The study area is located on the southern edge of the Gurbantunggut Desert in Northwest China. The Gurbantunggut Desert, characterised by a typical temperate continental arid climate, is the largest stabilised and semi-stabilised desert in China (Chen et al., 2019c; Lu et al., 2022; Zhang et al., 2022b). The mean annual temperature varies from 6°C to 10°C, with extreme high temperatures over 40°C (Liu et al., 2020a). The annual potential evaporation can exceed 2,000 mm, and the annual precipitation in the central region of the desert is only 70–100 mm, which mainly occurs in winter and spring (Liu et al., 2020a; Lu et al., 2022). Natural conditions in the Gurbantunggut Desert are extremely harsh and susceptible to human activity and extreme climatic events. However, compared with other desert regions in China, herbaceous plants in the Gurbantunggut Desert are relatively abundant in spring, especially in ephemeral plant communities (Erfan et al., 2023). E. oxyrhinchum is a typical annual ephemeral plant found in the Gurbantunggut Desert. It contributes to more than 50% of the aboveground productivity of ephemeral plants, is the dominant species in the Gurbantunggut Desert, and plays a crucial role in the desert ecosystem of China (Zang et al., 2020).

The survival percentage of E. oxyrhinchum was analysed using independent sample t-tests with a significance level of 0.05, and Levene’s test was used to confirm the homogeneity of variance. In addition, independent sample t-tests (consistent with the survival percentage analysis method) and two-way ANOVA were used to analyse plant height, root length, number of leaves, leaf area, individual biomass, reproductive biomass, root to shoot ratio, growth rate, and stoichiometric characteristics. The coefficient of variation (CV) was employed to examine the variability of leaf C, N, and P concentration and their ratios throughout the sampling period: 0≤CV<15% shows weak variability, 15%≤CV<35% indicates moderate variability, and CV ≥ 35% indicates high variability. Data analysis was conducted using SPSS27.0, and graphical representations were generated using R-4.3.2.

Statistical analysis revealed that the phenological stages of E. oxyrhinchum under drought pretreatment were significantly earlier than those under the control treatment (p < 0.001, Table 1). Specifically, under drought pretreatment, the full leaf expansion stage occurred 5.25 ± 1.2 days earlier, the initial flowering stage 3.13 ± 0.84 days earlier, the initial fruiting stage 4.75 ± 1.63 days earlier, and the withering stage 5 ± 1.38 days earlier than that of control treatment (Table 1). Notably, the entire life history of E. oxyrhinchum under drought pretreatment were also significant shorter than that of control treatment: DP (69.63 ± 0.75 days) < CK (74.63 ± 0.63 days).

On 20 March 2023 the seeds of E. oxyrhinchum exhibited an explosive germination strategy, with seedlings appearing uniform before entering the cotyledon stage (Figure 3). The survival percentages of E. oxyrhinchum in the control and drought treatments did not significantly decrease from the cotyledon stage to April 19. After the full leaf expansion stage, the survival percentage of E. oxyrhinchum under drought pretreatment showed an earlier and faster decline than that of the control treatment (Figure 3). By the end of May, plants exposed to the drought pretreatment rapidly entered the withering stage. By early June, E. oxyrhinchum had completed its reproductive output and ended its life history between the control treatment and drought pretreatment (Figure 3).

The results of the two-way ANOVA indicated significant differences in terms of plant height, root length, leaf area, and leaf number of E. oxyrhinchum at different phenological stages (p < 0.001, Table 2). There were significant differences between the control and drought treatments in terms of plant height and root length (p < 0.05, Table 2); however, no significant differences were observed in leaf number (p = 0.502, Table 2) or leaf area (p = 0.12, Table 2). Furthermore, there was a significant interaction effect between drought treatment and phenological stage on plant height and leaf number (p < 0.05, Table 2).

The height, root length, and leaf area of E. oxyrhinchum in the control and drought treatments increased gradually with the phenological stage (Figure 4). Prior to the fruiting stage, the number of leaves increases. From the initial leaf expansion stage to the full fruiting stage, plant height under drought pretreatment was significantly higher than that of the control treatment (p < 0.05, Figure 4). During the initial leaf expansion stage, root length under drought pretreatment was significantly longer than that of the control treatment. At the initial leaf expansion and flowering stages, the number of leaves under the drought pretreatment was significantly higher than that of the control treatment, but the difference did not reach a significant level during the full blooming and fruiting stages.

According to the results of the two-way ANOVA, the individual biomass, reproductive biomass, and growth rate of E. oxyrhinchum showed significantly increasing trends, but the R/S exhibited a significantly decreasing trend as the plants grew (p < 0.001, Table 2). Similarly, there were significant differences in the individual biomass, root to shoot ratio, and reproductive biomass of E. oxyrhinchum between the control and drought treatments (p < 0.05, Table 2). Throughout the reproductive season, the reproductive biomass under the drought pretreatment was consistently higher than that under the control treatment (Figure 5). Specifically, the reproductive biomass under drought pretreatment during the full fruiting stage was 1.41 times than that of the control. From the perspective of the entire life history, the individual biomass of E. oxyrhinchum under the drought pretreatment was greater than that of the control treatment (Figure 5). For example, during initial leaf expansion, the individual biomass of E. oxyrhinchum under the drought treatment was 1.29 times than that of the control treatment.

Two-way ANOVA revealed significant differences in the stoichiometric characteristics and ratios of C: N, and P in the leaves of E. oxyrhinchum between the control and drought treatments (p < 0.001, Table 2). Similarly, from the perspective of the entire life history, significant differences in the stoichiometric characteristics and C: N, and P in the leaves of E. oxyrhinchum were observed at different phenological stages (p < 0.01). From the initial leaf expansion stage to the full leaf expansion stage, the C concentration in the leaves of E. oxyrhinchum under drought pretreatment was significantly higher than that in the control treatment (Figure 6A). However, at the initial flowering and full blooming stages, the C concentration under drought pretreatment was significantly lower than that of the control treatment. In addition, the N concentration in the leaves of E. oxyrhinchum exhibited a trend of initially increased and then decreased (Figure 6C). Before the initial flowering stage, the N concentration in the leaves of E. oxyrhinchum under the drought pretreatment was significantly higher than that in the control treatment. Similarly, from the perspective of the entire life history, the P concentration in the leaves of E. oxyrhinchum under drought pretreatment was significantly higher than that in the control treatment (Figure 6E).

The C: N ratio in the leaves of E. oxyrhinchum initially decreased and then increased as the phenological stage progressed, whereas the C:P and N:P exhibited an increasing trend. From the cotyledon stage to the full leaf expansion stage, the C: N in the leaves of E. oxyrhinchum under drought pretreatment was significantly lower than that in the control treatment (p < 0.001, Figure 6B). From the perspective of the entire life history, the C:P in the leaves of E. oxyrhinchum under drought pretreatment was also significantly lower than that in the control treatment (Figure 6D). Similarly, from the cotyledon stage to the full blooming stage, the N:P under the drought pretreatment was significantly lower than that under the control treatment (p < 0.05, Figure 6F).

From the perspective of the entire life history, the N and P concentration in the leaves of E. oxyrhinchum exhibited greater variation than the C concentration. Specifically, the coefficient of variation for C concentration in the leaves of E. oxyrhinchum between the control treatment and drought pretreatment was less than 5%, indicating weak variation (Table 3). However, from the full blooming stage to the full fruiting stage, the coefficient of variation for the N concentration under the control treatment and the P concentration under the drought pretreatment was greater than 15%, indicating moderate variation.

Phenology represents cyclical changes in vegetation and is a sensitive biological indicator of climate change (Li et al., 2021). This study revealed that the phenological stages of E. oxyrhinchum subjected to drought pretreatment were markedly accelerated. For instance, the full leaf expansion stage under drought pretreatment was 5.25 ± 1.2 days earlier than that of the control treatment. Early leafing allows plants to initiate photosynthesis sooner, thereby enhancing biomass accumulation (Ganjurjav et al., 2021; Amini et al., 2023; Wu and Yang, 2024). Similarly, the onset of flowering and fruiting of E. oxyrhinchum under drought pretreatment also occurred significantly earlier than in the control treatment. This may be because earlier leafing under drought pretreatment induced plants to enter the reproductive stage. Previous studies also have indicated that desert ephemeral plants employ a “drought escape strategy” by accelerating phenological stages to complete life history (Franks, 2011; Ding et al., 2022). These findings not only highlight the positive response of ephemeral plants to drought stress but also demonstrate that drought pretreatment can accelerate the rapid transitions of phenological stages, thereby influencing the life history of ephemeral plants.

The survival percentage reflects the adaptability of plants to environmental stress and is a valuable indicator of drought resistance in seedlings (He et al., 2024; Chen et al., 2025). In this study, the survival percentage of E. oxyrhinchum from March 20 to April 19 did not show a significant decrease trend, but the survival percentage of E. oxyrhinchum significantly decline after the full leaf expansion stage. The survival percentage of E. oxyrhinchum under drought pretreatment exhibited a faster and earlier declining trend than that of the control treatment. With the rapid growth of E. oxyrhinchum, the difference in survival percentage may be related to the decrease in soil water content and intensification of intraspecific competition. Similarly, Guo et al. (2020) investigated the combined effects of drought and intraspecific competition on the growth of C. lanceolata, and the results indicated that intense competition imposed by neighbours was a great threat to the survival of young C. lanceolata plants under prolonged drought. Furthermore, the trend in soil water content revealed lower moisture levels under drought pretreatment (Supplementary Figure 1), further suggesting intensified intraspecific competition among plants. Therefore, ephemeral plants usually adopt a strategy of decreasing their survival percentage, reducing intraspecific competition, and shortening their life history to cope with intense drought stress.

Plants respond to environmental stress by adjusting their morphological traits and biomass allocation (Freschet et al., 2018; Geleta et al., 2024). Our study demonstrated that plants exposed to drought pretreatment showed more significant morphological changes than those in the control treatment (Figure 4). Specifically, from the initial leaf expansion to the full blooming stage, the plant height of E. oxyrhinchum under the drought pretreatment was significantly higher than that of the control treatment. Additionally, during the initial leaf expansion and flowering stages, the number of leaves under the drought pretreatment was significantly higher than that in the control treatment. The root length of E. oxyrhinchum under drought pretreatment was also significantly greater than that of the control treatment during the initial leaf expansion stage. This suggests that ephemeral plants under drought pretreatment could prioritise resource allocation to their roots, allowing them to extend deeper or wider soil layers to obtain more water and nutrients and improve their adaptability. Similarly, Xu et al. (2023) found that drought stress prompted Alhagi sparsifolia to allocate more resources for root growth, thereby enhancing its survival and adaptability in desert environments.

The positive correlation between morphological traits and biomass has been validated extensively (Lu et al., 2014; Chen et al., 2019b). In this study, correlation analysis also revealed a positive correlation between plant height, root length, leaf area, and leaf number, and the individual biomass of E. oxyrhinchum (Table 4). In addition, we found that the individual and reproductive biomass of E. oxyrhinchum under drought pretreatment was significantly higher than that of the control treatment. This suggests that moderate drought stress did not significantly decrease the biomass of E. oxyrhinchum but rather promoted its growth of E. oxyrhinchum, demonstrating an overcompensation effect. Similarly, in agricultural practices, managers often apply mild short-term drought stress to specific crop root zones, inducing a stress response through alternating partial root-zone irrigation, thereby promoting biomass accumulation and improving yield (Consoli et al., 2017; Fu et al., 2017; Liu et al., 2020b; Wang et al., 2024). The persistence of annual plant populations depends entirely on seed production (Lan and Zhang, 2008). In this study, drought pretreatment significantly increased the reproductive output of E. oxyrhinchum, which is likely related to the accelerated transition of ephemeral plants from vegetative to reproductive growth under drought stress. In the early spring in the Gurbantunggut Desert, soil moisture undergoes a gradual change from high to low, indicating that drought stress continues to intensify (Huang et al., 2015b; Chen et al., 2019d; Lu et al., 2022). To ensure population persistence, ephemeral plants must complete their reproductive output before the onset of extreme drought or hot summers (Mu et al., 2021; Lu et al., 2022). Therefore, the overcompensation effect in E. oxyrhynchum under drought stress is an important strategy for population persistence and community stability in the Gurbantunggut Desert.

C, N, and P are essential nutrients that play critical roles in plant growth and development, influencing key physiological and ecological processes such as photosynthesis, transpiration, and reproductive growth (Zhang et al., 2021; Costa et al., 2024). Our study revealed that the leaf C concentration under drought pretreatment was significantly higher than that of the control treatment during the initial and full leaf expansion stages. This suggests that drought pretreatment increases C accumulation in the early stages of growth, which is likely because plants adjust their energy storage structure under drought conditions to maintain survival and growth. A study on tropical plant seedlings found that plants allocate limited carbon resources to nonstructural carbohydrates under drought stress, which is beneficial for maintaining normal metabolic requirements and rapid recovery after drought stress (O’Brien et al., 2014). Additionally, the coefficient of variation for leaf C concentration under the control and drought treatments was less than 5%. This may be because carbon is a fundamental element of plant structure, and the C concentration in leaves generally remains relatively stable. Our study showed that the leaf N concentration of E. oxyrhinchum in the control and drought treatments exhibited a trend of first increased and then decreased. This may be because the growth rate of the leaves reached its maximum and required higher nitrogen absorption to synthesise proteins before the flowering stage. However, as plants grow, more N is allocated to the reproductive organs, resulting in a relative decrease in the leaf N concentration. In addition, the leaf P concentration under drought pretreatment was significantly higher than that of the control treatment. This may be due to the enhanced resistance and overcompensation effects in plants exposed to drought stress, which require more phosphorus for rRNA synthesis to support plant growth (Elser et al., 2010).

The leaf N:P ratio can be used as an indicator to assess whether plant productivity is limited by nutrients (Yuan et al., 2011; Wang et al., 2014; Lu et al., 2023). Generally, it is accepted that when N:P < 14, plant growth is more limited by nitrogen, when 14 < N:P < 16, plant growth is simultaneously limited by nitrogen and phosphorus, and when N:P > 16, phosphorus becomes the primary limiting factor for plant growth (Güsewell and Koerselman, 2002; Sardans et al., 2012; Yang et al., 2018). Our study found that leaf N:P under drought pretreatment was less than 14 during the cotyledon stage, 14 < N:P < 16 during the initial leaf expansion stage, and N:P > 16 from the full leaf expansion stage to the full fruiting stage (Figure 6F). Similarly, the leaf N:P in the control treatment was greater than 16 from the leaf expansion stage to the full fruiting stage. These findings indicated that the growth of E. oxyrhinchum during the cotyledon stage was primarily limited by N. After the initial leaf expansion stage, N and P limit plant growth simultaneously. The effect of P on plant growth gradually increased during the full leaf expansion, flowering, and fruiting stages. Therefore, it is necessary to optimise ecological restoration measures based on the nutritional needs of plants and improve the effectiveness of restoration.