Interactive effects of biochar amendment and lead toxicity on soil microbial community

Abstract

This study determined the interactive effects of biochar and lead toxicity on the soil microbial community in a phytoextraction experiment. Arranged with a completely randomized design in a greenhouse, banana liners were planted singly in a sandy soil spiked with Pb(NO₃)₂ at 0, 400 and 1200 mg kg⁻¹ and amended with bamboo biochar (pyrolyzing at 600 °C) at 0, 1, 3%. Soil samples were taken from triplicated pots five months after planting and measured for (i) content of lead and organic carbon; (ii) lead speciation; and (iii) microbial community composition through 16S rRNA gene sequencing. DNA sequencing results showed that lead and biochar treatments had significant individual and interactive effects on soil microbial dissimilarities from taxonomic levels of phyla to genera. While some specific taxa were lead resistant, biochar addition apparently alleviated lead toxicity and increased their richness (e.g., Alkanibacter, Muciaginibacter, Burkholderiaceae, and Beggiatoaceae). Soil analysis data indicated that biochar not only helped retain more lead in the soil matrix but created a soil environment inducive for transformation of lead into highly insoluble pyromorphite. This study highlights the effectiveness of biochar for lead remediation and the sensitivity of soil microorganisms in sensing changes in soil environment and lead bioavailability.

Graphical Abstract

1. Introduction

Lead (Pb) is one of the most frequently encountered heavy metal contaminants in the environment. Soil contaminated by anthropogenic Pb poses serious threats to human health, especially for children. About 1% of the total global burden of disease is attributed to mild intellectual developmental disorder and cardiovascular problems caused by Pb exposure (Fewtrell et al., 2004). Conventional physical and chemical remediation methods to clean up Pb contaminated sites include physical removal of material and engineering controls (site stabilization) to reduce Pb acute toxicity and control the risk of Pb exposure. However, the high cost of these remediation practices has limited their wide scale application. Industries and governmental agencies are seeking cost-effective and eco-friendly technologies for sustainable environmental remediation. Plant-based bio-remediation technologies have recently received considerable attention as alternatives of conventional cleanup of Pb-contaminated sites (Muthusaravanan et al., 2018). One of the plant-based remediation approaches is to apply biochar, a pyrolytic black carbon derived from thermal conversion of plant biomass in an oxygen-limited environment (Uchimiya et al., 2010, Mohan et al., 2014), as a soil amendment to reduce Pb bioavailability (Cao et al., 2011, Mukherjee et al., 2011, Uchimiya et al., 2012, Abdelhafez et al., 2014). Biochar has been reported to efficiently immobilize Pb in soil and water through a number of adsorption mechanisms such as electrostatic extraction, cation exchange, complexation, and precipitation (Cao et al., 2011, Mukherjee et al., 2011, Uchimiya et al., 2012, Ding et al., 2016). These sorption mechanisms are mostly associated with biochar’s large specific surface area, high cation exchange capacity, porous structure, and abundant functional groups. Significant positive effects on plant growth are also achieved with biochar amendment due to its ability to improve soil physical properties, nutrient status, pH, and soil organic matter content (Tomczyk et al., 2020).

While feedstock type and pyrolysis temperature are key engineering considerations for biochar production to suit for site specific soil conditions (Uchimiya et al., 2011, Ding et al., 2016), the application rate is perhaps the most important factor controlling the performance of biochar amendment on Pb remediation. Most studies evaluated heavy metal remediation efficiencies at application rate ≥ 5% (on a weight basis or ≥120 metric tons/ha) with Abdelhafez et al. (2014) reporting the maximum transferable Pb achieved at 10% application rate for sugar cane biogases and orange peel biochar, Cao et al. (2011) indicating superior performance of 5% manure- or litter-based biochar over biomass biochar for its high content of phosphate to form pyromorphite and hydroxypyromorphite, Xu et al. (2018) treating Pb-spiked soil with 5% macadamia nutshell biochar, and Cheng et al. (2018) noting that bioavailable Pb concentrations decreased by increasing biochar (tobacco stalk) application rate with the 5% amendment rate performing better than lower doses. At lower application rates, the biochar effects may be difficult to detect due to soil heterogeneity and uncertainty of conventional environmental measurements.

Soil microorganisms are perhaps the most sensitive soil biota to subtle changes in soil Pb bioavailability and soil environment. While it is generally accepted that biochar C is intrinsically low in biodegradability and largely unavailable to soil microbes (Zimmerman, 2010), biochar-induced changes in soil physicochemical properties may impact microbial biomass, respiration, and community structure (Smith et al., 2010). The impact on soil microbial community composition and structure can vary with soil type and biochar dose with enrichment of key taxa within the parent soil microbial community (Khodadad et al., 2010, Herrmann et al., 2019). In general, biochar amendment increased relative abundances of microbes that are related with C and N cycling and use of chemically bound phosphate in soils (Anderson et al., 2011, Xu et al., 2014, Wu et al., 2016). By contrast, the toxic effects of bioavailable Pb in soils may suppress microbial metabolism, reduce biomass production, and shift microbial community composition along with changes in diversity and relative abundance (Liu et al., 2018, Sobolev and Begonia, 2008, Xu et al., 2018). Liu et al. (2018) note that Pb resistance and metabolic activities of microorganisms were prompted upon exposure to Pb contamination. Xu et al. (2018) found that soil incubation with Pb(NO₃)₂ (5000 mg kg⁻¹ for 49 days) resulted in an increase of Gram-positive bacteria accompanied by a decrease in the abundance of fungi and Actinomycetes. The immediate inhibitory effects of Pb can be reflected in taxa related with C and N cycling (Sobolev and Begonia, 2008, Xu et al., 2019).

Despite an already overwhelming number of studies on the effects of biochar or Pb toxicity on soil microbes, relatively few studies have examined the interactive effects of these two factors on the soil microbial community at low dosage of biochar amendment whereby transformation and bioavailability of Pb can become complex and difficult to decern due to interactions among biochar, plant, and the soil environment. We investigated these interactions with a greenhouse study using banana as a test plant of phytoextraction growing in a sandy soil with three levels of biochar amendment (0%, 1%, and 3%) and three levels of spiked Pb (0, 400, and 1200 ppm). The objectives were (1) to examine the interactive effects of biochar amendment and Pb toxicity on the soil microbial community; and (2) to elucidate the roles of biochar in taxa-specific Pb resistance and soil Pb speciation changes.

2. Materials and Methods

2.1. The soil and biochar

About 60 kg of Myakka soil (Entisol), a signature flatwood soil in Florida, was taken from a farmland in Apopka, Florida (USA). The Myakka soil is comprised of 94% sand, 3% silt, and 3% clay with a pH of 5.3 and electrical conductivity of 0.09 dS m⁻¹. Mehlich-III extracted NO₃-N, NH₄-N, and potassium were 0.32, 0.37, and 29.79 mg kg⁻¹, respectively. Selected mineral elements that are beneficial to plant growth (Mehlich-III extraction) are included in Table 1.

Table 1.

Characterization of the Myakka soil and bamboo biochar.

	Bulk density (g cm−3)	pH	Cation exchange capacity (mmol 100 g−1)a	Specific surface area (m2 g−1)b	Total Pb (mg kg−1)c	Elemental composition (g kg−1)d
						Ca	Mg	K	Zn	Fe	Mn	P
Soil	1.65	5.8	–	–	15	0.53	0.09	0.03	0.02	0.23	0.02	0.20
Biochar	0.29	7.9	73.1	247.2	–	1.58	1.82	10.2	0.33	0.57	0.47	9.48

^a.Cation exchange capacity (CEC) of the biochar was determined using the point zero net charge method. CEC of the soil was not determined though it is expected to be much lower than for biochar given its sandy nature.

^b.Specific surface area (SSA) of the biochar was determined using the Brunauer–Emmett–Teller (BET) method on Quantachrome Autosorb1 at 77 K in the 0.01–0.3 relative pressure range of the N₂ adsorption isotherm. SSA of the soil was not determined though it is expected to be much lower than for biochar.

^c.Total Pb content in the soil was determined with a X-ray fluorescence spectrometer. Total Pb content in the biochar was not determined but should be low as well.

^d.Major elements were determined using Mehlich-III extraction for the soil and acid digestion for the biochar by inductively-coupled plasma atomic emission spectroscopic (ICP-AES, Perkin Elmer Optima 2100 DV) analysis.

A biochar was produced by pyrolyzing bamboo stems at 600 °C using a furnace apparatus (Bartlett 3K-CF, Fort Madison, IA, USA). Preliminary trials indicated that biochar produced at this temperature had higher pH, porosity, and specific surface area than biochar pyrolyzed at lower temperatures (Ding et al., 2016, Tomczyk et al., 2020). The resulting biochar was ground into fine particles < 2 mm. Key physical and chemical properties of the biochar were analyzed (Table 1). The biochar had a bulk density of 0.29 g cm⁻³ and a pH of 7.9. The specific surface area of the biochar was 247 m² g⁻¹ and cation exchange capacity about 74 mmol 100 g⁻¹. Major mineral elements of the biochar were included in Table 1.

2.2. Experimental design

The biochar was mixed with the soil at three rates: 0%, 1%, and 3% based on dry weight. One week after biochar amendment, each medium was spiked with Pb(NO₃)₂, at the rate of 0, 400, and 1200 mg Pb kg⁻¹ (or ppm), respectively. These Pb treatments were selected to correspond the US EPA’s standard for Pb in bare soil of play areas (400 mg kg⁻¹) and non-play areas (1200 mg kg⁻¹). A total of nine growing media were formulated with the naming convention taking the Pb and biochar treatments such that P0B0 stands for the control with no Pb and biochar amendment while P12B3 represents 1200 mg kg⁻¹ lead spike and 3% biochar amendment. The media were completely mixed and incubated at 24 °C for a week.

The nine media were respectively filled into 10-cm plastic pots, 500 g each, 12 pots per medium. Tissue culture liners of banana plant (Musa ‘Truly Tiny’), purchased from Agri-Starts, Inc. (Apopka, FL), were planted into the pots. There was no significant difference in the tissue culture liners of banana plant used in the experiment. Banana was selected as a test plant of phytoextraction for its rapid growth with high biomass accumulation and over 70% of the total root mass being confined in the upper 20 or 40 cm of the soil (Araya et al., 1998). The experiment was arranged as a completely randomized design with 12 replications. Plants were grown in a shaded greenhouse under a photosynthetically active radiation of 500 μmol m⁻² s⁻¹. Temperatures in the shaded greenhouse ranged from 20° to 32°C and relative humidity varied from 50% to 100%. All plants were fertilized with top dressing 4 g of a controlled-release fertilizer (Osmocote 19 N-2.18 P-7.47 K, 8–9 month; The Scotts Co., Marysville, OH) per pot one week after transplanting, at a rate of 0.75 g N per pot. Plants were irrigated three to four times a week. Plant growth was monitored by weekly visual examination for any growth disorders and monthly recording of canopy height, widest width, and width perpendicular to the widest width. Photosynthetic parameters including net photosynthetic rate, intercellular CO₂ concentration, transpiration rate, and stomatal conductance were measured using a Li-6800 Portable Photosynthesis System. These growth data along with biomass and phytoextraction efficiency will be reported separately (Chen et al., unpublished data).

2.3. Soil sampling and analyses

About five months after banana was planted, fresh soil samples were collected from triplicate pots of each treatment for soil microbial community analysis. Clean spatulas were used to scrape soil from the growth media at approximately upper, middle, and lower depths from the pot to obtain a homogenous composite sample. The sample was placed into a 50 mL sterile Falcon tube and transported on ice to the laboratory. The samples were held at 4 °C, and portions of soil samples were shipped cold to the DNA Environmental Sample Preparation and Sequencing Facility at Argonne National Laboratory. There, DNA was extracted from the soil, the V4 region of the 16S rRNA gene of Bacteria and Archaea was amplified by PCR with extracted DNA as template (Stackebrandt and Goebbel, 1994), and amplicons were sequenced on an Illumina MiSeq platform all according to procedures used by the Earth Microbiome Project (Thompson et al., 2017, Earth Microbiome Project Protocols and Standards, 2020a).

The remainder of soil samples were air dried and partitioned into two size fractions: 2–0.15 mm and < 0.15 mm using standard sieves to examine the partition of Pb and C between soil particle size fractions (Acosta et al., 2011). Total Pb content was determined using an X-ray fluorescence spectrometer (XRF, Thermo Scientific Niton XL3). Soil organic C was analyzed using a CE Elantech (Lakewood NJ) Flash EA elemental analyzer. Speciation of Pb in soil samples (<0.15 mm) of the 1200 ppm Pb treatments including P12B0, B12B1, and P12B3 was identified from X-ray absorption near edge structure (XANES) spectra collected at sector 10 (MR-CAT) beamline 10-ID (Segre et al., 2000) at the Advanced Photon Source of the Argonne National Laboratory in Lemont, IL. To prepare for XANES data collection, the individual soil sample (~100 mg) was mixed with 10 mg of polyvinylpyrrolidone gently ground with an agate mortar and pestle to homogenize, pressed into a 1-cm pellet, and mounted onto Kapton tape. The XANES data were collected across the Pb L_III edge (13035 eV) in fluorescence using a Lytle detector purged with pure argon (Ar) gas. Data were collected in step scan mode between 12835 eV and 13279 eV (8k). A Pb reference foil spectrum was collected concurrently with each scan using an ion chamber detector. The I₀, I_t and I_ref Ion chambers were all purged with pure nitrogen (N₂) gas. The data were merged in Larch and imported into ATHENA (Ravel and Newville, 2005) for normalization. A linear combination fitting (LCF) procedure was performed on the XANES region (−20 eV to +50 eV) using a library of over 20 known spectra of Pb minerals and adsorbed species. Analytical details are provided in the supplementary material (Figs. S1 & S2). The values of two replicated samples were both reported to show the variability.

2.4. Data analysis

DNA sequences were processed with mothur software using a standard set of commands (Schloss et al., 2009, Mothur, 2018). The mothur split.cluster command was used to construct the distance matrix and sequences having > 97% nucleotide identities were clustered within an Operational Taxonomic Unit (OTU), a conceptual unit of species or populations for applying statistical analyses first developed for plants and animals. This level encompasses species within a bacterial genus (Stackebrandt and Goebbel, 1994). Sequences were randomly subsampled to 7842 sequences for each of 27 samples. The modified SILVA non-redundant v135 16S rRNA sequence database file available through the mothur website was used as reference for identifying the OTUs with mothur by the classify.otu command (Mothur, 2018, SILVA, 2020b, Quast et al., 2013). The DNA sequences are available from GenBank under Bioproject PRJNA752169.

The number of DNA sequences OTU⁻¹ sample⁻¹, coverage as percent of all estimated sequences obtained from a soil sample, and alpha diversity indices were obtained using either mothur (Schloss et al., 2009) or Primer (Aukland, New Zealand). Two-way ANOVA (Minitab 19 software) was used to test for treatment effects on number of OTUs, Margalef’s species richness index, Pielou’s evenness index, Shannon’s diversity index, and Simpson’s diversity index. Tukey test was used to compare difference among means. The same analyses were performed on soil total Pb and organic C contents.

Sequences OTU⁻¹ sample⁻¹ were 4th root transformed and analyzed using Bray-Curtis similarities across samples or multivariate analyses based on rank order by permutation (Clarke et al., 2014; Primer-e). Analyses run were: (i) group averaged hierarchal clustering of Bray-Curtis similarities with SIMPROF tests for significance of clusters, (ii) PERMANOVA for analysis of similarity between microbial communities by treatment (Anderson et al., 2008), (iii) non-metric multidimensional scaling (nMDS) to display separation of microbial communities based on OTU ranks, (iv) Multidimensional Scaling (MDS) to display centroids of treatments determined with PERMANOVA, (v) ordered ANOSIM to test for significance of seriation of community differences by treatment doses, and (vi) SIMPER to rank OTUs by percent contribution of each to dissimilarity scores between treatment groups. The 25 OTUs identified from ranked 0 vs 1200 ppm Pb or 0 vs 3% biochar SIMPER analyses were identified with the Silva database and their relative contributions to sequence abundance within samples are visualized by matrix shade plots in Primer. PERMANOVA uses the term P(perm) to indicate the significance level of differences between samples. The significance level of all statistical analyses was set at α = 0.05.

3. Results

3.1. Soil lead and carbon

Two-way ANOVA indicated strong treatment effects on total Pb and soil organic C (p < 0.001) for the two particle size fractions upon biochar amendment and Pb spike (Fig. 1A & B). The < 0.15 mm size fraction contained much higher Pb and C than the coarser fraction (2–0.15 mm). For each size fraction, soil Pb or C means were significantly different with respect to Pb spike or biochar amendment treatments at α = 0.05. Biochar amendment increased soil Pb content marginally, but the effects were not significant for both size fractions (p = 0.15 and 0.35, respectively). However, under the 1200 ppm Pb treatments, total soil Pb of both size fractions did increase with increasing biochar amendment rate (Fig. 1C). Lead treatment did not have a significant effect on soil C content (p > 0.05). The interactions between Pb and biochar treatments on soil Pb and C contents were not significant (p > 0.05).

Fig. 1.

Soil total Pb and organic C content in 2–0.15 mm and < 0.15 mm size fractions. (A) Change of soil total Pb with Pb spike level; (B) Change of soil organic C with biochar amendment; and (C) change of soil total Pb with biochar amendment in the 1200 ppm Pb spike level. Error bars are standard errors.

XANES spectra indicated that Pb existed in the soil mainly as Pb minerals including anglesite, cerussite, and Pb phosphates, as well as lead adsorbed to ferrihydrite and organic materials (biochar in this case). Fig. 2 shows the percentage of each of these Pb species in the 1200 ppm Pb treatments with varying biochar amendments. Anglesite, which was present in all samples, ranged from about 30% in P12B0 to about 10% in P12B3, decreasing with increasing biochar amendment. In contrast, Pb sorbed by organic materials was not found in P12B0, which received no biochar additions, and increased to between 20% and 45% in P12B1 and P12B3, showing significant increase with biochar bound Pb. Also note that pyromorphite, which is a highly insoluble lead phosphate (K_sp= 10⁻⁸⁴), was only present in P12B3 (~26%) while cerussite was only present in P12B0 (11%).

Fig. 2.

Relative amounts of each Pb species in soil spiked with 1200 ppm Pb at varying levels of biochar amendment. Results were obtained from X-ray absorption near edge structure (XANES) spectra using the high energy resolution fluorescence detection. Replicated samples were analyzed.

3.2. Soil microbial communities

Comparisons of bacterial community composition are based on the OTUs obtained by grouping 16S rRNA gene sequences at the 97% identity level. Sequences obtained in the gene libraries were estimated with mothur to represent 90 – 97% of 16S rRNA sequence in each sample (Table S1). Archaeal sequences were a minor component (0.6%) of the sequence data and had little influence on outcomes of the statistical analyses. Thus, the results apply foremost to bacterial communities.

Bacterial community responses to Pb and biochar treatments were first compared using species richness and diversity values determined within each treatment (alpha diversity). Numbers of OTUs and calculated species richness values differed (p = 0.011) between Pb-treated and control soils, and not between biochar-treated and control soils (p = 0.151) due to high variability (Table 2). Soils with added Pb had fewer OTUs relative to soils with no added Pb. Neither treatment alone showed an effect (p > 0.05) on evenness, nor on the Shannon or Simpson diversity indices. By two-way ANOVA, however, interactions of Pb and biochar had significant effects on numbers of OTUs, species richness, evenness and community diversity indices (p = 0.003–0.037).

Table 2.

Two-way ANOVA and mean values of species numbers and diversity measurements of soil microbial communities under Pb and biochar treatment. Margalef’s species richness index, Pielou’s evenness index, and the Shannon and Simpson indices calculated with Primer-e 7. Standard errors of the means are in parentheses. Two-way ANOVA p-values indicating significant differences are in bold.

Treatment		OTU Total	Margalef’s Species Richness	Pielou’s Evenness	Shannon Diversity	Simpson Diversity	Inverse Simpson
ppm Pb	% biochar
		Mean values
0	0	1155 (143)	128.7 (16)	0.76 (0.02)	5.4 (0.2)	0.98 (0.01)	55.7 (16)
0	1	616 (46)	68.6 (5.1)	0.65 (0.03)	4.2 (0.2)	0.94 (0.02)	19.5 (4.8)
0	3	837 (151)	93.3 (17)	0.69 (0.05)	4.7 (0.5)	0.97 (0.01)	36.2 (12)
400	0	542 (50)	60.4 (5.5)	0.65 (0.02)	4.1 (0.2)	0.94 (0.02)	19.6 (4.2)
400	1	775 (35)	86.4 (3.9)	0.71 (0.02)	4.7 (0.1)	0.97 (0.00)	34.2 (4.2)
400	3	640 (64)	71.3 (7.2)	0.71 (0.02)	4.6 (0.2)	0.97 (0.01)	36.0 (6.9)
1200	0	616 (54)	68.6 (6.0)	0.67 (0.01)	4.3 (0.1)	0.95 (0.00)	20.6 (0.9)
1200	1	535 (30)	59.6 (3.3)	0.69 (0.01)	4.3 (0.0)	0.96 (0.01)	26.0 (4.3)
1200	3	551 (11)	61.4 (1.3)	0.69 (0.01)	4.3 (0.1)	0.96 (0.01)	29.3 (4.4)
		Two-way ANOVA
Pb		0.001	0.001	0.612	0.090	0.899	0.200
Biochar		0.151	0.151	0.758	0.612	0.404	0.507
Pb x Biochar		0.003	0.003	0.019	0.008	0.039	0.037

The significance of treatment effects was further explored by PERMANOVA first with OTU data and then at higher levels of taxonomy (Table 3). Both Pb spike and biochar amendment had significant effects on communities based on OTU analyses. Pair-wise tests indicated significant differences between all Pb treatment levels and between the 0% and 3% biochar amendment, consistent with the ordered ANOSIM test (Table S2) which indicated much more significant seriation in the effects of Pb treatment than biochar. Interactive effects of ppm Pb and % biochar on OTUs were also significant. These treatment effects and interactions significant at the OTU level were also detected at the higher taxonomic levels except for the main treatment effect of biochar at the phylum level.

Table 3.

P(perm) values for Pb x biochar PERMANOVA at different taxonomic levels. Treatments of % biochar are indicated in parentheses. Sequences sample⁻¹ were compiled into progressively higher taxonomic levels according to the silva taxonomic hierarchy. Taxonomic levels are numbered from lowest to highest, (n) for level 2, 37; for level 3, 103; for level 4, 344; for level 5, 407; for level 6, 3337 and include some unapproved names. Bolded values indicate significant effects in distributions between treatments.

		Taxonomic level
Source	df	2 (Phylum)	3 (Order)	4 (Class)	5 (Family)	6 (OTU)
Pb	2	0.0006	0.0001	0.0001	0.0001	0.0001
Biochar	2	0.1054	0.0198	0.0210	0.0198	0.0011
(0)v(1,3)	1	0.1596	0.0384	0.0054	0.0384	0.0204
(1)v(0,3)	1	0.2313	0.2147	0.3887	0.2147	0.0047
Pb x Biochar	4	0.0037	0.0003	0.0034	0.0003	0.0010
Pb x (0)v(1,3)	2	0.0204	0.0222	0.0297	0.0222	0.0038
Pb x (1)v(0,3)	2	0.0006	0.0015	0.0382	0.0015	0.1675
Residual	18
Total	26

Divergence of microbial communities by comparisons is seen in the nMDS and MDS plots (Fig. 3). Communities from soils with no Pb added were widely spaced along a diagonal in the nMDS plot, and communities from Pb spiked soils were separated from, and formed tighter groups than soils without Pb (Fig. 3A). The wide spacing of no Pb treatments, in contrast to the more tightly clustered 400 and 1200 ppm treatments, likely reflects the loss of species richness upon Pb exposure as described above (Table 2). Communities with biochar amendments were stratified to the upper part of the MDS graph in contrast to those without biochar in the lower part of the graph (Fig. 3B).

Fig. 3.

Separation of microbial communities based on Bray-Curtis similarity values. A, 3 D non-metric multidimensional scaling including all replicated measurements. B, metric multidimensional scaling of microbial community centroids. The same symbols in both plots represent the level of Pb amendment and the numbers in plot B refer to the % biochar addition.

3.3. Taxa specific changes

Variation in relative abundance of the top ten classes contributing most to Bray-Curtis dissimilarities between treatments is shown in Fig. 4. In the pristine soil (P0B0), the most abundant five classes were Bacteroidia, Gammaproteobacteria, Alphaproteobacteria, Verrucomicrobia, and Acidobacteriia. Biochar amendment tended to increase Gammaproteobacteria but reduced Bacteroidia and Verrucomicrobiae while Alphaproteobacteria and Acidobacterila remained about the same. In contrast, Pb treatment increased Gammaproteobacteria, Alphaproteobacteria, Verrucomicrobiae, and Acidobacteriia but reduced and Bacteroidia (Pb toxicity). These treatment effects were also shown in P12B3, which had more Gammaproteobacteria and Verrucomicrobia and less Bacteroidia and Acidobacteriia and less Alphaproteobacteria than in P0B0. Other notable classes that reflected treatment effects included Bacilli (increased by biochar), Actinobacteria (increased by Pb), and Clostridia (reduced by Pb).

Fig. 4.

Relative abundance of sequences within classes for each Pb treatment and level of biochar. Classes were ranked by SIMPER analysis of the P0B0 and P12B0 treatments, and the ten having the highest relative abundances plotted. Values are means of triplicate samples, each with 7842 sequences. Error bars represent the standard error of the mean.

SIMPER analysis, combined with shade plot displays, was used to show the OTUs that contribute most to Bray-Curtis dissimilarities between treatments. The most important 25 ranked OTUs from the 0 vs 1200 ppm Pb SIMPER comparison contributed 4.2% of the Bray-Curtis dissimilarity between treatments (Table S3), whereas the first 25 ranked OTUs for the 0 vs 3% biochar SIMPER comparison accounted for 3.5% of the dissimilarity (Table S4). Their relative contributions to sequence abundance within samples are visualized by the shade plot in Fig. 5. The 25 taxa clustered across samples by the Pb SIMPER analysis were divided between two large groups (Fig. 5A). The top cluster may be considered Pb resistant or tolerant while the lower cluster is Pb sensitive as demonstrated by gradation of the shade across treatments. Most of the identified taxa were unique to either the sensitive or resistant groups. Example OTUs corresponding to putative Pb-resistant bacteria in the upper group of the dendrogram were identified as OTU30 Alkanibacter, OTU59 Betaproteobacteriales, OTU29 Beggiatoaceae, OTU04 Muciliaginibacter, OTU45 Sphingobacateriales, OTU48 Microscillaceae, and OTU25 Burkholderiaceae among others. In contrast, sensitive taxa in the lower cluster include OTU0149 Lachnospiraceae, OTU40 Parvibaculum, OTU0171 Chirinophagaceae, OTU0208 Diplorickettsiaceae, OTU56 Opitutus and others. OTU30 Alkanibacter and OTU04 Muciliaginibacter were among the highest proportions of microbial communities in Pb-dosed soils. Higher percentages of apparent Pb-resistant bacteria in P12B3 than in P12B0 suggests a mitigating effect of biochar to Pb toxicity or bioavailability. Some of these OTUs identified at the family or genus level found among the resistant taxa are OTU45 Sphingobacateriales, OTU59 Betaproteobacteriales, OTU04 Muciliaginibacter, and OTU25 Burkholderiaceae.

Fig. 5.

Effects of Pb and biochar on bacterial communities among treatments. Identifications determined with the Silva v132 database and mothur. A. Most abundant 25 taxa within treatment identified from 0 vs 1200 ppm Pb SIMPER analysis. B. Most abundant 25 taxa within treatments of combined 0 vs 3% SIMPER analyses. Scale bars are 4th root transformed abundances within treatments for n = 7842 sequences. Dendrograms illustrate clustering of taxa among samples.

Proportions of the 25 taxa across treatments that contributed most to Bray-Curtis dissimilarity based on the 0 vs 3% biochar SIMPER analysis are displayed with the shade plot in Fig. 5B. The dendrogram divides co-occurrences of taxa over samples into two main groups with distinct sensitivity to biochar. One group of ten taxa such as OTU0280 Bryobacter, OTU312 Bacteroidia, and OTU40 Parvibaculum at the bottom of the shade plot with two subclusters demonstrates sensitivity to both biochar and Pb. The third subcluster of this group, which includes two taxa, OTU09 Mucilaginibacter and OTU38 Chitinophagaceae, was less biochar sensitive and can survive in a Pb contaminated niche. The other group in the upper part of the dendrogram formed two clusters, demonstrating high percentages upon biochar amendment with varying Pb sensitivity. Two types of taxa can be identified. The ones with Pb resistance such as OTU41 Lacunisphaera, OTU29 Beggiatoaceae, OTU05 Burkholderiaceae, OTU21 Pedobacter, and OTU104 Sphingobacteriales are found in greater percentages with biochar amendment in Pb-treated soils. The remaining taxa such as OTU259 Sphingomonadaceae, OTU0195 Bacteroides, OTU58 Herminiimonas, and OTU20 Diplorickerttsiaceae may be responsive to biochar amendment but cannot thrive with Pb toxicity. Thus, as visualized in the shade plots, Pb and biochar may exert synergistic effects on development of microbial communities, including perhaps a mitigating effect of biochar on Pb bioavailability.

4. Discussion

4.1. Effects of lead toxicity on soil microbial community

The effect of Pb toxicity as a main driver of changes in soil bacterial community is clearly observed in the diversity and richness indices and the composition at 5 taxonomic levels from Phylum to OTU (Table 2, Table 3), supporting similar Pb-spiked incubation experiments conducted in a green house or laboratory setting (Akmal et al., 2005, Sobolev and Begonia, 2008, Xu et al., 2019). Our results contrast well with Xu et al. (2019)’s 2500 and 5000 mg kg⁻¹ of Pb incubation. These authors, based on the microbial phospholipid fatty acids (PLFA) analysis, indicated that Pb toxicity had a great negative influence on bacteria with 8–32% reduction in abundances. Sobolev and Begonia (2008) conducted incubation at Pb levels of 1, 500, 1000, and 2000 mg kg⁻¹ and indicated that Pb had detectable effects on community diversity at relatively low concentrations, and that several concentration thresholds were observed for shifts in community diversity. Field studies comparing microbial communities between Pb contaminated and uncontaminated soils also reported distinct dissimilarity of microbial community diversity and structure (Hu et al., 2007; Beattie et al., 2018), albeit no significant heavy metal induced differences in diversity or richness indices between sites of 100 years of Pb and Zn mining and unmined locations were also noted (Xu et al., 2017).

While the specific effects on microbial community structure are often better elucidated at lower taxonomic levels, at the class level Pb toxicity clearly suppressed Bacteroidia while increasing the relative abundance of Alphaproteobacteria and Gammaproteobacteria (Fig. 4), both of which belong to Proteobacteria, a predominant heavy metal resistant phylum in many polluted environments. These classes were also identified on contaminated sites of mining operations located in humid temperate or subtropical regions in the US and China (Guo et al., 2017, Beattie et al., 2018). At the OTU level, Alkanibacter and Muciliaginibacter were among the highest proportions of microbial communities in Pb-dosed soils. Both have been found to be resistant to heavy metals including Pb (Fan et al., 2018, Li et al., 2018).

Significant treatment effects of soil Pb spiked at 400 and 1200 mg kg⁻¹ levels (Table 2) are consistent with soil microbes Pb toxicity threshold levels, reported generally between 200 and 600 mg kg⁻¹ (Akmal et al., 2005, Akmal and Xu, 2009). Much lower levels were also noted by Sobolev and Begonia (2008), who found detectable effects on community diversity for a soil spiked even by 1 mg Pb kg⁻¹. In general, low levels of Pb toxicity would create a soil environment conducive for microorganisms to develop Pb resistant strategies while high Pb concentrations mainly induce physiological adaptations (Renella et al., 2003). In this experiment, the top 200 ranked OTUs and top ten dominant classes are common among all treatment groups and represent typical soil bacteria, though Pb toxicity resulted in differences in Bray-Curtis values between communities. This may suggest that Pb toxicity in this experimental condition did not instigate significant lethal effects on most of these bacteria. While the higher order taxonomic composition of the community was retained, the lower order taxa within those may vary. Within the higher taxons, lower taxa may have been greatly diminished, and others responded adaptively by developing Pb-resistant mechanisms through extracellular and intracellular strategies or genetic exchange and others (George and Wan, 2019). Pb-resistant bacteria (e.g., Alkanibacter, Betaproteobacteriales, Beggiatoaceae, Muciliaginibacter, Sphingobacateriales, Microscillaceae, and Burkholderiaceae) and Pb sensitive taxa (e.g., Lachnospiraceae, Parvibaculum, Chirinophagaceae, Diplorickettsiaceae, Opitutus) were evidently affected by Pb treatment (Fig. 5A).

4.2. Effects of biochar amendment on soil microbial community

Notable soil physical/chemical properties influencing microbial community composition of Pb contaminated soils are soil organic matter and pH (Jiang et al., 2019). The biochar used in this study is high in pH, specific surface area, and cation exchange capacity, and contains abundant C and mineral nutrients that are poor in the soil (Table 1). Biochar amendment undoubtedly improved the physicochemical nature of the soil such as pH, C and mineral content, and soil water holding capacities with increases in soil pH following biochar application being the most frequently reported, especially in acidic soils (Lehmann et al., 2011). These improvements in soil properties were responsible for the shift in the microbial community, albeit with lesser statistical significance than with Pb treatment (Table 2; Fig. 3). Biochar amendment obviously increased the abundance of class Gammaproteobacteria, Bacilli, and Alphaproteobacteria but reduced Bacteroidia and Verrucomicrobiae.

The biochar effects on microbial community were also observed at the OTU level in Fig. 5B and by other workers (Smith et al., 2010, Zimmerman, 2010, Lehmann et al., 2011). Note that two OTUs responding to biochar were identified as Sphingobacteriales. Sphingobacteriales were considered to account for increased CO₂ emissions from nitrogen-fertilized wetland soil (Smith et al., 2010), and their presence in biochar amended soils is consistent with that finding. Xu et al. (2018) also indicated that 5% biochar treatment increased microbial respiration, microbial biomass C, and soil C use efficiency in a Pb-spiked silty loam. The relative abundance of OTU29 Beggiatoaceae increased with biochar added. Beggiatoa have a range of physiological capabilities, including chemolithotrophy with reduction of nitrate (Strohl, 2015). OTU42 Neorhizobium, which are capable of N fixation, also increased in abundance, suggesting a need for N fixation possibly in response to active C use activities (Mousavi et al., 2014, Abujabhah et al., 2018, Xu et al., 2018). Other studies have reported increase of N cycling bacteria with addition of biochar and nutrients, or biochar in N-rich soil (Xu et al., 2014, Wu et al., 2016, Wu et al., 2017). Biochar addition likely provides active carbon surfaces for these bacteria to enhance C and N cycling.

The impacts of biochar on the structure and function of soil microbial community depend on soil type and biochar application rate, which, in turn, determine the extent to which soil properties can be changed by biochar amendment. We used 1% and 3% application rates of biochar pyrolyzed at 600 °C in an acidic sandy soil. While the biochar effects on the number of OTUs and Shannon diversity index in our study (Table 2) are similar to Abujabhah et al. (2018), in which applications of biochar (pyrolyzed at ~700 °C) varied from 0% to 10% in three soil types (a black clay loam, a brown sandy loam, and a red loam) of decreasing soil organic C content and CEC, biochar-induced taxa specific changes are, however, different. For example, Abujabhah et al. (2018) indicated no significant differences with relative abundance of Proteobacteria among all biochar applications in all soils except at the highest level (10%) in the brown sandy loam, whereas we fund very sensitive response of Proteobacteria with biochar addition, even at 1% rate. In a recent meta-analysis of effects of biochar on soil microbial biomass and diversity, Li et al. (2020) concluded that biochar produced under high temperatures (600–700 °C) and high application rate tends to decrease microbial diversity because it can dramatically disrupt the microenvironment for microbial growth. This is consistent with the result show in Table 2 though our application rate was low. The impact on soil microbial community seems to be more significant in acidic and sandy soils amended by biochar with high pH, high porosity, and high specific surface area. In contrast, a clayey soil with high organic matter content may not experience as significant impacts on soil microbial community by biochar amendment.

4.3. Interactive effects of lead toxicity and biochar amendment on soil microbial community

While Pb toxicity and biochar amendment both influenced microbial communities, their interactive effects (Table 2, Table 3) may be partly attributable to higher percentages of apparent Pb-resistant taxa (e.g., Sphingobacateriales, Betaproteobacteriales, Muciliaginibacter, Beggiaceae, and Burkholderiaceae) in P12B3 than in P12B0 (Fig. 5A). The “preferential selection” of Pb resistant taxa by biochar suggests a mitigating effect of biochar to Pb toxicity or bioavailability. Several mechanisms could be at work together. First, biochar fragments, visually identifiable in both soil size fractions, may have created microhabitats (e.g., soil pH, moisture conditions, C substrate) that were niches for these Pb-resistant taxa. In other words, the microhabitats created by biochar, which were also found with addition of inorganic minerals into soil (Carson et al., 2009), may trigger certain microbes to adapt more rapidly and lead to fitness gain among the community. Typically, fine particles, especially clays, preferentially absorb Pb and C due to their larger specific surface area. This was also the case in this study even with added biochar (Fig. 1), which had a difference particle size distribution from the soil. While preferential partitioning of Pb and C in fine particle size fraction supports Acosta et al. (2011) and signifies the risk of human inhalation and ingestion of dust originated from contaminated sites, the heterogeneity induced by biochar addition (Fig. 1B) may also explain the variability in soil microbial community as shown in Fig. 3. Note that a more abundant distribution of these taxa was also found in P0B3 than in P0B0, though not as significant as in P12B3, suggesting that added biochar promoted the growth of these bacteria even without the presence of Pb toxicity. It is possible that the C compounds in biochar, albeit being more recalcitrant than soil organic matter in the natural system (Högberg et al., 2007), served as the energy source required by these taxa to develop Pb resistant strategy (Xu et al., 2018). The magnitude of soil organic C content relative to biochar application rate also suggests that some of the added biochar C went through microbial degradation within five months of incubation. The presence of heterogenous microhabitats created by biochar amendment may also partly explain the large variabilities in soil C contents and microbial community (Fig. 1, Fig. 4).

Second, biochar has been reported to efficiently immobilize Pb in soil through a number of adsorption mechanisms in additional to changes in soil properties (Cao et al., 2011, Mukherjee et al., 2011, Uchimiya et al., 2012). This preferential adsorption of Pb by biochar, albeit not statically significant due to the low biochar application doses, explains the marginal increase in Pb content with increasing biochar amendment (Fig. 1C). While the complexation of Pb by functional groups, such as carboxyl and hydroxyl in soil organic matter, may impede the formation of pyromorphite in soils (Hashimoto et al., 2009), our data suggested that biochar addition at 3% promoted formation of pyromorphite (Fig. 2). It is possible that biochar created a soil environment, i.e., improved soil pH and water holding capacity, that was inducive to formation of pyromorphite. Lead speciation data (Fig. 2) suggested that Pb salts and inorganic Pb-bearing minerals were transformed into pyromorphite (Pb₅(PO₄)₂Cl,OH), which is several orders of magnitude less soluble than the most commonly found Pb minerals in soils such as anglesite (PbSO₄) and cerussite (PbCO₃). This transformation is accompanied with reduced anglesite percentage, which can be induced by increasing soil pH upon biochar amendment (Zhang and Ryan, 1998). The transformation of soil Pb to pyromorphite would reduce the bioavailability and therefore Pb toxicity to soil bacteria, though soil total Pb content may still increase through Pb adsorption onto biochar (Fig. 1C). These microorganisms, as they develop Pb resistant strategies, may interact with Pb by releasing organic chelating agents and exopolymers which can precipitate Pb or reduce its activity at the cell surface (George and Wan, 2019). Furthermore, biochar may also add phosphate into the soil-biochar matrix (Table 1) that may facilitate the formation of pyromorphite.

Last but not least, the combined effects between biochar and Pb in promoting the thriving condition for Pb resistant taxa may, in turn, influence how the community interacts with soil processes and the plant. The toxic effects of Pb may inhibit enzyme activities or even damage cell membrane or DNA structure (Yang et al., 2016; Chen et al., 2018; Tipayno et al., 2018). The selection of microorganisms for Pb-resistant forms of nitrite reducers upon exposure to Pb was evidenced by changes in the community harboring nirK in Pb-spiked soil (Sobolev and Begonia, 2008). Biochar, on the other hand, may potentially enhance the growth of organisms involved in N cycling in the soil, specifically those that decrease the flux of N₂O (Xu et al., 2014), thus altering soil N dynamics. The detoxifying effect of biochar may have contributed to the microbially mediated soil processes of N utilization and transformation, which may be partly attributable to the greater vigor of banana visually observed in pots with biochar amendment. Epelde et al. (2015) indicated that genes involved in transposition and transfer of genetic material were significantly more abundant in highly contaminated soils whereas genes related to stress and starvation responses were more abundant in less contaminated samples. Thus, metatranscriptomic or metagenomic analyses of the genes would provide further information about how biochar detoxifies Pb contamination through microbially mediated soil processes.

5. Conclusions

Through microbial community and taxa specific analyses, it is evident that biochar amendments, even at low doses of 1–3%, have detoxifying effects over Pb contamination in shaping the structure of soil microbial communities. The potential mechanisms involve biochar’s ability to improve soil microhabitats and to immobilize Pb in the soil-biochar matrix through surface adsorption and chemical transformation to highly insoluble pyromorphite. This study highlights the effectiveness of biochar for Pb remediation and the sensitivity of soil microorganisms in sensing changes in soil environment and Pb bioavailability. Future research can take a metatranscriptomic or metagenomic approach to elucidate the relationship between nitrogen and carbon cycling functional genes and Pb resistance upon biochar amendment in Pb contaminated soils.