Introduction

Thinning is a purposeful anthropogenic disturbance of natural forests and plantations that aims to promote the growth of trees and improve their quality by removing a subset of trees (Portmann et al. 2022; Wang et al. 2022a, b). This management measure affects the composition and density of species, improves environmental conditions, and thus affects the productivity of forest stands and carbon sequestration (Osuri et al. 2019). Thinning improves the forest environment, promotes the healthy growth of trees, enhances the quality of plants, reduces strain on forest resources, and provides economic benefits; it thus makes a positive contribution to societal development (Wang et al. 2023a, b, c; Zhang et al. 2023a, b) promotes the sustainable development of China’s forestry industry. The use of appropriate forest thinning techniques improves the natural environment and ecosystem health (Shrestha et al. 2018; Rosano-Peña et al. 2021; Gao et al. 2023). Tending thinning alters the structure of forest stands, improves their quality and productivity, and increases the ecological utility of forest resources (Abrahamsson et al. 2013; Horváth et al. 2021; Zhou et al. 2023). The use of appropriate conservation thinning intensities enhances the growth of forest trees (Nanni and Grau 2017; Remy et al. 2021; Wang et al. 2023a, b, c). They can also reduce the incidence of forest pests and diseases (Bateni et al. 2021; Ma et al. 2019; Tanner and Strong 2023) and the possibility of forest fires (Belletti et al. 2012; Chen and Wu 2022; Gao and Li 2023).

Recent studies have examined the effects of thinning on the transformation of natural forests and the protection of existing forest resources. Nurturing thinning has been shown to improve ecosystem productivity, quality of forest resources, biodiversity, ecosystem stability (Štraus et al. 2023; Wang et al. 2023a, b, c; Zhang et al. 2023a, b). Liu et al. (2023) found that at 15% and 20% thinning intensities, stands with better stand structure were more ecologically viable due to a high degree of hybridization, appropriate size differentiation, and intense competition. In contrast, plots with 35% thinning intensity were weakly mixed, vertically simple, and lacked young trees in the understory, resulting in poor overall stability. Zhang et al. (2023a) found that thinning increased carbon stocks and species diversity in ecosystems compared to controls. The most significant increase in carbon stocks and species diversity was found at 30% thinning intensity, indicating that 30% thinning intensity favors the accumulation of carbon stocks and the enrichment of species diversity. Wang et al. (2022a, b) found that the heavier the thinning, the larger the crown. Trees were more evenly distributed under moderate thinning (thinning intensity of 15%–30%). All trees in the light thinning (thinning intensity less than 15%) and heavy thinning (thinning intensity more than 30%) treatments were highly mixed compared to the control sample plot CK. Both size differentiation and crowding decreased with increasing thinning intensity.

Research on forest conservation has mainly focused on the efficacy of different methods, such as the intensity of conservation thinning (Wang et al. 2023a, b, c), land preparation (Flechard et al. 2020; Hanssen et al. 2022), cutting and weed control (Keller and Fox 2019; Li et al. 2019; Sheehan et al. 2019), vegetation growth (Iacob et al. 2014; Chen et al. 2021), stand structure (Wu et al. 2018; Zhu et al. 2022), and canopy structure (Priano et al. 2014; Li et al. 2022). Most previous studies have used single indicators and focused on forest types in single stands; however, few studies of natural mixed coniferous and broadleaf forests have been conducted. The effects of forest management vary due to differences in the ecological environment (Pilli et al. 2015; Qie et al. 2017; Yu and Gao 2020), stand conditions, and vegetation types of natural forests (Du et al. 2014; Lutz and Howarth 2015; Bouriaud et al. 2016; Carrasco et al. 2017; Gale et al. 2017; Pretzsch et al. 2019). The most suitable forest management regime thus depends on local endemic types. In this study, we characterized the physical and chemical properties of soil, biodiversity, regeneration, deadwood characteristics, and canopy structure of natural mixed coniferous and broadleaf forests in Xiaoxing’an Mountains, China, under different thinning intensities. We used principal component analysis (PCA) to quantitatively evaluate the effects of different thinning intensities. The optimal forest management method was identified to aid the formulation of forest management regimes for natural mixed conifer and broadleaf forests in Xiaoxing’an Mountains, China.

Materials and methods

Study area

The study area was located in Dongfanghong Forest in Xiaoxing’an Mountains, Heilongjiang Province, China (128°37′46″ to 129°17′50″ E, 46°50′8″ to 47°21′32″ N). The average altitude is 600 m above sea level, and the slope (14°) is oriented to the northwest (Fig. 1). The study area is located on the east side of Eurasia and experiences a temperate humid monsoon climate with large seasonal temperature differences. Historical data show that the highest temperature can reach 37 °C in July, and the lowest temperature in early January can reach − 40 °C; the average annual temperature is 1.4 °C, and the average annual precipitation is 661 mm. The soils are dominated by dark brown loam, and a small amount of woodland occurs in areas with swampy soils and valley meadows; the soils are weakly acidic. Natural mixed coniferous and broadleaf forest is the dominant forest type.

Fig. 1
figure 1

Distribution map of Dongfanghong Forest farm sample plot in Dailing Forestry Bureau

Full size image

In 2011, an area with uniform geographic and microtopographic conditions was selected for fertilization. The understory nursery method was used. Non-target species and excess trees were removed, and downed trees, dying trees, trees with poor stem form, and dead and diseased trees were felled. Replanting operations were carried out at the same time. Six sample plots 100 m × 100 m were established with different thinning intensities: plot A (10%), B (15%), C (20%), D (25%), E (30%), F (35%), and the control sample area (0%)There were 7 standardized plots, and the conservation thinning intensity was the ratio of cutting amount to total stock. In July 2023, six kinds of tending thinning intensity were investigated for trees with DBH greater than 5 cm in the above seven plots. The sample summary is shown in Table 1.

Table 1 Overview of the sample plots
Full size table

Research methodology

The 50 index research methods

For forest carbon sequestration, there are three vertical structures from top to bottom, namely the forest canopy, the litter layer, and the soil layer. The forest canopy layer plays an important role in light and precipitation interception, affecting the photosynthesis of the forest stand and the water content within the forest; the forest deadwood has an extremely important impact on the study of forest hydrological characteristics, and plays an important role in soil and water conservation and water containment; soil fertility affects the supply of nutrients to the trees and species diversity is mainly a study of the stability of the forest ecosystem after the thinning interval.

Determination of soil physical and chemical properties: In each plot and the control plot, sampling was carried out in accordance with the “S” mixed sampling method. Five soil samples were selected for each tending and thinning plot, and soil samples with a thickness of 10 cm were taken from each sample point, and 1 kg of soil samples were taken in accordance with the quadratic mixed sampling method. Soil samples were taken to a laboratory, where they were naturally dried, ground and screened, and then analyzed for chemical properties, which were measured as shown in Table 2.

Table 2 Methods for determining soil chemical properties
Full size table

Deadwood collection and measurement: The “Z-type” sampling method was used to sample each transformed sample plot and the same plots, and the sample size was 30 cm × 30 cm. The undecomposed and semi-decomposed dead branches under the forest were collected and brought to the laboratory, and the accumulation of dead branches was obtained by drying method. First, the litter was dried in a drying oven at 85 °C, then the volume of litter was calculated using the dry matter mass, and the water content of the apoplast was determined using the soaking method. Calculate the maximum water holding capacity and effective storage capacity of the apoplastic material based on the measured apoplastic volume. The maximum water holding capacity and soaking time were 0.25, 0.50, 1.00, 2.00, 4.00, 8.00 and 24.00 h, respectively.

Canopy survey: The longitude, latitude and elevation of each tree were measured by GPS at 5 random points in 6 thinning intensity plots and parallel plots. After determining the northward direction, the leveling data acquisition device Mini-O-Mount 7 MP was leveled and the distance of the instrument lens from the ground was measured and observations were made from three to four different directions. The Winscanopy canopy analyzer was used to collect images from four different directions, east, south, west, and north, and the acquired data were processed with XLScanopy to analyze the canopy structure.

Biodiversity survey: A survey of trees, shrubs and herbaceous vegetation in each plot and in parallel plots to record the species and number of trees; Measure the DBH and height of the remnant tree with a DBH rule and altimeter. The growth rate of the residual trees was calculated in terms of DBH and height in successive years. The coordinates of tree species were determined by total station. According to the size of the sample plots, three shrub plots were randomly set up in the plot to determine the species and coverage of shrubs; five herb sample plots of 1 m × 1 m were set up in accordance with the “Z” shape within the shrub sample plots to investigate the species and cover of herb plants.

Seedling survey: Seedling height and ground diameter were measured using digital vernier calipers and steel rulers for all Pinus koraiensis Sieb. et Zucc., Pinus sylvestris var. mongolica Litv., Picea asperata Mast, Larix gmelinii (Rupr.) Kuzen seedlings in each plot; the same method was adopted in control plots by selecting young trees of comparable stand age.

Evaluation methodology

By using the principal component analysis method, the transformation effect of the natural needle and broad mixed forest in Xiaoxing’an Mountains was comprehensively evaluated. The basic steps were as follows: There were n different evaluation indicators and m evaluation indicators, and the set of them constituted the original matrix X:

$$X = \left( {x_{ij} } \right)_{m \times n} \;\;\;i = 1,2, \cdots ,m;j = 1,2, \cdots ,n$$

where, xij denotes the measured value of the i indicator for the j sample plot.

Standardized raw matrices

The raw data was standardized using SPSS26.0, eliminating the effects of orders of magnitude and dimensions. The forward index was standardized according to Eq. 1, and the reverse index was standardized according to Eq. 2.

$$X_{ij}^{*} = \frac{{X_{ij} }}{{\overline{{X_{j} }} }}$$
(1)
$$X_{ij}^{*} = \frac{{\overline{{X_{j} }} }}{{X_{ij} }}$$
(2)

Principal component determination

SPSS26.0 software was used to process the standardized data, and the first m principal components whose cumulative contribution rate of ANOVA was ≥ 85% were selected to establish the relationship between m principal components and standardized variables. The calculation formula was as follows:

$$\mathop Y\nolimits_{k} = \mathop b\nolimits_{k1} \mathop X\nolimits_{1}^{*} + \mathop b\nolimits_{k2} \mathop X\nolimits_{2}^{*} + ... + \mathop b\nolimits_{kp} \mathop X\nolimits_{p}^{*}$$
(3)

Weight determination

The weight of each principal component was the ratio of the contribution of the kth principal component to the total contribution of the selected m principal components:

$$w_{k} = \frac{{\lambda_{k} }}{{\sum\limits_{k = 1}^{m} {\lambda_{k} } }}$$
(4)

Wk and λk denoted the weight of the kth principal component and the kth principal component, respectively.

Comprehensive evaluation function

According to the first m principal components determined by Eq. 3 and the weights obtained by Eq. 4, a comprehensive evaluation function was established:

$$F = \sum\limits_{k = 1}^{m} {w_{k} Y_{k} }$$
(5)

F denoted the overall evaluation score of the sample plots with different conservation thinning intensities. Higher comprehensive evaluation scores indicate stronger carbon sequestration.

Results

A comprehensive evaluation was conducted on the carbon sink of natural coniferous and broad-leaved forests in the Xiaoxing’an Mountains. The model was established using as many evaluation indicators as possible. After sparse forests of different intensities in natural mixed coniferous and broad-leaved forests, the forest soil exhibits weak acidity (pH close to 6.0). When the pH value approaches 7.0, the neutralization of soil acidity is more conducive to the growth of forest vegetation. With pH value as a positive indicator, all positive indicators in this study were standardized according to Eq. 1. The bulk density of forest soil represents the degree of soil compaction, and the higher the bulk density of soil, the more unfavorable the growth of forest vegetation. The bulk density of forest soil was taken as the inverse index and standardized according to Eq. 2 in Table 3.

Table 3 Standardized treatment of various indicators
Full size table

Using SPSS26, the standardized data were subjected to PCA, and the results of the ANOVA for each indicator at each sample site are shown in Table 4. The cumulative contribution of the first five principal components was 93.90%, which indicates that they captured most of the variation in the data. These principal components could thus be used to determine how the thinning intensity affects the carbon sequestration of mixed natural coniferous and broadleaf forests of the Xiaoxing’an Mountains of China.

Table 4 ANOVA for each indicator across the different treatments
Full size table

The factor loadings of the first five principal components were extracted and are shown in Table 5. As can be seen from Table 5, the first principal component had large loadings on the indicators of total soil porosity, soil water content, direct fixation factor, indirect fixation factor, total N, total P, organic matter, and species richness index (S) of the herb layer. The second principal component loaded heavily on the growth and preservation rates of red pine, spruce, and larch, species richness index of the tree layer (S), the Shannon–Wiener species diversity index (H′), Pielou’s evenness index (J), maximum water-holding capacity and natural water-holding rate in the semi-decomposed layer of deadfall, and effective interception area. The third principal component loaded heavily on pH, quick-impact P, organic matter, maximum water-holding capacity in the undecomposed layer of deadfall, effective interception area, degree of openness, total fixation factor, and direct radiation fluxes, scattered radiation fluxes, and total radiation fluxes under the canopy. The fourth principal component loaded heavily on the shrub layer and herb layer species richness index (S), Shannon–Wiener species diversity index (H′), Pielou evenness index (J), deadfall undecomposed layer effective interception area, quick-impact N, quick-impact P, and maximum water-holding capacity of the deadfall semi-decomposed layer. The fifth principal component had a large loaded on the growth and preservation rates of red pine, spruce, and larch, deadfall semi-decomposed layer accumulation, and maximum water-holding capacity.

Table 5 Factor loadings of each indicator in each sample site
Full size table

The score coefficient matrices of each principal component under the different thinning intensities are shown in Table 6 and reveal the relative contributions of the evaluation indexes of each principal component.

Table 6 Matrix of coefficients of principal component scores of indicators under different thinning intensities
Full size table

The covariance matrix of the principal component scores of each indicator for each sample site under different conservation thinning intensities is shown in Table 7.

Table 7 Covariance matrix of principal component scores of indicators across the different treatments
Full size table

The factor scores of the five principal components were calculated, and the weights of each principal component were 0.39, 0.19, 0.16, 0.15, and 0.11.The comprehensive evaluation function was generated, and the formula was used to determine the effects of thinning intensity. The factors of the various principal components, the comprehensive scores, and the rankings of the comprehensive scores are shown in Table 8. The comprehensive score indicating the extent of transformation in each sample plot was highest in E (0.75), followed by D (0.63), F (0.50), C (0.48), B (0.22), A (0.11), and CK (0.06). Sample plot E had the highest composite score, indicating that a 30% thinning intensity was the most effective in promoting carbon sequestration in natural mixed conifer and broadleaf forests in Xiaoxing’an Mountains, Heilongjiang Province, China.

Table 8 Factor scores of the principal components and comprehensive scores for each plot
Full size table

Discussion

In this study, six different intensities (10%, 15%, 20%, 25%, 30%, and 35%) of intercropping were carried out to renovate the natural mixed coniferous and broadleaf forests in Xiaoxing’an Mountains, Heilongjiang Province, China, and replanting was carried out at the same time of thinning; the replanted species were spruce, red pine, and larch. Sample plots with similar stand conditions were selected as control plots. Sample plots had similar stand conditions as the CK, and 50 variables were measured, including soil physical and chemical properties, water-holding capacity of deadwood, canopy structure parameters, species diversity, and seedling regeneration of the different sample plots. Descriptive statistical analysis was conducted to establish a comprehensive evaluation model using PCA, and the carbon sequestration of the forest under different thinning intensities were determined.

Soil physical properties are important indicators for evaluating the water-holding capacity which directly affects the root growth of forest trees and their absorption of water and nutrients (Agathokleous and Yu 2023). At a thinning intensity of 30%, soil physical properties were optimized, the infiltration capacity of surface water was high, and the intensity of surface runoff was small. Soil porosity was also large, and the soil penetration resistance was low. This facilitates the extension and development of tree and sapling roots, the transport of nutrients, air, and water, and the growth and regeneration of forest floor vegetation (Basche and DeLonge 2019). Soil chemical properties have major effects on the productivity and health of forests, the content of soil organic matter, and many soil properties such as the content of nitrogen, phosphorus, and potassium as important indicators of soil fertility, as these are important nutrients required for tree growth (Wang et al. 2014).

Deadwood is an important component of forest ecosystems (Gao et al. 2023). The forest deadwood layer has a loose structure with a strong water-absorbing capacity and high water permeability that slows surface runoff (Huo et al. 2023). It can reduce soil decomposition, promote the cycling of substances, and maintain fertility. Comparison of the water-holding characteristics of the deadwood revealed that the 30% logging intensity was optimal and comparison of the water-holding characteristics of litter showed that the volume of litter storage and the maximum water-holding capacity were higher at a 30% thinning intensity, which enhances soil and water conservation (Yang et al. 2019).

The diversity of understory shrubs and herbs is positively associated with the intensity of thinning (Zhang et al. 2018). This mainly stems from the fact that the growth of both shrubs and herbs requires certain light conditions, and the reduction in the density of trees after thinning can provide these plants with necessary light which accelerates the development of understory plants and promotes the development of the plant community (Chen et al. 2023). Forest communities with high species diversity are highly resistant to exotic species, pests, and diseases and show strong resilience after being disturbed (Zhang et al. 2013), as forest communities with high species diversity contain more populations with different ecological characteristics and are more resistant to fluctuations. A large number of studies have shown that there is a positive correlation between understory species diversity and stability. Ecosystems with high species diversity likely include more species that are resistant to disturbance, which increases ecosystem stability.

The canopy structure of forest stands is mainly analyzed using canopy structure parameters and light environment characteristics (Guan et al. 2021). The most important index affecting canopy structure is the leaf area index, which directly affects the light energy below the canopy and the degree of rainwater retention. Thinning greatly promotes the growth of major tree species within the forest (Khatibi et al. 2018), and the leaf area index of forest species within the sample plot increases after thinning (Chou et al. 2023). Subcanopy light radiation flux was the most important variable, and the total subcanopy light radiation flux was mainly derived from subcanopy direct radiation flux. This indicated that the forest canopy structure and light distribution were optimal at a thinning intensity of 30%.

Conclusion

The carbon sequestration effects of six types of nursery inter-stocking intensities in natural mixed conifer and broadleaf forests in Xiaoxing’an Mountains, Heilongjiang Province, China, were comprehensively evaluated by principal component analysis, and 50 evaluation indexes, such as canopy structure of forest stands, water-holding performance of understory litter, species diversity, soil physical and chemical properties, and growth status of renewed seedlings, were screened out to reflect the carbon sequestration effects of natural mixed conifer and broadleaf forests, and growth status of regeneration were used to characterize the carbon sequestration of natural forests. These five sets of indicators were analyzed after conservation thinning, and PCA was used to establish a comprehensive evaluation model. The carbon sequestration effects of different thinning intensities were highest in plot E (− 0.75), followed by D (− 0.63), F (0.50), C (0.48), B (0.22), A (0.11), and the CK (0.06). The E sample plot had the highest composite score, indicating that the carbon sequestration was highest at a thinning intensity of 30%. This study provided valuable suggestions for the sustainable management of natural mixed conifer and broad forests in Xiaoxing’an Mountains, Heilongjiang Province, China, and will aid scientific research of this forest type. Generally, the methods used in the study could aid the development of improved approaches for comprehensively evaluating the efficacy of different thinning intensities on forest properties, including carbon sequestration.