Impacts of Forest Management on the Biodiversity and Sustainability of Carya dabieshanensis Forests

: Natural secondary Carya dabieshanensis forests comprise the natural wealth of the Dabie Mountains, which maintain the ecological balance of this region and enhance the incomes of farmers. However, forest ecosystems are being increasingly challenged due to increasing long-term anthropogenic development and management. Elucidating the impacts of management intensity and duration on the diversity and stability of vegetation communities in natural secondary Carya dabieshanensis forests is of great signiﬁcance toward achieving sustainable forest management. For this study, we compared the effects of three forest management intensities (no management (CK), extensive management (EM), and intensive management (IM)), and ﬁve management durations (0, 3, 8, 15, and 20 y) on species diversity and the stability of vegetation communities (trees, shrubs, and herbaceous layers) of a Carya dabieshanensis forest. Our results revealed that the diversity of the vegetation communities continued to decline following the onset of management activities. The diversity, dominance, evenness, and richness indices of the vegetation communities decreased by 53.37%–62.77%, 46.04%–59.17%, 32.58%–53.55%, and 50.18%–51.30%, respectively, after 20 y of forest management. Intensive management translated to species diversity loss more so than extensive management. All vegetation communities of the forest stands under study were generally unstable; however, the stability they did have was not at risk under greater management intensity and duration, and neither did greater species diversity enhance it. This study provides new evidence to support the non-synchronicity of species diversity and community stability in forest resident vegetation communities. Since the species diversity and structural composition of tree layers have a positive effect on community stability, so we suggest that both extensive and intensive forest lands need to retain or replant more tree species other than Carya dabieshanensis .


Introduction
Biodiversity is an essential foundation for ecosystem sustainability, and plays an important role in maintaining the structural stability and functional complexity of ecosystems [1,2], and to some extent reflects their productivity. The study and protection of biodiversity has become a critical task that aligns with the development of human destiny [3].
Forests cover more than 30% of the land area worldwide and are home to most terrestrial biodiversity on Earth. They play irreplaceable roles for the protection of ecosystem diversity and maintaining species richness [4]. A recent study revealed that protecting and enhancing the diversity of forest vegetation, particularly tree diversity, can promote soil carbon and nitrogen storage, as well as improve carbon capacities and soil nitrogen fertility [3]. A meta-analysis showed that on a global scale, the diversity of plant species ment (REA). From the perspective of the applicability of a given method, hyperspectral image estimation and model estimation are more suitable for large-scale applications and require richer databases [23,24]. REA do not require species identification, which greatly improves work efficiencies; however, the selection of variables used and the skills of field workers have higher requirements, and often necessitate professional training [25]. Sample surveys are basic methods that are extensively used in the study of species diversity in forest ecosystems worldwide. Although this work is time consuming and laborious, it provides accurate data for species and is universally applicable [2,26]. At present, there are two main strategies for the determination of community stability. One strategy involves the use bioecological methods to assess stability by quantifying the compositions of tree community species, age structures, and distribution patterns [8,12]. A second approach involves the use of quantitative ecological methods, such as the optimized M. Godron stability measurement method, community stability index, etc. [27,28]. Between them, the M. Godron stability measurement method has lower data requirements and has a more flexible discrimination of community stability; thus, it is widely used.
Carya dabieshanensis is prevalent in the Dabie Mountains of China, where its natural secondary forest is primarily distributed across the Dabie Mountains of subtropical Northern China. This species has flourished on a large scale over the last 20 years due to the rich nutritional value and high-quality taste of its fruit [29]. To enhance fruit yields, fertilization, as well as the regular removal of understory vegetation and miscellaneous trees (other than Carya dabieshanensis) have been the most common management methods used by operators [30]. More than 20 years of forest management have modified the original stand structures, as many species have disappeared in this community. Furthermore, the diversity of forest vegetation and sustainable management capacities in this area are facing challenges.
This study aims to elucidate the impacts of forest management intensity and duration on the diversity of vegetation species and community stability. Furthermore, to determine how the biodiversity of this area might be protected and maintained to sustain its unique stand community structures and ecological functions. To this end, we established 45 fixed plots in a natural secondary Carya dabieshanensis forest area. The diversity of tree, shrub, and herbaceous plant species in the plot communities was enumerated, and the stability of the community structures was quantified. We hypothesized that: (1) With increased management intensity and duration, the diversity of plant species in the natural secondary forest of Carya dabieshanensis continually decreases; (2) Intensive management has a greater effect on the stability of vegetation communities in the natural secondary Carya dabieshanensis forest than does extensive management; (3) The diversity of vegetation species plays a decisive role in community stability.

Study Area
The study site is situated in the Dabie Mountain Forest Area, Tiantianzhai Town, and Guanmiao Township, in Jinzhai County, of Anhui Province, China ( Figure 1, Table S1). This region is home to a humid subtropical monsoon climate with average temperatures that range from 26 • C in summer to 2 • C in winter, an altitude of 150-1353 m, and average annual precipitation of 1300 mm [30]. The soil thickness ranges from 30-100 cm, is mainly yellow-brown loam (70% medium loam and sandy loam), and slightly acidic (pH 4.5-6.5), with a slope of 20 • -30 • . The primary tree layer in the forest area is Carya dabieshanensis, accompanied by small quantities of Cunninghamia lanceolata, Castanea mollissima, and Fortunearia sinensis. Due to the regular removal of understory vegetation, there is a negligible shrub layer, with Diospyros lotus and Lindera glauca being the more common species, while Duchesnea indica, Erigeron annuus, Aster tataricus, and Stellaria chinensis are the dominant herbs. All study plots were established on shady and semi-shady slopes at altitudes of from 550 to 850 m. Referring to the definition of management strategies in other regions of bamboo forest, chestnut forest, and crop planting areas, we divided the management strategy of the natural secondary Carya dabieshanensis forest into three categories (unmanaged, extensive management, and intensive management) [31][32][33]. The unmanaged model was referred to as CK, as it could be seen as blank processing in terms of administration duration and management intensity. In the unmanaged stands, farmers only picked fruit in autumn and were not engaged in any other management activities. In extensively managed and intensively managed forest stands, during the first year of management activity, all trees except for Carya dabieshanensis were cleared from the forest. Subsequently, the differences in management included the number of times the understory cover was cleared each year, and whether fertilization was applied. Among them, in the extensively managed stands, the understory vegetation was cleared using a brush cutter in August each year without fertilization. In the intensively managed forest area, a brush cutter was used twice in June and August to clear the understory vegetation, and compound fertilizer (N: P2O5: K2O = 13:5:7) was applied (375 kg·hm −2 ) in June and August of each year. Organic fertilizer (organic matter content 45%) was applied (1500 kg·hm −2 ) once a year, from November to December.
Before 2002, none of the sites we chose had any human management activities. Next, we selected five time periods (0 y (CK), 3 y, 8 y, 15 y, and 20 y) as the durations of the management time series according to the actual onset of these activities.

Sampling Design
In August 2022, based on the information provided by local operators, a total of nine forest areas (CK, EM-3, EM-8, EM-15, EM-20, IM-3, IM-8, IM-15, and IM-20) were surveyed, and five duplicate plots were established in each forest area, for a total of 45 plots. For each plot, 20 m × 20 m subplots were randomly established and included in the tree Referring to the definition of management strategies in other regions of bamboo forest, chestnut forest, and crop planting areas, we divided the management strategy of the natural secondary Carya dabieshanensis forest into three categories (unmanaged, extensive management, and intensive management) [31][32][33]. The unmanaged model was referred to as CK, as it could be seen as blank processing in terms of administration duration and management intensity. In the unmanaged stands, farmers only picked fruit in autumn and were not engaged in any other management activities. In extensively managed and intensively managed forest stands, during the first year of management activity, all trees except for Carya dabieshanensis were cleared from the forest. Subsequently, the differences in management included the number of times the understory cover was cleared each year, and whether fertilization was applied. Among them, in the extensively managed stands, the understory vegetation was cleared using a brush cutter in August each year without fertilization. In the intensively managed forest area, a brush cutter was used twice in June and August to clear the understory vegetation, and compound fertilizer (N: P 2 O 5 : K 2 O = 13:5:7) was applied (375 kg·hm −2 ) in June and August of each year. Organic fertilizer (organic matter content 45%) was applied (1500 kg·hm −2 ) once a year, from November to December.
Before 2002, none of the sites we chose had any human management activities. Next, we selected five time periods (0 y (CK), 3 y, 8 y, 15 y, and 20 y) as the durations of the management time series according to the actual onset of these activities.

Sampling Design
In August 2022, based on the information provided by local operators, a total of nine forest areas (  each plot, 20 m × 20 m subplots were randomly established and included in the tree layer vegetation survey, which involved measuring and recording the species name, DBH (diameter at breast height), height, and crown width of all trees (DBH > 5 cm, height ≥ 6 m). In each tree subplot, three 5 m × 5 m shrub layer plots were set diagonally, for a total of 135 shrub plots, wherein all names, populations, heights, and crown widths of all shrubs were measured and recorded (e.g., all woody plants with a ground diameter of ≤5 cm, 5 m > height ≥ 0.3 m, where the ground diameter refers to the thickness of the stem at 10 cm aboveground). A total of ten 1 m × 1 m herbaceous layer sample plots (all vascular plants with heights of <0.3 m) were set diagonally in each tree subplot, for a total of 450 herb plots, for which the names, populations, and coverage of the herbaceous species where identified, measured, and recorded.

Data Analysis
The stand species diversity α was characterized by the species richness combined with the diversity, dominance, and evenness indices. The Margalef richness index (S) of each forest layer species was expressed by the species number, diversity was expressed by the Shannon-Wiener diversity index (H) and Simpson dominance index (D), while the species Pielou evenness was calculated based on the H and S indices (J).

Margalef richness index :
Pielou evenness index : J = H ln S where S represents the total number of species in the plot; P i is the proportion of the ith species to the total number of individuals, that is, P i = N i /N; N i is the number of individuals of the ith species; and N is the total number of individuals of all species in the plot.
An improved M. Godron [27] stability assay was employed to calculate the community stability [2]. The reciprocal cumulative percentage of the total number of species in the community, and the relative frequency accumulation percentage of species occurrence were used to simulate the curve equation, respectively. In addition, we used a straight-line equation (y = 100 − x) that intersected the smooth curve, where the closer the intersection point was to the stable point (20,80), the more stable the community. The converse meant an unstable community. The Euclidean distance (d) between the intersection coordinates and the stable point coordinates indicates the community stability status.
All data analysis was performed using R 4.1.3 and Excel 2016; α diversity was calculated using the "vegan" package and plotted with the "ggplot2" package. One-Way ANOVA was used to compare the effects of management intensity and duration on vegeta-tion community α diversity, and LSD tests were performed on the diversity of vegetation community α for each forest stand. Differences in the β diversity of vegetation communities in various stands were compared through primary coordinate analysis (PCoA). The "adonis" function in the "vegan" package was used to perform permutational multivariate analysis of variance (PerMANOVA) for the composition of plant communities in the tree, shrub, and herbaceous layers. For PerMANOVA analysis, the Bray-Curtis difference matrix was used to summarize the species composition, with 999 permutation tests to determine statistical significance. Finally, linear fitting was employed to observe the coupling relationship between α diversity and community stability for each forest layer.

Effects on Stand Characteristics
The management intensity and duration had significant impacts on the stand characteristics of the natural secondary Carya dabieshanensis forest (Table 1). Multiple comparisons revealed that the mean breast diameter and average tree height of the forest stands increased significantly following management measures (p < 0.05). However, the average density of the forest stands was generally lower than that of CK ( Figure 2). There were no significant differences in the average breast-height diameters of Carya dabieshanensis stands under different operating intensities during the same operational timeframe. The average tree height of Carya dabieshanensis forests under intensive management increased with prolonged management duration. After 20 years of intensive management, it was significantly higher than that of the average tree height under the extensive management model (21.35%). The stand basal area and stand density of extensively managed forests were almost all higher than those of intensively managed forests for the same management duration. Note: ***, correlation is significant at the 0.001 level.

Effects on the α Diversity Index of Species for Each Forest Layer
The management intensity and duration significantly impacted the diversity of vegetation community species (Table 2, Figure 3). The α diversity of species in each forest layer was significantly lower than that of CK (p < 0.05) under the two management intensities. The diversity and richness indices gradually decreased with management duration, but increased under extensive management. Under intensive management, the community species diversity decreased sharply and remained at a low level for a prolonged period. The diversity of vegetation community species dropped to similarly low levels under both management intensities 20 years after the onset of management activities. Compared with CK, after 20 years of extensive and intensive management, the diversity, dominance, evenness, and richness indices of vegetation communities decreased by 53.37%-62.77%, 46.04%-59.17%, 32.58%-53.55%, and 50.18%-51.30%, respectively.

Effects on the α Diversity Index of Species for Each Forest Layer
The management intensity and duration significantly impacted the diversity of vegetation community species (Table 2, Figure 3). The α diversity of species in each forest layer was significantly lower than that of CK (p < 0.05) under the two management intensities. The diversity and richness indices gradually decreased with management duration, but increased under extensive management. Under intensive management, the community species diversity decreased sharply and remained at a low level for a prolonged period. The diversity of vegetation community species dropped to similarly low levels under both management intensities 20 years after the onset of management activities. Compared with CK, after 20 years of extensive and intensive management, the diversity, dominance, evenness, and richness indices of vegetation communities decreased by 53.37%-62.77%, 46.04%-59.17%, 32.58%-53.55%, and 50.18%-51.30%, respectively.   The α diversity of species in the various vegetation layers of the Carya dabieshanensis forest was significantly affected by the management intensity and duration (Table 3). Compared with CK, regardless of the management mode, subversive changes in the vegetation species diversity of the tree layer occurred. Of these, changes in the species diversity of the tree layer under intensive management was the most significant (p < 0.05), where the management behavior led to a sharp decline in species diversity (Figure 4). The The α diversity of species in the various vegetation layers of the Carya dabieshanensis forest was significantly affected by the management intensity and duration (Table 3). Compared with CK, regardless of the management mode, subversive changes in the vegetation species diversity of the tree layer occurred. Of these, changes in the species diversity of the tree layer under intensive management was the most significant (p < 0.05), where the management behavior led to a sharp decline in species diversity (Figure 4). The species diversity of the shrub layer under different management durations varied under the two management intensities. It was worthy of note that in the early stages of extensive management (EM-3 and EM-8), the species diversity, dominance, and richness indices of the shrub layer were significantly higher than those for CK. Furthermore, the diversity and richness indices of the extensively managed forests were significantly higher than those of the intensively managed forests under the same management durations. Compared with CK, there were no significant changes in the diversity of the shrub layer species at the onset of intensive management (save for the richness index). Variations in the α diversity of herbaceous vegetation species with management duration were essentially identical under the two management intensities. The diversity of the herbaceous layer species increased three years following the initiation of management activities, albeit they decreased steadily under prolonged management duration. In general, higher management intensities accelerated this change. Note: H, D, J, and R are the Shannon-Wiener diversity index, Simpson dominance index, Pielou evenness index, and Margalef richness index of tree layer, shrub layer, and herb layer. *, correlation is significant at the 0.05 level; **, correlation is significant at the 0.01 level; ***, correlation is significant at the 0.001 level.

Effects on Species β Diversity at All Levels
The results of nonparametric multivariate variance analysis (perMANOVA) based on a distance matrix indicated that for distinct tree species, there were significant differences in the composition of plant species in the tree (p < 0.001), shrub (p < 0.001), and herbaceous (p < 0.001) layers of natural secondary Carya dabieshanensis forests in the Dabie Mountains under different management intensities and durations ( Table 4).
The results of a post-hoc comparisonsof species composition of the forest layers of different forest stands revealed that management activities led to significant differences between the compositions of vegetation community species and CK in each forest layer (p < 0.05). Except for the early stages of management (EM-3 and IM-3), there were significant differences in the species composition of the tree layers under different management intensities and the same management periods (p < 0.05) ( Table S2). The species compositions of the shrub and herbaceous community layers obviously changed, and there were significant differences between them across the nine different stands (p < 0.05). This indicated that changes in the management intensity and duration altered the compositions of shrub and herbaceous community species layers to some extent.

Effects on Species β Diversity at All Levels
The results of nonparametric multivariate variance analysis (perMANOVA) based on a distance matrix indicated that for distinct tree species, there were significant differences in the composition of plant species in the tree (p < 0.001), shrub (p < 0.001), and herbaceous (p < 0.001) layers of natural secondary Carya dabieshanensis forests in the Dabie Mountains under different management intensities and durations (Table 4).  The results of principal coordinate analysis (PCoA) revealed that the cumulative explanation rate of the differences in community species composition in the tree layer under different management intensities and durations was 74.52%. The species composition of the CK in the tree layer differed significantly from that of the other stands, which also implied that management behavior had a strong influence on the species composition of the tree layer ( Figure 5). The interpretation rates of the shrub and herbaceous layers were lower in the first two axes (31.8% and 43.3%, respectively), which indicated that the species composition of the shrub and herb layer in each stand remained uncertain, even under the same management intensity and duration. Within the 95% confidence interval, the distance between the stands under intensive and extensive management gradually increased with prolonged management. Based on the differences in species composition between stands (Table S2), the succession direction of community structures and the dominant species of each community was differentiating. differences in the species composition of the tree layers under different management intensities and the same management periods (p < 0.05) ( Table S2). The species compositions of the shrub and herbaceous community layers obviously changed, and there were significant differences between them across the nine different stands (p < 0.05). This indicated that changes in the management intensity and duration altered the compositions of shrub and herbaceous community species layers to some extent.
The results of principal coordinate analysis (PCoA) revealed that the cumulative explanation rate of the differences in community species composition in the tree layer under different management intensities and durations was 74.52%. The species composition of the CK in the tree layer differed significantly from that of the other stands, which also implied that management behavior had a strong influence on the species composition of the tree layer ( Figure 5). The interpretation rates of the shrub and herbaceous layers were lower in the first two axes (31.8% and 43.3%, respectively), which indicated that the species composition of the shrub and herb layer in each stand remained uncertain, even under the same management intensity and duration. Within the 95% confidence interval, the distance between the stands under intensive and extensive management gradually increased with prolonged management. Based on the differences in species composition between stands (Table S2), the succession direction of community structures and the dominant species of each community was differentiating.

Effects on Community Stability
The Euclidean distance (d) between the model crossover point (x, y) and the ideal stable point (20,80) varied under different management intensities and durations ( Figure 6). The model intersection of the CK vegetation community was closest to the stable point (d = 12.841), followed by EM-20 (d = 13.138). The model intersections of all forest stands with management activities were different degrees distant from the stable point (Table 5). The degrees of stability of the stand vegetation communities were as follows: CK > EM-20 > IM-20> IM-8 > EM-3 > IM-15 > EM-15 > IM-3 > EM-8. In this study, continuous forest management activities led to a general decrease in the stability of vegetation communities. Compared with CK, the damage to community stability in the early stages of management activities was more serious. It is worth noting that no changes in vegetation community stability correlated with management intensity and duration were found in the nine stands under study. model intersection of the CK vegetation community was closest to the stable point (d = 12.841), followed by EM-20 (d = 13.138). The model intersections of all forest stands with management activities were different degrees distant from the stable point (Table 5). The degrees of stability of the stand vegetation communities were as follows: CK > EM-20 > IM-20> IM-8 > EM-3 > IM-15 > EM-15 > IM-3 > EM-8. In this study, continuous forest management activities led to a general decrease in the stability of vegetation communities. Compared with CK, the damage to community stability in the early stages of management activities was more serious. It is worth noting that no changes in vegetation community stability correlated with management intensity and duration were found in the nine stands under study.

Vegetation Community Species Diversity
The versatility and ecosystem services of forest ecosystems benefit from the high biodiversity of vegetation communities [36]. Many studies have shown that anthropogenic forest management activities are the key factor that affects vegetation community species diversity and succession [37,38]. Contingent on multiple factors, including management objectives, location, intensity, and duration, some forests will exhibit a significant decline in species diversity following disturbances. However, they can recover over a relatively short period, or even become more robust. Other forests do not fully recover, as the disturbances may be too strong or last too long, which leads to a sharp decline in community species diversity [39,40].
In this study, vegetation community species diversity reflected different changes following the occurrence of forest management activities and was lower than CK regardless of the management intensity and duration. It was clear that management practices such as deforestation, understory vegetation removal, and fertilization led to the decline or even loss of certain species in the area. However, under lower management intensity (extensive management), the diversity of vegetation community species slowly and steadily decreased (Figure 3). Intensive management led to a rapid decline in vegetation community species diversity, but it remained relatively stable at a low level over a long period of time, which was consistent with the results of other studies [2,41]. According to Hrivnak [42], the competition between plants for soil and light resources is the main driving force that modifies plant species diversity and composition. The felling of miscellaneous trees other than Carya dabieshanensis will lead to a sharp decrease in the species diversity of the tree layer. Nevertheless, these trees will provide light and soil resources for the shrub and herbaceous layers, and some sun-loving herbs and shrubs may develop. Changes in the forest canopy drives the diversity of herb and shrub species in the early stages of management, which does not decrease or even increases [43,44]. However, under prolonged forest management, several neutral or shade-loving herbs will gradually withdraw from the community, leading to a steady decline in species richness ( Figure 4).
Although the understory vegetation (herb and shrub layers) was cleared with equal intensity, the changes in species richness varied between the shrub and herb layers, which may have been attributable to variations in their capacities to self-recover. Herbaceous plants have a robust vitality and can restore coverage in a short time after being disturbed. The capacity of shrubs to recover was not as good as that of herbs, where continuous logging compressed their habitat, which led to a more rapid decline in the richness and species diversity of the shrub layer in contrast to the herb layer [45]. Several studies have suggested that community species diversity generally decreases with the greater intensity and frequency of disturbances [42,46]. Denser understory vegetation clearance in intensively managed forests accelerated species replacement in vegetation communities. Fertilization increased the competition for nutrients between species, which further intensified the competitiveness of dominant species between populations. This eventually led to the long-term supremacy of some dominant species in the shrub and herbaceous layers [45]. Furthermore, it has been suggested that structural changes in the tree layer will affect soil porosity, soil nutrients, and surface runoff to indirectly affect the diversity of community species [2,47]. Consequently, further attention is warranted to learn whether changes in the soil characteristics of forest lands under different management intensities and durations drive those of species diversity in the vegetation communities of Carya dabieshanensis forests.

Vegetation Community Stability
The stability of forest ecosystem communities is generally considered as a key indicator of community anti-interference ability and resilience [48,49]. The relationship between community stability and species diversity has always been controversial. Many early researchers believed that species diversity had a positive impact on vegetation community stability, where the higher the species richness the stronger the community stability. This was purported to be because increased species diversity made the energy flows and material circulation in the community more complex [50][51][52]. However, a growing body of research has revealed that community diversity and stability are not simply positively correlated, as complex interspecific relationships may be more unstable than simple systems [7,53,54]. Studies have revealed that the influence of species dominance on community stability is greater than the number of species. This phenomenon is attributed to the uneven contribution of each species to community stability, where the increase in the number of species does not necessarily promote community stability [13,55]. Pillar's [56] research showed that the maintenance effect of species diversity on community stability derived from the insurance effect of functional redundancy. The spatial stability of plant communities is closely related to the supply capacity of land resources, while community functional redundancy is correlated with community stability, but not species diversity or redundancy [57].
In this study, changes in species diversity did not have a significant impact on community stability, and continuous forest management led to a steady decline in species diversity; however, this was not consistent with the changes in community stability. Species diversity and community stability were also inconsistent between different forest layers ( Figure 7). Community stability was weakly negatively correlated with α diversity in the shrub and herbaceous layers, but weakly positively correlated with α diversity in the tree layer. However, the ecological mechanisms behind this phenomenon are not well understood; thus, we suspected that this may have been related to the uncertainty of interspecific relationships. Some scholars believe that community stability is primarily determined by the proportion of dominant species in the community, and to a lesser extent is affected by the species richness [58]. The long-term natural succession of stands in unmanaged forests has kept the dominant species in the community relatively stable for a long time, where human management activities have once again upset this balance. In new vegetation communities, the increased complexity of interspecific relationships leads to more intense interspecific competition, which results in unstable communities [59]. As the dominant population of community succession gradually dominates and remains relatively constant, community stability is improved; thus, there is no diversity-stability relationship. In this study, except for CK, the other stands were practically single species forests, where the relationship between species diversity and community stability in the tree layer confirmed that mixed forests had higher stability than single species stands [52,60].

Conclusions
By studying the species diversity and stability of vegetation communities in natural  between different forest layers and community stability of Simpson dominance index, (C) relationships between different forest layers and community stability of Pielou evenness index, (D) relationships between different forest layers and community stability of Margalef richness index.

Conclusions
By studying the species diversity and stability of vegetation communities in natural secondary Carya dabieshanensis forests under different management intensities and durations, we were able to draw the following conclusions: 1.
Long-term forest management led to a continuous decline in species diversity in the plant communities of natural secondary forests in Dashan Mountain, which was expedited compared with intensive management. Community species diversity at both management intensities generally stabilized after 15 years, which did not align with our first hypothesis.

2.
Since there was no evidence that community species diversity translated to community stability, and intensive management did not lead to lower community stability under the same management durations, we rejected Hypotheses 2 and 3.
Finally, we believed that forest management activities would lead to a decline in community species diversity, although this did not reduce community stability. In this area, more attention may be required to focus on the coupling mechanism of community functional diversity and community stability and the driving role of the microenvironment on biodiversity and community stability. We recommend that operators appropriately retain or replant additional tree species to improve the species richness of the tree layer, which may be conducive to improving stand structures and community stability.