Soil microbial community and physicochemical properties together drive soil organic carbon in Cunninghamia lanceolata plantations of different stand ages

Study site and soil sample collection

Our study was conducted at Gaojingmiao Forest Farm (31°00′59″N, 119°12′08″E) in Langxi County, Xuancheng City, Anhui Province, China. Gaojingmiao Forest Farm is a state-owned forest farm, covering an area of 10.37 km², with large areas of Pinus massoniana, Pinus elliottii, and Cunninghamia lanceolata. This forest farm is located in the southeast of Anhui Province and has a humid subtropical monsoon climate with four distinct seasons and sufficient sunlight. The mean annual temperature in this area is 15.9 °C, and the coldest and warmest months are January (with a mean temperature of 2.7 °C) and July (with a mean temperature of 28.1 °C), respectively. The region receives an average total annual precipitation of 1,294.4 mm mostly occurring in summer. The area has 1784 h of sunshine and 228 days of Frost-free period per year. The soil is acidic yellow-red soil.

In December 2018, Cunninghamia lanceolata plantations of three stand ages (young plantation of 6 years, middle-aged plantation of 12 years, and mature plantation of 25 years) with basically the same site conditions (30–50 m above sea level, slope less than 30° ) were selected as the sample blocks. Three plots (10 m × 10 m) were randomly set in each of the stand ages of Cunninghamia lanceolata plantation. Five soil cores were randomly collected from the 0–10 cm soil layer in each plot using a corer six cm in diameter. These five soil cores were then combined and mixed thoroughly as one sample for each plot. The individual soil samples of each plot were collected in plastic bags and placed in a cooler in the field, then transported to the laboratory for further analysis.

After visible plant residues and stones were removed, each soil sample was divided into two subsamples. One subsample was used to determine pH, soil organic carbon (SOC), and total nitrogen (TN) content after air-dried; one subsample was stored at −20 °C and used to determine microbial community composition, ammonium nitrogen (NH₄⁺-N), nitrate nitrogen (NO₃⁻-N), and dissolved organic carbon (DOC) (Yuan et al., 2020). Cylinders were used to obtain four undisturbed soil cores from the 0–10 cm profile of each plot. One of these soil cores was used to determine soil bulk density (BD) and soil water content (SWC), while the other three of them were used to determine SOC physical fractions. The three soil cores used to determine SOC fractions were gently broken apart along the natural breakpoints and passed through a two mm sieve to remove visible organic debris and stones, and then air-dried for further analysis (Mitchell et al., 2018).

Soil physicochemical properties analysis

Both SOC and TN were measured using a C/N analyzer (Elementar, Vario Max CN, Germany) with a combustion temperature of 900 °C. Soil samples collected with cylinders were oven-dried to calculate BD and SWC according to the volume of the cylinder and the soil weight before and after drying. Soil pH was measured at a soil: water ratio of 1:2.5 using a pH meter (Mettler Toledo, Greifensee, Switzerland). Soil NH₄⁺-N and NO₃⁻N were extracted from 20 g of fresh soil with 1 mol L⁻¹ KCl (soil: extract, 1:5) and analyzed using a flow-injection autoanalyzer ((CFA)-AA3, SEAL, Germany). Soil DOC was extracted using deionized water (soil: extract = 1:5) and analyzed using a TOC analyzer (Vario TOC cube; Elementar, Hanau, Germany).

Wet sieving and density flotation methods were used to detect SOC fractions (Mitchell et al., 2018). First, soil samples were dispersed by low-energy sonication. Briefly, add 30 g of air-dried soil into 150 mL water and then put them into an ultrasonic treatment (KQ5200DE; Kunshan Ultrasound Instrument Co., Ltd., China) with the output energy of 22 J ml⁻¹. The macroaggregates were disrupted in this step and only the more stable microaggregates were left. Secondly, the dispersed suspension was then wet sieved over a 53 µm aperture sieve several times until the rinsing water was clear. Soil samples were separated into two parts in this step: the fraction >53 µm which was left on the sieve and the fraction <53 µm which passed through the sieve and remained in suspension. The fraction >53 µm was rinsed repeatedly with deionized water, dried in the oven at 40 °C, and weighed. Thirdly, density flotation was conducted to separate the fraction >53 µm from the second step into organic carbon protected within microaggregates (SA) and particulate organic matter (POM). The fraction >53 µm from the second step was stirred with sodium polytungstate (SPT) at a density of 1.8 g cm⁻³ and then centrifuged at 1,000 g for 15 min. The light part was POM, which could be decanted and then washed with deionized water to remove all SPT, dried at 40 °C, and weighed. The heavy part was also dried at 40 °C and weighed as SA. At last, the fraction <53 µm from the second step was filtered through a 0.45 µm aperture nylon mesh, and the material >0.45 µm was dried at 40 °C and weighed as mineral-associated organic carbon (SC).

Phospholipid fatty acid (PLFA) analysis

The PLFA contents of the samples were analyzed using the method described by Bååth & Anderson (2003). Briefly, lipids were extracted from freeze-dried soil (8 g) in a single-phase mixture of chloroform:methanol:phosphate buffer (1:2:0.5). After extraction, the lipids were separated into neutral lipids, glycolipids, and polar lipids (phospholipids) on a silicic acid column. The samples were analyzed using a Thermo ISQ gas chromatography-mass spectroscopy system (TRACE GC Ultra ISQ). The concentrations of the individual compounds were obtained by comparing the peaks with an internal standard (nonadecanoic acid methyl ester 19:0). The fatty-acid signatures 15:0, i15:0, a15:0, i16:0, 16:1 ω7c, 17:0, i17:0, cy17:0 and cy19:0 were used as bacterial biomarkers (B). The fatty acids 18:2 ω6 and 18:1 ω9c were used as fungal indicators (F) and 10Me16:0 and 10Me18:0 were used as indicators for the actinomycetes (ACT). The fatty acids i15:0, a15:0, i16:0, and i17:0 were used to represent Gram-positive bacteria (G⁺), while 16:1 ω7c, cy17:0, and cy19:0 were used to represent Gram-negative bacteria (G⁻). The total PLFAs (TPLFA) were the sum of bacterial, fungal, and actinomyces PLFA biomarkers.

Statistical analysis

A one-way ANOVA was performed to identify significant differences in SOC, soil physicochemical properties, and PLFAs among the three stand ages. A principal components analysis (PCA) was used to analyze soil microbial community structure. The Pearson correlation method was used to analyze the correlation between SOC and some important physicochemical properties and PLFAs. One-way ANOVA and Pearson correlation analysis was performed using SPSS 19.0 (SPSS Inc., Chicago, IL, USA). PCA was conducted with CANOCO for Windows 4.5. Other figures were generated using the Origin 9.5 package (Origin Lab Corporation, Northampton, MA, USA). P < 0.05 was considered to be statistically significant. Mean ±standard deviation was shown in tables and figures.

Variation partitioning analysis (VPA) was conducted to quantify the relative importance of soil microbial community and soil physicochemical properties to the variations of SOC among Cunninghamia lanceolata plantations of different stand ages. To avoid collinearity, variables with a large variance inflation factor (VIF) were removed in order until the VIF of the retained variables does not exceed 10. Next, principal component analysis was performed on microbial variables and physicochemical variables respectively (the first two principal components were retained). Then VPA was employed to quantify their explanation for the variations of SOC among different stand ages. VPA was conducted in R 4.0.2 (R Core Team, 2020) and the vegan package (Oksanen et al., 2020).

Soil organic carbon content

Soil organic carbon content exhibited large variability among different stand ages (Fig. 1). The middle-aged plantation had the highest SOC (10.63 g kg⁻¹) followed by the mature plantation (7.83 g kg⁻¹), and the young plantation had the lowest SOC (5.33 g kg⁻¹). Soil organic carbon content of the middle-aged plantation was 99.4% and 35.8% higher than those of the young and the mature plantation, respectively. The difference in SOC between the middle-aged and the young plantations was significant.

Soil physicochemical properties

Soil physicochemical properties of the Cunninghamia lanceolata plantations varied among different stand ages (Table 1). The middle-aged plantation had the highest pH, SWC and TN. The mature plantation had a significantly higher NH₄⁺-N content than the other stand ages. Soil BD, C/N, NO₃⁻-N, and DOC showed no significant difference among the three stand ages.

Soil organic carbon fractions were different among Cunninghamia lanceolata plantations of different stand ages (Fig. 2). The proportion of SA was significantly higher in the soil of the middle-aged plantation (40.4%) than those of the young (29.2%) and the mature (27.8%) plantation. The other two fractions, SC and POM, showed no significant difference among the three stand ages.

Soil microbial community

Soil PLFAs showed significant differences among different stand ages (Fig. 3). The middle-aged plantation had significantly higher soil total PLFAs, bacterial PLFAs, fungal PLFAs, actinobacterial PLFAs, Gram-positive and Gram-negative bacterial PLFAs than the young plantation and the mature plantation (Fig. 3A). Soil total PLFAs of the middle-aged plantation were 30.23 nmol g⁻¹, while those of the young plantation and the mature plantation were 72.5% and 60.3% lower than the middle-aged plantation. The ratio of F/B and the ratio of G⁺/G⁻ showed no significant difference among stand ages (Fig. 3B). Principal components analysis (PCA) also confirmed that soil microbial community composition was different among the three stand ages and 94% of the variance was explained by the first component of the PCA (Fig. 4). The soil PLFAs of the middle-aged plantation were separated from those of the other two stand ages along with the first principal component.

Associations of SOC with the microbial community and physicochemical properties

The results of VPA showed that the soil microbial community and physicochemical properties explained 82.7% of the variance in SOC of different stand ages (Fig. 5). Pearson correlation analysis also showed that some of the microbial and physicochemical factors were significantly correlated with SOC (Table 2). Among the microbial factors, TPLFA, B, F, ACT, G⁺, and G⁻were positively correlated with SOC while F/B was negatively correlated with SOC. Among the physicochemical properties, TN, NO₃⁻N, and SA were positively correlated with SOC.