Illuminating Plant Community Assembly on Karst Mountain Road Slopes through Plant Traits and Environmental Filters

Abstract

Understanding how assembly processes shape local plant assemblages from the potential species pool is crucial for biodiversity conservation and revegetation. Mountainous regions are global biodiversity hotspots with high levels of diversity, concentration, and vulnerability. Road construction in these areas poses ecological challenges, including habitat loss and reduced biodiversity. Feature-based ecology highlights non-biological filtering as a key driver of habitat-specific community formation. Analyzing trait structures and their association with the environment can reveal community assembly processes under specific environmental conditions. However, quantifying species-environment-traits interactions during community assembly on roadside slopes is still underexplored. In our study, 76 naturally recovered roadside slopes, 656 self-established plant communities and 113 plant species across ten functional traits, along with their environmental associations, in the karst mountain region of southwestern China, were examined. Our findings show that there are still abundant native plants with colonization potential settled on steep roadside slopes in karst mountain areas. Diffusion constraints stemming from distance to the core species pool, elevation, and differences in adjacent vegetation types emerged as key factors causing variations in species composition of self-established communities. The slope environment exerts strong selective pressures leading to a convergence pattern in traits related to dispersal and colonization while showing a divergence pattern in traits linked to competitive strategies and regeneration. These findings identify critical functional traits and environmental factors shaping roadside plant communities and illustrate the predictability of environmental filtering and fundamental community assembly. Overall, our study sheds light on the intricate interactions among assembly processes, functional traits, and environmental factors driving local plant assemblages in mountainous regions, providing insights for effective diversity conservation and revegetation strategies.

1. Introduction

Mountain ecosystems are widely recognized as cradles and sanctuaries for a diverse array of species [1]. However, both rural development and forest conservation management require essential infrastructure support, particularly linear projects like roads and railways [2]. Nonetheless, road construction inevitably inflicts irreversible damage on forest ecosystems, causing issues such as habitat fragmentation, soil erosion, and habitat loss. It has long served as the primary conduit for the continuous spread of invasive alien plants. Some reports showed that road slope expansion reaches 200–300 million square meters annually, exclusively in China [3]. However, restoring these slopes has long posed a challenge for ecological rehabilitation practitioners. Despite various innovative engineering techniques for restoration [4,5,6,7], these methods often lead to unnatural and unsustainable post-recovery vegetation [2]. A more effective approach lies in understanding spontaneous processes like diffusion, colonization, and succession, pivotal for rehabilitating disrupted sites [8,9,10,11].

Regional species composition primarily hinges on environmental factors such as climate, elevation, and soil [12,13,14,15]. It was not until recent advancements in trait-based community ecology that the intricate interplay between community composition and plant traits at the local scale came into sharper focus [16,17]. Communities assemble as both biotic and abiotic environments act as filters, allowing only those adapted to local conditions to establish from the potential colonizers, promoting establishment [18,19]. Plant traits interact with these filters, shaping trait distribution among coexisting species [16,20]. Biotic and abiotic filters drive trait convergence toward optimal conditions, shaping species assemblages [19,21], and competition partially restricts species with similar ecological niches, scattering plant traits [22]. Analyzing the interactions between the environment and traits, as well as trait distribution, can help decipher community assembly [21,23,24].

However, this interaction varies with environmental conditions, succession stages, and research scales [25,26,27]. Studies of roadside slope plant communities suggest dispersal, colonization, and drought adaptation traits play pivotal roles in early successional communities [10,28,29]. As succession proceeds, traits associated with resource competition gain significance [30,31]. However, research quantifying trait-environment relationships on roadside slopes remains limited, resulting in uncertainty in trait distribution patterns, whether divergent or convergent.

The karst mountainous region in southwestern China boasts distinct terrain and climate, abundant natural resources and biodiversity. Nonetheless, challenging topography, steep slopes, and sluggish plant growth hinder robust vegetation cover establishment. To comprehend community assembly, a specific karst mountain range with an ample potential species pool was chosen. We explored species composition in self-established vegetation on seven naturally regenerating road slopes to determine the interplay between trait structure and environmental attributes in determining community structure in succession.

Our threefold objectives are: (1) To comprehend self-established vegetation composition and distribution on karst mountain roadside slopes; (2) to delve into the contributions of the environment and plant traits in community assembly and overall patterns and (3) to comprehensively examine mechanisms facilitating self-sustaining community development and offer practical revegetation recommendations to mitigate negative impacts.

2. Methods

2.1. Study Area

The study was conducted in the Baima Mountains of Wulong, Chongqing, China (107°28′–107°40′ E, 29°10′–29°24′ N). This area is characterized by distinct karst topography and elevations ranging from 167 to 1900 m. It falls within a subtropical humid monsoon climate, with average annual temperatures ranging from 15 °C to 18 °C and annual precipitation ranging from 1200 mm to 1400 mm. The total annual sunshine duration is about 1100 h. Due to its distinctive climate and intricate topography, this area conserves extensive subtropical evergreen broad-leaved forests, establishing it as an ideal habitat for a wide variety of plants. Within the elevation range of 1200m to 1800m, it harbors a rich species pool, including various relict plants such as Cathaya argyrophylla, Davidia involucrata, and Taxus wallichiana. Designated as the Baimashan Municipal Nature Reserve, this area holds significant ecological importance (Figure 1).

2.2. Sampling Design

The investigation was conducted from July 2021 to July 2022. We selected seven mountainous roads for comprehensive research and analysis. These roads are currently in normal operation and have undergone natural recovery without intervention for over five years. Among these, R2, R5, and R7 exhibited an east-west orientation, while R1, R3, R4, and R6 predominantly followed a north-south direction (Figure 1). The slopes under study ranged in height from 5 to 20 m. At each sampling point, three perpendicular transects were established originating from the road, and a plot was established at the top, middle, and bottom, respectively. In total, nine plots were established per point. Within each plot, we documented species diversity and counted individuals, averaging the height of three randomly selected individuals per plant species, and visually assessing species coverage. The selection of plots on each road was based on the degree of variation in the species composition of indigenous plant communities along the slope, with a minimum of 5 points required. In total, 76 sample points and 684 plots were examined to complete our analysis (Table 1).

2.3. Materials and Methods

2.3.1. Species Identification

The identified species were classified utilizing the established resources ‘Flora of China’ (http://www.iplant.cn/frps, accessed on 1 September 2022). Additionally, information about alien species was sourced from the ‘Chinese Invasive Alien Species Database’ (http://www.iplant.cn/ias/, accessed on 1 September 2022). Our data analysis relies on the calculation of plant importance values (IV), which are determined using the formula: IV = (relative abundance + relative frequency + relative coverage)/3 [32].

2.3.2. Trait Measurement

After a year-long investigation, plants’ leaves, roots, and seeds were collected from the field to assess their traits [33]. Extreme cases of high abundance (with frequencies exceeding 95%) or low scarcity (with frequencies below 5%) were excluded. Ultimately, a comprehensive set of ten functional traits was selected for measurement, which are widely recognized for their pivotal role in plant colonization [16]. For each species, five individuals were randomly sampled for testing. Leaves from the outer canopy were enclosed in airtight bags along with accompanying branches and transported to the laboratory. Herbaceous were collected with their entire vegetation for subsequent analysis. We scanned ten mature and intact leaves of each species using a scanner, and then calculated the leaf area using ImageJ and AutoCAD 2016. Leaf biomass was determined by drying the leaves in an electric constant-temperature oven (model DHG-9070A) set at 75 °C for 48 h until a stable weight was achieved.

Subsequently, we calculated the specific leaf area (SLA), which is the ratio of leaf area to leaf dry mass. A substantial quantity of fresh leaves was collected from five carefully chosen plant specimens. After detaching the petioles, the leaves underwent a drying process, followed by grinding into a fine powder to assess leaf nutrient content (measured in mg g⁻¹). This analysis included leaf carbon concentration (LCC), leaf nitrogen concentration (LNC), leaf phosphorus concentration (LPC), and leaf potassium concentration (LKC). We evaluated the dry weight and length of ten lateral roots from five plants and calculated the specific root length (SRL), defined as the ratio of root dry weight to root length. Seeds collected were measured for the weight of a thousand seeds or a hundred seeds to determine the average dry seed weight. Additionally, ten seeds were randomly selected to measure their dimensions, including length (L), width (W, μm), and height (H, μm), using a dissecting microscope (Leica M125 C). These measurements enable us to calculate both the flatness index (FI) and the eccentricity index (EI) using the formulas FI = (L + W)/2H and EI = L/W, respectively [34]. Finally, we conducted a comprehensive examination of seed appendages, categorizing them into three types: winged (SSF.wing), hairy (SSF.hair), and mucilaginous (SSF.sucilage) [10]. Seeds lacking distinctive appendages were labeled as SSF.no, while ferns were identified as “spores”. The various types of trait indicators and their corresponding data acquisition methods are presented in Table 2.

2.3.3. Environmental Factor

We measured twelve environmental variables at each plot, including three established topographical factors known to often influence plant dispersion and colonization: altitude (ALT), aspect (ASP), and slope (SLO). Additionally, two variables related factors to the potential species pool were considered, encompassing adjacent vegetation types (VEG) and distance from the core species pool (DSP). The vegetation types were classified based on the composition and structure of vegetation communities, including coniferous forests (VEGcon), theropencedrymion (VEGthe), and broad-leaved forests (VEGbro). Seven soil variables, recognized for their influence on plant growth and trait variation, were examined: available soil nitrogen (SAN), available soil potassium (SAK), available soil phosphorus (SAP), soil organic matter content (SOM), soil thickness (STH), soil water content (SWC), and pH. All samples were collected from a depth range of 0–20 cm below the surface and were naturally air-dried. The fraction passing through a 0.15 mm mesh sieve was collected for subsequent analysis.

2.4. Statistical Analysis

2.4.1. Assessment of Plant Trait and Structural Attributes

We categorized all species into functional groups based on their respective families (families with frequencies exceeding 5 were listed separately, while others were grouped under ‘else’) and analyzed variations in plant traits among distinct functional groups using either one-way ANOVA or the Kruskal–Wallis test. To calculate the average Mean Pairwise Distance index (MPDobs) for plant traits across different environmental conditions on road slopes, we employed the “picante” package. For species located at the terminal nodes of the functional trait dendrogram underwent 999 random permutations. This procedure allowed us to calculate the average pairwise trait distance (MPDnull) between species pairs within the sample plots, based on the random null model. The observed values were then standardized using the results from the random distribution. To quantify the pattern of functional dispersion, we utilized the Standardized Effect Size of MPD (SESMPD), which is determined by the equation [35]:

(1)$\mathrm{S}\mathrm{E}\mathrm{S}\mathrm{M}\mathrm{P}\mathrm{D}=\frac{({\mathrm{M}\mathrm{P}\mathrm{D}}_{\mathrm{o}\mathrm{b}\mathrm{s}}-\mathrm{m}\mathrm{e}\mathrm{a}\mathrm{n}({\mathrm{M}\mathrm{P}\mathrm{D}}_{\mathrm{n}\mathrm{u}\mathrm{l}\mathrm{l}}\left)\right)}{\mathrm{S}\mathrm{D}\left({\mathrm{M}\mathrm{P}\mathrm{D}}_{\mathrm{n}\mathrm{u}\mathrm{l}\mathrm{l}}\right)}$

The absolute value of SESMPD indicates the extent of difference between the observed and simulated MPD. MPD_obs represents the observed MPD, while MPD_null and SD represent the mean and standard deviation, respectively, of the 999 simulated MPD values. A positive SESMPD suggests a convergence of functional trait patterns among species within the sample plot, highlighting the significant role of environmental filtering in driving community assembly. Conversely, a negative SESMPD indicates a dispersion of functional trait patterns among species within the sample plot, emphasizing the dominant influence of limiting similarity. An SESMPD value of 0 signifies a stochastic arrangement of functional trait patterns among species within the sample plot, showcasing the combined effects of both environmental filtering and limiting similarity.

2.4.2. Linking Species Distribution and Plant Traits with Environment Variables

To comprehend variations in self-established species composition across different environments, we employed redundancy analysis (RDA, scaling = 2) using the ‘rad()’ function in the ‘vegan’ package. This analysis helps us identify key environmental factors and quantify their effects on species distribution. We also conducted RLQ analysis (linking R-mode to Q-mode) and fourth-corner analysis to establish relationships between traits and environmental variations [36]. To assess overall correlations among the R, L, and Q tables, we explored these correlations between plant traits and the environment through four models (models 2, 4, 5, and 6). Model 2 tests the hypothesis that species distribution is independent of the environment. Model 4 tests the hypothesis that plant traits do not affect species composition within a specific environment. Model 5 examines the null hypothesis that species distribution is unaffected by both plant traits and the environment. Finally, model 6 (a combination of models 2 and 4) evaluates the null hypothesis that species distribution remains unaffected by the environment [37]. Certain less influential environmental factors were excluded from the analysis.

2.4.3. Mechanisms of Community Assembly

To discern the primary driving factors underlying community assembly, specifically the relative importance of niche-based and dispersal-based processes, the PER SIPER method was introduced [38]. This method involves calculating an empirical SIMPER profile along with three distinct null SIMPER profiles to dissect plant community assembly processes: (1) Niche-Controlled Model, In this model, the distribution of taxonomic units is exclusively constrained by the availability and extent of ecological niches within each sample, without considering the dispersal potential of these units; (2) Dispersal-Controlled Model, This model assumes that the distribution of taxonomic units is solely limited by the dispersal potential of these units, without accounting for the quantity and scope of accessible ecological niches within each sample; (3) Niche- and Dispersal-Controlled Model, in this model, the distribution of taxonomic units is concurrently influenced by both the quantity and breadth of ecological niches within each sample, as well as the inherent dispersal potential of the taxonomic units themselves. Subsequently, the sum of squared logarithmic deviations (E) is calculated between the empirical SIMPER and each null SIMPER model, resulting in three distinct sets of E values. A lower E value indicates a closer resemblance between the empirical SIMPER and its corresponding null SIMPER, while a higher E value suggests a more significant discrepancy. A reduced distribution of E values from a particular null model indicates a higher likelihood that the prevailing community assembly process aligns predominantly with the mechanism associated with that null model. A lower E value signifies a stronger resemblance between the empirical SIMPER and the null SIMPER, whereas a higher E value indicates a greater disparity. A lower distribution of E values from a specific null model suggests a higher probability that the corresponding assembly process largely governs community assembly.

(2)$\mathrm{E}={\mathrm{l}\mathrm{o}\mathrm{g}}_{10}\left({\sum }_{\mathrm{i}-1}^{\mathrm{i}=9}{\left({\overline{\mathsf{\gamma }}}_{\mathrm{i}\left(\mathrm{n}\mathrm{u}\mathrm{l}\mathrm{l}\right)}-{\overline{\mathsf{\gamma }}}_{\mathrm{i}\left(\mathrm{o}\mathrm{b}\mathrm{s}\right)}\right)}^2\right)$

We conducted all statistical analyses using the R 4.03 statistical platform. For investigations involving slope aspects, the ASP index is transformed into binary 0–1 data.

3. Results

3.1. Composition of Self-Established Plants on Road Slopes

In the study of 684 sampled plots, a total of 231 species spanned across 86 families and 170 genera, accounting for 17.76% of the regional wild species diversity. The species richness within each plot varies from 1 to 12, with an average of five species. These findings revealed a pronounced presence of valuable native plant resources on the steep slopes of karst mountainous terrain, indicating considerable potential for colonization. Moreover, indigenous species displayed heightened adaptability in contrast to their alien counterparts, constituting 98.3% of the overall composition. Among them, the three most prevalent plants are Miscanthus sinensis, Hydrangea strigosa, and Patrinia villosa, distributed across 18.6%, 10.53%, and 10.23% of plant communities, respectively. Remarkably, no species with an exceptionally high frequency (>95%) were identified, while, species with sporadic occurrences (<5%) comprise a significant proportion of 51%, totaling 118 species. The families most abundant in tree species encompass Ericaceae, Lauraceae, and Cupressaceae. Noteworthy diversity in shrub species is observed within Rosaceae, Adoxaceae, and Ericaceae. Herbaceous species showcase notable diversity in Asteraceae, Gramineae, Urticaceae, and Leguminosae.

3.2. Plant Distribution and Environmental Characteristics

The RDA plot illustrates the influential role of key environmental factors, including ALT (altitude), ASP (aspect), VEG (adjacent vegetation types, VEGthe, VEGcon, VEGbro), DSP (distance from the core species pool), STH (soil thickness), SAK (available soil potassium) and SMC (soil moisture content) on the variation in self-establishing plant community composition along road slopes within the study area (R² = 11.31%, p = 0.001, as depicted in Figure 2). The combined explanatory variance of the first and second axes total 53%. Notably, DSP (27.7%), ALT (22.39%), and VEG (19.73%) contribute significantly to explanatory percentages, whereas ASP (4.25%) and SAK (1.95%) fall below the 10% threshold. Certain species display distinct habitat preferences. For example, Rhododendron coeloneurum, Hydrangea strigosa, and Aster ageratoides primarily inhabit high-altitude broadleaf forest slopes where they own rich species diversity. These slopes typically exhibit elevated soil moisture levels and relatively higher concentrations of available soil potassium. In contrast, Rhododendron bachi, Pinus massoniana, and Dryopteris erythrosora are commonly found on mid-altitude slopes with thicker, moister soil profiles, contributing to diverse species communities within coniferous forests. Conversely, species such as Rhus chinensis, Eriophorum comosum, and Miscanthus sinensis tend to favor slopes associated with low-altitude coniferous or theropencedrymion forests, which have a more limited species range. These slopes typically possess shallower soil profiles and are prone to drought conditions. Meanwhile, Buddleja indleyana, Rubus setchuenensis, and Salix luctuosa predominantly occupy mid to low-altitude slopes with reduced species diversity within the species pool. These slopes typically feature thinner, arid soil profiles and are adjacent to the Theropencedrymion forest.

We conducted an in-depth analysis of three pivotal environmental factors: adjacent vegetation types (VEG), altitude (ALT), and distance from the regional core species pool (DSP), which play a central role in shaping the diversity of plant communities along road slopes. The analysis of PER-SIMPER results yielded three consistently distributed E values, as depicted in Figure 3. It illustrates that the E values for both the dispersal-controlled model and the niche- and dispersal-controlled model are notably low, influenced by various environmental variables—such as VEG, ALT and DSP. This observation indicates that the principal driver of community assembly is dispersal limitation, rather than niche limitation. Even after a restoration period spanning 5 to 20 years, this discovery highlights the critical role of seed dispersal in shaping the composition of plant communities on mountainous road slopes.

3.3. Plant Trait Characteristics and Structure

In addition to LCC, LNC and LPC, notable variations in plant traits are observed among different plant groups (Figure 4). When assessing functional trait dispersion using mean SESMPD, 70% of traits exhibited a strong tendency toward excessive dispersion (SESMPD > 0, Figure 5). Notably, traits such as SRL (55%), SLA (51%), EI (54%), LNC (54%), LKC (54%), and SSF (82%) primarily demonstrate overdispersion, supporting the concept of similarity constraints. Conversely, attributes including SM (87%), FI (66%), LCC (54%), and LPC (67%) show a tendency toward aggregation, as shown in Figure 5. This suggests a robust filtering mechanism in the road slope environment that affects plant seed traits and their carbon and phosphorus storage capacities.

3.4. Root Trait Distributions along Environmental Gradients

The RLQ analysis links plant traits, species distribution, and environmental variables, revealing distinct clusters along multiple ordination axes for species like Erigeron bonariensis, Artemisia argyi, Quercus serrata and Pueraria montana (Figure 6a,b). Plant species, including both seed-bearing plants with larger seeds and ferns, tend to favor slopes in coniferous forests with denser soils at lower elevations. In contrast, seeds with awns or trichomes are more abundant in areas with fewer species pools at lower elevations. Conversely, flattened and winged seeds are frequently observed on slopes with thinner, drier soil layers (refer to Figure 6c,d). Moreover, this connection becomes remarkably clearer in the quadrant analysis results (model 5, p = 0.001). The outcomes of the multivariate permutation test demonstrate that the distribution of native plant species on road slopes is influenced jointly by plant traits and the environment. A significant yet nuanced overarching correlation exists between traits and the environment (refer to Figure 7, modeltype = 5). The relevant result figures depict a positive correlation between SRL and ASP, SM and VEG.con, as well as STH, LPC with VEG.con and DSP. Additionally, a positive correlation is observed between SSF.hair and DSP, SSF.no and ALT, N.P and VEG.bro, along with ALT. Conversely, an inverse relationship is noticeable between LPC and VEG.bro, ALT, SSF.hair and ALT, SSF.no and DSP, as well as N.P and VEG.con and DSP.

4. Discussion

4.1. Community Composition and Distribution on Karst Mountain Road Slopes

Numerous studies have documented the composition and distribution patterns of indigenous plant communities in newly developed road slope habitats worldwide [2,10,39,40]. However, the primary emphasis has traditionally been on gentle slopes within arid regions. In contrast, our investigation centers on a mountainous karst landscape situated in a subtropical humid monsoon climate, with an impressive mean slope gradient of 72.8 degrees and an abundant reservoir of potential species. Data analysis indicates that plant families, including Rosaceae, Asteraceae, Ericaceae, Theaceae, Poaceae, Lauraceae, Urticaceae, and Fabaceae, collectively constitute 42% of the species composition within the native plant community thriving on slopes. This finding aligns with studies conducted by Son (2022) and Park (2021) regarding slope plant composition [2,41], highlighting the significant role of these botanical families in promoting the natural restoration of slope vegetation on a global scale. Our investigation unveils the sporadic occurrence of 118 species (frequency < 5%), making up a substantial 51% proportion. This discovery implies a notable turnover rate in the plant species composition of the slope community within the study area. It gains further support through Richness Difference (RichDiff) and Replacement (Repl) triangular diagrams utilizing the Jaccard similarity index. The graphical representation underscores that species replacement (Repl) accounts for 71% of the total β-diversity variation, with richness difference (RichDiff) constituting the remaining 29% (Figure 8). Figure 8 illustrates that these disparities predominantly stem from diffusion constraints (Figure 2), which are linked to fluctuations in species richness across altitudinal gradients, variations in vegetation types, and the prevalent climatic conditions in mountainous regions [42,43]. These findings highlight that roads can also serve as habitats for a wide variety of plant species, not just as pathways for alien plants. This role underscores their significant contribution to the conservation of mountainous biodiversity, contingent upon effective management practices [44,45].

4.2. Establishment of Plant Communities on Road Slopes

The multivariate permutation test results indicate that the distribution of native species on road slopes is jointly influenced by both plant traits and environmental factors (Figure 6 and Figure 7). Environmental selection constrains species establishment by shaping species traits, with crucial shifts from initial-stage seed dispersal ability to later-stage competitive prowess [46].In our study, we observed an increased significance of environmental factors correlated with species richness and traits related to seed dispersal and establishment (Figure 7). This finding sharply contrasts with the conclusions drawn by Lebrija-Trejos (2010), which emphasized the dominant influence of temperature and temperature-regulating leaf traits on the differentiation of tropical rainforest community composition [47]. This discrepancy arises from the unique conditions of steep roadside slopes with limited soil seed banks. The presence of an ample seed source and effective dispersal mechanisms assume paramount significance as the initial prerequisites for community establishment [10]. Among the examined environmental factors, specific variables, including pH, SOM, SAN, and SAP exhibited no discernible impact on the constitution of thriving plant communities along slopes. These outcomes could be attributed to either: (a) the uniformity of limestone bedrock across karst mountainous regions, resulting in similar nutrient profiles; or (b) the presence of more rigorous environmental or ecological filters impeding vegetation establishment. Surprisingly, slope gradient no longer emerges as the pivotal determinant governing plant colonization and growth, a departure from the findings of Bochet (2004) and Canton (2004) regarding the critical slope threshold for colonization in arid zones whose investigations suggested that plant growth ceases when slopes surpass a 45° threshold [48,49]. In contrast, our study revealed an average slope gradient of an impressive 72.8° at sampling sites. We hypothesize that the impact of slope on settler plant colonization is significantly reduced when the seed quality of colonizing plants is sufficiently small. Unfortunately, this received limited attention in the present study. Undoubtedly, the success of colonization is intricately linked to the abundant rainfall characteristic of the subtropical humid monsoon climate.

Furthermore, the impact of diverse trait indicators on the composition of slope-based plant communities is less prominent than initially hypothesized. Specifically, the traits of SLA, LCC, LNC, LKC and C.N exhibit diminished effects. SLA is commonly employed as a proxy for relative growth rate and plant competitiveness. Additionally, both LCC and LNC demonstrate strong associations with the carbon and nitrogen cycling of species [50]. This phenomenon can be attributed to robust environmental selection mechanisms, resulting in relatively subtle variations in competitive capacities and carbon-nitrogen cycling efficiency among different species during the succession process. Conversely, LPC and N.P exhibit robust correlations with species distribution and environmental factors. The connection between phosphorus content and rRNA synthesis is well-recognized [51,52], wherein elevated phosphorus content or reduced nitrogen-to-phosphorus ratios indicate accelerated plant growth rates [53,54].

4.3. Distributional Structure of Slope Plant Traits

The analysis of trait structure reveals a prevalence of functional overdispersion rather than functional clustering in the plant community across the entire study area (Figure 4). This finding suggests that during the advanced stages of succession in road slope communities, species with distinct functions coexist extensively. These divergent traits may be linked to resource acquisition, differentiation, or functional complementarity among species [55]. However, a distinct finding emerges from Figure 4, showing a broad convergence in essential seed traits (SM and FI) across all plant species, along with trait indicators linked to photosynthetic capacity and net primary productivity (LCC, LPC). This convergence emphasizes a robust selection mechanism imposed by road slope environments on seed dispersal, colonization ability, growth rate, and photosynthetic capacity of colonizing plants. This alignment with the conclusions of Chauvet (2017) and Kang (2017) suggests that substantial environmental filtration need not necessarily hinder trait differentiation. Instead, it leads to a coexistence of both trait convergence and divergence [25,56]. Conventionally, features associated with regeneration and competitive strategies often manifest a divergent pattern, while those linked to nutrient-driven growth tend to exhibit communal convergence [56,57]. However, our investigation reveals a propensity for both divergence and convergence to unfold during distinct stages. Particularly, traits linked to plant dispersal and colonization display a convergence pattern. Interestingly, this aspect has received limited attention in prior research endeavors.

4.4. Inspiration for Road Slope Revegetation

Numerous studies emphasize the significance of plant functional traits in predicting plant community assembly [21,58]. However, translating this knowledge into effective strategies poses challenges for restoration practitioners, leading to restoration complexities [59,60]. Throughout restoration, it’s vital to identify plant traits suited to the road slope environment. Our study shows that species within the Rosaceae, Asteraceae, Ericaceae, Theaceae, Poaceae, Lauraceae, Urticaceae and Fabaceae families with smaller seeds and ferns are well-adapted to slopes. The potential regional species pool and neighboring vegetation richness directly affect the composition and diversity of self-sustaining plant communities on restored slopes. Prioritizing seeds with strong dispersal and colonization abilities, including long-distance mechanisms (e.g., winged, hairy, bird-dispersed) and mucilage-secreting properties, is recommended during restoration. Quantitative evaluation of plant environmental suitability based on seed traits before restoration helps assess trait strength and predict outcomes [61]. Additionally, including suitable species based on seed traits and environmental factors can overcome dispersal limitations and seed loss [62]. Lastly, strategic combinations of plants with different trait structures should be developed based on traits at various succession stages [63]. Understanding self-sustaining plant community assembly dynamics, and plant responses to the environment, can greatly contribute to biodiversity conservation and revegetation efforts.

5. Conclusions

In conclusion, our study highlights the significant roles of the environment such as regional species pool richness and elevation in shaping the diversity and complexity of native plant populations along road slopes in karst mountain regions, which are primarily influenced by dispersal limitations. Favorable external conditions and ample species reservoirs help mitigate the impact of slope-specific variables like gradient, aspect, and soil thickness on road slope revegetation. Throughout the establishment of plant communities on road slopes, there is a reciprocal interplay between plant traits and the surrounding environment. In the early stages of succession, environmental factors guide the establishment of plants with similar seed traits. As succession advances, limitations arising from species similarities encourage the coexistence of organisms with notable differences in growth characteristics. Gaining a comprehensive understanding of the formation process governing native plant communities on road slopes holds the potential to drive forward restoration endeavors.

Author Contributions

K.Q., H.Q. and H.W.; Data curation, K.Q., Z.W. and H.Z.; Funding acquisition, H.W.; Investigation, K.Q., Z.W., L.L. and H.W.; Methodology, K.Q., H.Q. and H.W.; Software, K.Q., Z.W., L.L. and H.Z.; Supervision, H.Q. and H.W.; Writing—original draft, K.Q.; Writing—review and editing, K.Q., Z.W. and L.L. All authors have read and agreed to the published version of the manuscript.

Data Availability Statement

The data in this study are available from the authors upon request.

Acknowledgments

Sincere thanks to the Baimashan Municipal Nature Reserve for supporting this study.

Conflicts of Interest

The authors declare no conflict of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures and Tables

Figure 1. Geographical location of study area and sampling points.

Figure 2. Redundancy analysis (RDA) double sequence diagram of species and environmental factors (Only species with R2 > 0.1 are shown). Environmental variables include adjacent vegetation types (VEG), aspect (ASP), soil moisture content (SMC), altitude (ALT), aspect (ASP) and Soil available potassium (SAK), Soil thickness (STH), Distance from regional core Species pool (DSP).

Figure 3. Comparison of the empirical SIMPER profile generated by cellular automaton-like simulation with permutational SIMPER null distribution profiles (left-hand graph), and related distributions of the E-metric (right-hand graph): (a) Distribution model of plants at varying distances from regional core species pools based on PER-SIPER method; (b) Distribution model of plants at different altitudes based on PER-SIPER method; (c) Distribution model of plants adjacent to different vegetation types based on PER-SIPER method.

Figure 4. Analysis on the difference of functional characters of plants in different families. LCC, LNC, LPC and LKC are the contents of total carbon, total nitrogen, total phosphorus and total potassium in leaves. SRL, specific root length; SLA, Specific Leaf Area; SM, seed mass; EI, Eccentric index; FI, Flatness index, SSF, Special Functional Traits of Seeds. Spermatophyte functional groups including Adoxaceae (Ado), Asteraceae (Ast), Ericaceae (Eri), Rosaceae (Ros), and else (Ele), fern are marked by “Fer”.

Figure 5. Plant functional structure was evaluated based on the standardized effect size of the functional mean pairwise distances (SESMPD) across all plants (n = 113) determined using species means of traits: (a) Mean Standardized effect size for different traits (SESMPDmean, mean ± 95% confidence interval). Plant functional structure is evaluated based on determined using species means of traits. The positive SESMPDmean value represents excessive dispersion of functions, while a negative value represents functional aggregation; (b) The frequency of excessive dispersion (and aggregation) of different traits. The dashed line indicates the 50% quantile. Total, when all the traits were combined.

Figure 6. RLQ analysis results of species-trait-environment for the natural plant community on road slopes: (a) point coordinates (L); (b) species coordinates (Q); (c) environmental variables; (d) species traits.

Figure 7. Results of the fourth-corner analysis on the bivariate associations between environments and plant traits with adjusting p-values for multiple testing (model type = 5). Significantly (p-value < 0.05) positive associations are represented by red cells, and significantly negative associations correspond to blue cells. Nonsignificant associations are in gray. The different variables are separated by black lines. C.N is the ratio of leaf carbon content to nitrogen content, and N.P is the ratio of leaf nitrogen content to phosphorus content.

Figure 8. Two pairwise exponential triangle of Jaccard similarity index, replacement and richness difference (the big black dots is the average).

Table 1

Summary of basic information of surveyed roads in the study area.

Road Number	Construction Time (Year)	Length(km)	Elevation(mean ± SD, m)	Vegetation 1	Slope (°)	Sample Points	DSP 2(km)
R1	2013	6.51	1361 ± 129	VEGthe	63.83 ± 16.82	9	5.61
R2	2014	4.24	1283 ± 202	VEGthe	65.23 ± 10.99	9	12.18
R3	2016	4.63	1219 ± 133	VEGcon	67.86 ± 14.18	8	12.86
R4	2000	7.38	1588 ± 169	VEGbro	69.14 ± 13.96	31	1.17
R5	2017	21.22	795 ± 261	VEGbro	82.68 ± 5.63	6	14.73
R6	2006	8.57	1060 ± 98	VEGthe	77.5 ± 5.56	5	13.51
R7	2017	2.45	1570 ± 24	VEGbro	58.54 ± 9.61	8	9.82

¹ Vegetation: Including coniferous forests(VEGcon), theropencedrymion (VEGthe), and broad-leaved forests(VEGbro). ² DSP: Average distance from core species pool.

Table 2

Description of plant traits and abbreviations used in the paper, compilation sources and ecological relevance for plant colonization on road slopes.

Trait Category	Trait	Scale and Unit	Abbreviation	Ecological Relevance	Information Compilation Source
General plant traits	Total nitrogen content of leaves	Continuous (mg g−1)	LNC	Seedling growth	Field harvesting and laboratory measurements
	Total phosphorus content in leaves	Continuous (mg g−1)	LPC	Seedling growth	Field harvesting and laboratory measurements
	Total carbon content of leaves	Continuous (mg g−1)	LCC	Seedling growth	Field harvesting and laboratory measurements
	Total potassium content in leaves	Continuous (mg g−1)	LKC	Seedling growth, Response to stress	Field harvesting and laboratory measurements
	Specific Leaf Area	Continuous (mm2 mg−1)	SLA	Seedling growth, Response to stress	Field harvesting and laboratory measurements
	specific root length	Continuous (mm mg−1)	SRL	Seedling growth, Response to stress	Field harvesting and laboratory measurements
Seed-related traits	Seed mass	Continuous (mg)	SM	Dispersal, Germinate, Seedling growth	Field harvesting and laboratory measurements
	Seed eccentricity index	Continuous (dimensionless, ≥1)	FI	Settle, Response to stress	Field harvesting and laboratory measurements, and calculations
	Seed flatness index	Continuous (dimensionless, ≥1)	EI	Settle, Response to stress	Field harvesting and laboratory measurements, and calculations
	Seed accessory functional organs	Classified, including seeds with wings, seeds with hairs, seeds that secrete mucus, Seeds without appendages, and Spores of Pteridophyte	SSF.wing	Dispersal	Field harvesting and laboratory measurements
			SSF.hair	Dispersal, Germinate	Field harvesting and laboratory measurements
			SSF.sucilage	Settle, Response to stress	Field excavations and observations, and literature
			SSF.spore	Dispersal	Field excavations and observations, and literature
			SSF.no	Dispersal	Field excavations and observations, and literature

References

1. Rahbek, C.; Borregaard, M.K.; Antonelli, A.; Colwell, R.K.; Holt, B.G.; Nogues-Bravo, D.; Rasmussen, C.M.; Richardson, K.; Rosing, M.T.; Whittaker, R.J. et al. Building mountain biodiversity: Geological and evolutionary processes. Science; 2019; 365, pp. 1114-1119. [DOI: https://dx.doi.org/10.1126/science.aax0151] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/31515384]

2. Son, D.; Alday, J.G.; Chu, Y.; Lee, E.J.; Park, S.; Lee, H. Plant species colonization in newly created road habitats of South Korea: Insights for more effective restoration. Sci. Total Environ.; 2020; 719, 137476. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2020.137476] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32114236]

3. Wang, Q.; Li, M.; Xie, X. Optimal allocation model of plant community in slope ecological restoration. Glob. Nest. J.; 2020; 22, pp. 554-564.

4. Rivera, D.; Mejias, V.; Jauregui, B.M.; Costa-Tenorio, M.; Lopez-Archilla, A.I.; Peco, B. Spreading topsoil encourages ecological restoration on embankments: Soil fertility, microbial activity and vegetation cover. PLoS ONE; 2014; 9, e101413. [DOI: https://dx.doi.org/10.1371/journal.pone.0101413] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24984137]

5. Fu, D.; Yang, H.; Wang, L.; Yang, S.; Li, R.; Zhang, W.; Ai, X.; Ai, Y. Vegetation and soil nutrient restoration of cut slopes using outside soil spray seeding in the plateau region of southwestern China. J. Environ. Manag.; 2018; 228, pp. 47-54. [DOI: https://dx.doi.org/10.1016/j.jenvman.2018.08.108]

6. Xiao, H.; Huang, J.; Ma, Q.; Wan, J.; Li, L.; Peng, Q.; Rezaeimalek, S. Experimental study on the soil mixture to promote vegetation for slope protection and landslide prevention. Landslides; 2017; 14, pp. 287-297. [DOI: https://dx.doi.org/10.1007/s10346-015-0634-x]

7. Ai, X.; Sheng, M.; Su, X.; Ai, S.; Jiang, X.; Yang, S.; Huang, Z.; Ai, Y. Effects of frame beam on structural characteristics of artificial soil on railway cut-slopes in southwestern China. Land Degrad. Dev.; 2021; 32, pp. 482-493. [DOI: https://dx.doi.org/10.1002/ldr.3719]

8. Novak, J.; Prach, K. Vegetation succession in basalt quarries: Pattern on a landscape scale. Appl. Veg. Sci.; 2003; 6, pp. 111-116. [DOI: https://dx.doi.org/10.1111/j.1654-109X.2003.tb00570.x]

9. Matesanz, S.; Valladares, F.; Tena, D.; Costa-Tenorio, M.; Bote, D. Early dynamics of plant communities on revegetated motorway slopes from southern Spain: Is hydroseeding always needed?. Restor. Ecol.; 2006; 14, pp. 297-307. [DOI: https://dx.doi.org/10.1111/j.1526-100X.2006.00132.x]

10. Bochet, E.; Garcia-Fayos, P.; Tormo, J. Road slope revegetation in semiarid mediterranean environments. Part I: Seed dispersal and spontaneous colonization. Restor. Ecol.; 2007; 15, pp. 88-96. [DOI: https://dx.doi.org/10.1111/j.1526-100X.2006.00193.x]

11. Ulrich, W.; Zaplata, M.K. Trait driven or neutral: Understanding the change in functional trait diversity during early plant succession using Price partitions. Plant Ecol.; 2023; [DOI: https://dx.doi.org/10.1007/s11258-023-01344-y]

12. Holzapfel, C.; Schmidt, W. Roadside vegetation along transects in the Judean Desert. Isr. J. Bot.; 1990; 39, pp. 263-270.

13. Godefroid, S.; Tanghe, M. Influence of small climatic variations on the species composition of roadside grasslands. Phytocoenologia; 2000; 30, pp. 655-664. [DOI: https://dx.doi.org/10.1127/phyto/30/2000/655]

14. Guzman, P.; Benitez, A.; Carrion-Paladines, V.; Salinas, P.; Cumbicus, N. Elevation and soil properties determine community composition, but not vascular plant richness in tropical andean roadside. Forests; 2022; 13, 685. [DOI: https://dx.doi.org/10.3390/f13050685]

15. Backes, A.R.; Römermann, C.; Alexander, J.M.; Arévalo, J.R.; Keil, P.; Padrón-Mederos, M.A.; Trogisch, S.; Haider, S. Mechanisms behind elevational plant species richness patterns revealed by a trait-based approach. J. Veg. Sci.; 2023; 34, e13171.

16. Bochet, E.; Garcia-Fayos, P. Identifying plant traits: A key aspect for species selection in restoration of eroded roadsides in semiarid environments. Ecol. Eng.; 2015; 83, pp. 444-451. [DOI: https://dx.doi.org/10.1016/j.ecoleng.2015.06.019]

17. Beck, J.J.; Richards, J.H. Functional traits influence local plant distributions and spatial patterns of diversity within a heterogeneous bedrock glade. Plant Ecol.; 2023; 224, pp. 729-740. [DOI: https://dx.doi.org/10.1007/s11258-023-01337-x]

18. Cornwell, W.K.; Ackerly, D.D. Community assembly and shifts in plant trait distributions across an environmental gradient in coastal California. Ecol. Monogr.; 2009; 79, pp. 109-126. [DOI: https://dx.doi.org/10.1890/07-1134.1]

19. Lopez-Angulo, J.; de la Cruz, M.; Pescador, D.S.; Sanchez, A.M.; Escudero, A. A dimmer shade of pale: Revealing the faint signature of local assembly processes on the structure of strongly filtered plant communities. Ecography; 2021; 44, pp. 87-97. [DOI: https://dx.doi.org/10.1111/ecog.05405]

20. Silva, J.L.A.; Souza, A.F.; Santiago, L.S. Traits uncover quasi-neutral community assembly in a coastal heath vegetation. J. Plant Ecol.; 2019; 12, pp. 703-712. [DOI: https://dx.doi.org/10.1093/jpe/rtz007]

21. Kraft, N.J.B.; Adler, P.B.; Godoy, O.; James, E.C.; Fuller, S.; Levine, J.M. Community assembly, coexistence and the environmental filtering metaphor. Funct. Ecol.; 2015; 29, pp. 592-599. [DOI: https://dx.doi.org/10.1111/1365-2435.12345]

22. Bittebiere, A.K.; Mony, C. Plant traits respond to the competitive neighbourhood at different spatial and temporal scales. Ann. Bot.; 2015; 115, pp. 117-126. [DOI: https://dx.doi.org/10.1093/aob/mcu206]

23. May, F.; Giladi, I.; Ristow, M.; Ziv, Y.; Jeltsch, F. Plant functional traits and community assembly along interacting gradients of productivity and fragmentation. Perspect. Plant Ecol.; 2013; 15, pp. 304-318. [DOI: https://dx.doi.org/10.1016/j.ppees.2013.08.002]

24. Dubuis, A.; Rossier, L.; Pottier, J.; Pellissier, L.; Vittoz, P.; Guisan, A. Predicting current and future spatial community patterns of plant functional traits. Ecography; 2013; 36, pp. 1158-1168. [DOI: https://dx.doi.org/10.1111/j.1600-0587.2013.00237.x]

25. Chauvet, M.; Kunstler, G.; Roy, J.; Morin, X. Using a forest dynamics model to link community assembly processes and traits structure. Funct. Ecol.; 2017; 31, pp. 1452-1461. [DOI: https://dx.doi.org/10.1111/1365-2435.12847]

26. Thompson, K.; Petchey, O.L.; Askew, A.P.; Dunnett, N.P.; Beckerman, A.P.; Willis, A.J. Little evidence for limiting similarity in a long-term study of a roadside plant community. J. Ecol.; 2010; 98, pp. 480-487. [DOI: https://dx.doi.org/10.1111/j.1365-2745.2009.01610.x]

27. Backhaus, L.; Albert, G.; Cuchietti, A.; Jaimes Nino, L.M.; Fahs, N.; Lisner, A.; Kolář, V.; Kermavnar, J.; Widmer, S.; Zimmermann, Z. et al. Shift from trait convergence to divergence along old-field succession. J. Veg. Sci.; 2021; 32, e12986. [DOI: https://dx.doi.org/10.1111/jvs.12986]

28. Erfanzadeh, R.; Garbutt, A.; Petillon, J.; Maelfait, J.P.; Hoffmann, M. Factors affecting the success of early salt-marsh colonizers: Seed availability rather than site suitability and dispersal traits. Plant Ecol.; 2010; 206, pp. 335-347. [DOI: https://dx.doi.org/10.1007/s11258-009-9646-8]

29. Pescador, D.S.; Sierra-Almeida, A.; Torres, P.J.; Escudero, A. Summer freezing resistance: A critical filter for plant community assemblies in mediterranean high mountains. Front. Plant Sci.; 2016; 7, 194. [DOI: https://dx.doi.org/10.3389/fpls.2016.00194]

30. Janse-ten Klooster, S.H.; Thomas, E.J.P.; Sterck, F.J. Explaining interspecific differences in sapling growth and shade tolerance in temperate forests. J. Ecol.; 2007; 95, pp. 1250-1260. [DOI: https://dx.doi.org/10.1111/j.1365-2745.2007.01299.x]

31. Lohbeck, M.; Poorter, L.; Martinez-Ramos, M.; Rodriguez-Velazquez, J.; van Breugel, M.; Bongers, F. Changing drivers of species dominance during tropical forest succession. Funct. Ecol.; 2014; 28, pp. 1052-1058. [DOI: https://dx.doi.org/10.1111/1365-2435.12240]

32. van der, M.E.; Franklin, J. Vegetation Ecology; 2nd ed. Wiley: Chichester, UK, 2013; pp. 154-196.

33. Cornelissen, J.H.C.; Lavorel, S.; Garnier, E.; Díaz, S.; Buchmann, N.; Gurvich, D.E.; Reich, P.B.; ter Steege, H.; Morgan, H.D.; van der Heijden, M.G.A. et al. A handbook of protocols for standardised and easy measurement of plant functional traits worldwide. Aust. J. Bot.; 2003; 51, pp. 335-380. [DOI: https://dx.doi.org/10.1071/BT02124]

34. Cerda, A.; Garcia-Fayos, P. The influence of seed size and shape on their removal by water erosion. Catena; 2002; 48, pp. 293-301. [DOI: https://dx.doi.org/10.1016/S0341-8162(02)00027-9]

35. Luo, W.; Lan, R.; Chen, D.; Zhang, B.; Xi, N.; Li, Y.; Fang, S.; Valverde-Barrantes, O.J.; Eissenstat, D.M.; Chu, C. et al. Limiting similarity shapes the functional and phylogenetic structure of root neighborhoods in a subtropical forest. New Phytol.; 2021; 229, pp. 1078-1090. [DOI: https://dx.doi.org/10.1111/nph.16920] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32924174]

36. Dray, S.; Legendre, P. Testing the Species Traits-Environment Relationships: The fourth-corner problem revisited. Ecology; 2008; 89, pp. 3400-3412. [DOI: https://dx.doi.org/10.1890/08-0349.1] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/19137946]

37. Dray, S.; Choler, P.; Dolédec, S.; Peres-Neto, P.R.; Thuiller, W.; Pavoine, S.; ter Braak, C.J.F. Combining the fourth-corner and the RLQ methods for assessing trait responses to environmental variation. Ecology; 2014; 95, pp. 14-21. [DOI: https://dx.doi.org/10.1890/13-0196.1]

38. Gibert, C.; Escarguel, G. PER-SIMPER-A new tool for inferring community assembly processes from taxon occurrences. Glob. Ecol. Biogeogr.; 2019; 28, pp. 374-385. [DOI: https://dx.doi.org/10.1111/geb.12859]

39. Saikia, P.; Deka, J.; Bharali, S.; Kumar, A.; Tripathi, O.P.; Singha, L.B.; Dayanandan, S.; Khan, M.L. Plant diversity patterns and conservation status of eastern Himalayan forests in Arunachal Pradesh, Northeast India. For. Ecosyst.; 2017; 4, 28. [DOI: https://dx.doi.org/10.1186/s40663-017-0117-8]

40. Wani, Z.A.; Khan, S.; Bhat, J.A.; Malik, A.H.; Alyas, T.; Pant, S.; Siddiqui, S.; Moustafa, M.; Ahmad, A.E. Pattern of beta-diversity and plant species richness along Vertical Gradient in Northwest Himalaya, India. Biology; 2022; 11, 1064. [DOI: https://dx.doi.org/10.3390/biology11071064]

41. Park, S.; Kim, J.H.; Byun, C.; Hong, S.Y.; Lee, E.J. Identification of restoration species for early roadcut slope regeneration using functional group approach. Restor. Ecol.; 2021; 29, e134241. [DOI: https://dx.doi.org/10.1111/rec.13424]

42. Wang, X.; Wiegand, T.; Anderson-Teixeira, K.J.; Bourg, N.A.; Hao, Z.; Howe, R.; Jin, G.; Orwig, D.A.; Spasojevic, M.J.; Wang, S. et al. Ecological drivers of spatial community dissimilarity, species replacement and species nestedness across temperate forests. Glob. Ecol. Biogeogr.; 2018; 27, pp. 581-592. [DOI: https://dx.doi.org/10.1111/geb.12719]

43. Laiolo, P.; Illera, J.C.; Obeso, J.R. Stuck on top of a mountain: Consequences of dispersal limitations for alpine diversity. J. Biogeogr.; 2023; 50, pp. 282-290. [DOI: https://dx.doi.org/10.1111/jbi.14513]

44. Azcarate, F.M.; Robleno, I.; Seoane, J.; Manzano, P.; Peco, B. Drove roads as local biodiversity reservoirs: Effects on landscape pattern and plant communities in a Mediterranean region. Appl. Veg. Sci.; 2013; 16, pp. 480-490. [DOI: https://dx.doi.org/10.1111/avsc.12003]

45. Arenas, J.M.; Escudero, A.; Mola, I.; Casado, M.A. Roadsides: An opportunity for biodiversity conservation. Appl. Veg. Sci.; 2017; 20, pp. 527-537. [DOI: https://dx.doi.org/10.1111/avsc.12328]

46. Schleicher, A.; Peppler-Lisbach, C.; Kleyer, M. Functional traits during succession: Is plant community assembly trait-driven?. Preslia; 2011; 83, pp. 347-370.

47. Lebrija-Trejos, E.; Perez-Garcia, E.A.; Meave, J.A.; Bongers, F.; Poorter, L. Functional traits and environmental filtering drive community assembly in a species-rich tropical system. Ecology; 2010; 91, pp. 386-398. [DOI: https://dx.doi.org/10.1890/08-1449.1]

48. Bochet, E.; Garcia-Fayos, P. Factors controlling vegetation establishment and water erosion on motorway slopes in Valencia, Spain. Restor. Ecol.; 2004; 12, pp. 166-174. [DOI: https://dx.doi.org/10.1111/j.1061-2971.2004.0325.x]

49. Canton, Y.; Del Barrio, G.; Sole-Benet, A.; Lazaro, R. Topographic controls on the spatial distribution of ground cover in the Tabemas badlands of SE Spain. Catena; 2004; 55, pp. 341-365. [DOI: https://dx.doi.org/10.1016/S0341-8162(03)00108-5]

50. Garnier, E.; Cortez, J.; Billès, G.; Navas, M.L.; Roumet, C.; Debussche, M.; Laurent, G.; Blanchard, A.; Aubry, D.; Bellmann, A. et al. Plant functional markers capture ecosystem properties during secondary succession. Ecology; 2004; 85, pp. 2630-2637. [DOI: https://dx.doi.org/10.1890/03-0799]

51. Elser, J.J.; Acharya, K.; Kyle, M.; Cotner, J.B.; Makino, W.; Markow, T.; Watts, T.; Hobbie, S.E.; Fagan, W.; Schade, J. et al. Growth rate-stoichiometry couplings in diverse biota. Ecol. Lett.; 2003; 6, pp. 936-943. [DOI: https://dx.doi.org/10.1046/j.1461-0248.2003.00518.x]

52. Zhu, D.H.; Hui, D.F.; Wang, M.Q.; Yang, Q.; Yu, S.X. Light and competition alter leaf stoichiometry of introduced species and native mangrove species. Sci. Total Environ.; 2020; 738, 140304. [DOI: https://dx.doi.org/10.1016/j.scitotenv.2020.140301] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32806386]

53. Elser, J.J.; Sterner, R.W.; Galford, A.E.; Chrzanowski, T.H.; Findlay, D.L.; Mills, K.H.; Paterson, M.J.; Stainton, M.P.; Schindler, D.W. Pelagic C:N:P stoichiometry in a eutrophied lake: Responses to a whole-lake food-web manipulation. Ecosystems; 2000; 3, pp. 293-307. [DOI: https://dx.doi.org/10.1007/s100210000027]

54. Karpinets, T.V.; Greenwood, D.J.; Sams, C.E.; Ammons, J.T. RNA: Protein ratio of the unicellular organism as a characteristic of phosphorous and nitrogen stoichiometry and of the cellular requirement of ribosomes for protein synthesis. BMC Biol.; 2006; 4, 30. [DOI: https://dx.doi.org/10.1186/1741-7007-4-30] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/16953894]

55. Herben, T.; Goldberg, D.E. Community assembly by limiting similarity vs. competitive hierarchies: Testing the consequences of dispersion of individual traits. J. Ecol.; 2014; 102, pp. 156-166. [DOI: https://dx.doi.org/10.1111/1365-2745.12181]

56. Kang, S.; Niu, J.; Zhang, Q.; Li, D.; Ren, H.; Ren, J.; Zhang, X.; Dong, J. Environmental filtering does not necessarily prevent trait divergence: A case study of the Xilin River Basin in Inner Mongolia, China. J. Plant Ecol.; 2017; 10, pp. 497-509. [DOI: https://dx.doi.org/10.1093/jpe/rtw050]

57. Bernard-Verdier, M.; Navas, M.L.; Vellend, M.; Violle, C.; Fayolle, A.; Garnier, E. Community assembly along a soil depth gradient: Contrasting patterns of plant trait convergence and divergence in a Mediterranean rangeland. J. Ecol.; 2012; 100, pp. 1422-1433. [DOI: https://dx.doi.org/10.1111/1365-2745.12003]

58. Suter, M.; Edwards, P.J. Convergent succession of plant communities. is linked to species’ functional traits. Perspect. Plant Ecol.; 2013; 15, pp. 217-225. [DOI: https://dx.doi.org/10.1016/j.ppees.2013.05.001]

59. Jalonen, R.; Valette, M.; Boshier, D.; Duminil, J.; Thomas, E. Forest and landscape restoration severely constrained by a lack of attention to the quantity and quality of tree seed: Insights from a global survey. Conserv. Lett.; 2018; 11, e12424. [DOI: https://dx.doi.org/10.1111/conl.12424]

60. Reubens, B.; Moeremans, C.; Poesen, J.; Nyssen, J.; Tewoldeberhan, S.; Franzel, S.; Deckers, J.; Orwa, C.; Muys, B. Tree species selection for land rehabilitation in Ethiopia: From fragmented knowledge to an integrated multi-criteria decision approach. Agrofor. Syst.; 2011; 82, pp. 303-330. [DOI: https://dx.doi.org/10.1007/s10457-011-9381-8]

61. Kulpa, S.M.; Leger, E.A. Strong natural selection during plant restoration favors an unexpected suite of plant traits. Evol. Appl.; 2013; 6, pp. 510-523. [DOI: https://dx.doi.org/10.1111/eva.12038]

62. Pywell, R.F.; Bullock, J.M.; Roy, D.B.; Warman, L.I.Z.; Walker, K.J.; Rothery, P. Plant traits as predictors of performance in ecological restoration. J. Appl. Ecol.; 2003; 40, pp. 65-77. [DOI: https://dx.doi.org/10.1046/j.1365-2664.2003.00762.x]

63. Rinella, M.J.; James, J.J. A modelling framework for improving plant establishment during ecological restoration. Ecol. Model.; 2017; 361, pp. 177-183. [DOI: https://dx.doi.org/10.1016/j.ecolmodel.2017.08.005]