Leaf Stoichiometry of Potentilla Fruticosa Across Elevations Ranging from 2400 m to 3800 m in China’s Qilian Mountains (Northeast Qinghai-Tibetan Plateau)

Background: Plant species have developed their individual leaf stoichiometries to adapt to changes in the environment. Changes in plant leaf stoichiometry with elevation are largely undocumented, but could provide information critical to protecting or enhancing a species’ growth and development and manage the ecosystem housing it. We investigate the leaf stoichiometry of Potentilla fruticosa L. along with different elevations in China’s Qilian mountains (Northeast Qinghai-Tibetan Plateau). This study aims to reveal how elevations effect of the leaf stoichiometry of Potentilla fruticosa L. along with various soil properties in China’s Qilian mountains . Results: In our study, we selected seven elevations 2,400 m, 2,600 m, 2,800 m, 3,000 m, 3,200 m, 3,500 m, and 3,800 m elevation. We sampled leaves at top and middle of P. fruticosa from each of seven elevations. Maximum and minimum leaf carbon (C) concentrations ([C] leaf ) of 523.59 g kg -1 and 402.56 g kg -1 were measured at 2,600 m and 3,500 m, respectively. Showing a generally increasing trend with elevation, leaf nitrogen (N) concentration ([N] leaf ) peaked at 3,500 m (27.33 g kg -1 ). Leaf phosphorus (P) concentration ([P] leaf ) varied slightly over elevations of 2,400 m to 3,200 m, then dropped to a minimum (0.60 g kg -1 ) at 3800 m. While [C] leaf :[N] leaf , [C] leaf :[P] leaf and [N] leaf :[P] leaf varied little between 2,400 m and 3,000 m, at higher elevations they uctuated somewhat, the latter two showing a decrease at 3,200 m followed by an increase at higher elevations. The soil organic C, pH, and soil total P were the main factors inuencing P. fruticosa leaf stoichiometry. The limiting nutrients were P. slope

The soil organic C, pH, and soil total P were the main factors in uencing P. fruticosa leaf stoichiometry. The limiting nutrients were P.
Conclusions: We highlight the dependency of leaf stoichiometry on slope aspect and elevation. As P. fruticosa is a major alpine shrub in this region and plays an important role in maintaining ecological functions and services on the Qinghai-Tibetan Plateau, measures should be adopted to improve P. fruticosa growth by preventing P loss, especially at higher elevations where signi cant P losses occur due to high precipitation and sparse vegetation.
Leaf stoichiometry re ects the balance and limitations in the uptake of plant nutrients (C, N, P), thereby in uencing plants' growth rate and overall life history strategy (Baxter and Dilkes, 2012;Liu et al., 2015;Zhu et al., 2020) and global C, N, P biogeochemical cycles (Moe et al., 2005;Liu et al., 2018). The leaf stoichiometry P. fruticosa remains relatively underexplored. Leaf stoichiometry information is critical to understand nutrient cycling processes, in developing biogeochemical models, and in predicting plant responses to global changes in climate (Zhao et al., 2014;Zhao et al., 2018). Previous studies have shown that leaf stoichiometry can be affected by a wide range of edaphic, climatic and topographic factors (e.g., Cao et al., 2020) and other disturbances such as increased CO 2 and N availability (Esmeijer-Liu et al., 2009), livestock grazing and P addition (Bai et al., 2012;Scott et al., 2013).
Topography, an important factor in soil formation (Jenny, 1941), can directly (e.g., water distribution) or indirectly (e.g., microclimate) affect a plant's growing environment, particularly so in mountainous regions. However, how topography might affect soil nutrient content and leaf stoichiometry of P. fruticosa at different elevations is not well understood. Accordingly, the present study's overall objective was to examine the effects of elevation (from 2400 m to 3800 m) on leaf stoichiometry of P. fruticosa, a major alpine shrub. The study was conducted in the Qilian mountains of the Qinghai-Tibetan Plateau (QTP), the world's highest elevation plateau. Since mean annual precipitation (MAP) and mean annual temperature (MAT) and soil nutrients vary with elevation (Table 1), our hypothesis was that leaf stoichiometry of P. fruticosa would vary with elevation, and, based on Cao et al. (2020), that P would be a limiting nutrient for P. fruticosa growth. Leaf stoichiometry of P. fruticosa at different elevations At 3,500 m, the [C] leaf (402.56 g kg -1 ) was signi cantly lower than at any other elevation, while at 2600 m the [C] leaf (523.59 g kg -1 ) was signi cantly greater than at any other elevation except 3,000 m ( Fig 1A). The [N] leaf showed an increasing trend with increasing elevation. At 3,500 m, [N] leaf (27.33 g kg -1 ) was signi cantly greater than at other elevations, while [N] leaf (18.15 g kg -1 ) at 2,800 m was signi cantly lower than at any other elevation ( Fig 1B). While the [P] leaf changed slightly at elevations between 2,400 m and 3,200 m, there was a decreasing trend of [P] leaf at elevations of 3,500 m and 3,800 m, with the lowest value (0.60 g kg -1 ) recorded at 3,800 m, where it was signi cantly lower than at 2,600 m, 3,200 m or 3,500m ( Fig 1C). From 2,400 m to 3,000 m, [C] leaf : [N] leaf varied little, but was signi cantly greater than at ≥3,200 m. However, from 3,200 m to 3,800 m, [C] leaf : [N] leaf showed a decreasing trend rst and then an increasing trend, with the value at 3,500 m (14.74) being signi cantly lower than at other elevations ( Fig 1D) (Table 2).

Dominant factors in uencing leaf stoichiometry of P. fruticosaat different elevations
The eigenvalues of the rst and axes were 0.27 and 0.021, respectively, explaining about 29% of the total variation (Fig 2). Except STP, all other parameters were signi cantly related to two axes. Across elevations, SOC, STP, and pH had the greatest impact on leaf stoichiometry of P. fruticosa (Table 3). In the present study, all sampled sites were from the sunny slope aspects, but slope gradients and positions were different (data unpublished). Thus, soil properties in the study area were also affected by these micro-landforms.

Discussion
With changes of biotic and abiotic environments with elevation, leaf stoichiometry of P. fruticosa also varied with elevation ( Fig. 1), concurring with other studies (Badano et al., 2005; fruticosa showed a decreasing trend (Fig. 1A), which was in contrast with Zhao et al. (2014) and Rong et al. (2016), who found that [C] leaf increased as temperature decreased to balance the osmotic pressure of cells and resist freezing. This result may re ect that the fact that low temperatures inhibit photosynthesis in P. fruticosa. In contrast, [C] leaf and [N] leaf of P. fruticosa showed an increasing trend with a decrease in temperature (Fig. 1B), as reported by others (e.g., Oleksyn et al., 2010;Li and Sun 2016;Cao et al., 2020). This may be because The Dominant Environmental Factors In uencing Leaf Stoichiometry Of P.fruticosa Based on RDA (Fig. 2, Table 3), it is clear that SOC, STP, and pH had a greater effect on leaf stoichiometry of P. fruticosa than temperature or precipitation in the Qilian Mountains. This is in slight contradiction with others studies (Sardans et al., 2011;Cao et al., 2020); for example, Cao et al. (2020) found that, across various elevations in the Qilian Mountains, temperature signi cantly affected leaf stoichiometry of O. ochrocephala, as it could dictate or control nutrient availability in soils, root absorption, and the plant nutrient budget (Reich and Oleksyn, 2004;Isles et al., 2017;Liu et al., 2019). Likewise,  found that temperature and precipitation directly affected the spatial patterns of leaf elements across China, as precipitation regulates the mobilization of soil nutrients (Müller et ahl., 2017).
Although SOC, STP and pH were the main contributors to differences in leaf stoichiometry of P. fruticosa, STP was not related to any index of leaf stoichiometry (  (Fig. 1A).
This suggests that the relationship between [C] leaf of P. fruticosa in the Qilian Mountains may not represent a true causality. This may also suggest that the C in soil is the structural basis for plants (Schade et al., 2003;Liu et al., 2011) (Table 4). Generally, SOC and pH are negatively correlated, as acidic soil is bene cial to adsorption of organic C Hobara et al. 2016). Therefore, the relationships between pH and leaf stoichiometry of P. fruticosa were converse to relationships between SOC and leaf stoichiometry (Table 4). In the present study, the measured parameters can only explain about 30% the total variation of leaf stoichiometry of P. fruticosa (Fig. 2), indicating that other factors, such as plant community composition , may also control the variations. As the plant community in the study area changed with elevation, effects of intra-and inter-species competitions on leaf stoichiometry of P. fruticosa should also be considered to achieve a comprehensive understanding.

Limiting Nutrients For P. Fruticosa Across Elevations
It is well known that [N] leaf :[P] leaf rather than [N] leaf or [P] leaf individually, can provide a better assessment of plants' nutrient limitations (Li et al., 2018). In the present study, the [N] leaf :[P] leaf of P. fruticosa was > 16 (Fig. 1F), suggesting that by the criteria provided by Koerselman and Meuleman (1996), P was limiting to the growth of P. fruticose. Soil P de ciency is common across China (Han et al., 2005;Zhao et al., 2016), including the whole QTP (Niu et al., 2016) and the Qilian Mountains Xu et al., 2018aXu et al., , 2019Cao et al., 2020). However, Reich and Oleksyn (2004) concluded that plant growth in high elevations was more limited by N. This suggests that limitation of nutrient elements for plants is dependent on region.
Furthermore, at elevations ≥ 3,500 m, the [N]leaf:[P]leaf was > 50 (Fig. 1F), suggesting that P. fruticosa growth was greatly restricted by a lack of P. It is well known that in the Qilian Mountains, soil surface coverage by vegetation decreases as elevation increases. In combination with P leaching through the soil pro le (Chardon and Schoumans, 2007), the lack of vegetative cover at high elevations can easily increase P losses through erosion and surface run-off (Vanden Nest et al., 2014) and can make P more scarce. To reduce P loss, vegetation at these high elevations must be offered some protection.

Conclusions
In the Qilian Mountains of the QTP, elevations ranging from 2,400 m to 3,800 m affected P. fruticosa [C] leaf , [N] leaf and [C] leaf :[N] leaf mainly through the effects on SOC, STP and pH. From low elevation to high elevation, the [C] leaf and [P] leaf decreased, whereas [N] leaf increased, partly supporting the temperature-physiology hypothesis.
In the study area, P. fruticosa growth was commonly limited by soil P. Moreover, the areas at higher elevations were particularly lacking in soil P. Due to high precipitation and sparse vegetation at higher elevations, loss of soil P is greater. Accordingly, disturbances such as livestock grazing must be excluded to prevent soil erosion and reduce run-off in this area. As P. fruticosa is a major alpine shrub on the QTP, improving its growth conditions will play an important role in maintaining the ecologically integrated functions and services of the whole QTP.

Methods
Plant material and soil sampling

Study area
With a mean elevation of 4000 m (closer to 3000 m in the northeast), MAP of 400 mm, and MAT of < -4°C, the Qinghai-Tibetan Plateau covers 2.5 × 10 6 km 2 . Ranging in elevation from 2,200 to 5,500 m, and located in the northeastern portion of the QTP, the Qilian Mountains mainly present two slope aspects; south-facing and northfacing. On the south-facing slope aspects, grasslands growing on sandy-textured chestnut soils are the dominant vegetation type, while on the north-facing slope aspects, Qinghai spruce (Picea crassifolia Kom.), growing on silty-sand-textured grey cinnamon soils, is the dominant species (Qin et al., 2016).

Laboratory analyses
Using established laboratory methods, [C] leaf , [N] leaf , [P] leaf , soil organic carbon (SOC), soil total nitrogen (STN), and soil total phosphorus (STP) were measured (more details in Cao et al. (2020)), while soil water content (SWC) was determined from oven drying of samples at 105°C to a constant mass (Qin et al., 2019).
All data were expressed as mean and standard deviation (SD), and analyzed with SPSS 22.0 for Windows (SPSS, Inc., Chicago, IL). One-way ANOVA and univariate general linear model were used to explore differences in leaf stoichiometry and soil nutrients at different elevations, and the effects of elevation on soil nutrients, respectively. The Pearson Correlation Coe cient was used to determine the correlation between leaf stoichiometry and abiotic factors. Redundancy analysis (RDA) was performed to nd the dominant environmental variables in uencing leaf stoichiometry of P. fruticosa (Maccherini et al. 2011;Yang et al. 2018a) using the R 'vegan' package . The signi cance of the eigenvalues of the canonical axes was tested by a reduced Monte Carlo model with 270 unrestricted permutations (Yuan, 2017;Sun et al., 2017 Availability of data and materials The data related to this study were deposited in the my Institute.

Competing interests
The authors declare that they have no competing interests. The funder who is Wei Liu had roles in study design, data collection and analysis, and preparation of the manuscript.

Authors' contributions
Conception and design of the research: YQ; acquisition of data: XZ and YQ; analysis and interpretation of data: WL; statistical analysis: XZ; drafting the manuscript: WL, XZ and YQ; revision of manuscript for important intellectual content: JA and AB. All authors read and approved the nal manuscript.