Matures seeds were collected in late September 2020 from the wild S. corniculata populations in the Ertou wetlands in the MNNR (45°53′ – 45°55′ N, 123°36′ – 123°41′ E). The collected seeds were air-dried indoors under dark conditions and then stored in cloth bags at 4°C until the start of the experiment in January 2022. Healthy S. corniculata seeds were surface-sterilized using 75% ethanol for 10 min and then rinsed 9 times with distilled water, and the water was wiped off the seeds surface with sterile filter paper in the super clean bench before being used in the experiments.

The experiment was conducted in the Key Laboratory of Wetland Ecology and Environment of the Northeast Institute of Geography and Agroecology, Chinese Academy of Sciences. We selected six priming treatments: MT10 (10 μM melatonin), MT20 (20 μM melatonin), MT50 (50 μM melatonin), MT100 (100 μM melatonin), MT200 (200 μM melatonin), and CK (distilled water). Seeds were incubated in each priming solution for 12 h. For each treatment, 25 seeds were evenly placed in Petri dishes (diameter 9 cm) containing double Whatman No. 1 filter paper soaked with 2 ml of 30 mM NaHCO₃ solution (the salt solution concentration data comes from the MNNR), and sealed with Parafilm^® to prevent moisture evaporation. The seeds were then germinated in an artificial climate box (RDN-300B-4, China) under an alternating cycle of 12/12 h light and dark at an average temperature of 22.5/8.6°C (the climate data represents field conditions in May in the MNNR). For each germination treatment, five independent replicates were used (three petri dishes were used to test the germination rate of S. corniculata seeds, while two petri dishes were used to cultivate testable seed materials). Germinated seeds were recorded daily, and germination was considered having occurred when the radicle ≥ 2 mm. The entire experiment lasted for 7 days (from 6^th to 12^th January, 2022) and partial seeds were harvested on days 1, 3, 5, and 7, and immediately frozen in liquid nitrogen and stored at −80°C until further analysis of antioxidant enzymes and reserve substances.

In order to describe seed germination changes, the germination rate (GR), germination speed (GS), and germination index (GI), were measured according to following Biju et al. (2017):

where n is the number of germinated seeds, N is the total number of the tested seeds, Nt is the percentage of germinated seeds at the t^th days, G_t is the number of germinated seeds in t days, and Dt is the corresponding germination days.

The activities of superoxide (SOD), peroxidase (POD), and the malondialdehyde (MDA) content were determined via WST-8, visible-spectrophotometry and TBA-spectrophotometry by using detection kits (M0102B, M0105B, M0106B, Michy Biomedical Technology Co., Ltd., Suzhou, China), respectively. The contents of starch (ST), soluble sugar (SS), and soluble protein (SP) were determined by detection kits (M1101B, M1503B, M1806B, Michy Biomedical Technology Co., Ltd., Suzhou, China) which are based on anthrone-sulfuric acid colorimetry, anthrone colorimetry, and Lowry’s method, respectively. The fat content was determined by the supercritical fluid-Soxhlet extractor using petroleum ether as the extraction buffer (Soriano et al., 2011). All the samples were conducted in three independent biological replicates.

Statistical analysis was conducted using SPSS 22.0 (SPSS, Chicago, IL, United States) and Origin Pro 2022 (OriginLab, Northampton, MA, United States). The data normality and homogeneity were checked before further analyses; seed GR, SOD, and ST were log-transformed and square root transformed to meet the assumptions of homoscedasticity. Effects of melatonin priming on S. corniculata seeds under salinization stress were evaluated using one-way analysis of variance (ANOVA). Duncan’s multiple comparison test was used to determine the significant differences at the 0.05 significance level. Pearson’s correlation analysis was performed to identify the correlational relationship between antioxidant enzymes, reserve contents, as well as seed germination. Structural equation model (SEM) analysis was performed to evaluate the direct and indirect regulatory pathways of melatonin priming on S. corniculata seed germination in SPSSAU¹.

Among the different priming treatments, we observed significant differences in the S. corniculata seed GR (F = 20.64, p < 0.05), ranging from 51.33 to 87.33%, with the maximum and minimum being recorded in the M50 and M200 treatments, respectively (Figure 1A). Priming treatment significantly affected the GS (F = 13.20, p < 0.05), which peaked in M50 (7.47) and was 1.80 times greater than in the M200 treatment (Figure 1B). Additionally, the GI was significantly affected by the priming treatments (F = 19.75, p < 0.05). The GI values recorded in the M50 treatment (27.41 ± 0.89) were 61.05 and 92.28% higher than those in CK and M200 treatment, respectively (Figure 1B).

We identified significant differences in the activities of antioxidant enzymes of S. corniculata seeds at each stage under different melatonin priming treatments (Table 1 and Figure 2). The antioxidant enzymes activities showed a gradually increasing trend during germination stage, with both SOD and POD activities peaking in the M50 treatment on the 7^th day, and which were 2.59 and 3.67 times greater than the M200 treatment during the same period (Figures 2D,H). The degree of membrane lipid peroxidation in S. corniculata seeds was significantly affected by the melatonin priming treatments (Figure 2). The MDA content gradually decreased with germination time, with its lowest value appearing in the M50 treatment on the 7^th day, and it was 27.33% lower than that on the 1st day (Figures 2I,L).

Melatonin priming significantly affected the reserve mobilization process in S. corniculata seeds (Table 1). The trend of changes in the starch and fat contents was basically uniform across each stage, i.e., they decreased significantly with the increase in germination days, where the seed starch and fat contents under M50 treatment decreasing up to 40.36 and 8.87%, respectively (Figures 3A,D). The soluble sugar content continuously increased with the germination stage, with the increase ranging from 13.90 to 56.11% under each melatonin treatment (Figure 3B). Additionally, the soluble protein content initially increased before decreasing during the process of S. corniculata seed germination (Figure 3C). By the 7^th day, the soluble protein content of the M200 treatment was the highest, thus indicating that its mobilization level was significantly lower than in the other treatments.

We observed significant positive relationships between SOD, POD, and soluble sugar, while there were significant negative relationships with MDA, starch, soluble protein, and fat (Figure 4). Additionally, the GR was significantly correlated with antioxidant enzymes’ activity and reserve mobilization. The SEM analysis showed that the differences in melatonin concentration priming were directly related to the antioxidant enzymes’ activity, reserve mobilization, and seed germination (Figure 5). Meanwhile, melatonin priming also indirectly affected S. corniculata seed germination by affecting their antioxidant enzymes’ activity and reserve mobilization. Additionally, the degree of reserve mobilization directly affected the level of antioxidant enzymes’ activity.

Seed propagation is preferred for maintaining plant genetic variation, and it will also play an important role in the process of near-naturalized wetland restoration (Kim, 2019). The seed germination process is usually affected by salinization stress, which leads to the decrease in germination quality and even results in an incomplete plant life cycle (Jia et al., 2018). Multiple studies have shown that melatonin is a major regulator not only involved in seed germination, but also in diverse physiological and biochemical processes like stress resistance-related gene expression, boosting the antioxidant defense system, and reserve mobilization (Cao et al., 2019; Chen et al., 2021; Zhang T. G. et al., 2022). Our results showed that melatonin priming positively affected seed germination of S. corniculata in a concentration-dependent manner, while significantly increasing the GR, GS, and GI of S. corniculata seed under the M50 treatment. This finding was consistent with Qin and Zeevaart (2002) and Ogawa et al. (2003), where they found that exogenous melatonin priming upregulates the expression of gibberellins (GA) biosynthesis genes (GA20ox and GA30ox), while downregulating the key abscisic acid (ABA) biosynthesis genes (LbNCED1 and LbNCED3), both of which affect seed germination. Additionally, the appropriate melatonin priming concentration activated the SOS pathway to improve the seed’s salinity tolerance at the germination stage by upregulating the expression of chloride channel proteins (CLC1 and CLC2) (Li X. J. et al., 2017). Meanwhile, priming with the appropriate melatonin concentration could up-regulate the ethylene, kinetin, and brassinolide biosynthetic gene expression and the plant hormone signal pathways in response to salinization stress (Zhang Y. X. et al., 2021). However, the mitigative effect of melatonin priming treatment on salinization stress could be counterproductive at high concentration. Our results suggested that the M200 treatment inhibited the S. corniculata seed germination as compared with the CK treatment. This may be because high melatonin concentration promoted the synthesis of endogenous ABA, which regulated seed germination via ABI3, ABI4, and ABI5 and promotes dormancy via the response factors like SCHNARCHZAPFEN (SNZ) and ABA overly sensitive 5 (ABO5) (Liu et al., 2010; Zou et al., 2013). This explanation has been supported by the fact that 1,000 μM melatonin treatment significantly inhibited the Arabidopsis seed germination as compared with the untreated control in Lv et al. (2021).

The excessive accumulation of reactive oxygen species (ROS) not only has irreversible negative effects on seed germination, but also affects the future seedling growth (Goud and Kachole, 2011). Antioxidant system is one of the major physiological defense systems protecting against salinization-induced oxidative stress by improving the antioxidant enzyme activity and antioxidant contents (Ibrahim, 2016). In this sense, studying the antioxidant defense is crucial for a holistic understanding of priming treatments. Previous studies have demonstrated that melatonin get through the free radical scavenging cascade, extends the free radical scavenging capacity of antioxidants to its multi-stage metabolites. Especially N1-acetyl-N2-formyl-5-methoxykynuramine, a primary metabolite derived from melatonin, also has strong antioxidant capacity, thus effectively protecting organisms from oxidative stress (Balabusta et al., 2016). Simultaneously, melatonin helps scavenge free radicals through its receptors by modulating the antioxidant enzymes in the cells and tissues in a dose-dependent manner (Li et al., 2018; Cao et al., 2019). Our results showed that the SOD and POD activities of S. corniculata seeds continued increasing with the germination stage, which was significantly higher than others in the M50 treatment. Exogenous melatonin priming may have indirectly upregulated the expression of antioxidant enzyme and related genes, which then improves their efficiency as antioxidants (Arora and Bhatla, 2017; Wang et al., 2017). However, other studies have demonstrated that exogenous melatonin priming did not change the antioxidant enzymes’ activity of Lolium perenne (Zhang et al., 2016; Lu et al., 2022), and we suspect that melatonin possibly uses different physiological mechanisms to strengthen the antioxidant defense. MDA is the indicator of oxidative-stress induced membrane damage under salinization stress (Ibrahim, 2016). Our study results showed the MDA content in S. corniculata seeds continuously decreased during the experiment, and it decreased up to 27.33% in the M50 treatment. This result was in agreement with previous findings indicating that a negative correlation exists between the MDA content and the activity of antioxidant enzymes (Esfandiari et al., 2008; Younesi and Moradi, 2015). The possible explanation is that melatonin treatment increases the autophagy induction capacity and thereby reduces the accumulation of oxidized proteins, while also simultaneously controlling the ROS accumulation by upregulating the antioxidant enzymes activity and their transcript levels to mitigate membrane lipid peroxidation under salinization stress (Wang et al., 2015; Sun et al., 2020; Zhang T. G. et al., 2022). Additionally, it improves the activity of non-enzymatic antioxidants by regulating the activity of enzymes in the ASA-GSH cycle, thereby further improving the free radical scavenging capacity (Ni et al., 2018). Simultaneously, by up-regulating the genes related to glutathione metabolic pathway and activating the downstream signal transduction pathway, also may help explain how melatonin priming improves the antioxidant defense of seeds under salinization stress (Zhang Y. X. et al., 2021).

It is well known that the seed reserve is an important factor which affects germination characteristics, and reserve mobilization plays an important role in the energy supply and maintains osmotic balance (Zhao et al., 2018; Lei et al., 2021). Recent experimental data suggested that the priming treatment regulates the carbohydrate, protein, and fat metabolism, thereby resulting in increased seed germination (Li et al., 2019; Chen et al., 2021). However, the effect of melatonin priming treatment on reserve mobilization under abiotic stress still needs to be studied according to the target species. Our results showed that there is a strong negative correlation between starch and soluble sugar content variation. A plausible explanation for this phenomenon could be that melatonin priming upregulated the α-amylase gene expression, to ensure that starch is broken down into smaller and simpler molecules which will be used as an energy source by the organism for seed germination (Cao et al., 2019; Yan et al., 2020). Previous studies also found that the priming treatment induced a phased effect on the change of seed soluble protein content under abiotic stress (Adawy et al., 2003). Priming treatment induced seeds to adopt the defensive strategy of accumulating soluble protein for improving the cell’s water holding capacity and protect its membrane (Li et al., 2010). During the later stage of germination, hydrolysis of storage proteins provides the energy for both the hypocotyl growth and new protein synthesis (Wahid and Bounoua, 2013). Meanwhile, the fat metabolism also pushed the seed’s developmental switch from dormancy to germination (Footitt et al., 2002), which was similar to our data.

The action pathway of priming treatment is the basis for revealing the driving mechanisms. The SEM analysis showed that there are both direct and indirect effect pathways of melatonin priming on the S. corniculata seed germination under salinization stress conditions (Figure 5). On the one hand, melatonin priming promoted seed germination by regulating the expression of stress tolerance, hormone synthesis, and metabolism-related genes (Zhang Y. et al., 2021). On the other hand, priming treatment also upregulated the antioxidant enzyme activities and accelerated the level of reserve mobilization, ultimately improving the seed germination (Arora and Bhatla, 2017; Cao et al., 2019; Lu et al., 2022). This study is the first to reveal the effect pathways of melatonin priming on the S. corniculata seed germination under salinization stress, and also proved the feasibility of melatonin priming in improving the germination of S. corniculata seeds. It is also worth noting that the positive effects of the melatonin priming we found were concentration-dependent, which reconfirms our hypothesis.