1. Introduction

Nitrogen (N) and phosphorus (P) play critical roles in plant photosynthetic and metabolic processes (Liu, 2020), as well as ecosystem function and services (Reich and Oleksyn, 2004; Yuan et al., 2011; Tian et al., 2019; Du et al., 2020). For example, the N element in Rubisco and the P element in ribosomal RNA are the main essential nutrients that drive the photosynthesis and reproduction of plants (Tian et al., 2018), respectively. However, N and P availability have been shown to limit plant growth, development, and primary production at regional and global levels (Du et al., 2020; Hou et al., 2020). In view of this, plants have been shown to vary leaf N and P stoichiometric patterns (e.g., concentrations of N, P and N/P ratios) to adapt to changing environments (Han et al., 2005; Yuan and Chen, 2015), and several hypotheses have been proposed to explain patterns of plant leaf N and P. Specifically, stoichiometric patterns of leaf N and P are closely related to variations in climatic conditions (especially of temperature), biogeochemical gradients (especially of soil P availability) and plant species (evolutionary history) (Reich and Oleksyn, 2004; Yang et al., 2016; Tian et al., 2018).

The temperature-plant physiological hypotheses state that species with higher growth and reproduction rates require higher P-rich RNA to support rapid protein synthesis, thus leading to a low N/P ratio, especially at high-latitude and low-temperature environments, such as those on the Tibetan Plateau (Reich and Oleksyn, 2004; Tian et al., 2018). The biogeochemical hypothesis proposes that geographic patterns of stoichiometry (e.g., leaf N and P) is controlled by soil substrate age (Lambers et al., 2008; Quesada et al., 2010). Specifically, nutrient limitations differ between young soil found at higher latitudes and old soil found in the tropics (McGroddy et al., 2004; Reich and Oleksyn, 2004). This hypothesis predicts that P in older tropical soil is expected to be more leaching by rainfall than N, consequently, low soil P availability (Vitousek and Farrington, 1997; Aerts and Chapin, 2000) and high leaf N/P ratio in tropical ecosystems (Reich and Oleksyn, 2004). Variation in species composition and structure can also strongly affect concentrations of leaf N and P (Guo et al., 2017; Tian et al., 2018). Previous studies have shown that closely related plant functional groups (i.e., plants with short phylogenetic distance) may share general stoichiometric patterns of N and P because of similar evolutionary histories and adaptations to environmental stresses (Wright et al., 2005; Yang et al., 2016). Despite some support for these hypotheses (Han et al., 2011; Yang et al., 2016), none of them can completely explain the global patterns of leaf N and P. In other word, these previous studies have improved our understanding of biogeochemical cycles and factors that regulate leaf N and P stoichiometry (Yuan and Chen, 2015; Diaz et al., 2016; Tian et al., 2018), but it remains unclear whether there are generalizable nutrient balance strategies for different plant functional groups (e.g., monocots and dicots).

N and P limitation strongly affect ecological evolution and adaptation of species (Reich and Oleksyn, 2004; Kerkhoff et al., 2006; Tian et al., 2019). One important adaptation to the environment is the strategy plants use to maintain stoichiometric homeostasis. Homeostasis indicates the ability of organisms to maintain stable levels of nutrients in response to fluctuations in the environment, which reflects the physiological and biochemical adaptation of organisms to their surrounding environment (Jeyasingh et al., 2009). Plant functional groups differ in their ability to absorb, transport, distribute, utilize and release nutrients to adapt to the environment (Sistla et al., 2015; Guo et al., 2017). For example, studies have shown that the characteristics of N and P stoichiometric homeostasis vary among species, with heterotrophs significantly more homeostatic in terms of N/P ratio than autotrophs (Persson et al., 2010). Moreover, previous studies have demonstrated that plant species adapted to N-limited environments deploy a "conservative" nutrient utilization strategy to survive, including a higher C/N ratio, small leaf area and greater stress tolerance; in contrast, species that growth in less N-limited environments deploy an "opportunistic" approach, with a lower C/N ratio, large leaf area, flexible competitor-stress tolerator-ruderal strategy (Zhang et al., 2020; Zhou et al., 2021a,b).

Here, we speculate that monocots and dicots might have evolved into contrasting strategies to balance N and P in response to changes in climate and soil nutrient availability. Thus, monocots deploy a "conservative" nutrient utilization strategy, whereas dicots deploy an "opportunistic" nutrient utilization strategy. To test this hypothesis, we use a global database of leaf N and P that includes climate and edaphic factors to characterize global patterns of leaf N, P and N/P ratio in monocots and dicots. We also explore the sensitivity of stoichiometry to environment factors in monocots and dicots. We predict that monocots maintain high stoichiometric homeostasis and narrow climate-soil nutrient niches, and that their nutrient contents are less sensitive to environmental changes. In addition, we predict that dicots have low stoichiometric homeostasis and wide climate-soil niches because their strong adaptability in changing environments (Zhou et al., 2021a, 2021b). Our results will provide important insights into the nutrient balance of plants in various habitats, and will be beneficial to understanding the biological survival strategies of different plants to environmental changes.

2. Materials and methods 2.1. Data collection

The database used in our study was constructed in a previous study by Tian et al. (2019). This global database includes information on leaf N and P concentrations of terrestrial plants (including 201 families, 1265 genera and 3227 species from 1291 sites with 11, 354 individual records (Fig. S1). The database also contains information on habitat type (e.g., grassland, natural forest, wetland ecosystems and shrubland) and various environmental factors (e.g., mean annual temperature (MAT), mean annual precipitation (MAP), soil total N (STN) and soil total P (STP).

2.2. Data analysis

We created two databases, one for monocots and one of dicots according to the classification of cotyledons in the original database. These new databases included 1483 individual records of monocots and 9322 individual records of dicots.

We then determined the frequency distributions of leaf N, P and N/P ratio for monocots and dicots in SPSS 19.0 (SPSS Inc., Chicago, IL, USA) and plotted our findings in Origin (2021) (OriginPro, Version, 2021. OriginLab Corporation, Northampton, MA, USA). One-way ANOVA analysis was used to compare the relationships between leaf N, P and N/P ratio, STN, STP, MAT and MAP in monocots and dicots (p < 0.05).

We used linear model to analyse the relationships between leaf nutrients (N, P and N/P ratio) and climate of MAT and MAP (1483 individual records), between leaf nutrients and soil properties (STN and STP), between soil nutrients and climate, between MAT and MAP for both monocots and dicots. Notably, 508 individual records of monocots and 3223 individual records of dicots that included soil variables of N and P were used in regression analysis.

We further analysed the stoichiometric homeostasis of monocots and dicots. The stoichiometric homeostasis index (H) was calculated by the following Eq (Persson et al., 2010):

where y is the concentration of leaf nutrients (N, P and N/P ratio), x refers to the supply of nutrients in the environment, and c is a constant.

To determine the climate and soil nutrient niches of monocots and dicots, general principal component analysis (PCA) was employed with the whole set of climatic factors (MAT and MAP) and edaphic factors (STN and STP). Total climate and soil nutrient niche space of the monocots and dicots were assessed by the valuation of the n-dimensional hypervolume (de la Riva et al., 2017). To reduce the dimensions of the hypervolume, the first of the three PCA axes was used to calculate the hypervolume for monocots and dicots with a multidimensional kernel density estimation procedure (de la Riva et al., 2017). A rarefaction analysis was also used to control the effects of climate and soil factors on the hypervolume. Finally, the hypervolumes of monocots and dicots were calculated.

We also conducted a phylogenetic analysis on all species to determine whether there were significant phylogenetic signals in the stoichiometry of leaf N, P and N/P ratio (Blomberg et al., 2003). Phylogenies were created in the R package V.PhyloMaker (Jin and Qian, 2022). We calculated Pagel's λ values with the packages of phytools and 'Picante' in software R and assessed the strength of the phylogenetic signal for monocots and dicots (Huang et al., 2023). These analyses were conducted in R 4.2.0 (R Development Core Team, Vienna, Austria).

3. Results 3.1. Leaf N and P stoichiometry in monocots and dicots

Leaf N and N/P ratio were higher in dicots than in monocots, but leaf P was higher in monocots than in dicots (Figs. 1 and S2). In monocots, leaf N concentrations ranged between 4.50 and 59.13 mg/g (mean value = 20.25 mg/g), P concentrations between 0.10 and 7.37 mg/g, (mean value = 1.60 mg/g), and N/P ratio between 2.09 and 92.22 (mean value = 16.13). In dicots, leaf N concentrations ranged between 2.48 and 66.80 mg/g (mean value = 21.41 mg/g), P concentrations between 0.08 and 9.59 mg/g (mean value = 1.48 mg/g), and N/P ratio between 1.63 and 92.23 (mean value = 18.42).

3.2. Relationships between leaf N and P stoichiometry and climate

Our analyses indicated that climate was not strongly correlated with monocots leaf nutrient concentrations. Specifically, leaf N and MAT were weakly correlated (R²＝0.03, p < 0.0001, Fig. S3a), but there were no significant relationships between MAT and leaf P (Fig. S3b), MAT and leaf N/P ratio (Fig. 2a), MAP and leaf N (Fig. S3b), MAP and leaf P (Fig. S3d), or MAP and leaf N/P ratio (Fig. 2c). In contrast, in dicots, our analyses detected significant relationships between climate and leaf nutrients (p < 0.0001). Leaf N was negatively correlated to MAT (R²＝0.05, Fig. S4a) and MAP (R²＝0.03, Fig. S4c). Similarly, negative relationships between leaf P and MAT (R²＝0.15, Fig. S4b), between leaf P and MAP (R²＝0.10, Fig. S4d) were observed. Consequently, leaf N/P ratio was found to be positively correlated to MAT (R²＝0.13, Fig. 2b) and MAP (R²＝0.09, Fig. 2d). Taken together, these findings indicate that leaf N/P ratio is more sensitive to the dynamics of temperature and precipitation in dicots than in monocots.

3.3. Relationships between leaf N and P stoichiometry and soil nutrients

Soil P availability is more strongly correlated with the variation of leaf N/P in dicots than in monocots. For monocots, soil P availability was significantly correlated with leaf P (R²＝0.25, p < 0.0001, Fig. S5b) and with leaf N/P ratio (R²＝0.14, p < 0.0001, Fig. 2g). In dicots, soil P availability was significantly positively correlated with leaf P (R²＝0.18, p < 0.0001, Fig. S5d), but negatively correlated with leaf N/P ratio (R²＝0.20, p < 0.0001, Fig. 2h). Soil N has limiting effects on leaf N (Fig. S5a and c) and N/P ratio in both monocots and dicots (Fig. 2e and f).

3.4. Stoichiometric homeostasis, niche hypervolume and phylogenetic signal in monocots and dicots

Compared to monocots, dicots have low stoichiometric homeostasis, wide climatic-edaphic niches and high phylogenetic signals in stoichiometry. The estimated three-dimensional niche hypervolume of dicots is clearly larger than that of monocots (Fig. 3 a-c). In addition, the 1/Homeostasis index of leaf N/P ratio is lower in monocots (1/Homeostasis index = 0.20) than in dicots (1/Homeostasis index = 0.22, Fig. 3d). Moreover, significant (p < 0.05) phylogenetic signals of leaf N (λ = 0.65), P (λ = 0.57) and N/P (λ = 0.46) were observed in dicots, however, only significant phylogenetic signals of P were observed in monocots (Fig. 4).

4. Discussion 4.1. Divergent response of leaf nutrients to climate in monocots and dicots

Temperature-plant physiological hypotheses posit that plants acclimate and adapt to low temperatures by increasing N and P (Reich and Oleksyn, 2004). Here, we found that in dicots, variations in N, P and N/P ratio were significantly correlated with temperature and precipitation. However, our data indicate that in monocots variations in P and N/P ratio are not affected by temperature or precipitation. These results indicate that temperature-plant physiological hypotheses are unable to explain variations in plant N and P for all plant functional groups. This might be because the climate indirectly influences plant chemical traits by affecting the nutrients available in the soil and the species composition in the community (Yang et al., 2016; Guo et al., 2017). For example, soil P availability in soils where dicots grow was significantly correlated with temperature (R² = 0.33, p < 0.0001, Fig. S6c) and precipitation (R² = 0.21, p < 0.0001, Fig. S6d). However, no significant relationship was detected between soils where monocots grow and P and climate (temperature and precipitation). Taken together, these findings indicate that climate might have a weak modifying effect on plants nutrient balance (He et al., 2008; Yang et al., 2016).

4.2. Soil total P rather than N shapes the balance of leaf N/P ratio

The biogeochemical hypothesis suggests that nutrient content in plants is determined by the accessibility of soil nutrients (Reich and Oleksyn, 2004; Yang et al., 2016; Tian et al., 2018). Our results demonstrate that in monocots and dicots the concentration of P is significantly correlated with P availability in soil, but the concentration of N in monocots and dicots was not significantly affected by the N availability in soil. These findings not only indicate that soil P directly affects plant chemical traits, but also that P limits plant growth. Previous studies have also demonstrated that in natural terrestrial ecosystems the belowground and aboveground of plant production is limited more by P than by N (Du et al., 2020; Hou et al., 2020). Specifically, 43% and 18% of the global terrestrial ecosystems are limited by soil P and N (Du et al., 2020), respectively. Moreover, N/P ratios were higher in dicots (N/P＝18.42) than in monocots (N/P＝16.13), demonstrating that the growth of dicots is more limited by soil P availability. According to Liebig's law of the minimum (Hooker, 1917), plants are limited by resources with the smallest supply compared to demand. Consequently, we detected significant phylogenetic signals of leaf P, N and N/P ratio in dicots. The higher phenotypic plasticity of dicots may improve their adaptability to P-limiting environments. In summary, P is an integral component of global terrestrial ecosystem biogeochemistry that plays an important role in shaping the balance of plant nutrients (Tian et al., 2018; Du et al., 2020; Hou et al., 2020).

4.3. Niche segregation explains divergent stoichiometric homeostasis between monocots and dicots

N, P and N/P ratios are less sensitive to environmental dynamics in monocots than in dicots. In other words, monocots might be expected to have stronger stoichiometric homeostasis in response to environmental change. Our results confirm this conjecture. Specifically, we found that indexes of stoichiometric homeostasis for the N (Fig. S7a), P (Fig. S7b), and N/P ratio were higher in monocots than in dicots. Stoichiometric homeostasis of plants reflects the physiological and biochemical adaptation of organisms to environmental changes (Hessen et al., 2004; Elser et al., 2010), which is related to the ecological adaptability of species (Yu et al., 2011). Compared to dicots, our results further demonstrated that monocots grow in harsh environments with lower soil nutrients (Fig. S8 a and b), lower water-heat availability (Fig. S8 c and d) and lower hydrothermal concordance (R² = 0.26, p < 0.0001, Fig. S8e). Thus, the growth of monocots may be limited by environmental stresses, insufficient water-heat availability coupled with low soil N and P availability (Sun et al., 2020; Zhou et al., 2020). Together, these constraints may explain why monocots deploy a "conservative" nutrient utilization strategy (survival priority strategy) and maintain a relatively higher stoichiometric homeostasis (Persson et al., 2010; Zhang et al., 2020).

Dicots are more sensitive to environmental dynamics because of their wide climatic-edaphic nutrient niches and phylogenetic signals in stoichiometry. We found that dicots grow in an environment with more soil nutrients (Fig. S8 a and b), water-heat availability (Fig. S8 c and d), and higher hydrothermal concordance (R² = 0.66, p < 0.0001, Fig. S8f). Furthermore, our findings indicate that dicots exhibit relatively lower stoichiometric homeostasis and significant phylogenetic signals in stoichiometry under changing environments. These results support previous studies that showed dicots can adapt to resource- and water-deficient conditions with a growth priority strategy (Zhang et al., 2020; Zhou et al., 2021a). In other words, dicots can be distributed in both arid, nutrient-poor and humid, nutrient-fertile environments because of their wide climate and soil niches. These results are in accordance with our hypothesis that plants have evolved contrasting strategies to balance nutrients in response to changes in climate and soil nutrient availability (Fig. 5). In summary, the explanatory powers of climate conditions, biogeochemical and phylogenetic factors for plant nutrients are comparatively low. And the difference in environmental adaptability in plant nutrients may be determined by intrinsic genetic factors, physiological-ecological evolutionary history and niche segregation of plant functional groups (Sardans et al., 2015; Ma et al., 2018; Tian et al., 2018).

5. Conclusion

Here, we analyzed the variations of leaf N, P and N/P ratio in monocots and dicots along the environmental gradients. We found that plants have evolved contrasting strategies to balance nutrients in response to environmental variation. Monocots use a "conservative" nutrient utilization strategy and high stoichiometric homeostasis to tolerate dry-stressful environments with deficient water and soil nutrient availability, because of narrow climatic-edaphic nutrient niches and weak phylogenetic signals in stoichiometry. In contrast, dicots are more sensitive to the dynamics of climate and soil nutrients, and exhibit lower stoichiometric homeostasis that have allowed them to adapt to changeable environments with wide climatic and edaphic nutrient niches. The innovations of stoichiometric homeostasis, niche segregation and phylogenetic signals of plant functional groups along environmental gradients indicate that the ecological adaptation and evolution of plants played an important role in shaping biogeochemical cycles.

Acknowledgments

This research was supported by the National Science Foundation of China (Grant No. 32271774, 42301071), and the China Postdoctoral Science Foundation (Grant No. 2023M743633).

CRediT authorship contribution statement

Miao Liu: Writing – review & editing, Writing – original draft, Supervision, Resources, Methodology, Funding acquisition, Conceptualization. Tiancai Zhou: Writing – review & editing, Formal analysis. Quansheng Fu: Formal analysis.

Declaration of competing interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Appendix A. Supplementary data

Supplementary data to this article can be found online at https://doi.org/10.1016/j.pld.2024.08.002.