Dongtan as a coastal salt marsh wetland locates in the Chongming Island at the estuary of Yangtze River (121°50′E-122°5′E, 31°25′N-31°38′N). The area locates in the northern edge of the mid-subtropical zone with the marine monsoon climate. The average annual temperature is 15.3°C and the annual precipitation is 1100 mm (Jia et al., 2022).

The seeds of P. australis populations were collected from the Dongtan wetland at a tidal estuary with an irregular semi-diurnal tide type, and the daily tidal flat characterized by two tidal changes of day and night. According to the observations in recent decades, the mean sea level is 2.17 m, the mean low water level is 1.03 m, and the mean high water level is 3.5 m (Yuan et al., 2020). The area is densely covered with tidal creeks, and the zones of high, middle and low tidal flats are very obvious. Along the elevational gradient of the transect, three tide sampling locations were established including the high tidal flat (2.7–3.5 m), the middle tidal flat (2.4–2.7 m), and the low tidal flat (2.2–2.4 m). The P. australis populations at the middle and high tidal flats in the Dongtan marsh wetland were distributed in continuous bands, while at the low tidal flat, it was distributed in patches.

SPSS 22.0 was used for analyzing and comparing the data (mean ± standard error) and the significance test. In the absence of interaction, data from the germinated seeds was pooled, and the effects of salinity on the germination parameters were analyzed using one-way ANOVA and the Tukey test for the least significant difference (p=0.05). Two-way ANOVA was used to analyze the interaction of tidal level and treatment salinity on seed germination. Redundancy analysis (RDA) was used to analyze the differences in the correlation between the germination parameters of seeds stored under different temperatures and the soil physicochemical factors (pH, water content, salinity, TP, TN and SOC). Structural equation modeling (SEM) was used to evaluate and quantify the effects of soil physicochemical factors on the differences of germination parameters of seeds collected from the populations at different tidal flats and stored under different temperatures. RDA analysis and SEM used the “vegan” package (Dixon, 2003) and “lavvan” package (Rosseel, 2012) in R 4.1.2 version, respectively.

With the increase of soil salinity, the germination rates, germination indexes and germination speeds of P. australis seeds collected from populations at different tidal flats decreased compared with the control group at 0% salinity ( Figures 2a–e ; Supplementary Tables S1 , S2 ). When the treatment salinity was 0.5%, the seed germination rate of the P. australis population in the L tidal flat (91.1% ± 0.38) was significantly higher than that in the other several tidal flats(L(IN): 62.2% ± 1.5; M: 55.3% ± 0.5 and H: 66% ± 1.2). At 1% and 1.5% salinity, P. australis seeds collected from the populations at all tidal flats germinated, and the seed germination rates among the tidal flats showed no significant difference, but the germination rate of seeds collected from the populations at the L tidal flat outside the dike was higher than those of seeds collected from populations at the other sites ( Figures 2c, d ). At 2% salinity, only the seeds collected from the populations at the low tidal flat germinated, and the germination rates decreased by 90% compared with the control group ( Figure 2b ). The mixed linear model showed that soil salinity, tidal flat and their interactions affected the seeds germination, and the salinity was the main factor ( Table 1 ).

The germination rates of P. australis seeds collected from populations at the high, middle and low tidal flats inside the dike, and stored at the room temperature (20 ± 1°C) decreased compared to those stored at the low temperature (4°C). The germination rate of seeds collected from the populations at the L tidal flat and stored at the room temperature decreased by 8%-12% compared to those stored at 4°C condition ( Figures 3a–d ), but the germination rates of seeds were basically kept the same when the treatment salinity was 2% ( Figure 3 ). When the treatment salinity was 0.5%, the seed germination rates of the P. australis populations in the L, L(IN), M and H tidal flats under room temperature storage were 43.0% ± 12.23, 28.0% ± 14.42, 44.4% ± 6.36 and 49.2% ± 5.02, respectively. However, under the same treatment salinity, the seed germination rates of the P. australis populations in the L, L(IN), M and H tidal flats under low-temperature storage were 62.6% ± 6.63, 12.7% ± 3.05, 30.0% ± 2.0 and 35.05% ± 5.47, respectively. When the salinity was treated at 1%, 1.5% and 2% respectively, the seed germination rates of L(IN), M and H tidal flats were increased under room temperature storage conditions compared with low-temperature storage conditions ( Figures 3b–e ). Regardless of the treatment salinity conditions, the germination rate of L tidal flat seeds after low-temperature storage was higher than that of seeds under room temperature storage conditions ( Supplementary Tables S1 , S2 ).

Total nitrogen (TN), organic carbon content (Organic C), pH and available phosphorus (AVIP) of different soil layers in the original habitats of P. australis populations were significantly different. The contents of NO₃ ^–N (0.18 ± 0.07 mg/kg), NH₄ ⁺-N (1.02 ± 0.17 mg/kg) and TN (0.1 ± 0.05g/kg) show a trend that they were higher at high tidal flat than at middle (NO₃ ^–N (0.20 ± 0.07 mg/kg), NH₄ ⁺-N (0.91 ± 0.04 mg/kg) and TN (0.05 ± 0.02g/kg)) and low tidal flats (NO₃ ^–N (0.17 ± 0.04 mg/kg), NH₄ ⁺-N (0.96 ± 0.17 mg/kg) and TN (0.09 ± 0.02g/kg)). Soil physicochemical indexes of different tidal flats were all significantly different, except for total phosphorus (TP). 0–10 cm soil layer salinity at the low tidal flat (0.5 ± 0.18 PSU) was higher than that at the middle (0.33 ± 0.08 PSU) and high tidal flats (0.29 ± 0.14 PSU), respectively, and soil salinity at 0–10 cm layer was higher than that at 10–20 cm layer. The soil pH of the low-tidal flat (8.12 ± 0.24) and the middle-tidal flat (8.31 ± 0.39) were higher than that of the high-tidal flat (7.92 ± 0.38), while the soil salinity of the 0–10 cm layer was lower than that of the 10–20 cm layer. This phenomenon indicates that the soil gradually tends to become salinized from the high tidal flat to the low tidal flat. The soil available phosphorus (AVIP)content in the tidal flat was the highest, being 21.15 ± 3.55 mg/kg. The variation trends of soil water content and AVIP were consistent with those of soil salinity at the tidal flats, but were opposite with those of soil salinity between soil layers. There was no interaction between soil physicochemical properties of soil layers and tidal flats ( Table 2 ).

The correlation analysis between seed germination parameters under 4 °C seed storage condition and soil physicochemical factors of original habitats of P. australis populations showed that germination rates, germination potentials, germination indexes and germination speeds were significant positively correlated with soil salinity, AVIP and water content of original habitats of P. australis populations, while, they were negatively correlated with soil NO₃ ^–N in the same soil layer. Among them, seed germination rates and germination speeds had the greatest correlation with soil salinity of original habitats of P. australis populations, which was consistent with the results of seed germination experiment with salt addition treatment ( Figure 4a ). When seeds stored under the room temperature, seed germination rates were positively correlated with soil N/P and soil microbial biomass carbon in 10–20 cm layer, and negatively correlated with soil TP content of original habitats of P. australis populations. Seed germination potentials were positively correlated with soil pH value, microbial biomass carbon content in 0–10 cm layer, and AVIP in 10–20 cm layer soil, and negatively correlated with soil NH₄ ⁺-N in 10–20 cm layer of original habitats of P. australis populations. Seed germination indexes and germination rates were positively correlated with soil AVIP, salinity and soil moisture in 10–20 cm layer, and negatively correlated with soil NO₃ ^–N of original habitats of P. australis populations ( Figure 4b ).

In order to further explore the effects of interaction between germination parameters of seeds under different seed storage temperatures and the key soil physicochemical factors of original habitats of P. australis populations, RDA was conducted, which showed that germination rates of seeds collected from the populations at different tidal flats and stored under 4 °C were significantly different. Taking the germination rates of seeds stored at 4°C as response variables, and soil salinity, NO₃ ^–N, NH₄ ⁺-N, TN, TP, organic carbon content, moisture content, pH, AVIP, MBC and MBN of original habitats of P. australis populations as the explanatory variables, RDA was further conducted, which showed that the first and second ranking axes of RDA explained 65.77% of the impact of 0–10 cm layer soil physicochemical properties on the variation of seed germination performance. In the RDA ranking diagram, AVIP, C/N of 0–10 cm soil layer, C/P of 0–10 cm layer, and NO₃ ^–N of 10–20 cm layer of original habitats of P. australis populations were most closely related to the first axis. In addition, soil salinity, pH at 0–10 cm soil layer, and soil water content in 0–10 cm layer, MBN and NO₃ ^–N were also strongly correlated with the first principal component. In the soil physicochemical properties, soil TN had the greatest correlation with the second principal component, which followed by soil C/N. Cluster analysis showed that the germination rates of seeds collected from the populations at the middle and high tide flats had a strong correlation with NO₃ ^–N, C/N, N/P and NH₄ ⁺-N of 0–10 cm and 10–20 cm soil layers, and the germination rates of seeds collected from the populations at the low tidal flat were correlated closely with other soil physicochemical properties of original habitats of P. australis populations ( Figure 5a ).

The relationship between soil physicochemical properties of original habitats of P. australis populations and the germination rates of seeds collected from the populations at different tidal flats and stored at the room temperature was also analyzed through RDA, which showed that total RDA results explained 46.68% of the impact of soil physicochemical properties on the seed germination performance, and the first axis included most of the effective results ( Figure 5b ). Result showed that AVIP, C/N, C/P and NO₃ ^–N content of 0–10 cm soil layer, were most closely related to the first axis. In addition, soil pH, TP in 0–10 cm soil layer, and soil C/N, NH₄ ⁺-N in 10–20 cm soil layer were also strongly correlated with the second axis. Cluster analysis showed that germination parameters of seeds collected from the populations at the middle and high tidal flats, respectively, had strong correlation with soil NO₃ ^–N, TP, C/N and NH₄ ⁺-N in 10–20 cm soil layer. The germination rates of seeds collected from the populations at the low tidal flat were significantly correlated with other soil physicochemical properties ( Figure 5b ). The RDA analysis results on the relationships between germination rates of seeds collected from the populations at different tidal flats and stored at different temperatures and the soil physicochemical properties showed that seed germination rates had the same trend of correlation with soil NO₃ ^–N content, AVIP, pH and C/N of the soil layers of original habitats of P. australis populations ( Figure 5 ).

The germination rates, germination potentials, germination indexes and germination speeds measured in this study were represented as a germination parameter of the germination strategy to construct a structural equation model (SEM). The SEM results showed that under the seed storage condition at 4°C, soil water, N/P1, N/P2 and soil microbial biomass carbon content of original habitats of P. australis populations affected seed germination. The path coefficient between soil moisture, soil N/P1, N/P2, MBC and seed germination was 0.14, 0.33, 0.2 and 0.14, respectively ( Figure 6a ). The tide level directly reduced the soil organic carbon content of original habitats of P. australis populations, and then indirectly led to the decline of the germination rates of seeds stored at 4°C ( Figure 6a ). The tidal level promoted the increase of soil moisture content, nitrogen and phosphorus ratio of original habitats of P. australis populations, which indirectly led to the increase of seed germination rates under the same seed storage conditions ( Figure 6a ). Soil pH of original habitats of P. australis populations was the most important factor affecting seed germination, when seeds were stored at room temperature, and the path coefficients between soil water content, C/P1, pH2 and seed germination was 0.14, 0.17 and 0.3, respectively ( Figure 6b ). The path coefficient between tidal flats of original habitats of P. australis populations and germination of seeds stored at room temperature was -0.16, while the path coefficient between tidal flats and germination of seeds stored at 4°C was -0.97. The tide level directly increased soil pH, which indirectly led to increase in germination rates of seeds collected from the populations at different tidal flats and stored at room temperature, while, the decrease of soil moisture content and phosphorus content among different tidal flats indirectly lead to the decrease of germination rates of seeds under the same storage conditions ( Figure 6b ).

The two essential and critical developmental stages in the life cycle of an annual plant were regarded as flowering and seed germination (Ehrlén et al., 2015). Seed germination is usually the first step in the establishment of a plant population. Numerous studies demonstrated the effects of environmental factors such as temperature, salinity, soil moisture and other soil factors on seed germination (Gul et al., 2013; Fernandez-Torquemada and Sanchez-Lizaso, 2013; Gorai et al., 2006; Kettenring and Whigham, 2009). Soil salinity was found to be one of the vital factors directly or indirectly affecting the germination of seeds of P. australis populations along the Dongtan salt marsh wetland ( Figure 2 ). Generally, the germination rates and germination potentials of P. australis seeds decreased with the increase of soil salinity (Wijte and Gallagher, 1996; Zhao et al., 2021). Seed germination rates of P. australis populations in the Yellow River Delta were also found to be inhibited seriously under the condition of high salt (2% - 3%) (Yu et al., 2012). Our research revealed that P. australis seed germination rates topped 80% under 0 salinity conditions ( Figure 2 ). However, when exposed to 2% salinity, the germination rates of P. australis seeds from diverse tidal flats decreased to varying degrees, and some even failed to germinate. In addition, seeds at 2% salinity from middle tidal flats failed to germinate compared to those from the low tidal flat (80.9% ± 0.06%). These suggested that P. australis seeds from the low tidal flat possessed a higher tolerance and adaptability to salinity. Adaptation to a saline habitat that is the low-tidal zone is a likely explanation for our observation (Guan et al., 2016; Haraguchi, 2014). Under the two storage temperature conditions, the germination rate, germination index, germination potential and germination speed of L and L(IN) were all different. This might be due to the construction of the dike, which led to the reduction of hydrological connectivity, thus resulting in the differentiation of germination strategies between these two populations. Consequently, P. australis population could potentially establish and expand gradually based on the seeds of pioneer P. australis genets in low tidal habitats with relatively high soil salinity.

The germination rates of P. australis seeds in the Yangtze River Delta were slightly affected when salinity reached 1%. In contrast, the same salinity level led to a decreased germination rate for P. australis seeds collected from inland regions (Yu et al., 2012). Researchers studied the salt tolerance of P. australis seeds in salt marsh wetlands and inland estuaries, revealing the same pattern that the germination rates of P. australis seeds collected from the salt marsh wetland population were higher than those of seeds collected from the estuarine population (Haraguchi, 2014). The two-way ANOVAs testing the effects of salinity, the tidal flats and their interactions on the germination indexes of P. australis, results showed that salinity was the main factor affecting germination rate (Gt) and germination speed (Gs), while the tidal flat was the main factor affecting germination index (GI) and germination potential (GP) ( Table 1 ). It is well known that salinity is negatively correlated with Gt and GI, low concentrations of NaCl induce seed dormancy, and high concentrations of NaCl induce osmotic stress and ionic toxicity (accumulation of Na⁺ and Cl^-) and inhibit seed germination (Rajabi Dehnavi et al., 2020). In fact, for seeds in a salinity environment, due to the low osmotic potential of the germination medium, salinity can change the imbibition process of seeds to water, and water diffused from low salinity to high salinity, delaying the absorption of water by seeds, thus reducing germination (Farooq et al., 2015). Under salt stress, Na⁺ and Cl⁻ may be absorbed by seeds, and the toxic effect of NaCl may occur, thus slowing down the germination index and germination speed (Ashrafi and Razmjoo, 2015). The high germination rates, fast germination speeds and high germination vitality were found in P. australis seeds collected from the low tidal flat population growing in habitats with high soil salinity and high tide disturbance frequency (Baldwin et al., 2010; Briea, 2006). Our study also has similar results, low tidal flat P. australis seed Gt is higher, Gs is faster and GP is better than other tidal flats. Table 1 shows that there are significant differences in germination parameters (Gt, Gs, GP, and GI) between different tide levels. As shown in Figure 4 , different germination parameters are significantly correlated with some factors in soil physicochemical properties, which indicates that soil sediments and nutrient availability directly or indirectly affect seed germination traits of the reed populations. Such seed germination strategy may help accelerate the recruitment and establishment of P. australis populations at the low tidal flat of the Dongtan salt marsh wetland. Greenberg et al. (2001) revealed that most of the time plant seeds tended to adopt a “sit or wait” germination strategy, i.e., seeds sitting in their habitats and waiting for favorable opportunities for germination caused by environmental disturbances. However, the germination rates, speeds, and vigor of P. australis seeds from middle and high tide levels were inferior to those from low tidal flat populations. This suggested that the underground seeds of P. australis populations in middle and high tide levels might employ a “sit or wait” germination strategy, i.e., they could germinate when favorable opportunities arose. Along the Dongtan salt marsh wetland, the local adaptation of P. australis populations to the edaphic environmental conditions (especially the soil salinities) in the habitats shaped the sexual reproduction and seed germination strategies of P. australis populations as well as the seedling growth, showing the maternal effect in the local populations. In other words, the sexual progeny of P. australis populations at different tide levels in the Dongtan salt marsh wetland, apparently inherited the adaptive characteristics of the original ancestral populations, suggesting that when the sexual offspring of P. australis populations colonize the same habitats of their parental populations at different tide levels, they might perform well and maintain the long term structural and dynamic stability of the local populations.

70% of plant seeds have different types and degrees of dormancy, and plants need specific mechanisms to avoid adverse environmental conditions in nature, such as low temperature and dryness, and try to ensure that plants germinate in an environment suitable for seedling establishment (Afroze and O’Reilly, 2017; Baskin et al., 2003). Dormancy of gramineous seeds is a common phenomenon, and low temperatures in winter could efficiently break the dormancy of seeds of P. australis, suggesting the rejuvenation of seed germination vigor needs a short and proper cold treatment (Milberg, 1999; Li et al., 2013). The germination rates of P. australis seeds collected from the low tidal flat populations under 4°C storage condition for a while was significantly higher than those under the storage condition at room temperature (20 ± 1°C) for the same duration ( Figures 2 , 3 ), indicating that different storage temperatures affected the germination vigor of P. australis seeds. The low-temperature storage condition (4°C) promoted the germination rates. Whereas, when P. australis seeds collected from the populations at the middle and high tidal flats were stored in the room temperature condition, the germination rates of seeds were improved, compared with those stored in 4°C condition ( Figure 3 ), which implied that low temperature, as one of the environmental seize factors for seed germination, regulated the after ripening of seeds collected from the populations at the low tidal flats. Low temperature mainly showed slow germination speed and low germination rate (Akinnuoye and Modi, 2015). As a major environmental stress factor, low temperature can break the original metabolic balance of seeds, destroy the enzyme system, accumulate toxic substances, and thus affect the growth and development of plants (Greenwood and MacFarlane, 2006). Temperature is critical to plant growth and development, and many plants in temperate regions can remember past winter seasonal temperature drops and undergo changes in after ripening patterns to ensure the survival and/or reproductive success of offspring (Gao et al., 2022; He and Li, 2018). However, our findings demonstrated that low-temperature treatment enhanced the germination of P. australis seeds collected from low tidal flat populations in the Dongtan coastal salt marsh wetland in a subtropical climate zone. This improvement was notable despite the lack of the same effect on seeds from middle and high tide flats.

The Dongtan wetland (especially the frontier low tidal flat) exposes to frequent wind and sea tidal impact, whose standing water often freezes in winter whenever there is cold weather; thus, after long-term adaptation, the germination of seeds from the low tidal flat populations required moist chilling condition was quite understandable. The chilling environment near the coastal wetland, the metabolism of stored protein was changed by inducing ROS production and reducing the activity of protease in seed cells, thus affecting the seed germination rate (Shutov et al., 2003). Therefore, the seeds produced by the low tide flat P. australis population need after-ripening, breaking dormancy after low temperature environment, and successfully germinating in suitable condition. As for the fact that the germination of seeds collected from the middle and high tidal flats with less ecological stresses in the habitats no longer required after ripening might be attributed to the comprehensive impacts of multiple physiological and environmental factors. Our results showed that the contents of NH₄-N and NO₃-N in the soil were higher than those at the low tidal flat. The contents of Organic C, MBC and MBN in high tidal flat were higher than those in low tidal flat ( Table 2 ). This means that the middle and high tide zone environment can provide a relatively stable soil sediment and availability of plant nutrients compared to the low tide zone environment with high frequency tidal disturbance. Meanwhile, previous studies have found that the soil structure of the middle and high tide zone of Dongtan salt marsh was sandy or silty soil, while the soil quality of the low tide zone was sandy clay (Liu et al., 2021; Jia et al., 2022). Sandy soil and sandy loam with large pores and low organic matter content are suitable substrates for the germination of Jatropha curcas seeds. The germination rate and germination speed of seeds with clay texture were lower (Valdés-Rodríguez et al., 2013). The soil texture of P. australis population at different tide levels affected the proportion of soil pore water and the content of organic matter, which would change the germination of seeds and the growth and development of plants. The seeds collected from low tidal gradients germinate better when stored at low temperatures, while seeds from middle and high tidal gradients germinate better when stored at room temperature. This phenomenon is one of the ecological drivers for seed germination in this species, as at a low tidal gradient, seeds may experience low and fluctuating temperatures due to a low water table. In contrast, at a middle and high tidal gradient, seeds may experience warmer and constant temperatures due to higher water table (Xiao et al., 2009).

Seed germination is the first stage of plant life, but the abiotic stress environment during seed germination can reduce seed germination by increasing seed quality degradation and reducing seed germination potential and vigor (Wang et al., 2024). The low temperature was often the key factor stimulating the seed germination vigor, while other conditions, such as suitable water, oxygen and nutrients, are still necessary (Lombardi et al., 2019). SEM results showed that germination of P. australis seeds stored at 4°C was affected by soil MBC, N/P ratio and soil water content in the original population habitats from which seeds were collected ( Figure 6 ). Significant differences in soil pH, TN and organic carbon content in soil layers (p<0.05) were found among habitats from which P. australis seeds were collected, and the soil salinity, soil available nitrogen decreased with the increase of tidal flat, while other soil indexes showed no significant differences among the habitats ( Table 2 ). The availability of certain elements is related to soil pH (Al-Wabel et al., 2018), thus, indirectly influencing the plant growing in the habitat. Soil pH is closely related to soil microbial diversity and indirectly affects soil MBC (Naz et al., 2022; Wan et al., 2021), while soil N/P affects the mineralization rate of soil, and further affect soil salinity synergistically with soil moisture (Miller and Geisseler, 2018), and indirectly affects seed germination rate (Feng et al., 2021).

As the mixed results of SEM analysis suggested, the relationship between soil properties, nutrients, and germination may be more complex. Former results also revealed the coupling effects of water, nitrogen and phosphorus in regulating the germination of wetland plant seeds (Zhao et al., 2019). Nitrate plays a role as a signaling molecule to mediate seed germination (Zhang et al., 2023; Li and Wang, 2015). Both endogenous and exogenous nitrate and phosphorus may affect seed germination (Song et al., 2016; Plassmann et al., 2008; Sims et al., 2012). The genetic diversity of P. australis patches in the Rhodes River Basin in the United States shows that almost all population patches are composed of multiple genotypes, and many patches are genetically different from other patches, which indicates that cross-fertilization is one of the sources of genetic diversity (McCormick et al., 2010). Therefore, the possibility of cross-fertilization and gene exchange between different populations may also affect the seed traits such as salt tolerance and germination rates across tidal zones (Lambertini et al., 2020). Except that, such as the water pollution, sediment and soil texture of tidal flats, structure, and availability of nutrients for plants also have direct or indirect influences on the seed germination strategy (Antonio Alburquerque et al., 2014; Bocchese et al., 2008; Peng et al., 2022; Satyanti et al., 2019; Wang et al., 1994). Chongming Dongtan Wetland is located in the lower reaches of the Yangtze River Delta, bearing the pressure of water pollution caused by population activities in the upper reaches, and the pollutants such as nitrogen, phosphorus and COD in the water can be removed by P. australis (Ho et al., 2024). Water pollution may cause significant differences in soil sediments between different tidal levels, and then affect the seed germination rate of different tidal flat populations. Wetland soil is the source, sink and transfer source of chemical pollutants. Anthropogenic activities (such as wetland reclamation, fertilization, and sewage discharge) have exacerbated changes in soil properties (such as SOM, pH, and Eh) of wetland soils/sediments (Bai et al., 2011). Absorption by the roots of P. australis can increase the residence time of these pollutants in the wetland, and then effectively reduce the diffusion of pollutants to the surrounding environment (Wang et al., 2015). In this study, P. australis seeds in coastal wetlands with high salinity showed a higher germination rate, which may be a trade-off in the face of tidal frequency and harsher environment.