Although competition and facilitation both influence tree diversity1-5, their relative importance and variation with latitude remain poorly understood. Using data from 17 large forest plots, including around 2.7 million trees and over 5,400 species spanning 5° S to 47° N, we quantified the latitudinal trends of the relative importance of negative (competitive) and positive (facilitative) interactions among neighbouring tree species, accounting for three biotic and eight environmental factors. We examined whether the average neighbourhood species diversity around individuals of each focal species was larger or smaller than expected under null models. The results show that negative interspecific interactions prevailed across most plots. Near the equator, the relative proportions of species surrounded by a lower or higher than expected number of neighbours were roughly equal, but at higher latitudes, the proportions of species with a relatively higher number of neighbours declined, and those with fewer neighbours increased significantly. This latitudinal pattern can be attributed in part to reduced abundance of legumes, non-arbuscular mycorrhizal associations, and the weaker canopy nursing effect towards higher latitudes, but it was mediated by mean annual temperature. These findings reveal a previously unrecognized relative decline in facilitative interactions and increase in competitive interactions with latitude and suggest that rising temperatures could enhance facilitative effects and promote tree community diversity at higher latitudes.

Question Local species pool and assembly processes play crucial roles in maintaining soil microbial and woody plant diversity in forest ecosystems. However, their effects on beta diversity patterns remain poorly understood along an altitudinal gradient.Location Fanjing Mountain, Guizhou Province, China.Methods We established 8 1-hectare plots along an altitudinal gradient (600 m-2100 m), assessing alpha and beta diversity, species pool of woody plants, and soil microbial community biotic and abiotic conditions. Community assembly was estimated using the beta Nearest Taxon Index, beta diversity via the Jaccard index, and species pool via species richness of woody plants and soil microbes in each 1-hm2 plot. A total of 200 sampling plots (20 m & times; 20 m) were analyzed.Results There was a decrease in the alpha diversity of woody plants and soil bacteria, but an increase in the beta diversity of soil fungi and bacteria, along the altitudinal gradient. Composition dissimilarities of woody plants and soil microbes, as well as both species turnover and nested components of beta diversity, expanded with increasing altitudinal distance. Deterministic processes, particularly heterogeneous selection, primarily governed soil bacterial community assembly. In contrast, stochastic processes, mainly dispersal limitation and drift, governed woody plants and fungal communities. The local species pool emerged as a primary driver of beta diversity, indirectly influencing it through assembly processes in woody plants and soil bacterial communities. While assembly processes of woody plants and soil bacterial communities significantly positively affected their respective beta diversity, soil fungal assembly processes showed no such effect.Conclusion These findings highlight that the local species pool contributed more to beta diversity patterns than assembly processes along altitudinal gradients, advancing our understanding of biodiversity maintenance mechanisms in montane ecosystems.

The carbon economy of alpine plants, particularly the allocation of non-structural carbohydrates (NSCs), is fundamental to survival in extreme environments. However, how iconic giant herbs partition these resources among organs across steep environmental gradients remains poorly understood. We investigated the giant herb Rheum nobile at five sites on the Tibetan Plateau, quantifying NSC concentrations in roots, stems, leaves, and its prominent bracts to elucidate organ-specific allocation patterns and their key drivers. Overall, NSC concentrations were highest in roots (28.5%), followed by stems (25.1%), bracts (23.6%), and leaves (22.9%). Crucially, the drivers of NSC variation differed markedly among organs. Stem, root, and bract NSC concentrations were predominantly shaped by abiotic factors; for example, stem NSCs showed a hump-shaped relationship with elevation and mean annual temperature, while root NSCs displayed a U-shaped response to a soil fertility gradient. In contrast, leaf NSCs were mainly governed by the plant's morphological traits, exhibiting complex nonlinear relationships with plant height and stem diameter. These findings reveal a finely tuned, organ-level carbon allocation strategy in Rheum nobile , where investment in storage and support organs is strongly coupled to the external environment, while carbon status in photosynthetic leaves is regulated by plant size and architecture. This provides novel insights into the physiological mechanisms underpinning plant adaptation to high-altitude ecosystems.

Macadamia is a high-quality forest-food tree species and has now become one of the pillar industries in economic forestry in Southwest China. Understanding the characteristics of incompatibility in pairwise combinations of female and male parents, and further enriching the research cases of cultivars will help screen more high-quality parental hybrid combinations, thereby providing a basis for increasing yields. In our study, artificial pollination was performed on 12 reciprocal cross combinations of four cultivars: HAES 344, Own Choice, Hinde, and Guang 11. Fruit set, pollen germination, pollen tube growth, and embryo development were investigated. We found that the average number of fruit per inflorescence of the pollination combination 'HAES 344' d 3 'Own Choice' 2 at the fruit maturity stage was 5.13 +/- 1.88, which was significantly greater than that of other pollination combinations. In contrast, 'Own Choice' d 3 'HAES 344' 2, 'Own Choice' d 3 Guang 11' 2, and the reciprocal crosses between 'Hinde' and 'Guang 11' exhibited no effective fruit set under the tested sample size. Moreover, macadamia intercultivar incompatibility involves both gametophytic incompatibility and sporophytic incompatibility stages, manifested as pollen tube growth inhibition and abnormal endosperm development, leading to seed abortion. For the attainment of high yields, it is advisable to avoid random or indiscriminate intercultivar combinations in cultivation.

Community completeness, an emerging concept based on dark diversity, is gaining increasing attention in ecological research. However, how community completeness changes during secondary succession remains poorly understood. This study investigated patterns of observed diversity, measured as species richness (SR), dark diversity (DD), and community completeness index (CCI), as well as phylogenetic community structure, quantified as the standardized effect size of the mean pairwise phylogenetic distance (SES.MPD) and mean nearest taxon phylogenetic distance (SES.MNTD), across different second successional stages (30-, 45-, 60-, and 80-year-old stands) in Pinus yunnanensis forests. We also examined the effects of abiotic factors (soil water content, SWC; soil pH, SpH), biodiversity metrics (SR; functional dispersion, FDis; mean nearest taxon phylogenetic distance, MNTD), and soil nutrients (soil total nitrogen, STN; soil total phosphorus, STP; soil hydrolysable nitrogen, SHN; soil available phosphorus, SAP) on CCI. The results showed that SR and CCI followed similar trends, both increasing initially and peaking at the 60-year stage, followed by a decline at the 80-year stage. In contrast, DD did not differ significantly across the four successional stages. As succession proceeded, the phylogenetic structure of the community shifted from overdispersion to clustering: stands aged 30 to 60 years exhibited significant phylogenetic overdispersion, while the 80-year-old stands showed phylogenetic clustering. Ordinary least squares (OLS) regression revealed significant positive relationships between CCI and SR, FDis, STN, STP, and SHN. In contrast, SpH and MNTD were significantly negatively related with CCI. No significant relationships were observed between CCI and SWC or SAP. The final general linear model (GLM) identified MNTD and FDis as the most important predictors of CCI, accounting for 40.61% and 44.02% of the variation, respectively. From the perspective of DD and community completeness, this study demonstrates that secondary succession in P. yunnanensis forests is generally a benign process. However, progression beyond a certain stage (e.g., 80 years) may result in a decline in community completeness, which could subsequently impair ecosystem functioning. These findings provide a scientific basis for the management and conservation of P. yunnanensis forest ecosystems.

Spatial patterns in plant communities provide valuable insights into the ecological processes that sustain species coexistence. However, studies on uneven-aged mixed Pinus kesiya over-mature forests remain relatively limited. Using census data from three 1-ha permanent plots in the Weiyuanjiang Provincial Nature Reserve, we integrated spatial distribution patterns, neutral community model, and the Godron stability method to examine the spatial structure and interspecific associations of nine dominant tree species, and to infer the underlying mechanisms maintaining the natural uneven-aged P. kesiya forest. Based on importance values, nine dominant species shared among plots were identified and size-class structures of most tree generally exhibited an increasing trend. As upper-canopy dominants with stable diameter-class structures, P kesiya and Betula alnoides define the overstory niche with a stable size-class structure, while Vaccinium exaristatum and Rapanea nerifolia occupy the lower canopy and the remaining species are largely distributed within the mid-canopy layer. Most species exhibited aggregated spatial distributions, except for B. alnoides and Castanopsis fleuryi, which were largely random. Interspecific associations P. kesiya with Schima wallichii were predominantly positive or non-significant; associations with B. alnoides were mostly non-significant; and those with R. nerifolia, C. fleuryi, and Anneslea fragrans were mainly negative. Associations with V. exaristatum and Lithocarpus fenestratus included both positive and negative correlations. These patterns are closely linked to the stability and assembly processes of community across the three plots. Overall, the P. kesiya forest represents a relatively stable community in which neutral processes play a dominant role. All plots were primarily shaped by stochastic processes, with Plot B exhibiting lower stability and a higher neutral model R2 value compared to the other plots. This study enhances the understanding of the structural traits, diversity maintenance, and succession dynamics of P. kesiya forests, and provides a scientific basis for their effective conservation and sustainable management.

IntroductionAccurately assessing the natural recovery processes of forest ecosystems remains a key challenge in restoration ecology. The concept of dark diversity—the set of species absent from a site but belonging to its habitat-specific species pool—provides a novel lens for this assessment.MethodsIn this study, we developed and applied an integrated diagnostic framework that synthesizes dark diversity, functional traits, and diagnostic species. We applied this framework to a chronosequence of recovering forest ecosystems in subtropical China, representing early, middle, and late recovery stages.ResultsOur results demonstrated that the Community Completeness Index (CCI), derived from dark diversity, increased significantly during recovery, with its stabilization indicating the approach to a stable state. The framework identified stagespecific early-warning species: the absence of light-demanding, acquisitive transitional species in the mid-stage signaled successful progression, while the absence of shade-tolerant, conservative climax species in the late-stage signaled potential degradation. Crucially, analysis using Dark Diversity Affinity (DDA) revealed that the functional traits of species (e.g., seed mass, mycorrhizal type, leaf economics) were the primary filters determining species absence, exhibiting a stronger influence than local environmental conditions. These filters shifted predictably across stages, from dispersal and establishment limitations early on to competitive interactions later.DiscussionThe proposed framework translates dark diversity theory into an actionable tool for restoration. It moves beyond simple observation to diagnose recovery success, pinpoint specific bottlenecks, and inform targeted interventions such as assisted dispersal or canopy management. This provides a mechanism-based approach for guiding precision restoration in forest ecosystems.

Dry-hot valleys represent underutilized marginal lands for industrial crop afforestation, where non-structural carbohydrates (NSCs) are critical for sustaining plant productivity and resilience to extreme aridity and nutrient limitation. However, how afforestation alters NSC dynamics to support industrial crop performance following savanna conversion remains poorly understood. We investigated this in China’s Yuanmou dry-hot valley using 70 sampling plots, evaluating the effect of six high-value industrial plantation species (Leucaena leucocephala, Eucalyptus camaldulensis, Azadirachta indica, Dodonaea viscosa, Jatropha curcas, Dalbergia sissoo) alongside native savanna reference vegetation (Albizia kalkora) on NSC allocation in leaves and roots based on interactive effects of soil properties, microbial diversity, and leaf functional traits. Afforestation significantly enhanced leaf NSCs in all industrial species compared to Albizia kalkora, with Dodonaea viscosa exhibiting the highest levels. In contrast, effects on root NSC content varied considerably among species. Soil microbial diversity strongly influenced NSC allocation: leaf NSCs showed unimodal relationships with bacterial α-diversity (Chao1 index) and fungal β-diversity, and a U-shaped response to bacterial β-diversity. Acquisitive leaf traits correlated with these microbial-mediated leaf NSC patterns, while conservative traits drove root NSC variation. Soil fertility modulated microbial effects, enhancing benefits of bacterial β-diversity but reversing fungal diversity impacts. Structural equation modeling identified bacterial α-diversity and fungal β-diversity as primary drivers of leaf and root NSC variation, respectively, with bacterial β-diversity exerting a strong positive direct effect on leaf NSCs. Specific microbial phyla (e.g., Firmicutes, Chytridiomycota) and functional groups (e.g., nitrifying bacteria, plant saprotrophs) were key contributors to leaf NSC regulation, whereas sulfate-respiring bacteria and animal pathogens influenced root NSCs. Keystone bacterial taxa, but not fungal, differentially regulated leaf NSC content. Our findings reveal microbial-mediated mechanisms of NSC allocation and suggest that managing specific soil microbial communities can optimize carbon strategies in dryland afforestation.

Intraspecific trait variation (ITV) drives plant species performance in natural communities and can have substantial influences on multiple ecological processes. However, current studies on plant belowground (root) ITV are limited in trait coverage, spatial scale and biome type. Here, by conducting in-situ measurements of root traits in three forests from temperate to subtropics and compiling a large-scale dataset of root traits, we explore the latitudinal pattern of root ITV among species within natural plant community, its drivers and implications. Our results show that intraspecific variation constitute a substantial component of the total variation in root traits, and the physiological trait, root exudation, exhibit the highest ITV among traits. The extent of root ITV tends to increase with root diameter across species, with thick-root species having higher root ITV than thin-root species. We also find a significant latitudinal gradient for species' intraspecific root trait variation in natural communities, with increasing ITV towards the equator. These findings can simplify the incorporation of root ITV in trait-based ecology, and contribute to understanding of plant species performance and community assembly under environmental changes.

Forest conversion to agroforestry systems has the potential to impact soil ecosystem functions in terrestrial ecosystems and optimize crop management, which may play a critical role in sustaining soil ecosystem multifunctionality (EMF, defined as simultaneously provision of multiple ecosystem functions) in a subtropical agroforestry system. However, the impact of sowing crops on soil EMF and its relationship with soil microbe remain poorly understood within macadamia-based agroforestry systems. Here, we investigated 50 plots, including macadamia monocultures and macadamia-based agroforestry systems intercropping with dasheen, konjac, and maize, as well as adjacent primary forest in the southwest of Yunnan Province, China. Our main objective was to assess the effects of soil microbial community structure and abiotic and biotic factors on soil EMF following forest conversion. We found that forest conversion significantly decreased soil EMF and multiple individual functions. Interestingly, the macadamia intercropping dasheen system exhibited advantages in maintaining soil EMF resilience. Soil EMF was negatively correlated with increasing soil bacterial diversity, soil bulk density, and soil pH, while positively correlated with increasing soil bacterial network complexity and woody aboveground biomass. Further analyses indicated that soil bacterial network complexity was a primary contributor to soil EMF, mediating the effect of woody aboveground biomass on soil EMF. Acidobacteriota and Basidiomycota were identified as important predictors of declines in soil EMF and most individual ecosystem functions. In contrast, others like Gemmatimonadota, Firmicutes, and Chytridiomycota were associated with increases in soil EMF and most individual ecosystem functions. Furthermore, macadamia-based agroforestry systems were found to result in lower soil bacterial network complexity and woody aboveground biomass, and higher soil bulk density, leading to reduced soil EMF. These findings highlight the potential diminishment of soil EMF associated with the development of macadamia-based agroforestry systems but also suggest the effective management of microbial taxa could enhance soil bacterial network complexity, ultimately promoting resilience in soil EMF.

A thorough understanding of tree growth responses to drought across temperature and moisture gradients is crucial for assessing forest health and growth dynamics under ongoing climate change. This study investigated the growth-climate responses of Pinus yunnanensis and its resilience to drought with varying intensities across four distinct climate zones (hot-dry, warm, cold-dry, and humid) on the Yunnan Plateau, southwestern China. The radial growth of P. yunnanensis exhibited stronger moisture limitations and greater drought susceptibility in hot-dry and warm regions compared to cold-dry and humid regions. Severe drought during the main growing season diminished growth resistance and resilience, particularly in hot-dry regions. In contrast, drought during the early growing season had a positive effect on tree recovery and resilience in warmer regions. Post-drought growth recovery varied across climate zones, with a longer recovery time in drier regions compared to wetter regions. The sensitivity of growth to seasonal climate affects growth resilience, particularly in xeric areas. With increased moisture restrictions, the effects of seasonal temperature sensitivity on growth resilience were gradually being replaced by the early growing season moisture sensitivity. These findings improve our understanding of the habitat-specific growth resilience of P. yunnanensis to extreme drought events and provide quantitative guidance for implementing forest management strategies in the context of regional climate change.

Understanding the drivers of soil microbial diversity along altitudinal gradients is critical for ecosystem functioning responses to environmental change. This study, conducted in the Fanjing Mountain Nature Reserve, China, investigated the relative roles of plant functional composition, plant species richness, phylogenetic diversity, and soil nutrients in shaping soil microbial diversity and community composition across an altitudinal gradient. Our findings revealed that plant leaf traits exhibited distinct shifts, with resource acquisition strategies (e.g., higher leaf nitrogen (LN), leaf phosphorus (LP), and specific leaf area (SLA)) dominating below 1000 m, while resource conservation strategies (e.g., higher leaf dry matter content (LDMC) and leaf mass per area (LMA)) prevailed at higher altitudes. Soil bacterial alpha-diversity decreased with increasing altitude, while soil fungal alpha-diversity exhibited a unimodal pattern, peaking at intermediate altitudes (similar to 1000 m). Plant functional composition, particularly community-weighted mean of LP (CWM (LP)), emerged as a key driver of soil microbial diversity. CWMSLA positively influenced bacterial alpha-diversity, while CWMLN, and CWMLMA showed strong negative effects. Notably, CWMLC, CWMLN, and CWMLP had the greatest impact on fungal beta-diversity, surpassing the effects of soil nutrients, FD, PD, and SR. Furthermore, plant functional traits strongly affected the composition of dominant microbial phyla, including Proteobacteria, Actinobacteriota, Basidiomycota, and Mortierellomycota, highlighting their role in mediating plant-microbe interaction. These findings highlight the critical role of plant functional traits in shaping soil microbial communities, independent of soil nutrient levels and plant diversity in mountain ecosystem and emphasize the importance of incorporating plant functional traits into biodiversity and ecosystem function studies.

Soil microbial communities are key factors in maintaining ecosystem multifunctionality (EMF). However, the distribution patterns of bacterial diversity and how the different bacterial taxa and their diversity dimensions affect EMF remain largely unknown. Here, we investigated variation in three measures of diversity (alpha diversity, community composition and network complexity) among rare, intermediate, and abundant taxa across a latitudinal gradient spanning five forest plots in Yunnan Province, China and examined their contributions on EMF. We aimed to characterize the diversity distributions of bacterial groups across latitudes and to assess the differences in the mechanisms underlying their contributions to EMF. We found that multifaceted diversity (i.e., diversity assessed by the three different metrics) of rare, intermediate, and abundant bacteria generally decreased with increasing latitude. More importantly, we found that rare bacterial taxa tended to be more diverse, but they contributed less to EMF than intermediate or abundant bacteria. Among the three dimensions of diversity we assessed, only community composition significantly affected EMF across all locations, while alpha diversity had a negative effect, and network complexity showed no significant impact. Our study further emphasizes the importance of intermediate and abundant bacterial taxa as well as community composition to EMF and provides a theoretical basis for investigating the mechanisms by which belowground microorganisms drive EMF along a latitudinal gradient.

Traditional biodiversity measures, such as species richness (SR), are inadequate for accurately capturing the completeness of local plant communities. In contrast, community completeness, derived from the concept of dark diversity offers a more robust framework for cross-ecosystem biodiversity comparisons. Dark diversity and community completeness are gaining prominence in both theoretical and applied ecology. In this study, we assessed SR, dark diversity, and community completeness across four primary forest types, Pinus yunnanensis, Quercus oxyodon, Tsuga dumosa, and Abies georgei, along an elevation gradient of 2300-3588 m. We further investigated the effects of various abiotic factors, multiple biodiversity metrics, and soil nutrients on the community completeness index (CCI) in the Baima Snow Mountain National Nature Reserve. Our results revealed that distinct distribution patterns of SR, dark diversity, and CCI among vegetation types. For example, although both SR and dark diversity were relatively low in P. yunnanensis forest, its CCI was high. Overall, CCI was highest in P. yunnanensis and T. dumosa forests, followed by Q. oxyodon forest, with A. georgei forest having the lowest value. Linear mixed-effects models (LMMs) identified functional diversity (Functional Dispersion index [FDis]), phylogenetic diversity (standardized effect size of mean pairwise distance [SES.MPD]), and soil pH (SpH) as the primary drivers of CCI. Hierarchical partitioning analysis further indicated that biodiversity dimensions-FDis and SES.MPD-jointly explained most of the variation in CCI. For a management perspective, integrating CCI with other ecological indicators can support more effective conservation strategies. We recommend prioritizing T. dumosa forest for protection and focusing, restoration and monitoring efforts on A. georgei forest. Enhancing community completeness will require attention to multiple dimensions diversity and SpH. This study provides new insights into local biodiversity patterns and offers a theoretical scientific basis for biodiversity management in protected areas.

Monsoon evergreen broad-leaved forests (MEBFs) represent one of the most species-rich and structurally complex vegetation types, and one of the most widely distributed forests in Yunnan Province, Southwest China. However, they have yet to undergo a comprehensive analysis on their community diversity, spatial differentiation patterns, and underlying drivers across Yunnan. Based on extensive field surveys during 2021–2024 with 548 MEBF plots, this study employed the Unweighted Pair Group Method for forest community classification and Non-metric Multidimensional Scaling for ordination and interpretation of community–environment association. A total of 3517 vascular plant species were recorded in the plots, including 1137 tree species, 1161 shrubs, and 1219 herbs. Numerical classification divided the plots into 3 alliance groups and 24 alliances: (1) Castanopsis–Schima (Lithocarpus) Forest Alliance Group (16 alliances), predominantly distributed west of 102°E in central-south and southwest Yunnan; (2) Castanopsis–Machilus (Beilschmiedia) Forest Alliance Group (6 alliances), concentrated east of 101°E in southeast Yunnan with limited latitudinal range; (3) Castanopsis–Camellia Forest Alliance Group (2 alliances), restricted to higher-elevation mountainous areas within 103–104° E and 22.5–23° N. Climatic variation accounted for 81.1% of the species compositional variation among alliance groups, with contributions of 83.5%, 57.6%, and 62.1% to alliance-level differentiation within alliance groups 1, 2, and 3, respectively. Precipitation days in the driest quarter (PDDQ) and precipitation seasonality (PS) emerged as the strongest predictors of community differentiation at both alliance group and alliance levels. Topography and soil features significantly influenced alliance differentiation in Groups 2 and 3. Collectively, the interaction between the monsoon climate and topography dominate the spatial differentiation of MEBF communities in Yunnan.

Accurately estimating aboveground biomass (AGB) is crucial for understanding terrestrial carbon cycling and informing climate policy. China’s diverse topography and rich vegetation types make it a significant contributor to the global carbon stock. However, existing AGB products often lack sufficient spatial resolution, data consistency, and accessibility to fully capture biomass patterns across the country’s varied ecosystems. Here, we present a 30 m-resolution, nationwide dataset of AGB density (AGBD) for China in 2020, which integrates multiple vegetation types. This product was developed using openly accessible, multi-source remote sensing data including LiDAR, optical, and radar imagery and enables consistent mapping of forests, grasslands, shrublands, croplands and wetlands. Systematic validation using field observations, national statistical yearbooks, and spatial distribution patterns, alongside comparisons with existing AGB products, demonstrates the model’s high accuracy with an average R2 of 0.85, RMSE of 31.26 Mg/ha and rRMSE of 50.04%, estimating a total carbon stock of 20.20 Pg across China’s vegetation areas. This dataset provides an updated, comprehensive baseline for biomass assessment, supporting carbon accounting and biodiversity monitoring in China.

Macadamia (Macadamia integrifolia)-based agroforestry systems,widely adopted in Yunnan Province, China since 1981, are critical for soil organic carbon (SOC) sequestration. However, their effects on SOC stocks and coupling mechanisms with soil water dynamics remains poorly understood. This study quantified changes in SOC and soil water stocks following the conversion from primary forest to macadamia monoculture and intercropping systems (dasheen, konjac, and maize) in Yongde County, southwestern Yunnan province. Soil profiles (0-100 cm, 0-20 cm, and 20-100 cm) were analyzed to assess vertical SOC and soil water stocks interactions and driving factors. Results revealed that macadamia-based agroforestry systems reduced significantly SOC stock compared to primary forest (261 t/hm2) but enhanced soil water stock. Among intercropping systems, macadamia+dasheen plantation exhibited the highest recovery of SOC and soil water stock. Surface soil layers (0-20 cm) showed strong SOC-water coupling, whereas this relationship weakened in deeper layers (20-100 cm). Simultaneously, woody above-ground biomass directly increased SOC stock, while specific leaf area and leaf phosphorus content regulated SOC stock directly or indirectly via woody above-ground biomass. Conversely, herb above-ground biomass negatively influenced SOC stock in surface soil layer, but had no significant effect in the deeper layer. Woody above-ground biomass exhibited a consistent negative influence on soil water stock across all soil layers, with the standardized path coefficients of -1.23 (0-100 cm), -1.153 (0-20 cm), and -1.23 (20-100 cm). This relationship was further modulated by indirect effects mediated through specific leaf area and leaf phosphorus content. This study provides mechanistic insights into the interdependencies of SOC and water stock in the subtropical macadamia-based agroforestry systems, providing valuable guidance for sustainable carbon management practices in China's subtropical mountainous region.

Studying the relationship between plant functional groups and soil microbial functional groups, abiotic and biotic factors are important for understanding the mechanisms of species coexistence and ecological processes among biological communities, especially in species-rich tropical rainforests. This study utilized the Competitor, Stress-tolerator, and Ruderal (CSR) theory for plant classification, and FAPROTAX and FUNGuild for soil bacterial and fungal functional guild analysis, and combined with abiotic and biotic factors. We use correlation analysis, multivariate regression analysis, random forest model, redundancy analysis and variance decomposition to analyze the data. (1) A total of 316 woody plant species were classified into four functional groups. The intermediate (Int, 53.48

Diversity-biomass relationships (DBRs) in terrestrial ecosystems tend to vary across spatial scales, but, particularly in hyperdiverse forests, the mechanisms driving these trends remain uncertain. Until now, few have simultaneously investigated the connections between tree species diversity, stand structural diversity, mycorrhizal associations, and ecosystem functioning. In addition, DBRs have only been studied at limited spatial scales, with limited focus on the direct and indirect effects of environmental factors. We addressed these research gaps using a 30-ha forest dynamics plot located in Pu’er City, Southwest China. Through piecewise structural equation models, we quantified the direct effects of tree species diversity (α, β, γ), stand structural diversity, mycorrhizal associations (AM, EcM), and the environmental factors (soil fertility and topography), as well as the indirect effects of the environmental factors on aboveground tree biomass across spatial scales ranging from 400 to 230,400 m2. We hypothesized that complex interactions among these factors underpin the variation in DBRs in natural ecosystems across spatial scales. Our results showed that environmental conditions indirectly affected the tree biomass via changes in tree species diversity, and these effects became stronger as the spatial scale increased. At small to moderate spatial scales, environmental factors were more predictive of tree biomass than tree species diversity (or its components); the effects of stand structural diversity on biomass also gradually increased with spatial scale. Conversely, from the intermediate to the largest spatial scales, mycorrhizal associations gradually became the best predictors of DBR dynamics. Our research offers novel empirical evidence demonstrating the importance of environmental conditions, structural diversity, and mycorrhizal associations in shaping cross-scale DBRs. Future comprehensive studies should consider these factors to assess the mechanisms shaping scale-dependent DBRs in complex natural ecosystems.

Soil microbes mediate soil organic carbon (SOC) storage by affecting microbial residue carbon (C) in terrestrial ecosystems, a main source of SOC. However, how microbial residue C accumulation changes during the conversion of subtropical primary forest to plantations and the role of soil microbial community structure in regulating these changes remain unclear. Here, effects of conversion from forest on microbial residue C accumulation in three plantations, including tea (Camellia sinensis var. assamica), walnut (Juglans regia), and macadamia (Macadamia integrifolia). We assessed amino sugars content and the relative importance of soil microbial community structure, soil fertility, and aboveground biomass in four forest types in southwest Yunnan province, China. Our results show that total microbial residue C accounted for 27.8 % of SOC on average across all forest types, and fungal residue C contributed more to SOC than bacterial residue C across different forest types. Conversion from primary forest decreased fungal, bacterial, and total microbial residue C content in all plantation types except tea. Microbial residue C contents was significantly lower in the macadamia plantation than in the primary forest and the tea plantation. Fungal and total microbial residue C content increased with increasing soil microbial community composition, soil fungal abundance, and soil fertility. Soil fertility regulated fungal, bacterial, and total microbial residue C accumulation, which was major determinant of microbial residue C via its influence on soil fungal and bacterial abundance. Overall, we found that soil fungal abundance contributed more to microbial residue C accumulation than change in soil microbial community composition. We also found that soil bacterial abundance was negatively correlated with fungal and microbial residue C, while soil fungal abundance was positively related to microbial residue C accumulation. Ascomycota was most important phylum regulating microbial residue C. These findings suggest that soil fungal and bacterial abundance and soil fertility are the principal regulators of microbial residue C accumulation during the conversion from primary forest to different plantations.