2.1 Study site
The study site was located at the Maoxian Mountain Ecosystem Research Station CERN (Chinese Ecosystem Research Network), Chinese Academy of Sciences, in the Hengduan mountainous region, Sichuan Province, China (31°41′ N, 103°53′ E, 1,800–2300 m). The geomorphology is mainly characterized by high mountains with deep valleys (Zhang 1992). In most cases the slope of the mountain is greater than 25°. The climate is classified as warm temperate, with a humid and rainy summer and a cold and dry winter. Annual precipitation average is 760.8 mm, 65% of which occurring during the rainy season from June to September. Annual mean temperature is 9.8°C, with the maximum occurring in July (19.1°C) and minimum in February (-1.0°C). The mean annual frost-free period is 219 days, with an early frost in October and late frost in March (2005–2015) (Yu 2019). The original natural vegetation is a deciduous broadleaved forest. We obtained monthly climate data from the nearest meteorological station, located < 300 m from the study site.
We conducted the study in two plots: a HISD plantation with active restoration, and a natural secondary forest with passive restoration. Before the establishment of the experiment, both sites were bush woods and had experienced with similar levels of anthropogenic disturbance for firewood extraction. The bush woods were dominated by Corylus heterophylla and Corylus mandshurica, with Carexmitrata var. aristata, Parathelypteris nipponica, and Aruncus sylvester in the herb layer. Both plots were established in the autumn of 1987 with similar community structure. The distance between them is 346 m and site characteristics are presented in Table 1.
The HISD plantation was created by the technology of strip-cutting shrubs on mountain slopes based on the principle of biodiversity-oriented silviculture (Fig. 1). In the autumn of 1987, two parallel belts were set along contour lines. One belt (about 3 m width) is called preserved belt, where all shrubs and herbs remained intact. The other belt (about 2 m width), where aboveground part of vegetation except native tree propagule was cleared, is called cutting belt. In the spring of 1988, three to four years old seedlings (about 85% Pinus armandii, 10% Pinus tabuliformis, 5% Larix kaempferi) were planted in the cutting belts with about 3 m apart from each other. Therefore, the stand age of the HISD was from 20 to 36 years old during the litterfall observation period from 2005 to 2021. After planting, all seedlings were watered 5 times for the first month and kept intact afterwards. The density of planted seedlings was 1500–2000 in./ha. The ecological function of the preserved belt was preserving biodiversity of native species and preventing soil erosion. Clearing shrubs and herbs on the cutting belt increased available light penetrated through canopy and reduced the competition between tree seedlings and native shrubs and herbs, consequently lead to higher survival ratio and growth rate of tree seedlings. There was no anthropogenic disturbance after establishment on the natural secondary forest plot.
Table 1
Site information and soil properties (mean values) at the high indigenous-species diversity plantation (HISD plantation) and the secondary forest.
| Plot size | Slope (°) | Altitude (m) | Soil pH (H2O) | Soil C(mg/g) | Soil N(mg/g) | Soil P(mg/g) |
HISD plantation | 50 m*50 m | 34 | 1891 | 5.42 | 21.29 | 2.9 | 0.35 |
Secondary forest | 30 m*50 m | 30 | 1860 | 5.39 | 16.30 | 2.1 | 0.30 |
Soils were collected at 0–20 cm depth in August of 2020.
2.2 Soil samples
Soil samples were collected at 0–20 cm depths in the HISD plantation and the secondary forest in August of 2005, 2010, 2015, 2020 to compare the soil fertility. Each soil sample was mixed and divided to two samples. One sample was stored at 4°C prior to analysis, and the other was air-dried and ground and then sieved through sifter with a 2 mm mesh. Soil pH was measured with a pH meter (Mettler Toledo, China) using a 1:2.5 soil/water ratio. Soil C and N concentrations were analyzed with a Vario MACRO cube elemental analyzer (Elementar Analysensysteme GmbH, Germany), and soil total P (TP) was analyzed colorimetrically using the ascorbic acid molybdate method. Colorimetric methods were used to measure NO–3-N and NH4+-N using KCl leaching (AA3 Auto Analyzer 3, Seal, Jena, Germany). After samples were extracted with HCL-NH4F, available phosphorus (AP) was measured by molybdenum blue colorimetric method.
2.3 Tree community parameters
The vegetation structure was measured in both sites in July of 2010, 2012, 2015, 2017, and 2020. All trees > 1.3 m in height were identified and measured. The diameter at breast height (DBH) was measured with a diametric tape and the height with a graduated fishing rod. We calculated tree community parameters per plot: basal area (m2/ha), stem density (ind. /ha), height (km/ha).
2.4 Litterfall collection
Litterfall was collected in each study plot, monthly over 17 conservative years from 2005–2021 with stand ages from 20 to 36 years. The traps were cones of 0.5 m2 area with 80 cm diameter at the top. They were made with 0.1 mm nylon netting funnels, attached to circular stainless steel frames and fixed 30cm above the ground on iron stakes.
We randomly distributed 30 traps in each plot with two traps in each 10×10 m subplot of the HISD plantation and the secondary forest. We collected litter monthly typically at the end of each month. The litter samples were oven-dried for 48 h at 65°C and then divided into: leaves, twigs (including bark), reproductive parts (flowers and fruits (including seeds)), and debris (unidentified materials).
2.5 Carbon and Nutrient returns
The sum of leaves and twigs accounted for 85.8% and 87.7% of the total litterfall in the HISD plantation and in the secondary forest, respectively, over the 17-yearstudy period (Fig. 3). So we used the amount of leaf and twig litterfall to substitute for the total litterfall production in the two forests.
P. armandii and Quercus aliena var. acutiserrata were dominant species in the HISD plantation and the Secondary forest, respectively. We used fresh litter of leaves and twigs of P. armandii and Q. aliena var. acutiserrata collected in September 2020 for determining the C and nutrient concentration. Carbon and N were quantified using an elemental analyzer (Vario MACRO cube CN, Germany). Phosphorus, Ca and Mg were extracted by sulfuric acid-perchloric acid digestion. Phosphorus was determined by spectrophotometric with the molybdenum blue method, and Ca and Mg were measured by ICP-AES (Atomscan Advantage, Thermo Jarrell Ash). The Annual C and nutrients (N, P, Ca and Mg) return (kg/ha/yr) were estimated by the sum of nutrients return of litter leaves and those of twigs. Nutrients return of leaves and twigs were determined by multiplying nutrient concentration by the weight of their litter production, respectively.
2.6 Data analysis
All data were log-transformed prior to analysis to achieve normal distribution. We used two-way ANOVA to test the effects of forest type, year and their interaction on litter production and tree community parameters. Multiple comparison of means were performed using Tukey’s HSD test and we considered P < 0.05 as significant. One-way ANOVA was conducted to determine the significant difference of forest type on C and nutrients return (N, P, Ca and Mg), soil nutrient, and the difference of nutrient concentrations between P. armandii and Q. aliena var. acutiserrata. To evaluate factors regulating litterfall production, Pearson correlation coefficients were calculated between annual litterfall production and climate variables, soil nutrients and tree community parameters for each forest type. All data processing and analysis were performed at α = 0.05 with IBM SPSS Statistics for Windows version 21 (SPSS Inc., Chicago, IL).