2. Materials and methods

2.1 Study area

The study was conducted at the Longyandong Forest Farm (113°21’ E–113°27’ E and 23°10’ N–23°18’ N) in southern China. The study area has a tropical monsoon climate that is characteristic of the area. Throughout the year, the average temperature is 21 °C and the relative humidity is 80%. There are approximately 1, 900 mm of precipitation per year in the area, during the wet season (April–September), the region receives approximately 80% of rainfall; while, in the dry season (October–March), it receives approximately 20% (Liu et al., 2013). The region has experienced considerable N deposition (34 kg N ha−1 yr−1) due to rapid urbanization since 1978 (Huang et al., 2015). The soil type is lateritic red soil and is mainly composed of granite and sand shale.
C. hystrix plantations were planted in 1986, 1990, 1995, 2005, 2010, and 2014 after clear cutting. All plantations created in different years were subjected to the same artificial cultivation measures, including fertilization and the removal of understory vegetation, in the third and fifth years, followed by the cessation of artificial disturbance. All lands had similar geological and land use history before planting; all lands were planted with Acacia mangium using the same methods beginning in the 1970s, before which all lands were covered with evergreen broad-leaved forest, dominated by C. hystrix , Castanopsis chinensis , Syzygium rehderianum , andSchima superba (Zhou et al., 2013). The main shrub species wereFicus hirta , Psychotria asiatica , and Melicope pteleifolia , and the main herb species were Lophatherum gracile ,Woodwardia japonica , and Blechnum orientale .

2.2 Plot design

After carefully selecting and replicating plots, a chronosequence approach was used to survey forest stands (Walker et al., 2010). In August 2020, a sample survey was conducted in six stands of different ages (6, 10, 15, 25, 30, and 34 years old). Three plots (20 m × 20 m) were randomly set up in each stand (a total of 18 plots; three plots × six stand ages). A distance of at least 20 m was maintained between each plot and the forest edge, and the slope positions and aspects of all selected plots were similar. Plots between different stand ages were spaced at least 1 km apart to minimize spatial autocorrelation. The data of all stand ages were obtained from the records of the Longyandong Forest Farm (Table 1).

2.3 Field measurements

In August 2020, mature leaves and litter (1 m × 1 m on the ground) were collected at each stand age of the C. hystrix plantation (three replicates × six stand ages = 18 samples). All live trees were measured simultaneously to determine the diameter at breast height (DBH, cm; ≥ 5 cm) and tree height (m). The species richness and stand density (tree ha−1) were calculated by summing the stand basal area (m2 ha−1) for each species at the plot level (Feng et al., 2017; Yang et al., 2021).
Seven subsamples were collected from each of the three soil layers (0–10, 10–20, and 20–40 cm) using a 4-cm diameter auger, and the three subsamples from each soil layer were combined into a single composite sample. The soil samples were quickly refrigerated using ice packs and handheld storage boxes so they could be tested in the laboratory within 72 hours.

2.4 Aboveground productivity of C. hystrix plantations

The tree and understory biomass were measured in each plot of theC. hystrix plantation in August 2020. The tree biomass was estimated using allometric equations that correlated biomass with the tree height and DBH (Zhou et al., 2018). All understory vegetation (shrubs and herbs) was harvested from three sub-quadrats (2 m × 2 m) randomly located in each plot. Litter samples were collected from three sub-quadrats (1 m × 1 m) located within each plot. Then, all plant samples were dried at 65°C to obtain a constant weight for determining the biomass. The annual aboveground productivity of the C. hystrix plantation was calculated as the aboveground (tree + understory + litter) biomass/stand age.

2.5 Laboratory assessments

A minimum of 72 hours were required for the oven drying of the leaf and litter samples. The Walkley–Black wet digestion method as described by Nelson and Sommers (1982) was used to determine the concentrations of C in leaf and litter samples (g kg−1). The Kjeldahl method (Bremner and Mulvaney, 1982) was used to determine the N concentrations (g kg−1) in leaf and litter samples. The leaf and litter P concentrations (g kg−1) were measured using a photometer after digesting the samples with H2SO4–H2O2.
An ultraviolet spectrophotometer was used to determine the concentrations (mg kg−1) of NO3–N and NH4+–N in soil that had been extracted with KCl solution at 1 M. A solution containing 0.03 M NH4F and 0.025 M HCl was used to extract the soil available P concentration (mg kg−1) (the ratio between soil and extractant was 1:7). Then, an ultraviolet spectrophotometer was used to measure the results. Dichromate oxidation and titration with ferrous ammonium sulfate as described by Nelson and Sommers (1982) were used to determine the soil organic C (SOC, g kg−1).
The activities of soil enzymes involved in the cycling of C, N, and P were measured, as well as four hydrolytic enzymes, namely cellobiohydrolase (CBH), β -glucosidase (BG), AP, and NAG, and two oxidases, namely peroxidase and PhOx. The methodology described by Lie et al. (2019) was used to measure the BG, CBH, NAG, and AP activities. Hydrolytic enzymes were measured colorimetrically using a Multiskan EX (Thermo Scientific, Waltham, MA, USA) at 405 nm, and oxidase enzymes were measured at 450 nm (Tabatabai, 1994). To calculate the specific enzyme activities, this study followed the methodology described by Trasar-Cepeda et al. (2007) and divided the enzyme activities by the soil microbial biomass carbon (MBC).
The fumigation–extraction method was used to analyze the soil MBC, MBN, and MBP (Vance et al., 1987). The organic C and N of moist soil were extracted using a solution containing 0.5 M K2SO4, while P was extracted with 0.03 M NH4F and 0.025 M HCl. Following fumigation with chloroform for 24 hours at 25°C, the same extraction methods were conducted. A Total Organic Carbon Analyzer (TOC-VCSH, Shimadzu, Japan) was used to determine the extracted C and N in K2SO4, while the extracted P was determined using inductively coupled plasma optical emission spectroscopy. Based on the differences between the fumigated and unfumigated subsamples multiplied by their conversion factors, the microbial biomass values were calculated (Jenkinson et al., 2004).

2.6 Statistical analysis

To understand the mechanisms associated with the indirect and direct effects of altered leaf, litter, and soil nutrients triggered by stand ages, linear regression was used to examine the associations between stand ages and leaf, litter, soil characteristics, and microbial parameters. The relationships between variables (leaf, litter, and soil C:N:P concentrations and ratios) were represented by Pearson correlation coefficients. To determine the effects of stand ages, soil depth, and their interactions on soil characteristics and microbial parameters, two-way analyses of variance (ANOVAs) were used. All analyses were performed in R 4.0.4. and SPSS 20.0. The level of statistical significance was set at 0.05.