We conducted a literature search in the Web of Science resource for journal articles published between January 2000 and December 2023, using the following keywords: “nonstructural carbohydrate*” or “NSC” or “TNC” or “soluble sugar” and “fertiliz*” or “nutrition*” or “nitrogen*” or “ phosphorus”. This search identified a total of 13,866 articles. The selection criteria for the included studies were as follows: (1) The studies must be based on manipulation experiments with control and experimental groups, with research subjects being either woody or herbaceous plants. (2) Only field-based nutrient addition studies were included, excluding laboratory incubation studies. (3) The fertilizers must contain only nitrogen (N), phosphorus (P), or both nitrogen and phosphorus, with no other nutrients. (4) The studies must provide both means and sample sizes for the reported data. Ultimately, 30 studies were selected for inclusion in our dataset (see Supplementary File for details), with their geographic distribution shown in Figure 1 . The dataset comprises 62 plant species and 1,313 independent organ-level observations of NSCs. The content of different NSC components across various plant organs are presented in Supplementary Figures S1 and S2 .

From these articles, we extracted the NSC, soluble sugar, and starch contents in leaves, branches, stems, and roots, and standardized the units to mg/g. Additionally, we recorded research species, type of fertilizer, nitrogen and phosphorus addition rates, experimental duration, geographical location, leaf habit, and other relevant background information. Data were either directly retrieved from tables provided in the articles or extracted from images using WebPlotDigitizer software (He et al., 2020).

For each study, we prioritized the original classification of fertilization intensity if provided in the publication. If a study did not supply a classification, we standardized the fertilization amounts to grams per hectare per year (g/ha yr). Based on the overall distribution of application rates across our dataset, we defined concentrations as low (< 50 g/ha yr), medium (50–100 g/ha yr), and high (> 100 g/ha yr). While formal methods like breakpoint analysis can identify precise thresholds, they require a consistent continuous predictor, which was unavailable due to our hierarchical approach of honoring original study classifications. Therefore, our categorical analysis represents a robust strategy for identifying general patterns from diverse experimental designs, aligning with the primary goal of this synthesis.

Furthermore, we classified fertilization durations as short-term (≤ 6 months), medium-term (> 6 months to ≤ 3 years), and long-term (> 3 years). If a study reported data at multiple sampling dates, we utilized the data corresponding to the actual sampling times. Finally, due to limited data for finer classifications, we analyzed plant life forms based on two broad categories: herbaceous and woody plants, which are known to differ significantly in their carbon supply and storage strategies.

The natural logarithm of the response ratio represents the effect size, assessing the relative change between the nutrient addition treatment and the control (Li et al., 2020). The formula for calculating the log response ratio is as follows (Equation 1):

where X ¯ t is the mean NSC value in the treatment and X ¯ c is that in the control (He et al., 2020; Liu et al., 2016).

The formula for calculating the variance of each effect value is as follows (Equation 2):

where n t and n c are the sample sizes for the treatment and control groups, respectively, and S t and S c denote their corresponding standard deviations. X ¯ t and X ¯ c are the mean values of the variable in the treatment and control groups, respectively.

The weighting factor ( w ) for each observation is calculated by the inverse of the variance ( v ) (Equation 3):

If a study includes multiple results for a single variable, we adjusted the weights based on the total number of observations in each study to reduce the influence of repeated measurements from the same site. The final weighting factor ( w ′ ) and effect size ( ln R R ¯ ) were calculated using the following formulas (Equations 4, 5):

The distributions of all effect size results are shown in Supplementary Figures S3 and S4 . If the 95% confidence interval (CI) for a variable overlap with zero, it indicates that nutrient addition has no significant effect on the variable. Otherwise, the effect is considered statistically significant. The percentage change relative to the control was calculated as (Equation 6):

We conducted a meta-analysis using the “metafor” package in R 4.4.1 with the REML method. Data following a normal distribution were analyzed using one-way ANOVA followed by Tukey’s HSD method, while data not following a normal distribution were analyzed using the Kruskal-Wallis test followed by Dunn’s test.

The effects of nitrogen addition on the NSC and soluble sugar contents in woody plant leaves were minor, with a significant increase in NSC pool observed only under low nitrogen addition. However, nitrogen addition significantly decreased starch content in leaves across all concentrations. Notably, the magnitude of this negative effect was greatest at low nitrogen levels and diminished as the concentration increased. In branches, nitrogen addition had a significant positive effect on NSC content but a significant negative effect on starch content. For woody plant stems and roots, nitrogen addition generally had a positive effect on NSC and its components, but the effects were not statistically significant ( Figure 2 ).

In contrast, herbaceous plants showed a significantly stronger response to nitrogen addition, predominantly resulting in decreased NSC content. Specifically, in leaves, NSC, soluble sugar, and starch contents decreased progressively with increasing nitrogen levels. In stems, nitrogen enrichment at all concentrations significantly reduced NSC content. In roots, while NSC content generally decreased, the changes exhibited high variability and uncertainty ( Figure 2 ).

Phosphorus enrichment had highly variable effects on the NSC content of different organs in woody plants. For leaves, phosphorus enrichment generally exerted a negative effect on NSC and its components, with the exception of medium-level phosphorus enrichment, which significantly increased starch content. In branches, phosphorus enrichment generally increased NSC content, but none of these effects were statistically significant. In stems, the effects of phosphorus enrichment on NSC content fluctuated significantly, showing a trend of promotion at low levels, suppression at medium levels, and promotion again at high levels. In roots, NSC and starch contents increased with increasing phosphorus concentrations, while soluble sugar content decreased as phosphorus concentration increased ( Figure 3 ).

Due to the scarcity of experimental data on the response of herbaceous plant NSC content to phosphorus enrichment, our analysis only allowed for discussion of NSC responses in certain herbaceous plant organs. In leaves, phosphorus enrichment at all concentrations reduced NSC content. For stems, low-level phosphorus enrichment reduced NSC content, but as phosphorus concentration increased to medium or high levels, the effect shifted from suppression to promotion ( Figure 3 ).

Nitrogen enrichment in woody plants generally showed a trend of initial promotion followed by suppression for NSC, soluble sugars, and starch contents in leaves, branches, and stems. Short-term nitrogen treatments significantly increased NSC content, but as the treatment duration extended to one year or more, this promotive effect diminished and even turned inhibitory. In roots, the trend differed slightly, with NSC content showing a similar initial increase and subsequent decrease, whereas soluble sugars and starch exhibited a pattern of initial decline, followed by an increase, and then another decrease ( Figures 4A-D ).

In herbaceous plants, nitrogen enrichment had little to no promotive effect on NSC accumulation. In leaves, NSC content gradually decreased with prolonged treatment. In stems, NSC content dropped significantly at first and then slowly recovered over time ( Figures 4E-G ).

Phosphorus enrichment showed an initial promotive effect followed by suppression on NSC and starch contents in the leaves and roots of woody plants, while soluble sugar content exhibited a reverse trend of initial suppression followed by promotion. However, no significant trends were observed in the NSC content of branches and stems ( Figures 5A-D ). In herbaceous plants, prolonged phosphorus enrichment significantly reduced NSC content in leaves, while NSC content in stems showed a trend of initial decrease followed by an increase ( Figures 5E, F ).

As is well known, nitrogen is a critical element in the construction of chloroplasts, and nitrogen availability is a key limiting factor for photosynthesis (Yue et al., 2017). Nitrogen enrichment can alleviate nitrogen limitation in the local soil, optimize the physiological state and structural integrity of leaves, and enhance photosynthetic activity (Borghetti et al., 2017; Hacke et al., 2010; Zhang et al., 2018). Consequently, the non-structural carbohydrate (NSC) pool enters a state where supply exceeds consumption, leading to an increase in NSC content. Generally, NSC transport adheres to the principle of proximal transport, moving from the upper morphological regions to the lower ones (Dietze et al., 2014; Tixier et al., 2018). As the primary site for NSC production (Chapin et al., 1990), leaves are the most directly and significantly affected by nitrogen enrichment. Nitrogen enrichment induces plants to adopt a “growth-driven mode,” increasing carbon allocation towards respiration and structural growth. Soluble sugars, which serve as direct substrates for respiratory and other metabolic activities (Gibon et al., 2009), are rapidly consumed in this “growth-driven mode.” However, their overall content still increases due to the substantial breakdown of starch replenishing the soluble sugar pool (Hoch and Körner, 2003). Therefore, the specific pattern of NSC component changes involves the accumulation of soluble sugars in the leaves, while starch is significantly depleted ( Figure 2 ; Supplementary Figure S3 ). Similar results have been observed in studies on Quercus mongolica (Zhang Y.-L. et al., 2020). However, this pattern is not constant across all growth stages. When the study subjects are seedlings, nitrogen enrichment drives them into a more pronounced “growth-driven mode,” leading to the rapid consumption of all NSC components, resulting in a reduced NSC pool and accelerated biomass growth (Li et al., 2020).

Downward in the plant, the distribution patterns of NSC and its components in branches and stems are similar to those in leaves, but the magnitude of changes gradually decreases with increasing distance from the carbon source organs (leaves). Notably, the NSC content in leaves and branches, as well as leaf starch content under medium or low nitrogen enrichment, exhibit very clear “growth-driven mode” carbon allocation strategies. However, when nitrogen enrichment concentrations reach high levels, this strategy disappears, supporting our hypothesis that the impact of nitrogen enrichment on the NSC pool exhibits a threshold effect (Du et al., 2020; Zhou X. et al., 2024; Zhu et al., 2021). In contrast, the NSC variation pattern in roots is entirely different from that in other organs. Under nitrogen enrichment, the NSC pools and their components in roots are expanded. Subterranean NSC reserves are crucial for tree recovery from stress (He et al., 2020), serving as energy sources for growth and development following adverse conditions such as drought, thereby ensuring the energy balance of the trees (Hartmann et al., 2013). Thus, under nitrogen enrichment, the aboveground parts adhere to a “growth-driven” carbon allocation model, while the underground parts follow a “stress-storage” carbon allocation model.

Recent nitrogen addition experiments on three tree species have demonstrated that, for Acer pseudoplatanus, nitrogen enrichment initially increases and subsequently decreases its NSC content (Zhou X. et al., 2024). Another study similarly found that nitrogen addition first promotes and then inhibits NSC content in Syzygium bullockii and Carallia brachiate (Chen et al., 2022). Researchers typically focus on the seasonal variations of NSC content (Martínez-Vilalta et al., 2016) and NSC dynamics under prolonged drought stress (Rowland et al., 2015), paying less attention to the temporal threshold effects of nutrient enrichment. Consequently, studies addressing these threshold effects are rare and valuable. Our meta-analysis, integrating various studies as hypothesized, revealed that the duration of nitrogen enrichment also exhibits a significant threshold effect on NSC.

Firstly, no significant promoting effect of phosphorus enrichment on the NSC pool was observed ( Figure 3 ; Supplementary Figure S4 ). Alleviating phosphorus limitation did not drive plants into a growth-driven mode as nitrogen alleviation does, indicating that phosphorus limitation is not as prevalent in ecosystems as nitrogen limitation. A long-term experiment also yielded similar results, where phosphorus addition slowed plant growth rates and prompted more conservative resource allocation strategies (Báez and Homeier, 2018). Unlike nitrogen, which directly mediates plant photosynthetic processes (Fleischer et al., 2013) and thereby influences the NSC pool (Wiley and Helliker, 2012), phosphorus affects the NSC pool indirectly. Consequently, the gradient effect from the upper to the lower parts of the plant, as seen with nitrogen enrichment, was not observed with phosphorus enrichment. Instead, fluctuations in NSC content were markedly greater in stems and roots. Phosphorus fertilization can stimulate xylem growth and reduce conduit wall thickness (Thomas et al., 2006; Yi et al., 2022). Rapidly growing xylem with thinner conduit walls and pit membranes allows small air bubbles to pass through more easily, thereby increasing the likelihood of xylem embolism (Cochard et al., 2010; Delzon et al., 2010). Stems and roots are the largest biomass organs in woody plants and have the longest water transport distances (Cochard et al., 2010). Different species exhibit significant variations in their tolerance to phosphorus effects, as well as in the conditions governing water transport efficiency and safety (Villagra et al., 2013). Therefore, the indirect effect of phosphorus variation on NSC content through the regulation of water transport is most pronounced in these two organs.

Notably, as phosphorus enrichment concentrations increase, soluble sugars in leaves exhibit a clear upward trend, whereas the opposite trend is observed in roots ( Figure 2 ). The reason for this is twofold. With higher phosphorus enrichment concentrations, soil water potential increases, leading plants to experience physiological drought (Hasanuzzaman et al., 2018) and necessitating the accumulation of soluble sugars to regulate internal osmotic pressure (Secchi and Zwieniecki, 2011). Meanwhile, aboveground parts, especially the canopy, serve as the primary sites for carbon assimilation and fruit development, thereby possessing a higher priority for carbon allocation (Minchin and Thorpe, 1996). Soluble sugars, acting as carbon currency, are transported from underground to aboveground parts, resulting in a negative correlation between soluble sugar contents in leaves and roots. In summary, the effects of nitrogen and phosphorus enrichment exhibit significant differences in woody plants.

Herbaceous plants respond to nutrient enrichment in their NSC content in a manner that is completely different from woody plants. Herbaceous plants possess smaller biomass and are typically annual or biennial (Liu et al., 2016). Additionally, herbaceous plants are characterized by rapid growth and weaker physiological regulatory capabilities (Zhao and Dixon, 2011), making them highly sensitive to environmental changes, including nutrient enrichment. Studies have demonstrated that nitrogen addition significantly affects the richness, diversity, and stoichiometric ratios of herbaceous plants (Miao et al., 2024; Soons et al., 2017; Wu et al., 2021). In this context, the growth-promoting effects of nitrogen enrichment are more pronounced in short-lived herbaceous plants, resulting in extensive consumption of NSC for growth. This is evidenced by a substantial reduction in NSC pools and their components across almost all concentrations of nitrogen enrichment, with NSC and its component contents decreasing as nitrogen enrichment concentration increases ( Figure 2 ).