Introduction

Carbon (C), nitrogen (N), and phosphorus (P) are the essential elements utilized by plants1. Ecological stoichiometry, which explores the balance of energy and chemical elements in biological systems, provides a powerful framework for understanding nutrient cycling and the functioning of terrestrial ecosystems2,3. The study of leaf ecological stoichiometry in particular has become an effective method for determining plant growth rates, nutrient utilization efficiency, and the availability of soil nutrients for plant growth4,5. Among these essential elements, C is the most important for dry matter accumulation and serves as the substrate and energy source for various critical physiological processes6. The C: N and C: P ratios represent the ability of plants to assimilate C while simultaneously absorbing both N and P; in contrast, the N: P ratio reflects the dynamic balance between soil nutrients and plant nutrition demands7,8. Ecosystem C: N:P stoichiometry varies significantly among ecosystem types, land-use categories, and environmental gradients9,10. Recently, most studies on C: N:P stoichiometry have focused on the plant organ level, with a scarcity of studies conducted at the plant functional group or community levels in grasslands11.

A plant community is formed through the assemblage of species adapted to a specific environment and through mutual competition. Thus, plants with different strategies compete for limited resources, such as water, light, and nutrients12,13. Changes in community traits are often weighted by the relative abundances of species and are mostly drive by dominant species rather than other species14. Therefore, any changes in community composition can have significant implications for community-level nutrient resources15,16. Previous studies in high-altitude grasslands have shown that net plant–plant interactions shift from competitive to facilitative in response to environmental changes17,18. However, little is known about the changes in soil nutrient availability and plant C: N:P stoichiometry associated with the progression of community succession, which limits our understanding of nutrient geochemical cycles in ecosystems19.

The Qinghai–Tibetan Plateau, popularly known as the “Roof of the World,” contains unique and fragile ecosystems that harbor diverse plant species adapted to extreme environmental conditions20. Alpine meadows, a dominant vegetation type of the plateau, support a variety of plant functional groups, including grasses, sedges, legumes, and forbs, which each play crucial roles in maintaining ecosystem services, such as nutrient cycling, carbon sequestration, and forage production21. However, in recent decades, the plateau has experienced profound changes caused by climate warming and overgrazing, leading to widespread degradation of alpine meadows22. These ecological pressures often promote the invasion of exotic plant species, which can in turn create positive feedback loops that maintain and exacerbate the degraded state by which they are caused23. On one hand, toxic plants in grassland are harmful to grazing animals which induces substantial losses of animal husbandry24. Moreover, these toxic weeds can form intraspecific aggregations that enhance their competitive ability against other species16,25. Therefore, the wide expansion of toxic plants under climate change and human activities has been concerned globally23,26. Recent research on the spread of toxic plants has focused mainly on key contemporary factors such as grazing livestock and climate change27. However, how toxic plants adapt to the poor soil conditions of degraded grasslands remains unknown.

One of the most notorious invasive species in the Qinghai–Tibetan Plateau is Ligularia virgaurea, a perennial herbaceous plant native to Central Asia that has gradually spread across alpine grasslands28. Ligularia virgaurea can become dominant in grazed alpine communities owing to its toxicity to grazers. Its invasion is associated with significant shifts in species composition, community structure, and ecosystem processes6. The allelochemicals released by L. virgaurea alter soil microbial communities, thus affecting soil nutrients and creating positive feedback effects25,29. There is also evidence that allelopathic effects of volatile or aqueous leaf or root extracts of L. virgaurea inhibit the seed germination and growth of forb species native to the Qinghai–Tibet Plateau28. Consequently, it alters the growth of associated plant species and the overall functioning of invaded alpine meadows.

Previous studies have documented the effects of L. virgaurea on nutrient cycling, plant community composition, and soil properties26,29. However, a comprehensive understanding of how the spread of L. virgaurea impacts the ecological stoichiometry of associated plants remains elusive. However, addressing this knowledge gap is crucial for developing effective management strategies to mitigate the impacts of invasive species and restore the resilience of alpine meadows on the Qinghai–Tibetan Plateau. This study was conducted to investigate the effect of L. virgaurea spread on the ecological stoichiometry of associated plants in alpine meadows of the Qinghai–Tibetan Plateau. Specifically, we aimed to (1) quantify the changes in C, N, and P contents and stoichiometric ratios (C: N, C: P, and N: P) of different plant functional groups in response to increasing L. virgaurea density; (2) analyze the impacts of L. virgaurea spread on soil nutrient contents and their relationships with plant stoichiometric traits; and (3) identify the underlying mechanisms responsible for the observed changes in plant ecological stoichiometry under L. virgaurea invasion.

Materials & methods

Study site

The study was conducted in an alpine meadow located at an elevation of 3,532 m in Henan County (34°42′ N, 101°36′ E), Huangnan Tibetan Autonomous Prefecture, Qinghai Province, China. The region has a typical plateau continental climate, with a mean annual solar radiation of 2,580 h, an average annual temperature of 0.4–3.4 °C, and an annual precipitation accumulation of 579–616 mm (most of which falls between May and September). The vegetation is typical of nearby alpine meadows, dominated by Carex alatauensis, C. capillifolia, and Stipa aliena. Other commonly occurring species include C. atrofusca, Poa pratensis, Melissitus ruthenica, and Lancea tibetica. The study site is utilized as winter rangeland by local herders (grazing from September to May) with a heavy grazing intensity of about 7.94 sheep units per hectare. Over the past few decades, Ligularia virgaurea has invaded the grassland and gradually become the dominant species (reaching an average density of up to 133.6 plants per m2)30, posing serious degradation challenges to the grassland. The permission (granted by Grassland Working Station of Henan county, Huangnan, China and local herdsmen) had been obtained for field study and plant materials’ collection.

Sampling and chemical analysis

To focus on the effect of L. virgaurea density and minimize the influence of confounding environmental gradients, the study was deliberately conducted within a contiguous area characterized by uniform topography, consistent soil type, and a homogenous macro-climate. To isolate the effect of L. virgaurea invasion from the confounding effects of contemporary grazing, the entire study area was fenced in May 2022, one year prior to sampling, to exclude livestock. Three plots (each 80 × 80 m) with varying naturally occurring densities of L. virgaurea were established, ensuring a sufficient distance of approximately 30–40 m between them. In August 2023, within each plot, four levels of L. virgaurea density were selected based on plant counts per square meter: LN (no L. virgaurea), LL (low density, 30–40 plants/m2), LM (moderate density, 100–120 plants/m2) and LH (high density, 200–240 plants/m2) (Fig. 1). These density categories were defined according to the Local Standards of Qinghai Province (Grassland Toxic Weed Management Guidelines, DB63/T 241–2021), which align with the invasion intensities commonly observed in local grasslands. Three quadrats (0.5 m × 0.5 m) were randomly established for each density level. Aboveground tissues of all species within each quadrat were sampled and sorted into five mutually exclusive functional groups: grasses, sedges, legumes, forbs, and L. virgaurea. Species identification was undertaken by collectors according to Hou31 et al. To minimize sampling bias and ensure the representativeness of the chemical analysis, the aboveground tissues for each plant functional group collected from the three quadrats of the same L. virgaurea density level within a plot were pooled to form one composite sample per group per density per plot. Each composite sample was then homogenized and split into three analytical subsamples for subsequent nutrient determination. At the community level, another three quadrats were randomly surveyed within each L. virgaurea density level, and aboveground tissues of all observed species were collected and pooled as one sample per quadrat. Root samples were collected using a root auger (7-cm diameter) from each quadrat from depths of 0–10 cm and 10–20 cm and then washed to remove soil and other impurities. After sampling, all tissues and roots were oven-dried at 85 °C to a constant mass, weighed, and ground for further nutrient analysis at the laboratory facilities of Qinghai University, Xining, Qinghai, China. Soil samples were also collected from 0 to 20 cm deep using a soil auger (3.5-cm diameter) within each quadrat, and all the soil samples of three quadrats in a plot were pooled into one sample. After air-drying, soil samples were passed through a 1-mm sieve for nutrient analysis.

Fig. 1
Fig. 1
Full size image

The sets for different density levels of Ligularia virgaurea. LN: None of L. virgaurea, LL: Low density of L. virgaurea, LM: Moderate density of L. virgaurea, LH: High density of L. virgaurea.

The total soil and plant organic carbon content was determined using the oil bath-K2CrO7 titration method (oxidization with dichromate in the presence of H2SO4, heated to 180 °C for 5 min followed by titration with FeSO4)32. The total nitrogen content of soil and plant samples was analyzed using the Kjeldahl digestion method with a Nitrogen Analyzer System (Kjeltec 2300 Auto System II, Foss Tecator AB, Höganäs, Sweden). Total phosphorus content was determined using the molybdate blue colorimetric method with a spectrophotometer (SP-723; Analytikjena, Germany) after digestion with H2SO4 and H2O2. Soil NH4+-N and NO3−-N levels were measured using a FIAstar 5000 Analyzer (Foss Tecator AB). Available phosphorus in soil was analyzed using the Molybdenum-antimony anti-colorimetric method16,32.

Statistical analysis

Data analyses were conducted using SPSS (version 17.0; SPSS Inc., Chicago, IL, USA). As necessary, data were transformed to meet the homogeneity of variance and normality assumptions of ANOVA. Independent sample t-tests were used to determine significant differences between various L. virgaurea density levels for all parameters. For all two-way ANOVAs that revealed significant main or interaction effects (P < 0.05), Tukey’s Honest Significant Difference (HSD) post-hoc test was subsequently applied to identify significant pairwise differences among the levels of functional groups and L. virgaurea density. Statistical significance was determined at the 95% level (P < 0.05). To examine the overall relationships between the multivariate profile of plant stoichiometric traits and soil nutrient parameters, Mantel tests were performed using the linkET package in R. Separate Euclidean distance matrices were constructed for the dataset of plant stoichiometric traits (including nitrogen, phosphorus, and carbon contents, and C: N, C: P, N: P ratios in both aboveground and root tissues) and the dataset of soil variables (including TN, NH₄⁺-N, NO₃⁻-N, TP, AP, TOC, MBC, MBN, MBP). The correlation between these two distance matrices was assessed using the Spearman rank correlation coefficient. The statistical significance of the Mantel correlation statistic (r) was tested with 999 permutations. A significance threshold of P < 0.05 was applied, with correlations at P < 0.01 considered highly significant.

Results

The biomass and stoichiometry of different functional group plants under different L. virgaurea density levels

With increasing L. virgaurea density, the total aboveground biomass in the alpine meadow increased significantly (Fig. 2). However, when the biomass of L. virgaurea was excluded, the total aboveground biomass of other plant functional groups declined gradually with L. virgaurea density. Specifically, as L virgaurea density increased, the biomass of forbs and legumes showed a decreasing trend; in contrast, sedge biomass increased, and grasses biomass showed an initial decline followed by an increase.

Fig. 2
Fig. 2
Full size image

Aboveground biomass of different functional groups in the community with different levels of Ligularia virgaurea.

The nitrogen content in aboveground tissues of grasses and legumes decreased with increasing L. virgaurea density (Fig. 3a). In high-density L. virgaurea (LH) conditions, the nitrogen content of grasses and legumes was significantly lower than that in the no L. virgaurea (LN) and low-density L. virgaurea (LL) conditions (P < 0.05). However, L. virgaurea density had no significant effect on the nitrogen content of sedges, forbs, or L. virgaurea itself (P > 0.05). In the LH community, the nitrogen content of L. virgaurea was higher than that of both grasses and sedges. The phosphorus content in plant tissues of grasses, legumes, forbs, and L. virgaurea initially increased and then decreased with increasing L. virgaurea density (Fig. 3b). The highest phosphorous content values for all sample types was observed in the moderate-density L. virgaurea (LM) community. In contrast, phosphorus content in sedge tissues increased gradually with increasing L. virgaurea density. In the LH community, phosphorus content in L. virgaurea was higher than that in forbs, grasses, and sedges. As L. virgaurea density increased, the organic carbon content of plants in the other four functional groups decreased (Fig. 3c). Ligularia virgaurea had higher organic carbon content than the other four functional groups, and the difference in organic carbon content widened as L. virgaurea density increased. Overall, both the functional group and L. virgaurea density significantly affected plant nitrogen, phosphorus, and organic carbon contents. Additionally, the interaction effect on carbon, nitrogen, and phosphorous contents between functional groups and L. virgaurea density was significant (Table 1). Phosphorus content was significantly higher in the LH and LM treatments than that in the LN and LL treatments. Nitrogen and carbon contents were significantly lower in the LH treatment than that in the LN, LL and LM treatments (Tukey’s HSD, P < 0.05, Table 1; Fig. 3).

Fig. 3
Fig. 3
Full size image

The changes of nitrogen (a), phosphorus (b) and carbon (c) contents of different functional groups with the increasing of levels of Ligularia virgaurea. Different lowercase letters represent significant differences (P < 0.05) among different density of L. virgaurea. Different colors of lowercase letters represent different functional groups. Red: grasses; Yellow: sedges; Green: legumes; Blue: forbs; Purple: Ligularia virgaurea.

Table 1 Two-way ANOVA results of Ligularia virgaurea density (D) and functional group (G) and post-hoc tukey’s HSD results of L. virgaurea density on nitrogen (N), phosphorus (P) and carbon (C) contents, and C: N, C: P and N: P ratios of plants.
Full size table

As L. virgaurea density increased, the C: N ratio in aboveground tissues of grasses gradually increased (Fig. 4a). Under LH conditions, the C: N ratio of grasses was significantly higher than that under LN conditions (P < 0.05). Conversely, the C: N ratio of sedges decreased as L. virgaurea density increased, being significantly lower in LH conditions relative to LN and LL conditions (P < 0.05). There were no significant differences in plant C: N ratios for legumes, forbs, and L. virgaurea as L. virgaurea density changed. Thus, while L. virgaurea density had no significant effect on plant C: N, the interaction between functional group and L. virgaurea density was significant (Table 1). Except for legumes, the C: P ratio of the other groups, i.e., grasses, sedges, forbs, and L. virgaurea, decreased as L. virgaurea density increased (Fig. 4b). Significant differences in the C: P ratio were observed between LH and LN conditions among grasses and sedges. Additionally, the C: P ratio of forbs in LM conditions was significantly lower than that in LN and LL conditions, and the C: P of L. virgaurea in LM and LH conditions was significantly lower than that in LL conditions (P < 0.05). Therefore, both functional group and L. virgaurea density significantly affected plant C: P, and their interaction was significant (Table 1). As L. virgaurea density increased, the N: P ratio of grasses, forbs, and L. virgaurea first decreased and then increased (Fig. 4c). In contrast, sedges and legumes showed a gradual decrease in N: P ratio as L. virgaurea density increased. Both L. virgaurea density and functional group significantly affected plant N: P, but their interaction was not significant (P > 0.05, Table 1). The difference of C: N ratio was significant between LH and LM treatments. C: P and N: P ratios were significantly lower in the LM and LH treatments than that in the LN and LL treatments (Tukey’s HSD, P < 0.05, Table 1; Fig. 4).

Fig. 4
Fig. 4
Full size image

The changes of C: N (a), C: P (b) and N: P (c) ratios of different functional groups with the increasing of levels of Ligularia virgaurea. Different lowercase letters represent significant differences (P < 0.05) among different density of L. virgaurea. Different colors of lowercase letters represent different functional groups. Red: grasses; Yellow: sedges; Green: legumes; Blue: forbs; Purple: Ligularia virgaurea.

Stoichiometry of aboveground and belowground tissues under different L. virgaurea densities

As L. virgaurea density increased, the nitrogen content of community plants decreased, resulting in a general increase in C: N ratios (Fig. 5a, b). In LM and LH conditions, nitrogen content was significantly lower and C: N ratios were significantly higher than in LN conditions (P < 0.05). Conversely, plant phosphorus content increased with L. virgaurea density, with the highest phosphorous contents observed in LM conditions, significantly exceeding those in LN and LL conditions (Fig. 5c). Consequently, plant C: P ratios decreased as L. virgaurea density increased, and C: P in LH and LM conditions was significantly lower than that in LN and LL conditions (P < 0.05, Fig. 5d). Plant organic carbon content did not significantly differ for any of the plant groups among LN, LL, and LM conditions (P > 0.05, Fig. 5e). However, in LH conditions, plant carbon content was significantly lower than that in the three lower L. virgaurea densities. Among N: P ratios, there was a declining trend as L. virgaurea density increased. The N: P ratios of plants in LH and LM conditions were significantly lower than those in LN and LL conditions (P < 0.05, Fig. 5f).

Fig. 5
Fig. 5
Full size image

Nitrogen (a), phosphorus (c) and carbon (e) contents, and C: N (b), C: P (d) and N: P (f) ratios of plants in the community with different levels of Ligularia virgaurea. Different lowercase letters represent significant differences (P < 0.05) among different density of L. virgaurea.

The changes in nitrogen content and C: N ratio of roots as L. virgaurea density increased showed no consistent pattern (Fig. 6a, b). In LH conditions, root nitrogen content in the 0–10 cm soil layer was the highest and significantly higher than that in LL conditions. Correspondingly, the C: N ratio in LH conditions was the lowest and significantly different from that under LM conditions (P < 0.05). There was no significant difference in root nitrogen and C: N ratios in the 10–20 cm soil layer across the different L. virgaurea densities (P > 0.05). Root phosphorus content in the 0–10 cm soil layer increased with L. virgaurea density, while C: P ratios decreased gradually (Fig. 6c, d). This difference between LN conditions and each of the L. virgaurea-containing conditions was significant. In LN conditions, root phosphorus content in the 0–10 cm soil layer was significantly lower, and the C: P ratio was significantly higher than that in the 10–20 cm layer (P < 0.05). The highest root organic carbon content in both the 0–10 cm and 10–20 cm soil layers was observed under LM conditions (Fig. 6e). Roots under LN and LL conditions in the 0–10 cm layer and under LH conditions in the 10–20 cm layer had significantly lower organic carbon content than those under LM conditions. For N: P ratios, LN conditions had a significantly higher value than those of LL and LM conditions in the 0–10 cm root layer (Fig. 6f). However, there was no significant difference in N: P ratios of the 10–20 cm root layer across the different L. virgaurea density levels (P > 0.05).

Fig. 6
Fig. 6
Full size image

Nitrogen (a), phosphorus (c) and carbon (e) contents, and C: N (b), C: P (d) and N: P (f) ratios of plant roots in the community with different levels of Ligularia virgaurea. Different lowercase letters represent significant differences (P < 0.05) among different density of L. virgaurea in 10–20 cm soil. Different capital letters represent significant differences (P < 0.05) among different density of L. virgaurea in 0–10 cm soil.

Correlations in soil nutrient contents and plant stoichiometry during the spread of L. virgaurea

Soil total nitrogen (TN) and total phosphorus (TP) contents showed no significant change as L. virgaurea density increased (P > 0.05, Table 2). However, ammonium nitrogen (NH₄⁺-N) content decreased significantly (P < 0.05), while available phosphorus (AP) content increased with L. virgaurea density. Soil nitrate-N (NO₃⁻-N) was highest under LL conditions (15.17 µg·g⁻¹) but was sharply lower under both LM and LH conditions. Additionally, total organic carbon (TOC), microbial biomass nitrogen (MBN), microbial biomass carbon (MBC), and microbial biomass phosphorus (MBP) contents increased significantly with L. virgaurea density (P < 0.05, Table 2).

Table 2 The nutrient contents of soil in the community with different levels of Ligularia virgaurea.
Full size table

Mantel test results showed significant negative correlations both between NH₄⁺-N and TOC and between NO₃⁻-N and TP, MBN, and MBC (P < 0.05, Fig. 7). MBC, MBP, and MBN were significantly positively correlated with each other. Additionally, MBC exhibited significant positive correlations with TOC, AP, and TP, while MBP had significant positive correlations with TN only. MBN exhibited significant positive correlations with both TP and AP (P < 0.05). For plant stoichiometry, nitrogen content and C: N ratio in plants had highly significant correlations with soil AP (P < 0.01) and significant correlations with soil TOC and MBC (P < 0.05). Plant phosphorus and C: P exhibited significant correlations with soil MBC and highly significant correlations with both NH₄⁺-N and TOC (P < 0.01). Plant carbon content exhibited a significant correlation with soil MBN and highly significant correlations with both soil AP and MBC (P < 0.01). Plant N: P exhibited highly significant correlations with both soil NH₄⁺-N and soil AP (P < 0.01) and significant correlations with soil TOC, MBN, and MBC (P < 0.05, Fig. 7).

Fig. 7
Fig. 7
Full size image

Mantel’s results of soil nutrients on the stoichiometry traits of plants in the Ligularia virgaurea spread community. N: plant nitrogen content and C: N in aboveground biomass and roots; P: plant phosphorus content and C: P in aboveground biomass and roots; C: plant carbon content in aboveground biomass and roots; NP: plant N: P in aboveground biomass and roots; TN: soil total nitrogen; NH4+-N: soil ammonium nitrogen; NO3--N: soil nitrate nitrogen; TP: soil total phosphorus; AP: soil available phosphorus; TOC: soil total organic carbon; MBC: microbial biomass carbon; MBP: microbial biomass phosphorus; MBN: microbial biomass nitrogen.

Discussion

The present study comprehensively examined the ecological consequences of L. virgaurea invasion in a typical alpine meadow of the Qinghai–Tibetan Plateau. Our findings revealed several important insights into the intricate interactions between invasive species and native plant communities, as well as their implications for nutrient cycling and ecosystem functioning. Studies have suggested that overgrazing provides a competitive advantage to L. virgaurea by altering plant community structure, which is closely related to the soil environment29. In our study, the aboveground biomass of different functional groups changed as L. virgaurea density increased. The total biomass of other plant functional groups declined gradually as L. virgaurea density increased, primarily owing to the observed decrease in biomass of legumes and forbs. As a species reported to possess allelopathic properties28, L. virgaurea is likely to displace high-quality forage and affects the seed germination and plant establishment of cooccurring species33. Thus, legumes and forbs may spatially compete with L. virgaurea, resulting in their lower dominance after L. virgaurea invasion. Previous studies have shown that volatile and aqueous leaf and root extracts of L. virgaurea influence the germination and growth of native forb species28,34. Our present results are consistent with these previous findings. While plants with different life strategies compete for limited resources (i.e., water, light, and nutrients), they may also gain facilitative shelter from neighboring plants against severe climatic events in alpine grasslands12,16; this phenomenon could explain the observed increase in sedge biomass associated with L. virgaurea density. Although grasses suffered some observable competitive suppression effects, it is possible that they receive some similar benefit from the aggressive spread of L. virgaurea.

It is well known that carbon, nitrogen, and phosphorus are the principal chemical elements and the stoichiometric C: N:P ratios of plant have been widely used as indicators of nutrient limitations on plant growth8,35,36. The pathway of carbon assimilation differs from the absorption of nitrogen and phosphorus, as carbon is obtained from atmospheric CO₂ through photosynthesis37,38. In this study, the carbon contents of grasses, sedges, and legumes were significantly lower under the highest L. virgaurea density (i.e., LH) relative to the other density levels (Fig. 3c). This might be owing to limited sunlight reaching plants under high L. virgaurea density, which reduces the photosynthetic rate. Additionally, photosynthetic capacity is strongly correlated with nitrogen allocation to the photosynthetic apparatus39, so the lower photosynthetic efficiency of grasses and legumes would result in reduced nitrogen absorption. However, the nitrogen content of L. virgaurea remained unchanged across its density levels (Fig. 3a). This could be explained by the stronger capacity for nitrogen absorption of L. virgaurea compared to grasses and legumes. This aligns with the observed decline in soil inorganic nitrogen content (NH₄⁺-N and NO₃⁻-N, Table 2) as L. virgaurea density increased. Thus, L. virgaurea can spread rapidly by more efficiently absorbing nitrogen relative to other species in the community. Our results are consistent with Suter40 et al., who found that low nitrogen input promotes weed (Senecio jacobaea) invasion by reducing competition from desirable species. In association with changes in carbon and nitrogen contents, the C: N ratio increased in grasses and decreased in sedges as L. virgaurea density increased (Fig. 4a). Many studies have found that sedges have greater root biomass and root-to-shoot ratios than grasses41,42. Therefore, we suggest that sedges have a greater ability to take up nitrogen from belowground compared to grasses under the same conditions associated with inefficient photosynthesis. This is consistent with the observed change in biomass (Fig. 2), indicating that sedges might be more competitive than grasses under L. virgaurea invasion. While our study focused on stoichiometric outcomes and did not measure root depth or water-use efficiency, these factors could be significant contributors to the observed competitive suppression. Differences in root architecture or water competition may have further limited the nitrogen uptake capacity of grasses relative to the more deeply rooted L. virgaurea or sedges42,43.

Except for nitrogen, the phosphorus content of an organism is largely associated with the allocation of phosphorus to ribosomal RNA, which is related to its growth rate7. Thus, a change in N: P ratio is thought to alter the competitiveness of a species in a manner that depends on their growth rate and life history8,35. In this study, the plant phosphorus contents of most functional groups increased firstly and then declined with L. virgaurea density increasing (Fig. 3b). This may be attributed to the high biomass production and persistent litterfall of L. virgaurea that stimulate microbial activity, particularly phosphorus-solubilizing bacteria (PSB) and fungi, which secrete phosphatases to hydrolyze organic phosphorus44. This enhanced mineralization would directly increase soil phosphorus availability. Consequently, phosphorus acquisition of neighboring plants would increase within a certain range of L. virgaurea densities. As an increasingly dominant plant in alpine grasslands, much like Stellera45, L. virgaurea might create islands of fertility by this microbial activation, thus promoting its own continued expansion. However, when L. virgaurea density reaches a certain level, the phosphorus concentration in some plants (such as grasses, legumes, and forbs) appeared to decline owing to increased phosphorus demand. The different stoichiometric responses among functional groups likely reflect their distinct ecological strategies. For example, the sustained biomass and lower C: P and N: P ratios in sedges (Figs. 2 and 4) are consistent with a resource-conservative strategy and efficient belowground nutrient foraging, traits well-documented for alpine Kobresia species41,43. However, the competitive ability of forbs was weaker under the highest L. virgaurea density compared to the other densities, according to the value of N: P (Fig. 4c). This may relate with the similar lifestyle between forbs and L. virgaurea. The over intensity of L. virgaurea influences the ability of forbs to capture resources around. On the other hand, grasses suffer stronger nitrogen limitations induced by the spread of L. virgaurea but initiatively enhance their competition in this community by adjusting the value of N: P. Our observation that plants adjusted C: N, C: P and N: P ratios to maintain competitiveness under invasion supports the idea that plant functional groups differentially regulate stoichiometry in response to environmental change46.

At the community level, the carbon content in 0–10-cm deep roots increased with L. virgaurea density (Fig. 6e). This is related to the greater amount of litter falling back to the rhizosphere soil under higher L. virgaurea density, which is consistent with the findings of previous studies45,47. However, the nitrogen content of plant aboveground tissues declined significantly with increasing L. virgaurea density, thus increasing the C: N ratio (Fig. 5a, b). This may be attributed to the substantial increase in L. virgaurea plants, which absorb more nitrogen to support their comparatively high aboveground biomass growth. Consequently, nitrogen availability for other species (such as grasses and legumes) was more limited, causing a decline in community plant nitrogen concentration. The lower nitrogen absorption capacity of other species compared to L. virgaurea was also reflected in root nitrogen concentration (Fig. 6a, b). In comparison with native species, L. virgaurea demanded more nitrogen and thus absorbed more nitrogen, as indicated by its greater root biomass and root biomass nitrogen stock47. Similar to the findings of Calluna vulgaris in N-saturated heathlands, our study suggests that nitrogen limitation can persist even under conditions of high nitrogen availability, possibly due to efficient N uptake by invasive species or microbial N immobilization48. On the other hand, the significant increase in plant C: N ratios predict the production of poorer quality litter that may decompose more slowly, potentially leading to a longer-term accumulation of soil organic carbon and a reduction in nitrogen mineralization rates49,50,51. In contrast with nitrogen, phosphorus content in aboveground tissues and 0–10-cm deep roots increased with L. virgaurea density. This directly reduced C: P and N: P ratios in both aboveground and belowground tissues (Figs. 5c, d and f and 6c, d and f). In particular, aboveground N: P ratios were between 14 and 16 in LN and LL conditions, indicating co-limitation of nitrogen and phosphorus52. As L. virgaurea density increased, communities gradually became more nitrogen-limited (N: P < 14). Thus, the increased density of L. virgaurea may have increased the phosphorus uptake of surrounding plants while also exacerbating nitrogen limitation of this plant community. This phenomenon has also been reported in alpine meadows in Gansu province, China, where soil nitrogen limitation was identified as a dynamic factor controlling L. virgaurea expansion33. Simultaneously, the widespread decrease in C: P and N: P ratios (Figs. 5d and f and 6d and f) suggests a relative enrichment of phosphorus in plant tissues and litter, which could accelerate phosphorus cycling. This decoupling of nitrogen and phosphorus cycles—towards greater nitrogen limitation and phosphorus availability—may create a positive feedback that stabilizes the dominance of L. virgaurea and other phosphorus-efficient species, thereby altering long-term nutrient cycling trajectories in these alpine meadows.

Plants can dramatically modify the composition, biomass, and activity of soil microbial communities, largely through rhizodeposition53. Microbial biomass and activity in soils associated with L. virgaurea, as observed among many other invasive plants, were higher compared to native species45,54. In our study, microbial biomass carbon, nitrogen, and phosphorus each increased significantly as L. virgaurea density increased (Table 2). This was related to the greater aboveground litter input and belowground root biomass compared to soils associated with other species47, suggesting that L. virgaurea benefits microbial communities reliant on C-rich exudates. The significant increase in soil microbial biomass phosphorus (MBP) under higher L. virgaurea density (Table 2) indicates a strengthened microbial capacity for phosphorus transformation. This suggests that the high biomass and root exudates of L. virgaurea likely stimulate the microbial community, particularly phosphorus-solubilizing microorganisms, which are known to secrete phosphatases to hydrolyze organic phosphorus44. The significant increases in microbial biomass carbon (MBC) and nitrogen (MBN) (Table 2) hold broader ecological significance beyond indicating greater microbial abundance. They signal a shift in the belowground ecosystem towards a state of enhanced nutrient retention and accelerated turnover within the microbial loop. This ‘primed’ microbial community likely plays a crucial role in mineralizing nutrients that support L. virgaurea’s dominance. While our data cannot confirm a permanent microbial regime shift, the establishment of this self-reinforcing feedback—where the invader promotes a microbial community that in turn facilitates its nutrition—suggests a stable, alternative state for the invaded meadow that could be resistant to reversion.

The increase in microbial biomass nutrients indirectly altered soil element concentrations. Specifically, total organic carbon (TOC) in soil increased with L. virgaurea density (Table 2). This is consistent with previous studies45,55. The rise in total organic carbon highlights the capacity of L. virgaurea to augment soil carbon stocks through its high biomass production and recalcitrant litter—which together serve as a feedback mechanism that stabilizes its dominance in invaded grasslands26,55. On the other hand, soil ammonium-N (NH₄⁺-N) decreased significantly in L. virgaurea-invaded communities compared to non-invaded controls, indicating intensified competition for bioavailable nitrogen as L. virgaurea established its dominance. This aligns with the documented efficient nitrogen acquisition of invasive plants that suppresses native competitors by limiting critical nutrients56,57. However, nitrate-N (NO₃⁻-N) was highest under LL conditions but declined sharply under higher-density conditions (Table 2), likely reflecting a shift in plant–microbe interactions. Early invasion stages may temporarily elevate nitrate-N owing to disrupted microbial nitrification, while higher densities drive rapid plant uptake, as observed in the alteration of nitrogen cycling caused by other invasive species57. Soil available phosphorus (AP) increased progressively with L. virgaurea density, indicating enhanced phosphorus mobilization occurred in invaded soils. Plant-derived organic inputs and microbial activity jointly drive the transformation of phosphorus from its recalcitrant to bioavailable forms, thus enhancing phosphorus accessibility without altering total phosphorus pools58. Meanwhile, high-P litter from L. virgaurea decomposes rapidly, thus releasing phosphorus back into the soil and sustaining available phosphorus levels59. This phenomenon also explains the observed increases in plant phosphorus and soil available phosphorus contents, which occurred alongside negligible change in total soil phosphorus. Similar to Hieracium invasion in New Zealand grasslands60, L. virgaurea may reduce soil mineral N and alter microbial biomass, potentially creating a nutrient environment that favors its persistence over native species.

According to the results, we found that significant correlations exist between plant stoichiometric traits and soil nutrient dynamics in L. virgaurea communities. The significant correlations between plant N-related metrics (nitrogen content, C:N) and soil available phosphorus (P < 0.01), organic carbon (P < 0.01), and microbial biomass carbon (P < 0.05) reflect the adaptive plasticity of L. virgaurea to alleviate N limitation through root exudation to enhance phosphatase or microbial activity, a strategy documented in other invasive species that exploit nutrient imbalances26,61. Notably, the linkage between plant N: P ratios and both soil AP and ammonium-N indicates a co-limitation scenario, in which L. virgaurea adjusts its N: P homeostasis to exploit transient pulses in nutrient availability—a trait linked to invasion success in fluctuating environments62. The “microbial priming” strategy mirrors mechanisms observed in other allelopathic invasive species that reengineer soil communities to favor their own mutualistic microbes29,45. In general, the success of L. virgaurea can be partially attributed to its distinct stoichiometric signature compared to the native species it displaces. Our data show that L. virgaurea maintains higher tissue concentrations of nitrogen and phosphorus than key native functional groups under high invasion pressure. This superior nutrient acquisition ability, potentially driven by more efficient uptake or stronger associations with soil microbes, allows it to outcompete natives like forbs and legumes that suffer from nutrient suppression, thereby facilitating its dominance. The competitive dynamics observed are likely driven by the combined effects of shading and allelopathy. The declines in nitrogen content and photosynthetic capacity (inferred from higher C: N) in grasses and legumes are classic responses to light limitation under shading. In contrast, the specific vulnerability of forbs and the significant alteration of the soil microbial biomass (Table 2) provide indirect support for the role of allelopathy, a mechanism previously confirmed for L. virgaurea in controlled bioassays28,34. A synthesis of our results with this literature suggests that shading may be the dominant mechanism driving reduced nitrogen status across the community, while allelopathy may be more critical for specific plant-soil feedback and the suppression of certain competitors like forbs. Future studies employing factorial designs (e.g., using shade cloth to mimic light reduction and activated carbon to adsorb allelochemicals in the soil) are needed to quantitatively disentangle these intertwined mechanisms, and studies that directly link stoichiometric shifts with a wider suite of plant functional traits, such as SLA and root: shoot ratio, would provide even deeper mechanistic insights into the strategies plants employ during invasion.

Conclusions

In summary, Ligularia virgaurea invasion disrupts biogeochemical cycling of carbon, nitrogen, and phosphorus in alpine meadows. It achieves dominance through a superior nutrient acquisition strategy, characterized by efficient nitrogen uptake and the promotion of microbially-mediated phosphorus conversion, which collectively suppress native competitors. Different plant functional groups exhibit distinct stoichiometric responses: grasses face intensified nitrogen limitation, while sedges maintain competitiveness through efficient nitrogen and phosphorus use. At the ecosystem level, the invasion enhances carbon storage and microbial biomass but exacerbates nitrogen limitation, creating a feedback loop that likely stabilizes the invaded state. These findings underscore the critical role of plant-soil-microbe interactions in driving invasion success and altering ecosystem stoichiometry.

Building on our findings, a critical next step is to test the specific hypothesis that climate warming exacerbates the positive plant-microbe feedback underpinning L. virgaurea’s dominance. A future multi-factorial experiment, coupling climate manipulation (e.g., warming and altered precipitation) with detailed profiling of the rhizosphere microbiome (e.g., metagenomics and enzyme activities), would elucidate whether the invader’s stoichiometric advantages are stable under future climates. This approach could identify key microbial functional groups that are sensitive to climate change, revealing novel targets for ecological management aimed at disrupting the core mechanism of invasion. Beyond plant-soil feedbacks, investigating how such stoichiometrically-driven changes in plant quality influence herbivory and food web dynamics represents a critical next step for understanding the comprehensive ecosystem-level impact of L. virgaurea invasion.