Introduction

Seeds are the foundation for agricultural production. Normal seed germination and seedling establishment are critical to the crop growth and yield formation. As an important component of seed quality, seed vigor reflects the ability of seeds to emerge quickly and neatly under complex field conditions. High vigor seeds are in strong association with robust plants and higher yields (Catusse et al. 2011).

Plants undergo a series of complex changes in morphology and physiology from flowering to pollination, eventually developing into mature seeds (Kumar et al. 2018). Changes in moisture content, reserved nutrients and enzymatic activity are closely connected to the formation of seed vigor (Zhu et al. 2016; Chen et al. 2020, 2021). Carbohydrates, lipids and proteins are important nutrients accumulating during seed development. Oil crop seeds contain substantial free fatty acids at the early stage of development, which decrease gradually with seed maturation to cause synthesis of complex oils (Norton and Harris 1983). In addition, metabolic changes in plant hormones (e.g. gibberellin (GA), abscisic acid (ABA) and auxin) and plant growth substances [e.g. polyamines (PAs)] also generate vital roles in seed development. The above-mentioned plant hormones may function in the growth arrest before seed maturity, reserve substances accumulation, external seed tissues development, as well as the seed germination (Monpara et al. 2019).

Seeds must be harvested quickly within the most appropriate period whether it is for economic or seed purposes. Lodging and shattering may easily occur when the harvesting is delayed, while hot and humid weather usually results in germination on the ears (Zhu et al. 2016; Li et al. 2004). Additionally, premature harvesting results in underdeveloped, insufficiently plumped, poorly vigorous seeds. It is of great practical significance to investigate the maturity-vigor relationship of crop seeds, and to determine the best harvesting time, thus ensuring the quality and safety of crop production and seed multiplication.

Sunflower (Helianthus annuus L.), originated in North America, has seeds rich in unsaturated fatty acids, multivitamins and trace elements, which is one of the major oil crops globally (Buti et al. 2011; Celus et al. 2018). In China, with a planting area of 1.15 million hm2 and a total output of 299 t, sunflower is the third largest oil crops after rapeseed and peanut (Zhang and Zhang 2018). Due to its high ornamental, edible and oil-use values, recent years have witnessed yearly increasing planting area of sunflower in China. However, research on its seed maturation mechanism and optimal harvest period has been poorly reported. Our study clarified the optimal harvest period of sunflower seeds, and established the relationship of physiological and biochemical changes with seed vigor during the development of sunflower seeds by measuring the dynamic metabolic changes in nutrients, antioxidant enzymes and phytohormones. Moreover, the obtained findings can provide theoretical and practical guidance for the development of sunflower industry.

Materials and methods

Experimental materials

The sunflower (Helianthus annuus L.) hybrid F1 seeds produced by hybridization of varieties S175 (♀) and T101 (♂) were used in this study. The female parent floral disc was bagged prior to pollination, which was then artificially pollinated to obtain hybrid F1 seeds. Consistent field management was implemented during growth. Meanwhile, the experiments were carried out at the Huzhou experimental farm.

Sampling of sunflower seeds

Samples were collected at 3 days intervals 10–31 days after following pollination. Ten female parent sunflower plants were randomly selected for each sample, whose seeds were peeled off, treated with liquid nitrogen and then stored in a − 80 °C ultra-low temperature freezer for subsequent analysis. Apart from that, fresh sunflower seeds were dried at room temperature to a target moisture content of 9%, and then applied for germination test.

Determination of moisture content, dry/fresh weight ratio and volume of sunflower seeds

The moisture content of seeds was measured with a fast grain moisture analyzer and the volume of seeds was measure with a graduated cylinder. The fresh weight of seeds was weighed with a thousandth balance. After that, the seeds were dried in an 80 °C oven for 24 h, and then measured to 0.01 g accurately. Three replicates were set up for each treatment, with 50 seeds per replicate.

Germination test

Standard germination test was performed with sunflower seeds, which were dried at room temperature to a target moisture content of 9%. The seeds were sterilized with 0.1% NaClO solution for 15 min, and later placed on 6 cm × 12 cm wet paper rolls for germination. After that, the germination paper rolls were placed in a constant 25 °C incubator under an illumination cycle of 12 h light/12 h dark. Four replicates were set up, with 50 seeds per replicate. Germination criterion was seed radicle breaking through the seed coat by 1 mm. The germination number was recorded every day. In addition, the germination energy (GE) and germination percentage (GP) were calculated on the 4th and 7th days of germination, respectively. The computational formula for germination index (GI) = ∑(Gt/Tt), where Gt denotes the germination number on day t, and Tt represents the corresponding number of days. The computational formula for vigor index (VI) = GI × seedling dry weight.

Determination of seed component contents

Determination of lipid contents

Soxhlet extraction was adopted for extracting lipids from sunflower seeds. After methyl esterification of the samples (0.5 g), lipid analysis was performed on an Agilent 7890A gas chromatograph equipped with a 19091N-133 column (30 m × 0.25 mm × 0.25 μm; Hewlett-Packard, USA). The lipid contents were determined by comparison of peak areas against standard curves.

Determination of starch and soluble sugar contents

The sunflower seed samples (0.5 g) were ground, homogenized in 5 mL of ethanol, and centrifuged at 5000×g for 10 min after being placed in an 80 °C water bath for 30 min. The supernatants were collected for exploring the soluble sugar contents by employing anthrone method. The precipitates were added with 2 mL of distilled water, placed in a boiling water bath for 15 min, cooled, and subsequently added with 2 mL of perchloric acid and 4 mL of distilled water, followed by centrifugation at 4000×g for 10 min. The supernatants were collected for analyzing the starch contents by adopting spectrophotometry. The contents of starch and soluble sugar were measured by comparing the O.D. of the samples to the standard curve.

Determination of protein contents

The sunflower seed samples (0.5 g) were ground, homogenized in distilled water (5 mL), and centrifuged at 5000×g for 10 min, followed by collection of supernatants. The soluble protein contents in the supernatants were determined by spectrophotometry against BSA standard. Besides, the total protein contents of seeds were determined by Kjeldahl method. The determinations of seed component contents were performed with four biological replications.

Determination of antioxidant enzyme activities

By consulting the method proposed by Sheteiwy et al. (2017), the seed samples (0.5 g) were homogenized with 8 mL of potassium phosphate buffer (50 mmol L−1, pH 7.8), and centrifuged at 12,000×g and 4 °C for 20 min. Then, the supernatants were collected for determining the activities of superoxide dismutase (SOD), catalase (CAT), peroxidase (POD) and ascorbate peroxidase (APX) with the methods of Celus et al (2018). The determination of antioxidant enzyme activities was performed with four biological replications.

Determination the contents of GA, ABA and PAs

The contents of GA, ABA and PAs were determined by high performance liquid chromatography (HPLC). GA and ABA were extracted by referring to Huang et al. (2017). The extraction and derivatization of PAs, including putrescine (Put), spermidine (Spd) and spermine (Spm), were performed with the method of Flores and Galston (1982). Elution (10 μL) was performed through a C18 Shim-Pack CLC-ODS column (6.0 mm × 150 mm; 5 µm particle size) at room temperature via a HPLC pump (Shimadzu). Methanol solution (methanol/water = 64/36, v/v) was applied as the mobile phase, whose flow rate was 1.0 mL/min. The contents of GA, ABA and PAs were determined with the SPD-20A UV–Vis detector (Shimadzu) by comparison of peak areas against standard curves. The determination the contents of GA, ABA and PAs were performed with four biological replications.

Determination of phytohormone metabolism-related enzyme activities

ELISA was employed to determine the activities of relevant enzymes as per the instructions of kit (Mlbio, Shanghai, China). The chromogenic reaction was carried out, and the color change was measured spectrophotometrically at a wavelength of 450 nm with enzyme mark instrument. The activities of fatty acid and glycometabolism metabolism-related enzymes in seeds were then measured by comparing the O.D. of the samples to the standard curve. Four biological replicates were performed in the determination of phytohormone metabolism-related enzyme activities.

Quantitative real-time PCR (qRT-PCR)

Total RNA was extracted from each sample and first-strand cDAN was reverse transcribed using the Prime Script™ RT kit (Takara, Dalian, China). qRT-PCR was performed on a CFX96TM Real-Time Fluorescence Quantitative PCR System (Bio-Rad, Hercules, CA, USA). Sunflower Actin was used as an internal reference gene and relative gene expression was calculated using the 2−ΔΔCT method (Livak and Schmittgen 2001). Three technical replicates and three biological replicates per experiment were performed. The gene-specific primer sequences are listed in Supplementary Table 1.

Statistical analysis

SAS software was used to perform statistical analysis of the obtained data. Multiple comparisons were made by LSD (α = 0.05). At the same time, the percentage data were arcsine transformed (y = arcsin [sqrt (x/l00)]) prior to analysis.

Results

Vigor formation of sunflower seeds during development

Sunflower seed vigor increased gradually after pollination. At the early stage of development (DAP10 and DAP13), sunflower seeds fail to germinate. Then, four seed germination indexes (GE, GP, GI and VI) were increased gradually, peaking at day 25 or 28 after pollination. Afterwards, the germinability of sunflower seeds were significantly decreased at 31 DAP (day after pollination) (Table 1).

Table 1 The GE, GP, GI and VI of sunflower seeds at different maturation stages
Full size table

Changes in physical characteristics of sunflower seeds during development

To better understand the physical characteristics underlying the seed vigor formation, we measured the moisture content and dry weight of sunflower seeds during the development. The moisture content was 41.25% at 10 DAP, which then increased to 44.38% (13 DAP). Afterwards, the moisture content decreased continuously with the lowest value being 24.17% at 31 DAP (Table 2). Accordingly, the dry weight of seeds increased continuously as the development progressed, peaking at 28 DAP, and no longer decreased among 28–31 days after pollination.

Table 2 Differences of moisture content at different maturation stages of sunflower seeds
Full size table

Dynamic changes in nutrient accumulation during seed development

Given that the nutrients play important roles in seed germination processes, we further compared the accumulation patterns of different nutrients during the sunflower seed development. The results showed that seeds contained substantial soluble sugars at the early stage with a relative content of about 35%, but the content decreased gradually to remain at approximately 5% among 28–31 days after pollination. Meanwhile, the relative contents of lipids and total proteins in seeds presented upward trends. Among them, the lipid content accumulated significantly, reaching a peak of 37% at 28 DAP, and then decreased slightly while the total proteins was found to increase steadily from the initial 11% (10 DAP) to 25% (31 DAP). The relative soluble protein content increased slightly at 19 DAP and remained stable at nearly 9% for the rest of the time. Notably, the relative starch content of seeds increased slightly after pollination (Fig. 1).

Fig. 1
figure 1

Variation of the relative contents of lipid, total protein, soluble protein, soluble sugar and starch in sunflower seeds at different maturation stages

Full size image

Variations in antioxidant enzyme activities during seed development

We next assayed the variation trends of POD, CAT, SOD and APX at different maturation stages. The POD activity continued to increase from days 10 to 22 after pollination, and then decreased gradually (Fig. 2A). The activities of both CAT and APX decreased among 13–22 days after pollination, while the APX activity showed a cliff-like decline (Fig. 2B, D). However, the SOD activity exhibited a mild fluctuation during seed development, which was maintained at around 12 U/g FW among 10–31 days after pollination (Fig. 2C).

Fig. 2
figure 2

Variations in POD (A), CAT (B), SOD (C) and APX (D) activities at different maturation stages in sunflower seeds. POD peroxidase, CAT catalase, SOD superoxide dismutase, APX ascorbate peroxidase

Full size image

Changes in GA and ABA metabolism during seed development

Given the key roles of phytohormones ABA and GA in seed germination processes, we also compared the metabolic changes of these two hormones. As shown in Fig. 3A, the GA content decreased continuously from days 10 to 22 after pollination, and remained at low levels among 22–31 days after pollination. Both activities of GA3ox (gibberellin-3-β-dioxidase) and GA20ox (gibberellin-20-oxidase), which act as key enzymes in GA synthesis, decreased significantly during the seed development process (Fig. 3B, C). The activity of another enzyme GA2ox (gibberellin-2-β-dioxidase), which catalyzes the inactivation reaction of GA, continued to increase from 10 to 19 days after pollination, and decreased subsequently (Fig. 3D).

Fig. 3
figure 3

Changes of GA content (A) and activities of GA3ox (B), GA20ox (C), and GA2ox (D) at different maturation stages of sunflower seeds. GA gibberellin, GA3ox GA-3-oxidase, GA20ox GA-20-oxidase, GA2ox GA-20-oxidase. Different lowercase letters above the bars indicate significant differences between treatments at the 0.05 probability level

Full size image

RT-qPCR assays indicated that the expressions of GA metabolism-related genes were consistent with the enzymatic activities data during the sunflower seed development (Fig. 4). The transcription of GA synthesis genes HaGA3ox, HaGA20ox1 and HaGA20ox2 remained high levels from days 10–16 after pollination and then decreased. Differently, the expression of GA inactivation gene HaGA2ox2 was up-regulated gradually and peaked at 19 and 22 DAP. It was shown that HaGA2ox1 was only slightly up-regulated at 19 and 22 DAP, while HaGA2ox3 was expressed at low levels throughout the development of seeds (Fig. 4).

Fig. 4
figure 4

The levels of GA metabolism-related genes at different maturation stages of sunflower seeds. GA3ox GA-3-oxidase, GA20ox GA-20-oxidase, GA2ox GA-20-oxidase

Full size image

Notably, the content of ABA in sunflower seeds accumulated significantly from days 10 to 16 after pollination, and then decreased rapidly, reaching the minimum at 28 and 31 DAP (Fig. 5A). We next examined the activity of ABA synthetase NCED (9-cis-epoxycarotenoid dioxygenase) and found the same variation rule with the ABA content data (Fig. 5B). Furthermore, among 25–31 days after pollination, the NCED activity decreased to a extremely low level, which was only around 3.0 U/mg FW (Fig. 5B). There existed insignificant difference in AAO activity among 10–25 days after pollination (Fig. 5C). However, ZEP (zeaxanthin epoxidase) activity, another key enzyme in ABA anabolism, continued to decrease from days 10 to 31 after pollination (Fig. 5D). In contrast, the activity of ABA8ox (ABA 8′-hydroxylase), an ABA catabolic enzyme, increased continuously during late stage of seed development (DAP19–DAP31) (Fig. 5F).

Fig. 5
figure 5

The ABA content (A) and activities of NCED (B), AAO (C), ZEP (D) and ABA8ox (E) at different maturation stages of sunflower seeds. ABA abscisic acid, NCED 9-cis-epoxycarotenoid dioxygenase, AAO abscisic aldehyde oxidase, ZEP zeaxanthin epoxidase, ABA8ox ABA-8′-hydroxylases. Different lowercase letters above the bars indicate significant differences between treatments at the 0.05 probability level

Full size image

For genes expression, RT-qPCR analysis further indicated that, ABA anabolic genes HaNCED and HaZEP were highly expressed at the early stage (DAP10–DAP16), while the transcription levels of ABA catabolic genes ABA8ox1 and ABA8ox2 up-regulated as the seed development progressed (Fig. 6).

Fig. 6
figure 6

Changes of the transcriptional level of ABA metabolism-related genes at different maturation stages of sunflower seeds. NCED 9-cis-epoxycarotenoid dioxygenase, AAO abscisic aldehyde oxidase, ZEP zeaxanthin epoxidase, ABA8ox ABA-8′-hydroxylases

Full size image

Changes in PAs metabolism during seed development

As shown in Fig. 7, the variation trends of putrescine (Put), spermidine (Spd) and spermine (Spm) differed during development of sunflower seeds. The Put content fluctuated slightly during seed development (Fig. 7A). Spd accumulated gradually during the seed development, peaking among 25–28 days after pollination, with a content of 185 μg/g FW, which then decreased significantly (Fig. 7B). Similarly, Spm content increased gradually from days 10 to 22 after pollination, and decreased significantly among 25–31 days after pollination (Fig. 7C).

Fig. 7
figure 7

The Put (A), Spd (B) and Spm (C) contents at different maturation stages of sunflower seeds. Put putrescine, Spd spermidine, Spm spermine. Different lowercase letters above the bars indicate significant differences between treatments at the 0.05 probability level

Full size image

Arginine decarboxylase (ADC), ornithine decarboxylase (ODC), S-adenosylmethionine decarboxylase (SAMDC), spermidine synthase (SPDS) and spermine synthase (SPMS) are considered as key enzymes in the polyamine synthesis pathways. We found that ADC activity in sunflower seeds was enhanced significantly at 13 DAP, while decreased gradually among 13–31 days after pollination (Fig. 8A).The activities of SAMDC and SPDS showed a similar variation pattern of initial increase and subsequent decrease (Fig. 8C, D) while ODC and SPMS activities remained at a stable level throughout the development of sunflower seeds (Fig. 8B, F).

Fig. 8
figure 8

The activities of ADC (A), ODC (B), SAMDC (C), SPDS (D) and SPMS (E) at different maturation stages of sunflower seeds. ADC arginine decarboxylase, ODC ornithine decarboxylase, SAMDC S-adenosylmethionine decarboxylase, SPDS spermidine synthase, SPMS spermine synthase

Full size image

The gene expression analysis revealed that HaADC, HaSAMDC2, HaSPDS and HaSPMS all tended to increase initially and then decrease during the development of sunflower seeds. Among these genes, HaSPDS was highly expressed among 16–22 days after pollination. The expression level of HaSAMDC1 remained extremely low after pollination, indicating its nonparticipation in sunflower seed development (Fig. 9).

Fig. 9
figure 9

The transcriptional level of polyamines metabolism-related genes at different maturation stages of sunflower seeds. ADC arginine decarboxylase, ODC ornithine decarboxylase, SAMDC S-adenosylmethionine decarboxylase, SPDS spermidine synthase, SPMS spermine synthase

Full size image

Correlation analysis between physiological/biochemical indices and seed vigor

The seed vigor was closely associated with the physiological and biochemical indices during the development of sunflower seeds. As shown in Table 3, the dry weight and moisture content of seeds exhibited extremely significant positive and significant negative correlations, respectively, with the seed vigor-related indices. During the seed development, the ABA content was extremely significantly negatively correlated with the GE, GP, GI and VI of seeds (r = − 0.943, − 0.904, − 0.899 and − 0.906, respectively), while the GA content was extremely significantly negatively correlated only with the seed GE (r = − 0.902). In terms of a variety of components, the lipid content was extremely significantly positively correlated with the seed GE (r = 0.903). The total protein content was extremely significantly positively correlated with the GE, GP and VI of seeds (r = 0.921, 0.889 and 0.883, respectively). Besides, the soluble sugar content was extremely significantly negatively correlated with the GE and VI (r = − 0.918 and − 0.879, respectively). Moreover, the contents of three components were significantly correlated with all of the rest vigor indices. Obviously, the Spd content was extremely significantly positively correlated with all of the GE, GP, GI and VI (r = 0.937, 0.921, 0.942 and 0.932, respectively). Thus, the above indices can be adopted for the evaluation of sunflower seed vigor.

Table 3 The correlation between seed physiological parameters and seed vigor of sunflower during seed development
Full size table

Discussion

As an important indicator of seed quality and seed value, seed vigor is closely associated with the field seed emergence (Huang et al. 2017). More importantly, seed vigor represents the quality of seedlings in the field, especially under unfavorable environmental conditions, which can better reflect the seed quality than the GP. With the development of seeds, the seed vigor begins to form gradually reaching the highest when the seeds are physiologically mature. The physiological state of seeds during development directly affects their vigor. Generally, the seed vigor is comprehensively measured by GE, GI, VI, seed size, dry/fresh weight ratio, as well as some vigor-related physiological and biochemical indices (Hu 2014). However, studies on the molecular mechanism of vigor formation during the seed development remain scarce, most of which concentrated on model crops rice and Arabidopsis, with little attention to sunflower.

In the current experiments, the moisture content of sunflower seeds decreased significantly at 16 DAP, which tended to decrease overall in the late period. At the same time, the dry weight and volume of seeds continued to increase, which peaked at 28 DAP, and did not decrease afterwards, suggesting that the nutrient storage reached a steady state upon maturation of seeds (Table 2). During seed development, the changes in soluble sugars, lipids and total proteins were obvious. Initially, the soluble sugar content was high, which could satisfy the high energy demand of seeds at the early development stage. Thereafter, the content decreased gradually and eventually stabilized. Lipids and proteins were primarily accumulated at the late stage of seed maturity (Fig. 1). In the current work, the GP, GE, GI and VI of seeds reached optimal values among 25–28 days after sunflower pollination, which was the best harvesting time since the seed vigor was the highest. Thereafter, the seed vigor decreased evidently (Table 1).

Development and maturation of seeds respond to endogenous and exogenous signals such as trophic state, energy and phytohormones. The phytohormones playing important roles in crop seed development include GA, ABA, auxin and PAs (Li et al. 2017). The GA metabolism is closely associated with seed development and germination. Inconsistent gibberellin metabolic responses were observed during seed development and seed germination process. The increase in GA level was crucial for normal seed germination during the early imbibition process (Liu et al. 2016). However, the GA content during seed development shows a gradual downward trend, while ABA gradually accumulates in the later stage of seed development (Zhu et al. 2016). Consistently, we found that the GA content in sunflower seeds decreased gradually among 10–22 days after pollination, presenting a significant negative correlation with the seed vigor. As shown in Fig. 3, the decrease in GA content was attributed to reduced activities of GA synthetases (GA3ox and GA20ox) and an elevated activity of GA catabolic enzyme (GA2ox). In addition, the corresponding genes expressions were also consistent with the changes in GA content and metabolic enzyme activities (Fig. 4). Generally, ABA content of seeds is extremely low during early embryogenesis, which increases significantly during embryogenesis transition to maturity stage. And it decreases drastically during later period of seed development, especially at maturity and drying stages (Cheng et al. 2014). Xiong and Zhu (2003) and Lefebvre et al. (2006) found that ABA anabolic genes (NCED6, NCED9) were highly expressed at the middle stage of Arabidopsis seed development. During sunflower seed development herein, the ABA accumulated abundantly among 10–16 days after pollination and then decreased significantly (Fig. 5A). The activities of NCED and ZEP were significantly decreased at the middle and late stages of seed development, while the activity of ABA8ox was enhanced significantly at the late development stage. It suggests that the de novo pathway of ABA may be responsible for the surge of ABA content at the middle stage of seed development, while the declined ABA content at the late stage is generated by the surge of ABA8ox activity. In addition, this conclusion was also verified by the RT-qPCR results. Additionally, correlation analysis revealed the presence of a very significant negative correlation between ABA content and seed vigor. At the late stage of seed maturation, ABA may exert an important role in dehydration tolerance and post-ripening of seeds (Tuan et al. 2021).

Shiozaki et al. (2000) found that the content of PAs in seeded grapes increased on days 20–30 after flowering, but not in seedless grapes. Free PAs may stimulate the development of grape seeds by facilitating the cellular differentiation of outer integument and endosperm. Endogenous PAs in abortive seeds of maize (Zea mays L.) hybrids were significantly lower than those in normal seeds at 4 DAP, hinting at a probable connection between PAs and early endosperm development (Liang and Lur 2002). Uranoa et al. (2005) concluded that AtADC, the PA synthesis gene, is essential for the normal formation of Arabidopsis thaliana seeds. Apart from that, a study by Chen et al. (2000) showed that the Spd and Spm contents of lychee are closely associated with the embryo abortion. The Spd content in the seeds of Arabidopsis spds1 and spds2 double mutants decreased significantly, and the embryo morphogenesis stopped at the torpedo stage (Imai et al. 2004). In the current work, the Spd content showed a significant positive correlation with seed vigor based on the correlation coefficient above 0.9. Meanwhile, the Put and Spm contents were insignificantly correlated with the seed vigor compared with other crop seeds (Table 3).

Reactive oxygen species (ROS) are produced during the development of seeds. Adequate concentrations of ROS play a role in seed development, dormancy breakage and resistance to environmental stress, while excessive ROS causes toxicity to cells and plant tissues. However, the correlation between antioxidant enzymes and seed vigor formation during seed development remains unclear. In sunflower seeds, the activities of SOD, POD and CAT show insignificant correlations with the seed vigor indices (Table 3). APX retains a high activity at the early stage of seed development, whose activity decreases rapidly at the middle stage, and remains a low level at the late stage (Fig. 2). Consistent with the obtained result, Zhu et al. (2016) found that APX activity decreased gradually during the development of hybrid rice seeds, and its correlation coefficients with GE and GP also reached − 0.857 and − 0.850, respectively. We propose the close correlation between APX and sunflower seed vigor formation (Table 3) while the unknown regulatory mechanism requires further investigation.

Conclusion

Sunflower seeds have the highest vigor among 25–28 days after pollination, which is the best harvesting period. Variables significantly negatively correlated with seed vigor contain the moisture content, APX activity, as well as GA and ABA contents. Dry weight and contents of lipids, total protein and Spd are significantly positively correlated with seed vigor, which can serve as key predictors for seed vigor. Our study reveals a valuable foundation for the sunflower seed vigor formation, and provide an important theoretical basis for the quality and safety control of seed production and multiplication.

Author contribution statement

Shuyu Cai designed and performed most of the experiments. Liyuan Wu and Guofu Wang performed the experiments of qRT-PCR analyses. Jianxin Liu, Hua Xu and Jiangping Song contributed to the enzymes and HPLC analysis. Shuyu Cai reviewed and edited the whole manuscript. All authors read and approved the final version of the manuscript.