Abstract
Slope position, as an important topographic factor, profoundly influences the productivity and carbon allocation strategies of Ochroma lagopus Sw. plantations. In this study, three-year-old O. lagopus trees were subjected to one control (no water-soluble fertilizer, Control) and four fertilization treatments (secondary macronutrients water-soluble fertilizer applied at 600 g/plant, F1; 800 g/plant, F2; 1000 g/plant, F3; and 1200 g/plant, F4) at both upper and lower slope positions to investigate the effects of secondary macronutrients t fertilization on growth and non-structural carbohydrate (NSC) accumulation and allocation. Results showed that: (1) Fertilization significantly promoted tree height growth, whereas diameter at breast height (DBH) exhibited an initial increase followed by a decline, and further increases in the fertilization rate did not yield additional gains. (2) At the upper slope position, carbon was preferentially allocated to leaves and roots during early to middle stages, followed by accelerated translocation to branches for storage during later stages. At the lower slope position, low to moderate fertilization promoted carbon accumulation across organs, whereas excessive fertilization reduced leaf and root carbon pools. (3) Allometric growth and principal component analysis further revealed that carbon allocation tended to be integrated at the upper slope, with a relatively strong correlation between growth and branch carbon storage; in contrast, carbon allocation was relatively dispersed at the lower slope, and the association between growth and carbon storage was weakened. Integrating the variation trends of growth and physiological indices, a relatively high fertilization rate at the upper slope and a moderate fertilization rate at the lower slope may be more beneficial for the growth of O. lagopus. This study reveals a dynamic pattern in which slope position and secondary macronutrients fertilization jointly regulate carbon allocation in O. lagopus, providing a theoretical basis and practical reference for implementing topography-based differentiated precision fertilization in mountain plantations.
Supplementary Information
The online version contains supplementary material available at https://doi.org/10.1186/s12870-026-09286-z.
Keywords: Ochroma lagopus Sw. (balsa tree), Fertilization, Non-structural carbohydrates, Slope heterogeneity
Introduction
Fertilization is a key silvicultural practice for regulating tree growth and development. It directly supplements soil available nutrients, maintains the dynamic balance between tree demand and site supply, and thereby ensures forest productivity. Water-soluble fertilizers are widely used in modern intensive forestry due to their readily absorbable nutrient forms, high use efficiency, and precise dose controllability, representing an important technical means for achieving high-efficiency nutrient management [1]. Secondary macronutrients, mainly including Ca, Mg, and S, are essential mineral nutrients required for plants to complete their life cycle [2], with their demand levels intermediate between macronutrients and micronutrients. In long-term plantation management, due to continuous tree uptake and nutrient removal, the depletion of secondary macronutrients in soils has become increasingly common. This depletion is now a key limiting factor affecting crop growth, wood formation, and forest health [3]. Targeted application of water-soluble secondary macronutrient fertilizers can promote more balanced nutrient uptake in plants [4] and serves as an effective approach to rapidly correct deficiencies, maintain internal ion homeostasis, and enhance stress resistance [5].
Mountain ecosystems are typically characterized by pronounced slope position differentiation [6]. This topographic gradient can profoundly influence tree nutrient acquisition and utilization strategies by mediating water movement, soil erosion, and the spatial redistribution of nutrients. Studies have demonstrated that hilly terrain induces significant spatial heterogeneity in soil secondary macronutrients [7]. At upper slope positions, strong surface runoff and shallow soils render cations such as Ca²⁺ and Mg²⁺ susceptible to continuous depletion through water and soil loss, leading to decreased soil cation exchange capacity and exacerbated acidification [8]. In contrast, at lower slope positions, the deposition of organic matter and clay particles generally enhances the availability of secondary macronutrients, yet these sites may concurrently face risks of inter-element antagonism and combined pollution involving sulfur and heavy metals [9]. However, conventional uniform fertilization practices often overlook such topography-driven spatial heterogeneity in soil nutrients, potentially creating a dual dilemma—insufficient fertilizer efficiency at upper slopes versus excessive environmental loading at lower slopes—which ultimately impedes the balance between economic benefits and ecological safety.
Non-structural carbohydrates (NSC) are an important form of carbon storage in plants, mainly composed of soluble sugars and starch. Their content reflects the level of carbon supply available for plant growth [10]. As the primary products of photosynthesis, NSC participate in various metabolic processes in plants, and their components can be interconverted [11]. Soluble sugars (SS) play crucial roles in regulating cellular osmotic balance, participating in signal transduction, and enhancing cold tolerance [12, 13]. Starch (ST), in contrast, functions primarily as a long-term energy reserve; it can be degraded into sugars under environmental stress to supply energy and provide osmotic protection for damaged tissues [14]. Existing literature has proposed that secondary macronutrients may regulate the accumulation and allocation of NSC in plants through multiple pathways. Ca²⁺ is thought to stabilize cell walls and membrane structures and maintain phloem function, thereby potentially influencing photosynthate transport, and its deficiency is speculated to cause NSC retention in leaves [15]. Mg²⁺, as the central atom of chlorophyll and an enzyme activator, may directly drive photosynthetic carbon assimilation and sucrose synthesis, thereby potentially promoting NSC translocation to sink organs [16]. Sulfur, as a constituent of sulfur-containing amino acids, is speculated to modulate the carbon–nitrogen metabolic balance and affect the activities of sugar metabolism-related enzymes, thereby regulating inter-organ NSC allocation [17]. Therefore, the content and allocation pattern of NSC among various organs not only intuitively reflect the overall carbon supply status of the plant and reveal the allocation rules of photosynthates, but their dynamic changes can further embody the plant’s responses to environmental spatiotemporal heterogeneity [18, 19].
Ochroma lagopus Sw., commonly known as balsa wood, is an evergreen tree species of the family Malvaceae. Native to low-altitude areas of tropical America, it produces the lightest commercial timber in the world [20]. The species is characterized by rapid growth and a short rotation period [21], and its wood is both lightweight and mechanically strong [22], offering broad application prospects in aerospace, navigation, and industrial sectors [23]. With the expansion of its introduction and cultivation, O. lagopus has become an important fast-growing timber species in tropical and subtropical regions. However, as a newly introduced tree species in China, its plantation management system remains underdeveloped, with nutrient management being particularly inadequate. Current practices rely heavily on empirical fertilization, lacking precise regulation based on site conditions and tree physiological demands, which often leads to low fertilizer use efficiency, increased environmental risks, and unstable economic returns.
Recent research on O. lagopus has focused primarily on wood property improvement [20], suitable growing environments [24, 25], branching mechanisms [26], seasonal effects [27], abiotic stress responses [28], and nutrient effects [29–32]. Nevertheless, it remains unclear how topographic factors regulate the synthesis, transport, and storage strategies of carbon assimilation products (i.e., NSC allocation) in O. lagopus by altering soil environmental conditions. To address this knowledge gap, the present study proposes the following scientific hypothesis: slope position and secondary macronutrient fertilization jointly regulate the accumulation and allocation of NSC among different organs of O. lagopus, and this regulatory effect varies dynamically with growth time. Furthermore, plant responses to nutrient supply are dynamic: short-term responses may manifest as rapid adjustments in soluble sugars, whereas long-term effects are reflected in changes in starch reserves. Therefore, in this study, a three-factor controlled experiment incorporating slope position, fertilization rate, and time was conducted to systematically measure growth indicators and the contents of soluble sugars and starch in various organs of O. lagopus. The objectives were to reveal the main effects of slope position on NSC content and inter-organ allocation, to clarify the regulatory role of secondary macronutrients fertilization and its interactive effects with slope position, and to elucidate the temporal dynamics of NSC allocation over the growth period. The results are expected to provide a theoretical basis for developing precision fertilization regimes tailored to slope position differences in O. lagopus plantations, thereby promoting the high-quality development of characteristic economic forest industries in tropical regions.
Materials and methods
Study area
The experiment was conducted from May 15 to December 15, 2024, in a balsawood plantation base at Mengxing Farm, Menglun Town, Mengla County, Xishuangbanna Dai Autonomous Prefecture, Yunnan Province (21°54′–21°55′ N, 101°10′–101°30′ E, altitude approximately 550 m). The site experiences a southwestern tropical monsoon climate with distinct dry and wet seasons. The mean annual temperature is 21.9 °C, with the lowest monthly mean of 16.6 °C in January and the highest of 25.5 °C in June. The annual precipitation is 1520 mm, 85% of which falls between May and October. The soil is a typical Yunnan red soil, classified as Acrisols according to the IUSS Working Group WRB (2015) classification system [33]. Soil analysis showed that it was strongly acidic (pH 4.28), and the basic physicochemical properties are shown in Table 1.
Table 1.
Soil nutrient contents (0–20 cm) at different slope positions (mean ± SD)
| Slope position | Organic carbon/(g·kg⁻¹) | Total nitrogen/(g·kg⁻¹) | Total phosphorus/(g·kg⁻¹) | Total potassium/(g·kg⁻¹) |
|---|---|---|---|---|
| Upper slope position | 15.62 ± 3.54a | 1.61 ± 0.26a | 0.38 ± 0.05a | 11.22 ± 2.07a |
| Lower slope position | 20.10 ± 6.63a | 1.68 ± 0.16a | 0.36 ± 0.07a | 11.63 ± 2.07a |
Different lowercase letters within a column indicate significant differences between slope positions (P < 0.05); n(upper) = 9, n(lower) = 9
Experimental materials and experimental design
The plant materials used in this study were obtained from plantation forests at Mengxing Farm in Mengla County, Xishuangbanna Prefecture, Yunnan Province, China. All sampling activities were permitted by the local administrative authorities. Field identification of the samples was performed by Forestry Engineer Wang Wenbing of Mengxing Farm, and further authentication was conducted by Associate Professor Wu Junwen of Southwest Forestry University. All voucher specimens have been deposited in the Herbarium of Southwest Forestry University, though specific voucher numbers are currently unavailable.
In this study, a field experiment with water‑soluble fertilizers was conducted on three‑year‑old O. lagopus seedlings planted at a spacing of 3 m × 5 m from Mengxing Farm, Menglun Town, Mengla County, Xishuangbanna. The experimental site was located on a south‑facing slope with an inclination of approximately 40°. Two slope positions were selected: an upper slope position at 640 m above sea level and a lower slope position at 590 m. At each slope position, four fertilization treatments (F1: 600 g plant⁻¹; F2: 800 g plant⁻¹; F3: 1000 g plant⁻¹; F4: 1200 g plant⁻¹) plus one unfertilized control were established. The water‑soluble fertilizer (Shandong Runkang Fertilizer Industry Co., Ltd., China) was a secondary macronutrients fertilizer containing Mg (≥ 10%) and S (≥ 10%), supplemented with micronutrients including calcium, iron, zinc, and boron. Each treatment was replicated three times, with 30 trees per replicate, resulting in 450 trees per slope position and 900 trees in total across both slopes. The fertilizer rates were determined based on the product characteristics, local climate, common agronomic practice, and previous studies on fast‑growing tropical tree species [34]. Due to high soil moisture during the rainy season, fertilization was performed by placing the fertilizer directly into the root zone and covering it with soil. Each fertilization treatment was established with three spatial replicates (i.e., three independent plots), each containing 30 trees. Consequently, each treatment comprised a total of 90 trees, each slope position (including five treatments) contained 450 trees, and the total number of trees across the two slope positions was 900. The first fertilizer application was conducted on 15 May 2024, and the second (final) application was completed on 15 June 2024, with an interval of approximately 30 days between the two applications. The total application rates for the four fertilization treatments were 600, 800, 1000, and 1200 g per tree, respectively, and half of the total rate was applied at each application. The control treatment received no fertilizer. To comprehensively capture the dynamic changes in NSC allocation after fertilization, plant samples were collected at 120, 150, and 180 days after the final fertilization. These three sampling time points covered the late stage of the main growing season and the subsequent critical physiological adjustment phase of O. lagopus, effectively reflecting the dynamic transition of fertilization effects from the peak growth period to the post-growing season. Therefore, this study employed a three-factor controlled experiment (slope position × fertilization rate × time), with three plots serving as spatial replicates and three sampling time points serving as temporal replicates.
Indicator measurements
Tree height and DBH were measured prior to the experiment and at 120, 150, and 180 days after the final fertilization. At each measurement, nine representative trees were randomly selected from all plots within each treatment. Height was determined using a hypsometer with a precision of 0.1 cm, while DBH was measured with a diameter tape to the nearest 0.1 mm [35]. To rule out the effect of initial tree size on the final results, initial tree height and DBH were measured before the experiment (Table 2). One‑way ANOVA conducted within each slope position (upper and lower) showed no significant differences among fertilizer treatments (P > 0.05), indicating that the treatments started with comparable tree sizes.
Table 2.
Initial tree height and DBH at different slope positions
| Slope position | Fertilization treatment | Initial tree height (m) | Initial DBH (cm) |
|---|---|---|---|
| Upper slope position | Control | 8.68 ± 0.31a | 11.59 ± 1.04a |
| F1 | 8.66 ± 0.26a | 11.53 ± 0.84a | |
| F2 | 8.69 ± 0.24a | 11.60 ± 0.74a | |
| F3 | 8.65 ± 0.25a | 11.48 ± 0.82a | |
| F4 | 8.69 ± 0.25a | 11.58 ± 0.78a | |
| Lower slope position | Control | 11.53 ± 0.23a | 15.68 ± 1.26a |
| F1 | 11.51 ± 0.19a | 15.50 ± 1.11a | |
| F2 | 11.51 ± 0.21a | 15.56 ± 1.23a | |
| F3 | 11.53 ± 0.19a | 15.60 ± 1.09a | |
| F4 | 11.51 ± 0.20a | 15.53 ± 1.10a |
Different lowercase letters indicate significant differences among different fertilizer treatments within the same slope position (P < 0.05)
Simultaneously with each growth measurement, nine sample trees were randomly selected from all plots within each treatment (i.e., nine trees per treatment) for the collection of plant tissue samples used in physicochemical analyses. Prior to chemical analysis, the samples from each individual tree were determined independently without pooling. From each tree, sun-exposed current-year shoots (including leaves) from the mid-canopy were collected using pruning shears. Corresponding root systems were also excavated. All collected samples were labeled, transported to the laboratory, and sequentially processed by washing, drying, and grinding. The protocol involved an initial oven-drying step at 120 °C for 30 min to deactivate enzymes rapidly. Subsequently, samples were dried at a constant temperature of 80 °C until a constant weight was achieved, ensuring complete moisture removal. Finally, the dried material was ground using a mill, passed through a 100-mesh sieve, and stored in sealed containers for subsequent analysis.
The contents of soluble sugars and starch were determined using the phenol-sulfuric acid method [36]. Soluble Sugar Determination: For each of the sampled types, leaves, branch and roots, SS were measured by accurately weighing 0.1 g of sample into a 10 mL centrifuge tube. After adding 8 mL of distilled water, the mixture was extracted in a boiling water bath for 30 min. Following cooling, the extract was centrifuged at 4000 r min⁻¹ for 10 min. The supernatant was collected and diluted to a final volume of 50 mL. For the assay, 1 mL of the diluted extract was mixed with 1 mL of distilled water, 1 mL of 9% phenol solution, and 5 mL of concentrated sulfuric acid. The mixture was thoroughly vortexed and allowed to stand at room temperature for 30 min. Absorbance was then measured at 485 nm using a UV-Vis spectrophotometer (UV-2600, Shimadzu, Japan). Starch Determination: The residue remaining after soluble sugar extraction was used for starch analysis. To this residue, 6 mL of distilled water was added, and the mixture was gelatinized in a boiling water bath for 15 min. After cooling, 2 mL of perchloric acid was added for hydrolysis. The mixture was then centrifuged, and the supernatant was diluted to 50 mL. The same colorimetric procedure described above for soluble sugars was applied to determine the starch content. Soluble sugar content was calculated based on a glucose standard curve. The starch content, obtained from the absorbance reading, was multiplied by a conversion factor of 0.9. Total NSC content was calculated as the sum of soluble sugars and starch.
Statistical analysis
Data were organized using Microsoft Excel 2016. One-way ANOVA and Levene’s test were performed in SPSS 27.0. All other statistical analyses were conducted in the R environment (version 4.3.1).
To examine the effects of slope position, fertilization, and time on tree height and DBH of balsa wood, linear mixed-effects models were constructed using the lme4 package. The experimental design included three blocks, each containing 30 trees. Because the same trees were measured repeatedly throughout the experiment, tree was treated as a random effect nested within block. The model formula was: Y = slope position × fertilization × time + Initial_size + (1 | block/plant_id), where the fixed effects comprised the three-way interaction among slope position, fertilization, and time, along with the covariate initial size (initial height for tree height, initial DBH for DBH). Random effects consisted of block and tree individuals nested within block. Inference for fixed effects was conducted using type III ANOVA with Satterthwaite approximation for degrees of freedom via the lmerTest package, yielding numerator degrees of freedom, denominator degrees of freedom, F-values, and exact P-values.
NSC determination involved destructive sampling; trees sampled at 120, 150, and 180 days were different individuals. Therefore, time was treated solely as a fixed effect, and tree ID was not included as a random effect. For each NSC component (soluble sugars, starch, total NSC, and sugar-to-starch ratio in leaves, branches, and roots), the following model was fitted: log(Indicator) = slope position × fertilization × time + (1 | block), where the fixed effects were the three-way interaction of slope position, fertilization, and time; block served as the only random intercept. Because no corresponding initial values were available for NSC data, no covariate was included in the model. When heteroscedasticity was detected for an indicator, natural log transformation was applied to satisfy the assumption of homogeneity of variance, such that the left-hand side of the model became log(Indicator).
For all response variables, linear mixed-effects models were first fitted on the original scale. Homogeneity of variance was assessed using plots of standardized residuals versus fitted values, boxplots of residuals against treatment combinations, and Levene’s test (α = 0.05). If Levene’s test was significant (P < 0.05), the response variable was natural log-transformed, and diagnostics were repeated until the model satisfied the homogeneity assumption (P > 0.05). The results showed that tree height met the homogeneity assumption on the original scale; for DBH and all NSC components, variances stabilized after natural log transformation, with Levene’s tests being non-significant (P > 0.05). Normality was qualitatively evaluated using Q–Q plots, and no serious deviations were observed. Residual plots for all transformed models exhibited random scatter without systematic patterns (Supplementary Fig. S1).
Type III ANOVA from the lmerTest package was used to obtain numerator degrees of freedom, denominator degrees of freedom, F-values, and exact P-values (presented in scientific notation with three significant digits) for fixed effects. When interaction terms were significant, estimated marginal means were computed using the emmeans package, and post-hoc multiple comparisons were performed using Tukey’s HSD test. For log-transformed indicators, marginal means were back-transformed to the original scale (mg/g or mm) using type = “response”, and significant groupings were denoted by compact letter display. All statistical inferences were made at the 0.05 significance level.
To evaluate the robustness of the results to heteroscedasticity, fixed effects were additionally inferred using cluster-robust standard errors (CR2 type, clubSandwich package). Clustering was set at the block level to correct for potential heteroscedasticity. This analysis employed exactly the same model structure as the main analysis (log-transformed scale for DBH and NSC components). P-values obtained with robust standard errors were compared with those from the main linear mixed models to verify the consistency of conclusions. Detailed results of the sensitivity analysis are presented in the Results section and Supplementary Material (Supplementary Table S1).
Pearson correlation analysis was performed separately for growth indicators and NSC components at 180 days after fertilization for each slope position. P-values were adjusted for multiple comparisons using the false discovery rate (FDR) correction in R. Significant results after correction are presented in tabular form, with complete data available in Supplementary Table S2. Principal component analysis (PCA) was performed based on the correlation matrix (variables standardized) using Origin 2024, and biplots were generated to reveal the association strengths among indicators and the comprehensive response patterns under treatments. Eigenvalues, proportions of variance explained, cumulative contributions, and principal component loading matrices are provided in Supplementary Tables S3a and S3b. Line graphs were constructed using marginal means ± standard errors extracted from the linear mixed models. For log-transformed indicators, means represent geometric means on the original scale after back-transformation.
To explore the proportional relationships of NSC among different organs of balsa wood, allometric equations y = axb were fitted based on individual plant data (n = 9 per treatment combination, total N = 90). The equation was log‑transformed into a linear form: log y = b log x + log a, where x is the NSC content of a given organ and y is the total NSC content (sum of leaf, branch, and root NSC). In this model, a is the allometric constant (intercept of the linear relationship) and b is the allometric exponent (slope). An exponent of b = 1 indicates isometric scaling (constant proportion) between the two components, whereas b < 1 or b > 1 indicates allometric scaling. The smatr package in R was used to first test for heterogeneity of slopes among treatments [37]. If slopes were not significantly different (P > 0.05), differences in elevation (intercept) and shift along the common slope were further tested.
Results
Effects of fertilization on tree height and DBH inO. lagopusplantations at different slope positions
Across the 180-day observation period, growth dynamics of O. lagopus in both height and DBH were independently and significantly driven by the fixed main effects of slope position, fertilization, and time (P < 0.01). Notably, a significant two-way interaction between slope position and fertilization was detected exclusively for height growth (P < 0.01), indicating a synergistic or antagonistic modulation that was organ-specific; no such interactive effect was observed for diameter increment. All interactions involving the temporal dimension—including Slope × Time, Fertilization × Time, and the three-way Slope × Fertilization × Time—failed to reach statistical significance (Table 3). The absence of significant interactions with time suggests that the treatment-induced hierarchies in growth rates remained largely consistent and did not exhibit crossover reversals across repeated measures. Random-effect variance components for height and DBH models are presented in Table 4. In both models, block variance (σ²_block) was considerably smaller than plant and residual variances, indicating negligible differences among blocks; plant-level variance was an important source of growth variation, supporting the inclusion of plants nested within blocks as random effects.
Table 3.
Linear mixed-effects model analysis of slope position, fertilization, and time effects on growth and NSC traits of O. lagopus
| Index | Effect | NumDF | DenDF | F value | P-value |
|---|---|---|---|---|---|
| Height of tree | Slope position | 1 | 77.0 | 24.819 | 3.77 × 10 6*** |
| Fertilization | 4 | 77.0 | 46.308 | 9.13 × 10 20*** | |
| Time | 2 | 160.0 | 23.271 | 1.35 × 10 9*** | |
| Slope position×Fertilization | 4 | 77.0 | 3.753 | 7.66 × 10 3** | |
| Slope position×Time | 2 | 160.0 | 0.146 | 0.865 | |
| Fertilization×Time | 8 | 160.0 | 0.134 | 0.998 | |
| Slope position×Fertilization×Time | 8 | 160.0 | 0.134 | 0.998 | |
| DBH | Slope position | 1 | 77.3 | 50.675 | 4.86 × 10 10*** |
| Fertilization | 4 | 77.0 | 10.903 | 4.68 × 10 7*** | |
| Time | 2 | 160.0 | 30.853 | 4.65 × 10 12*** | |
| Slope position×Fertilization | 4 | 77.0 | 2.391 | 0.058 | |
| Slope position×Time | 2 | 160.0 | 0.084 | 0.920 | |
| Fertilization×Time | 8 | 160.0 | 0.547 | 0.819 | |
| Slope position×Fertilization×Time | 8 | 160.0 | 1.146 | 0.336 | |
| Leaf SS | Slope position | 1 | 238.0 | 185.908 | 1.14 × 10 31*** |
| Fertilization | 4 | 238.0 | 2.480 | 0.045* | |
| Time | 2 | 238.0 | 1316.607 | 2.01 × 10 129*** | |
| Slope position×Fertilization | 4 | 238.0 | 5.984 | 1.32 × 10 4*** | |
| Slope position×Time | 2 | 238.0 | 7.217 | 9.06 × 10 4*** | |
| Fertilization×Time | 8 | 238.0 | 8.540 | 3.24 × 10 10*** | |
| Slope position×Fertilization×Time | 8 | 238.0 | 10.650 | 8.89 × 10 13*** | |
| Leaf ST | Slope position | 1 | 240.0 | 171.982 | 5.48 × 10 30*** |
| Fertilization | 4 | 240.0 | 3.550 | 0.008** | |
| Time | 2 | 240.0 | 577.097 | 2.03 × 10 92*** | |
| Slope position×Fertilization | 4 | 240.0 | 14.227 | 1.96 × 10 10*** | |
| Slope position×Time | 2 | 240.0 | 16.610 | 1.75 × 10 7*** | |
| Fertilization×Time | 8 | 240.0 | 6.211 | 2.70 × 10 7*** | |
| Slope position×Fertilization×Time | 8 | 240.0 | 9.170 | 5.31 × 10 11*** | |
| Leaf NSC | Slope position | 1 | 238.0 | 290.559 | 4.04 × 10 43*** |
| Fertilization | 4 | 238.0 | 2.560 | 0.039* | |
| Time | 2 | 238.0 | 1551.009 | 3.07 × 10 137*** | |
| Slope position×Fertilization | 4 | 238.0 | 3.519 | 0.008** | |
| Slope position×Time | 2 | 238.0 | 4.438 | 0.013* | |
| Fertilization×Time | 8 | 238.0 | 8.538 | 3.27 × 10 10*** | |
| Slope position×Fertilization×Time | 8 | 238.0 | 10.432 | 1.62 × 10 12*** | |
| Leaf SS/Leaf ST | Slope position | 1 | 240.0 | 8.037 | 0.005** |
| Fertilization | 4 | 240.0 | 3.075 | 0.017* | |
| Time | 2 | 240.0 | 602.019 | 2.99 × 10 94*** | |
| Slope position×Fertilization | 4 | 240.0 | 14.264 | 1.84 × 10 10*** | |
| Slope position×Time | 2 | 240.0 | 20.419 | 6.47 × 10 9*** | |
| Fertilization×Time | 8 | 240.0 | 8.259 | 7.07 × 10 10*** | |
| Slope position×Fertilization×Time | 8 | 240.0 | 9.351 | 3.18 × 10 11*** | |
| Branch SS | Slope position | 1 | 238.0 | 504.060 | 1.06 × 10 60*** |
| Fertilization | 4 | 238.0 | 22.559 | 8.21 × 10 16*** | |
| Time | 2 | 238.0 | 17.631 | 7.24 × 10 8*** | |
| Slope position×Fertilization | 4 | 238.0 | 5.604 | 2.51 × 10 4*** | |
| Slope position×Time | 2 | 238.0 | 347.345 | 2.59 × 10 71*** | |
| Fertilization×Time | 8 | 238.0 | 4.807 | 1.69 × 10 5*** | |
| Slope position×Fertilization×Time | 8 | 238.0 | 17.088 | 5.24 × 10 20*** | |
| Branch ST | Slope position | 1 | 238.0 | 173.892 | 3.56 × 10 30*** |
| Fertilization | 4 | 238.0 | 6.161 | 9.83 × 10 5*** | |
| Time | 2 | 238.0 | 418.936 | 1.08 × 10 78*** | |
| Slope position×Fertilization | 4 | 238.0 | 14.133 | 2.31 × 10 10*** | |
| Slope position×Time | 2 | 238.0 | 43.039 | 1.11 × 10 16*** | |
| Fertilization×Time | 8 | 238.0 | 11.979 | 2.42 × 10 14*** | |
| Slope position×Fertilization×Time | 8 | 238.0 | 7.077 | 2.18 × 10 8*** | |
| Branch NSC | Slope position | 1 | 238.0 | 670.877 | 3.38 × 10 71*** |
| Fertilization | 4 | 238.0 | 29.670 | 5.01 × 10 20*** | |
| Time | 2 | 238.0 | 126.175 | 4.39 × 10 38*** | |
| Slope position×Fertilization | 4 | 238.0 | 8.910 | 1.01 × 10 6*** | |
| Slope position×Time | 2 | 238.0 | 205.295 | 1.54 × 10 52*** | |
| Fertilization×Time | 8 | 238.0 | 11.265 | 1.66 × 10 13*** | |
| Slope position×Fertilization×Time | 8 | 238.0 | 19.126 | 4.04 × 10 22*** | |
| Branch SS/Branch ST | Slope position | 1 | 238.0 | 88.753 | 4.09 × 10 18*** |
| Fertilization | 4 | 238.0 | 12.262 | 4.42 × 10 9*** | |
| Time | 2 | 238.0 | 234.212 | 5.94 × 10 57*** | |
| Slope position×Fertilization | 4 | 238.0 | 7.550 | 9.65 × 10 6*** | |
| Slope position×Time | 2 | 238.0 | 301.535 | 5.72 × 10 66*** | |
| Fertilization×Time | 8 | 238.0 | 7.275 | 1.23 × 10 8*** | |
| Slope position×Fertilization×Time | 8 | 238.0 | 12.853 | 2.37 × 10 15*** | |
| Root SS | Slope position | 1 | 238.0 | 670.100 | 3.74 × 10 71*** |
| Fertilization | 4 | 238.0 | 39.867 | 1.53 × 10 25*** | |
| Time | 2 | 238.0 | 535.280 | 8.21 × 10 89*** | |
| Slope position×Fertilization | 4 | 238.0 | 14.517 | 1.27 × 10 10*** | |
| Slope position×Time | 2 | 238.0 | 181.628 | 1.27 × 10 48*** | |
| Fertilization×Time | 8 | 238.0 | 48.294 | 1.02 × 10 45*** | |
| Slope position×Fertilization×Time | 8 | 238.0 | 6.154 | 3.22 × 10 7*** | |
| Root ST | Slope position | 1 | 238.0 | 28.769 | 1.93 × 10 7*** |
| Fertilization | 4 | 238.0 | 5.435 | 3.33 × 10 4*** | |
| Time | 2 | 238.0 | 156.024 | 5.07 × 10 44*** | |
| Slope position×Fertilization | 4 | 238.0 | 3.167 | 0.015* | |
| Slope position×Time | 2 | 238.0 | 34.084 | 9.62 × 10 14*** | |
| Fertilization×Time | 8 | 238.0 | 19.997 | 5.36 × 10 23*** | |
| Slope position×Fertilization×Time | 8 | 238.0 | 7.259 | 1.29 × 10 8*** | |
| Root NSC | Slope position | 1 | 238.0 | 735.810 | 9.07 × 10 75*** |
| Fertilization | 4 | 238.0 | 30.059 | 3.01 × 10 20*** | |
| Time | 2 | 238.0 | 643.049 | 1.08 × 10 96*** | |
| Slope position×Fertilization | 4 | 238.0 | 10.895 | 3.97 × 10 8*** | |
| Slope position×Time | 2 | 238.0 | 241.861 | 4.64 × 10 58*** | |
| Fertilization×Time | 8 | 238.0 | 43.838 | 9.95 × 10 43*** | |
| Slope position×Fertilization×Time | 8 | 238.0 | 4.830 | 1.58 × 10 5*** | |
| Root SS/Root ST | Slope position | 1 | 238.0 | 265.106 | 1.47 × 10 40*** |
| Fertilization | 4 | 238.0 | 25.457 | 1.44 × 10 17*** | |
| Time | 2 | 238.0 | 142.148 | 2.40 × 10 41*** | |
| Slope position×Fertilization | 4 | 238.0 | 15.540 | 2.61 × 10 11*** | |
| Slope position×Time | 2 | 238.0 | 40.461 | 7.48 × 10 16*** | |
| Fertilization×Time | 8 | 238.0 | 26.002 | 1.07 × 10 28*** | |
| Slope position×Fertilization×Time | 8 | 238.0 | 9.578 | 1.74 × 10 11*** |
Leaf SS, Leaf ST, Leaf NSC, and Leaf SS/Leaf ST refer to leaf soluble sugar, starch, total non-structural carbohydrate content, and the soluble sugar-to-starch ratio, respectively; the same applies to branch and root organs. Exact P-values are reported in scientific notation, with asterisks serving only as significance indicators: * indicates P<0.05,** indicates P<0.01, ***indicates P < 0.001. Model formulas for all linear mixed models are as follows: for height, H ~ altitude × fertilizer × time + Initial_Height + (1 | block/plant_id); for DBH, log(DBH) ~ altitude × fertilizer × time + Initial_DBH + (1 | block/plant_id); for all NSC traits, log(Indicator) ~ altitude × fertilizer × time + (1 | block). Sample size: n = 270 observations (90 individual trees × 3 sampling times)
Table 4.
Random-effect variance components of linear mixed models for height and DBH
| Indicator | σ²_block | σ²_plant: block | σ²_residual |
|---|---|---|---|
| Height of tree | 3.18 × 10 4 | 0.0322 | 0.1970 |
| DBH(log scale) | 1.03 × 10 4 | 0.0032 | 0.0061 |
σ²_block = variance among blocks; σ²_plant: block = variance among plants nested within blocks; σ²_residual = residual variance
After reanalysis using cluster-robust standard errors (CR2), the significance conclusions for fixed effects of height and all NSC components were fully consistent with the primary analysis. However, for DBH, the slope position × fertilization interaction was not significant in the primary analysis (P = 0.058) but became significant under robust standard errors (P = 0.0044). The significance of all other fixed effects remained consistent between the two methods. Given that robust standard errors do not rely on the assumption of homoscedasticity, we consider this interaction effect to be genuine, while the conventional model may have produced a conservative inference due to heteroscedasticity. Complete results of the sensitivity analysis are provided in Supplementary Table-S1.
The growth-promoting effects of fertilization on O. lagopus exhibited distinctly divergent patterns between the two slope positions (Fig. 1). The three sampling time points (120, 150, and 180 days after the final fertilization) covered the late stage of the main growing season and the subsequent critical physiological adjustment phase. At the upper slope, tree height displayed a consistently increasing trend along the fertilization gradient, reaching its peak in the F4 high-fertilization treatment at all observation time points and significantly exceeding the control. In contrast, the peak response of DBH consistently occurred in the moderate-fertilization F3 treatment. At the lower slope, both tree height and DBH showed a consistent nonlinear response pattern, characterized by a unimodal curve that increased first and then declined with increasing fertilization rate, both attaining their maximum values in the F2 treatment and being significantly higher than the control. Under the F3 and F4 treatments, growth indicators remained higher than those of the control but did not surpass the F2 level, indicating that further increases in fertilization rate did not yield additional growth gains. In terms of temporal dynamics, growth indicators of all treatments at the upper slope generally exhibited a steadily rising trend, reflecting the cumulative effect of fertilization; at the lower slope, temporal changes among treatments were relatively moderate, with the superiority of the F2 treatment remaining stable across all three time points. Overall, tree height and DBH were significantly greater at the lower slope than at the upper slope.
Fig. 1.
Effect of fertilization on tree height and DBH of O. lagopus at different slopes. Figure (A) refers to the height of tree and Figure (B) refers to the diameter at breast height. Data are marginal means, with error bars representing standard error (SE). Control (no water-soluble fertilizer), and F1, F2, F3, and F4 indicate 600, 800, 1000, and 1200 g plant⁻¹ water-soluble fertilizer, respectively. Sample size: n = 270 observations (90 individual trees × 3 sampling times). Statistical method: linear mixed model (LMM) with fixed effects of slope position × fertilization × time, initial value as a covariate, and random effects of block and plant individual. Post-hoc multiple comparisons were performed using Tukey's HSD test. Different lowercase letters indicate significant differences between different fertilization treatments at the same sampling time (P<0.05)
Effects of fertilization on leaf NSC in O. lagopus at different slope positions
The responses of leaf NSC components to treatment factors exhibited a clearly stratified pattern (Table 3). Fertilization treatment had significant main effects on leaf soluble sugars, total NSC content, and the SS/ST ratio (P < 0.05), and the fertilization × time interaction significantly affected leaf NSC (P < 0.05). In contrast, slope position, sampling time, and all two-way and three-way interactions exerted highly significant regulatory effects on all leaf NSC components—SS, ST, total NSC, and the SS/ST ratio (P < 0.01). These results indicate that although the dynamics of the leaf NSC pool are independently driven by fertilization, they are dominated to a greater extent by the combined effects of environmental factors and their temporal evolution.
The effects of fertilization on leaf NSC components in O. lagopus exhibited pronounced dynamic differentiation with slope position and observation time (Fig. 2). At the upper slope, leaf SS content displayed a unimodal response to increasing fertilization rate, first rising and then declining, at both 120 and 150 days after fertilization, reaching its maximum in the F3 treatment at 120 days; by 180 days after fertilization, no consistent pattern was observed among fertilization treatments. The dynamics of total leaf NSC content closely mirrored those of SS, also exhibiting a unimodal curve during the early-to-middle observation period, with maximum values in the F3 treatment at 120 days and in the F2 treatment at 150 days, while no clear trend among treatments was detected at 180 days. Leaf ST content showed an increasing trend at 120 days, no obvious trend at 150 days, and a pattern of first decreasing and then increasing at 180 days, with the minimum occurring in the F1 treatment. The leaf SS/ST ratio exhibited a consistent trend of initially increasing and then decreasing at all time points after fertilization, reaching its maximum in the F2 treatment at 180 days.
Fig. 2.
Effect of fertilization on NSC of O. lagopus leaves at different slopes. Figure (A) refers to the leaf soluble sugar content, Figure (B) refers to the leaf starch content, Figure (C) refers to the leaf NSC content and Figure (D) refers to the leaf soluble sugar/leaf starch. Data are marginal means, with error bars representing standard error (SE). Control refers to no water-soluble fertilizer treatment, 600g/ water-soluble fertilizer applied to F1, 800g/ water-soluble fertilizer applied to F2, 1000g/ water-soluble fertilizer applied to F3 and 1200g/ water-soluble fertilizer applied to F4. Sample size: n = 270 observations (90 individual trees × 3 sampling times). Statistical method: linear mixed model (LMM) with fixed effects of slope position × fertilization × time, and random effects of block. Post-hoc multiple comparisons were performed using Tukey's HSD test. Different lowercase letters indicate significant differences between different fertilization treatments at the same sampling time (P<0.05)
At the lower slope, leaf SS content exhibited a unimodal pattern, increasing first and then declining, at 150 days after fertilization, reaching its maximum in the F3 treatment. In contrast, at 120 and 180 days, SS displayed a decreasing-then-increasing trend; at 180 days, the F3 treatment was significantly lower than the Control. Total leaf NSC content largely mirrored the trajectory of SS, peaking in the F3 treatment at 150 days (significantly higher than the Control) and declining to its minimum in F3 at 180 days. Leaf ST content showed a decreasing-then-increasing trend at 120 days, with the minimum occurring in the F4 treatment; at 180 days, it reached its maximum in the F1 treatment, while no significant differences among treatments were detected at 150 days. The leaf SS/ST ratio generally increased at 120 and 150 days, peaking in the F4 and F3 treatments respectively, and then decreased at 180 days.
In terms of temporal dynamics, leaf SS and total NSC content in all treatments at both slope positions accumulated continuously over time, with the rate of accumulation markedly accelerating during the 150–180 day period. The leaf SS/ST ratio showed an overall increasing trend throughout the observation period. In contrast, leaf ST content exhibited a V-shaped fluctuation (decreasing then increasing) at both slope positions, reaching a relatively low level around 150 days. Overall, the contents of all leaf NSC components were generally higher at the upper slope than at the lower slope, a pattern opposite to the faster tree growth observed at the lower slope (Fig. 1).
Effects of fertilization on branch NSC in O. lagopus at different slope positions
All branch NSC components exhibited highly significant overall responses to all treatment factors and their interactions (P < 0.01; Table 3), indicating that branch carbon pool dynamics are driven by multi-level environmental perturbations.
The responses of branch NSC components to fertilization displayed pronounced spatiotemporal differentiation with slope position and observation time (Fig. 3). At the upper slope, branch SS content was significantly suppressed under the low fertilization treatment (F1) at both 120 and 150 days after fertilization. By 180 days, this response underwent a directional reversal, shifting to a positive, continuously increasing trend with rising fertilization rate, with the F3 and F4 treatments being significantly higher than the Control. The dynamics of total branch NSC content closely mirrored those of SS, reaching minimum values in the F2 and F1 treatments at 120 and 150 days, respectively, and significantly lower than the Control; at 180 days, branch NSC exhibited an increasing trend, reaching its maximum in the F4 treatment. Branch ST) content showed a suppressed pattern at all observation time points, displaying a decreasing trend at 120 days, and a decreasing-then-increasing trend at 150 and 180 days, with the low-to-moderate fertilization treatments being significantly lower than the Control. The branch SS/ST ratio exhibited a decreasing-then-increasing trend at 120 days after fertilization, reversed to an increasing-then-decreasing trend by 180 days with a maximum in the F3 treatment, while no significant changes were observed at 150 days.
Fig. 3.
Effect of fertilization on NSC of O. lagopus branches at different slopes. Figure (A) refers to the branch soluble sugar content, Figure (B) refers to the branch starch content, Figure (C) refers to the branch NSC content and Figure (D) refers to the branch soluble sugar/branch starch. Data are marginal means, with error bars representing standard error (SE). Control refers to no water-soluble fertilizer treatment, 600 g/ water-soluble fertilizer applied to F1, 800 g/ water-soluble fertilizer applied to F2, 1000 g/ water-soluble fertilizer applied to F3 and 1200 g/ water-soluble fertilizer applied to F4. Sample size: n = 270 observations (90 individual trees × 3 sampling times). Statistical method: linear mixed model (LMM) with fixed effects of slope position × fertilization × time, and random effects of block. Post-hoc multiple comparisons were performed using Tukey’s HSD test. Different lowercase letters indicate significant differences between different fertilization treatments at the same sampling time (P < 0.05)
At the lower slope, branch SS content displayed an increasing-then-decreasing trend at 120 and 150 days after fertilization, reaching maximum values in the F2 and F3 treatments, respectively, with no significant changes detected at 180 days. Branch ST content showed an increasing trend at 150 days, peaking in the F4 treatment, while no clear trends among treatments were observed at 120 and 180 days. Branch NSC content exhibited an increasing-then-decreasing trend at 120 days, with a maximum in the F2 treatment, an increasing trend at 150 days, and no apparent trend by 180 days. The branch SS/ST ratio followed an increasing-then-decreasing trend at both 120 and 180 days, reaching maxima in the F2 treatment at both time points, whereas no clear trend was evident at 150 days.
In terms of temporal dynamics, branch SS, NSC, and SS/ST ratio in all fertilization treatments at the upper slope generally increased over time, with the rate of increase markedly accelerating during the 150–180 day period. Notably, the Control exhibited an opposite temporal trend, characterized by an initial increase followed by a decline. Branch ST content exhibited a V-shaped fluctuation (decreasing then increasing) at both slope positions. At the lower slope, temporal changes in branch SS and total NSC content were relatively moderate, and the branch SS/ST ratio showed an increasing-then-decreasing trend. Similar to the pattern observed in leaves, the contents of all branch NSC components were generally higher at the upper slope than at the lower slope. However, at 150 days, the branch SS/ST ratio was generally higher at the lower slope than at the upper slope, indicating a greater relative proportion of SS in branches at the lower slope. This pattern further corroborates the divergent characteristics whereby the upper slope tends to accumulate carbon reserves, whereas the lower slope tends to favor growth.
Effects of fertilization on root NSC in O. lagopus at different slope positions
The responses of root NSC components to treatment factors exhibited a stratified pattern: the slope position × fertilization interaction significantly affected root starch content (P < 0.05), while slope position, fertilization treatment, sampling time, and all two-way and three-way interactions exerted highly significant effects on all root NSC components (P < 0.01; Table 3). In the mixed models for all NSC indicators, the block variance (σ²_block) was substantially smaller than the residual variance, and for most indicators the block variance approached zero (Table 5), indicating that the variation in each indicator under the treatment factors originated primarily from within blocks, with very limited variation among blocks. Consequently, block as a random effect had limited influence on model inference but was retained in accordance with the experimental design.
Table 5.
Random-effect variance components of linear mixed models for NSC traits in different organs
| Indicator | σ²_block | σ²_residual |
|---|---|---|
| log(Leaf SS) | 2.12 × 10⁻⁴ | 0.0296 |
| log(Leaf ST) | 0 | 0.0172 |
| log(Leaf NSC) | 3.66 × 10⁻⁴ | 0.0179 |
| log(Leaf SS/Leaf ST) | ~ 0 | 0.0487 |
| log(Branch SS) | 9.89 × 10⁻⁴ | 0.0625 |
| log(Branch ST) | 2.29 × 10⁻³ | 0.0407 |
| log(Branch NSC) | 4.50 × 10⁻⁵ | 0.0246 |
| log(Branch SS/Branch ST) | 6.42 × 10⁻³ | 0.0981 |
| log(Root SS) | 5.77 × 10⁻⁴ | 0.0336 |
| log(Root ST) | 4.09 × 10⁻⁴ | 0.0273 |
| log(Root NSC) | 4.13 × 10⁻⁵ | 0.0193 |
| log(Root SS/Root ST) | 2.36 × 10⁻³ | 0.0563 |
σ²_block = variance among blocks; σ²_residual = residual variance. All NSC models used destructive sampling and included only block as a random intercept. ~0 indicates that the block variance was estimated near zero
The response patterns of root NSC components to fertilization varied with slope position and observation time (Fig. 4). At the upper slope, root SS content showed an increasing trend with rising fertilization rate at 120 days after fertilization, reaching its maximum in the F4 treatment; at 150 days, it exhibited an increasing-then-decreasing trend, with the F2 treatment being significantly higher than the Control; no significant changes were detected at 180 days. Root total NSC content followed a trajectory consistent with that of root SS, attaining maximum values in the F2 treatment at both 120 and 150 days, whereas no significant changes were observed at 180 days. Root ST content displayed no clear directional trend at any observation time point; at 120 and 150 days, peaks occurred in the moderate fertilization treatments and were significantly higher than the Control, but at 180 days, moderate fertilization treatments were instead significantly lower than the Control. The root SS/ST ratio exhibited an increasing trend at 120 days after fertilization, reaching its maximum in the F4 treatment; at 150 days, it showed an increasing-then-decreasing trend with a maximum in F2; at 180 days, differences among treatments were not significant.
Fig. 4.
Effect of fertilization on NSC of O. lagopus roots at different slopes. Figure (A) refers to the root soluble sugar content, Figure (B) refers to the root starch content, Figure (C) refers to the root NSC content and Figure (D) refers to the root soluble sugar/root starch. Data are marginal means, with error bars representing standard error (SE). Control refers to no water-soluble fertilizer treatment, 600 g/ water-soluble fertilizer applied to F1, 800 g/ water-soluble fertilizer applied to F2, 1000 g/ water-soluble fertilizer applied to F3 and 1200 g/ water-soluble fertilizer applied to F4. Sample size: n = 270 observations (90 individual trees × 3 sampling times). Statistical method: linear mixed model (LMM) with fixed effects of slope position × fertilization × time, and random effects of block. Post-hoc multiple comparisons were performed using Tukey’s HSD test. Different lowercase letters indicate significant differences between different fertilization treatments at the same sampling time (P < 0.05)
At the lower slope, root SS content displayed a continuously increasing trend at 120 days after fertilization, reaching its maximum in the F4 treatment; at 150 and 180 days, it showed a decreasing-then-increasing trend, with the minimum in the F3 treatment, which was significantly lower than the Control. Root ST content exhibited an increasing-then-decreasing trend at 120 and 150 days, reaching maximum values in the F2 and F3 treatments, respectively, and showed no significant changes at 180 days. Root total NSC content followed a gradually increasing trend at 120 days, with the F4 treatment significantly higher than the Control; it shifted to a continuously decreasing trend at 150 days and fluctuated in a decreasing-then-increasing pattern at 180 days. The root SS/ST ratio showed an increasing trend at 120 days, with the F4 treatment significantly higher than the Control, and then shifted to a trough-shaped response (decreasing then increasing) at both 150 and 180 days.
In terms of temporal dynamics, root SS and total NSC content at the upper slope generally increased over time, with the rate of increase markedly accelerating during the 150–180 day period. The temporal trajectories of root ST and SS/ST ratio among fertilization treatments showed pronounced inter-treatment divergence, with no consistent pattern. At the lower slope, the temporal changes in root SS and total NSC content were relatively moderate overall. Notably, the temporal dynamics of all root NSC components at both slope positions exhibited strong treatment specificity, lacking a unified temporal evolution pattern. Comparisons between slope positions revealed that the contents of all root NSC components were also significantly higher at the upper slope than at the lower slope, consistent with the results for leaves and branches. Therefore, regardless of organ type, the upper slope maintained a higher carbon reserve level, whereas the lower slope allocated more photosynthates to structural growth.
Allometric analysis of NSC in O. lagopus under fertilization at different slope positions
At 180 days after fertilization, allometric analysis based on standardized major axis (SMA) regression, with organ-level NSC content as the dependent variable and total NSC content as the independent variable, revealed pronounced organ-specific patterns in carbon allocation of balsa wood. These patterns were jointly driven by slope position and fertilization intensity (Table 6; Fig. 5). All regressions were conducted on individual-plant data.
Table 6.
Allometric analysis of NSC of O. lagopus in different slope positions under sampling detection for 180d after fertilization
| Slope position | Fertilization treatment | Organ | Slope (95% CI) | Intercept | R 2 | P |
|---|---|---|---|---|---|---|
| Upper slope position | Control | Leaf | 1.191(0.739, 1.92) | -2.285 | 0.693 | 0.005** |
| Branch | 1.056(0.816, 1.367) | -1.411 | 0.9149 | <0.001*** | ||
| Root | 1(0.997,1.003) | -1.099 | 0.99999 | <0.001*** | ||
| F1 | Leaf | 1.071(0.772, 1.485) | -1.526 | 0.861 | <0.001*** | |
| Branch | 1.197(0.75, 1.911) | -2.258 | 0.706 | 0.005** | ||
| Root | 1(0.993, 1.007) | -1.099 | 0.99994 | <0.001*** | ||
| F2 | Leaf | 1.815(0.92, 3.582) | -6.172 | 0.327 | 0.108 | |
| Branch | 1.049(0.805, 1.367) | -1.384 | 0.910 | <0.001*** | ||
| Root | 1(0.981,1.02) | -1.100 | 0.9995 | <0.001*** | ||
| F3 | Leaf | 1.105(0.769, 1.588) | -1.739 | 0.828 | <0.001*** | |
| Branch | 3.004(1.397, 6.458) | -13.596 | 0.112 | 0.38 | ||
| Root | 1(0.996, 1.004) | -1.099 | 0.99998 | <0.001*** | ||
| F4 | Leaf | 1.443(0.791, 2.633) | -3.866 | 0.490 | 0.036* | |
| Branch | 4.148(1.892, 9.095) | -20.986 | 0.056 | 0.54 | ||
| Root | 1(0.99, 1.01) | -1.099 | 0.99987 | <0.001*** | ||
| Lower slope position | Control | Leaf | 1.732(0.876, 3.428) | -5.457 | 0.321 | 0.112 |
| Branch | 1.032(0.823, 1.293) | -1.255 | 0.935 | <0.001*** | ||
| Root | 1(0.995,1.005) | -1.099 | 0.9997 | <0.001*** | ||
| F1 | Leaf | 1.096(0.77, 1.561) | -1.67 | 0.837 | <0.001*** | |
| Branch | 1.08(0.771, 1.513) | -1.497 | 0.852 | <0.001*** | ||
| Root | 1.002(0.94,1.069) | -1.111 | 0.995 | <0.001*** | ||
| F2 | Leaf | 1.107(0.774, 1.583) | -1.729 | 0.833 | <0.001*** | |
| Branch | 1.475(0.8, 2.722) | -3.388 | 0.469 | 0.042* | ||
| Root | 1.006(0.915, 1.107) | -1.131 | 0.989 | <0.001*** | ||
| F3 | Leaf | 1.387(0.773, 2.489) | -3.274 | 0.521 | 0.028* | |
| Branch | 1.836(0.908, 3.711) | -5.204 | 0.271 | 0.151 | ||
| Root | 1(0.988, 1.012) | -1.099 | 0.99983 | <0.001*** | ||
| F4 | Leaf | 1.101(0.76, 1.597) | -1.681 | 0.819 | <0.001*** | |
| Branch | 1.284(0.74, 2.227) | -2.531 | 0.560 | 0.017* | ||
| Root | 1(0.989, 1.011) | -1.099 | 0.99985 | <0.001*** |
Allometric regressions were based on individual plant data (n = 9 per treatment combination, total N = 90). The model was log(organ NSC) ~ log(total NSC), where total NSC = leaf + branch + root NSC. * indicates P<0.05,** indicates P<0.01, ***indicates P<0.001. R² values are presented with four decimal places. Treatments with P ≥ 0.05 are not marked for significance
Fig. 5.
Allometric scaling slopes of non-structural carbohydrate fractions in O. lagopus at different slope positions 180 days after fertilization
Root carbon allocation exhibited a strict isometric relationship (b ≈ 1.00) across all treatments, with exceptionally high model fit (R² > 0.999, P < 0.001). It should be noted that because root NSC is a component of total NSC, an inherent part–whole correlation exists, which inflates the R² values to some extent. Nevertheless, the consistently near-unity slopes and excellent fit across all treatments indicate that the root carbon pool maintains a highly conservative fixed proportion within the whole-plant carbon economy.
Branch carbon allocation displayed treatment-dependent allometric fluctuations. At the upper slope, the allometric slopes for branches under the control and low-fertilization treatments (F1, F2) were close to 1.0 (b = 1.05–1.20) with good model fit (R² = 0.71–0.92, P < 0.01), indicating near-isometric growth of branch NSC relative to total NSC. However, under the high-fertilization treatments (F3, F4), model fit was extremely low (R² = 0.11 and 0.06, P > 0.05), and no stable allometric relationship was detected. At the lower slope, branch allometric slopes generally ranged between 1.0 and 1.5, but model fit was low for most treatments (especially F2–F4; R² < 0.60, with some P > 0.05), indicating weak relationships and suggesting that branch carbon allocation may be governed by more stochastic factors.
Leaf NSC allometric relationships lacked regularity. Leaf slopes fluctuated between 1.07 and 1.81 across treatments, but model fit for several treatments (e.g., upper-slope F2, lower-slope control) was very low (R² < 0.33) and statistically non-significant (P > 0.05). Even for treatments with significant slopes, R² values mostly ranged from 0.49 to 0.86, substantially lower than those for roots. This indicates a weak quantitative relationship between leaf NSC and total NSC, consistent with the functional role of leaves as transient carbon turnover pools.
Correlation analysis
At 180 days after fertilization, correlations were observed between growth indicators and the SS and ST contents of various organs in O. lagopus (Table 7). After false discovery rate (FDR) correction for multiple comparisons (Benjamini–Hochberg procedure, Q < 0.05), the following significant correlations were identified. At the upper slope, tree height was highly significantly positively correlated with branch SS (Q < 0.01), DBH was highly significantly positively correlated with branch SS (Q < 0.01), and leaf ST was highly significantly positively correlated with branch ST (Q < 0.01). At the lower slope, tree height was significantly negatively correlated with both leaf SS and root SS (Q < 0.05); leaf ST was significantly negatively correlated with branch SS (Q < 0.05); and branch ST was significantly positively correlated with both root SS and root ST (Q < 0.05). All other correlations were not significant after FDR correction (see Supplementary Table-S2 for complete results).
Table 7.
Correlation analysis between growth traits and NSC components of O. lagopus under different slope positions at 180 days after fertilization (Pearson correlation coefficients, FDR-adjusted, Q < 0.05)
| Slope position | Variable A | Variable B | Correlation (r) | FDR-adjusted P (Q) |
|---|---|---|---|---|
| Upper slope position | Height of tree | Branch SS | 0.669 | < 0.001*** |
| Upper slope position | DBH | Branch SS | 0.516 | 0.0027** |
| Upper slope position | Leaf ST | Branch ST | 0.601 | <0.001*** |
| Lower slope position | Height of tree | Leaf SS | -0.488 | 0.0107* |
| Lower slope position | Height of tree | Root SS | -0.401 | 0.0351* |
| Lower slope position | Leaf ST | Branch SS | -0.407 | 0.0351* |
| Lower slope position | Branch ST | Root SS | 0.453 | 0.0167* |
| Lower slope position | Branch ST | Root ST | 0.484 | 0.0107* |
Only significant correlations with Q < 0.05 after FDR (Benjamini–Hochberg) correction are shown. Q is the adjusted P‑value
For the full correlation results (including non‑significant pairs), see Supplementary Table-S2
Abbreviations: DBH diameter at breast height, SS soluble sugars, ST starch
Principal Component Analysis (PCA)
At 180 days after fertilization, principal component analysis (PCA) revealed that the loading patterns of the indicators differed between the upper and lower slope positions (Fig. 6). For both slope positions, the first two principal components cumulatively explained 52.7% of the total variance. At the upper slope, PC1 was primarily loaded on tree height and branch SS content, while PC2 was dominated by branch ST content. At the lower slope, the main contributing variables to PC1 were root SS and root ST contents, whereas PC2 was primarily loaded on leaf ST content. These results indicate that variables such as tree height, branch SS and ST, and root SS and ST had high loadings on the principal components, suggesting their potentially important statistical contributions under different slope positions.
Discussion
Slope position modulates fertilization effects: divergent growth patterns with continuous increase on the upper slope and a unimodal response on the lower slope
Linear mixed model analysis in this study revealed that slope position, fertilization, and time all had highly significant independent main effects on tree height and DBH of O. lagopus (P < 0.01; Table 3). The two-way interaction between slope position and fertilization was highly significant for tree height but had no significant effect on DBH growth, indicating that the regulation of height growth by fertilization exhibited slope-position dependence, whereas the regulation of radial growth was statistically consistent across slope positions. None of the interaction terms involving time reached significance, suggesting that the rank order of growth rates among treatment combinations remained relatively stable over the 180 days observation period.
Tree height and DBH are key indicators for assessing tree growth status [38]. Previous studies have indicated that appropriate application of secondary macronutrients water-soluble fertilizer can effectively promote plant growth and increase yield [39]. In the present study, the growth response of O. lagopus at the upper slope showed a continuously increasing trend in tree height with increasing fertilization rate, reaching its optimum under the highest fertilization treatment. This is similar to the findings of Kulmann et al. [40] on fertilization promoting the growth of Pinus taeda plantations. This phenomenon may be related to nutrients becoming the key limiting factor for growth under conditions of relative water scarcity [41], and is consistent to some extent with the “co-limitation of resources” theory: when one critical resource (e.g., water) is limited, increasing the availability of another resource (e.g., nutrients) can still significantly enhance the resource acquisition and utilization efficiency of plants, thereby promoting growth [42]. In contrast, DBH growth reached its maximum in the F3 treatment and then leveled off, statistically suggesting that radial growth may be more sensitive to excessive fertilization. At the lower slope, both tree height and DBH exhibited a unimodal curve with increasing fertilization rate, increasing first and then declining, and both reached their maximum values in the F2 treatment. Although the F3 and F4 treatments were still significantly higher than the control, they did not exceed the F2 level, indicating that further increases in fertilization rate did not bring additional growth gains. This trend is similar to the findings of Ou and Kang [43] regarding the modulation of tree growth by topography. At the lower slope, the topographic position may be more conducive to water interception and nutrient accumulation, and the inherent nutrient supply capacity of the soil may already meet most of the growth requirements. Additional moderate fertilization likely played a role in optimizing nutrient balance and stimulating growth potential, whereas excessive fertilization was speculated to potentially reduce rhizosphere water potential [44] or cause ionic imbalances [45], thereby attenuating further growth promotion. From the perspective of temporal dynamics, growth indicators of all treatments at the upper slope generally showed an upward trend, whereas temporal changes at the lower slope were relatively moderate. This is speculated to be related to the sustained dependence on nutrient inputs at the upper slope and the earlier stabilization of growth at the lower slope. The non-significance of time interaction effects indicates that the rank order of growth among treatments did not reverse significantly over the observation period.
Topography-driven carbon allocation strategy: adaptive NSC dynamics in O. lagopus in response to slope-specific fertilization
NSC serve as the central carriers of energy storage in plants, and their contents and allocation patterns among organs directly reflect the balance between carbon supply and demand, making them crucial indicators for elucidating carbon allocation strategies and environmental adaptation mechanisms [46, 47]. Integrating the aforementioned growth response patterns, this study further attempted to decipher the potential mechanisms by which slope position and fertilization affect the growth of O. lagopus from the perspective of carbon allocation. Linear mixed model analysis revealed pronounced hierarchical differences in the responses of NSC in different organs to treatment factors: all components of leaf and branch NSC exhibited highly significant overall responses to all main effects and interactions, whereas root NSC displayed a more complex stratified pattern.
At the upper slope, carbon allocation in O. lagopus exhibited distinct temporal adjustments. During the early-to-middle period after fertilization, the SS and total NSC contents in leaves and roots displayed a unimodal response to increasing fertilization rate, peaking at moderate fertilization levels. This may indicate enhanced photosynthetic carbon assimilation and carbon reserve accumulation following improved nutrient supply [47], consistent with the report by Liu et al. [48] on enhanced carbon accumulation in Betula platyphylla Sukaczev seedlings after fertilization. In contrast, branch SS and total NSC contents under low fertilization were significantly lower than those of the control during the same period, and branch ST was also generally low. It is speculated that limited photosynthates were preferentially allocated to leaves and roots to sustain assimilation and absorption functions, while branches may not have been the priority carbon sink at this stage. By 180 days, the allocation pattern had shifted: differences among treatments converged for leaves, with no regular pattern observed, whereas branch SS and total NSC contents increased continuously with increasing fertilization rate, suggesting potential compensatory accumulation during the later stage. Roots exhibited transitional characteristics during this process, as differences in root SS and total NSC among treatments dissipated at 180 days, and root ST was significantly lower under moderate fertilization than in the control, implying that root ST may have been mobilized during the later stage to support branch carbon demand. From the temporal trajectory perspective, SS and total NSC contents in all organs at the upper slope increased over time, while ST generally showed a decreasing-then-increasing trend. These inter-organ temporal adjustments in allocation—more to leaves and roots during the early-to-middle period, and increased accumulation in branches during the later stage—corresponded to the growth pattern of continuously increasing tree height and stabilizing DBH at the upper slope, suggesting that upper-slope plants may adopt an adaptive strategy of maintaining growth through temporal optimization of carbon allocation [49, 50].
The response of NSC in balsa wood at the lower slope to fertilization exhibited a nonlinear trend characterized by initial promotion followed by inhibition. During the early-to-middle period after fertilization, both leaf and branch SS and total NSC contents showed a unimodal response, peaking at moderate fertilization treatments; root SS and ST also displayed increasing or unimodal trends during the same period. It is speculated that under the relatively nutrient-sufficient conditions of the lower slope, appropriate exogenous nutrient supply can further stimulate sugar synthesis and starch reserves in various organs, similar to the findings of Wang et al. [51] on fertilization promoting carbon accumulation in Triadica sebifera. However, by 180 days, the inhibitory effect of excessive fertilization became statistically evident: leaf and root SS and total NSC contents in the F3 treatment dropped to their minimum values and were significantly lower than the control, exhibiting a directional reversal from promotion to inhibition. As a hypothesis, this shift may be related to the relatively abundant nutrient supply at the lower slope. Moderate fertilization stimulated metabolic activity in the early stage, but the cumulative effect of excessive fertilization over time may have triggered rhizosphere ionic imbalance or osmotic stress [44], thereby potentially impairing carbon assimilation efficiency [52]. It should be noted that this study did not measure soil electrical conductivity, osmotic potential, root damage, nutrient antagonism, or tissue ion concentrations; therefore, the above speculation regarding ionic imbalance or osmotic stress remains a hypothesis, and its validity awaits future experimental verification. Notably, branch NSC did not differ significantly among treatments at 180 days and did not show the marked decline observed in leaves and roots, suggesting that as structural organs, branches possess relatively stable carbon pools when roots face stress [53]. From the temporal trajectory perspective, the temporal changes in NSC in all organs at the lower slope were generally more moderate than those at the upper slope, lacking a unified temporal evolution pattern, which is speculated to reflect the complexity of carbon dynamics being synergistically regulated by multiple factors at the lower slope. The aforementioned nonlinear trend of NSC—promotion in the early-to-middle stage and suppression in leaves and roots in the later stage—are consistent with the growth pattern in which both tree height and DBH peaked under moderate fertilization at the lower slope and the growth-promoting effect weakened under excessive fertilization, indicating that excessive fertilization at the lower slope may result in diminishing marginal returns.
This study found that although tree height and DBH at the lower slope were significantly greater than those at the upper slope (Fig. 1), the NSC contents in leaves, branches, and roots were generally lower at the lower slope than at the upper slope (Figs. 2, 3 and 4), whereas the SS/ST ratio in branches at 150 days was higher at the lower slope. This contrasting pattern may be related to a “growth–storage trade-off” strategy: trees at the upper slope tend to accumulate more NSC as “carbon reserves” to cope with adverse environments, whereas trees at the lower slope allocate more photosynthates to structural growth rather than storage, thereby exhibiting higher growth rates but lower NSC storage levels. The higher branch SS/ST at 150 days at the lower slope may reflect a greater relative proportion of soluble sugars in the branches there. This explanation is consistent with resource allocation theory: when resources are limited, plants prioritize reserves; when resources are abundant, they prioritize investment in growth.
In summary, the responses of growth and carbon allocation to fertilization in balsa wood exhibited significant slope position differentiation. The upper slope was characterized by a sustained response trend with increasing fertilization rate and inter-organ temporal allocation adjustments, whereas the lower slope exhibited a nonlinear trend of initial promotion followed by inhibition and diminishing marginal returns. The contrasting pattern of NSC content and growth between the upper and lower slopes suggests that carbon allocation patterns may differ with slope position, although the underlying physiological mechanisms require further investigation.
Trait integration and mechanistic insights based on allometric growth, correlation, and principal component analyses
Allometric analysis revealed distinct organ-specific patterns in carbon allocation across organs of O. lagopus (Table 6; Fig. 5). Roots maintained a strict isometric relationship across all treatments, with exceptionally high model fit. Although part–whole correlations may inflate the R² values to some extent, the root allometric slope consistently remained close to 1 with excellent fit, strongly indicating that the root carbon pool retains a highly conservative fixed proportion within the whole-plant carbon economy, and this conservatism was unaffected by slope position or fertilization intensity. For branch carbon allocation, stable allometric relationships were detected only in the control and low-fertilization treatments at the upper slope, where branch NSC exhibited near-isometric growth with total NSC, reflecting coordination between the branch carbon pool and whole-plant carbon accumulation under these conditions. In the high-fertilization treatments at the upper slope and most treatments at the lower slope, model fit was weak or no significant relationship was detected; therefore, no explicit biological interpretation of branch allometric patterns was made for these treatments. Overall, no consistent and stable allometric pattern was found between leaf NSC and total NSC, as most treatments showed low model fit or statistical non-significance. This result indicates that leaf carbon allocation is weakly influenced by changes in total whole-plant NSC and may be regulated more by transient source–sink dynamics.
Correlation analysis revealed statistical association patterns between growth indicators and organ carbon fractions in O. lagopus (Table 7). At the upper slope, both tree height and DBH were significantly positively correlated with branch SS, indicating coordinated variation between aboveground growth and the branch soluble sugar pool [54]. Leaf ST was positively correlated with branch ST, suggesting that leaf and branch starch reserves may accumulate synchronously. At the lower slope, tree height was negatively correlated with leaf SS and root SS, whereas branch ST was positively correlated with root SS and root ST, and leaf ST was negatively correlated with branch SS. These results imply potential trade-off relationships between growth and certain organ carbon fractions at the lower slope; for example, the accumulation of SS in leaves and roots was inversely associated with height growth, while ST reserves in branches and roots exhibited positive coordination.
PCA further distilled the statistical patterns of O. lagopus responses to fertilization at different slope positions (Fig. 6). At the upper slope, tree height, branch SS, and branch ST had high loadings, indicating that these variables contributed substantially to the overall variation at the upper slope. At the lower slope, root SS, root ST, and leaf ST had high loadings. This difference in loading patterns suggests that slope position may influence the association patterns among variables: variation at the upper slope was primarily driven by growth traits and the branch carbon pool, whereas variation at the lower slope was mainly attributable to root and leaf carbon reserves.
Fig. 6.
Principal component analysis of each index of O. lagopus in different slope positions under 180 days of sampling detection for 180 days after fertilization. Figure (A) refers to the upper slope position and Figure (B) refers to the lower slope position. PCA was performed on a correlation matrix (variables standardized). Sample sizes: n = 45 for the upper slope and n = 45 for the lower slope. Complete data are provided in Supplementary Table-S3a and Table-S3b
In summary, carbon allocation and growth in O. lagopus displayed clearly differentiated patterns between slope positions. Variation in upper-slope plants mainly originated from the coordinated adjustment of aboveground growth and the branch carbon pool, suggesting a preferential allocation of carbon resources to height growth and mobilizable carbon reserves. In contrast, variation in lower-slope plants was more strongly contributed by root and leaf carbon reserves, reflecting a carbon allocation strategy that may center more on the trade-off between belowground carbon pool accumulation and leaf metabolic regulation. It should be noted that these conclusions are based on statistical association patterns (part–whole coupling may have inflated allometric R² values, correlations were adjusted by FDR correction, and PCA cumulatively explained only 52.7% of the variance); the underlying physiological adaptation mechanisms require independent experimental verification.
Several limitations should be acknowledged. First, baseline soil data were limited to organic carbon and total nitrogen, phosphorus, and potassium; we did not measure exchangeable Ca, Mg, S, available nutrients, cation exchange capacity, pH, or soil moisture in the two slope positions. The available soil data showed no significant differences between slope positions, so the specific drivers of slope position effects remain incompletely understood and may involve multiple interacting factors such as water redistribution and microclimatic variation. Nevertheless, this study has clearly demonstrated significant main and interactive effects of fertilization, slope position, and time on growth and non‑structural carbohydrate allocation, as well as organ‑specific response patterns. Future research should systematically monitor the above‑mentioned soil and environmental variables to further elucidate the key pathways through which slope position influences O. lagopus growth and carbon allocation.
Conclusion
Slope position and fertilization jointly shaped the carbon allocation strategies among organs of O. lagopus. Our results demonstrated that moderate fertilization effectively promoted photosynthetic carbon assimilation and significantly increased tree height; although DBH growth also benefited from fertilization, further increases in fertilization rate did not yield additional gains. Under different slope positions, O. lagopus exhibited distinct carbon utilization patterns: at the upper slope, carbon allocation displayed a temporal compensatory pattern, with carbon supply preferentially directed to leaves and roots during the early-to-middle stages and subsequently accelerated translocation to branches and stems for storage at the later stage; at the lower slope, carbon allocation showed a unimodal response, with low to moderate fertilization promoting carbon accumulation in various organs, whereas excessive fertilization led to declines in leaf and root carbon pools. Allometric growth and correlation analyses further revealed that at the upper slope, growth was positively correlated with the branch soluble sugar pool and allometric relationships were relatively integrated; at the lower slope, associations between growth and some organ carbon fractions were weaker, and allometric constraints among organs were relatively relaxed. Principal component analysis identified slope-position-specific key response trait combinations, with tree height and the branch carbon pool as the primary contributing variables at the upper slope, and root and leaf carbon reserves as the core at the lower slope. These findings provide a theoretical basis for differentiated fertilization management in O. lagopus plantations based on site conditions: a moderately higher fertilization intensity can be applied at the upper slope to promote the synergy between carbon accumulation and growth, whereas fertilization dosage should be controlled at the lower slope to avoid diminishing carbon gains caused by excessive inputs. It should be noted that this study did not directly measure soil secondary macronutrients contents or plant physiological indicators; therefore, the specific mechanisms underlying the slope position effects require further investigation and validation.
Supplementary Information
Acknowledgements
We are very grateful to every scientific researcher who supports and helps us in our work.
Authors’ contributions
J.W.: Conceptualization, Data curation, Funding acquisition, Project administration, Writing – original draft, Writing – review & editing. J.L.: Writing – review & editing, Writing – original draft, Investigation, Data curation, Formal analysis, Conceptualization. W.S.: Investigation, Data curation, Conceptualization. J.C.: Investigation, Software, Methodology. W.H. and G.C.: Investigation. All authors have read and agreed to the published version of the manuscript.
Funding
This research was funded by the National Natural Science Foundation of China, grant number 32071688.
Data availability
All data generated or analyzed during this study are included in this paper, and materials used in the experiments are available upon request.
Declarations
Ethics approval and consent to participate
This study did not involve endangered or protected species. The field sampling of Ochroma lagopus and the associated fertilization experiments were carried out on plantation forests within Mengxing Farm, Mengla County. All activities were performed with the formal permission of the local land management authority. The research protocol adhered to the standard operational guidelines of Southwest Forestry University for field-based agricultural and forestry trials.
Consent for publication
Not applicable.
Competing interests
The authors declare no competing interests.
Footnotes
Publisher’s note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
References
- 1.Guang M, Liao LL. Application Effect of Amino Acid Water-soluble Fertilizer on Eggplant. J. Modern Agricultural Science and Technology. 2023;(18):67–70.
- 2.Cai SH. A preliminary study on the effects of spraying medium–element water–soluble fertilizer on rice yield and economic benefits. Hubei Plant Protection. 2023;(1):37–8.
- 3.Xu X, Zhao C, Qian K, Sun M, Hao Y, Han L, Wang C, Ma C, White JC, Xing B. Physiological responses of pumpkin to zinc oxide quantum dots and nanoparticles. J Environ Pollut. 2022;296:118723. [DOI] [PubMed] [Google Scholar]
- 4.Wei QQ, Zhang M, Gu XF, Liu LL, Ying BK, Pan WY, Yao J, Gou JL. Effects of spraying medium–element water–soluble fertilizer on sorghum yield, nutrient uptake and quality. Agricultural Technol Service. 2023;40(07):23–6. [Google Scholar]
- 5.Qi JF. Effect of Medium Element Water Soluble Fertilizer on the Growth and Yield of Solanum lycopersicum. Agricultural Technology & Equipment. 2025;(03):172–4.
- 6.BrubaKer SC, Jones AJ, Lewis DT, Frank K. Soil properties associated with landscape position. J Soil Sci Soc Am J. 1993;57:235–9. [Google Scholar]
- 7.Hawthorne S, Miniat CF. Topography may mitigate drought effects on vegetation along a hillslope gradient. J Ecohydrology. 2017;10(1):e1825. [Google Scholar]
- 8.Pang HD, Li L, Cui HX, Zeng XF, Tang WP. Effects of Different Site Factors on the Growth and Soil Physical and Chemical Properties of Paulownia Plantation. J J Northeast Forestry Univ. 2021;49(07):28–32. [Google Scholar]
- 9.Xie M, Zhang P, He Y, Hong D. Combined effects of rain-fall pattern and fertilization on subsurface nitrogen transport and loss from sloped soil: a laboratory experiment. J Arab J Geosci. 2023;16(1):31. [Google Scholar]
- 10.Sun T, Mao ZJ, Dong LL, Hou LH, Song X, Wang YT. Further evidence for slow decomposition of very fine roots using two methods: Litter bags and intact cores. Plant Soil. 2012;366:633–46. [Google Scholar]
- 11.Dielze MC, Sala AN, Carbone MS, Czimczik CI, Mantooth JA, Richardson AD, Vargas R. Nonstructural carbon in woody plants. Annu Rev Plant Biol. 2014;65:667–87. [DOI] [PubMed] [Google Scholar]
- 12.Graham D, Patterson BD. Responses of plants to low, nonfreezing temperatures: proteins, metabolism, and acclimation. Annu Rev Plant Physiol. 1982;33:347–72. [Google Scholar]
- 13.Savage JA, Clearwater MJ, Haines DF, Klein T, Mencuccini M, Sevanto S, Turgeon R, Zhang CK. Allocation, stress tolerance and carbon transport in plants: how does phloem physiology affect plant ecology. Plant Cell Environ. 2016;39(4):709–25. [DOI] [PubMed] [Google Scholar]
- 14.Thalmann M, Santelia D. Starch as a determinant of plant fitness under abiotic stress. New Phytol. 2017;214(3):943–51. [DOI] [PubMed] [Google Scholar]
- 15.White PJ, Broadley MR. Calcium in plants. Ann Bot. 2003;92(4):487–511. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Cakmak I, Kirkby EA. Role of magnesium in carbon partitioning and alleviating photooxidative damage. Physiol. Plant. 2010;133(4):692–704. [DOI] [PubMed] [Google Scholar]
- 17.Kopriva S. Regulation of sulfate assimilation in Arabidopsis and beyond. Ann Bot. 2006;97(4):479–95. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Shi PL, Körner C, Hoch G. End of season carbon supply status of woody species near the treeline in western China. Basic Appl Ecol. 2006;7(4):370–7. [Google Scholar]
- 19.Wang WN, Li JN, Wang HR, Gu JC. Seasonal dynamics of leaf non-structural carbohydrates in different tree species. J Northeast Univ. 2014;42(4):46–9. [Google Scholar]
- 20.Borrega M, Gibson LJ. Mechanics of balsa (Ochroma pyramidale) wood. Mech Mater. 2015;84:75–90. [Google Scholar]
- 21.Borrega M, Ahvenainen P, Serimaa R, Gibson L. Composition and structure of balsa (Ochroma pyramidale) wood. Wood Sci Technol. 2015;49(2):403–20. [Google Scholar]
- 22.Kotlarewski N, Derikvand M, Lee M, Nolan G, Jiao H. Bifenthrin treatment for balsa: susceptibility of Papua New Guinea-grown Ochroma pyramidale to attack by Coptotermes acinaciformis in an Australian context. Int Biodeterior Biodegrad. 2019;137:153–7. [Google Scholar]
- 23.Yang ZJ. Lecture series on aviation model and flight science aircraft and material chemistry. Scientific Inquiry: Rural Youth. 2009;(7):43–5.
- 24.Samuel ILT, Alejandro MR. Differences in growth and survival of two varieties of Ochroma pyramidale in rustic plantations in southern Mexico. Trees People. 2024;17:100652. [Google Scholar]
- 25.Vahedi N, Tiago C, Vassilopoulos AP. Thermophysical properties of Ochroma lagopus used as core of sandwich composite bridge decks exposed to external fire. J Constr Build Mater. 2022;329:127–64. [Google Scholar]
- 26.Qian Y, Bai X, Huang P, Wang L, Zhu W, Su Y, Tang M. Transcriptome analysis reveals the mechanism of early branching of balsa (Ochroma lagopus Swartz). Physiol. Plant. 2025;177(5):e70509. [DOI] [PubMed] [Google Scholar]
- 27.Gu JY, Liu CY, Duan HJ, Du GB, Wu JW. Rainy season dynamics of carbon, nitrogen and phosphorus and their stoichiometric characteristics in Ochroma lagopus. BMC Plant Biol. 2025;25(1):1–9. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28.Liu YX, Sun JL, Dai C, Du GB, Shi R, Wu JW. Physiological and biochemical adaptations to repeated drought–rehydration cycles in Ochroma lagopus Swartz: implications for growth and stress resilience. Plants. 2025;14(7):1636. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Wen CJ, Duan HJ, Wu JW, Zhao ZJ, Liu YX, Chen G, Duan GH, Li ZQ, Du GB. Response of Ochroma lagopus Swartz growth and nutrient uptake to urea and slow-release fertilizer addition. Forests. 2025;16(2):278. [Google Scholar]
- 30.Chen G, Zhao ZJ, Li ZQ, Duan GB, Wen CJ, Du GH, Wu JW. Response of non-structural carbohydrates and carbon, nitrogen and phosphorus stoichiometric characteristics of Ochroma lagopus leaves to nitrogen addition. Forests. 2025;16(1):117. [Google Scholar]
- 31.Wen CJ, Wu JW, Zhao ZJ, Chen G, Li ZQ, Duan GH, Du GB, Liang W. Response of growth and organ carbon, nitrogen and phosphorus stoichiometry to nitrogen addition in a Ochroma lagopus plantation forest. J Plant Nutr Fertilizers. 2025;31(01):188–200. [Google Scholar]
- 32.Duan GH, Wu JW, Zhao ZJ, Li ZQ, Chen G, Wen CJ, Du GB. Effect of fertilizer application on growth and non-structural carbohydrates of Ochroma pyramidale. Journal of Northwest Forestry University. 2025;1–13.
- 33.IUSS Working Group WRB. World Reference Base for Soil Resources 2014 update 2015 International soil classification system for naming soils and creating legends for soil maps. World Soil Resources Reports No 106 FAO Rome; 2015.192 p. ISBN 978-92-5-108369-7.
- 34.Zhang K, Zheng H, Ouyang ZY, Li RD, Yang M, Lan J, Xiang XW. Effects of nitrogen fertilization on greenhouse gas fluxes of soil-atmosphere interface in growing and non-growing season in Eucalyptus plantations in southern China. J Chin J Ecol. 2015;34(7):1779–84. [Google Scholar]
- 35.Chen JL, Zhu WS, Liu YX, Chen G, Han JC, Zhang WH, Wu JW. Effects of Fertilizer Application on Growth and Stoichiometric Characteristics of Nitrogen, Phosphorus, and Potassium in Balsa Tree (Ochroma lagopus) Plantations at Different Slope Positions. Plants. 2025;14(14):2221. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Dubois M, Gilles KA, Hamilton JK, Rebers PA, Smith F. Colorimetric method for determination of sugars and related substances. Anal Chem. 1956;28(3):350–6. [Google Scholar]
- 37.Warton DI, Duursma RA, Falster DS, Taskinen S. Smatr 3 – an R package for estimation and inference about allometric lines. Methods Ecol Evol. 2012;3(2):257–9. [Google Scholar]
- 38.Li X, Bo H, Zhu J, Zhang J, Hu J, Mu F, Nie L. Reducing nutrient loss caused by thinning: effects of four composts of forest thinning shreds on soil nutrients and tree growth in semimature Pinus tabuliformis, Beijing, China. Forests. 2022;13(5):702. [Google Scholar]
- 39.Fan XM, Li JH. Field experiment on fertilizer effect of medium element water-soluble fertilizer on tomato. Northwest Horticulture. 2021;(1):48–51.
- 40.Kulmann MSS, Deliberali I, Schumacher MV, Stahl J, Figura MA, Ludvichak AA, Stape JL. Can fertilization and stand uniformity affect the growth and biomass production in a Pinus taeda plantation in southern Brazil. Ecol Manag. 2023;541:121075. [Google Scholar]
- 41.Gessler A, Schaub M, McDowell NG. The role of nutrients in drought-induced tree mortality and recovery. New Phytol. 2016;214(2):513–20. [DOI] [PubMed] [Google Scholar]
- 42.Chapin FS III, Bloom AJ, Field CB, Waring RH. Plant responses to multiple environmental factors. Bioscience. 1987;37(1):49–57. [Google Scholar]
- 43.Ou JD, Kang YW. The annual growth dynamics and growth rhythm of Manglietia yuyuanensis saplings at different slope positions. J Northeast Univ. 2024;52(09):8–12. [Google Scholar]
- 44.Munns R, Tester M. Mechanisms of salinity tolerance. Annu Rev Plant Biol. 2008;59:651–81. [DOI] [PubMed] [Google Scholar]
- 45.Kopittke PM, Blamey FPC. Theoretical and experimental assessment of nutrient solution composition in short-term studies of aluminium rhizotoxicity. Plant Soil. 2016;406:311–26. [Google Scholar]
- 46.Pan QM, Han XG, Bai YF, Yang JC. Advances in physio-ecological studies on plant non-structural carbon reserves. Chin Bull Bot. 2002;19(1):30–8. [Google Scholar]
- 47.Martínez-Vilalta J, Sala AN, Asensio D, Galiano L, Hoch G, Palacio S, Piper FI, Lloret F. Dynamics of non-structural carbohydrates in terrestrial plants: a global synthesis. Ecol Monogr. 2016;86(4):495–516. [Google Scholar]
- 48.Liu TY, Hao LF, Wang QC, Zhang DG, Jia HX, Cui XY. Effects of different light substrates and fertilization treatments on nutrient carrying capacity of Betula platyphylla seedlings. J Northeast Univ. 2019;47(10):16–9. [Google Scholar]
- 49.Sala A, Woodruff DR, Meinzer FC. Carbon dynamics in trees: feast or famine? Tree Physiol. 2012;32(6):764–75. [DOI] [PubMed] [Google Scholar]
- 50.Richardson AD, Carbone MS, Keenan TF, Czimczik CI, Hollinger DY, Murakami P, Schaberg PG, Xu X. Seasonal dynamics and age of stemwood nonstructural carbohydrates in temperate forest trees. New Phytol. 2013;197(3):850–61. [DOI] [PubMed] [Google Scholar]
- 51.Wang TT, Zhou Q, Zhuge F, Xu ZL, Zhang XM, Hu DH. The effects of rootstocks and slow-release fertilizer on the growth and ornamental traits of Triadica sebifera ‘Hongzi Jiaren’. Acta Hortic Sin. 2025;52(04):1037–50. [Google Scholar]
- 52.Manter DK, Kavanagh KL, Rose CL. Growth response of Douglas-fir seedlings to nitrogen fertilization: importance of Rubisco activation state and respiration rates. J Tree Physiol. 2005;25(8):1015–21. [DOI] [PubMed] [Google Scholar]
- 53.Wang LY, Liao D, Rengel Z, Zhang FS, Shen JB. Soil-plant-microbe interactions in the rhizosphere: incremental amplification induced by localized fertilization. Front Agric Sci Eng. 2025;12(1):57–68. [Google Scholar]
- 54.Martínez-Vilalta J, Sala A, Asensio D, et al. Dynamics of non-structural carbohydrates in terrestrial plants: a global synthesis. Trends Plant Sci. 2016;21(10):811–23. [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
All data generated or analyzed during this study are included in this paper, and materials used in the experiments are available upon request.