Soil pH amendment alters the abundance, diversity, and composition of microbial communities in two contrasting agricultural soils

Abstract

Soil microorganisms are the most active participants in terrestrial ecosystems, and have key roles in biogeochemical cycles and ecosystem functions. Despite the extensive research on soil pH as a key predictor of microbial community and composition, a limitation of these studies lies in determining whether bacterial and/or fungal communities are directly or indirectly influenced by pH. We conducted a controlled laboratory experiment to investigate the effects of soil pH amendment (+/- 1-2 units) with six levels on soil microbial communities in two contrasting Chinese agricultural soils (pH 8.43 in Dezhou, located in the North China Plain, Shandong vs pH 6.17 in Wuxi, located in the Taihu Lake region, Jiangsu, east China). Results showed that the fungal diversity and composition were related to soil pH, but the effects were much lower than the effects of soil pH on bacterial community in two soils. The diversity and composition of bacterial communities were more closely associated with soil pH in Wuxi soils compared to Dezhou soils. The alpha diversity of bacterial communities peaked near in situ pH levels in both soils, displaying a quadratic fitting pattern. Redundancy analysis and variation partition analysis indicated that soil pH affected bacterial community and composition by directly imposing a physiological constraint on soil bacteria and indirectly altering soil characteristics (e.g., nutrient availability). The study also examined complete curves of taxa relative abundances at the phylum and family levels in response to soil pH, with most relationships conforming to a quadratic fitting pattern, indicating soil pH is a reliable predictor. Furthermore, soil pH amendment affected the transformation of nitrogen and the abundances of functional genes involved in the nitrogen cycle, and methane production and consumption. Overall, results from this study would enhance our comprehension of how soil microorganisms in contrasting farmlands will respond to soil pH changes, and would contribute to more effective soil management and conservation strategies.

Importance: This study delves into the impact of soil pH on microbial communities, investigating whether pH directly or indirectly influences bacterial and fungal communities. The research involved two contrasting soils subjected to a 1-2 pH unit amendment. Results indicate bacterial community composition was shaped by soil pH through physiological constraints and nutrient limitations. We found that most taxa relative abundances at the phylum and family levels responded to pH with a quadratic fitting pattern, indicating that soil pH is a reliable predictor. Additionally, soil pH was found to significantly influence the predicted abundance of functional genes involved in the nitrogen cycle as well as in methane production and consumption processes. These insights can contribute to develop more effective soil management and conservation strategies.

Keywords: agricultural soil; bacterial community; composition; diversity; fungal community; soil pH.

Fig 1

Spearman correlations between soil physicochemical properties at the end of the experiment and amended pH at the Dezhou (a) and Wuxi (b) sites. pH_A represents the amended soil pH value through the addition of acidic or alkaline solution before the start of the experiment, whereas pH_E represents the soil pH value at the end of the experiment. Color gradient and circle size denote Spearman’s correlation coefficients. The pH value was the negative logarithm transformation of the hydronium ion concentration. Asterisks denote different significance levels: *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig 2

Soil bacterial and fungal alpha diversity affected by site and treatment. The Shannon indices were affected by site (n = 18) or by treatment (n = 3) for bacterial (a or b) and fungal (c or d) community, respectively. Significant differences were based on the Kruskal–Wallis pairwise test. The treatments in the dotted box are significantly different from the treatment of the connecting line. Boxes represent the interquartile range (IQR, 25%–75% of the data). The median values are represented by the bar inside each box, and whiskers extend to values within 1.5 times the IQR. Outliers, data points lying beyond this range, are depicted as gray points. Asterisks denote different significance levels: *P < 0.05, **P < 0.01, and ***P < 0.001. D, Dezhou; H, H₂SO₄; Na, NaOH; and W, Wuxi.

Fig 3

Variation in microbial communities induced by pH manipulation and the impacts of soil physicochemical properties on microbial communities. PCoA plot based on weighted UniFrac distance for bacterial (a) and fungal (b) communities at the ASV level. The statistical significance of PERMANOVA, PERMDISP, and ANOSIM across all treatments was assessed via 999 permutations test. Relationships between the community compositional structures and soil physicochemical properties for bacteria (c) and fungi (d) using tb-RDA based on weighted UniFrac distance at the genus level. The percentage in parentheses represents the variation explained by each axis. The confidence ellipse represents a 95% confidence level. Individual impact of each soil physicochemical property to bacterial (e) and fungal (f) communities in the two sites or single site. pH_A represents the amended soil pH value of each treatment through the addition of acidic or alkaline solution before the experiment began. The pH value was the negative logarithm transformation of the hydronium ion concentration. D, Dezhou; H, H₂SO₄; Na, NaOH; and W, Wuxi.

Fig 4

Relative abundances of the dominant bacterial (a and b) and fungal (c and d) phyla in two agricultural soils. Thirteen and nine of the most abundant taxa are displayed for bacterial and fungal communities, respectively. D, Dezhou; H, H₂SO₄; Na, NaOH; and W, Wuxi.

Fig 6

Relative abundances of dominant bacterial and fungal phyla in relation to amended soil pH at the Dezhou (a) and Wuxi (b) sites. Lines represent the best-fit quadratic models to the data. The coefficients of determination (R²) are shown for each taxon with P values. Shadow represents a 95% confidence level. The pH value was the negative logarithm transformation of the hydronium ion concentration. Asterisks denote different significance levels: *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig 5

Correlations between the relative abundances of dominant bacterial phyla and amended soil pH at the Dezhou (a, b, and c) and Wuxi (d) sites. Lines represent the best-fit linear models to the data. Pearson correlations coefficients (R) are shown for each taxon with P values in yellow and blue at the Dezhou and Wuxi sites, respectively. Shadow represents a 95% confidence level. The pH value was the negative logarithm transformation of the hydronium ion concentration. Asterisks denote different significance levels: *P < 0.05, **P < 0.01, and ***P < 0.001.

Fig 7

Predicted abundances of functional genes that participate in the nitrogen cycle (a), methane production and consumption (b) by PICRUSt2 across treatments and sites. The gene abundances were standardized by Z-Score. Hierarchical clustering of the genes was used to create blocks in which the values are close. Enzymes encoded by these genes (a) perform the nitrogen transformations according to KEGG (Kyoto Encyclopedia of Genes and Genomes) Orthology: assimilatory nitrate reductase (NAS, nasAB); membrane-bound (NAR, narBGHIVYZ) and periplasmic (NAP, napAB) dissimilatory nitrate reductases; nitrite oxidoreductase (NXR, nxrAB); nitrite reductases (NIR, nirABDKS); dissimilatory periplasmic cytochrome c nitrite reductase (ccNIR, nrfAH); nitric oxide reductase (NOR, norBC); nitrous oxide reductase (NOS, nosZ); molybdenum-iron nitrogenases (MoFe, nifHDK); ammonia monooxygenase (AMO, amoCAB); and hydroxylamine oxidoreductase (HAO, hao). The purple font indicates the genes relevant with methane production, and the orange font indicates the genes relevant with methane consumption (b). Significant differences in the two sites were based on MaAsLin 2. Asterisks denote different significance levels: *P < 0.05, **P < 0.01, and ***P < 0.001. D, Dezhou; H, H₂SO₄; Na, NaOH; and W, Wuxi.