Skip to main content
NCBI home page
Proceedings of the Royal Society B: Biological Sciences logoLink to Proceedings of the Royal Society B: Biological Sciences
. 2024 Jul 24;291(2027):20240206. doi: 10.1098/rspb.2024.0206

Global leaf sulfur stoichiometry and the relationships with nitrogen and phosphorus: phylogeny, growth form and environmental controls

Liangjian Zhang 1,†, Zhenjun Zuo 1,†, Xiujuan Qiao 2,✉, Yixuan Liu 3,4, Rui Qu 1, Haocun Zhao 1, Youxin Wang 3, Peidong Zhao 1, Lin Zhang 5, Zhigang Wu 6, Zhong Wang 1,3,4,✉
PMCID: PMC11265873  PMID: 39043235

Abstract

Sulfur (S) is an essential bioelement with vital roles in serving regulatory and catalytic functions and tightly coupled with N and P in plants. However, globally stoichiometric patterns of leaf S and its relationships to leaf N and P are less well studied. We compiled 31 939 records of leaf-based data for 2600 plant species across 6652 sites worldwide. All plant species were divided into different phylogenetic taxa and growth forms. Standard major axis analysis was employed to fit the bivariate element relationships. A phylogenetic linear mixed-effect model and a multiple-regression model were used to partition the variations of bioelements into phylogeny and environments, and then to estimate the importance of environmental variables. Global geometric mean leaf S, N and P concentrations were 1.44, 15.70 and 1.27 mg g−1, respectively, with significant differences among plant groups. Leaf S–N–P positively correlated with each other, ignoring plant groups. The scaling exponents of LN–LS, LP–LS and LN–LP were 0.64, 0.76 and 0.79, respectively, for all species, but differed among plant groups. Both phylogeny and environments regulated the bioelements. The variability, rather than mean temperature, controlled the bioelements. Phylogeny explained more for the concentrations of all the three bioelements than environments, of which S was the one most affected by phylogenetic taxa.

Keywords: climatic variability, ecological stoichiometry, phylogenetic signal, scaling relationship, spatial variation, stress tolerance

1. Introduction

Ecological stoichiometry focuses on the scaling relations of bioelements in organisms and the patterns along environmental gradients [1]. Over the past decades, carbon (C), nitrogen (N) and phosphorus (P) have attracted the most attention and achieved numerous valuable insights in biogeography and functional ecology at multiple study scales [2–9]. However, information from only C, N and P is increasingly insufficient in more comprehensive researches. Other bioelements should also be included and some have been well used in exploring species coexistence within the community [10,11], plant adaptive strategies within the context of plant economics [12], biogeochemical cycles of elements [13], the evolutionary association of species distribution [7,14,15], the responses of plants to global change [16], etc., of which sulfur (S) is one of the preferred candidates.

S is one of the most essential bioelements for plant survival and growth [17], which represents the fourth macroelement in crops [18] and the ninth macroelement, by concentration, in all plants [19]. S is not only incorporated into structural components, but also more crucial in serving regulatory and catalytic functions, especially in stress conditions [20], and thus is also considered as the fourth most important mineral element after N, P and potassium (K) in plants [21]. S is absorbed mainly in the form of sulfate (SO42−) and is synthesized into a wide variety of S-containing compounds, for example, amino acids (cysteine and methionine), oligopeptides (glutathione and phytochelatins), vitamins, cofactors and various secondary products, taking the job of antioxidative, catalytic and defensive functions and so on [17,19,22]. Physiologically, S and N are taken up by plant roots with similar strategies [23], assimilated in well-coordinated pathways and tightly coupled with each other [24,25]. Therefore, S concentrations and the correlations of S to N and P in plants should be species-specific, which are largely determined by the physiological requirements of plants and vary among species [18,19]. The variations in plant S–N–P stoichiometries can be attributed to the differences in plant taxa, as well as the local elemental conditions in environments, as demonstrated by the regional studies [12,14,25–27]. However, compared with the well-documented leaf N–P stoichiometry, the globally stoichiometric patterns of plant leaf S concentrations and their relationships to N and P are less studied.

In nature, S exists in two forms, inorganic S and organic S, and cycles through the global ecosystem (reviewed in [17]). Sulfate is the most common form of S absorbed by plants, which is reduced to sulfide and then incorporated into organic metabolites in assimilatory processes. The content of sulfate in soil is related to soil age and the mineralization process of organic remains by soil microorganisms [28]. Therefore, the plant S concentrations are affected by local soil S availability, which is determined by soil parent material and climatic conditions [26,29]. Another source of S is the volatile compounds from the volcanoes and oceans, which are oxidized to sulfate in the atmosphere and fall on the surface through precipitation. Under global change regimes, anthropogenic fertilizer S inputs in agricultural ecosystems, atmospheric S deposition in natural ecosystems induced by waste gas emissions and sewage discharge from heavy industries enrich S in environments chronically [30–32] and then alter the elemental stoichiometry and S-related physiological processes in organisms [33].

Compared with leaf N (LN) and leaf P (LP), no studies have focused on global leaf S (LS) stoichiometry by far. Based on the few findings that reported the consistently positive leaf S–N relationships at the regional scales [12,15,34], together with the well-documented leaf N–P correlations [4,6,9], we hypothesized that, at the global scale, (i) bioelemental concentrations of leaf S, N and P were species-specific (that is, they differed among different plant growth forms and phylogenetic taxa), and (ii) LS, LN and LP concentrations varied along geographic and environmental gradients, but (iii) the positive leaf S–N–P correlations were universal across different biomes. To verify our hypotheses and meanwhile fill the basic knowledge gap of global leaf S stoichiometry, we compiled a comprehensive dataset of plant LS, LN and LP concentrations to summarize S-N-P stoichiometric characteristics and provide evidence of phylogenetic and environmental controls on plant S at the global scale.

2. Material and methods

(a). Data collection and compilation

The global dataset of LS, LN and LP concentrations consists of data from published studies, books and online databases (references in electronic supplementary material), plus our unpublished data. We mainly focused on the variation patterns of LS and its relationships with LN and LP across different phylogenetic taxa and plant growth forms worldwide. In total, we collected 31 939 records of data from 6652 sites covering six continents (figure 1), and 2600 plant species belonging to 1186 genera, 210 families, 69 orders, 7 classes and 4 divisions (phyla; electronic supplementary material, figure S1). In the raw data we collected, 7 of the overall 2600 plant species (involving 123 records of bioelemental concentration data) were not given scientific names and thus could not be assigned to any group. To retain as much data as possible, all LS, LN and LP data of the 2600 species were used to analyse the correlations among the three bioelements and their variations along environmental gradients on the overall level, but the data of 2593 species were used in the plant group analysis. In order to confirm the taxonomic classifications of the species and identify the plant growth forms, we employed the Angiosperm Phylogeny Group IV (APG IV) classification, the World Flora Online (http://www.worldfloraonline.org/), Useful Tropical Plants (http://tropical.theferns.info/), Wikipedia (https://en.wikipedia.org/wiki), the Flora of China (http://www.efloras.org) and Catalogue of Life China (http://www.sp2000.org.cn/). Since 127 of the 6652 sites did not provide geographic coordinates in the original references, we used the Map Location and the Environment for Visualizing Images to obtain the latitude, longitude and altitude of the sites. The environmental variables (BIO1–BIO19, electronic supplementary material, table S1) were obtained from the WorldClim (https://worldclim.org/) based on the geographic coordinates of each site. We tested the collinearity of the environmental variables using Pearson correlation analysis. The results showed that BIO2 and BIO15 had the weakest correlations with the other environmental variables (electronic supplementary material, table S1). Considering the significant effects of extreme (max and min) values and the fluctuation of environmental variables on plant growth, we finally selected six variables, including mean diurnal range of temperature (BIO2), maximum temperature in the warmest month (BIO5), minimum temperature in the coldest month (BIO6), precipitation in the wettest month (BIO13), precipitation in the driest month (BIO14) and precipitation seasonality (BIO15), together with the two most commonly used variables (mean annual temperature, MAT, and mean annual precipitation, MAP), as independent variables to analyse the environmental controls of bioelemental variations.

Figure 1.

A total of 6652 sites containing all 31 939 records of data from the six continents in this study

Open in a new tab

A total of 6652 sites containing all 31 939 records of data from the six continents in this study. Europe and Central and East Asia, which contain the majority of data, are shown separately.

(b). Data analysis

In group analysis, overall 2593 species were divided into different phylogenetic groups and growth forms. The phylogenetic groups included angiosperms (2447), gymnosperms (82), pteridophytes (35) and bryophytes (29). Considering the small amount of data of pteridophytes and bryophytes, we focused on angiosperms and gymnosperms. For growth forms, we divided all species into herbaceous (1040) and woody species (1553). Among woody species, we subdivided all species into broad-leaved trees (1340, including broad-leaved species of gymnosperm) and coniferous trees (69) with needle-like leaves by leaf type, and then the broad-leaved trees were further grouped into evergreen species (tree: 732; shrub: 247) and deciduous species (tree: 264; shrub: 97) by leaf lifespan. In coniferous trees, 5 of 69 species (involving 75 of all 16 684 records of data) were deciduous species and thus were not subdivided by leaf lifespan. Herbs were subdivided into aquatic (140) and terrestrial herbs (900) by habitats and into graminoids (141) and forbs (899) by leaf types. All records of data for each group are listed in table 1.

Table 1.

Stoichiometric characteristics of LS, LN and LP concentrations in the overall observations and different plant groups. n, the number of species/observations; AM, the arithmetic mean; GM, the geometric mean. Different lowercase letters denote significant difference among the above groups at p < 0.05 level. The overall observations were divided into four phylogenetic groups: angiosperm (ANG), gymnosperm (GYM), pteridophyte (PTE) and bryophyte (BRY); two growth forms: woody and herb; five growth forms within woody plants: coniferous tree, deciduous broad-leaf (DB) tree, evergreen broad-leaf (EB) tree, deciduous (D) shrub and evergreen (E) shrub; and two growth forms within herbaceous plants by habitat (aquatic and terrestrial) and by leaf type (graminoid and forbs), respectively.

groups n a S (mg g−1) N (mg g−1) P (mg g−1) N:S P:S N:P
n AM/GM n AM/GM n AM/GM n AM/GM n AM/GM n AM/GM
 overall 2600/31 939 2453/29 485 1.90/1.44 1705/27 492 16.80/15.70 1206/27 323 1.41/1.27 1417/25 096 12.60/11.51 908/24 892 1.14/0.98 1015/26 549 13.55/12.11
 phylogeny
 ANG 2447/14 745 2303/12 602 2.94/2.14a 1576/10 548 21.49/20.09a 1095/10 327 1.52/1.25b 1292/8458 11.54/9.56c 802/8205 0.82/0.64b 922/9660 17.80/15.58b
 GYM 82/16 707 79/16 406 1.10/1.06b 65/16 633 13.85/13.43c 61/16 654 1.36/1.29a 62/16 334 13.15/12.68b 56/16 354 1.31/1.23a 46/16 601 10.84/10.37c
 PTE 35/171 35/161 1.33/0.98b 33/116 13.06/11.95d 19/147 0.77/0.61d 32/109 15.12/12.91a 19/138 0.87/0.68b 16/93 25.50/20.63a
 BRY 29/193 29/193 1.73/1.65a 29/193 17.78/17.05b 29/193 1.07/0.84c 29/193 10.76/10.35d 29/193 0.65/0.51c 29/193 28.80/20.27a
 growth form
 woody 1553/27 560 1436/25 412 1.47/1.26b 774/24 182 15.94/15.06b 801/24 544 1.32/1.23b 531/22 077 13.23/12.53a 545/22 400 1.19/1.06a 653/24 013 13.43/12.19a
 herb 1040/4256 1010/3950 4.59/3.37a 929/3308 23.13/21.29a 403/2777 2.28/1.74a 884/3017 8.00/6.18b 361/2490 0.70/0.51b 360/2534 14.71/11.42b
 woody group
 conifers 69/16 684 66/16 385 1.10/1.06c 54/16 616 13.85/13.44d 56/16 640 1.36/1.30a 51/16 319 13.15/12.68b 51/16 342 1.31/1.23a 43/16 592 10.84/10.37d
 DB tree 264/1187 238/808 2.06/1.75a 145/685 22.98/21.76a 130/711 1.49/1.31a 104/328 13.85/12.23b 84/335 1.16/0.79b 102/621 17.82/16.06c
 EB tree 732/7956 670/6779 1.86/1.63a 328/6006 20.51/19.44b 365/6140 1.22/1.10b 179/4837 13.79/12.96a 212/4963 0.88/0.78c 312/5979 19.04/17.50a
 D shrub 97/294 92/262 2.56/1.80b 54/123 21.91/20.16ab 44/128 1.45/1.20a 49/92 12.99/10.51c 39/96 0.92/0.55cd 31/99 16.87/15.82c
 E shrub 247/950 233/808 3.79/2.16ab 142/519 17.81/15.87c 161/698 1.09/0.88c 123/387 10.83/7.48d 140/556 0.66/0.38d 120/495 21.25/17.72b
 herb group (habitat)
 aquatic 140/1708 140/1704 5.94/5.33a 137/1677 25.29/23.77a 92/1489 3.02/2.62a 137/1673 5.04/4.41b 92/1485 0.58/0.47b 90/1465 9.81/8.89b
 terrestrial 900/2548 870/2246 3.57/2.38b 792/1631 20.91/19.00b 311/1288 1.43/1.08b 747/1344 11.68/9.43a 269/1005 0.89/0.58a 270/1069 21.43/16.10a
 herb group (leaf type)
 graminoid 141/612 135/594 3.08/2.42b 130/295 19.00/16.81b 73/283 1.70/1.35b 122/280 9.09/7.30a 66/269 0.99/0.64a 64/216 14.62/10.98a
 forbs 899/3644 875/3356 4.86/3.57a 799/3013 23.54/21.78a 330/2494 2.35/1.79a 762/2737 7.89/6.08b 295/2221 0.67/0.49b 296/2318 14.72/11.47a
Open in a new tab
a

Seven of the overall 2600 plant species were not given a scientific name but had bioelemental concentration in the raw data (involving 123 of the overall 31 939 records of data). Therefore, all 2600 species were included on the overall level, but 2593 species were used in the plant group analysis. In the woody group, 144 of the 1553 species failed in subdivision, involving 489 of the 27 560 records of woody data.

We first drew the histograms to illustrate the frequency of data distribution for the overall data and each group and found that most of the datasets were in a skewed distribution (electronic supplementary material, figures S2–S12). Shapiro–Wilk normality tests and Bartlett tests were then performed to check the normality of residual and homogeneity of variances, respectively. Non-parametric methods (Kruskal–Wallis test and Dunn’s post hoc multiple comparisons) were used to compare the differences in bioelemental concentrations and ratios among the plant groups.

All data of LS, LN and LP concentrations were log10-transformed and then standard major axis (SMA) analysis [35] was conducted to fit the bivariate relationships between the three bioelements for the overall data and different plant groups. Then, we regressed the bioelements against latitude, altitude, MAT and MAP, to determine the geographic patterns and environmental influences, respectively.

Phylogenetic signal is the pattern in which evolutionarily related species tend to resemble each other in their organismal characteristics [36]. Pagel’s λ was used to detect the phylogenetic signal of LS, LN and LP [37], with a value of λ equal to 0 indicating evolutionary independence of a trait from phylogeny and a value of λ close to 1 indicating a strong phylogenetic signal (evolutionary conservatism) for the trait [38]. We used the ‘phytools’ [39] and ‘S.PhyloMaker’ package [40] in R to construct a phylogenetic tree (without bryophytes) and visualized it in the Interactive Tree of Life (ITOL). Pagel’s λ values and their significance were calculated using the ‘Picante’ package in R [41].

We employed the phylogenetic linear mixed effect model (PLMM) and variance partitioning analysis to partition the variations in bioelements into phylogeny and environments by using residual maximum likelihood estimation [42,43]. A hierarchically nested structure ‘order/family/species’ represented the phylogenetic effects, which enabled us to investigate the degree of variation at each phylogenetic level [44]. The ‘site’ component of the model was composed of variations in leaf elemental concentrations affected by environmental variables [43–45]. We defined the overall random term within PLMM as ‘site + [(order/family/species)]’. All such analyses were performed for all species, woody plants and herbaceous plants, respectively. We used the ‘lmer’ function in the ‘lme4’ package in R [46].

To further investigate the effects of environmental variables on the variations in LS, LN and LP concentrations, we firstly regressed the three bioelements against the environmental variables, respectively, and then introduced the multiple regression model with variance partitioning analysis to estimate the importance of the selected environmental variables in explaining the variation of the three bioelements. The analyses were performed using the ‘lm’ function in ‘stats’ package and the ‘calc.relimp’ function in ‘relaimpo’ package in R [47,48].

3. Results

(a). Global LS, LN and LP stoichiometric characteristics

For all pooled data (31 939 records), the arithmetic (geometric) mean concentrations of plant LS, LN, LP, N:S, P:S and N:P were 1.90 (1.44) mg g−1, 16.80 (15.70) mg g−1, 1.41 (1.27) mg g−1, 12.60 (11.51), 1.14 (0.98) and 13.58 (12.12), respectively (table 1).

The three bioelements showed significant differences among different phylogenetic taxa and growth forms (table 1). The highest LS, LN and LP concentrations were found in angiosperms (2.94 mg S g−1, 21.49 mg N g−1 and 1.52 mg P g−1), while the lowest LS concentrations in gymnosperms (1.10 mg S g−1) and the lowest LN and LP concentrations were in pteridophytes (13.06 mg N g−1 and 0.77 mg P g−1), respectively. The N:S, P:S and N:P ratios varied from 10.76 (bryophyte) to 15.12 (pteridophyte), 0.65 (bryophyte) to 1.31 (gymnosperm) and 10.84 (gymnosperm) to 28.80 (pteridophyte), respectively.

Herbaceous species showed significantly higher concentrations than woody species in all the three bioelements (4.59 versus 1.47 mg S g−1, 23.13 versus 15.94 mg N g−1 and 2.28 versus 1.32 mg P g−1), but lower N:S (8.00 versus 13.23) and P:S (0.70 versus 1.19) and similar N:P (14.71 versus 13.43) (table 1). Within woody species, deciduous species accumulated more LN and LP than evergreen species but LS did not show such a pattern. Shrubs had the highest LS, while coniferous trees had the lowest LS and LN. Within herbaceous species, all the three bioelements in aquatic herbs were significantly higher than those in terrestrial herbs, resulting in lower N:S, P:S and N:P in aquatic herbs. Forbs had higher LS, LN and LP concentrations than graminoid, but lower N:S and P:S and similar N:P ratios.

(b). The scaling relationships among LS, LN and LP

For all pooled data, LS, LN and LP positively correlated with each other (figure 2a-c), with scaling exponents of LN–LS 0.64 (R2 = 0.29, p < 0.001), LP–LS 0.76 (R2 = 0.10, p < 0.001) and LN–LP 0.79 (R2 = 0.17, p < 0.001) (table 2), but differed among phylogenetic taxa (figure 2d-f) and growth forms (figure 2g-l).

Figure 2.

The bivariate trait relationships among LS, LN and LP concentrations at overall observations (a–c) and different plant groups (dl)

Open in a new tab

The bivariate trait relationships among LS, LN and LP concentrations at overall observations (a–c) and different plant groups (d–l). Angiosperm (ANG), gymnosperm (GYM), pteridophyte (PTE) and bryophyte (BRY); two growth forms: woody and herb; five growth forms within woody plants: coniferous tree, deciduous broad-leaf (DB) tree, evergreen broad-leaf (EB) tree, deciduous (d) shrub and evergreen (e) shrub; two growth forms within herbaceous plants by habitat (aquatic and terrestrial) and by leaf type (graminoid and forbs), respectively.

Table 2.

Summary of standardized major axis (SMA) regression results for LS, LN and LP concentrations at overall observations and different plant groups. n, the number of observations. The overall observations were divided into four phylogenetic groups: angiosperm (ANG), gymnosperm (GYM), pteridophyte (PTE) and bryophyte (BRY); two growth forms: woody and herb; five growth forms within woody plants: coniferous tree, deciduous broad-leaf tree (DB), evergreen broad-leaf tree (EB), deciduous (D) shrub and evergreen (E) shrub; two growth forms within herbaceous plants by habitat (aquatic and terrestrial) and by leaf type (graminoid and forbs), respectively. The bivariate scaling exponents (α with 95% confidence interval, CI) were calculated by SMA analysis using log10-transformed data of bioelements concentrations.

group N–S P–S N–P
n scaling exponent n scaling exponent n scaling exponent
αSMA (95% CI) R 2 p αSMA (95% CI) R 2 p αSMA (95% CI) R 2 p
overall 25 096 0.64 (0.64–0.65) 0.29 <0.001 24 892 0.76 (0.75–0.77) 0.10 <0.001 26 549 0.79 (0.79–0.80) 0.17 <0.001
phylogeny
 ANG 8458 0.55 (0.54–0.56) 0.10 <0.001 8205 0.80 (0.79–0.82) 0.17 <0.001 9660 0.62 (0.61–0.63) 0.27 <0.001
 GYM 16 334 0.97 (0.96–0.99) 0.16 <0.001 16 354 1.16 (1.14–1.18) 0.05 <0.001 16 601 0.83 (0.82–0.84) 0.23 <0.001
 PTE 109 0.58 (0.50–0.69) 0.26 <0.001 138 1.13 (0.96–1.33) 0.08 <0.001 93 0.54 (0.45–0.66) 0.08 <0.01
 BRY 193 0.97 (0.86–1.09) 0.27 <0.001 193 2.53 (2.20–2.91) 0.03 <0.05 193 0.38 (0.34–0.43) 0.26 <0.001
growth form
 woody 22 077 0.87 (0.86–0.88) 0.23 <0.001 22 400 0.90 (0.88–0.91) 0.01 <0.001 24 013 0.88 (0.87–0.89) 0.11 <0.001
 herb 3017 0.54 (0.52–0.56) 0.18 <0.001 2490 0.91 (0.87–0.94) 0.23 <0.001 2534 0.54 (0.52–0.56) 0.33 <0.001
woody group
 conifers 16 319 0.97 (0.96–0.99) 0.16 <0.001 16 342 1.16 (1.14–1.18) 0.05 <0.001 16 592 0.83 (0.82–0.84) 0.23 <0.001
 DB tree 328 0.61 (0.55–0.67) 0.14 <0.001 335 0.76 (0.69–0.85) 0.02 <0.01 621 0.68 (0.63–0.72) 0.24 <0.001
 EB tree 4837 0.99 (0.97–1.02) 0.12 <0.001 4963 1.14 (1.11–1.17) 0.05 <0.001 5979 0.74 (0.73–0.76) 0.26 <0.001
 D shrub 92 0.49 (0.41–0.60) 0.12 <0.001 96 0.69 (0.56–0.84) >0.05 99 0.75 (0.67–0.84) 0.70 <0.001
 E shrub 387 0.58 (0.53–0.64) 0.04 <0.001 556 0.65 (0.60–0.70) 0.06 <0.001 495 0.77 (0.71–0.83) 0.19 <0.001
herb group (habitat)
 aquatic 1673 0.78 (0.74–0.82) 0.04 <0.001 1485 1.29 (1.23–1.36) 0.01 <0.01 1465 0.66 (0.64–0.69) 0.37 <0.001
 terrestrial 1344 0.60 (0.57–0.63) 0.18 <0.001 1005 0.83 (0.78–0.88) 0.11 <0.001 1069 0.57 (0.54–0.61) 0.14 <0.001
herb group (leaf type)
 graminoid 280 0.69 (0.62–0.76) 0.26 <0.001 269 0.80 (0.71–0.90) 0.09 <0.001 216 0.78 (0.69–0.89) 0.03 <0.05
 forbs 2737 0.52 (0.50–0.54) 0.15 <0.001 2221 0.94 (0.90–0.97) 0.23 <0.001 2318 0.51 (0.49–0.53) 0.38 <0.001
Open in a new tab

Among different phylogenetic taxa, the LN–LS exponents varied greatly in angiosperms (0.55), gymnosperms (0.97), pteridophytes (0.58) and bryophytes (0.97), while the LP–LS exponents of the four groups were 0.80, 1.16, 1.13 and 2.53 and those of LN–LP were 0.62, 0.83, 0.54 and 0.38, respectively (table 2 and figure 2d-f). Woody plants had higher LN–LS (0.87 versus 0.54) and LN–LP (0.88 versus 0.54) exponents but similar LP–LS (0.90 versus 0.91) exponents to herbaceous plants (table 2 and figure 2g-l). Within woody species, deciduous species showed lower LN–LS and LN–LP exponents than evergreen species. Conifers had higher LP–LS and LN–LP exponents than broad-leaf species. LN–LP exponents showed less variations (0.68–0.83) than LN–LS (0.49–0.99) and LP–LS exponents (0.65–1.16) among growth forms of woody plants (table 2). Within herbaceous species, exponents of LN–LS, LP–LS and LN–LP of aquatic plants were higher than those of terrestrial plants (0.78 versus 0.60, 1.29 versus 0.83, 0.66 versus 0.57), while graminoid showed higher LN–LS (0.69 versus 0.52) and LN–LP (0.78 versus 0.51) but lower LP–LS (0.80 versus 0.94) exponents than forbs.

(c). Biogeographic and environmental patterns of LS, LN and LP

For all species pooled together, LS and LN negatively but LP positively correlated with latitude (figure 3a I–III and electronic supplementary material, table S2). LS of gymnosperms showed similar latitudinal patterns to the overall species, whereas that of angiosperms had no significant trends with latitude (figure 3a IV). LN and LP of both angiosperms and gymnosperms increased with increasing latitude (figure 3a V–VI). All the three elements of all data and angiosperms positively but that of gymnosperms negatively correlated with altitude (figure 3b I–VI). Among growth forms, woody plants had less bioelements but herbaceous plants concentrated more bioelements (LS, LN and LP) toward high latitude and high altitude (figure 3a,b VII–XII).

Figure 3.

Patterns of LS, LN and LP concentrations along the geographic gradients

Open in a new tab

Patterns of LS, LN and LP concentrations along the geographic gradients. (a) latitude (°) and (b) altitude (km). Angiosperm (ANG), gymnosperm (GYM), pteridophyte (PTE) and bryophyte (BRY); two growth forms: woody and herb; five growth forms within woody plants: coniferous tree, deciduous broad-leaf tree (DB), evergreen broad-leaf tree (EB), deciduous (d) shrub and evergreen (e) shrub; two growth forms within herbaceous plants by habitat (aquatic and terrestrial) and by leaf type (graminoid and forbs), respectively.

In terms of environmental patterns, LS showed the opposite patterns with respect to MAT (positive, figure 4a I) and MAP (negative, figure 4b I). LN of overall data positively but LP of overall data negatively correlated with MAT and MAP (figure 4a,b II–III). LS and LN of gymnosperms showed positive correlations with MAT (figure 4a IV–V), whereas the three bioelements of angiosperms negatively correlated with MAT and MAP (figure 4a,b IV–VI). LS and LN of woody and herbaceous plants showed opposite patterns but LP of the two plant growth forms had consistent trends with MAT and MAP (figure 4a,b VII–XII).

Figure 4.

Patterns of LS, LN and LP concentrations along the climatic gradients

Open in a new tab

Patterns of LS, LN and LP concentrations along the climatic gradients. (a) Mean annual temperature (MAT, ℃), (b) Mean annual precipitation (MAP, mm). Angiosperm (ANG), gymnosperm (GYM), pteridophyte (PTE) and bryophyte (BRY); two growth forms: woody and herb; five growth forms within woody plants: coniferous tree, deciduous broad-leaf tree (DB), evergreen broad-leaf tree (EB), deciduous (d) shrub and evergreen (e) shrub; two growth forms within herbaceous plants by habitat (aquatic and terrestrial) and by leaf type (graminoid and forbs), respectively.

(d). Relative effects of phylogeny and environments on variations of LS, LN and LP

Phylogenetic signal detection showed that all of LS, LN and LP had significantly phylogenetic signals (electronic supplementary material, figure S1). Both phylogeny and environments regulated the variations in the three bioelemental concentrations. For all species, degrees of explanation for LS, LN and LP by phylogenetic levels (incorporating effects of order, family and species levels) were 86.87, 79.70 and 69.23% and by environments (sites) were 4.68, 10.26 and 18.45%, respectively (figure 5; electronic supplementary material, table S3). That is, phylogeny explained more variances than environments for LS, LN and LP of all, woody and herbaceous species. LS was more evolutionarily conservative than LN and LP. In addition, bioelements of woody species were more controlled by phylogeny than herbs (degrees of explanations for LS 93.90% versus 51.62%, LN 79.07% versus 47.65% and LP 58.15% versus 33.06%).

Figure 5.

Variance partitioning for LS, LN and LP variations using a phylogenetic linear mixed effect model (PLMM) for all species (a), woody species (b), and herbaceous species

Open in a new tab

Variance partitioning for LS, LN and LP variations using a phylogenetic linear mixed effect model (PLMM) for all species (a), woody species (b), and herbaceous species (c). The model including phylogenetic effects (hierarchically nested structures: order/family/species levels), environmental effects (site level) and residual. Columns with different colours represent different variance components in PLMM.

Further analysis by partitioning the environmental variables showed that the three bioelements of all species and the two growth forms were more affected by temperature than precipitation (figure 6; electronic supplementary material, table S4). The variability of temperature and precipitation, rather than the mean values, had the most significant effects on LS of all species (figure 6a). For different growth forms, the variability of temperature and precipitation controlled the variations in LS of woody species (figure 6b), while those of herbaceous species were governed by the variability of temperature and the lowest temperature (figure 6c). Of the three bioelements, LN was least affected by environments in all, woody and herbaceous species (figure 6).

Figure 6.

Contributions of environmental variables to LS, LN and LP concentrations for all species (a), woody species (b), and herbaceous species (c) based on correlation and multiple regression mode

Open in a new tab

Contributions of environmental variables to LS, LN and LP concentrations for all species (a), woody species (b), and herbaceous species (c) based on correlation and multiple regression model. Circle size represents the variable importance (that is, proportion of explained variability calculated via multiple regression model and variance partition analysis). MAT, mean annual temperature; BIO02, mean diurnal range of temperature (monthly maximum minus monthly minimum); BIO5, maximum temperature in the warmest month; BIO6, minimum temperature in the coldest month; MAP, annual precipitation; BIO13, precipitation in the wettest month; BIO14, precipitation in the driest month; BIO15, precipitation seasonality (coefficient of variation).

4. Discussion

(a). Global LS stoichiometry and the relationships with LN and LP

As mentioned above, a large number of studies reported the stoichiometric characteristics of plant N–P at multiple scales and the related hypotheses of plant ecological stoichiometry based on N–P over the past decades (reviewed in [49]). The geometric mean LN and LP concentrations of our pooled data were 15.70 mg g−1 and 1.27 mg g−1, differing slightly from the recent reports of N 18.9 mg g−1 and P 1.2 mg g−1 by Tian et al. [9] and N 17.1 mg g−1 and P 1.57 mg g−1 by Reich et al. [6] at the global scale. These two bioelements varied significantly among the different plant groups. Unlike LN and LP, no global data on LS are available for comparison. We reported a global geometric mean concentration of leaf S, N:S and P:S of 1.44 mg g−1, 11.51 and 0.98, with significant differences among phylogenetic taxa and growth forms (table 1). At the regional scale, Dalle Fratte et al. [12] measured 740 vascular plant species in northern Italy and reported LS as 2.1 mg g−1. Terrestrial plants in China contained 1.58 mg g−1 (1900 species [27]), 1.55 mg g−1 (702 species [14]), 1.2 mg g−1 (348 species [50]), 2.13 mg g−1 (2207 species [15]) and 2.32 mg g−1 (2745 species [34]) LS, while in the Tibetan Plateau, the mean LS concentration was 1.68 mg g−1 at the community level (2040 communities [51]). We believed that the differences in the plant species they measured explained the variations in leaf S concentrations at the regional scales. For example, the mean leaf S concentrations in the forests was 1.69 mg g−1, while in the deserts it was 12.22 mg g−1 [15]. The low LS concentrations in Wu et al. [50] were probably owing to most of the study species being conifers. Similarly, relatively lower LS concentrations in our study were caused by the high proportion of gymnosperms (mainly conifers), because gymnosperms contained the minimum LS concentrations of all phylogenetic taxa and thus pulled down the mean values of woody plants (table 1). The herbaceous species, especially aquatic plants, had the highest LS concentrations in all growth forms, verifying the crucial role of S in stress tolerance, because most of the aquatic plant species were collected from saline habitats in the Tibetan Plateau and arid and semi-arid regions in northwest China [25].

The LS, LN and LP were positively correlated with each other, ignoring the phylogenetic taxa and growth forms (figure 2). The scaling exponent of N–P correlation was 0.79 for all pooled data in this study, slightly higher than the general 2/3-power law of N–P relations [6] and also varied among different plant groups [9]. Based on our global dataset, we reported a scaling exponent for N–S and P–S relationships as 0.64 and 0.76, respectively, but showed significant differences among phylogenetic taxa (0.55–0.97) and growth forms (0.49–0.99; table 2). Compared with LN and LP, LS in gymnosperms were much lower than that in angiosperms (N: 13.43 versus 20.09, 1.5 times; P: 1.29 versus 1.25, equal; S: 1.06 versus 2.14, 2 times, table 2), resulting in more gentle slopes of N–S and P–S relations. Both leaf S and N concentrations of conifers were the lowest in all growth forms. In this study, samples from deciduous conifers (e.g. genus Larix) accounted for less than 1% (43 in 16 385 of S and 73 in 16 616 samples of N), whereas evergreen conifers (long leaf lifespan species) dominated the vast majority of the total coniferous species. Leaf elemental concentrations decrease with increasing leaf lifespan [52,53], that is, the longer the leaf lifespan, the lower the element concentrations are. Therefore, longer leaf lifespan of conifers explained the lower leaf S and N concentrations in this plant group. Interestingly, LP concentrations were highest in gymnosperms, i.e. the conifers (table 1), resulting in higher P–S and N–P exponents than other groups (table 2). Conifers accumulate more phospholipids in leaves to resist the severe cold conditions, explaining the higher leaf P concentrations in conifers [9].

(b). Phylogenetic and environmental controls on variations in LS, LN and LP

As plant functional traits formed by long-term evolution, leaf elemental concentrations are controlled by both phylogeny and environments [7,54]. Although previous studies have found that phylogeny affected bioelements to varying degrees [15,43,55], all three bioelements in this study showed significant phylogenetic signals, of which leaf S was the one most affected by phylogenetic taxa (figure 5). Aquatic plants contained the highest LS and LN concentrations, while conifers had the lowest values in this study (table 1). Considering the distribution regions of different species, we suggested that the globally stoichiometric patterns of LS, LN and LP were formed by plant distribution patterns. For example, aquatic plants in this study were collected in the Tibetan Plateau (low latitude but high altitude) and conifers were from Europe (high latitude but low altitude), resulting in negative latitudinal patterns and positive altitudinal patterns of leaf S and N (figure 3). Relatively high LP concentrations both in conifers and in aquatic plants produced positive latitudinal and altitudinal patterns of these bioelements, respectively (figure 3). Interestingly, latitudinal LN patterns of plant groups of phylogenetic taxa and the main growth forms (positive; figure 3a V, VIII) were opposite from that of all species (negative; figure 3a II). It suggested that, when the plant groups were separated for analysis, the effects of environments on LN concentrations were obvious. That is, plants in cold regions accumulated more leaf N to maintain survival and growth [4,56,57]. LP of different plant groups showed the same positively latitudinal patterns with LN (figure 3a VI, IX), whereas LS did not (figure 3a IV, VII), indicating that LS was less affected by the environment (figure 5). In terms of growth forms, degrees of explanations for the three bioelements of woody plants by phylogenetic taxa were higher than that of herbs (figure 5b,c), providing evidence of the greater evolutionary conservatism of woody plants [58–61] in the perspective of elemental stoichiometry.

Within environmental variables, temperature had greater effects on the three bioelements of all, woody and herbaceous species than precipitation (figure 6; electronic supplementary material, table S4). All the three bioelements of different plant groups, except LS of gymnosperms (conifers), decreased with increasing temperature (figure 4a IV–VI), indicating that the colder the habitats, the higher the plant bioelemental concentrations. For LN and LP, Reich & Oleksyn proposed the temperature–plant physiology hypothesis, which suggested that plants accumulated more bioelements to counterbalance the depressed biochemical efficiency caused by low temperature, to explain the increased concentrations of plant bioelements in cold regions. Patterns of LS–temperature of gymnosperms and conifers (positive) were opposite to that of angiosperms and herbs (negative), reflecting the differences of life history strategies of different plant groups [12]. Gymnosperms (conifers) showed adaptive strategies of long leaf lifespan, low growth rate, low bioelemental concentrations and so on [62]. Considering the negative correlations between leaf lifespan and bioelemental concentrations, increasing temperature shortened the leaf lifespan [63] but improved the bioelemental concentrations. Furthermore, the LS–temperature patterns of gymnosperm and conifers were also opposite to those of LN–temperature and LP–temperature (figure 4a IV–IX). Although LS positively correlated with LN and LP, different functions of S and N/P in plants might lead to the differences in the environmental patterns of the three bioelements, but need further studies to explain them physiologically.

The diurnal temperature variability and extremes (the minimum temperature in this study) explained more regarding the variations in the three bioelements, especially in herbs, than that of mean temperature (figure 6; electronic supplementary material, table S4). The variability and the extremes of temperature represent the extent to which the plants have to face the extreme conditions [64–66]. The lower the minimum temperature, or the greater the temperature variability, the higher the bioelemental concentrations. The results reflected the important role of bioelements in plant stress resistance. We suggest that further consideration of the effects of climatic variability and extreme temperature on plant survival and growth is necessary in future climate change regimes.

5. Conclusions

By compiling a dataset of plant S, N and P covering six continents, we provided a global perspective and quantitative basis for plant leaf S stoichiometry and its scaling relationships with N and P. We found that leaf S, N and P concentrations significantly differed among phylogenetic taxa and growth forms, but showed robust positive bivariate correlations in all plant groups. Both phylogeny and environments regulated the variations in the leaf S, N and P concentrations, with phylogeny explaining more than environments. Of the three elements, S was the one most affected by phylogenetic taxa. Within climatic variables, the variability and extremes, rather than the mean values of temperature, were the key drivers of elemental variations. Considering the crucial roles of S in plants and its relationships with N and P, we proposed that S should be incorporated into the global ecological stoichiometric research framework, and be stressed in studies on plant–environment interactions.

Acknowledgements

We would like to thank all the authors of the data we have compiled in this study for their invaluable data. This work was supported by the TRY initiative on plant traits (http://www.try-db.org). Also, we thank two anonymous reviewers for their constructive comments and suggestions, which have helped to improve this manuscript.

Contributor Information

Liangjian Zhang, Email: zhanglj@whu.edu.cn.

Zhenjun Zuo, Email: zuozhenjun@whu.edu.cn.

Xiujuan Qiao, Email: xjqiao@wbgcas.cn.

Yixuan Liu, Email: yixuan.liu@utibet.edu.cn.

Rui Qu, Email: QRqurui@whu.edu.cn.

Haocun Zhao, Email: hczhao@whu.edu.cn.

Youxin Wang, Email: changyon0939@126.com.

Peidong Zhao, Email: Zhaopd@whu.edu.cn.

Lin Zhang, Email: zhanglin@itpcas.ac.cn.

Zhigang Wu, Email: wuzg@ihb.ac.cn.

Zhong Wang, Email: wangzhong@whu.edu.cn.

Ethics

This work did not require ethical approval from a human subject or animal welfare committee

Data accessibility

The data and code from this manuscript can be found at [67].

Supplementary material is available online [68].

Declaration of AI use

We have not used AI-assisted technologies in creating this article.

Authors’ contributions

Lia.Z.: data curation, formal analysis, investigation and writing—original draft; Z.Z.: data curation, formal analysis, investigation and writing—original draft; X.Q.: funding acquisition, supervision and writing—review and editing; Y.L.: writing—review and editing; R.Q.: data curation and investigation; H.Z.: data curation, formal analysis and investigation; Y.W.: data curation and investigation; P.Z.: data curation and investigation; Lin.Z.: writing—review and editing; Z.Wu.: writing—review and editing; Z.Wa.: conceptualization, funding acquisition, supervision and writing—original draft.

All authors gave the final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare that we have no competing interests.

Funding

This work was funded by the National Natural Science Foundation of China (Grant numbers 32360287 and 32171536).

References

  • 1. Sterner RW, Elser JJ. 2003. Ecological stoichiometry: the biology of elements from molecules to the biosphere. Princeton, NJ: Princeton University Press. ( 10.1515/9781400885695) [DOI] [Google Scholar]
  • 2. Ågren GI. 2004. The C:N:P stoichiometry of autotrophs – theory and observations. Ecol. Lett. 7, 185–191. ( 10.1111/j.1461-0248.2004.00567.x) [DOI] [Google Scholar]
  • 3. McGroddy ME, Daufresne T, Hedin LO. 2004. Scaling of C:N:P stoichiometry in forests worldwide: implications of terrestrial redfield-type ratios. Ecology 85, 2390–2401. ( 10.1890/03-0351) [DOI] [Google Scholar]
  • 4. Reich PB, Oleksyn J. 2004. Global patterns of plant leaf N and P in relation to temperature and latitude. Proc. Natl Acad. Sci. USA 101, 11001–11006. ( 10.1073/pnas.0403588101) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 5. Güsewell S. 2004. N: P ratios in terrestrial plants: variation and functional significance. New. Phytol. 164, 243–266. ( 10.1111/j.1469-8137.2004.01192.x) [DOI] [PubMed] [Google Scholar]
  • 6. Reich PB, Oleksyn J, Wright IJ, Niklas KJ, Hedin L, Elser JJ. 2010. Evidence of a general 2/3-power law of scaling leaf nitrogen to phosphorus among major plant groups and biomes. Proc. R. Soc. B 277, 877–883. ( 10.1098/rspb.2009.1818) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 7. Sardans J, et al. 2015. Foliar elemental composition of European forest tree species associated with evolutionary traits and present environmental and competitive conditions. Glob. Ecol. Biogeogr. 24, 240–255. ( 10.1111/geb.12253) [DOI] [Google Scholar]
  • 8. Tang Z, et al. 2018. Patterns of plant carbon, nitrogen, and phosphorus concentration in relation to productivity in China’s terrestrial Ecosystems. Proc. Natl Acad. Sci. USA 115, 4033–4038. ( 10.1073/pnas.1700295114) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 9. Tian D, et al. 2018. Global leaf nitrogen and phosphorus stoichiometry and their scaling exponent. Natl. Sci. Rev. 5, 728–739. ( 10.1093/nsr/nwx142) [DOI] [Google Scholar]
  • 10. Peñuelas J, Fernández-Martínez M, Ciais P, Jou D, Piao SL, Obersteiner M, Vicca S, Janssens IA, Sardans J. 2019. The bioelements, the elementome, and the biogeochemical niche. Ecology 100, e02652. ( 10.1002/ecy.2652) [DOI] [PubMed] [Google Scholar]
  • 11. Sardans J, et al. 2021. Empirical support for the biogeochemical niche hypothesis in forest trees. Nat. Ecol. Evol. 5, 184–194. ( 10.1038/s41559-020-01348-1) [DOI] [PubMed] [Google Scholar]
  • 12. Dalle Fratte M, Pierce S, Zanzottera M, Cerabolini BEL. 2021. The association of leaf sulfur content with the leaf economics spectrum and plant adaptive strategies. Funct. Plant Biol. 48, 924–935. ( 10.1071/FP20396) [DOI] [PubMed] [Google Scholar]
  • 13. Tipping E, Somerville CJ, Luster J. 2016. The C:N:P:S stoichiometry of soil organic matter. Biogeochemistry 130, 117–131. ( 10.1007/s10533-016-0247-z) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 14. Zhang SB, Zhang JL, Slik JWF, Cao KF. 2012. Leaf element concentrations of terrestrial plants across China are influenced by taxonomy and the environment. Glob. Ecol. Biogeogr. 21, 809–818. ( 10.1111/j.1466-8238.2011.00729.x) [DOI] [Google Scholar]
  • 15. Zhao WZ, Xiao CW, Li MX, Xu L, He NP. 2022. Variation and adaptation in leaf sulfur content across China. J. Plant Ecol. 15, 743–755. ( 10.1093/jpe/rtac021) [DOI] [Google Scholar]
  • 16. Tian DS, Reich PB, Chen HYH, Xiang YZ, Luo YQ, Shen Y, Meng C, Han WX, Niu SL. 2019. Global changes alter plant multi-element stoichiometric coupling. New. Phytol. 221, 807–817. ( 10.1111/nph.15428) [DOI] [PubMed] [Google Scholar]
  • 17. Takahashi H, Kopriva S, Giordano M, Saito K, Hell R. 2011. Sulfur assimilation in photosynthetic organisms: molecular functions and regulations of transporters and assimilatory enzymes. Annu. Rev. Plant Biol. 62, 157–184. ( 10.1146/annurev-arplant-042110-103921) [DOI] [PubMed] [Google Scholar]
  • 18. Shah SH, Islam S, Mohammad F. 2022. Sulphur as a dynamic mineral element for plants: a review. J. Soil Sci. Plant Nutr. 22, 2118–2143. ( 10.1007/s42729-022-00798-9) [DOI] [Google Scholar]
  • 19. Saito K. 2004. Sulfur assimilatory metabolism. The long and smelling road. Plant Physiol. 136, 2443–2450. ( 10.1104/pp.104.046755) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 20. May MJ, Vernoux T, Leaver C, Montagu MV, Inze D. 1998. Glutathione homeostasis in plants: implications for environmental sensing and plant development. J. Exp. Bot. 49, 649–667. ( 10.1093/jxb/49.321.649) [DOI] [Google Scholar]
  • 21. Nazar R, Iqbal N, Masood A, Syeed S, Khan NA. 2011. Understanding the significance of sulfur in improving salinity tolerance in plants. Environ. Exp. Bot. 70, 80–87. ( 10.1016/j.envexpbot.2010.09.011) [DOI] [Google Scholar]
  • 22. Leustek T, Martin MN, Bick JA, Davies JP. 2000. Pathways and regulation of sulfur metabolism revealed through molecular and genetic studies. Annu. Rev. Plant Phys. 51, 141–165. ( 10.1146/annurev.arplant.51.1.141) [DOI] [PubMed] [Google Scholar]
  • 23. Legay N, Personeni E, Slezack-Deschaumes S, Piutti S, Cliquet JB. 2014. Grassland species show similar strategies for sulphur and nitrogen acquisition. Plant Soil 375, 113–126. ( 10.1007/s11104-013-1949-9) [DOI] [Google Scholar]
  • 24. Kopriva S, Rennenberg H. 2004. Control of sulphate assimilation and glutathione synthesis: interaction with N and C metabolism. J. Exp. Bot. 55, 1831–1842. ( 10.1093/jxb/erh203) [DOI] [PubMed] [Google Scholar]
  • 25. Zuo Z, Zhao H, Yang L, Lv T, Li X, Ma F, Wang Z, Yu D. 2022. Salinity induces allometric accumulation of sulfur in plants and decouples plant nitrogen‐sulfur correlation in alpine and arid wetlands. Global Biogeochem. Cycles 36, e2022GB007372. ( 10.1029/2022GB007372) [DOI] [Google Scholar]
  • 26. Tallec T, Diquélou S, Fauveau C, Bataillé MP, Ourry A. 2008. Effects of nitrogen and sulphur gradients on plant competition, N and S use efficiencies and species abundance in a grassland plant mixture. Plant Soil 313, 267–282. ( 10.1007/s11104-008-9699-9) [DOI] [Google Scholar]
  • 27. Han WX, Fang JY, Reich PB, Ian Woodward F, Wang ZH. 2011. Biogeography and variability of eleven mineral elements in plant leaves across gradients of climate, soil and plant functional type in China. Ecol. Lett. 14, 788–796. ( 10.1111/j.1461-0248.2011.01641.x) [DOI] [PubMed] [Google Scholar]
  • 28. Turner BL, Condron LM, France CAM, Lehmann J, Solomon D, Peltzer DA, Richardson SJ. 2016. Sulfur dynamics during long-term ecosystem development. Biogeochemistry 128, 281–305. ( 10.1007/s10533-016-0208-6) [DOI] [Google Scholar]
  • 29. Luo WT, et al. 2016. A threshold reveals decoupled relationship of sulfur with carbon and nitrogen in soils across arid and semi-arid grasslands in northern China. Biogeochemistry 127, 141–153. ( 10.1007/s10533-015-0174-4) [DOI] [Google Scholar]
  • 30. Bailey JS, Deng YJ, Smith RV. 2001. Changes in soil organic carbon storage under grassland as evidenced by changes in sulphur input-output budgets. Chemosphere 42, 141–151. ( 10.1016/s0045-6535(00)00119-3) [DOI] [PubMed] [Google Scholar]
  • 31. Yang YH, Ji CJ, Ma WH, Wang SF, Wang SP, Han WX, Mohammat A, Robinson D, Smith P. 2012. Significant soil acidification across northern China’s grasslands during 1980S-2000s. Glob. Chang. Biol. 18, 2292–2300. ( 10.1111/j.1365-2486.2012.02694.x) [DOI] [Google Scholar]
  • 32. Giweta M, Dyck MF, Malhi SS, Puurveen D, Robertson JA. 2014. Long-term S-fertilization increases carbon sequestration in a sulfur-deficient soil. Can. J. Soil. Sci. 94, 295–301. ( 10.4141/cjss2013-022) [DOI] [Google Scholar]
  • 33. Norici A, Hell R, Giordano M. 2005. Sulfur and primary production in aquatic environments: an ecological perspective. Photosyn. Res. 86, 409–417. ( 10.1007/s11120-005-3250-0) [DOI] [PubMed] [Google Scholar]
  • 34. Zhao WZ, Xiao CW, Li MX, Xu L, Li X, He NP. 2022. Spatial variation and allocation of sulfur among major plant organs in China. Sci. Total Environ. 844, 157155. ( 10.1016/j.scitotenv.2022.157155) [DOI] [PubMed] [Google Scholar]
  • 35. Warton DI, Duursma RA, Falster DS, Taskinen S. 2012. Smatr 3-an R package for estimation and inference about allometric lines. Methods Ecol. Evol. 3, 257–259. ( 10.1111/j.2041-210X.2011.00153.x) [DOI] [Google Scholar]
  • 36. Blomberg SP, Garland T, Ives AR. 2003. Testing for phylogenetic signal in comparative data: behavioral traits are more labile. Evolution. 57, 717–745. ( 10.1111/j.0014-3820.2003.tb00285.x) [DOI] [PubMed] [Google Scholar]
  • 37. Pagel M. 1999. Inferring the historical patterns of biological evolution. Nature 401, 877–884. ( 10.1038/44766) [DOI] [PubMed] [Google Scholar]
  • 38. Freckleton RP, Harvey PH, Pagel M. 2002. Phylogenetic analysis and comparative data: a test and review of evidence. Am. Nat. 160, 712–726. ( 10.1086/343873) [DOI] [PubMed] [Google Scholar]
  • 39. Revell LJ. 2012. Phytools: an R package for phylogenetic comparative biology (and other things). Methods Ecol. Evol. 3, 217–223. ( 10.1111/j.2041-210X.2011.00169.x) [DOI] [Google Scholar]
  • 40. Qian H, Jin Y. 2016. An updated megaphylogeny of plants, a tool for generating plant phylogenies and an analysis of phylogenetic community structure. J. Plant Ecol. 9, 233–239. ( 10.1093/jpe/rtv047) [DOI] [Google Scholar]
  • 41. Kembel SW, Cowan PD, Helmus MR, Cornwell WK, Morlon H, Ackerly DD, Blomberg SP, Webb CO. 2010. Picante: R tools for integrating phylogenies and ecology. Bioinformatics 26, 1463–1464. ( 10.1093/bioinformatics/btq166) [DOI] [PubMed] [Google Scholar]
  • 42. Broadley MR, Bowen HC, Cotterill HL, Hammond JP, Meacham MC, Mead A, White PJ. 2004. Phylogenetic variation in the shoot mineral concentration of angiosperms. J. Exp. Bot. 55, 321–336. ( 10.1093/jxb/erh002) [DOI] [PubMed] [Google Scholar]
  • 43. Watanabe T, Broadley MR, Jansen S, White PJ, Takada J, Satake K, Takamatsu T, Tuah SJ, Osaki M. 2007. Evolutionary control of leaf element composition in plants. New Phytol. 174, 516–523. ( 10.1111/j.1469-8137.2007.02078.x) [DOI] [PubMed] [Google Scholar]
  • 44. Zhao N, et al. 2016. Coordinated pattern of multi-element variability in leaves and roots across Chinese forest biomes. Glob. Ecol. Biogeogr. 25, 359–367. ( 10.1111/geb.12427) [DOI] [Google Scholar]
  • 45. Hao Z, Kuang YW, Kang M. 2015. Untangling the influence of phylogeny, soil and climate on leaf element concentrations in a biodiversity hotspot. Funct. Ecol. 29, 165–176. ( 10.1111/1365-2435.12344) [DOI] [Google Scholar]
  • 46. Bates D, Mächler M, Bolker B, Walker S. 2015. Fitting linear mixed-effects models using lme4. J. Stat. Softw. 67, 1–48. ( 10.18637/jss.v067.i01) [DOI] [Google Scholar]
  • 47. Gröemping U. 2006. Relative importance for linear regression in R: the package relaimpo. J. Stat. Softw. 17, 925–933. ( 10.18637/jss.v017.i01) [DOI] [Google Scholar]
  • 48. Field A, Miles J, Field Z. 2012. Discovering Statistics using R. London, UK: Sage Publications. [Google Scholar]
  • 49. Tian D, Yan ZB, Fang JY. 2021. Review on characteristics and main hypotheses of plant ecological stoichiometry. Chin. J. Plant Ecol. 45, 682–713. ( 10.17521/cjpe.2020.0331) [DOI] [Google Scholar]
  • 50. Wu Y, Liu H, Song Z, Yang X, Li Z, Hao Q, Liu L. 2017. Ecological stoichiometry of nitrogen, phosphorous, and sulfur in China’s forests. Acta Geochim. 36, 525–530. ( 10.1007/s11631-017-0225-8) [DOI] [Google Scholar]
  • 51. Zhao WZ, Xiao CW, Li MX, Xu L, Li X, Zhu XY, Cheng CJ, He NP. 2023. Biogeographic patterns of sulfur in the vegetation of the Tibetan Plateau. J. Geophys. Res. 128, e2022JG007051. ( 10.1029/2022JG007051) [DOI] [Google Scholar]
  • 52. Reich PB, Walters MB, Ellsworth DS. 1992. Leaf life‐span in relation to leaf, plant, and stand characteristics among diverse ecosystems. Ecol. Monogr. 62, 365–392. ( 10.2307/2937116) [DOI] [Google Scholar]
  • 53. Reich PB, Ellsworth DS, Walters MB, Vose JM, Gresham C, Volin JC, Bowman WD. 1999. Generality of leaf trait relationships: a test across six biomes. Ecology 80, 1955–1969. ( 10.1890/0012-9658(1999)080[1955:GOLTRA]2.0.CO;2) [DOI] [Google Scholar]
  • 54. Wright IJ, et al. 2004. The worldwide leaf economics spectrum. Nature 428, 821–827. ( 10.1038/nature02403) [DOI] [PubMed] [Google Scholar]
  • 55. Valverde-Barrantes OJ, Freschet GT, Roumet C, Blackwood CB. 2017. A worldview of root traits: the influence of ancestry, growth form, climate and mycorrhizal association on the functional trait variation of fine-root tissues in seed plants. New Phytol. 215, 1562–1573. ( 10.1111/nph.14571) [DOI] [PubMed] [Google Scholar]
  • 56. Wang Z, Xia CX, Yu D, Wu ZG. 2015. Low-temperature induced leaf elements accumulation in aquatic macrophytes across Tibetan Plateau. Ecol. Eng. 75, 1–8. ( 10.1016/j.ecoleng.2014.11.015) [DOI] [Google Scholar]
  • 57. Wang Z, Wu ZG, Wang Y, Yu D. 2018. Variations in species-level N:P stoichiometry of charophytes and aquatic angiosperms on the Tibetan Plateau. Front. Plant Sci. 9, 870. ( 10.3389/fpls.2018.00870) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 58. Smith SA, Donoghue MJ. 2008. Rates of molecular evolution are linked to life history in flowering plants. Science 322, 86–89. ( 10.1126/science.1163197) [DOI] [PubMed] [Google Scholar]
  • 59. Smith SA, Beaulieu JM. 2009. Life history influences rates of climatic niche evolution in flowering plants. Proc. R. Soc. B 276, 4345–4352. ( 10.1098/rspb.2009.1176) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 60. Hawkins BA, Rodríguez MÁ, Weller SG. 2011. Global angiosperm family richness revisited: linking ecology and evolution to climate. J. Biogeogr. 38, 1253–1266. ( 10.1111/j.1365-2699.2011.02490.x) [DOI] [Google Scholar]
  • 61. Flores O, et al. 2014. An evolutionary perspective on leaf economics: phylogenetics of leaf mass per area in vascular plants. Ecol. Evol. 4, 2799–2811. ( 10.1002/ece3.1087) [DOI] [PMC free article] [PubMed] [Google Scholar]
  • 62. Aerts R, Chapin FS. 2000. The mineral nutrition of wild plants revisited: a re-evaluation of processes and patterns. Adv. Ecol. Res. 30, 1–67. ( 10.1016/S0065-2504(08)60016-1) [DOI] [Google Scholar]
  • 63. van Ommen Kloeke AEE, Douma JC, Ordoñez JC, Reich PB, van Bodegom PM. 2012. Global quantification of contrasting leaf life span strategies for deciduous and evergreen species in response to environmental conditions. Glob. Ecol. Biogeogr. 21, 224–235. ( 10.1111/j.1466-8238.2011.00667.x) [DOI] [Google Scholar]
  • 64. Lobell DB. 2007. Changes in diurnal temperature range and national cereal yields. Agric. For. Meteorol. 145, 229–238. ( 10.1016/j.agrformet.2007.05.002) [DOI] [Google Scholar]
  • 65. Vázquez DP, Gianoli E, Morris WF, Bozinovic F. 2017. Ecological and evolutionary impacts of changing climatic variability. Biol. Rev. Camb. Philos. Soc. 92, 22–42. ( 10.1111/brv.12216) [DOI] [PubMed] [Google Scholar]
  • 66. Wang J, Liu D. 2023. Larger diurnal temperature range undermined later autumn leaf senescence with warming in Europe. Glob. Ecol. Biogeogr. 32, 734–746. ( 10.1111/geb.13674) [DOI] [Google Scholar]
  • 67. Zhang LJ, et al. 2024. Data from: Global leaf sulphur stoichiometry and the relationships with nitrogen and phosphorus: phylogeny, growth form and environmental controls. Dryad Digital Repository ( 10.5061/dryad.fj6q5741c) [DOI] [PubMed]
  • 68. Zhang LJ, et al. 2024. Supplementary material for: Global leaf sulphur stoichiometry and the relationships with nitrogen and phosphorus: phylogeny, growth form and environmental controls. Figshare ( 10.6084/m9.figshare.24354259) [DOI] [PubMed]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The data and code from this manuscript can be found at [67].

Supplementary material is available online [68].

Articles from Proceedings of the Royal Society B: Biological Sciences are provided here courtesy of The Royal Society

RESOURCES