Succession of bacterial community structure in response to a one-time application of biochar in barley rhizosphere and bulk soils

Biochar is increasingly used as an amendment to improve soil fertility (Enders et al., 2012; Gul et al., 2015), especially as anthropogenic activities degrade soils at unprecedented rates (Ibrahim et al., 2019; Sarfraz et al., 2019; Ge et al., 2020). Biochar not only affects soil physicochemical properties (Liang et al., 2010; Sohi et al., 2010; Van Zwieten et al., 2010; Lehmann et al., 2011; Gul et al., 2015; Tomczyk et al., 2020) but also directly or indirectly influences soil microorganisms through changes in soil microbial community diversity and abundance (Lehmann et al., 2011; Gul et al., 2015; Palansooriya et al., 2019). In particular, soil bacteria mediate multiple biogeochemical processes, which, in turn, affect soil fertility and plant growth (Dai et al., 2016; Yu et al., 2018; Yu et al., 2020b), suggesting that understanding how biochar alters soil bacterial community composition can help inform biochar management plans and application rates.

Biochar may affect soil bacterial communities by altering soil physicochemical properties like pH (Lehmann et al., 2011; Zhu et al., 2017), nutrient availability (Farrell et al., 2013), and carbon (C) sources (Dai et al., 2016; Hütsch et al., 2002), which can affect their growth and maintenance. Because microbial-mediated soil nutrient cycling can influence soil fertility (Kidinda et al., 2023), changes in soil bacterial community structure may influence the effectiveness of biochar as a soil conditioner. Thus, to understand how changes in bacterial communities may affect nutrient cycling, it is necessary to assess changes in bacterial abundance and diversity, as well as the response of specific functional communities for taxa at the phylum or genus level (Dai et al., 2016; Whitman et al., 2016; Xu et al., 2023), and their interactions with other factors that influence microbial growth (e.g., soil pH, nutrient availability, carbon sources).

The region of soil surrounding plant roots—the rhizosphere—is often studied to understand interactions between plants and microbes relative to bulk soils and help infer key factors that influence plant growth (Orozco-Mosqueda et al., 2022). As a soil amendment, biochar may help plants acquire nutrients through changes in the soil environment that are specific to the rhizosphere, but not to bulk soils (Yu et al., 2020b; Shetty et al., 2021). For example, root exudates add labile C to the rhizosphere that can affect microbial activity (Shi et al., 2013; Wang et al., 2022), but adding biochar can further accelerate inputs of labile C by accelerating root growth and subsequent inputs of labile C (Farrell et al., 2015). Thus, understanding how bacterial communities respond to biochar inputs in the rhizosphere relative to bulk soils may help forecast changes in nutrient cycling and availability and overall soil fertility.

With the development of high-throughput sequencing technology, in-depth studies into the succession of soil microbial community structure have been made possible. For example, studies have shown that the abundance of specific functional bacterial taxa may increase after adding biochar (Palansooriya et al., 2019; Li et al., 2020). Similarly, other studies have documented changes in bacterial community structure in response to biochar inputs (Rutigliano et al., 2014; Singh et al., 2016; Liu et al., 2019). However, few studies have investigated the different responses of soil bacteria to biochar inputs in the rhizosphere relative to bulk soils and how these responses may change over time with changes in plant phenology (i.e., as plants transition from the seedling stage to the mature stage). Given that labile C inputs to the soil change as a function of plant phenology (Adair and Burke, 2010; Debnath et al., 2020) and biochar inputs (Farrell et al., 2015), understanding the interactions between biochar and plant phenology can help inform the application of biochar as soil conditioner to increase site fertility and maximize agricultural productivity.

In this study, we focus on differential responses of bacteria to biochar amendments in both space and time by contrasting rhizosphere and bulk soils over 3 stages of plant growth: the seedling stage, Year 1 mature stage, and Year 2 mature stage. Using DNA extractions and Hiseq sequencing analysis, we evaluate how soil functional bacterial taxa respond to biochar amendments by focusing on richness, diversity, and the relative abundance of bacteria sensitive to biochar inputs. We define sensitive bacterial communities as those expressing variation in community abundance or diversity indices after adding biochar (Dai et al., 2016). We hypothesized that (1) biochar would increase bacterial richness and diversity, especially in the rhizosphere relative to bulk soils; (2) some genera would be more sensitive responders to biochar addition relative to others, with their sensitivity varying across space and time between rhizosphere and bulk soils, and (3) the effect of biochar bacterial community structure would persist for 2 years despite a one-time biochar application at the start of our experiment.

We conducted a 2-year pot experiment using 0.05-µm pore-size root bags to exclude barley (Hordeum vulgare L.) roots and separate bulk soils from the rhizosphere. The pots were filled with soil obtained from Quzhou, Zhejiang Province, China, and classified as Ultisols (key soil properties are listed in Table S1).

A wheat-straw-derived biochar was prepared for the pot experiment using 3 treatments with biochar (N0C2, 30-g kg⁻¹ biochar equivalent to a biochar: soil ratio of 3%; N1C2, 0.23-g kg⁻¹ urea and 30-g kg⁻¹ biochar; N2C2, 0.46-g kg⁻¹ urea and 30-g kg⁻¹ biochar) and 3 treatments without biochar (N0, control; N1, 0.23-g kg⁻¹ urea; N2, 0.46-g kg⁻¹ urea; Table S2). Each treatment was replicated 3 times. We used 0.23-g kg⁻¹ urea (N1) because it corresponds to the typical application rate in barley agriculture used in Quzhou city (Yu et al., 2020b) and 0.46-g kg⁻¹ urea (N2) to observe biochar effects under excess N supply.

Biochar was prepared at 450°C with a raising rate of 26°C min⁻¹ and kept at this temperature for 2 h. After cooling, biochar samples were ground into powder and sieved (2 mm). The basic physicochemial properties of biochar are listed in Table S1 and described in detail by Yu et al. (2017) and (2020b). Based on technical and economic considerations, we did a pre-experiment by setting a concentration gradient of biochar additives (0, 1%, 3% w:w) and confirmed 3% did not have an adverse effect on soil while meeting the requirements for fundamental research. Biochar was only applied at the beginning of the pot experiment and mixed thoroughly with soils, while urea was applied as a fertilizer at the beginning of both the first and second years according to the dosage of each treatment (N0 = control, N1 = 0.23-g kg⁻¹ urea, and N2 = 0.46-g kg⁻¹ urea).

The pot experiment along with the 6 treatments was conducted in a ventilated and sheltered laboratory over the length of 2 barley growth cycles from winter 2014 through summer 2016. Over this period, we collected samples representative of 3 different growth stages: the barley seedling stage during the first year (Year 1 seedling stage), the barley mature stage during the first year (155 days after first sampling, Year 1 mature stage), and the barley mature stage during the second year (520 days after first sampling, Year 2 mature stage). At each sampling time, the root bags made of nylon material with 0.05-µm pore size were gently removed to prevent damage to the plant roots. After gentle shaking, soil adhering to roots was brushed off into the mesh bag, which represented rhizosphere soils after removing all visible roots (He et al., 2013; Yu et al., 2020b). Bulk soils were sampled 1 cm away from the nylon bags.

Basic physicochemical properties of rhizosphere and bulk soils were analyzed immediately after sampling. Soil dissolved organic carbon (DOC) was determined by MilliQ water (w:v, 1:10) extraction and measured on a TOC analyzer. The air-dried soil was measured for total carbon (C_total), pH, exchangeable base ions (Ca²⁺, Mg²⁺, K⁺, and Na⁺), and exchangeable Al³⁺. C_total was measured in an elemental analyzer (Flash EA 1112, Thermo Finnigan). Soil pH was determined using a pH meter by mixing with deionized water (w:v, 1:2.5). The exchangeable Ca²⁺, Mg²⁺, K⁺, and Na⁺ were extracted with 1-M ammonium acetate (pH 7.0) and determined by flame atomic absorption spectrometry (Analytikjena, Germany). Total exchangeable base cations (EBCs) were calculated by summing Ca²⁺, Mg²⁺, K⁺, and Na⁺ concentrations (Sartori et al., 2007). The cation exchange capacity was determined by the ammonium acetate exchange method (Matula et al., 2009). The concentration of soil exchangeable Al³⁺ was determined by the NaOH titration method and calculated from the difference between the exchangeable acidity and hydrogen (Baquy et al., 2018). Soil properties of Year 1 seedling stage, Year 1 mature stage, and Year 2 mature stage are listed in Table 1 with more details available elsewhere (Yu et al., 2020b). Multifactor analysis of variance (ANOVA) of soil physicochemical properties was carried out using a general linear model after passing the homogeneity test for variance. The effects of biochar, the rhizosphere, and plant phenology are summarized in Table S3, as well as their interaction effects on physicochemical properties.

Both rhizosphere and bulk soils from the 3 sampling stages were prepared for DNA extraction. Soil nucleic acid extraction was performed on 0.60-g soil using the FastDNA spin kit for soil (MP Biomedicals, OH) following the manufacturer’s protocol. DNA quality was measured by gel electrophoresis, and DNA concentration was measured by Nanodrop spectrophotometer (NanoDrop Technologies, Wilmington, DE).

The V4–V5 regions of 16S rRNA gene of soil samples at 3 sampling stages were analyzed by Hiseq sequencing with universal primers 515F (5′-GTG CCA GCM GCC GCG GTA A-3′) and 907R (5′-CCG TCA ATT CCT TTG AGT TT-3′). DNA amplification was conducted by the Beijing Genomics Institute, China. Each sample was amplified in triplicate, and PCR products were then purified using MinElute PCR Purification Kit (Qiagen, Germany) and sequenced with a Hiseq PE250 sequencing platform (Illumina, San Diego, CA).

Low-quality reads were first removed from the raw data with 2 paired-end reads. The sequences were quality-filtered by setting the maximum expected error threshold to 1.0. Thereafter, the paired-end reads were merged to tags by FLASH (Magoč and Salzberg, 2011). Sequences were analyzed using the QIIME software package (Caporaso et al., 2010). Reads were filtered by QIIME quality filters. Sequences with ≥97% similarity were assigned to the same operational taxonomic units (OTUs). We selected representative sequences for each OTU and use the RDP classifier (Wang et al., 2007) to annotate taxonomic information for each representative sequence based on SILVA-132 database (v 2020.02; Caporaso et al., 2010). The chimera check was performed during the process. All sequences have been deposited in the National Center for Biotechnology Information Sequence Read Archive with accession number SRP161910 (Yu et al., 2020b).

The annotated OTU table was further refined, and alpha diversity analysis was performed for samples both with and without biochar in 3 stages of Years 1 and 2. Alpha diversity of species was illustrated by the Chao1 estimator indicating community richness and the Shannon diversity index indicating community diversity. Following a one-way ANOVA, we used the least significant difference test to identify significant differences between the group with and without biochar in both rhizosphere and bulk soils over 3 plant growth stages (last 2 years), obtaining the degree of freedom (df). We estimated degrees of freedom using the following equation: df = Sample size (n) – Constrain factor (k). Since we compared the group with biochar against the group without biochar, n = 2, while k = 1; therefore, df = 1.

The Chao1 estimator was used to estimate the total number of species in ecological studies and was first proposed by Chao (1984): $Schao 1=Sobs+n1(n1−1)2(n2+1)$⁠, where S_chao1 is the estimated richness, S_obs is the observed number of species, n₁ is the number of OTUs with only 1 sequence (i.e., singletons), and n₂ is the number of OTUs with only 2 sequences (i.e., doubletons; Chao, 1984).

The Shannon diversity index was used to estimate the microbial diversity by analyzing 1,797 OTUs in each soil sample. The calculation formula used in this analysis is as follows: $Hshannon=−∑i=1SobsniN ln niN$⁠, where S_obs is the number of OTUs actually measured, n_i is the number of OTUs containing i sequence, and N is the total number of individuals in the community.

Bray–Curtis distance was calculated in R software using the Vegan package and described nonmetric multidimensional scaling (NMDS) to represent differences between groups (Oksanen et al., 2012). Permutational multivariate ANOVA (PERMANOVA) based on Bray–Curtis distance was performed to evaluate the effect of biochar treatment, rhizosphere effect, and plant growth stages on bacterial community structure. The bacterial relative abundance at phylum level was assessed with one-way ANOVA. LefSe analysis was carried out on the relative abundance of microorganisms with LDA score >10 in http://huttenhower.sph.harvard.edu/lefse/ (Segata et al., 2011), so as to find taxa with significant differences in abundance between groups with biochar and groups without biochar (namely biomarker). R package “DESeq2” was used to calculate the differential abundance (log2-fold change in relative abundance of each genus) for each sample with biochar as compared with samples without biochar (Love et al., 2014). Then we selected the responders with log2-fold change in relative abundance >1 and an adjusted P value of <0.1 for further community analysis. Redundancy analysis (RDA) of environmental factors and dominant phylums was conducted by using the “vegan” package in R (Oksanen et al., 2012), whose P value was annotated based on Monte Carlo substitution test. Pearson correlation analysis was calculated to illustrate the relationships between dominant phyla and environment factors.

Both the Chao1 estimators for bacterial community richness and the Shannon diversity index were used to assess alpha diversity. Overall, biochar had no significant effect on the Chao1 richness estimator at the beginning of Year 1 seedling stage in either rhizosphere or bulk soils (Figure 1A and B). However, during Year 1 mature stage in bulk soils and during Year 2 mature stage in the rhizosphere, adding biochar increased the Chao1 richness estimator (P < 0.001, P = 0.037; Figure 1A and B). In contrast to richness, adding biochar decreased the Shannon diversity index during the Year 1 seedling stage in both rhizosphere and bulk soils (P < 0.001; Figure 1C and D), but the effect dissipated thereafter.

Actinobacteria, Chloroflexi, Proteobacteria, Acidobacteria, Firmicutes, and Bacteroidetes were the 6 most abundant phyla in all treatments regardless of biochar application (Figure S1). Among these phyla, the relative abundance of Chloroflexi and Actinobacteria decreased from Year 1 seedling stage to Year 2 mature stage (P < 0.05), while the relative abundance of Proteobacteria increased during this same period (P < 0.01; Figure S1).

Adding biochar affected the above 6 phyla from Year 1 seedling stage to Year 2 mature stage. Specifically, the abundance of Actinobacteria increased by 14%–39% and Proteobacteria by 2%–46% (P < 0.01). In contrast, adding biochar decreased the abundance of Firmicutes by 24%–75%, Chloroflexi by 6%–38%, and Acidobacteria by 13%–44% (P < 0.05; Figure S1).

We used PERMANOVA to assess variations in the bacterial community with respect to biochar treatment, effect of rhizosphere versus bulk soils, and plant growth stage, both individually and in combination (Table S4). Plant growth stage explained the largest portion of the variation in bacterial community (38.3%, P < 0.001), followed by biochar treatment (11.5%, P < 0.001). The loadings of treatments with biochar grouped separately from those without biochar in the NMDS plot, indicating biochar-shaped bacterial community structure (Figure 2). Moreover, bacterial community loadings in the rhizosphere grouped separately during different barley growth periods while those in bulk soils overlapped (Figure 2). These results were consistent with those of PERMANOVA, where we observed interaction effects of biochar treatment and plant growth stage (4.1%, P < 0.001) and rhizosphere effect and plant growth stage (5.8%, P < 0.001) on bacterial community structure (Table S4).

Based on the phyla sensitive to biochar amendment, we plotted the abundance of sensitive bacteria after a log2 conversion in Figure 3 (Love et al., 2014). During the first year, the abundance of responding genera in bulk soils decreased from 180 (seedling stage) to 121 genera (mature stage) but then increased to 348 genera during the mature stage of the second year. In contrast to bulk soils, the abundance of responding genera in the rhizosphere decreased from 292 to 169 during the second year.

LEfSe analysis showed that Actinobacteria, Proteobacteria, and Gemmatimonadetes were sensitive to adding biochar in the rhizosphere. In contrast, Actinobacteria, Acidobacteria, Firmicutes, Chloroflexi, and Proteobacteria were sensitive to adding biochar in bulk soils (Figures 4 and S2). Specifically, at the end of first year, the biochar-sensitive bacterial communities at the genus level in the rhizosphere included Pseudarthrobacter, Leifsonia, Sinomonas, Streptomyces, Nesterenkonia, Blastococcus, Gemmatimonas, and Arthrobacter. In bulk soils, biochar-sensitive genera included Pseudarthrobacter, Mizugakiibacter, Sphingomonas, Blastococcus, Gemmatimonas, Jatrophihabitans, and Bryobacter (Figure S2). During the Year 2 mature stage in the rhizosphere, the biochar-sensitive communities included genera or families belonging to Actinobacteria, Proteobacteria, and Bacteroidetes, including Kribbella, Pseudarthrobacter, Streptomyces, Terrabacter, Nocardia, Sphingomonas, Nitrosospira, and Flavisolibacter. In contrast, biochar-sensitive responders in bulk soils included Pseudarthrobacter, Jatrophihabitans, Blastococcus, Amycolatopsis, Mycobacterium, Pseudolabrys, Flavisolibacter, and Gemmatimonas belonging to Actinobacteria, Proteobacteria, Bacteroidetes, and Gemmatimonadetes (Figure 4).

RDA indicated that the relationship between environmental parameters and dominant phyla was similar in both rhizosphere and bulk soils. In particular, Actinobacteria were positively correlated with soil pH, EBC, and K; Proteobacteria were positively correlated with NO₃⁻; and Chloroflexi were positively correlated with Al³⁺ and DOC (Figure 5A and B). In the RDA map, samples without biochar were positively correlated with Acidobacteria, Firmicutes, and Chloroflexi, while samples with biochar were correlated with Gemmatimonadetes, Bacteroidetes, Actinobateria, and Proteobacteria (Figure 5A and B).

Pearson correlations indicated that, in the rhizosphere, soil pH-related indices including pH, EBC, and K were positively correlated with Actinobacteria and Chloroflexi but negatively correlated with Acidobacteria, Firmicutes, Proteobacteria, Thermomicrobia, and Bacteroidetes. Soil Al³⁺ was negatively correlated with Actinobacteria but positively correlated with Firmicutes and Acidobacteria. Nutrient-related indices (C_total and DOC) were positively correlated with Actinobacteria (Figure 6A). Like the rhizosphere, similar relationships between the above environmental factors (pH, EBC, K, Al³⁺, C_total, and DOC) and dominant phyla were found in bulk soils.

We used our study to understand (1) whether adding biochar would affect bacterial community structure in both rhizosphere and bulk soils and (2) whether environmental factors changing with biochar amendment would affect bacterial functional taxa. In general, we found that Proteobacteria and Actinobacteria were dominant phyla during the 2-year plant growth period studied, especially in the rhizosphere. Adding biochar not only increased nutrient-related indices (DOC, total N, NH₄⁺, NO₃⁻, K, Ca, and Mg) but also increased soil pH, with these changes correlating with changes in bacterial community structure. Overall, our measurements suggest that the effects of biochar on bacterial sensitive responders were relatively short-lived—mostly dissipating within 2 years postbiochar application—but that changes in soil fertility observed after amending soils with biochar were related to changes in bacterial community composition in response to combined biochar, plant phenology, and rhizosphere effects.

During the first year of barley growth, many genera belonging to the Actinobacteria responded positively to biochar in the rhizosphere relative to genera in bulk soils, including Rhodococcus, Blastococcus, Arthrobacter, Microbacterium, Sinomonas, Prauserella, Micrococcus, Actinoallomurus, and Actinomycetospora (Figure 4B). In fact, the relative abundance of Actinobacteria in the rhizosphere was larger than any other phyla in pots with biochar (Figures 3, S1, and S2), which may have implications for plant growth. For example, Actinobacteria could promote plant growth via nitrogen fixation, solubilization of mineral nutrients, and production of siderophores and plant growth hormones (Jog et al., 2014; Yadav et al., 2018; Yu et al., 2020a). Moreover, Actinobacteria may indirectly promote plant growth via production of antagonistic compounds that help protect plants against pathogens (Yadav et al., 2018), overall suggesting that an increase in the relative abundance of Actinobacteria after adding biochar may benefit plants.

Proteobacteria gradually became most abundant in both rhizosphere and bulk soils from the first-year mature stage to the second year. Given their high species richness and genetic diversity (Figure S2), it is also possible they may benefit plant growth since they participate in several biogeochemical processes including nitrification, nitrogen fixation, and sulfate reduction. For example, Rhizobacteria, mostly belonging to the α-Proteobacteria class, often settle around plant roots, helping plants acquire soil nutrients (Jorquera et al., 2014). Proteobacteria have also been previously linked with increasing soil nutrient availability and crop yields (Janssen, 2006; Yu et al., 2020b; Mickan et al., 2022). Consistent with these studies, Proteobacteria abundance increased significantly over time in both rhizosphere and bulk soils (Figure S1), perhaps explaining why we observed a positive correlation between NO₃⁻ and Proteobacteria abundance (Figures 5 and 6).

Despite the potential positive effects of biochar amendments on plants, the abundance of biochar-sensitive responders in the rhizosphere significantly decreased during the Year 2 mature stage, though it increased for bulk soils (Figure 3). It is possible that by Year 2, the rhizosphere bacterial community structure may have become less dependent on biochar inputs, whereas those in bulk soils began to see the effects of adding biochar. As our data suggest, the Shannon index decreased in all treatments during the plant seedling stage in Year 1 (Figure 1), but there was little difference between treatments with or without biochar relative to changes in diversity over time, suggesting plant characteristics during different growth stages likely played a more important role on bacterial community diversity than adding biochar (Figure 1C and D). Nevertheless, previous studies found that adding biochar stimulates root growth and elongation (Wardle et al., 1998; Yamato and Iwase, 2005), producing more root exudates (Yu et al., 2018; Liao et al., 2021) that may become more important to soil bacterial community structure than direct biochar inputs (Hütsch et al., 2002; Paterson et al., 2007).

Although the community diversity decreased by the end of Year 2, sensitive responders belonging to Actinobacteria and Proteobacteria were still active in both rhizosphere and bulk soils. Nitrosospira (Proteobacteria phylum) in the rhizosphere was still more sensitive to biochar than in bulk soils (Figure 4), likely owing to active nitrification near roots as we and others have found (Taylor and Bottomley, 2006; Yu et al., 2020b). Sphingomonas (belonging to the Proteobacteria) were also sensitive responders in the rhizosphere (Figure 4C), which have been linked with degradation of aromatic compounds originating from both root exudates and biochar (Song et al., 2021; Lin et al., 2022). Altogether, the succession of sensitive responders was affected by biochar amendments with implications for site fertility.

Soil bacterial abundance was affected by soil pH-related indices (pH, EBC, K, and Al³⁺; P < 0.01; Figure 5). In particular, adding biochar increased soil pH, which can affect the relative abundance of bacterial communities (Luo et al., 2018; Yu et al., 2020b). Indeed, the observed increase in pH after adding biochar correlated with an increase in the abundance of Actinobacteria (Figure S1), suggesting they are favored at high pH. This is further supported by our RDA map and correlation analysis, which identified tight relationships between soil pH and Actinobacteria (Figures 5 and 6). In contrast to Actinobacteria, biochar-induced increases in pH lowered the abundance of Acidobacteria in both rhizosphere and bulk soil; Acidobacteria are tolerant of acidic soils and were positively correlated with soil Al³⁺ but negatively correlated with EBC and pH (Figure 5). These relationships suggest that biochar-induced increases in soil pH were an effective means for altering bacterial community changes in both rhizosphere and bulk soils.

Besides soil pH-related indices, nutrient-related indices (e.g., C_total, DOC, N_total, NH₄⁺, and NO₃⁻) also affected bacterial community structure. Our RDA results identified a positive correlation between Choroflexi and DOC in the rhizosphere (Figure 5), suggesting Choroflexi might favor environments rich in available C or are involved in processing C (Gupta, 2013). For example, Ktedonobacteria, a class belonging to Choroflexi (Figure 4), is considered as green bacteria capable of photosynthesis, which could take part in energy transfer and electron transport between biochar and soil (Oh-oka et al., 2021). Our previous study indicated the increase in available C was related to root growth (Yu et al., 2020b), which could have influenced bacterial metabolism and increased bacterial abundance in the rhizosphere. However, the DOC index decreased in the Year 2 mature stage in this study (Table 1), which could be ascribed to bigger bacterial abundance utilizing available carbon source and assimilating to microbial biomass, resulting in lower DOC concentrations in the soil.

Soil nutrient availability is thought to govern the activities of functional microbial community members (Smith et al., 2002). In this sense, microbes with a high nutrient demand are more likely to grow better in a postbiochar environment as we observed with Bacteroidetes—we found a positive correlation between Proteobacteria and NO₃⁻ availability (P < 0.001; Figures 5 and 6). Besides nutrients, increased availability in polyaromatic compounds from biochar may also influence the activity of different microbial taxa in charge of decomposing complex C, such as Actinobateria and Gemmatimonadetes (Thies and Rillig, 2012; Xu et al., 2023). In support of this understanding, both Actinobateria and Gemmatimonadetes responded positively to biochar inputs in the rhizosphere (Figures 5 and S2), with Gemmatimonadetes being one of the fastest phyla responding to changing soil nutrients induced by agricultural practices and biochar amendments (Whitman et al., 2016; Cesarano et al., 2017), and our RDA results showed that Actinobateria were positively correlated with C_total (P < 0.01; Figure 5A). Our observations suggest adding biochar can induce changes in soil nutrients and C availability that ultimately shape bacterial community structure.

During our 2-year plant growth experiment, observed variations in bacterial community structure resulted from interactions among biochar, rhizosphere, and plant phenology effects. In general, biochar inputs favored pH-sensitive phyla like Actinobacteria in the short term (i.e., 1 month after adding biochar), whereas in the long term (>520 days after adding biochar), a postbiochar environment favored Proteobacteria over other phyla, likely due to increased access to C and nutrients. Overall, a single application of 3% biochar was enough to support 2 consecutive years of barley growth without supplemental inputs of biochar. In addition to soil fertility indicators (e.g., pH, EBC, C_total, DOC, N_total), the observed bacterial responses over plant growth stages suggest that biochar can affect soil fertility directly, as well as indirectly via changes in bacterial community structure, which could help inform the application of biochar as a soil conditioner in agroecosystems.

All data generated or analyzed during this study are included in this article or listed in the supplemental material. DNA sequences are publicly available through National Center for Biotechnology Information Sequence Read Archive with accession number SRP161910.

The supplemental files for this article can be found as follows:

Figures S1 and S2. Tables S1–S4. Docx

The authors appreciate Dr Zhao Haochun for graph processing.

This work was financially supported by the National Natural Science Foundation of China (42007126), National Key R&D Program of China (2019YFC1803702), and Beijing Natural Science Foundation (8232019).

PMH is an associate editor at Elementa but did not have a role in the reviewing or handling of this manuscript. The authors declare no other competing interests.

Contributed to funding acquisition, project administration, supervision, conceptualization, original draft, data analysis, and review and editing: LY.

Contributed to investigation, validation, visualization, and review and editing: PMH.

Contributed to methodology, software, and visualization: LL.

Contributed to review and editing: HG.

How to cite this article: Yu, L, Homyak, PM, Li, L, Gu, H. 2023. Succession of bacterial community structure in response to a one-time application of biochar in barley rhizosphere and bulk soils. Elementa: Science of the Anthropocene 11(1). DOI: https://doi.org/10.1525/elementa.2022.00101

Domain Editor-in-Chief: Steven Allison, University of California Irvine, Irvine, CA, USA

Associate Editor: Stephanie Kivlin, University of Tennessee Knoxville, Knoxville, TN, USA

Knowledge Domain: Ecology and Earth Systems