Halophytic succulence is a driver of the leaf non-structural carbohydrate contents in plants in the arid and hyper-arid deserts of northwestern China

Abstract

Background and Aims

Non-structural carbohydrates (NSC), primarily sugars and starch, play a crucial role in plant metabolic processes and the ability of a plant to tolerate and recover from drought stress. Despite their importance, our understanding of NSC characteristics in the leaves of plants that thrive in hyper-arid and saline environments remains limited.

Methods

To investigate the variations in leaf NSC across different species and spatial scales and to explore their possible causes, we collected 488 leaf samples from 49 native plant species at 115 sites in the desert area of northwestern China. The contents of soluble sugars (SS), starch and total NSC were then determined.

Key Results

The average contents of SS, starch and total NSC were 26.99, 60.28 and 87.27 mg g⁻¹, respectively, which are much lower than those reported for Chinese forest plants and global terrestrial plants. Herbaceous and woody plants had similar NSC levels. In contrast, succulent halophytes, a key component of desert flora, showed significantly lower leaf SS and total NSC contents than non-succulent plants. We observed a strong negative correlation between leaf succulence and SS content, suggesting a role of halophytic succulence in driving multispecies NSC pools. Environmental factors explained a minor portion of the spatial variation in leaf NSC, possibly owing to the narrow climatic variation in the study area, and soil properties, particularly soil salinity, emerged as more significant contributors.

Conclusions

Our findings increase the understanding of plant adaptation to drought and salt stress, emphasizing the crucial role of halophytic succulence in shaping the intricate dynamics of leaf NSC across diverse plant species in arid and hyper-arid environments.

INTRODUCTION

Carbohydrates are the primary products of photosynthesis, serving as the crucial substrates for plant growth and metabolism. In plants, carbohydrates can be classified into structural carbohydrates and non-structural carbohydrates (NSC). Structural carbohydrates, including lignin and cellulose, are mainly used for building cell walls and maintaining the physical integrity of plant tissues (Chapin et al., 1990). NSC, primarily composed of soluble sugars (SS; including glucose, sucrose, fructose, etc.) and starch, serve as the main source of carbon and energy reserves in plants (Blumstein et al., 2023). SSs provide immediate energy and substrates for various plant functions, including respiration, defence, phloem transport and osmoregulation. Starch, in contrast, serves as a long-term carbon store and a rapidly available energy source (Martínez-Vilalta et al., 2016). During periods of high photosynthesis, surplus SS are converted into starch for storage, and this starch must be broken down into SS when the plant requires additional energy or a substrate for osmotic adjustment (Signori-Müller et al., 2021). Thus, both the total amount of NSC and its composition (e.g. the SS:starch ratio) are crucial for understanding the balance between carbon sources and sinks in plants (Yang et al., 2023).

Given their essential roles, the content and variability of NSC differ among plant organs and life forms (e.g. trees, shrubs and herbs). Typically, the content and variability of NSC are higher in leaves than in stems and roots (Martínez-Vilalta et al., 2016; Fangel et al., 2024). The leaves are the primary site of photosynthesis (production of NSC) and exhibit higher metabolic activity (consumption of NSC) and greater responsiveness to environmental conditions than stems and roots (Luo et al., 2021). Studies have also demonstrated that leaf NSC differ significantly among plant functional groups, with herbaceous plants having higher levels than woody plants (such as trees and shrubs) owing to a greater proportion of living, metabolically active tissues in herbaceous leaves (Li et al., 2016). In contrast, leaves of woody plants have a higher content of structural carbohydrates (e.g. lignin, cellulose), making them more rigid, durable and resistant (Martínez-Vilalta et al., 2016). Therefore, understanding the roles and variations of leaf NSC is crucial for understanding plant resource allocation strategies and their responses to environmental changes.

The variation of leaf NSC results from complex physiological processes that are affected by multiple environmental factors, such as climate conditions (Li et al., 2016; Martínez-Vilalta et al., 2016), light intensity (Zepeda et al., 2022), soil properties (Xie et al., 2022) and water availability (Signori-Müller et al., 2021). On a large scale, leaf NSC usually increase with decreasing temperature and precipitation, which indicates that insufficient water and low temperatures constrain the demand by the plant for structural carbon and instead promote the accumulation of NSC in leaves (Li et al., 2016; Blumstein et al., 2023). At regional scales, gradients of climate, environmental stress and, especially, water availability become primary abiotic factors that affect leaf NSC (Signori-Müller et al., 2021). During periods of drought stress, starch becomes a crucial carbon reserve that buffers the plant against energy demand that exceeds the available SS, whereas SS are key osmolytes that contribute to osmotic adjustment (Tsuji et al., 2022). Therefore, significant attention has been devoted to determining the role of NSC in mediating plant drought tolerance and survival under drought stress. Both experimental studies and surveys have demonstrated that the enrichment of NSC and rapid conversion of starch to SS in leaves constitute important mechanisms by which trees tolerate drought stress (O’Brien et al., 2014; Tsamir-Rimon et al., 2021). However, previous studies have mainly focused on forest ecosystems; as a result, the association between leaf NSC and drought tolerance of plants in the driest environments on Earth remains unclear, and this limits our understanding of the mechanisms underlying plant adaptation to extremely arid conditions.

Deserts are arid and hyper-arid resource-poor environments, in which water availability is the greatest constraint on plant growth and distribution (Wang et al., 2022). Desert plants can survive and develop under severe drought and play a vital role in maintaining the structure and function of the entire desert ecosystem (Fei et al., 2022). However, empirical studies of leaf NSC in arid deserts are scarce, with research restricted to a single species at a given site. A previous study has found that when Haloxylon ammodendron (a typical desert shrub that exists in the arid regions of northwestern China) is subjected to drought stress, depletion of starch in its leaves does not necessarily coincide with an increase in SS, which suggests that the SS might be consumed to generate energy (e.g. for active transport) rather than becoming the primary osmotic contributors in some desert plant species (Yang et al., 2023).

Unlike plants in more humid ecosystems, plants in arid and hyper-arid deserts are usually exposed to extended and intense drought conditions; consequently, they have evolved a range of morphological and physiological traits to cope with drought (Li et al., 2021). Among these traits, halophytic succulence is a remarkable feature (Ogburn and Edwards, 2010; Griffiths and Males, 2017; Grace, 2019). Here, we use the leaf succulence index (LSI; water content per unit leaf area) to define the degree of succulence, whereby an LSI of >500 indicates that the leaves of a plant are thickened and fleshy (Medina et al., 2015; Wang et al., 2022). Different from the ‘iconic’ succulents (e.g. cacti and aloes), halophytic succulents are able to accumulate a substantial amount of salt ions, creating an osmotic gradient that lets them extract water from extremely dry or highly saline soils (Wang et al., 2017). Given that inorganic ions can act as osmoticum in halophytic succulents (Flowers et al., 2015), it is reasonable to expect that there might be a trade-off between leaf NSC and leaf succulence in desert plants. However, the leaf NSC characteristics of plants in arid deserts and their differences among plant functional groups are still poorly understood.

To address this knowledge gap, we performed extensive sampling of leaf NSC across one of the driest areas of China. This region is also one of the most representative desert ecosystems in Central Asia (Lu et al., 2019). The arid climate and diverse topography generate a range of desert types, including sandy deserts, gravel deserts (Gobi deserts) and saline deserts (Luo et al., 2021). In particular, soils of saline desert generally contain toxic levels of soluble salts (Wang et al., 2020). Plants that thrive in this region typically exhibit xerophytic and halophytic adaptations to minimize water loss and tolerate salt toxicity, with succulent halophytes being key components of the plant communities (Supplementary Data Figs S1 and S2). We collected 488 leaf samples from 49 native species (Supplementary Data Table S1) at 115 sites across a range of desert habitats within the study area in northern China (Fig. 1; Supplementary Data Table S2). To the best of our knowledge, this study is the first systematic survey of leaf NSC contents in plants from arid and hyper-arid deserts. In addition to the total NSC contents, we also quantified the starch and SS contents, because these components all play distinct roles in stress responses. Our goals were as follows: (1) to document the characteristics of leaf NSC (starch, SS, SS:starch ratio and total NSC) and their differences among species and plant functional groups; (2) to investigate variations in leaf NSC among desert habitats; and (3) to quantify the influence of environmental factors on leaf NSC and ascertain whether there are spatial patterns of leaf NSC within the study area. Based on our literature review, we hypothesized that:

Fig. 1.

(A) Locations of the study area and the 115 sampling sites superimposed on a digital elevation model of the Xinjiang region at a spatial resolution of 500 m. (B–D) Examples of gravel desert (B), sandy desert (C) and saline desert (D) in the study area. The photographs were taken by the authors in July 2021.

1. Given that most of the succulent plants in the study area are halophytic, they are expected to have lower leaf SS content compared with non-succulent plants.

2. Owing to the toxic levels of salts in saline desert soils, plants might require additional energy to counteract ion toxicity, resulting in distinct features in leaf NSC content and composition compared with plants in other habitats.

3. Given the relatively small climatic gradient in our study area, the spatial variation in leaf NSC would be driven primarily by soil properties that could create environmental constraints.

MATERIALS AND METHODS

Study area

Our research was conducted in the Xinjiang Uygur Autonomous Region, one of the driest regions in China, with extremely low annual precipitation. Xinjiang accounts for >60 % of the total desert area of China and encompasses a diverse array of desert types, including the expansive Taklamakan Desert and the rocky terrain of several Gobi Deserts (Lu et al., 2019). In conjunction with limited precipitation, high evapotranspiration leads to frequent accumulation of salts in the soil; consequently, Xinjiang also hosts extensive areas of salinized desert (Luo et al., 2021). These desert types offer a diverse glimpse into some of the most extreme and captivating desert landscapes in the world. Our study area encompasses gravel, sandy and saline deserts that span >1000 km from east to west in Xinjiang (Fig. 1). The climate in the study area is arid to hyper-arid, with mean annual precipitation (MAP) ranging from 40 to 200 mm and mean annual temperature (MAT) ranging from 5.3 to 12.1 °C. The soils were primarily Aridosols and Entisols according to the United States Department of Agriculture soil taxonomy. Owing to the harsh climate, the vegetation structure is fairly simple and consists primarily of small shrubs and herbs, with a majority being xerophytes and halophytes (Supplementary Data Figs S1 and S2).

Design and field sampling

In July 2021, we established 115 sampling sites in the study area (Fig. 1). At each site, we established a 10 m × 10 m plot in a flat area to minimize variations in vegetation owing to topography. To address time and budget constraints, we carefully selected representative plots based on species abundance and composition. This approach was more suitable than random and replicated plots for desert habitats with sparse vegetation (Wang et al., 2022). We identified plant individuals to the species level and estimated the canopy cover of each species in each plot.

We collected sun-exposed, mature leaves from five to ten adult individuals of each species, with a total fresh weight of >30 g. These leaves were then divided into two parts: one was stored in an icebox until the leaf area and water content could be measured, and the other was placed in a paper bag for chemical analysis. In total, we collected 488 leaf samples from 49 species (Supplementary Data Table S1), which accounted for >65 % of the common native desert plant species in the study area. We used a soil auger with a diameter of 5 cm to collect three random soil samples in each plot to a depth of 20 cm. Soil samples from each plot were mixed evenly to produce a single composite sample, then divided into two parts: one was stored in a cloth bag, and the other was packed in a polyethylene bag and stored in an icebox.

Measurements

We randomly chose 15–30 fresh leaves from each species and measured the fresh weight (FW) immediately, and then scanned them (we rotated the succulent leaves by 90° to obtain area measurements for two of the four sides) using a flatbed scanner. We calculated the leaf area (LA) using the software ImageJ (https://imagej.net/ij/index.html). Next, we oven-dried these leaves and measured their dry weight (DW). We quantified the degree of leaf succulence using the leaf succulence index (LSI), which is calculated as LSI = (FW − DW)/LA (Medina et al., 2015). Leaf samples in the paper bags were oven-dried at 60 °C until constant weight to determine the moisture content and were then ground into a fine powder using a ball mill. We used the standard anthrone colorimetric method to measure leaf NSC content (Li et al., 2016). Specifically, 0.1 g of ground leaf sample was mixed with 10 mL of 80 % (v/v) alcohol in a 20 mL test tube and boiled twice for 30 min to extract SS. The resulting solution was filtered, diluted and used to measure soluble sugar concentration. The solid residues were boiled with distilled water to extract starch, which was then hydrolysed with 30 % (v/v) perchloric acid (HClO₄). After filtration and dilution, the starch concentration was determined using a Ultraviolet–visible spectrophotometer. SS and starch concentrations were then converted to the dry matter ratio (in milligrams per gram) using leaf water content. We calculated the total NSC concentration by summing the SS and starch contents (Martínez-Vilalta et al., 2016).

The soil samples stored in cloth bags were air-dried, sieved through a 2 mm mesh, and ball-milled into a fine powder for chemical analysis. The soil samples in the icebox were sieved through a 2 mm mesh and stored at 4 °C to keep them fresh. Soil pH and electrical conductivity (EC) were determined in 1:2.5 v/v and 1:5 v/v soil:water extracts, using pH and conductivity meters (PHSJ-6L and DDSJ-319L, Leici, Shanghai, China), respectively. Total nitrogen (TN) was determined by an elemental analyser (ECS4010, Costech Analytical, Valencia, CA, USA). Total phosphorus (TP) was measured using the ammonium molybdate method. Total potassium (TK) was determined by flame photometry (WGH6400, Changxi, Shanghai, China). Available nitrogen (AN) was determined by a flow-injection analyser (FIAstar5000, FOSS, Hilleroed, Denmark) after extraction with of 2 mol L⁻¹ KCl. Available phosphorus (AP) was measured by the molybdenum blue–ascorbic acid method. Available potassium (AK) was determined by the flame photometer after extraction with neutral normal ammonium acetate.

For the climatic variables, we obtained the MAP and MAT at each sampling site from the Data Center for Resources and Environmental Sciences, Chinese Academy of Sciences (http://www.resdc.cn). These variables were uniformly converted into raster data with a 500 m spatial resolution using v.10.3 of the ArcMap GIS software (http://www.esri.com).

Statistical analysis

We analysed the data at two levels: the species level, using mean values for each species across the study area, and the plot level, using the cover-weighted values for all species present in each plot. The plot-level leaf trait values were calculated as:

$${\displaystyle \begin{array}{l}{\displaystyle \mathrm{P}\mathrm{l}\mathrm{o}\mathrm{t}\mathrm{l}\mathrm{e}\mathrm{v}\mathrm{e}\mathrm{l}\mathrm{t}\mathrm{r}\mathrm{a}\mathrm{i}\mathrm{t}={\displaystyle \underset{i=1}{{\displaystyle \sum ^n}}}[c_i\times \mathrm{t}\mathrm{r}\mathrm{a}\mathrm{i}{\mathrm{t}}_i]}\end{array}}$$

where c_i is the relative cover of species i in the plot, and trait_i is the trait value of species i. We used the Kolmogorov–Smirnov test to test for a normal distribution for each variable; if the data were not normally distributed, they were log-transformed before analysis (Supplementary Data Table S3). We used independent-sample t-tests to analyse the differences in leaf NSC between the succulent and non-succulent plant groups. We used Pearson’s correlation coefficient (r) to analyse the relationships between leaf NSC and the degree of leaf succulence (measured by LSI) and the soil properties. We performed one-way ANOVA to test for differences of leaf NSC and soil properties among the different desert types. We also performed two-way ANOVA, with succulent/non-succulent and desert type as the two levels, but given that we found no significant interaction, we have presented only the one-way ANOVA results. We used hierarchical partitioning analysis to examine the effects of environmental factors on community-level leaf NSC. We used P < 0.05 as the threshold for significance in this study. All statistical analyses were performed using v.3.6.1 of the software R (R Development Core Team; www.r-project.org), and we performed the hierarchical partitioning analysis using the ‘hier.part’ package (v.1.0-6; https://cran.r-project.org/src/contrib/Archive/hier.part/).

RESULTS

Leaf NSC across all species

The leaf SS, starch and total NSC contents ranged widely at the species level. When considering all species collectively, the SS content ranged from 4.79 to 74.39 mg g⁻¹, with a mean of 26.99 mg g⁻¹; the starch content ranged from 27.52 to 109.07 mg g⁻¹, with a mean of 60.28 mg g⁻¹; and the total NSC content ranged from 35.83 to 151.39 mg g⁻¹, with a mean of 87.27 mg g⁻¹ (Fig. 2; Supplementary Data Table S1). The coefficient of variation (CV) of SS (0.53) was higher than those of starch (0.31) and total NSC (0.31).

Fig. 2.

Frequency distributions for soluble sugars (SS), starch and total non-structural carbohydrate (total NSC) concentrations in the leaves of the 49 plant species across the study area.

Our results differed significantly from previous results for the leaf SS, starch and total NSC contents. Specifically, leaf SS content in our study was significantly lower than that of Chinese forest plants (Li et al., 2016) and global terrestrial plants (Martínez-Vilalta et al., 2016) (Fig. 3), whereas the leaf starch content was significantly higher than that for Chinese forest plants and not significantly different from that of global terrestrial plants. In addition, the leaf total NSC content was not significantly different from that of Chinese forest plants but was significantly lower than that of global terrestrial plants (Fig. 3).

Fig. 3.

Comparison of the leaf non-structural carbohydrates [soluble sugars, starch and total non-structural carbohydrate (NSC)] contents of plants in the present study with the results for a north–south forest transect in eastern China (NSTEC; Li et al., 2016) and for terrestrial plants worldwide (Martínez-Vilalta et al., 2016). Values are means ± s.e.m. *P < 0.05 and **P < 0.01; NS represents no significant difference.

Leaf NSC in different plant functional groups

The mean concentrations of SS, starch and total NSC in the leaves of herbaceous plants were 24.00, 57.91 and 81.97 mg g⁻¹, respectively. In comparison, the leaves of woody plants exhibited mean concentrations of 29.86 mg g⁻¹ for SS, 62.56 mg g⁻¹ for starch and 92.41 mg g⁻¹ for total NSC. Overall, the SS, starch and total NSC levels in woody plants were slightly higher than those in herbaceous plants, but these differences were not statistically significant (Fig. 4A).

Fig. 4.

Contents of leaf non-structural carbohydrates [soluble sugars, starch and total non-structural carbohydrates (NSC)] in woody vs. herbaceous plants (A) and succulent vs. non-succulent leaves (B). Box-and-whisker plots show the mean (dotted line), the median (solid line) and the interquartile range (25th–75th percentiles) in the boxes, and the 5th and 95th percentiles (whiskers). **P < 0.01; NS represents no significant difference.

Plants with succulent leaves had significantly lower SS (18.63 mg g⁻¹) and total NSC (75.71 mg g⁻¹) than plants with non-succulent leaves (35.02 mg g⁻¹ for SS and 98.37 for mg g⁻¹ for total NSC). However, there was no significant difference in starch concentration between plants with succulent leaves (57.09 mg g⁻¹) and those with non-succulent leaves (63.35 mg g⁻¹; Fig. 4B).

Across all species, we observed a weak but significant positive correlation between SS and starch (Fig. 5). The SS:starch ratio ranged from 0.10 to 1.20, with a mean value of 0.49. Of the 49 species examined, only four species had a SS:starch ratio greater than one, and the remaining 45 species had ratios below one (Supplementary Data Table S1). The SS:starch ratio did not differ significantly between herbaceous (0.47) and woody plants (0.52); in contrast, plants with succulent leaves had a significantly lower SS:starch ratio (0.36) than non-succulent plants (0.63; Fig. 5).

Fig. 5.

Relationship between leaf soluble sugars (SS) and starch. The value of Pearson’s correlation coefficient (r) is shown. *P < 0.05. The inset graph shows the differences of the soluble sugars:starch ratio for all species combined and between the different plant functional groups. **P < 0.01; NS represents no significant difference.

Relationships between leaf NSC and leaf succulence

We found statistically significant correlations between NSC contents and LSI. The LSI exhibited a significant negative correlation with the leaf SS concentration (r = −0.58, P < 0.001), the SS:starch ratio (r = −0.52, P < 0.001) and the total NSC concentration (r = −0.40, P < 0.001) across all species (Fig. 6A, C, D). However, there was no significant correlation between LSI and leaf starch concentration (r = −0.13, P = 0.38; Fig. 6B).

Fig. 6.

Correlations (Pearson’s r) between the leaf succulence index (LSI) and soluble sugars (A), starch (B), the soluble sugars to starch ratio (C) and total non-structural carbohydrates (NSC) (D). The vertical dashed lines at log LSI = 2.69 (corresponding to LSI = 500) in each panel distinguish succulents (left) from non-succulents (right). Data were log-transformed. **P < 0.01 and ***P < 0.001; NS represents no significant difference.

Soil properties and leaf NSC in different desert habitats

Among the three habitats, saline desert had higher soil fertility levels, with significantly higher soil TN, AN, AP and AK contents than in the gravel desert and sandy desert habitats (Fig. 7); however, TP was only significantly higher than in the gravel desert. In addition, the saline desert also had the highest soil EC, with the much higher value confirming that this desert had the highest soil salinity levels (Fig. 7). In contrast, the gravel and sandy desert habitats did not differ significantly in soil salinity and most nutrient levels, with the exception of a significantly lower TP and AK in the gravel desert (Fig. 7). Additionally, gravel desert had significantly higher relative cover of succulent plants than saline and sandy deserts (Supplementary Data Fig. S3).

Fig. 7.

Soil properties in different habitats: (A) pH; (B) electrical conductivity (EC); (C) soil total nitrogen (TN); (D) available nitrogen (AN); (E) total phosphorus (TP); (F) available phosphorus (AP); (G) total potassium (TK); and (H) available potassium (AK). *P < 0.05 and **P < 0.01; NS represents no significant difference.

Although soil properties differed greatly among the desert types, the NSC contents in plant leaves did not differ significantly. Tamarix ramosissima, the species most widely distributed throughout the study area, showed no significant variation in leaf SS, starch and total NSC contents among habitats (Fig. 8A). Likewise, when we considered all species within a plot, we found no significant differences in plot-level leaf NSC contents among the desert types (Fig. 8B).

Fig. 8.

Species-level (A) and plot-level (B) contents of leaf non-structural carbohydrates (NSC) (soluble sugars, starch and total NSC) in the three desert types. Box-and-whisker plots show the mean (dotted line), the median (solid line) and the interquartile range (25th–75th percentiles) in the boxes, and the 5th and 95th percentiles (whiskers). NS represents no significant difference.

Effects of environmental factors on leaf NSC

Environmental variables explained only a minor portion of the spatial variation in leaf NSC. The hierarchical partitioning analysis revealed that the combined effects of all environmental factors explained 8.6, 11.1, 6.3 and 12.5 % of the spatial variation in the leaf SS, starch, total NSC and SS:starch ratio, respectively (Table 1). The impact of environmental factors on leaf SS and the SS:starch ratio was driven primarily by soil EC, whereas the influence on leaf starch was mainly attributed to soil nutrients. In contrast, climatic factors did not significantly affect the leaf NSC (Table 1). In addition, soil EC had a significantly negative effect on leaf SS and the SS:starch ratio, although the correlation was weak (Fig. 9).

Table 1.

Results of the hierarchical partitioning analysis for the effects of soil properties (AK, available potassium; AN, available nitrogen; AP, available phosphorus; EC, electrical conductivity; TK, total potassium; TN, total nitrogen; TP, total phosphorus) and climatic factors (MAP, mean annual precipitation; MAT, mean annual temperature) on community-level leaf non-structural carbohydrates.

Carbohydrate	Full model (R2 as a %)	Soil EC	Soil total nutrients			Soil available nutrients			Climatic factors
			TN	TP	TK	AN	AP	AK	MAP	MAT
SS	8.57	63.07*	1.55	8.50	0.65	3.08	5.09	5.18	9.33	3.55
Starch	11.09	4.70	5.99	1.84	31.18*	31.87*	10.40	4.87	2.31	6.84
Total NSC	6.30	7.64	5.99	10.38	32.07	21.00	8.69	8.81	3.02	2.41
SS: Starch	12.47	38.44*	3.98	1.02	3.67	15.11	14.28*	2.17	12.78	8.55

*P < 0.05.

Fig. 9.

Relationships between soil electrical conductivity (EC) and community-level: (A) leaf soluble sugar contents; and (B) the leaf soluble sugar to starch ratio. Data were log-transformed, and values of Pearson’s correlation coefficient (r) are shown. **P < 0.01.

DISCUSSION

Leaf NSC characteristics and their differences among plant functional groups

Our study represents, to the best of our knowledge, the first systematic survey of the leaf NSC of plants thriving in the driest regions of China. The average leaf SS, starch and total NSC contents across all 49 species were 26.99, 60.28 and 87.27 mg g⁻¹, respectively (Fig. 2). We found that the mean leaf SS content in the present study was much lower than the means for the forest plants of China (Li et al., 2016) and the means for the world’s terrestrial plants (Fig. 3; Li et al., 2016; Martínez-Vilalta et al., 2016). Although it is not possible to compare our results fairly with those of these previous studies owing to differences in scale and the number of species, the comparison nonetheless provides context for our results. The plants in the present study did not exhibit higher leaf NSC contents than plants growing in more humid environments. According to the general framework of plant drought responses, increasing SS levels through starch decomposition is an effective strategy for maintaining osmotic balance under drought stress (Signori-Müller et al., 2021; Tsamir-Rimon et al., 2021; Blumstein et al., 2023). However, under severe drought stress, lethal drought could break the balance between SS and starch, potentially leading to the depletion of NSC pools (Hartmann et al., 2013).

Succulence is an essential evolutionary strategy for plants to overcome drought stress by storing water primarily in the symplastic compartments (Fradera-Soler et al., 2022a, b). Additionally, some mucilaginous succulent plants, such as aloes, can enhance their water-retentive capacity by using polysaccharides secreted into the intercellular spaces (Isager Ahl et al., 2023; Fangel et al., 2024). Despite these adaptations, most succulents are scarce in the arid deserts and are more commonly found in semi-arid zones with regular and predictable rainfall (Griffiths and Males, 2017). In fact, such succulents are drought-avoiding plants with limited capacity for osmotic adjustment (Grace, 2019). In contrast, the succulents in this study are primarily halophytes, whose ecophysiological adaptations differ significantly from those of ‘iconic’ succulents. This special type of succulence enables halophytes to achieve a very low leaf osmotic potential by absorbing inorganic salts rather than synthesizing organic osmolytes (Wang et al., 2022). It also highlights a knowledge gap in exploring the function of cell walls in succulent halophytes (Cárdenas Pérez et al., 2024). Consistent with our first hypothesis, the SS content in leaves of succulent plants (LSI > 500) was significantly lower than in non-succulent plants (Fig. 4B). In contrast, starch contents did not differ significantly between succulent and non-succulent plants (Fig. 4B). This might be because starch is primarily a storage compound not directly involved in osmoregulation (Hultine et al., 2021), and the accumulation of inorganic salts further reduces the need for starch conversion to organic osmolytes (Jiménez-Becker et al., 2019). The moderately strong negative relationship between LSI and the leaf SS content further suggests that halophytic succulence is an effective physiological adaptation (Fig. 6), allowing plants to use inorganic salts as ‘cheap’ osmolytes while avoiding excessive consumption of their NSC pool.

Previous studies have shown that the SS and starch contents are usually higher in herbaceous than in woody species (Li et al., 2016). This is mainly because leaves of herbaceous plants have a higher proportion of living and metabolically active tissues, whereas the leaves of woody plants contain more lignin and cellulose, making them harder and more durable (Martínez-Vilalta et al., 2016). However, our results revealed that both leaf NSC content (i.e. SS and starch) and its composition (i.e. SS:starch ratio) did not differ significantly between herbaceous and woody plants in our study area (Fig. 4A). On the one hand, the leaves of woody and herbaceous plants are structurally similar, with both containing many succulent species and probably having no major differences in the amount of lignified tissues. On the other hand, plants that thrive in extremely arid conditions usually adopt the ‘slow return’ strategy (Wang et al., 2017), which might lead to similar NSC metabolic demands in the leaves of both woody and herbaceous plants. Therefore, these findings emphasize that the unique ecological conditions in the study area, characterized by extreme aridity, might have influenced plant strategies, leading to a convergence in the leaf NSC contents between herbaceous and woody species.

Variations of leaf NSC among different desert habitats

Although our data were obtained in similar climatic conditions, the soil properties differed significantly among the three desert types. Specifically, the saline desert soils contained a substantial amount of salt, leading to markedly elevated soil salinity levels compared with the gravel and sandy deserts (Fig. 7). Drought stress primarily involves a water deficit, and soil salinity not only induces osmotic stress but is also associated with ion toxicity. Ion toxicity can significantly enhance the leaf SS content and reduce the leaf starch content (Yang et al., 2024). On this basis, we hypothesized that plants in saline deserts would exhibit a higher leaf SS content and a lower starch content than the same plants in a gravel or sandy desert, because they might require additional energy to cope with the associated effects of osmotic stress and ion toxicity. In contrast to our expectation (hypothesis 2), we found no significant differences in leaf SS, starch and total NSC contents among the three desert types, either at the species level or at the plot level (Fig. 8). This can probably be attributed to the fact that both succulent and non-succulent plants in this study are primarily halophytes, equipped with effective mechanisms to cope with ion toxicity, including compartmentalization, exclusion and excretion (Wang et al., 2017).

Previous research has shown that halophytic succulents growing in gravel deserts also accumulate substantial amounts of salt ions in their leaves, suggesting that their salt accumulation is an active process (Wang et al., 2020). Interestingly, we found that the dominance of halophytic succulents in gravel deserts was higher than in saline deserts (Supplementary Data Fig. S3). This is probably because saline deserts have higher moisture and nutrient availability, which supports a wider variety of halophytes (including excluders and secretors) and reduces the dominance of succulents. In contrast, halophytic succulence might be a more effective strategy for plants to achieve osmotic adjustments in gravel desert habitats. Moreover, the succulent species in the study area mainly belong to the Amaranthaceae family, which includes a variety of halophytic succulents worldwide (Flowers et al., 2010). Anatomical evidence shows that leaves of these plants usually contain many specialized achlorophyllous cells, which are responsible not only for sequestering the toxic ions but also for protecting photosynthetic cells from dehydration (Ogburn and Edwards, 2010). Therefore, the strict ion compartmentalization mechanism ensures that the salt accumulation, while providing an osmotic advantage, does not interfere with other biochemical processes (Kronzucker et al., 2013; Wang et al., 2017).

Recreto-halophytes are plants that excrete excess salts through specialized glands on their leaf surface (Yuan et al., 2016). Tamarix ramosissima is the most widespread species thriving in all three desert habitats across the study area. Tamarix ramosissima is also a typical salt-secreting plant, featuring glandular structures on its leaf surface (Wei et al., 2020). Our results indicated that the leaf NSC contents of T. ramosissima did not differ significantly between saline deserts and the other two desert habitats (Fig. 8A). It is likely that the salt-secreting properties enable T. ramosissima to avoid ionic toxicity in high-salinity environments. In contrast, pseudo-halophytes are plants that use a salt-exclusion mechanism (Wang et al., 2017). They can accumulate salts only in the parenchyma organs of the roots, thereby limiting the entry of toxic ions into above-ground parts (Meng et al., 2018). In this study, halophytes from the Fabaceae and Poaceae families are pseudo-halophytes; their leaves are neither succulent nor equipped with salt-secreting glands. Although these plants have limited tolerance to salinity, leguminous species, such as Alhagi camelorum, possess very well-developed root systems that help them to avoid the highly stressed environment of the shallow soil (Pirasteh-Anosheh et al., 2022). Overall, our results suggest that leaf NSC of the wild plants in the study area did not exhibit additional responses to saline conditions. This might highlight the resilience of desert plants in maintaining desirable leaf NSC profiles, potentially driven by strict salt-coping strategies.

Responses of leaf NSC to environmental factors

Previous research indicated that the leaf SS, starch and total NSC contents in the forest plants of China exhibited weak negative correlations with MAT and MAP, which suggested not only that climatic factors directly influenced the spatial variation of leaf NSC, but also that accumulation of leaf NSC was an adaptive strategy that allowed plants to cope with drought and changing temperature (Li et al., 2016). Similar results were observed in temperate and tropical forests (Blumstein et al., 2023). However, in our study, we found that leaf SS, starch and total NSC were not significantly related to MAT and MAP (Table 1). This might be attributed to the fact that our study area is characterized by extreme aridity, with the narrow ranges of MAP and MAT potentially limiting the ability to detect significant climatic correlations. However, most plant species in the study area were xerophytes and halophytes and were unique to the arid regions of China; as a result, they do not thrive in more humid climates. From this perspective, the distinctive leaf NSC patterns in our study area can be viewed as an outcome of the specialized plant adaptations to the extremely arid climate, because climatic factors are important driving forces for vegetation composition and plant species diversity across spatial and temporal scales (Feeley et al., 2020). At the spatial scale of the study area, the combination of all environmental factors accounted for only 8.6, 11.1 and 6.3 % of the variation in leaf SS, starch and total NSC, respectively, indicating that environmental factors exert a relatively weak effect on the spatial variation of leaf NSC. These findings might support the idea that desert plants have developed relatively stable physiological and ecological survival strategies during long-term adaptation to the harsh environment of the region (Abella et al., 2019).

The combination of scarce precipitation and high evaporation often induces soil salinity in arid deserts, which creates a serious constraint on plant growth and reproduction (Bui, 2013). In the study area, soil EC had stronger spatial variability than the narrow ranges of MAT and MAP. Although the overall impact of environmental factors on the spatial variation of leaf NSC was low, soil EC exerted a significant impact on the leaf SS and the SS:starch ratio (Fig. 9). This result supports our third hypothesis, that soil properties, and particularly salinity, would play a more direct role in influencing leaf NSC than climatic factors. Similar results were found in previous studies conducted in arid deserts, where soil salinity rather than climatic factors had a direct impact on plant nutrient uptake and physiological processes (Luo et al., 2021; Wang et al., 2022). Notably, we found that both leaf SS and the leaf SS:starch ratio decreased significantly with increasing soil EC, indicating that soil salinity does not promote the accumulation of leaf SS. This agrees with our statement in the first section of the Discussion that the osmotic adjustment can be provided by inorganic salts rather than organic compounds. However, given that desert plants adopt strict salt regulation strategies, the negative correlation was relatively weak.

Conclusion

To the best of our knowledge, our study offers the first comprehensive large-scale assessment of the leaf NSC of desert plants growing in arid and hyper-arid environments. Owing to the prevalence of succulent halophytes, plants in the driest regions of China did not exhibit elevated leaf NSC contents compared with plants in more humid environments. Notably, there was a strong negative correlation between leaf succulence (LSI) and the leaf SS content, highlighting halophytic succulence as a driver of leaf NSC pools in desert plants. In contrast, environmental factors played a weak role in influencing the variation of leaf NSC, probably owing to the relatively narrow range of these factors in our study area. Climatic factors made the smallest contribution to the spatial variation of leaf NSC, whereas soil properties, especially soil salinity, had a stronger and more direct influence. Contrary to previous research, we found that as soil salinity increases, the leaf SS content decreases. Our results provide valuable insights into the complex interplay of environmental factors and plant adaptations in arid desert ecosystems and emphasize the importance of halophytic succulence in influencing leaf NSC in desert plants.

SUPPLEMENTARY DATA

Supplementary data are available at Annals of Botany online and consist of the following.

Figure S1: photographs of the widely distributed succulent halophytes in the study area. Figure S2: photographs of the common plant species in the study area. Figure S3: the relative cover of succulent plants vs. non-succulent plants in each of the three desert habitats. Table S1: species information and mean values of leaf soluble sugars (SS), starch and total non-structural carbohydrates (Total NSC). Table S2: information of environmental factors and plot-level leaf traits of the 115 sampling sites across the study area. Table S3: results of Kolmogorov–Smirnov test for variables at different levels.

FUNDING

This research was supported by the Second Tibetan Plateau Scientific Expedition and Research (STEP) program (grant no. 2019QZKK0305) and the Central Government Guides Local Science and Technology Development Plans (grant no. 2022ZY0145).

AUTHOR CONTRIBUTIONS

Lilong Wang and Yuqiang Li conceived and designed the experiments. Lilong Wang, Xuyang Wang and Yulong Duan performed the sample collection. Lilong Wang and Chengzhuo Zheng performed the laboratory analysis. Lilong Wang analysed the data and wrote the manuscript. All authors contributed to the paper and approved the submitted version.

CONFLICT OF INTEREST

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

DATA AVAILABILITY

All relevant data are within the manuscript and its supplementary files.