Cinnamomum migao fruits were collected from Luodian County, Guizhou Province, China (25°26′40″N, 106°31′51″; 667 m altitude) in October–November 2017. The ripened fruits were transported to the laboratory immediately after collection. The fruits were then peeled, washed under running water, and placed in a ventilated dark room. The seeds enclosed by the endocarp were stored in the dark at 20 ± 2°C and approximately 60% relative humidity for 1 week before use.

The transverse axis diameter, longitudinal axis diameter, and thickness of C. migao seeds were measured using a vernier caliper (0.01 mm). The caliper was slowly clamped along the seed ridge to separate the embryo from the endocarp, and embryo development was observed and photographed under a fluorescence microscope (Leica M205 FA). Three replicates of 30 seeds were measured.

Gas chromatography (GC)-mass spectrometry (MS) was performed using HP6890/5975c GC–MS (Agilent, United States). For sample determination, 2 μl of the extract [ethyl acetate and methanol extracts of the endocarp, 1 g tissue (g ml⁻¹)] was injected with a microsyringe. The chromatographic conditions were as follows: capillary column, hp-5 ms (60 m × 0.25 mm × 0.25 μm); initial temperature, 70°C (hold for 2 min), which was increased to 270°C at a rate of 5°C min⁻¹ and then to 310°C at a rate of 8°C min⁻¹ (hold for 9 min) for a total operation time of 56 min; vaporization chamber temperature, 250°C; carrier gas, high purity He (99.999%); pressure, 7.65 psi; carrier gas flow rate, 1.0 ml min⁻¹; split ratio, 10:1; and solvent delay time, 6.0 min. The mass spectrum conditions were as follows: EI ion source with 230°C ion source temperature, 150°C quadrupole temperature, 70-EV electron energy, 34.6 μA emission current, 1,894 v multiplier voltage, 280°C interface temperature, and 29–450 amu mass range (Chen et al., 2010). Peaks in the total ion flow diagram were retrieved via the MS computer data system and were compared with the National Institute of Standards and Technology 2014 and Wiley 275 standard mass spectra to identify the volatile chemical components. The relative mass fraction of each chemical component was determined using the peak area normalization method.

Final germination percentage, germination energy, and germination index (GI) were calculated using the following formulas:

where Gt is the number of germinated seeds per day and Dt is that of germination time (d).

The SPSS 21.0 statistical software package (IBM, Chicago, IL, United States) was used for data processing. One-way ANOVA and Duncan’s multiple comparison analysis were used to compare the treatment outcomes. All data were expressed as means ± standard deviations (SD). The data were inverse sine-transformed using the asin() function; data were expressed as means ± SDs and were analyzed using the SPSS 17.0 software (IBM, United States). Statistical differences among the treatments were analyzed using one-way ANOVA followed by Duncan’s post-hoc test. Statistical significance was set at p < 0.05. Origin 2019b (Origin Lab, Northampton, Ma, United States) and Adobe Illustrator 17.0 (California, ADBE, United States) software were used for drawing and processing.

The mature C. migao fruits (Figure 1A) were baccate, 1.2–1.3 cm in diameter, and spherical with a slightly flat top (Figure 1B); these grew on the top of a goblet-shaped receptacle (Supplementary Figures S1A–D). The fresh seeds (encapsulated by the endocarp) were black and had a unique aroma. The seeds dried in the dark were brown and pea-shaped (Figure 1C).

The transverse and longitudinal axis diameters of the seeds were 10.73 ± 0.19 mm and 9.73 ± 0.09 mm, respectively, and the seed thickness was 9.31 ± 0.20 mm. The thousand-grain weight of the dry seeds was 421.34 ± 5.32 g, moisture content was 12.19 ± 1.21%, and viability was 98.67 ± 1.15%.

The endocarp was ossified and closely adhered to the testa, resulting in the separation of the testa from the embryo (Figure 2A). Inside the endocarp were two thick cotyledons (Figure 2B). The radicle, hypocotyl, and germ were small and cone-shaped, connecting the two cotyledons at the apex (Figure 2C). The obvious differentiation of the radicle, hypocotyl, and each part of the embryo indicates that these parts develop completely as the seeds mature (Figure 2D). The endocarp had a layer of stone cells (Figures 2E–H).

The imbibition data of the C. migao seeds with the endocarp, with the partial endocarp, and without the endocarp are shown in Supplementary Figure S3. The seed weigh rapidly increased in the first 6 h. After 2 h of imbibition, the mass increases were 10.51, 12.38, and 21.65% for the seeds with the endocarp, with the partial endocarp, and without the endocarp, respectively; after 6 h, these were 21.47, 23.58, and 32.29%, respectively (Figure 3). Mass increase began to plateau after 12 h. The seeds with the endocarp, with the partial endocarp, and without the endocarp attained saturation after 72, 60, and 48 h of imbibition, respectively. Final mass increases were 40.64, 41.21, and 41.76%, respectively, with no significant differences across treatments (p > 0.05; Figure 3).

To further investigate how the endocarp influences germination, we treated the seeds with NaOH, H₂SO₄, and H₂O₂ before the germination test. Owing to corrosion or oxidation by chemical solvents, the endocarp thinned or the outer structure carbonized, blackened, and peeled off. The final germination percentage of seeds pretreated with 20% H₂O₂ was the highest at 34% (Figure 4A), which was significantly greater than that of the unstratified seeds (p < 0.05). The final germination percentages did not differ significantly (p > 0.05) in seeds pretreated with NaOH (Figure 4B). In seeds pretreated with H₂SO₄, the final germination percentage was 31.33% after pretreatment with 50% H₂SO₄ (Figure 4C), which was significantly higher than that of the unstratified seeds (p < 0.05). However, pretreatment with 98% H₂SO₄ significantly (p < 0.05) reduced the final germination percentage of the seeds.

Cinnamomum migao endocarp and embryo extracts inhibited the germination of Chinese cabbage seeds (Supplementary Figure S3). Ethyl acetate endocarp extract, when applied at concentrations of 5, 10, 15, and 20%, significantly reduced the seed germination percentage of Chinese cabbage by 8.16, 17.69, 22.45, and 33.33%, respectively, compared with the control (Figure 5A). At the 5% concentration, acetone, methanol, and water extracts significantly (p < 0.05) inhibited the germination of cabbage seeds (Figure 5B). The methanol extract had the strongest inhibitory effect; concentrations of 5, 10, 15, and 20% reduced the final germination percentages of cabbage seeds by 10.20, 21.09, 29.93, and 44.90%, respectively, compared with the control (Figure 5B). Ryegrass seeds were treated with the C. migao extract showing the strongest inhibitory effects (endocarp: ethyl acetate extract; embryo: methanol extract) to confirm the inhibitory effects. The ryegrass seeds were significantly inhibited as the extract concentration increased (Figure 5C), and even the lowest concentration (5%) significantly reduced the final germination percentages. In addition, the growth of the cabbage and ryegrass seedlings was influenced by the C. migao seed extracts (Supplementary Tables S1–S3), with the ethyl acetate and methanol extracts of the endocarp and embryo, respectively, displaying the strongest inhibitory effects on the growth of cabbage seedlings.

The ethyl acetate and methanol extracts of C. migao seed endocarps were analyzed using GC-MS (Supplementary Figure S4). We identified 58 endogenous chemical compounds, including 25 in the ethyl acetate extract and 33 in the methanol extract. The primary components were as follows: 4 phenolic compounds, 11 aldehydes, 2 ethers, 8 ketones, 6 alcohols, 9 fatty acids, 11 lipids, 1 benzene, 1 pyrazine, 1 anhydride, and 1 alkene (Figure 6).

The temperature and light treatments influenced the initial seed germination and resulted in significant differences (p < 0.05) in the final germination percentage, germination vigor, and GI of C. migao seeds (Figure 7A). Seeds incubated at 30°C did not germinate until day 46, and the final germination percentage was the lowest at 4.0% (Figure 7B). ANOVA results (Supplementary Table S5) demonstrated that light (F = 0.706, p = 0.408) or the interaction between light and temperature (F = 0.157, p = 0.986) had no significant influence on germination. Nonetheless, the effect of temperature on the final germination percentage was significant (F = 55.33, p < 0.001). Thus, temperature is the primary factor in seed germination.

Pretreatment with GA₃—a dormancy-releasing plant hormone—can effectively reduce the initial germination time of C. migao seeds. The final germination percentage, germination vigor, and GI improved with an increase in the GA₃ concentration, reaching a peak at 200 mg L⁻¹ and then decreasing (Figure 8). After the treatment with 200 mg L⁻¹ GA₃, the initial germination time was reduced to 19 days, which was 5 days less than that of the control seeds. The final seed germination percentage, germination vigor, and GI were 44.67 ± 3.06%, 23.33 ± 1.15%, and 0.64 ± 0.05 (representing growths of 91.47, 84.14, and 113.33%), respectively, compared with the control group (p < 0.05).

The results of the stratification experiment revealed that the final germination percentage of C. migao seeds was higher under the cold/warm stratification + GA₃ treatment than that under the stratification treatment alone (Figure 9). In addition, the germination percentage was higher in the warm-temperature stratification treatment (15°C) than that in the cold-temperature treatment (4°C). The germination ability of the seeds treated at 15 and 4°C with GA₃ peaked at 60 (13 days for initial germination) and 90 (12 days for initial germination) days of stratification, respectively, which was significantly less than that of the unstratified seeds (Figure 9A). The final germination percentage, germination vigor, and GI were significantly higher (p < 0.05) in the 15°C + GA₃- and 4°C + GA₃-treatment groups than in the unstratified seeds by 73.11, 89.96, and 128.13% as well as 57.12, 82.90, and 123.44%, respectively (Figures 9B–D). The germination ability of seeds treated at 15 and 4°C peaked after 120 and 150 days of stratification, respectively. The initial germination time in both treatments was 15 days, which was 9 days less than that of the unstratified seeds. When 15°C + GA3 and 4°C + GA3 treatment compared with treatment in the absence of GA₃, the final germination percentage, germination vigor, and GI of the unstratified seeds increased significantly by 188.60, 216.67, and 333.33% and 174.32, 194.63, and 306.67%, respectively (p < 0.05). After reaching the peak, the germination of the seeds under different treatments exhibited a downward trend with prolonged stratification time. The final germination percentage and vigor of seeds in 120–180 days declined in a nonsignificant manner (p > 0.05). In the stratification of 180–240 days, the seeds in each treatment group showed different degrees of reduction in the germination. Seeds in the 15°C + GA₃ and 4°C + GA₃ treatment groups had significantly reduced final germination percentages than those in the same temperature treatment groups but without GA₃. Overall, stratification can effectively increase the final germination percentage of C. migao seeds, and GA₃ treatment can significantly shorten the stratification period and can thus reduce the initial germination time. The warm-temperature (15°C) treatment was more effective in releasing the dormancy of the seeds than the low-temperature (4°C) treatment, and adding GA₃ (15°C + GA₃) to the warm-temperature stratification treatment further increased the final germination percentage.

The complex and diverse factors responsible for seed dormancy are associated with changes in the pericarp, seed coat, and embryo (Meyer, 2008). Therefore, we systematically investigated the possible causes of C. migao seed dormancy from the aspects of morphology, anatomy, and physiology.

MD refers to the dormancy due to the small embryo size, incomplete development, or lack of differentiation after seed maturation. The morphology of the seed embryo is generally spatulate, rudimentary, or linear; in some species, it occupies ≤1% of the seed volume (Jaganathan, 2020; Liu et al., 2020). For example, the Lamprocapnos spectabilis seed endosperm comprises the entire internal tissue of the seed, and the embryo is almost invisible (Cho et al., 2020). Similarly, the Acer yangjuechi seed embryo exhibits poor development and low vigor, which lead to a germination failure under suitable conditions (Chen et al., 2017). However, our results show that the mature embryos of C. migao seeds are fully developed, indicating the absence of MD (Figure 2; Supplementary Figure S1).

The endocarp of C. migao is water-permeable, and thus they do not have PY (Figure 2). Furthermore, no significant difference in the final percentages of imbibition was detected among the seeds with the endocarp, with the partial endocarp, and without the endocarp (Figure 3). The imbibition of C. migao seeds was similar to that of Phoebe bournei seeds belonging to the same family, in which the pericarp does not hinder imbibition (Li and Min, 2020). C. migao seeds with the partial endocarp absorbed water faster than those with the intact endocarp; seeds without the endocarp absorbed water even faster than those with the partial endocarp. This finding is similar the results obtained for the endocarp of Prunus mahaleb (Murphy et al., 2010) and Phillyrea angustifolia (Mira et al., 2015). However, in the present study, the intact seeds were capable of absorbing the same amount of water as that of the seeds without the endocarp, although the imbibition was slower. Hence, the endocarp was not impermeable, which eliminated the PY of C. migao (Figure 3).

To assess PD, we examined whether the C. migao endocarp and embryo extracts contain compounds that inhibit the germination of cabbage and ryegrass seeds. Specifically, the ethyl acetate extract of the endocarp and the methanol extract of the embryo showed significant inhibitory effects. The methanol extract completely inhibited ryegrass germination at concentrations of >0.05 g ml⁻¹ (Figure 5B). The primary components of the extracts with the strongest inhibitory effects were lipids, fatty acids, ketones, aldehydes, and phenolic compounds. Previous studies have demonstrated that inhibitory compounds in the endocarp or the hard middle layer of the testa contribute to low and unstable seed germination (Hamilton and Carpenter, 1976; Richmond and Ghisalberti, 1994; Wu and Shen, 2021). Phenolic and lipid compounds present in the seeds have been widely shown to induce seed dormancy and prevent germination (Graeber et al., 2012; Inácio et al., 2013; Smýkal et al., 2014). Inhibitory compounds present in the seeds can cause physiological inhibition by hindering the growth potential or breakthrough ability of the embryo; as a result, it fails to break the mechanical resistance of the endocarp, thereby negatively influencing seed germination (Baskin and Baskin, 2014). Similarly, previous studies have reported that phenolic compounds present in Lycopersicon esculentum (tomato), Fagus sylvatica, and Pinus laricio in the seeds inhibit the metabolism of key enzymes involved in seed germination (Muscolo et al., 2001; El-Araby et al., 2006). Therefore, according to the classification system of Baskin and Baskin (2004), we determined that C. migao seed dormancy is physiological in nature and is due to endogenous inhibitors and low growth potential of the embryo.

For seeds with obvious dormancy, its release involves changes in the external environment and the physiology and biochemistry of the seed. The release of seed dormancy primarily occurs through the regulation of light, temperature, hormones, and stratification (Cosmas et al., 2019; Hou et al., 2019; Yang et al., 2019). Therefore, we evaluated the effects of light, temperature, an exogenous hormone, and stratification on C. migao seed dormancy and germination.

No significant differences in dormancy release were observed in the seeds under the same temperature but different light treatments. Light neither reduced the initial germination time nor influence the final germination percentage, germination vigor, or GI (Supplementary Table S5). Light-requiring seeds are generally small (Koutsovoulou et al., 2014; Yang et al., 2020), which allows the external light to be perceived by the seed pigments, integrating the metabolism and signal transduction of abscisic acid and GA to control seed dormancy and germination (Bentsink and Koornneef, 2008; Shu et al., 2016). However, C. migao seeds are relatively large and are encapsulated by the firm endocarp (Figures 1, 2), which prevents the internal pigments from perceiving external light (Liu and Lin, 1995).

The low constant temperature and alternating temperatures reduced the initial germination time and significantly increased the final germination percentage, germination vigor, and GI, with the treatment effect being the strongest at 10/20°C. Compared with the interaction of light and temperature, temperature alone significantly promoted the release of dormancy (Figure 7). The natural distribution of C. migao is limited to dry and hot valleys, and the 10/20°C alternating temperature condition is similar to the habitat temperature in spring (Tian et al., 2015). The changing temperature stimulates enzymatic activities in these seeds, which in turn promotes the transformation of the stored materials and releases the dormancy. The 10/20°C alternating temperature treatment may be in line with the natural temperature change in the environment (Chahtane et al., 2016; Kildisheva et al., 2019).

The role of the exogenous hormone GA₃ in releasing seed dormancy and increasing the final germination percentage has been extensively studied (Linkies and Leubner-Metzger, 2012). GA₃ treatment has been demonstrated to effectively enhance the final germination percentage of Buglossoides arvensis and Cyclocarya paliurus seeds (Fang et al., 2006; de Las Mercedes et al., 2021). GA₃ improves α-amylase hydrolysis and glyoxylate cycle enzymatic activities in seeds, thereby activating energy metabolism and enhancing the growth potential of the embryo (Song et al., 2019). In addition, the cold/warm stratification treatment has been shown to release PY (Wolkis et al., 2020). Similar to physical polishing, stratification increases the growth potential of the embryo, thus allowing it to overcome the mechanical resistance to its growth (Baskin and Baskin, 2014). Our results showed that seed treatment with 200 mg L⁻¹ pure exogenous GA₃ significantly reduces the initial germination time and the final germination percentage (Figure 8). The 15°C stratification was significantly better than the 4°C stratification, and stratification after GA₃ pretreatment resulted in earlier germination peaks than those observed with stratification alone. The combination of 15°C + GA₃ was particularly effective, inducing initial germination 6 days earlier than the control seeds. The final germination percentage was as high as 77.33 ± 4.16%, which is significantly higher than those noted in the alternating temperature, light, single hormone, and single stratification treatments (Figure 9). Therefore, combining GA₃ pretreatment with stratification was most effective in the rapid release of C. migao seed dormancy (Figure 9B). The hormone treatment activates and accelerates the metabolic activities of the seeds, promoting cell division and radicle growth (Yang et al., 2019), and the stratification treatment influences the hormone contents and metabolic processes. The combination of exogenous hormones and stratification treatment thus accelerated the physiological and biochemical processes.

Temperature and cold stratification treatment were effective in releasing the dormancy of C. migao seeds (Figure 9); however, light and dark treatments had no significant effect on dormancy release (Figure 7). The habitats of C. migao—the forests—are heavily impacted by human activities, including fruit harvesting as well as understorey vegetation and leaf litter removal. Therefore, although a large number of fruits fall on the ground after ripening, these are not covered by litter (Zhou, 2003). Owing to the relatively dry surface soil conditions in winter, C. migao seeds may be unable to absorb sufficient water for effective alternating temperature stratification. This phenomenon has been observed in several species. Exposure to the soil surface increases the risk of animal (e.g., rodents) foraging in winter (Gao et al., 2021). In fact, we noted several C. migao seeds that were apparently chewed by rodents. Therefore, the lack of an effective stratification treatment because of human disturbance may significantly hinder the release of dormancy and lead to an extreme shortage of C. migao seedlings. The results provide important insights into the successful regeneration of the C. migao population. For example, covering the seeds under the litter to ensure effective stratification in winter will facilitate seed germination. On the other hand, collecting seeds to artificially release seed dormancy can prevent animal foraging (Chen et al., 2020). Furthermore, the reintroduction of C. migao seedlings into wild population should be considered an effective method to increase the population of this arbor (Wright et al., 2009).