Study site
The study site was located in Guanzhuang National Forest Farm (117°43’E, 26°30’N), Sanming City, Fujian Province in southeastern China, at an altitude of approximately 200m. This area has a mid-subtropical monsoon climate with abundant rainfall, characterized by an average annual temperature of 20.1±1.96℃, precipitation of 2777±40.2 mm a-1 (>80% of which falls during May-October, Fig. S1), and frost-free period 271d (climatology based on measurements over 14 years from 2004 to 2017).
C.lanceolata plantation forest was planted in 1992 in an area of 6 hm2 at a density of 1660 individual trees per ha-1. In December 2003, twelve plots (20 m×20 m each and with a 15 m×15 m central area) were randomly selected within the forest,with the minimum distance between plots was 10 m. The average tree height of the whole was 12 m and the mean diameter at breast height (DBH, 1.3 m from the ground) was 16.1 cm. A survey conducted at this time revealed that understorey layer was sparse, with coverage between 3% and 5%, and dominant understorey species included Miscanthus floridulus, Dicranopteris olichotoma, and Pteridium aquilinum var. Latiusculum (Wu et al. 2013; Shen et al. 2019). We also measured soil physical and chemical properties at that time. Soil is acidic (pH=4.67), with an organic carbon content of 18.39 g kg-1 and a total N content of 0.79 g kg-1, and was classified as an Acrisol (Fan et al. 2007; Shen et al. 2019; Wu et al. 2021).
N fertilizer treatment
N fertilizer treatment in the twelve selected plots was initiated in January 2004 and continued for 14 years, until 2017. The N fertilizer treatments consisted of four levels,0, 60, 120, and 240 kg N hm-2a-1, referred to as N0 (control), N60 (low-N), N120 (medium-N), and N240 (high-N), respectively. Each treatment was applied to three replicate plots. According to the N concentration of each treatment, urea was dissolved in 20 L of water, and the solution was sprayed with a backpack sprayer on the forest floor in each plot every month.
Sampling
Sampling was conducted on December 15, 2017. According to the results of a per-wood inspection of all Chinese firs in plots in the previous period (2004-2017), we selected a standard tree (normal growth, no pests or diseases, and few scars) in each plot for wood test, thus, a total of twelve standard trees were selected for biomass measurements. We determined the orientation of the selected trees with a compass. Before felling, we marked the orientation of the stem at breast height. The central cross-sectional differentiation quadrature method was used to intercept the stem disc (5 cm thick) at the midpoint of each zone segment (every 2 m of the whole tree calculate initial from the coarse root of the plant). After felling, all leaves, branches, bark, and stem segments were taken from various point on the trees in the field and weighed. The root system biomass was measured following total root excavation. Root samples were divided into coarse roots (>10 mm) and fine roots (<2 mm, 2-5 mm, and 5-10 mm) and were cleaned with deionized water and weighed. Then all subsamples, 5 kg of each fresh weight, were transported to the laboratory.
Forest litter was sampled using a 1 m×1 m frame (three random samples per plot); the samples were separated into an undecomposed layer (L layer) and a semidecomposed layer (F layer), and the layers were then mixed into composite samples. We distinguished these two layers as follows: in the L layer, the original shape of the leaves is maintained, there is no significant change in colour and there is no superficial evidence of decomposition; by contrast, the leaves of the F layer do not have a complete outline, most of the litter has been crushed, and the mesophyll has decomposed into debris.
Soil samples were collected from three soil layers (0-20, 20-40, and 40-60 cm) using an auger (2.5 cm inner diameter). Subsamples of the soil cores (six cores randomly per plot) collected from the same layers in each plot were mixed into one composite soil sample. Soil bulk density was estimated using the core (5 cm inner diameter) method: the soil sample was oven-dried (105°C for 48 h) and bulk density was estimated as the mass of the oven-dried soil divided by its volume.
In order to capture the understorey diversity when more seasonal vegetation was visible, we investigated the diversity of understorey layer On September 16, 2017. Three 5 m×5 m survey subplots were randomly set in each plot, and the name, height and crown width of species as well as the number small trees, lianas, herbs and shrubs taller than 5 cm in each plot were recorded. The biomass of understorey layer was determined by harvesting; small trees and shrubs were collected from 2 m×2 m sampling areas, and herbs were collected from 1 m×1 m sampling area. Whole understorey plant were excavated and divided into above- and belowground components. The aboveground components consisted of leaves, branches and stems in the small tree, and leaves and branches in the herb layer, and the belowground components consist of roots. The fresh weight were measured, and transported to the laboratory to determine the water content and to calculate the dry biomass and N content of the understorey layer.
The stem disc samples taken at breast height were air-dried in ventilated conditions after which they were sanded with sandpaper until the annual rings were visible. The discs were then cross-dated to the tree core samples at each sample point. After that, the tree-ring width was determined using LinTab 6 (TSAP-Win, Germany).
Chemical analysis
The water content (%) of all the subsamples was determined in the laboratory. The soil samples were sieved through a 2-mm mesh screen to remove stones and roots and the soil water content was determined by drying at 105℃ to a constant mass. Fresh leaves were first washed with 10% dilute hydrochloric acid and then repeatedly washed with deionized water to completely remove dust and particulates adsorbed on the surface, then killed in an oven at 105℃ for 2 h, and then dried at 60℃ to a constant mass. The remaining samples were cleaned and then dried at 60℃ to over-dring. The dry biomass of each organ of the tree was calculated per unit area based on the water content.
All the subsamples were crushed with a microplant crusher, passed through a 100-mesh sieve, and kept in a dry environment prior to measurement. The total N and total C contents of the samples were analysed by an elemental analyser (PerkinElmer Corporation, 2400II, USA).
Calculation
(1) Basal area increment
We calculated the basal area increment (BAI, cm2) using the following equation:
where n is the number of tree rings and R is the radius of the tree (cm).
- Soil N storage
Soil N density (Nd, kg m-2) refers to the storage of N in the soil layer at a specific depth per unit area and was calculated in the following equations:
Ndi =0.1 × TNi × γ × Hi × (1-δ/100)
|
(2)
|
where 0.1 is the conversion factor, i represents the soil layer, cm; TNi represents the total soil N content, %; γ represents the soil bulk density, g cm-3; H represents the soil depth, cm; and δ indicates the percentage of gravel in soil with a diameter>2 mm, %.
Soil N storage (NS, t hm-2) was calculated as:
where A represents area, hm-2.
- Plant N uptake and recovery rate
REN was calculated by the difference method developed by Davis et al. 2003 and is defined as:
UN= Biomass × N %
|
(4)
|
REN= (UN-U0)/RN × 100%
|
(5)
|
where N (%) is the N content of major ecosystem components; biomass (t hm-2) is the plant biomass of a plot in this study; UN (kg N hm-2) is the total plant N uptake measured in above- plus belowground biomass in a plot that received N dose (RN) at i.e., 0, 60,120 or 240 kg N hm-2a-1; and U0 is the total plant N uptake in unfertilized N plots (N0).
Statistical analysis
We used two-way ANOVA to test the effects of N fertilizer treatments, tree components, and soil layers on biomass, N concentration, and N storage. One-way ANOVA with Dunnett’s post-hoc test was used to test the effects of the N fertilizer treatments on plant biomass, N content, and N uptake, understorey species, soil N content, and N storage in the soil layers. All analyses were conducted using SPSS software (version 19.0; SPSS Inc., Chicago, IL, USA). Figures were created with SigmaPlot 13.0 software (Sysat software, USA). Results were considered as significant at P < 0.05.