Site description
This study was conducted in a P. massoniana plantation in southern Guiyang, Guizhou Province, China (106°398″E, 26°2815″N, alt.1145 m). The study area has a subtropical monsoon climate, and the main landforms are mountainous and hilly. The mean annual temperature, mean annual sunshine duration, relative humidity, and annual average rainfall are 14.9°C, 1354 h, 75.5%, and 1178.3 mm, respectively. At the time of study, the P. massoniana plantation was approximately 35-year-old and no artificial logging took place during the entire growth period. The canopy density was 0.8 with a tree density of 853 stem ·hm− 2. The average tree diameter and height were 32.2 cm and 28.32 m. The dominant understory species were Ligustrum lucidum Ait., Viburnum dilatatum Thunb., Rhapis excelsa (Thunb.) Henry ex Rehd., and Hedera nepalensis K. Koch var. sinensis (Tobl.) Rehd. According to the Chinese classification system, the soil type is yellow soil and was developed from red clay in the quaternary. The soil properties for 2021 are summarized in Table 1.
Table 1
Soil properties of P. massoniana plantation in 2021.LR, litter removal; CT, control; LA, litter addition. Different lowercase letters indicate significant differences among the litter treatments (p < 0.05).
treatment
|
The annual mean Moisture of 5cm soil depth (%)
|
The annual mean Temp of 5 cm soil depth (℃)
|
pH
|
Soil organic carbon (g·kg-1)
|
Total N
concentration
(g·kg-1)
|
C:N
|
Microbial biomass carbon (MBC)
(mg·kg-1)
|
Microbial biomass nitrogen (MBN)
(mg·kg-1)
|
MBC:MBN
|
LR
|
29.07±
0.15a
|
14.97±
0.05a
|
4.86±
0.04a
|
19.07±
1.58a
|
2.14±
0.28a
|
9.90±
0.89a
|
381.53±
41.84a
|
54.06±
5.49a
|
7.30±
0.89a
|
CT
|
28.45±
0.12a
|
15.24±
0.06a
|
4.81±
0.04a
|
18.75±
2.01a
|
2.15±
0.09a
|
8.87±
0.90a
|
316.41±
54.25a
|
54.19±
14.80a
|
6.54±
0.97a
|
LA
|
29.79±
0.13a
|
15.26±
0.06a
|
4.95±
0.11a
|
21.73±
0.29a
|
2.09±
0.28a
|
10.81±
1.31a
|
344.85±
41.02a
|
47.65±
5.75a
|
7.36±
0.69a
|
Experimental design
Our experiment began in October 2018 using randomized block design. Three similar blocks in the P. massoniana plantation were established and three litter treatments: LR, CT, and LA, were randomly arranged in each block. Litterfall on the forest floor in LR plots was removed monthly by hand and transferred to LA plots. The LR and LA plots were paired in space. Each plot was 10 × 10 m, was enclosed with a rope, and had a 10 m buffer zone.
Litter sampling
In May 2020, 17 months after the plots were established, three litter traps were randomly placed in each plot to monitor litterfall. Each trap was placed 0.2 m aboveground level and had a collection area of 0.25 m2 (0.5 m × 0.5 m). Litter was then collected monthly from June 2020 to May 2021. When large leaves lay across the trap, the portion of the leaf lying within the frame area was collected and the remainder was discarded. The collected litter was dried to a constant weight at 65 ℃ and weighed. The litter was then ground with a plant crusher (FW100, Taisite, China) and the C content was measured using the potassium dichromate external heating method. One year was divided into four seasons according to the temperature and precipitation condition in this region: (1) spring (March to May); (2) summer (June to August); (3) autumn (September to November); (4) winter (December to February). Litter C input is calculated as:

where Li is litterfall (g·m− 2) and Ci is litter C content (g·kg− 1).
Fine root sampling
Fine root (< 2 mm diameter) were collected using a 5-cm diameter soil core sampler in July, October, and December 2020 and April 2021, representing summer, autumn, winter, and spring, respectively. Three 40 cm deep soil cores were randomly collected in each plot. Roots were removed from the cores with tweezers and washed with deionized water. Roots with a diameter < 2 mm were classed as fine root and were dried to constant weight at 65 ℃ and weighed.
Fine root biomass (FRB) is calculated as:

where d is the diameter of the soil core sampler (5 cm).
Annual fine root mortality (FRM) is calculated according to the calculation methods of Vogt (Vogt et al. 1998)and Mei (Mei, 206):

where FRB is the annual average fine root biomass, T is the annual turnover rate of fine root (T = 1.05 was used in this study according to the turnover rate in P. massoniana plantation (Wang et al. 2012)).
Fine root C input is calculated as:

Root exudation measurements
Two target trees with similar growth conditions in each plot were randomly selected for root exudates collection. Root exudates were collected using an in-situ collection device (Phillips et al. 2008) in July, October, December and 2020, and April 2021, representing summer, autumn, winter, and spring, respectively. Three terminal fine roots (2 mm average diameter with laterals) in every target tree were excavated from the topsoil (0–10 cm). The excavated roots were washed thrice with deionized water and twice with a nutrient solution (0.2 mM K2SO4, 0.3 mM CaCl2⋅2H2O, 0.1 mM KH2PO4, and 0.2 mM MgSO4⋅7H2O). The clean roots were placed into 30 mL syringe with sterile 1-mm-diameter glass beads to make up the volume. Glass wool was placed at the bottom to prevent the glass beads from clogging syringe barrel. 15 ml of nutrient solution was injected into the syringe to meet the growth and activity demands of the fine root. The syringe was then covered with aluminum foil and placed in the original soil environment. After a 24-h equilibration period, the solutions in the syringe were flushed using a vacuum pump to remove any soluble C. Then, 15 ml of nutrient solution was injected to continue cultivation as a ‘trap solution’ for the root exudates. After a 24-h incubation period, the trap solutions containing exudates were collected by a vacuum pump. Root exudates were collected three times from each syringe via the same method and cultivation lasted for three days. Each plot also contained six syringes without roots as blanks. The collected solutions were mixed for three days and transferred to brown bottles. Then, they were filtered through a filter (0.22 µm) and stored in a refrigerator (-20°C) for preservation. The C in the root exudation was analyzed using a total organic carbon analyzer (Vario, Germany). For root exudation collection, root samples in the syringe were cut and scanned at 400 dpi using an Epson scanner (Seiko Epson Corporation, Japan), and the root surface area and length were analyzed using WinRHIZO Pro 2019a (Regents Instruments Inc., Quebec, Canada). Subsequently, the roots were oven-dried at 65°C and weighed. The root C exudation rate was calculated as the mass of C (µg) flushed from each root system (minus the average C concentration in the control cuvettes) over the 24 h incubation period. The exudation rates I (µg C g− 1 root biomass h− 1), II (µg C cm− 1 root length h− 1), and III (µg C cm− 2 root area h− 1) were calculated by dividing the total amount of C flushed from the root system by the total fine root biomass, root length, and root area, respectively.
Seasonal root exudate C input and annual exudate C input were calculated using the following equation based on mass-specific root exudation rates, fine root biomass, and the number of days in a season and year.
Seasonal root exudate C input = root C exudation rate per root biomass × root biomass × days
Annual root exudate C input = annual root C exudation rate per root biomass × root biomass × days
Statistical analyses
All data were tested for normal distribution and homogeneity of variance before further analyses. Two-way analysis of variance was applied to assess the effects of litter treatments, season, and their interaction on C input. Significant differences between treatments were examined using Tukey’s test when variances were equal, whereas Dunnett’s test was used for unequal variances. Differences were considered statistically significant at p < 0.05. Statistical analyses were performed using SPSS software v22.0 (Chicago, IL, USA) and diagrams were drawn using Origin 9.0.