Abstract
Variations in plant traits are indicative of plant adaptations to forest environments, and studying their relationships with tree growth provides valuable insights into forest regeneration. The spatial arrangement of plant seeds within the forest litter or soil critically influences the variations of root–leaf traits, thereby affecting the adaptive strategies of emerging seedlings. However, our current understanding of the impacts of individual root–leaf traits on seedling growth in different relative position, and whether these traits together affect growth, remains limited. This study focuses on the dominant tree species, Castanopsis kawakamii, within the Sanming C. kawakamii Nature Reserve of China. The present experiment aimed to examine the variations in root–leaf traits of seedling, focus on the relative positions of seeds within different layers: beneath or above the litter layer, or within the bare soil layer (without litter). Our findings provided evidence supporting a coordinated relationship between root and leaf traits, wherein leaf traits varied in conjunction with root traits in the relative positions of seeds. Specifically, we observed that seedlings exhibited higher values for specific leaf area and average root diameter, while displaying lower root tissue density. The mixed model explained 86.1% of the variation in root–leaf traits, surpassing the variation explained by the relative positions. Furthermore, soil nitrogen acted as a mediator, regulating the relationship between seedling growth and root–leaf traits, specifically leaf dry matter content and root tissue density. Therefore, future studies should consider artificially manipulating tree species diversity based on root–leaf traits characteristics to promote forest recovery.
摘要
植物性状的变化反映了植物对森林环境的适应，研究它们与树木生长的关系为森林更新提供有效的见解。种子在森林凋落物或土壤层的空间位置显著影响植物根、叶性状的变化，进而影响新生幼苗的适应策略。然而，目前对不同种子相对位置的个体根、叶性状以及相关性状是否共同影响幼苗生长的理解仍然有限。为此，本研究以中国三明格氏栲自然保护区的优势树种格氏栲(Castanopsis kawakamii)为研究对象，设置种子在凋落物上层、凋落物下层与土壤表层(无凋落物层)等不同试验处理，揭示幼苗根、叶性状在种子相对位置不同处理下的变化规律。结果表明，叶性状在种子相对位置不同处理下随根性状的变化而变化，为支持根、叶性状存在协调关系提供了证据。具体而言，幼苗具有较高的比叶面积和根平均直径，同时表现出更低的根组织密度。混合模型的根、叶性状解释了幼苗相对生长86.1%的变异，高于种子相对位置解释的变异。此外，土壤氮调节了幼苗生长与根、叶性状，特别是与叶片干物质含量、根组织密度之间的关系。因此，未来研究可考虑基于植物幼苗根、叶性状特征，人工调控森林树种的多样性，以促进森林更新。
INTRODUCTION
Seedling regeneration is a bottleneck stage in forest dynamics, representing a highly sensitive period that mirrors the forest environment (Comita et al. 2009). It ultimately governs the tree species composition within the forest ecosystem (Spasojevic et al. 2014). Plant traits serve as manifestations of a plant’s adaptive capacity to environmental changes (Reich 2014), playing a pivotal role in elucidating the life history strategies of plant species through alteration in their physiological and morphological characteristics (Adler et al. 2014). Furthermore, plant root and leaf traits are intricately coordinated and serve as key sites for the allocation and exchange of elemental resources within plants (Wright et al. 2010). Under specific environmental conditions, plants regulate the allocation of roots and leaves to adapt to variations in their surroundings environments (Fortunel et al. 2012). This adaptation includes the augmentation of specific leaf area or root length to enhance the interception of available resources (Li et al. 2018). Therefore, gaining insight into the response of root and leaf traits to environment variations and recognizing the significance of these traits in plant growth are vital steps in understanding the process of forest regeneration.
Plant leaf traits have a significant impact on seedling growth, primarily by influencing the capacity of leaves for photosynthesis and respiration (Wright et al. 2010). Additionally, these traits also play a crucial role in determining the adaptability of leaf to the surrounding environment (Shen et al. 2019). Research has indicated that plants enhance their ability to capture light resources by increasing their leaf nitrogen content or stem length, subsequently impacting the process of plant photosynthesis and metabolism (Luo et al. 2021; Jiang et al. 2022). This augmentation in resource capture contributes to the plants’ improved survival capacity (Jiang et al. 2022). Additionally, the impacts of plant root traits on seedling growth are centered around their function in regulating the seedling’s ability to absorb soil resources (Boonman et al. 2020) and delivering the necessary soil nutrient resources for seedling growth (Umaña et al. 2021a). Moreover, understanding seedling growth is enhanced when considering the combined traits of root and leaf, rather than studying each organ individually (Shen et al. 2019; Weemstra et al. 2021). The combination of root and leaf traits strongly influences plant growth and survival (Shen et al. 2019; Weemstra et al. 2021). In regions where soil nutrient resources are abundant but light is limited, leaf growth is dependent on the ability of the root system to deliver soil resources (Martini et al. 2020). Conversely, in relatively resource-poor regions, plant root growth often relies on the absorption of photosynthetic products from leaves (Martini et al. 2020). However, the coordination of root and leaf traits does not always follow this pattern, and the interaction between these traits influences plant growth in relation to soil-related factors (Weemstra et al. 2021). Hence, further exploration is warranted to understand the impacts of multiple factors (leaf traits, root traits, and soil variables) on seedling growth and identify the relative importance of these factors in shaping plant growth.
Plants that propagate through seeds initially encounter either litter or soil as their physical environment (Zhang et al. 2022). This surrounding environment plays a crucial role in the entire regeneration process of vegetation (Sonkoly et al. 2020). The specific position of the seedling within the litter or soil layer influences the adjustment of leaf and root traits to adapt to their respective positions (Zhu et al. 2023) (Fig. 1). Previous studies have shown that the relative position of seeds in relation to litter can be either on top of or beneath the litter layer (Zhang et al. 2022). When seedlings emerge on top of the litter layer, the litter layer prevents direct contact between their roots and the soil. This is due to the presence of the litter layer, which acts as a physical barrier inhibiting the roots penetration into the soil (Sayer 2006). Conversely, when forest litter is deposited, seeds may become covered by the leaf litter layer, indicating their position beneath this layer (Sayer 2006). The litter layer fulfills several essential functions including insulation and moisture retention, nutrient release through decomposition, and subsequent replenishment of the soil (Berg 2014; Sonkoly et al. 2020). Additionally, it provides protection against excessive light exposure, which in turn facilitates optimal seedling growth (Visscher et al. 2016). When positioned above the soil surface, the litter layer creates a relatively stable environment that benefits seedling performance (Zhang et al. 2017). Previous observations have indicated that a well-coordinated relationship between the root and leaf of seedlings (Shen et al. 2019; Zhu et al. 2023). However, when the thickness of litter exceeds a certain threshold, it creates physical barriers that inhibit the upward growth of seedlings (Sayer 2006). In response to this, seedlings expand their root system to absorb nutrients from the soil and transport them to their leaves, supporting growth (Zhu et al. 2022c). Consequently, the combined effects of root and leaf traits on tree growth are significantly influenced by the soil environment (Umaña et al. 2021a). When seedling emerge above the forest floor, resource allocation is prioritized toward the roots to enhance the absorption of water and nutrients from the soil (Zhu et al. 2023). Consequently, this leads to reduced allocation to leaf tissues. However, the relationships between root and leaf traits are often disregarded in terms of their impact on plant growth under different relative positions. Therefore, it is crucial to comprehend the underlying mechanisms that govern the relationship between root–leaf traits, soil characteristics, and plant growth to accurately predict forest tree composition and coexistence.
A illustration of how variations in leaf and root traits, and soil factors are associated with seedling growth in different seed relative positions (developed based on Zhu et al. 2023). Litter or soil is the initial physical environment that seeds contact after falling from a mother tree, directly affect the relative position of seeds. Gray arrows illustrate the chemical effects influenced by the decomposition of litter, and the blue arrows illustrate the potential shifts in position that can occur due to physical effects affected by litter. The numbers ①–③ indicate the growth strategies (coordination of traits) that emerged seedlings may adopt under different relative positions of seeds (seeds placed on the forest floor, litter layer, or between soil and litter layer). In the variations of relative positions, number ④ indicates that whether the coordination relationship between root and leaf traits can affect the growth of seedling.
Castanopsis species are widely recognized as a significant vegetation type that plays a crucial role in promoting the sustainable development of subtropical forests (Yang et al. 2015). The Castanopsis kawakamii Nature Reserve encompasses a diverse range of typical species, predominantly dominated by the C. kawakamii tree species (He et al. 2020) which holds a pivotal role in shaping the composition of the forest community (He et al. 2020). The forest ecosystem is currently encountering uncertain regeneration dynamics as a result of various factors, including the presence of over-mature trees with a canopy age exceeding 100 years, limited regeneration of young seedlings, and a low turnover rate from the seedling to sapling stage (He et al. 2020; Zhu et al. 2022a). Previous research has demonstrated that litter decomposition enhances soil characteristics (Chen et al. 2023). Moreover, the spatial arrangement of seeds in relation to the litter layer has significant implications for the allocation of root and leaf traits in plants, ultimately influencing forest composition (Zhu et al. 2022b, 2023). However, there is a knowledge gap regarding the influence of different seed positions on root and leaf traits, soil factors, and their interaction among them that affect plant growth. To fill this research gap, we conducted a study to analyze the relationship between root and leaf traits in seedling of C. kawakamii and their impact on the relative growth rate of seedlings in different seed relative positions. Specifically, we aimed to answer the following questions: (i) Are root and leaf traits of seedlings well coordinated in different seeds relative positions? (ii) Does the positioning of seeds have an impact on root and leaf traits, thus influencing overall seedling growth? (iii) How can we assess the effects of seed relative position, root traits, and leaf traits on seedling growth, and differentiate their respective contributions to the overall growth of the seedling (refer to Fig. 1 for a visual representation)? Our hypotheses are as follows: (i) There is a well-coordinated relationship between root and leaf traits of seedlings, reflecting different resource allocation strategies adopted by plants in response to changes in seeds relative positions (Shen et al. 2019). (ii) Traits have a stronger impact on seedling growth than the relative positions of seeds, as changes in traits directly affect a plant’s resource acquisition ability (Poorter et al. 2012). (iii) The correlation relationship between root–leaf traits and relative growth rate is noteworthy in different relative positions, and the influence of soil nitrogen mediates the relationship between these traits and seedling growth. To address these questions, we examined essential root and leaf traits, along with soil characteristics across different seeds relative positions.
MATERIALS AND METHODS
Material collection
Materials (seeds, litter, and soil) were collected from C. kawakamii Nature Reserve in southeastern of China (Zhu et al. 2022b, 2022c, 2023). This region is characterized as a typical subtropical forest, with an annual mean temperature of 19.5°C and annual mean precipitation of approximately 1500 mm (He et al. 2020; Zhu et al. 2022a). The nature reserve covers an undisturbed forest of 700 ha, primarily dominated by the tree species C. kawakamii (He et al. 2020). In this study, we initially established 10 sample plots within the region for the collection of plants seeds, litter, and soil. Seed collection occurred during the peak period of fall (October to December 2018), with careful removal of insect-damaged seeds. For more detailed information on the seeds, litter, and soil, please refer to Zhu et al. (2022a, 2022b, 2022c).
Experimental design
The experiment was conducted in a spacious indoor laboratory in January 2019. We utilized germination pots measuring 38.5 cm × 27.5 cm × 14 cm for the experiment and determined the litter weight in accordance with the volume of the germination pot. We conducted a completely randomized experimental design, considering the relative position of seeds on either the litter or soil layer (Zhu et al. 2022b, 2022c, 2023). The relative position of seeds in relation to the litter layer encompasses both seeds located above and beneath the litter layer (Fig. 1). Seeds positioned above the litter layer were placed at heights of 2 cm (about 40 g of litter) and 4 cm (about 80 g of litter). The seeds beneath the litter layer were positioned at different depths, with depths of 2 cm (about 40 g of litter), 4 cm (about 80 g of litter), 6 cm (about 120 g of litter), and 8 cm (about 160 g of litter) of litter layer covering them, respectively (Zhu et al. 2022b, 2022c, 2023). Likewise, seeds placed directly on the forest floor without any litter were considered to be at a relative position of 0 cm. Each relative position includes three experimental pots, with each pot containing 50 seeds of uniform size (about 2.062 ± 0.06 g). The entire experiment comprised a total of 21 experimental pots (Zhu et al. 2022b, 2022c, 2023). The seeds began to germinate in mid-February 2019, with varying germination rates spanning from 5% to 64% based on their relative positions (Zhu et al. 2023). For a more comprehensive understanding of the experimental design, please refer to Zhu et al. (2022b, 2022c, 2023).
Growth data
Starting in April 2019, emerged seedlings, defined as those that successfully penetrated through the litter layer, were counted every 3 days until the number of emerged seedlings ceased to increase for three consecutive times (Zhu et al. 2023). Following seedlings emergence, we recorded the rates of seedlings emergence, which ranged from 16% to 54.7% (Zhu et al. 2022c). Subsequently, we extended the cultivation period 2 months and tagged three seedlings as representatives of the average plant height in the experimental pots. These marked seedlings were then continuously monitored for further growth evaluation. We initially measured their height at the time of selection, and after 6 months of seedling growth, conducted a second measurement to assess the cumulative growth of the seedlings. In each treatment, we harvested a total of nine survival seedlings. The average value of the three survival seedlings in each pot was calculated as the valid value, and each treatment consisted of three pots. The total height of each seedling was measured from the base (litter or soil surface) to the farthest tissue on the main stem. To determine the relative growth rate of the seedlings, we divided the height increment by the number of months between two consecutive censuses. That is, relative grow rate = (ln(H1) − ln(H0))/6, H1 signifies the last most recent measurement of seedling height data, while H0 signifies that the initial measurement of seedling height, 6 was a 6-month growth period (Jiang et al. 2022).
Root and leaf traits
In December 2019, we collected three survival seedlings mentioned above for further measurement for root and leaf traits. The soil attached to each plant tissue was thoroughly cleaned using deionized water. Afterward, the seedlings were carefully packed in sealed plastic bags to preserve moisture during transportation to the laboratory. We utilized filter paper to dry the water present on the plant surface, and promptly separated the root, stem, and leaf tissue.
Three relatively healthy and fully expanded leaves were selected from each survival seedling for further measurement of leaf traits. Leaf area was determined by scanning the leaves using a portable leaf area meter (Yaxin-1242, Beijing YaXin LiYi Technology, Co., Ltd, China). The scanned leaves were weighed using an electronic balance (accurate to 0.0001 g) to record their fresh weight. Subsequently, they were dried in an oven at 65°C for 72 h until a constant weight was achieved to determine their leaf dry mass. Specific leaf area was calculated by dividing the leaf area by the dry mass, while leaf dry matter content was calculated as leaf dry mass divided by the leaf fresh weight (Shen et al. 2019). The nitrogen content of the dry leaves of seedlings was determined by crushing them using a mortar and ball mill, followed by sieving through 0.15-mm mesh. The nitrogen content was then analyzed using an elemental analyzer (VARIO MAX CN, Elementar, Germany). Similarly, the phosphorus content of the leaf tissue was determined using an inductively coupled plasma emission spectrometer (ICP-OES, PE OPTIMA 8000, America).
The roots were scanned using the EPSON root scanner (EPSON Expression ll000XL, Seiko Epson Corp., Japan) to obtain sample image. From the root image, the total root length, root volume, and root average diameter were obtained using Win-RHIZO Pro software. The scanned roots were then packed into the marked envelopes, and dried in an oven at 65°C for 72 h until a constant weight was reached to determine their root dry mass. Specific root length was calculated by dividing the total root length by the root dry mass, while root tissue density was calculated by dividing the root dry mass by the root volume (Shen et al. 2019; Boonman et al. 2020). The dried roots of seedlings were crushed using a mortar and ball mill, and then sieved through a 0.15-mm mesh to determine the total nitrogen content. The nitrogen content was analyzed using an elemental analyzer (VARIO MAX CN, Elementar, Germany). Similarly, the total phosphorus content of the roots tissue was determined using an ICP-OES (PE OPTIMA 8000, America).
Soil data
Soil samples were collected from the root tissue of the aforementioned seedling and subsequently mixed in their corresponding positions within each experimental pot. The mixed soil samples were then transported to the laboratory, where they were divided into three soil subsamples from each relative position. These subsamples were air-dried and sieved through a 0.15-mm mesh. The soil’s total nitrogen and carbon content was determined using an elemental analyzer (VARIO MAX CN, Elementar, Germany). The soil’s total phosphorus content was determined using an ICP-OES (PE OPTIMA 8000, America).
Data analysis
To investigate the impact of root–leaf traits and their combination on the relative grow rate of seedling, we examined a comprehensive set of traits associated with the root and leaf economics spectrum and resource use strategies (Wright et al. 2004; Weemstra et al. 2021). Specially, we considered the following traits: specific leaf area, leaf dry matter content, leaf nitrogen content, leaf phosphorus content, specific root length, average root diameter, root tissue density, root nitrogen content, and root phosphorus content.
We initially conducted a normality test on the data using “shapiro.test” function before performing the relevant analysis. To ensure comparability, all variables were subsequently scaled using the maximum difference normalization method. This entailed calculating the maximum and minimum value for each observation, then determining the difference between them (subtracting the minimum value from maximum value). The scaled data were generated by subtracting each observation by the maximum value and dividing it by the difference between the maximum and minimum values. All scaled data ranged from 0 to 1. To assess the differences in relative growth rate, root-leaf traits of seedling, soil factors among different seed relative positions, the “aov” function from the agricolae package was employed. Additionally, to investigate the relationship between the relative growth rate and key traits, as well as soil characteristic, we utilized principal component analysis. The first two dimensions were extracted using the “prcomp” function from the factoextra package (Kassambara and Mundt 2017). Subsequently, the variations of traits from root to leaf along single or multiple dimension were analyzed to assess the coordination relationship between traits of different tissues.
Furthermore, in order to understand the effect of root–leaf traits, soil characteristic, and seed relative positions on the relative growth rate of seedling, a linear mixed-effect model (Bolker et al. 2009) was employed to identify the key drivers of tree growth. According to the Akaike information criteria (AIC), we used the “stepAIC” function to select the best growth model using maximum likelihood (AIC = −171.06). The relative growth rate of seedling was used as the response variable, with key traits of roots and leaves, and soil characteristic serving as the fixed explanatory variables. The seed relative positions were included as the random variables in the analysis. A linear mixed-effect model was fitted using the “lmer” function in the lme4 package. The correlation of the fixed factors was computed using the lmerTest package. Furthermore, the effectsize package was used to calculate the normalized regression coefficient values. Finally, the “r.squaredGLMM” function within the MuMIn package was employed to estimate the marginal R2 () and conditional R2 (). The marginal R2 represents the variance explained by all fixed effect (fixed explanatory variables), whereas conditional R2 represents the variance explained by both fixed and random effects. The variance of the random effect (seed relative positions) was obtained by subtracting the variance explained by the marginal R2 from the conditional R2. Moreover, we utilized the glmm.hp package (Lai et al. 2022, 2023) to calculate the individual effects explained by the fixed effect in the linear mixed-effects model.
Then, in order to understand the relationship between traits in the roots and leaves and soil factors in relation to relative growth rate, we employed the “dredgedfit” function in the MuMIN package to perform a backward stepwise regression analysis and determine the optimal model. Subsequently, based on the optimal model obtained earlier, we identified the key traits for fitting a linear relationship between the relative growth rate and these traits through a linear regression model. We also considered the interaction among these factors within the best model, where the impact of one variable on the relative growth rate differs across the various levels of the other variable. In order to determine the interaction effect of these variables, we conducted a binary linear regression analysis using the effects package. In addition, we categorized one of the variables into low, medium, and high categories, and utilized “factor (category variable)” upload formula (e.g. relative growth rate ~ lm(variable 1 * factor (variable 2))) to analyze the data. Within this formula, another variable was considered as the categorical variable (low, medium, and high), with the categories low, medium, and high representing the seedlings positioned on the forest floor, beneath the litter layer, and above the litter layer, respectively. All figures presented in this study were generated using the ggplot2 package, while all the statistical analyses were performed in R v.3.4.1 (R Core Team 2020).
RESULTS
Relative growth, root and leaf traits, and soil factors
The relative positions of seeds had a significant impact on the growth rate of seedling, as well as on their root and leaf traits, and soil characteristics, with exception of soil phosphorus content (P < 0.05, Supplementary Table S1). Notably, the variation in specific root length was greater than that of other factors (Supplementary Table S1).
Relationship between relative growth rate and root–leaf traits
The first two dimension of the principal component analysis accounted for 56.6% of the variance (Supplementary Fig. S1). Negative scores on Dimension 1 were associated with a high specific leaf area, root nitrogen content, average root diameter, and root phosphorus content, while positive scores were associated with a high specific root length, leaf dry matter content, and root tissue density. Dimension 2 (Dim 2) explained an additional 21.1% of the variance, with negative scores being correlated with high soil nitrogen content and leaf phosphorus content, and positive scores being correlated with a high relative growth rate.
A significant correlation was observed between the relative growth rate and various factors, such as specific leaf area, leaf dry matter content, specific root length, average root diameter, and soil nitrogen content (Supplementary Table S2). Additionally, we also found that the fixed effects explained 86.1% of the variation in seedling growth (marginal R2) and accounted for 89.3% of the variation in both fixed and random effects (Table 1). Soil nitrogen content was identified as the most important factor in determining seedling growth, followed by leaf dry matter content (Supplementary Fig. S2).
Linear mixed models of the relationship between root, leaf traits, soil factors, and seedling growth.
| Random effect | Response variable | Fixed effect | ||||||
|---|---|---|---|---|---|---|---|---|
| Seed relative position | Relative growth rate | SLA | LDMC | LPC | SRL | AvgD | 0.893 | 0.861 |
| 0.031* | 0.003** | 0.169 | 0.005** | 0.038* | ||||
| RTD | RPC | Soil N | Soil P | Soil C | ||||
| 0.131 | 0.012* | 0.002** | 0.201 | 0.089 | ||||
| Random effect | Response variable | Fixed effect | ||||||
|---|---|---|---|---|---|---|---|---|
| Seed relative position | Relative growth rate | SLA | LDMC | LPC | SRL | AvgD | 0.893 | 0.861 |
| 0.031* | 0.003** | 0.169 | 0.005** | 0.038* | ||||
| RTD | RPC | Soil N | Soil P | Soil C | ||||
| 0.131 | 0.012* | 0.002** | 0.201 | 0.089 | ||||
AvgD, average root diameter; LDMC, leaf dry matter content; LPC, leaf phosphorus content; RPC, root phosphorus content; RTD, root tissue density; SLA, specific leaf area; Soil C, soil carbon content; Soil N, soil nitrogen content; Soil P, soil phosphorus content; SRL, specific root length. , conditional R2; , marginal R2. Signifcance at P < 0.05 and P < 0.01 are indicated by * and **, respectively.
Linear mixed models of the relationship between root, leaf traits, soil factors, and seedling growth.
| Random effect | Response variable | Fixed effect | ||||||
|---|---|---|---|---|---|---|---|---|
| Seed relative position | Relative growth rate | SLA | LDMC | LPC | SRL | AvgD | 0.893 | 0.861 |
| 0.031* | 0.003** | 0.169 | 0.005** | 0.038* | ||||
| RTD | RPC | Soil N | Soil P | Soil C | ||||
| 0.131 | 0.012* | 0.002** | 0.201 | 0.089 | ||||
| Random effect | Response variable | Fixed effect | ||||||
|---|---|---|---|---|---|---|---|---|
| Seed relative position | Relative growth rate | SLA | LDMC | LPC | SRL | AvgD | 0.893 | 0.861 |
| 0.031* | 0.003** | 0.169 | 0.005** | 0.038* | ||||
| RTD | RPC | Soil N | Soil P | Soil C | ||||
| 0.131 | 0.012* | 0.002** | 0.201 | 0.089 | ||||
AvgD, average root diameter; LDMC, leaf dry matter content; LPC, leaf phosphorus content; RPC, root phosphorus content; RTD, root tissue density; SLA, specific leaf area; Soil C, soil carbon content; Soil N, soil nitrogen content; Soil P, soil phosphorus content; SRL, specific root length. , conditional R2; , marginal R2. Signifcance at P < 0.05 and P < 0.01 are indicated by * and **, respectively.
Relative growth rates are determined by interactions between key traits and soil factors
The linear mixed-effect model (Table 2) identifies the combination of soil nitrogen content and root tissue density as the best model for fitting the relative growth rate of seedlings, offering the highest degree of interpretability. This is followed by the combination of leaf dry matter content and soil nitrogen content.
Best growth model based on AIC.
| Fixed effects | Loglik | AICc | |ΔAICc| | Wt |
|---|---|---|---|---|
| Soil N + 1| Treatment | 8.097 | −5.7 | 0 | 0.161 |
| Soil N + RTD + 1| Treatment | 9.565 | −5.1 | 0.6 | 0.122 |
| LDMC + Soil N + 1| Treatment | 9.152 | −4.3 | 1.4 | 0.081 |
| AvgD + Soil N + SRL + 1| Treatment | 10.931 | −3.9 | 1.8 | 0.065 |
| Soil N + SRL + 1| Treatment | 8.802 | −3.6 | 2.1 | 0.057 |
| LDMC + Soil N + SRL + 1| Treatment | 10.771 | −3.5 | 2.2 | 0.055 |
| Fixed effects | Loglik | AICc | |ΔAICc| | Wt |
|---|---|---|---|---|
| Soil N + 1| Treatment | 8.097 | −5.7 | 0 | 0.161 |
| Soil N + RTD + 1| Treatment | 9.565 | −5.1 | 0.6 | 0.122 |
| LDMC + Soil N + 1| Treatment | 9.152 | −4.3 | 1.4 | 0.081 |
| AvgD + Soil N + SRL + 1| Treatment | 10.931 | −3.9 | 1.8 | 0.065 |
| Soil N + SRL + 1| Treatment | 8.802 | −3.6 | 2.1 | 0.057 |
| LDMC + Soil N + SRL + 1| Treatment | 10.771 | −3.5 | 2.2 | 0.055 |
The model only selects the fixed effect combination of |ΔAICc| < 3; AICc, Akaike information criterion corrected for small sample size; ΔAICc, difference between the AIC value of the given model and the AIC value of the best model; Treatment, seed relative positions; Wt, weight of AIC.
Best growth model based on AIC.
| Fixed effects | Loglik | AICc | |ΔAICc| | Wt |
|---|---|---|---|---|
| Soil N + 1| Treatment | 8.097 | −5.7 | 0 | 0.161 |
| Soil N + RTD + 1| Treatment | 9.565 | −5.1 | 0.6 | 0.122 |
| LDMC + Soil N + 1| Treatment | 9.152 | −4.3 | 1.4 | 0.081 |
| AvgD + Soil N + SRL + 1| Treatment | 10.931 | −3.9 | 1.8 | 0.065 |
| Soil N + SRL + 1| Treatment | 8.802 | −3.6 | 2.1 | 0.057 |
| LDMC + Soil N + SRL + 1| Treatment | 10.771 | −3.5 | 2.2 | 0.055 |
| Fixed effects | Loglik | AICc | |ΔAICc| | Wt |
|---|---|---|---|---|
| Soil N + 1| Treatment | 8.097 | −5.7 | 0 | 0.161 |
| Soil N + RTD + 1| Treatment | 9.565 | −5.1 | 0.6 | 0.122 |
| LDMC + Soil N + 1| Treatment | 9.152 | −4.3 | 1.4 | 0.081 |
| AvgD + Soil N + SRL + 1| Treatment | 10.931 | −3.9 | 1.8 | 0.065 |
| Soil N + SRL + 1| Treatment | 8.802 | −3.6 | 2.1 | 0.057 |
| LDMC + Soil N + SRL + 1| Treatment | 10.771 | −3.5 | 2.2 | 0.055 |
The model only selects the fixed effect combination of |ΔAICc| < 3; AICc, Akaike information criterion corrected for small sample size; ΔAICc, difference between the AIC value of the given model and the AIC value of the best model; Treatment, seed relative positions; Wt, weight of AIC.
Based on our best models, we identified soil nitrogen content, root tissue density, and leaf dry matter content as key indicators for predicting variables. Our analysis revealed a significant linear relationship between soil nitrogen content, leaf dry matter content, and relative growth rate of seedling (Fig. 2). However, there is no obvious linear relationship between root tissue density and seedling growth (Supplementary Fig. S3).
Linear relationship of soil nitrogen content (a), leaf dry matter content (b) and seedling relative growth rate. All variables were scaled. Gray area indicates 95% confidence interval; Colored points indicate the observed value of relative growth rate. U2 and U4 indicate that seeds positioned above the litter layer were placed at heights of 2 and 4 cm, respectively. D2, D4, D6, and D8 indicate that seeds beneath the litter layer were positioned at different depths, with depths of 2, 4, 6, and 8 cm of litter layer covering them, respectively. CK indicates that seeds placed directly on the forest floor without any litter.
The interaction between soil nitrogen content and root tissue density had a significant effect, suggesting that the impact of root tissue density varied depending on the soil nitrogen content (Fig. 3a; Supplementary Fig. S4a). Seedling growing in low soil nitrogen content exhibited faster growth rates when they had low root tissue density (CK treatment) compared with other root tissue densities. As nitrogen content increased, the relative growth rate of seedling with high (U treatment) and medium (D treatment) root tissue density expressed a decreasing trend. The same pattern was observed for low leaf dry matter content (Fig. 3b; Supplementary Fig. S4b), where seedling in high soil nitrogen content exhibited a slower relative growth rate. However, as leaf dry matter content increased, the growth rates of seedlings with medium-level soil nitrogen content showed an increasing trend.
Predicted growth (regression lines) for different values of (a) soil nitrogen content and root tissue density, and (b) leaf dry matter content and soil nitrogen content. Different colored (red, blue, and green) regression lines mark model predictions under (a) low, medium, and high root tissue density, (b) low, medium, and high soil nitrogen content, respectively. In this study, root tissue density and soil nitrogen content of emerged seeding from position on the forest floor, beneath the litter layer, and above the litter layer was defined as the low, medium, and high, respectively. y1 (High), y2 (Med), and y3 (Low) indicate the regression formula fitted based on the observed data, respectively.
DISCUSSION
Coordination relationship of between root and leaf traits in different seed relative positions
The coordination relationship of between root and leaf traits is generally considered foundational to our theoretical understanding of plants’ strategies for coping with environmental changes (Shen et al. 2019). A plant resource use strategy, which ranges from conservative to acquisitive, integrates root and leaf traits and proposes to explain individual ecological strategies (Reich 2014). Shen et al. (2019) demonstrated a strong correlation between root and leaf traits, thus providing evidence for the existence of the plant economics spectrum. Additionally, plants experience increased growth by employing contrasting leaf and root resource strategies (Umaña et al. 2021b). Specifically, plants with high specific leaf area and leaf nitrogen content tend to acquire more light and have higher photosynthetic rates, ultimately facilitating accelerated tree growth (Gibert et al. 2016; Umaña et al. 2021b; Weemstra et al. 2021). Here, the first dimension of principal component analysis revealed that leaf tissue displays a resource use strategy that transitions from fast acquisition to conservation across different relative positions. However, in contrast to leaf tissue, plant roots exhibit distinct functional or usage strategies due to the constrains imposed by the complex soil environment (Weigelt et al. 2021). In this study, the first dimension does not clearly depict a transition from conservative to acquisitive strategy for root tissue, as shown in Supplementary Fig. S1. Consequently, it is unlikely to be aligned with the leaf strategy. We speculate that the inconsistency in the coordination of root–leaf traits emphasizes the multidimensional relationship between plant traits (Han and Zhu 2021). Furthermore, the inconsistency raises questions regarding the coordination of root and leaf traits in relation to the relative position of seedlings (Asefa et al. 2022). The relative position beneath the deep litter layer creates a physical barrier for plant growth (Zhang et al. 2022). In this study, seedling that emerge in this position adopt a relatively acquisitive strategy to intercept more available soil resources with their roots (Zhu et al. 2022c). This adaptation enables successful penetration through the litter layer and facilitates growth. In contrast, the relative position beneath the moderate litter layer provides a favorable environment for plant growth. Seedlings that emerge in this position face less competition for soil resources and tend to allocate more energy toward leaf growth in order to capture available sunlight (Zhu et al. 2023). Therefore, in light of the multidimensionality of root traits, it is imperative to consider both the soil environment and plant properties when evaluating resource use strategies (Ding et al. 2020).
Root–leaf traits mediate the growth of seedlings in different seed relative positions
In this study, we founded that the effects of coordination among various plant organs on growth can be predicted by leaf dry matter content and root tissue density, rather than the relative positions of the seeds. Leaf dry matter content, root tissue density, and soil nitrogen content exhibited a negative correlation with the relative growth rate of seedlings (Fig. 2; Supplementary Fig. S3), with soil nitrogen content being the most influential factor (Supplementary Fig. S2). This is primarily attributed to the significance of nitrogen as an essential element for plant growth, and its limitation directly impacts the rate of plant photosynthesis (Zhu et al. 2017; Martini et al. 2020). Previous research has demonstrated that plants exhibit heightened sensitivity to nitrogen elements (N:P < 14) during the seedling stage (Zhu et al. 2023). Furthermore, the supplementation of nitrogen directly facilitates the augmentation of leaf size, thereby promoting a more rapid enhancement of photosynthesis (Martini et al. 2020). Similarly, in a comprehensive study encompassing over 2000 seedlings from 166 species, Martini et al. (2020) determined that soil nitrogen had a detrimental impact on seedling survival in four subtropical forests located in China. Nevertheless, it is crucial to acknowledge that plant roots play a fundamental role in the soil environment. It is also important to recognize that the complexity of soil function and its impact on plant growth cannot be fully encapsulated by a single soil element. Plants exhibiting higher root tissue density tend to adopt slower growth strategies in response to poor soil conditions (Kramer-Walter et al. 2016). Additionally, higher root tissue density indicates enhanced structural flexibility and defensive capabilities of roots, as well as longer lifespan (Kramer-Walter et al. 2016). In regions of the forest floor where nutrient resources are comparatively scarce, seedling typically develop higher root tissue density to enhance their nutrient absorption capacity from the soil. Consequently, this results in a reduction in leaf tissue allocation. The relatively diminished growth rate in such conditions further supports the notion that fast plant growth is inversely associated with low root tissue density (Kramer-Walter et al. 2016). Furthermore, research has indicated that plants with higher leaf dry matter content tend to adopt a more conservative and defensive strategy (Shen et al. 2019; Song et al. 2021). As a result of this strategy, there is a negative correlation between leaf dry matter content and the relative growth rate of plants (Song et al. 2021). This may explain the resistance observed in plants with high leaf dry matter content, especially in the relative position beneath the deep litter layer (Supplementary Fig. S1). However, it is important to note that the coordination of root and leaf traits does not always result in enhanced tree growth. In fact, a medium-level increase in nitrogen content or leaf dry matter content specifically in the seedling position beneath the litter layer has been observed to promote faster growth rates compared with other levels of root tissue density or soil nitrogen content. This positive effect can be attributed to the improvement of the soil environment facilitated by a suitable litter layer (Zhang et al. 2022). Emerged seedling in this position have the ability to coordinate their root and leaf traits in order to optimize the utilization of limited resources. Our findings underscore the notion that the growth rate of seedlings cannot be solely determined by a single trait (Umaña et al. 2021a; Weemstra et al. 2023). Instead, the coordination between root and leaf traits, which significantly influences plant growth, is attributable to the combination of key traits and soil conditions.
Implications of forest regeneration
Here, we have made efforts to elucidate the impacts of the relationship between root and leaf traits and soil factors on the growth of dominant species in subtropical forests. The linear mixed-effects model accounted for approximately 89.3% of the variation in both fixed and random effects, with the individual effect of soil nitrogen being greater than that of root–leaf traits. Hence, the addition of nitrogen may be crucial for seedling growth in this study. In addition, the growth of seedlings in subtropical forests is affected by numerous other factors, such as tree species (Roscher et al. 2018), soil conditions (Song et al. 2018), and microclimate (Song et al. 2018). Moreover, the intricate interplay between plant growth, microclimate, and competition from other species can additionally contribute to the sluggish and unpredictable dynamics of forest regeneration in subtropical forests (Yang et al. 2015). To accelerate the forest regeneration and recovery, it is suggested that several measures be taken into consideration. Firstly, the manual removal of the litter layer prior to seed dispersal should be implemented to enhance the contact between seeds and soil, thereby improving germination and establishment of tree species (Zhang et al. 2022). Furthermore, the artificial cultivation of seedlings should be carried out based on their specific trait’s requirements, with an emphasis on promoting diversity among tree species. Furthermore, when planning a new experiment on species of interests for forest regeneration, it is essential to consider the response to soil nutrients level, particularly nitrogen and phosphorus, which are commonly utilized in subtropical forests (Li et al. 2018). In addition, a long-term monitoring of tree species should be further considered. By implementing these measures, we can effectively evaluate the effectiveness of regeneration strategies and make well-informed decisions for future management practices.
CONCLUSIONS
This study presents an examination of key factors, such as root–leaf traits and soil factors, that influence seedling growth in subtropical forest and their impact on overall plant growth. Our results provide evidence for the existence of a coordination relationship between root and leaf traits of seedlings in different relative positions, with leaf traits displaying a notable variation in response to root traits. Furthermore, we have established a significant correlation between soil nitrogen content and seedling growth, with individual effect of soil nitrogen being relatively greater than that of root–leaf traits. Additionally, the coordination relationship that influences seedling growth, as mentioned earlier, can be predicted by examining root and leaf traits. Gaining an understanding of the relationship between the coordination of root–leaf traits and soil nitrogen content is crucial for predicting forest dynamics and regeneration in subtropical forests. Therefore, we propose the continuation of long-term monitoring to enhance our understanding of the mechanisms that govern the relationship between plant traits and soil factors in plant regeneration within subtropical forests.
Supplementary Material
Supplementary material is available at Journal of Plant Ecology online.
Table S1: Seedling relative growth rate, root-leaf traits, and soil factors in different seed relative positions.
Table S2: Correlation of fixed factors (root-leaf traits and soil factors).
Figure S1: Relationship of relative growth rate, root and leaf traits, and soil factors.
Figure S2: The relative importance of key functional traits and soil environmental factors.
Figure S3: Linear relationship of root tissue density and seedling relative growth rate.
Figure S4: Estimate the interaction effect between soil nitrogen content and root tissue density (a), leaf dry matter content and soil nitrogen content (b).
Acknowledgements
We wish to express our thanks for the support received from the Castanopsis kawakamii Nature Reserve in Sanming, Fujian Province to allow us to collect samples. Meanwhile, we also thank Xin-Guang Gu, Cong Xing, Meng-Ran Jin, Mei-Hua Jia, and Qian-Ru Xiao for the help during the experimental work.
Authors' Contributions
Jing Zhu (Conceptualization, Formal analysis, Investigation, Methodology, Writing—original draft, Writing—review and editing), Xue-Lin Wang (Investigation, Formal analysis), Xing Jin (Investigation), Lan Jiang (Investigation), Hong-Yu Lin (Investigation), Yang Hu (Investigation), Jin-Fu Liu (Conceptualization, Supervision, Writing—review and editing, Funding acquisition), and Zhong-Sheng He (Conceptualization, Supervision, Writing—review and editing, Funding acquisition)
Funding
This work was sponsored by National Natural Science Foundation of China (NSFC) (31700550, 31770678), Fujian Province Forestry and Technology Project of China (2022FKJ21), and Forestry Peak Discipline Construction Project of Fujian Agriculture and Forestry University of China (72202200205).
Conflict of interest statement
The authors declare that they have no conflict of interest.
