Biochar regulates the functions of keystone taxa to reduce p-coumaric acid accumulation in soil

Abstract

Introduction: Applying biochar (BC) to reduce toxic substance accumulation in soil, either through direct adsorption or modulation of the microbial community, has received considerable attention. However, a knowledge gap exists regarding how BC regulates microbial community structure and functions to mitigate toxic substance accumulation.

Methods: We previously identified p-coumaric acid (p-CA) as a representative autotoxin in tobacco rhizosphere soil. On this basis, this study simulated a soil environment with p-CA accumulation to investigate the impacts of BC on p-CA, soil physicochemical properties, and microbial community structure and function.

Results: The results showed that p-CA could be directly adsorbed onto BC, which followed the pseudo-second-order kinetic model (R ² = 0.996). A pot experiment revealed that BC significantly reduced soil p-CA, altered soil microbial composition, and enhanced bacterial community diversity. A weighted correlation network analysis showed a close association between taxon 1 in the microbial network and p-CA, suggesting a pivotal role for this taxon in reducing p-CA, with Devosia and Nocardioides identified as potential key contributors to this process. The prediction of possible keystone taxa functions showed that BC increased the relative abundances of aromatic compound degraders. Mantel tests indicated that soil organic matter exerted the greatest influence on keystone taxa functions and hub genera.

Discussion: These findings suggest that BC may either directly chemisorb p-CA or indirectly facilitate p-CA degradation by regulating the functioning of keystone taxa. The results of this study provide a novel perspective for further investigation of the mechanisms through which BC reduces the accumulation of toxic substances in soil.

Keywords: adsorption characteristics; biochar; keystone taxa; microbial function; p-coumaric acid.

Figure 1

Direct adsorption of p-coumaric acid by biochar. (A) shows the regression analysis of biochar adsorption kinetics for p-CA using the pseudo-second-order kinetic model (PKE), Elovich model (EM), and intra-particle diffusion model (IDM). (B) illustrates the isothermal adsorption process of p-CA onto biochar, fitted using the Langmuir and Freundlich equations.

Figure 2

Changes in soil physicochemical properties. NB, treatment without biochar; AB, treatment with biochar; p-coumaric acid, p-coumaric acid content in soil; AN, available nitrogen in soil; SOM, soil organic matter; * denotes significant differences according to the t-test (p < 0.05), and **** denotes highly significant differences according to the t-test (p < 0.0001). The meaning of these symbols is the same for the following figures.

Figure 3

Effects of BC on soil microbial diversity. NB, treatment without biochar; AB, treatment with biochar. Panels (A), (B), and (C) illustrate the α diversity indices. Panel (D) shows the β diversity analysis and Adonis test; the r value ranging from −1 to 1 indicates within- and between-group divergence, with an r value close to 1 indicating a larger difference between the groups compared to the difference within the groups; and the p value indicates the significance of difference. Panel (E) shows the stacked plot of the relative abundances of the top 10 genera in the NB and AB treatments.

Figure 4

Identification of keystone taxa and differences in their functions. NB, treatment without biochar; AB, treatment with biochar; WGCNA, weighted gene co-expression network analysis; OTU, operational taxonomic unit. Panel (A) shows the results of WGCNA. Microbial sequencing data were filtered, and OTUs with zero tag counts were removed. Scale preprocessing was performed for data standardization. A hierarchical clustering tree was constructed without the outlier sample AB3. The soft threshold (power) was determined to be 6 based on the scale-free topology fit index >0.9 and mean connectivity >10. The network and taxa were established using a mergeCutHeight of 0.25. Panel (A)(a) shows the correlation network constructed using the weights calculated using WGCNA. Panel (A)(b) shows the correlations between taxa and p-CA. Panel (A)(c) shows the relationships between genera in taxon 1 and p-CA. Hubs are genera showing high correlations with other genera in the taxa and with p-CA. Taxon hubs are genera showing high correlations with other genera in the group. Connectors are genera showing high correlations with p-CA. Peripherals are peripheral genera showing low correlations with both p-CA and other genera in the group. Panel (B) shows the effects of NB and AB on keystone taxa (taxon 1). Panel (B)(a) shows the correlation network of taxa 1 and other groups. Spearman’s pairwise correlations with p > 0.9 were selected for network construction. Panel (B)(b) shows the different functions of taxon 1 in NB and AB. Panel (B)(c) shows the relative abundances of aromatic compound degradation bacteria in NB and AB.

Figure 5

Correlations between environmental factors and microorganisms. The bottom left section shows pairwise comparisons of environmental factors, with the color gradient representing the Spearman’s correlation coefficient. The top right section displays the correlations of the microbial α-diversity (Sobs, Shannon, Simpson, Chao, ACE, Goods_coverage, Pielou, PD), key species (Devosia and Nocardioides), and the abundance of aromatic compound degradation functions with each environmental factor based on Mantel’s test. The edge width corresponds to the Mantel’s r statistic for distance correlation, and the edge color indicates statistical significance based on 1,000 permutations.

Figure 6

Structural equation modeling (SEM). Path coefficients were estimated using 1,000 bootstraps. Positive and negative effects are indicated by green and orange arrows, respectively; *p < 0.05, **p < 0.01, and ***p < 0.001. The model was evaluated using the goodness of fit (GOF) value. Biochar application was quantified using a binary coding method, with “0” indicating no application and “1” indicating application. Soil properties include TN, TP, TK, SOM, AP, AK, pH, and AN. Microbial diversity metrics include sobs, Shannon, Simpson, Chao, ACE, goods_coverage, Pielou, and PD. Hub genera refer to the absolute abundance of Devosia and Nocardioides. ACD function represents the relative abundance of aromatic compound degradation in taxon 1. p-CA indicates the residual p-CA content after 30 days of soil incubation.