phoD-harboring bacterial community composition dominates organic P mineralization under long-term P fertilization in acid purple soil

Abstract

Introduction: A better understanding of the regulatory role of microorganisms on soil phosphorous (P) mobilization is critical for developing sustainable fertilization practices and reducing P resource scarcity. The phoD genes regulate soil organic P (Po) mobilization.

Methods: Based on the long-term P application experiments in acid purple soil of maize system in Southwest China (started in 2010), the experiment included five P levels: 0, 16, 33, 49, and 65.5 kg P hm^-2 (P0, P16, P33, P49, and P65.5, respectively). The molecular speciation of organic P in soil was determined by 31P-nuclear magnetic resonance (NMR), high-throughput sequencing technology, and real-time qPCR were used to analyze the bacterial community and abundance of phoD-harboring bacterial genes, exploring the bacterial community and abundance characteristics of phoD gene and its relationship with the forms of Po and alkaline phosphatase (ALP) activity in the soil.

Results: The results showed that the orthophosphate monoesters (OM) were the main Po speciation and varied by P fertilization in acid purple soil. ALP activity decreased as P fertilization increased. Co-occurrence network analysis identified the overall network under five P fertilizations. The keystone taxon base on the network showed that Collimonas, Roseateles, Mesorhizobium, and Cellulomonas positively correlated with both OM and Po. The random forest showed that Cellulomonas, Roseateles, and Rhodoplanes were the key predictors for ALP activity. The keystone taxon was a more important predictor than the dominant taxon for ALP, OM, and Po. The structural equation model (SEM) showed that soil organic matter (SOM), available P (AP), and OM were the main factors influencing the ALP by reshaping phoD-harboring bacteria alpha diversity, community composition, and phoD abundance.

Discussion: The phoD-harboring bacterial community composition especially the keystone taxon rather than alpha diversity and abundance dominated the ALP activity, which could promote P utilization over an intensive agroecosystem. These findings improve the understanding of how long-term gradient fertilization influences the community composition and function of P-solubilizing microorganisms in acid purple soil.

Keywords: alkaline phosphatase; co-occurrence network; keystone taxa; long-term P fertilization; organic P speciation; phoD-harboring bacteria.

FIGURE 1

The concentration of the P speciation under gradient P fertilization rates. (A) Solution ³¹P NMR (nuclear magnetic resonance) spectra on NaOH-EDTA extracts for the five P fertilization rates. Nuclear magnetic resonance can be used to identify the exact molecular forms of P in soils. These include phosphate (δ 5.83 ppm), orthophosphate monoester (OM) (δ 4.75 ppm), orthophosphate diester (OD) (δ –0.44 ppm), and others (δ –4.76 ppm). (B) Line charts based on ³¹P NMR. The data are based on the concentration of organophosphorus molecular form at five P fertilization rates. P0, P16, P33, P49, and P65.5 represent 0, 16, 33, 49, and 65.5 kg P ha^–1 applied, respectively.

FIGURE 2

phoD gene abundance (A) and alkaline phosphatase (ALP) (B) in soils treated with five phosphorus fertilization rates. Small letters indicate a significant difference (P < 0.05 after the Duncan test) between P levels. P0, P16, P33, P49, and P65.5 represent 0, 16, 33, 49, and 65.5 kg P ha^–1, respectively.

FIGURE 3

Alpha diversity (A,B) and redundancy analysis (RDA) (C) of phoD-harboring bacteria under gradient P fertilization. In (A,B), small letters indicate significant differences (P < 0.05 after the Duncan test) between P levels. In (C), RDA was used to explore the relationship between the entire phoD-harboring bacterial community and selected soil properties. **Correlation is significant at the 0.01 level. P0, P16, P33, P49, and P65.5 represent 0, 16, 33, 49, and 65.5 kg P ha^–1, respectively.

FIGURE 4

Network of phoD-harboring bacteria based on Spearman correlation analysis from OTU profiles. The network consists of five gradient phosphorus fertilization rates (including P0, P16, P33, P49, and P65.5). The color of the nodes represents the taxa of the genus classification. Red lines represent a positive correlation, and green lines indicate a negative correlation. Node, edge, positive rate, density, average clustering coefficient, and average path length of the topological parameters were shown below the network. P0, P16, P33, P49, and P65.5 represent 0, 16, 33, 49, and 65.5 kg P ha^–1, respectively.

FIGURE 5

Bar chart of relative abundance of the keystone taxa at genus level under different phosphorus fertilization rates. (A) Pseudomonas; (B) Collimonas; (C) Rhodoplanes; (D) Roseateles; (E) Ralstonia; (F) Mesorhizobium; (G) Cellulomonas; (H) Herbaspirillum; (I) Friedmanniella; (J) Verminephrobacter. The key species were selected by ranking the top 10 “Degree” values based on the network. P0, P16, P33, P49, and P65.5 represent 0, 16, 33, 49, and 65.5 kg P ha^–1, respectively.

FIGURE 6

Correlation between alkaline phosphatase activity and relative abundance of keystone taxa which include (A) Pseudomonas, (B) Collimonas, (C) Rhodoplanes, (D) Roseateles, (E) Ralstonia, (F) Mesorhizobium, (G) Cellulomonas, (H) Herbaspirillum, (I) Friedmanniella, and (J) Verminephrobacter.

FIGURE 7

Random Forest analysis to identify the main predictors of keystone taxa on (A) alkaline phosphatases (ALP), (B) organophosphorus (Po), and (C) orthophosphate monoester. The mean decrease in accuracy (%IncMSE) was used to indicate the relative importance of each variable for predicting the soil’s total P concentration. *Correlation is significant at the 0.05 level; **correlation is significant at the 0.01 level.

FIGURE 8

Correlation among alkaline phosphatases (ALP), orthophosphate monoester (OM), and organophosphorus (Po) with the keystone (A) and dominant taxon (B). Pairwise comparisons of the keystone and dominant taxon are separately shown in the heatmap with a color gradient to represent Spearman’s correlation coefficients. Statistically significant correlations among ALP, OM, and Po with the keystone and dominant taxon are shown in the corresponding lower left. Edge width corresponds to Mantel’s coefficients. The orange line means P < 0.01, the green line means 0.01 < P < 0.05, and the gray line means no significant effect. A random forest analysis to identify the importance of keystone taxon and dominant taxon to ALP (C), OM (D), and Po (E).

FIGURE 9

Structural equation model (SEM) (A) and bar plots of effect values (total, direct, indirect) on alkaline phosphatases (ALP) activity (B) showing the influence of various factors on the correlationship among soil properties [including soil organic matter (SOM), available P (AP), and orthophosphate monoesters (OM)], alpha diversity, phoD abundance, community structure, and ALP activity of soil under long-term gradient phosphorus fertilization rates. The width of the arrow indicates the strength of the causal effect. The red and blue arrows indicate the positive and negative relationships between the indicators. The number above the arrow indicates the path coefficient. “*”, “**”, and “*” represent significant path. The percentage above each indicator represents the R² value, which is the variance-explained ratio of each variable. The final model fits the data well. The model is: (χ² = 13.343, df = 1.026, CFI = 0.997, IFI = 0.997, RMSEA < 0.043, AIC = 59.343).