Introduction

Nitrogen (N) cycling has been dramatically changed due to anthropogenic activities such as the combustion of fossil fuels and the use of agricultural fertilizers, with N inputs to terrestrial ecosystems doubled in the past century1,2. Previous study showed that the recent rate of N deposition has been 3–5 folds higher than the last century3. In addition, N deposition over the continental ecosystems would vary between 60 and 100 Tg N yr−1 by 21004. Available records of δ15N in soil sediments suggest that terrestrial ecosystems can continuously absorb reactive N released by human activities5. However, we still do not know how much N can be retained in terrestrial ecosystems and its underlying mechanisms5.

Many changes in the structure, function and stability of ecosystem may occur when the N inputs into terrestrial ecosystems increase6,7. As a primary nutrient for plant growth in many terrestrial ecosystems8, N fertilizers have been widely used to enhance the productivity of farmlands and pasturelands. However, due to the different responses to increasing N among individual species and communities determined by their inherently different N use efficiency and strategies6, experiments have indicated that increasing N input can lead to cascading effects such as long-term species composition change and biodiversity loss9,10. As a result, a better understanding of biodiversity, species, and community structure is critical for further ecosystem management and protection under the background of increasing N input8.

Changes in ecological conditions caused by N deposition may exert an important influence on ecosystem properties and biogeochemical cycling. For instance, N deposition promotes the uptake and assimilation of atmospheric CO2 by plants, accelerates the release of greenhouse gases, and affects nitrification and denitrification11,12,13. As defined by nutrient limitation, an ecosystem is limited if N addition leading to the increase of biomass or the rate of biological processes14. Therefore, under the circumstance of N limitation, exogenous N input can stimulate plant growth, increase net primary productivity and then promote C and N accumulation11,15. On the contrary, when exogenous N exceeds the capacity for plants, soil and microbes to uptake and retention, the negative effects of redundant N are obvious: photosynthesis and productivity decline and then plants stop growing or tend to die16,17. The N saturation point is defined as the state of an ecosystem when the supply of reactive N (mainly ammonia and nitrate) in the environment exceeds the demand of plants and soil microorganisms16,18, and critical rates are the amount of fertilizer input at this point19,20. Assessing the critical N rate (Ncr) at saturation point is an important goal for land managers19.

Although N limitation is a widespread phenomenon in terrestrial ecosystems11, the knowledge about limitation of grassland ecosystems is unclear, so a better understanding of N limitation and the maximum N retention and C gain at Ncr is essential to our comprehension of ecosystem N and C cycle. Many indicators are used to assess the status of N limitation, such as δ15N of plants and soil, foliar N:P ratios and plant N stress index, etc.14,17. All of them have advantages and limitations, and cannot evaluate N limitation of exogenous N addition effectively21. Fertilization is a key measurement to restore the ecological functions of degraded grasslands22. While previous work has often focused on the response of aboveground net primary productivity (ANPP) to N addition and discussed the issue of ecosystem N saturation threshold17,23,24, little attention has been paid to the response of plant N retention to N addition. Therefore, fertilization experiments are still widely regarded as a method to assess N limitation through observing whether plants NPP increase and how much C and N can be accumulated in the ecosystem11,21.

Known as the third pole of the earth, the Tibetan Plateau (TP) covers an area of about 2.57 million km2 with more than 60% (~ 1.54 million km2) covered by alpine grassland25 and most grassland is above 4000 m above sea level26. However, due to the particularity of its spatial characteristics, Tibetan alpine ecosystems are much more sensitive to climate change and are known as the “pre-warning region”, playing a vital role in not just Asia but also in global climate behavior27. Under the influence of climate change and long-term grazing, TP alpine grasslands are facing degradation28,29. The plant and soil N isotope measurement indicated that N limitation in this region might be aggravated30. The TP is now experiencing an intensely increase in N deposition rates with the range from 0.32 to 1.23 g N m−2 yr−131,32 and N fertilization has also been applied to restore degraded grasslands and promote grass productivity for grazing33,34,35. These situations significantly affect the biochemical processes of Tibetan alpine ecosystems36. Given the uncertainty of future N deposition, we do not know how much exogenous N can be retained by plants in alpine grasslands and how much N input (Ncr) is required to reach the saturation point (Fig. 1).

Figure 1
figure 1

Patterns of plant aboveground net primary productivity (ANPP), N retention capacity and N retention fraction (NRF) response to N addition. ANPPck is the initial ANPP without exogenous N addition. ANPPmax is the saturated ANPP at critical N rate (Ncr). MNRC is the maximum N retention capacity at Ncr.

Full size image

We assumed the greater N limitation, the greater retention of external N in the plant. To understand the impact of elevating N deposition on alpine grasslands community composition and quantify N limitation, we conducted a 6-year field experiment with six levels of N addition (0, 1, 2, 4, 8 and 16 g N m−2 yr−1) at an alpine steppe site on the TP. We also synthesized 89 (Appendix S2: Fig. S1) experiments with at least four N input levels in worldwide grasslands for comparison, in which 27 sites were located on the TP. The aim of this study was to quantify how much exogenous N can be retained in the plant pool of alpine grassland.

Methods

Study site

The field experiment was conducted at Namco Station (30°47’N, 90°58’E, altitude 4730 m) of the Institute of Tibetan Plateau Research, Chinese Academy of Sciences (ITPCAS), which is located in the alpine steppes of TP in China. The experiment was permitted by ITPCAS, complied with local and national guidelines and regulations. From 2006 to 2017, the mean annual temperature (MAT) and mean annual precipitation (MAP) was about − 0.6 °C and 406 mm, respectively. Monthly mean temperature varied from − 10.8 °C in January to 9.1 °C in July and most of the precipitation occurred from May to October37,38. During our six-year observations (2010, 2011, 2012, 2013, 2015 and 2017), climate change during the growing season from May to September varied differently, with the annual precipitation ranged from 255.9 mm to 493.8 mm and the MAT from 6.7 to 7.4 °C. Androsace tapete, Kobresia pygmaea, Stipa purpurea and Leontopodium pusillum were the dominant plant species at the alpine steppe.

Experimental design and treatments

The long-term experiment began in May, 2010. Three homogenous plots were randomly arranged as replicates at the alpine steppe and six subplots (~ 13 m2) were distributed in each plot by a cycle, with a 2 m buffer zone between each adjacent subplot (Appendix S1: Fig. S1). In this experiment, six treatments of N fertilization rate (0, 1, 2, 4, 8, and 16 g N m−2 yr−1) were clockwise applied in each subplot. The subplots of 0 g N m−2 yr−1 were control group. We sprayed NH4NO3 solution on the first day of each month in the growing season (from May to September) each year. After fertilizing, we rinsed plant residual fertilizer with a little deionized water (no more than 2 mm rainfall). For the control groups, we added equivalent amount of water. The experiment was conducted from 2010 to 2017 (it should be pointed out that there was no fertilization in 2014 and 2016).

Sampling and measurements

The samples were collected with the training and permission of ITPCAS and involved plants that are common species and not endangered or protected. The identification of the plants was done by referring to a book of Chen and Yang39. Pictures of the corresponding specimens can be seen on the website of ITPCAS (http://itpcas.cas.cn/kxcb/kxtp/nmc_normal_plant/).

Vegetation samples were collected in August in 2011 and repeated at the same time in 2012, 2013, 2015 and 2017. We established one 50 × 50 cm quadrat in each subplot, clipped aboveground biomass (AGB) and sorted species by families. The biomass was used to measure ANPP (g m−2 yr−1). Following aboveground portion collected, we used three soil cores (5 cm diameter) to collect the belowground roots at 0–30 cm depth and mixed into one sample, which were used to assess belowground net primary productivity (BNPP, g m−2 yr−1). The roots were cleaned with running water to remove sand and stones.

Both plant and root samples were dried at 75 °C for 48 h and then ground into powder (particle size ~ 5 μm) by a laboratory mixer mill (MM400, Retsch). To determine N and C content of plants, we weighed the samples into tin capsules and measured with the elemental analyzer (MAT253, Finnigan MAT GmbH, Germany).

Estimation of the critical N rate (Ncr), N retention fraction (NRF), N retention capacity and N-induced C gain

According to the N saturation hypothesis, plant productivity increases gradually during N addition, reaches a maximum at the Ncr, and eventually declines16,17. We considered the Ncr to be the rate where ANPP no longer remarkably changed with N addition (Fig. 1).

We defined plant N retention fraction (NRF, %; Eq. 1) as the aboveground N storage caused by unit N addition rate, and N retention capacity (g N m−2 yr−1; Eq. 2) was the increment of N storage due to exogenous N addition compared to the control40. The equations are as following:

$$N\;retention\;fraction = \frac{{ANPP_{tr} \times N\;content_{tr} - ANPP_{ck} \times N\;content_{ck} }}{N\;rate}$$
(1)
$$N\;retention\;capacity = ANPP_{tr} \times N\;content_{tr} - ANPP_{ck} \times N\;content_{ck}$$
(2)

where ANPPtr and N contenttr (%) refer to those in the treatment (tr) groups, and ANPPck and N contentck refer to those in the control (ck) groups. These expressions are also used in the following equations (Eqs. 3–5).

The N-induced C gain (g C m−2 yr−1; Eq. 3) was estimated by the increment of C storage owing to exogenous N addition compared to the control40. Maximum N retention capacity (MNRC, Eq. 4) and maximum N-induced C gain (Eq. 5) mean the maximum N and C storage increment in plant caused by exogenous N input at Ncr, respectively. The formulas are as following:

$$N{\text{-}}induced\;C\;gain = ANPP_{tr} \times C\;content_{tr} - ANPP_{ck} \times C\;content_{ck}$$
(3)
$$MNRC = ANPP_{\max } \times N\;content_{\max } - ANPP_{ck} \times N\;content_{ck}$$
(4)
$$Maximum\;N{\text{-}}induced\;C\;gain = ANPP_{\max } \times C\;content_{\max } - ANPP_{ck} \times C\;content_{ck}$$
(5)

where ANPPmax, N contentmax and C contentmax refer to the value of ANPP, N content and C content at Ncr, respectively.

Data synthesis

To evaluate N limitation and saturation on the TP more accurately, we searched papers from the Web of Science (https://www.webofscience.com) and the China National Knowledge Infrastructure (https://www.cnki.net). The keywords used by article searching were: (a) N addition, N deposition or N fertilization, (b) grassland, steppe or meadow. Article selection was based on the following conditions. First, the experimental site must be conducted in a grassland ecosystem. Second, the experiment had at least three N addition levels and a control group. Third, if the experiment lasted for many years, we analyzed data with multi-year average. Based on the above, we collected 89 independent experimental cases. Among these, 27 cases were located on the TP alpine grasslands, 25 in the Inner Mongolia (IM) grasslands and 37 in other terrestrial grasslands (detailed information sees Appendix S2: Table S1).

We extracted ANPP data and N addition rate of these cases and estimated Ncr and ANPPmax (Appendix S2: Fig. S2). We then calculated NRF, N retention and C gain of each group of data for further analysis (Appendix S2: Table S2). Most of the 89 cases did not have data on N and C content. To facilitate the calculation, we summarized N and C content from 40 articles in the neighboring areas of the cases and divided the N and C content into seven intervals according to the N addition rate (Appendix S2: Table S3 and Fig. S3). The unit of N addition rate was unified to “g N m−2 yr−1”. All the original data were obtained directly from texts and tables of published papers. If the data were displayed only in graphs, Getdata 2.20 was used to digitize the numerical data. For the estimation of N retention and C gain of the TP at current N deposition rates and future at Ncr, we fitted the exponential relationship to the data from 27 cases on the TP, and then substituted N rates into the fitted equations (Eq. 6):

$$y = a \times \left[ {1 - \exp \left( { - bx} \right)} \right].$$
(6)

We also included MAT, MAP, soil C:N ratio, fencing management (fencing or grazing) and grassland type (meadow, steppe and desert steppe) of the experiment sites for exploring the drivers affecting N limitation (Appendix S2: Table S1). When climatic data were missing from the article, MAT and MAP were obtained from the WorldClim (http://www.worldclim.org).

Species were usually divided into four functional groups (grasses, sedges, legumes and forbs) to study the response of species composition to N addition in previous study41. We synthesized 13 TP experimental cases (including our field experiment) from the data synthesis and each case included at least three functional groups (detailed references see Appendix S2).

Statistical analysis

There were 42 species in our field experiment. We divided them by family into eleven groups: Asteraceae (forbs), Poaceae (grasses), Leguminosae (legumes), Rosaceae (forbs), Boraginaceae (forbs), Caryophyllaceae (forbs), Cyperaceae (sedges), Labiatae (forbs), Primulaceae (forbs), Scrophulariaceae (forbs) and Others. Due to species in the group of Others contributed only 1.22% of AGB, we analyzed AGB and foliar stoichiometry among other ten families (Appendix S1: Table S1). In Namco steppe, forbs, grasses, sedges and legumes accounted for 78.0%, 7.4%, 8.2% and 5.2% of the AGB respectively (Appendix S1: Table S1 and Fig. S2). Such a large number of forbs suggested that our experiment was conducted on a severely degraded grassland.

For our field data, two-way ANOVAs were used to analyze the effects of year, N fertilization rate and their interactions on species AGB. One-way ANOVAs were used to test the response of ANPP, BNPP, root:shoot ratio, species foliar C content, N content and C:N ratio to N addition rate. Duncan’s new multiple range test was used to compare the fertilization influences at each rate in these ANOVAs. Prior to the above ANOVAs, we performed homogeneity of variance test and transformed the data logarithmically when necessary. Simple regression was used to estimate the relevance among ANPP, NRF, N retention capacity and C gain with N addition rates.

Structural equation modeling (SEM) was used to explore complex relationships among multiple variables. To quantify the contribution of drivers such as climate and soil to Ncr, ANPP, NRF and MNRC, we constructed SEM based on existing ecological knowledge and the possible relationships between variables. We considered environmental factors (MAT, MAP and soil C:N) and ANPPck as explanatory variables, and Ncr, NRF and MNRC as response variables. We included the ANPPck in the SEM rather than the ANPPmax because we wonder whether there was a relationship between ANPP in the absence of exogenous N input and the ecosystem N retention in the presence of N saturation. This has important implications for assessing N input. Before constructing the SEM, we excluded collinearity between the factors. In addition, Student’s t-test and one-way ANOVAs were performed to explain the effect of fencing management and grassland type on above response variables, respectively. The SEM was constructed using the R package “piecewiseSEM”42. Fisher’s C was used to assess the goodness-of-model fit, and AIC was for model comparison.

Given the influence of extreme values in the data synthesis, we calculated the geometric mean of Ncr, NRF, N retention and N-induced C gain. All statistical analyses were performed with SPSS 26.0 and RStudio (Version 1.2.1335) based on R version 3.6.2 (R Core Team, 2019).

Results

Species composition and foliar C:N ratio response to N addition

In our field experiment, the response of biomass to N addition differed between all the ten families, and these differences varied greatly across years (Appendix S1: Table S2). Asteraceae (forbs) occupied the largest proportion of biomass (28.89%). AGB of it typically increased with N addition and then decreased, reaching its maximum at the Ncr of 4 g N m−2 yr−1 (Fig. 2a).

Figure 2
figure 2

Aboveground biomass (AGB) of species from ten families in response to long term N addition rate. Biomass was collected from 2011 to 2013, 2015, and 2017 in Namco. Data are shown as mean ± SE (n = 3). Bars with different letters indicate significant differences (P < 0.05) by Duncan’s new multiple range test among fertilization levels. Full statistical results for two-way ANOVA (N, N fertilization; Y, year) are presented in Appendix S1: Table S2.

Full size image

The proportion of some forbs disappeared with some grasses increased with N addition. Primulaceae (forbs) in some treatment groups began to vanish since 2013 (Fig. 2b). Rosaceae (forbs) significantly increased in 2011 and decreased in 2015 (Fig. 2d). Caryophyllaceae (forbs), Boraginaceae (forbs) and Scrophulariaceae (forbs) had all been nearly absent since 2012 (Fig. 2h-j). Poaceae (grasses) increased with N addition and no clear Ncr was found even under the maximum N fertilization level in 2011 (Fig. 2e). For Cyperaceae (sedges) and Leguminosae (legumes), we only observed an upward trend in 2011, but they showed little change or even a downward trend in the following years (Fig. 2f,g).

N addition significantly influenced foliar stoichiometry (Appendix S1: Table S3). Foliar N content was enhanced significantly with N addition (F5,138 = 9.07, P < 0.001, Appendix S1: Fig. S3b) while C content did not change obviously (F5,138 = 0.80, P = 0.551; Appendix S1: Fig. S3a). As to species C:N ratios, high N addition led to lower C:N ratio (F5,138 = 7.78, P < 0.001; Appendix S1: Fig. S3c), particularly for Cyperaceae, Asteraceae, Rosaceae and Labiatae, while Leguminosae remained stable. Moreover, C:N ratio of Primulaceae was the largest in our experiment (Appendix S1: Fig. S3c).

ANPP, BNPP and the root:shoot ratio response to N addition

ANPP showed an increasing trend under the influence of N addition during the field experiment, with significant difference between 16 g N m−2 yr−1 N addition and the control in 2011 and 2015 (Appendix S1: Fig. S4a). Compared to the control, N addition clearly enhanced ANPP by 158.7% (29.8–77.1 g m−2 yr−1) and 105.5% (41.8–85.9 g m−2 yr−1) in 2011 and 2015, respectively. The trend that ANPP increased with N addition was consistent with the results demonstrated in the data synthesis which exhibited that ANPP saturated at Ncr in most grasslands (Appendix S2: Fig. S2). We also found a significant linear correlation between ANPPmax and initial ANPP (ANPPck). The ANPPmax was about 1.67 times higher than that in ANPPck (R2 = 0.97, F1,88 = 2427.58, P < 0.001; Fig. 3a and Appendix S2: Table S4). The BNPP had no significant change in response to N addition compared to ANPP, the ratio of root to shoot showed a reduce trend in 2011 and 2015 but an increased trend in 2013 and 2017 (Appendix S1: Fig. S4).

Figure 3
figure 3

The relationship between the (a) maximum aboveground net primary productivity (ANPPmax) and initial aboveground net primary productivity (ANPPck), (b) critical N rate (Ncr) and mean annual precipitation (MAP), (c) maximum N retention capacity (MNRC) and ANPPck, and (d) maximum N-induced C gain and ANPPck. The red points represent Tibetan Plateau (TP) sampling plots, green points represent Inner Mongolia (IM) sampling plots, and blue points represent other global grassland sampling plots. Simple regression analyses (Pearson) were implemented. Full regression results are shown in Appendix S2: Table S4. The gray areas indicate the 95% confidence interval.

Full size image

Spatial variations of NRF

For Namco alpine steppe (our study site), the maximum NRF was 37.16 ± 6.21% at the level of 1 g N m−2 yr−1, and then decreased with N addition rate until stabilized at 15.75 ± 5.29% at the Ncr of 4 g N m−2 yr−1 (Appendix S1: Fig. S5).

At a larger scale of TP alpine grassland and IM grassland, the NRF showed a trend of decreasing before reaching Ncr (Appendix S2: Fig. S4c,d). NRF at Ncr was generally higher in the eastern TP (e.g., Haibei, Maqu and Lanzhou) of the alpine grasslands than in the western TP (e.g., Damxung, Gerze, Nagqu, Namco and Nyima) (t = 4.85, df = 25, P < 0.001, Appendix S2: Fig. S4c). Grassland in the eastern TP retained nearly 76.08% of the N (NRF range: 13.30–313.86%) and the western TP retained nearly 13.91% (NRF range: 3.18–36.88%; Table 1). In addition, NRF at Ncr in IM (mean: 34.91%) was similar to that in the whole TP (mean: 40.54%; Appendix S2: Fig. S4c,d).

Table 1 Global differences in critical N rate (Ncr), N retention fraction (NRF) at Ncr, maximum N retention capacity (MNRC) and maximum N-induced C gain.
Full size table

NRF at Ncr varied considerably across the global grasslands, with an average NRF of 39.30% at an average Ncr of 9.79 g N m−2 yr−1 (F7,66 = 6.83, P < 0.001; Table 1). In the SEM, factors explained 49% of variance in NRF (Appendix S2: Fig. S5). Among them, initial ANPP and soil C:N were the strongest drivers of NRF, and MAT had a direct positive effect on NRF. In the simple regressions, NRF exhibited a positive linear relationship with initial ANPP (R2 = 0.54, F1,88 = 103.81, P < 0.001) and a negative non-linear relationship with soil C:N (R2 = 0.40, F2,78 = 26.21, P < 0.001) (Appendix S2: Table S4 and Fig. S6a,b).

Spatial variations of MNRC

In Namco steppe, N retention rapidly increased to 0.63 ± 0.21 g N m−2 yr−1 at the Ncr of 4 g N m−2 yr−1, after which this increasing trend became slower (Fig. 4a). Like ANPP, the N retention of plant exhibited a positive nonlinear response with N addition until the input N exceeded plant demand (Appendix S2: Fig. S7).

Figure 4
figure 4

Effects of N addition on N retention capacity and N-induced C gain. (a, c) N retention and C gain of Namco field experiment in 2011 (circle points) and 2015 (triangle points). (b, d) N retention and C gain of Tibetan Plateau (TP) experiments in the data synthesis. The left axes indicate the annual N retention and C gain per unit area of the TP. The right axes indicate the total N retention and C gain of the TP per year (based on 1.54 million km2 of grassland on the TP). Single regression analyses were implemented. Full regression results are shown in Appendix S2: Table S7. The red dashed line indicates N retention and C gain of Namco and the entire TP grassland at the current rate of N deposition (0.32–1.23 g N m−2 yr−1). The blue dashed line indicates N retention and C gain under the future N saturation rate (at the Ncr of 5.53 g N m−2 yr−1).

Full size image

Notably, the MNRC in Namco steppe was much lower than that of grasslands elsewhere in the world (Fig. 5a). We also observed that MNRC of eastern TP (e.g., Haibei, Maqu, and Lanzhou; mean: 4.10 g N m−2 yr−1) at elevations of 3000–3500 m was generally higher than that of the western TP (e.g., Bangoin, Damxung, Gerze, Nagqu, Namco and Nyima; mean: 0.80 g N m−2 yr−1) at elevations greater than 4000 m (Fig. 6a), and IM had a similar MNRC (mean: 3.92 g N m−2 yr−1; Fig. 6b) as the eastern TP (e.g., Haibei, Maqu and Lanzhou).

Figure 5
figure 5

Eight examples showing the response of (a) N retention capacity and (b) N-induced C gain to N addition rate. USA, America; NL, Netherlands; CN, China; AU, Australia. Each colored line indicates the regression line fitted with the exponential function. Full regression results are shown in Appendix S2: Table S8.

Full size image
Figure 6
figure 6

The response of N retention capacity and N-induced C gain in Tibetan Plateau (TP) alpine grassland and Inner Mongolia (IM) grassland. The left panels represent TP sample plots and the right panels represent IM sample plots. Single regression analyses were implemented separately for the western TP (Xizang, n = 49), eastern TP (Qinghai, n = 36; Gansu, n = 29; and Sichuan, n = 10) and four regions in IM (Xilingol, n = 95; Hulun Buir, n = 11; Ordos, n = 7; and Hohhot, n = 30). Full regression results are shown in Appendix S2: Table S9. Points of each color represent a sampling plot and fitting lines of each color represent a region.

Full size image

For global grasslands, under elevated N addition, the MNRC in aboveground biomass ranged from 0.35 to 42.59 g N m−2 yr−1, with an average of 3.85 g N m−2 yr−1 (Table 1). Compared to the eastern America, Wageningen and western Australia, TP had a lower MNRC (mean: 2.24 g N m−2 yr−1) (Table 1). The SEM explained 69% (Appendix S2: Fig. S5) of variances in MNRC. Both ANPPck and Ncr had strong positive effects on MNRC. ANPPck had a positive linear relationship with MNRC (R2 = 0.78, F1,88 = 320.11, P < 0.001) (Fig. 3c and Appendix S2: Table S4).

Spatial variations of N-induced C gain

The response of N-induced C gain to N addition was similar to that of N retention capacity (Fig. 4). In Namco steppe, C gain increased to 7.81 ± 5.76 g C m−2 yr−1 at the Ncr of 4 g N m−2 yr−1, and then gradually leveled off (Fig. 4c). Similarly, C gain of Namco was relatively low in the world (Fig. 5b). The maximum N-induced C gain of global grassland ranged from 4.86 to 583.73 g C m−2 yr−1, with an average of 56.93 g C m−2 yr−1 (Table 1). Eastern TP (mean: 53.27 g C m−2 yr−1) had a higher maximum C gain than the western TP (mean: 12.73 g C m−2 yr−1), and the IM (mean: 51.49 g C m−2 yr−1) was similar to that in the eastern TP (Fig. 6c,d; Table 1). Like MNRC, ANPPck also showed a positive linear relationship with the maximum N-induced C gain (R2 = 0.82, F1,88 = 388.59, P < 0.001) (Fig. 3d and Appendix S2: Table S4).

Spatial variations and regulating factors of Ncr

For global grassland, Ncr ranged from 1 to 61.3 g N m−2 yr−1, with an average of 9.79 g N m−2 yr−1 (Appendix S2: Table S2). There was a significant difference of Ncr on global scale (F7,66 = 6.38, P < 0.001; Table 1). For example, the mean Ncr of TP was 5.53 g N m−2 yr−1, which was significantly lower than that of eastern America, Wageningen and western Australia (Table 1). Environmental factors significantly influenced the variation of Ncr. In the SEM, all factors could explain 43% of variance in Ncr (Appendix S2: Fig. S5). In particular, MAP could positively regulate the Ncr directly or indirectly through soil C:N (Appendix S2: Fig. S5). Ncr demonstrated positive linear correlations with MAP (R2 = 0.73, F1,88 = 239.80, P < 0.001; Fig. 3b) and soil C:N (R2 = 0.65, F1,80 = 147.71, P < 0.001; Appendix S2: Table S4 and Fig. S6c) in the simple regressions.

Effects of fencing management and grassland types on Ncr, NRF and MNRC

In our data synthesis, 54 experiments were carried out under the enclosure and the other 35 cases were carried out under grazing (Appendix S2: Table S1). The Ncr of enclosure (mean: 8.06 g N m−2 yr−1) was lower than grazing (mean: 13.21 g N m−2 yr−1; Appendix S2: Table S5). Fencing management only reduced Ncr (t = − 2.69, df = 87, P = 0.01) and had no statistically significant effect on plant NRF (t = 1.54, df = 86.98, P = 0.13) or MNRC (t = − 1.04, df = 87, P = 0.30) (Appendix S2: Table S5).

Grassland types (meadow, steppe and desert steppe) significantly influenced Ncr (F2,81 = 4.72, P = 0.01). Ncr for meadow, steppe and desert steppe were 6.19, 11.65 and 10.00 g N m−2 yr−1, respectively (Appendix S2: Table S6). NRF was marginally significant (F2,81 = 2.73, P = 0.07) across the different grassland types, with the highest NRF (mean: 47.19%) in meadow, followed by steppe (mean: 35.20%) and desert steppe (mean: 17.90%; Appendix S2: Table S6).

Discussion

N addition altered species composition in alpine grassland

Combining our field experiment with 13 multi-level studies conducted on the Tibetan, we found that functional groups respond similarly to N addition across alpine grasslands: N addition reduced the AGB proportion of forbs (Appendix S2: Fig. S8c), which is similar to Zong, et al.43. N addition increased AGB proportion of grasses except sedges (Appendix S2: Fig. S8a,b), consistent with a meta-analysis by Wang, et al.35. For legumes, N addition reduced their AGB proportion (Appendix S2: Fig. S8d), which is also demonstrated by Huang, Liu and Zhou44 and Xu, et al.45.

The above results indicated that the resource allocation strategies differ between species after fertilization46. Several reasons may account for these changes. First, ecosystems constrained by N may reach N saturation and plant growth will convert soil N competition to light competition under eutrophication conditions47. The dominant species (such as Asteraceae) and grasses (Poaceae) are competitive in light competition due to their higher stems and dense leaves, thus they could grow faster under conditions with sufficient N and water supply. The small and slow-growing forbs (such as Boraginaceae) and sedges (Cyperaceae) are weak in light competition because they are at the lower layer in the vertical structure of the community. Therefore, even if short-term N addition could promote their growth, competition on light may lead to inferior species to decrease sharply or disappear eventually48,49. Second, legumes can promote plant growth through N fixation with rhizobia under low N level. At high N level, this ability may be inhibited and their inherent advantage over other non-legume species may be diminished50,51, but it is beneficial to the growth and reproduction of grasses52.

In our field experiment, before and after the Ncr, biomass of Asteraceae changed obviously (Fig. 2a). In addition to excessive N inhibiting the growth of Asteraceae, this change may also be related to the stability of plant. The study of Huang, Liu and Zhou44 reported that N addition significantly reduced the stability of Asteraceae, but had no effect on other species. This suggests that Asteraceae should be noted when studying the effects of N addition on plants, as they may be able to indicate the N saturation points in alpine grassland.

Our synthesis revealed that ANPP first increased with N addition and then saturated at Ncr (Appendix S2: Fig. S2). This could be illustrated as follows: N addition improves soil N availability through alleviating N limitation and thus boosts plant growth33,53, but excess N inputs to the soil may shift limitation factor from N to light54, so the ANPP saturated at Ncr. Moreover, we found the ratio of ANPPmax to initial ANPP was approximately 1.67 (the ratio in TP grasslands was about 1.45; Fig. 3a) and this coefficient was greater than the study of LeBauer and Treseder11 (1.29). This may be because LeBauer and Treseder11 chose ANPP at the highest rate of N addition while we chose ANPP at Ncr, and the highest rate reduces ANPP due to oversaturation which inhibits plant growth.

N addition increased N retention and C gain until reached Ncr

Not all the exogenous N could be retained in the plant pool. NRF would decrease with elevated N and stabilized at Ncr. A global average of less than half of N addition can be retained by plants (NRF: 39.30%; Table 1). At high altitudes on the Tibetan Plateau, N retention capacity is even lower (NRF: 13.91%; Table 1). The Ncr is considered to be the lowest N input that leads a major harmful effect on the sensitive ecological indicator, and it is essential to determine the Ncr for alpine grasslands in order to provide an early warning before harmful effects occur in these ecosystems19,43 (Fig. 1). When the N addition level is lower than local Ncr, the ANPP and N retention in the plant will increase with N addition, along with C accumulation. Once the N addition level is higher than the local Ncr, plant N retention and C gain will reach its maximum value. In previous study, for the alpine steppe, when the N addition level is higher than Ncr level (which is about 4 g N m−2 yr−1), ANPP will reach a maximum, and ecosystems will be saturated, inorganic N will be accumulated38,55, the C and N mineralization rate will reach the maximum12, the ecosystems function might even turn from a C sink to source12. Therefore, it is quite crucial to constrain the reasonable N application level in the grassland restoration management to avoid the soil N pollution and keeping the C sink function for the alpine grassland.

In this study, given the large spatial variability of Ncr and MNRC, we found Ncr was highly correlated with MAP. This indicates the wetter the environment, the greater Ncr, which is similar to the previous study of Peng, Chen and Yang17. While the initial ANPP is a reliable predictor for MNRC. The significant linear correlation between ANPPck and MNRC suggests that we might predict the MNRC of local grassland by observing local ANPP (Fig. 3c).

Fencing management and grassland types influenced Ncr, NRF and MNRC

Fencing and grazing may affect belowground C-N cycling and thus the uptake of exogenous N by plants56. Fencing management is often used to restore degraded grasslands57. In our study, we found that the Ncr was significantly lower in fencing grassland than in grazing grassland (Appendix S2: Table S5), indicating that grazing grassland might need more N to reach its saturation. In the case of fencing, N was accumulated and recycled in the plant-soil system because no biomass was harvested and relatively little exogenous N was required56. For the grazing sample plots, the plants consumed by livestock were comparable to the biomass harvested, so N did not accumulate in the soil and more N was required for the plants to reach saturation of ANPP35,57.

Although Peng, Chen and Yang17 also compared the Ncr of biomass with harvest and without harvest, there was no significant differences in Ncr between the above two. The distinct results between our work and Peng, Chen and Yang17 can be explained in two aspects: First, for some experimental cases of fencing in the growing season and grazing in non-growing season, we treated them as the fencing sample plots, while Peng, Chen and Yang17 considered them as the grazing sample plots. We applied our own criteria to the case of Peng, Chen and Yang17 and found a marginal significance (t = − 1.91, df = 53, P = 0.06) between fencing and grazing. Consequently, the inconsistency of criteria for judging the experimental cases may have an account for the inconsistent results. Second, we collected 89 cases and Peng, Chen and Yang17 collected 55 cases, the larger amount of data may have caused the difference between fencing and grazing.

Grassland type also potentially influenced Ncr. Previous works have shown that the Ncr of a semiarid grassland was 9.17 g N m−2 yr−158 and a temperate grassland was 10.5 g N m−2 yr−124. Our data synthesis also revealed that the steppe had a higher Ncr (11.65 g N m−2 yr−1) and MNRC (4.10 g N m−2 yr−1) than meadow (Ncr: 6.19 g N m−2 yr−1, MNRC: 2.92 g N m−2 yr−1; Appendix S2: Table S6). This suggests that steppe might require more N to reach its N saturation when compared to meadow and desert steppe, steppe had the highest MNRC, which was mainly attributed to its highest ANPP (Appendix S2: Table S6). From another perspective, most of the meadow sites were located on the Tibetan Plateau, where the relatively low precipitation further reduced the Ncr (Fig. 3b).

Implications

Under N limitation, N addition promoted the function of C and N absorption in grassland11,15. N application exceeding grassland carrying capacity will reduce grassland biodiversity and may reverse grassland C sink function12,16, alter species composition, groundwater resources and threaten ecological security20. Therefore, a better understanding of N retention capacity in plant pool and reasonable assessment of Ncr is important for ecosystem.

Based on a six-year multi-level (0, 1, 2, 4, 8 and 16 g N m−2 yr−1) N fertilization experiment and a data synthesis included 89 cases of N addition experiments, plant N retention showed a universal nonlinear pattern with N addition. Although N addition increased ANPP and N retention, the increasing trend disappeared at Ncr. The fixed ratio of ANPPmax to ANPPck (about 1.67) indicated that the N retention of grassland does not increase linearly with N input (Fig. 3a). When external N addition exceeded Ncr, the exogenous N stored by plants reached its maximum. The positive MNRC-ANPPck linear relationship revealed that the MNRC might be directly predicted by initial ANPP (Fig. 3c). Compared with other grassland in the world, Tibetan alpine grassland had the lowest MNRC and Ncr, implying the lowest N deficiency (Fig. 5; Table1).

Given the current level of atmospheric N deposition on the TP (0.32 to 1.23 g N m−2 yr−1)31,32, the N retention and C gain of Namco steppe ranged from 0.07 to 0.26 g N m−2 yr−1 (Fig. 4a) and 1.00 to 3.66 g C m−2 yr−1 (Fig. 4c), respectively. For the entire TP grassland covering about 1.54 million km2, TP alpine grasslands could retain 0.33–1.21 Tg N yr−1 (0.22–0.79 g N m−2 yr−1; Fig. 4b) and gain 5.65–20.11 Tg C yr−1 (3.67–13.06 g C m−2 yr−1; Fig. 4d). With the aggravation of N deposition, the capacity of N and C absorption over the TP grasslands will increase with the increase of ANPP until it reaches the Ncr. Considering the average Ncr of the whole TP was 5.53 g N m−2 yr−1, N retention and C gain of the TP grassland will increase to 4.41 Tg N yr−1 (2.86 g N m−2 yr−1; Fig. 4b) and 64.39 Tg C yr−1 (41.81 g C m−2 yr−1; Fig. 4d) in the future.

The key parameters obtained from this study include the ratio of ANPPmax to ANPPck, Ncr, NRF, MNRC, maximum N-induced C gain and their climate driven factors. These parameters are crucial to constrain the C-N coupling model for further constructing robust predictions of N input on ecosystem C balance and find out the optimal N application rate for N resource management in the restoration of degraded grasslands on the TP.