correlated with the bioavailable Cu concentration (R² = 0.883, P ≤ 0.05), and the Pb contents of C. dactylon stems and leaves in sample B exhibited a significant positive correlation with the corresponding bioavailable Cu concentrations (R² = 0.894, P ≤ 0.05; R² = 0.974, P ≤ 0.05).

The plant growth results of redundancy analysis are shown in Supplementary Data Fig. S1. The leaf width of type B plants was significantly and positively correlated with the Zn content of their leaves (R² = 0.887, P ≤ 0.05), but there was no significant correlation in the type A plants.

Microbial community structure and heavy metal content

Simpson’s diversity index (Simpson), the abundance index (ACE) and a sequencing depth index (Coverage) were applied to reflect the diversity of soil microbes, their richness and the coverage of the sample libraries, respectively (Table 2). The coverage of the sample libraries was one, indicating that the sample sequence information could adequately reflect the real profile of the soil bacteria in the area studied. The diversity indices of the rhizosphere microbial communities of types A and B were calculated at the OTU level.

Table 2.

Rhizosphere soil microbial community diversity indices.

Sample ID	ACE	Simpson	Shannon	PD whole tree
Type A	393.2 ± 58.127b	0.986 ± 0.004a	7.353 ± 0.266b	21.694 ± 2.825b
Type B	499.2 ± 67.132a	0.990 ± 0.003a	7.896 ± 0.227a	28.713 ± 2.858a

Superscript lowercase letters indicate significant differences between groups in the same row; numbers following ‘±’ represent standard deviations.

The sequences of all the prokaryotic microorganisms found were clustered according to inter-sequence distance. Sample A was divided into 697 OTUs based on similarity, and sample B was divided into 1254 OTUs. The α-diversity analysis produced an abundance-based coverage estimator (ACE) for the soil bacteria in sample B of 401–516. The Shannon–Wiener index (Shannon) was 7.55–8.15, and the PD whole tree index was 25.2–30.9. The ACE index was significantly higher in sample B than in sample A (P ≤ 0.05), and the Shannon and PD whole tree indices were also significantly greater (P ≤ 0.01), hence the soil bacteria of C. dactylon type B are more complex than those of the creeping variant (type A).

Proteobacteria accounted for 44.4–68.3 % of the bacteria in samp A (type A plants), Actinobacteriota accounted for 13.3–14.6 %, and Bacteroidota accounted for 3.6– 16.1 % (Fig. 6). In sample B (type B plants), Proteobacteria (35.8–58.3 %), Actinobacteriota (4.3–13.1 %), Bacteroidota (6.5–11.1 %), Acidobacteriota (6.7–10.6 %) and Patescibacteria (2.2–7.5 %) were all detected. Proteobacteria, Actinobacteriota and Bacteroidota dominated in both the samples A and B, with the relative abundance of Actinobacteriota being significantly higher in sample A (P ≤ 0.01).

Fig. 6.

Soil bacterial community structure distribution map at the phylum level (A) and at the genus level (B).

Sample A contained bacteria from 116 genera, whereas in sample B 189 were represented. In sample A, Sphingomona (5.6–12.3 %), Massilia (1.3–14.3 %), Sphingoaurantiaca (1.5–16.9 %) and unclassified bacteria (4.7–5.9 %) predominated. In sample B, it was Sphingomona (6.4–10.5 %), Noviherbaspirilla (2.2–12.6 %) and unclassified Sphingomonadaceae (1.8–7.1 %).

PICRUST2 software was applied in a predictive analysis of prokaryotic microflora. The secondary functional level of the predicted genes was analysed, and the functions included carbohydrate metabolism, amino acid metabolism, cardiovascular diseases, replication and repair, folding, sorting and degradation, cell motility, signal transduction and 43 other sub-functions. The relative abundance of the 14 sub-functions differed significantly between the two samples (Supplementary Data Table S2). Under the metabolism primary function, amino acid metabolism, carbohydrate metabolism, energy metabolism, and sugar biosynthesis and metabolism were significantly greater in sample B than in sample A (P ≤ 0.05). Under the human disease primary function, infectious/parasitic diseases, cancer and immune diseases were significantly less in sample B than in sample A (P ≤ 0.05). The excretory system in the first-level function organic system was more obviously expressed in sample A (P ≤ 0.05).

Differences in Cu, Pb and Zn contents and plant phenotypes in rhizosphere soil bacteria and soil bioavailable states between the two sample sites

The correlation between the relative abundance of the top ten genera and the bioavailable Cu, Pb and Zn in the two types of rhizospheres is presented as a heat map (see Supplementary Data Fig. S2). The relative abundance of Sphingoaurantiacus in sample plot A had a significant positive correlation with bioavailable Pb, whereas in plot B it was positively correlated with the bioavailable Cu.

A redundancy analysis was conducted to examine the relationship between the relative abundance of the top ten genera of rhizosphere soil bacteria and the morphology of the plant phenotypes (Supplementary Data Fig. S3). It showed that the leaf length and leaf width of creeping C. dactylon (type A) and upright C. dactylon (type B) were positively correlated with the relative abundance of Massilia, and internode length and plant height were positively correlated with the relative abundance of Noviherbaspirilla. The plant height of upright C. dactylon (type B) was positively correlated with the relative abundance of Fonticella, and leaf length and width were positively correlated with the relative abundance of Phenylobacteria.

DISCUSSION

Environmental advantages of the different C. dactylon types

Root lengths and internode lengths differed significantly between types A and B of C. dactylon (P ≤ 0.05), and those variables strongly affect the growth of the plants. Clonal plants with a longer stem node (spacer) length have better fitness in heterogeneous habitats through better resource acquisition, propagation of progeny and clonal amplification (Zhang and Zhang, 2013). Huang et al. (2018) showed that a taller plant has better access to light, and Díaz and Cabido (2001) concluded that plant height and leaf area can reflect the regulatory ability of a plant in terms of access, utilization, compensation and balance. Although the difference in leaf width between C. dactylon types A and B was not significant, the overall phenotypic trend was that type B had wider leaves than type A, and the leaves of type B were significantly longer (P ≤ 0.05), hence the leaf area of C. dactylon type B was generally larger than that of type A.

Mantill-Perez’s study of the morphology of the strongest flag leaf confirmed its effects on leaf length and the leaf area index, which are key to optimizing light interception and photosynthesis. They are affected by plant height, which is caused by pleiotropy or related genes (Anderson and Denmead, 1969; Mantilla-Perez and Salas Fernandez, 2017). The increased leaf length in the modified plants was mainly attributable to longer cells rather than a greater number of cells. Additionally, the plants were shorter in height, which was attributable to a reduction in the distance between stem internodes rather than a reduction in the number of internodes or the length of one or more specific internodes (Zanella et al., 2023).

Cynodon dactylon, as a shade-intolerant plant, tends to present as type B in open copper smelting slag dumps, where its taller plant height and larger leaf area promote sunlight capture. In contrast, type A plants are shorter, have a smaller leaf area and are less able to capture light. Li et al. (2018) showed that with increasing stolon internode length, the survival and growth of cloned fragments increased significantly, which promoted the invasiveness of the plant. It can be argued that C. dactylon type B is a ‘modified variety’ of type A produced under environmental stress, a phenomenon known as phenotypic plasticity, i.e. environmental modification of genotype expression (De Kroon et al., 2005). In summary, upright type B can capture sunlight better, whereas creeping type A is more invasive (Li et al., 2018).

Cu, Pb and Zn remediation potential of C. dactylon

Zinc is an essential element for many plants. It is critical for the functioning of >300 enzymes involved in metabolism and resistance to stress and is closely linked to many DNA and RNA syntheses (Gupta et al., 2016; Choi et al., 2018; Kaur and Garg, 2021). The Zn transport in both types of C. dactylon is root–stem transport. The Zn content in the leaves of upright C. dactylon (B) is greater than that in creeping C. dactylon (A). Pb transport in creeping C. dactylon (A) is also root–stem transport, which is different from that in creeping C. dactylon (B). Bioavailable Pb mainly exists in the roots, which reduces the effect of Pb on the toxicity of the leaves and its effect on photosynthesis. It also explains why the C. dactylon B in the study area was more dense than the C. dactylon A.

The bioavailable Cu, Pb and Zn in the soil of the B samples (2296.23, 3137.64 and 5193.82 mg kg⁻¹, respectively), was much more concentrated than in the soil of the A samples (1667.92, 2485.14 and 4689.196 mg kg⁻¹). Upright C. dactylon had the highest BCF_root for Cu (0.71) and showed some potential for enriching Cu, but that did not reach hyper-enrichment. Mohammed et al. (2024) planted C. dactylon to alleviate heavy metal pollution in soil caused by oil extraction. In that study, the Cu enrichment coefficient of dogbane root reached 3.46, and for Zn it reached 4.56. For Pb it reached 17.30, which is far beyond the standard value for ultra-enriching plants. Khan et al. (2023) showed that the Cu removal rate of C. dactylon can reach 59 % over 90 days, which is a good removal effect. The observed differences in BCF_root might be attributable to the differing content of heavy metals in the areas studied. Despite the differences among the available studies, it cannot be denied that C. dactylon has enrichment potential for Cu, Pb and Zn.

Differences in soil microorganisms between the two C. dactylon types

Both the A and B soil samples were dominated by Proteobacteria, Actinobacteriota and Bacteroidota. Many studies have found that those three phyla of microorganisms dominate in heavy metal-contaminated soil, indicating that these three types of bacteria can adapt better to heavy metal contamination and that they might reduce the toxicity of heavy metals through growth and metabolism (Liu et al., 2020; Wu et al., 2022). In addition to those three phyla, Acidobacteriota and Patellibacteria dominated in certain areas of plot B. Previous studies have shown that Acidobacteriota can produce a variety of cellulose-degrading enzymes, such as cellulase and xylanase, and can form a symbiotic relationship with plant root systems to facilitate the uptake of soil nutrients and metabolites by the plants (Zeng et al., 2019; Jiang et al., 2022).

Sphingoaurantiacus can assist phytoremediation by promoting biosorption and active efflux transport and reducing toxicity (Gatheru Waigi et al., 2017). Brevundimonas was significantly and positively correlated with the bioavailable Cu content of the soil in the B samples. Brevundimonas functions as a nitrogen-fixing and phosphorus-solubilizing agent that promotes plant growth (Naqqash et al., 2020).

Feng et al. (2016) showed that the 6NM-7T strain of Massilia is tolerant of Cu, Pb, Zn and Cd in tailings, where it can produce dimethyl disulphide, indicating that this strain could be used to control cell damage in plant leaves and promote plant growth by increasing the effectiveness of reduced sulphur. Noviherbaspirillum, as a denitrifying bacterium, promotes the conversion of nitrogen and phosphorus from organic to inorganic forms in high-Zn conditions by increasing the activities of hydrolytic enzymes, such as urease, protease, acid phosphatase, sucrase, cellulase and dehydrogenase. It also increases the availability of soil nitrogen (Lv et al., 2022). In addition, Noviherbaspirillum can control the morphology and bioavailability of Cu in the environment by metabolizing cysteine to form copper sulphide nanoparticles with inorganic sulphur oxidation and reduction, and it is, potentially, Cu tolerant (Xu et al., 2023). Fonticella responds significantly to Pb and Zn fixation (Dan et al., 2023).

Amino acid metabolism controls various cellular processes, including cell proliferation, differentiation and homeostasis (Suzuki and Iwata, 2021), and it is also closely related to energy and carbohydrate metabolism, carbon and nitrogen homeostasis, hormone and secondary metabolism and the stress response (Qingqing et al., 2020). Such activity is predicted by the functional genes of a bacterium. In this study, the B samples showed more evidence of amino acid metabolism, carbohydrate metabolism, energy metabolism and sugar biosynthesis and metabolism than the A samples. This probably reflects better stress relief by the B-type C. dactylon rhizosphere soil bacteria, and better proliferation and differentiation (Wei et al., 2018). They could adapt better to the complex environment. The low expression of infectious parasitic and immune system diseases and cancer in the B samples and the strong expression of antimicrobial agents suggest that C. dactylon type B rhizosphere soil bacteria are better suited for tolerance, resistance and long-term environmental adaptation. Type B C. dactylon is thus likely to be more stable in ecological restoration applications.

The rhizosphere soil bacteria were more abundant in the B samples than in the A samples, and the relationships between microorganisms and plants and between microorganisms and soil physicochemical properties were more complex. Also, prophylactic bacteria and genera that can fix heavy metals were found in the B samples that were not present in the A locations. Nitrogen fixation, phosphorus solubilization, carbohydrate metabolism and energy metabolism are the core functions of soil bacteria (Tang et al., 2020). In a copper slag dump, C. dactylon type B rhizosphere soil bacteria better express genes for resistance than those associated with type A.

Conclusion

Our analysis of the relationship between phenotypic differences, growth patterns and the soil micro-environment of two different types of C. dactylon from a copper smelter’s slag dump has shown that upright C. dactylon is better able to reduce the toxicity of Pb in the slag. Its leaf area is greater than that of the creeping type, favouring photosynthesis and growth. The upright type also exhibited stronger bioconcentration of Cu, Pb and Zn than the creeping variant. The microbial composition of the soil in the rhizosphere of the upright variant was more complex and stable, and the metabolic functioning was stronger. It can be concluded that the upright variant is more suitable for heavy metal enrichment in the remediation of such tailings. It is the upright variant that should be the primary subject of future research.

SUPPLEMENTARY DATA

Supplementary data are available at Annals of Botany online and consist of the following.

Figure S1: plant growth in terms of redundancy analysis. Figure S2: a heat map of the microbial communities in the soil of the rhizosphere in relationship to Cu, Zn and Pb. Figure S3: redundancy analysis of the ten most prevalent inter-root bacteria and Cynodon dactylon growth traits. Table S1: physical and chemical indexes of the rhizosphere soils. Table S2: prediction of the microbial function of rhizosphere soils. Method S1: determination of Cu, Pb and Zn by microwave digestion with flame atomic absorption. Method S2: the European Community Bureau of Reference continuous extraction of Cu, Zn and Pb.

FUNDING

This study was supported by China’s National Key Research and Development Program (grant 2023YFC3708103), as a Regional Innovation Cooperation Project of Sichuan Province (no. 2022YFQ0081), by the doctoral research initiation program of Xichang University (grant YBZ202301) and as a key R&D project in the science and technology plan of Liangshan Yi Autonomous Prefecture of China’s Sichuan Province (no. 23ZDYF0142).

DATA AVAILABILTY

The data related to microbiology are available under accession number PRJNA1081184.