https://orcid.org/0000-0002-2008-9626
Abstract
Precipitation significantly influences the composition and structure of grassland ecosystems, particularly in arid desert steppes. Stipa breviflora, as a keystone species, plays a crucial role in maintaining the stability of the desert steppe. However, the response of S. breviflora’s succession strategy to changes in precipitation within the community remains uncertain. Since 2016, this research was conducted in a desert steppe in Inner Mongolia, China, involving control precipitation (PCK), and increases of 50% (P50) and 100% (P100) in natural precipitation. We measured biomass, height and canopy cover, calculated the importance value (IV) by species, and assessed the photosynthetic parameters and leaf elemental content of S. breviflora in 2021 and 2022. Results showed that the increase of precipitation significantly reduced the IV of S. breviflora. The net photosynthetic rate, transpiration rate, stomatal conductance, aboveground biomass carbon content and aboveground biomass nitrogen of S. breviflora leaves grew considerably in experimental plots receiving more precipitation, while δ13C value of leaves decreased significantly. Linear regression analysis and structural equation model showed that although the increase of precipitation improved the adaptability of S. breviflora functional traits and increased its IV, a higher transpiration rate significantly contributed to the decrease in its IV. Consequently, our research reveals the succession strategy of S. breviflora and provides a theoretical basis for studying the response mechanisms of desert steppe plant communities to climate change.
摘要
降雨是影响草地生态系统组成和结构的主要因素,特别是在干旱的荒漠草原。短花针茅(Stipa breviflora)作为荒漠草原的建群种，对于维持荒漠草原的稳定至关重要。然而，短花针茅的演替策略对降雨变化的响应尚不明确。为此，我们于2016年起在内蒙古荒漠草原开展模拟降雨试验，包含对照(PCK)、增雨50% (P50)和增雨100% (P100)处理。在2021和2022年，本研究调查了群落内物种生物量、高度和盖度并计算了重要值，测定了短花针茅的光合参数和叶片元素含量。结果表明，降雨量的增加显著降低了短花针茅的重要值；随着降雨量的增加，短花针茅叶片的净光合速率、蒸腾速率、气孔导度、地上生物量碳含量和氮含量显著增加，δ13C值显著降低。线性回归分析和结构方程模型表明，虽然降雨量的增加提高了短花针茅功能性状的适应性并增加了其重要值，但较高的蒸腾速率是导致其重要值下降的主要原因。总之，本研究揭示了短花针茅的演替策略，为研究荒漠草原植物群落对气候变化的响应机制提供了理论依据。
INTRODUCTION
Climate change in mid-latitudes is primarily reflected by increased annual average rainfall and more frequent extreme rainfall events, particularly intensifying in northwest China (Chen et al. 2023; IPCC 2021). Simultaneously, the increased frequency of extreme climatic events heightens the uncertainty in grassland communities’ responses to precipitation changes (Knapp et al. 2016). As a crucial abiotic factor, changes in precipitation significantly regulate the functionality and stability of community ecosystems directly through plant physiological and ecological processes, and indirectly by influencing species diversity and morphology, thereby maintaining assembly processes and interactions within plant communities (Michaletz et al. 2018). In response to climate change, plants interact with the environment, forming physiological and morphological adaptation strategies, and reducing the adverse effects of environmental changes through changes in community vegetation features and functional traits (Ma et al. 2019; Sun et al. 2023; Wang et al. 2018b).
Precipitation plays an important role in regulating plant community structure and species composition, and affects ecosystem function and its potential feedback (Li et al. 2021; Yang et al. 2011). However, different functional plants have different responses to precipitation change. Chen et al. (2018) found that increasing precipitation significantly increased the relative abundance of annual herbs (AHs) in the community, but had no effect on the relative abundance of perennial grasses (PGs) and semi-shrub (SS) plants, thus changing the plant community structure. Some studies also showed that shrubs absorbed water more easily than perennial herbs with the increase of precipitation, which increased their coverage (Gherardi and Sala 2015). In addition, the increase in the proportion of forbs in the community due to the increase in precipitation was significant due to their ability to obtain water quickly (Roumet et al. 2016). Thus, understanding how precipitation changes affect plant communities requires knowledge of the different response patterns of different functional plants.
Precipitation dominates the changes in functional traits of plants in grasslands (Ma et al. 2022). Among various functional traits, leaf nutrient resorption (Killingbeck 1996) and photosynthetic characteristics (Ikkonen et al. 2021) have been considered important for evaluating plant adaptation to nutrient-limited habitats. A study in California grassland showed that precipitation was positively correlated with plant leaf nitrogen content (Sandel and Low 2019), while a precipitation control experiment in Inner Mongolia showed that precipitation increased plant leaf carbon content and decreased leaf nitrogen content (Cheng et al. 2021). Some scholars have found that there was no significant linear correlation between plant leaf nitrogen content and precipitation (Zhang et al. 2018). Additionally, the variation in photosynthetic pathways leads to differing plant responses to precipitation change, and C3 plants have photosynthetic advantages and higher water use efficiency in cool and humid conditions, whereas C4 plants favor warm, dry environments (Wu et al. 2024). However, higher transpiration in humid environments (Wang et al. 2018a) leads to higher water loss, which may reduce the response of vegetation production to increased precipitation. Ye et al. (2020) demonstrated that the plant communities’ adaptability to precipitation change primarily relied on the functional characteristics of dominant species. It is thus crucial to explore the functional characteristics of dominant species to predict how the desert steppe will adapt to future climate changes (Funk et al. 2017).
Stipa breviflora is a keystone species crucial for maintaining the stability of the ‘fragile’ desert steppe ecosystem (Cui et al. 2024). However, the response of S. breviflora’s succession strategy to community-level precipitation changes remains uncertain. With these in mind, in 2016, we initiated a ‘long-term’ precipitation experiment in Inner Mongolia’s eastern part of the Eurasian temperate steppe. After 6 years of simulated precipitation enhancement (i.e. CK, +50% and +100% of ambient precipitation), we measured biomass, height and canopy cover and calculated the importance value (IV) by species. We found that while increases in precipitation enhanced S. breviflora’s biomass, height and canopy cover, its IV decreased. To explain this phenomenon, we further measured the morphological and physiological characteristics of S. breviflora in 2021 and 2022. We hypothesize that increased precipitation is beneficial to the increase of S. breviflora leaf nutrient content and photosynthesis, given it is a C3 plant(Lü et al. 2021), but it is also accompanied by an increase in transpiration (Zhou et al. 2020), which weaken
MATERIALS AND METHODS
Study site
The study was carried out in a desert steppe (41°46ʹ43.6ʹʹ N, 111°53ʹ41.7ʹʹ E) in Siziwang Banner, Inner Mongolia. Annual precipitation was primarily concentrated from May to September, totaling 173.26 mm in 2021 and 168.2 mm in 2022, and the average temperature in both years was 4.6 °C (Supplementary Fig. S1). The vegetation is sparse, with low grass coverage and a relatively simple plant species composition. Stipa breviflora is a keystone species, and Cleistogenes songorica and Artemisia frigida are dominant species. The main associated species are Convolvulus ammannii, Caragana stenophylla, Kochia prostrata, Leymus chinensis and Neopallasia pectinata. All plants were divided into four functional groups based on their life forms: PGs, SS, perennial forbs (PFs) and AHs (Bai et al. 2004) (Supplementary Table S1).
Experimental design
We considered that the precipitation in western parts of northern China would be increasing, the average annual precipitation in the experimental site for many years was 223 mm (2004–2016) and the maximum precipitation was 468.2 mm (2012), which was about twice the average precipitation. In 2016, we initiated a simulated precipitation experiment on a flat plot with uniform vegetation. The experiment involved three treatments: natural precipitation (PCK), a 50% increase (P50) and a 100% increase (P100) in natural precipitation, each replicated three times. To prevent water runoff, we separated each treatment using 40 cm high galvanized iron sheets buried underground. According to the calculation of native precipitation (Gro-Weather® software version 12, Davis Instruments Corporation, USA), the water in the rainwater collection bucket was sprayed on P50 and P100 plots manually and evenly every 15 days from May to October each year. Precipitation addition amount from 2017 to 2022 is given in Supplementary Table S2.
Determination of soil water content
At the end of July, mid-August and end of August, PR2 (DELTA-T DEVICES LTD, UK) was used to measure the soil water content (SWC) of the 0–10 cm soil layer.
Determination of characters of plant community
In 2016, we established a 1 m × 1 m fixed quadrat in each plot to assess species richness based on the number of species present. In mid-August 2021 and 2022, we measured the height and coverage of all species. Then we gathered five plants with similar shapes outside the fixed quadrat and calculated the average dry weight to estimate the biomass of each species within the fixed quadrat.
Determination of leaf photosynthetic parameters of S. breviflora
In mid-August of 2021 and 2022, three healthy and of consistent size S. breviflora specimens from each quadrat were selected randomly to measure net photosynthetic rate (Pn), stomatal conductance (Gs), intercellular carbon dioxide (Ci) and transpiration (Tr) using portable open system infrared gas analyzer ((Li-Cor, Inc., Lincoln, NE, USA) on the clear and cloudless morning (9:00 a.m. to 11:00 a.m.). After preheating the instrument, we placed and secured the chosen leaves in the leaf chamber. Then, we exposed them to an artificial light source (6400-02B) with a red/blue ratio of 9:1. The light intensity, CO2 concentration, temperature and relative humidity in the chamber were maintained at 1500 μmol m−2 s−1, 380 μmol m−2 s−1, 25 °C, and 50%–60% (Yu et al. 2014).
Determination of leaf element contents of S. breviflora
Plants previously assessed for photosynthetic parameters were sampled and transported to the laboratory for further analysis. Initially, we selected disease and pest-free leaves, rinsed them three times with ultrapure water and then dried them at a constant 65 °C for 48 h until they reached a constant weight. Dried plant samples were then ground with a ball mill.
After the above treatment, the total carbon (TC), total nitrogen content (TN) and stable carbon isotope values (δ13C) of plant leaves were determined by a stable isotope mass spectrometer (Isoprime100, Elementar, Germany) connected to an element analyzer (Vario Isotope Select).
The aboveground biomass carbon (AGBC) and aboveground biomass nitrogen (AGBN) were calculated as follows:
δ13C is a relative ratio, and the formula is explained as follows:
Statistical analysis
Using Curtis and McIntosh (1951) formula 1 for calculating the IV, and adapting it to the actual growth conditions of desert steppe plants, we modified the original formula by removing frequency and density and incorporating biomass and height. The original formula is shown in equation (A.4) in the appendix.
Where Br, Cr and Hr are relative biomass, relative coverage and relative height, respectively. The calculation formulas for Br, Cr and Hr are shown in equations (A.1–3) in the appendix.
The comparison of the IV, photosynthesis parameters (Pn, Tr, Gs and Ci) and elemental content (AGBC, AGBN and δ13C) of S. breviflora of different precipitation treatments all used one-way ANOVA (SAS Institute Inc., Cary, NC, USA), and linear regression analysis was used to explore the relationships between the above indicators and precipitation amount. Simple linear regressions assess the relationships between IV and Pn, Tr, Gs, Ci, AGBC, AGBN and δ13C, we only displayed the indicators that were significant with IV. Then, we used a stepwise regression model to expose the critical elements that affected IV. The above graphics were all drawn in Origin Pro 2021 (OriginLab, MA, USA).
Given the above results, we constructed a structural equation model (SEM) in AMOS 21.0 (Amos Development Co., Armonk, NY, USA) to understand the direct and indirect effects of precipitation changes on the IV. We made a prior model (Supplementary Fig. S2). Then we removed the non-significant paths in turn until we had the final model. Chi-squared (χ2) test, P value and root mean square error of approximation (RMSEA) were employed to optimize the final model of SEMs.
RESULTS
Changes of IV of S. breviflora in the community
The increase of precipitation significantly increased the SWC, biomass, height and coverage of S. breviflora, but there was no difference between years (Supplementary Table S3). With the increase of precipitation gradient, the IV of S. breviflora decreased, but it did not reach a significant level in 2021 and 2022 (Fig. 1a). However, linear regression analyses indicate there was a significant negative correlation between precipitation amount and IV of S. breviflora (P < 0.05, Fig. 1b). At the same time, the decrease in the proportion of PG represented by S. breviflora in the community was accompanied by the increase in the other three functional groups (PF, SS and AH). According to the types of photosynthesis, the proportion of C3 plants in the three precipitation gradients was 62.51%, 70.85%, and 72.56% and that of C4 plants was 37.49%, 29.15%, and 27.45%, respectively (Supplementary Table S4).
Changes of importance value of Stipa breviflora under different precipitation gradients (a). CK: natural precipitation; P50: increasing precipitation by 50%; P100: increasing precipitation by 100%. Simple linear regression analyses between importance value and precipitation amount (b). Regression coefficients (R2) and P value are provided for simple linear regression.
Response of physiology parameters of S. breviflora to precipitation treatments
Increased precipitation significantly affected most photosynthetic parameters of S. breviflora leaves. We did observe that the net photosynthetic rate, transpiration rate and stomatal conductance of S. breviflora leaves grew considerably in experimental plots receiving more precipitation in both 2021 and 2022 (Fig. 2a–f); however, intercellular CO2 concentration was not sensitive to increased precipitation neither 2021 and 2022 (Fig. 2g and h).
Effects of precipitation changes on the net photosynthetic rate (a), transpiration rate (c), stomatal conductance (e) and intercellular CO2 concentration (g) of Stipa breviflora. Simple linear regression analyses between net photosynthetic rate (b), transpiration rate (d), stomatal conductance (f), intercellular CO2 concentration (h) and precipitation amount. Significance level: ***P < 0.001, **P < 0.01 and *P < 0.05.
In these 2 years, significant increases were observed in the AGBC content of S. breviflora leaves across different precipitation treatments (Fig. 3a and b). Although in 2021, the AGBN of S. breviflora reached the maximum at P50 (Fig. 3c), according to the comprehensive data of 2 years, the AGBN showed an upward trend along with the increase of precipitation (P < 0.05, Fig. 3d). We also observed a significant negative correlation between δ13C value and precipitation (P < 0.05, Fig. 3e and f).
Effects of precipitation changes on the aboveground biomass carbon (a), aboveground biomass nitrogen (c) and stable carbon isotope (e) of Stipa breviflora. Simple linear regression analyses between aboveground biomass carbon (b), aboveground biomass nitrogen (d),stable carbon isotope (f) and precipitation amount. Significance level: *P < 0.05.
Factors causing the change of IV
Net photosynthetic rate, transpiration rate and stomatal conductance of S. breviflora leaves were negatively correlated with its IV (P < 0.05, Fig. 4). Stepwise regression models revealed that coverage, transpiration rate and AGBN were the most influential factors on IV (Table 1).
Summary of stepwise regression models to reveal IV determined by morphological and physiological properties
| Model | Equation | R2 | F | P |
|---|---|---|---|---|
| 1 | IV = −3.368Tr + 48.908 | 0.348 | 10.057 | ** |
| 2 | IV = −4.865Tr + 22.235AGBN + 36.480 | 0.746 | 25.916 | *** |
| 3 | IV = −5.230Tr + 18.33AGBN + 0.677C + 26.545 | 0.820 | 26.869 | *** |
| Model | Equation | R2 | F | P |
|---|---|---|---|---|
| 1 | IV = −3.368Tr + 48.908 | 0.348 | 10.057 | ** |
| 2 | IV = −4.865Tr + 22.235AGBN + 36.480 | 0.746 | 25.916 | *** |
| 3 | IV = −5.230Tr + 18.33AGBN + 0.677C + 26.545 | 0.820 | 26.869 | *** |
AGBN, aboveground biomass nitrogen; C, coverage; IV, importance value; Tr, transpiration rate. Significance level: ***P < 0.001 and **P < 0.01.
Summary of stepwise regression models to reveal IV determined by morphological and physiological properties
| Model | Equation | R2 | F | P |
|---|---|---|---|---|
| 1 | IV = −3.368Tr + 48.908 | 0.348 | 10.057 | ** |
| 2 | IV = −4.865Tr + 22.235AGBN + 36.480 | 0.746 | 25.916 | *** |
| 3 | IV = −5.230Tr + 18.33AGBN + 0.677C + 26.545 | 0.820 | 26.869 | *** |
| Model | Equation | R2 | F | P |
|---|---|---|---|---|
| 1 | IV = −3.368Tr + 48.908 | 0.348 | 10.057 | ** |
| 2 | IV = −4.865Tr + 22.235AGBN + 36.480 | 0.746 | 25.916 | *** |
| 3 | IV = −5.230Tr + 18.33AGBN + 0.677C + 26.545 | 0.820 | 26.869 | *** |
AGBN, aboveground biomass nitrogen; C, coverage; IV, importance value; Tr, transpiration rate. Significance level: ***P < 0.001 and **P < 0.01.
Simple linear regression analyses between importance value and net photosynthetic rate (a), transpiration rate (b) and stomatal conductance (c). Regression coefficients (R2) and P values are provided for simple linear regression.
SEM (chi-square = 44.114, P = 0.533, RMSEA = 0.000, R2 = 0.90) fitted the variance best and explained 90% of the variance in IV of S. breviflora. The increase in precipitation promoted the coverage and AGBN of S. breviflora and then increased its IV. In addition, the increase in precipitation had a directly strong negative effect on the IV. We also found that the reinforcement of leaf transpiration rate with the increase of precipitation was the decisive reason for the decrease of the IV of S. breviflora, and stomatal conductance had a significant positive correlation with transpiration (Fig. 5).
(a) The structural equation model (SEM) assesses how influencing factors impact importance value (IV). The thicknesses of arrows indicate the strength of the relationship, supplemented by a standardized regression coefficient, and the dotted line indicates that the relationship is not significant. Significance level: ***P < 0.001, **P < 0.01 and *P < 0.05. (b) The total standardized effects of the driving factors on IV. P, precipitation changes; C, coverage; Gs, stomatal conductance; Tr, transpiration rate; AGBN, aboveground biomass nitrogen.
DISCUSSION
This study explored the influence of long-term simulated precipitation changes on the succession strategy of keystone species in the desert steppe. Our results showed that the S. breviflora functional traits had good adaptability to precipitation changes; however, the significant increase of leaf stomatal conductance caused by the precipitation addition resulted in greater water loss, which was a leading reason for the decrease of the IV of S. breviflora.
The primary factor constraining plant growth in desert steppes is the insufficient SWC (Song et al. 2024). A large amount of rainwater input eased the resource limitation and provided living space for more species (Cheng et al. 2023). In our study, increased precipitation weakened the IV of PGs represented by S. breviflora, and was accompanied by the increase of IV of SS, PFs and AHs in the community. This is similar to the results of Liu et al. (2018), that is, the proportion of constructive species is nonlinearly negatively correlated with precipitation, and the increase of precipitation provides living space for other functional group species. Studies have shown that AHs can complete their basic life cycle under the condition of 5 mm effective precipitation, and the increase of precipitation is beneficial to activate the dormancy seed germination of AHs (Bai et al. 2021), thus increasing its IV in the community. Gherardi and Sala (2015) found that shrubs are more competitive than grasses, and the increase of precipitation is beneficial to the increase of shrub importance. In addition, precipitation addition treatments increased the proportion of C3 plants and decreased C4 plants in the community, which showed that C3 plants have stronger photosynthetic capacity in the humid conditions.
AGBC and AGBN are important parts of terrestrial carbon and nitrogen storage, reflecting the balance between carbon and nitrogen retention and carbon and nitrogen loss in ecosystem (Hu et al. 2022). In our study, the biomass carbon and nitrogen of S. breviflora leaves were positively correlated with precipitation amount; this finding proves that increased precipitation is beneficial to the nutrient storage of S. breviflora leaves. Leaf nitrogen is one of the essential nutrients for the synthesis of photosynthetic-related enzymes and structural proteins; higher leaf nitrogen content means a stronger photosynthesis rate (Ordoñez et al. 2009). In our study, the net photosynthesis of S. breviflora was more intense in the plots receiving more rain, which was the same as most of the research results (Diao et al. 2023; Lü et al. 2021). The increase in photosynthesis rate led to the production of organic matter and enhanced the growth of S. breviflora. Therefore, AGBN and net photosynthetic rate promoted the IV of S. breviflora.
However, in the desert steppe, the supply of soil nitrogen to plants is limited. Our previous research showed that the increase in precipitation significantly reduced soil inorganic nitrogen (Cui et al. 2024). Actually, the higher stomatal conductance in humid environments leads to higher transpiration and therefore can improve soil nutrient uptake (Wang et al. 2018a). Concurrently, we observed a decrease in the δ13C value of S. breviflora with increased precipitation, indicating reduced water use efficiency as precipitation levels rise (Cao et al. 2020; Ehleringer, 1993; Zhou et al. 2021). Therefore, we conceived a conceptual framework that plants may absorb more soil nutrients in humid environments because high transpiration is accompanied by water loss. Our framework was confirmed by N and P nutrient addition experiments in meadows with high precipitation (Ren et al. 2009; Song et al. 2017). These indicated that plant growth is more limited by soil nutrients than by water under high precipitation conditions in the desert steppe.
CONCLUSIONS
This study demonstrated the impact of precipitation levels on the role of keystone species in the desert steppe. Results showed that with the increase in precipitation, the IV of S. breviflora decreased. The increase in precipitation harmed the IV of S. breviflora by boosting its transpiration rate. At the same time, the decrease in the proportion of PGs represented by S. breviflora is accompanied by the increase in the proportion of SS, PFs and AHs. And the increase of precipitation may promote the transformation of C4 plants to C3 plants. Our research focuses on the changes in both the morphological and physiological characteristics of S. breviflora, providing a theoretical basis for predicting desert steppe plant succession under precipitation change.
Supplementary Material
Supplementary material is available at Journal of Plant Ecology online.
Table S1: Community species composition.
Table S2: Growing season precipitation (mm) and precipitation applied in each experimental treatment from 2017 to 2022.
Table S3: Two-way ANOVAs is used to evaluate the effects of year (Y) and precipitation (P) and their interaction on the soil water content (SWC) and morphological characteristics of Stipa breviflora.
Table S4: Response of species composition to precipitation changes, divided by functional groups and photosynthetic types.
Figure S1: Growing season monthly precipitation and temperature from 2017 to 2022.
Figure S2: The prior structural equation model relating precipitation amount, coverage, stomatal conductance, transpiration rate, net photosynthetic rate, aboveground biomass nitrogen, and importance value of Stipa breviflora.
Authors’ Contributions
Yuan-Yuan Cui (Conceptualization, Investigation, Methodology, Software, Visualization, Writing—original draft), Liu Bai (Formal analysis, Investigation, Visualization), Dong-Jie Hou (Data curation, Visualization), Zhong-wu Wang (Conceptualization, Funding acquisition, Writing—review & editing), Jing Wang (Data curation, Visualization), Zhi-Qiang Qu (Data curation, Visualization), Yun-Bo Wang (Data curation, Visualization), Guo-Dong Han (Project administration, Supervision), Zhi-Guo Li (Methodology, Validation), Hai-Yan Ren (Methodology, Validation), and Hai-Ming Wang (Data curation, Resources)
Funding
This work was supported by the National Natural Science Foundation of China (32460353, U23A2001, 31560140), the Key Project of Science and Technology in Inner Mongolia of China (2021ZD0044) and the Interdisciplinary Fund Project of Inner Mongolia Agricultural University (BR22-14-04).
Acknowledgements
We would like to thank Dr Daniel Petticord at the University of Cornell for his assistance with English language and grammatical editing of the manuscript. We thank the anonymous reviewers, as well as team members for collecting data in the field.
Conflict of interest statement. The authors declare that they have no confict of interest.
