Introduction
Since Crocker’s (Reference Crocker1916) first detailed work on seed dormancy, thousands of research articles and hundreds of books have been published on this topic, which increased our understanding significantly in several areas. Now, we have a more acceptable definition of dormancy. Baskin and Baskin (Reference Baskin and Baskin2004, Reference Baskin and Baskin2014, Reference Baskin and Baskin2021) define “a dormant seed (or other germination unit) is one that does not have the capacity to germinate in a specified period of time (30 days) under any combination of normal physical environmental factors (temperature, light/dark, etc.) that otherwise is favourable for its germination, i.e. after the seed becomes non-dormant.” Likewise, despite several attempts that had been made to classify seed dormancy in the past (Bewley et al., Reference Bewley, Bradford, Hilhorst and Nonogaki2013; Harper, Reference Harper1977; Vleeshouwers et al, Reference Vleeshouwers, Bouwmeester and Karssen1995), the use of Russian seed physiologist Marianna G. Nikolaeva’s (Reference Nikolaeva1969) classification system modified by Baskin and Baskin (Baskin and Baskin, Reference Baskin and Baskin2004, Reference Baskin and Baskin2021) has become the most useful system because it not only refers to what causes dormancy (hampers germination) but also what conditions have to be met in the field before dormancy is broken.
The five classes proposed by Baskin and Baskin (Reference Baskin and Baskin2004, Reference Baskin and Baskin2014, Reference Baskin and Baskin2021) include: (i) physiological dormancy (PD), in which embryos lack the potential to elongate due to either innate constraints or thick water-permeable seed coat/endocarps preventing elongation; (ii) morphological dormancy (MD), where seeds are dispersed with underdeveloped embryos requiring time to mature and develop, despite having distinct cotyledons and hypocotyl-radicles; (iii) morphophysiological dormancy (MPD), characterized by seeds with underdeveloped embryos that also possess PD, necessitating specific dormancy-breaking conditions to break both MD and PD before germination; (iv) physical dormancy (PY), where the palisade layer of lignified Malpighian cells in the seed/fruit coat becomes impermeable during maturation drying, preventing water uptake and hydration; and (v) combinational dormancy (PY + PD), where seeds exhibit both PY and PD, requiring conditions that break bothdormancy classes before germination can occur. However, not all the spermatophytes produce dormant seeds, particularly those inhabiting tropical-wet and aquatic environments, produce non-dormant (ND) seeds that germinate almost immediately after dispersal when they receive favourable conditions for seed germination (Baskin and Baskin, Reference Baskin and Baskin2014, Reference Baskin and Baskin2021). Furthermore, some species in extreme environments such as mangroves, produce seeds that germinate prematurely while they are still attached to the parent plant or within the fruit, even before dispersal (Farnsworth, Reference Farnsworth2000). This phenomenon involves the continuous growth of the embryo without entering a dormant state. These seeds are known as viviparous or crypto-viviparous (Wijayasinghe et al, Reference Wijayasinghe, Jayasuriya, Gunatilleke, Gunatilleke and Walck2019).
The transition from seed to seedling, with a specific focus on the epicotyl, has been a subject of longstanding interest in plant biology (e.g., Engelmann, Reference Engelmann1880). Although the phenomenon of epicotyl dormancy, characterized by a temporal separation between radicle emergence and subsequent shoot development, has been recognized for some time, this particular form of dormancy has received comparatively limited attention across a diverse range of species. In the recently updated dormancy classification system, Baskin and Baskin (Reference Baskin and Baskin2021) included epicotyl dormancy under PD subclass 2 and recognized two levels: Level 1. Nondeep, in which the seeds are non-dormant, i.e., radicle emerges quickly, but the shoot emergence is delayed; and Level 2. Deep, which contains seeds that show both root emergence delay and delay for shoot emergence after root emerged. In addition, the palm seeds with MDare included under Class 2. Morphological [B] subclass 6. Specialized E (palms) also have known cases of epicotyl dormancy (the alphabets and subclass numbers refer to the dormancy classes proposed by Baskin and Baskin, Reference Baskin and Baskin2021). Further, MPDalso includes three levels, namely nondeep epicotyl, deep epicotyl, and double dormancy. Epicotyl dormancy can be present independently in species otherwise characterized by ND or occur in species with other established dormancy classes. Regardless of this, the defining feature of epicotyl dormancy is the delayed development of the shoot system, frequently necessitating specific environmental cues or treatments to facilitate shoot establishment. This intricate relationship underscores the need for more rigorous investigations into the physiological and ecological understanding of epicotyl dormancy.
Most seed researchers terminate their germination experiments abruptly after the first visible signs of germination, which is radicle emergence to 2 mm, based either on the International Seed Testing Association recommendation (ISTA, 2024) or a general understanding that a germinated seed is no longer part of “seed research.” A major problem with this approach is that studies concluding the dispersal units are ND underestimate the presence of dormancy, particularly epicotyl dormancy. Obviously, this oversight needs to be addressed for various reasons. Firstly, understanding epicotyl dormancy informs us of the adaptive strategy used by some species to survive adverse environmental conditions (Jaganathan, Reference Jaganathan2021; Jaganathan and Phartyal, Reference Jaganathan and Phartyal2024). Desiccation sensitive (DS) seeds avoid desiccation stress in the seed state, but the growing radicle can penetrate deep into the soil, which retains moisture from rainfall for a long time. Secondly, epicotyl dormancy might be a risk-minimizing strategy to reduce competition during the growing season. A delayed emergence in the shoot may allow the growing seedlings to absorb more nutrients when all other species have already started photosynthesis. Thirdly, our understanding of the distribution of dormant seeds in several habitats often ignores epicotyl dormancy, and a reassessment might offer a varied perspective on the ecological significance of dormancy in relation to climate. For example, in DS species, epicotyl dormancy may synchronize shoot emergence with the wet season. Finally, the hormonal pathways, mainly, gibberellins (GAs) and abscisic acid (ABA) in ND and comparatively in dormant seeds, can be better understood. This is especially crucial when epicotyl dormancy occurs with other forms of dormancy, and our understanding of the biochemical and molecular mechanisms involved is not complete (Chen et al, Reference Chen, Liu, Wang, Xu, Li, Pei, Zhang, Li and Yang2024).
The traditional definition of germination, focusing solely on radicle emergence, is insufficient to capture the complexities of epicotyl dormancy. This limitation is particularly significant as epicotyl dormancy is a key component of various dormancy classes including ND, MD, PD, and MPD. Thus, by relying solely on radicle emergence, researchers may overlook or misclassify important dormancy mechanisms, potentially leading to an incomplete understanding of seed biology and ecology. To address this problem in all species, including agricultural and wild species, this work proposes that continuing seed germination studies until the first leaf emergence is critical for understanding dormancy complexity and capturing novel insights into evolutionary ecology and trait coexistence. We call this as “full germination” period, i.e., from sowing to first leaf emergence. To this end, we have developed a detailed methodology from the onset of germination experiments (Fig. 1). Our method involves germinating fully matured seeds (immediately after collection) across a wide range of temperature and light conditions, separating them into three distinct groups based on germination timing: (1) Viviparous seeds, which by definition exhibit visible germination at the time of dispersal, as evidenced by radicle emergence; (2) ND or MD seeds that achieve more than 50% germination under any single combination of temperature and light within 30 days; and (3) PD or MPD seeds that fail to achieve 50% germination under any combination of temperature and light within 30 days. More importantly, our goal here is not to attempt an exhaustive review of epicotyl dormancy, but to show as many examples as possible that monitoring seed germination until the emergence of first leaves is essential.
Figure 1. A decision tree for classifying seed dormancy types based on radicle emergence and embryo development patterns. Seeds are initially sorted into three groups according to the time required for germination, 1) radicle emerged on the mother plant (Viviparous); 2) radicle emerged within 30 days at any temperature or light conditions (Non-dormant, ND or Morphological dormant, MD); and 3) radicle did not emerge within 30 days at any temperature or light conditions (Physiological dormant, PD or morphophysiological dormant, MPD). A cut test is then used to distinguish PD and MPD seeds. Depending on the interval between radicle emergence and shoot emergence, seeds are redefined as truly ND or as having epicotyl dormancy. If epicotyl dormancy is observed, seeds are assigned to the appropriate class, subclass, and levels. In PD and MPD seeds, if shoot emergence is delayed the growing seedlings should be monitored until first leaves and assigned to the appropriate class, subclass and levels (also see Baskin and Baskin, Reference Baskin and Baskin2021).
Group 1- viviparous/ cryptoviviparous seeds
The viviparous seeds with emerged radicles must be monitored for germination across a wide range of temperature and light conditions. If seeds are truly non-dormant, shoot emergence occurs within 30 days. If shoot emergence is delayed beyond 30 days, the seeds exhibit PD-subclass 2: epicotyl dormancy. Further, because the radicle already emerged indicating the absence of PD, these species have nondeep PD with subclass epicotyl dormancy. While this type of dormancy remains less explored in the literature, some true mangrove species have viviparous seeds where the hypocotyl elongates and emerges, piercing both the seed and fruit coats before the dispersal. However, the epicotyl emerges once the seed is firmly established on a stable shoreline. During the dispersal, shoot stays dormant. Wijayasinghe et al (Reference Wijayasinghe, Jayasuriya, Gunatilleke, Gunatilleke and Walck2023) unequivocally reported a significant delay in shoot emergence following radicle emergence in Aegiceras corniculatum (Primulaceae), a crypto-viviparous (radicle emerge piercing the seed coat but not piercing the fruit coat, prior to dispersal) mangrove species. This finding suggests that viviparous and cryptoviviparous species may exhibit epicotyl dormancy, overturning the standard notation that all viviparous seeds are non-dormant. Thus, viviparous seeds are either (truly) ND or PD with subclass-epicotyl Level 1 (i.e., radicle ND).
Group 2- non-dormant and morphological dormancy
The second group of seeds that germinated within 30 days are either ND or MD. Post-radicle emergence, germinated seeds must be evaluated at a wide-range of temperature and light gradients. This is because, shoot emergence may require specific conditions and testing at one or two conditions may fail to provide the precise set of conditions. If shoot emergence is delayed for more than 30 days under any temperature-light combination, these seeds have some component of epicotyl dormancy. Therefore, seeds initially classified as ND are not (truly) ND and may possess PD with subclass-epicotyl of varying levels, necessitating leaf emergence monitoring for definitive non-dormancy classification. For example, Farmer (Reference Farmer1977) and Allen and Farmer (Reference Farmer1977), showed that acorns of Quercus alba, Q. prinus, and Q. ilicifolia germinate in autumn with immediate root growth, but shoot emergence is delayed until spring. When freshly collected acorns germinate at typical fall soil temperatures (10°-15°C), root growth occurs, while shoot emergence requires a period of cold stratification. This adaptation establishes roots before winter while protecting the vulnerable shoot from frost damage. Thus, the roots of these acorns are non-dormant, but the epicotyl is dormant, i.e., Level 1: Non-deep, Subclass 2- Epicotyl of Class Physiological Dormancy. Jaganathan and Phartyal (Reference Jaganathan and Phartyal2024) stated that many Fagaceae acorns reported to be ND could have epicotyl dormancy, and long-term monitoring of the shoot emergence is important.
Although Fagaceae species have been discussed extensively for the presence of epicotyl dormancy, an increasing number of studies show this could be much more common in other families. Jayasuriya et al (Reference Jayasuriya, Wijetunga, Baskin and Baskin2010) reported that seeds of Humboldtia laurifolia needed approximately 2 weeks for radicle emergence, but subsequent shoot development take about 1-2 months. A similar pattern was described for Cynometra cauliflora, which requires approximately 2 weeks for radicle emergence, followed by shoot development that occurs after an additional 4-6 weeks (Jayasuriya et al, Reference Jayasuriya, Wijetunga, Baskin and Baskin2012). Thus, both these species of Fabaceae have non-deep epicotyl PD due to their delayed shoot emergence after root development. In contrast, Brownea coccinea (Fabaceae) demonstrates a more pronounced epicotyl dormancy, where the radicle emerges within about 2 weeks, but the shoot may take up to 8 weeks or more to appear (Jayasuriya et al, Reference Jayasuriya, Wijetunga, Baskin and Baskin2012). For Chionanthus retusus (Oleaceae), which exhibits deep epicotyl dormancy, the radicle typically emerges within 1 week, but shoots may take up to 3 months to develop fully (Chien et al, Reference Chien, Kuo-Huang, Shen, Zhang, Chen, Yang and Pharis2004). Likewise, Lecythis ampla (Lecythidaceae) and Platonia insignis (Clusiaceae) also germinated within 30 days, yet shoot growth required up to 240 days (Mourão & Beltrati, (Reference Mourão and Beltrati1995c)).
In some ND and dormant seeds, the seed reserves are moved to the extending radicle, creating a bulge or corm-like structure, from which the shoot emerges after some time. For instance, in Polygonatum sibiricum (Asparagaceae) seeds, radicle emerges at 25 °C (over 30 to 60 days) with the development of a corm structure, but 60 days of cold-stratification is necessary to break epicotyl dormancy, with longer root lengths (2.5 cm) significantly enhancing shoot emergence timing at 25 °C (Liao et al, Reference Liao, An, Wei, Wang, Li and Qi2021).
We propose that long-term studies tracking complete seedling ontogeny in species exhibiting MD are essential for a comprehensive understanding of their germination and growth processes. It has been described in Bhesa nitidissima (Centroplacaceae) and Gaertnera walkeri (Rubiaceae) (Athugala et al, Reference Athugala, Jayasuriya, Gunarathne and Baskin2018); and Strychnos benthamii (Loganiaceae) ((Muthuthanthirige et al, Reference Muthuthanthirige, Wijetunga and Jayasuriya2020), that the underdeveloped embryos, indicated by an embryo-to-seed length ratio of ≤ 0.5 had T50 (time to 50% germination) of less than 30 days. When radicle emergence is considered the criterion for germination, these species could be categorized as having MD. However, despite optimal germination conditions, the emergence of the shoot from seeds of these three species takes longer than 30 days. This extended period before shoot emergence suggests that they should be classified as having MD epicotyl dormant seeds. However, more studies are needed to further divide them into levels and types.
Group 3- physiological dormancy (PD) or morphophysiological dormancy (MPD)
In ND seeds, terminating germination experiments immediately after radicle emergence could lead to incorrect dormancy classification; otherwise, species having PD with subclass epicotyl (Fig. 1). A similar problem could be evident in seeds with PD and MPD, where one could have applied dormancy-breaking treatment(s), and terminating the experiments with successful radicle emergence underestimates the presence of epicotyl dormancy. For instance, in Nothapodytes nimmoniana (Icacinaceae) and Symplocos cochinchinensis (Symplocaceae) warm stratification at 25 °C combined with gibberellic acid (GA3) treatment led to the radicle emergence within 46-100 days and 28 days, respectively. However, epicotyl emergence was significantly delayed in both species. If the radicle has PD along with epicotyl having dormancy, these species would be Level 2. Deep, Subclass 2 Epicotyl of Class PD, with Type 1 or 3. This has also been shown to occur in several Fagaceae species, particularly in Quercus (Jaganathan and Phartyal, Reference Jaganathan and Phartyal2024) and Castanopsis species (Li et al, Reference Li, Jaganathan, Kang and Liu2023). Thus, studying “full germination” is critical to understand the dormancy mechanisms.
Several species with MPD also have been reported to have epicotyl dormancy. At family level, this has been studied extensively in Rannunculaceae. Seeds of Helleborus thibetanus exhibits MPD with epicotyl dormancy requiring sequential treatment: 6 weeks at 15°C (or GA3) for radicle emergence and 4 weeks at 4°C (or GA3) for epicotyl emergence, albeit only in > 13% seeds (Gong et al, Reference Gong, Cao, Chen, Peng, Zhang, Shi and Du2024). The requirement of cold and/or warm stratification for radicle and shoot emergence appears to be wide-spread in species with MPD with epicotyl dormancy. Similarly, working with Anemone nemorosa from northern Italy, Mondoni et al (Reference Mondoni, Probert, Rossi, Hay and Bonomi2008) reported that the seeds have underdeveloped embryos requiring growth before germination and a physiological component delaying germination, involving epicotyl emergence. Thus, these seeds have deep, simple epicotyl MPD, which require cold-stratification for shoot emergence.
Studies in three Crocus ser. Verni (Iridaceae) of Mediterranean and temperate climate showed that the seeds exhibit nondeep simple MD, requiring warm stratification (20–25°C) followed by cool temperatures (∼10°C) for embryo growth and radicle emergence in November, while shoot emergence in March depended on species-specific processes (Carta et al, Reference Carta, Probert, Moretti, Peruzzi and Bedini2014). For C. neapolitanus, adapted to temperate climates, shoot emergence and epicotyl dormancy breaking required cold stratification, whereas the sub-Mediterranean C. etruscus and Mediterranean C. ilvensis relied on gradual development under natural temperature fluctuations. These variations in dormancy mechanisms reflect ecological adaptations to different climatic conditions, highlighting the influence of phylogeny on dormancy evolution within the Crocus genus. Indeed, nondeep simple MPD has been reported in several other Iridaceae species, e.g., Crocus alatavicus (Fu et al, Reference Fu, Tan, Baskin and Baskin2013).
Many temperate Viburnum (Viburnaceae) species exhibit deep simple epicotyl MPD, requiring a sequence of warm and cold stratification. Initial warm temperatures, typically around 20/10°C to 25/15°C, promote embryo growth, with embryo length increasing significantly before radicle emergence (Hidayati et al, Reference Hidayati, Baskin and Baskin2005; Baskin et al, Reference Baskin, Chien, Chen and Baskin2008; Phartyal et al, Reference Phartyal, Kondo, Fuji, Hidayati and Walck2014). Radicle emergence itself is also optimized at approximately 20/10°C, often requiring several weeks to months, e.g., 24 weeks for 50% emergence in V. betulifolium (Chien et al, Reference Chien, Chen, Tsai, Baskin, Baskin and Kuo-Huang2011). Subsequent shoot emergence is strictly dependent on a prolonged period (≥12 weeks) of cold stratification at 5°C following radicle emergence. In contrast, some species, like V. tinus (Karlsson et al, Reference Karlsson, Hidayati, Walck and Milberg2005) and V. odoratissimum (Baskin et al, Reference Baskin, Chien, Chen and Baskin2008), exhibit non-deep simple epicotyl MPD, requiring shorter periods of warm stratification for both radicle and shoot dormancy release. If the experiments were terminated after radicle emergence, the seeds would have been misclassified as MPD, and the critical role of the epicotyl would have remained uncharacterized.
In Paeonia ostii (Paeoniaceae), radicle emergence occurs in autumn, but shoot emergence is delayed until late winter, requiring cold stratification (Zhang et al, Reference Zhang, Pan, Baskin, Baskin, Xiong, Cao, Yao, Tang, Zhang and Tao2022). Seeds of Paeonia peregrina exhibit MPD, which requires warm stratification and GA3 for growth of the embryo and radicle emergence, followed by cold stratification for shoot development (Prijić et al, Reference Prijić, Mikić, Peškanov, Zhang, Guo, Dragumilo, Filipović, Anačkov and Marković2024). Many Paeoniaceae species, including P. ludlowii seeds require fluctuating temperatures – cold, warm, and cold – to break dormancy, with the radicle emerging in the first year and the plumule (shoot) developing in the second year (Zhang et al, Reference Zhang, Yao, Zhang, Baskin, Baskin, Xiong and Tao2019). Furthermore, it appears that epicotyl becomes sensitive to cold stratification only after the root has reached a certain length, and that exogenous application of gibberellic acid (GA₃) or scarification can enhance embryo growth and germination rates, but this requires to be tested in a wide range of species (Hao et al, Reference Hao, He, Li, Shi and Tang2014).
Investigations on MPD with subclass double dormancy have been less studied, but available evidence shows root and shoot emergence can take up to two years (Adams et al, Reference Adams, Baskin and Baskin2003). For instance, Convallaria keiskei (Asparagaceae) exhibits deep simple double MPD, requiring two years for complete germination: roots emerge in the first spring, followed by embryo growth, and shoots emerge in the second spring, with cold stratification being essential for shoot emergence (Kondo et al, Reference Kondo, Narita, Phartyal, Hidayati, Walck, Baskin and Baskin2015). Seeds of Trillium camschatcense (Melanthiaceae) also have simple double deep MPD, which requires cold stratification for radicle emergence, followed by a second winter for shoot emergence (Kondo et al, Reference Kondo, Walck and Hidayati2019). These cases further exemplify the need for reporting “full germination.” Furthermore, it is not easy to infer which species might have epicotyl dormancy, as some families have very less species documented (Fu et al, Reference Fu, Tan, Baskin and Baskin2013; Rhie et al, Reference Rhie, Lee and Kim2019) compared with well-studied genera or families (Baskin et al, Reference Baskin, Chen, Chien and Baskin2009).
Physical dormancy
One would argue that the germination failure in the protocol developed may occur due to the presence of a water-impermeable seed coat. However, to the best of our knowledge, epicotyl dormancy has not been reported in PY species (Jaganathan, Reference Jaganathan2025). While we recommend conducting an imbibition test when seeds fail to germinate within 30 days, monitoring seeds until first leaf emergence is critical (Fig. 1). The imbibition test quantifies water uptake by comparing initial seed mass with subsequent mass changes under controlled moisture conditions. Seeds with PY exhibit complete impermeability; consequently, showing no measurable water absorption. Thus, slower water absorption could be caused by the physical barrier of thick endocarps. In this case, physical barrier is not equal to PY, and should not be misinterpreted as PY.
Conclusion
The complexity of seed dormancy mechanisms necessitates a paradigm shift in germination studies methodology, specifically advocating for the extension of experimental observations until the first true leaf emergence. Many studies articulated that casual observation of leaf emergence led to subsequent detailed investigations on epicotyl dormancy (Jayasuriya et al, Reference Jayasuriya, Wijetunga, Baskin and Baskin2010; Kondo et al, Reference Kondo, Narita, Phartyal, Hidayati, Walck, Baskin and Baskin2015). Our methodological refinement makes casual observation in to systematic process, thereby bridging the temporal dissociation between radicle protrusion and shoot emergence in species exhibiting epicotyl dormancy. We also suggest reporting the “full germination” period, and standardizing germination reporting protocols to include the comprehensive germination period, defined as the interval from seed sowing to first true leaf emergence. This is particularly crucial to enhance our understanding of germination ecophysiology and facilitate more accurate comparisons across species and studies. Whenever possible, we suggest documenting distinct temporal parameters for both radicle protrusion and shoot emergence, as these represent physiologically distinct developmental phases.
In seeds lacking epicotyl dormancy, whether initially dormant or non-dormant at dispersal, shoot emergence typically occurs within 30 diurnal cycles post-radicle protrusion (Baskin and Baskin, Reference Baskin and Baskin2021). However, delayed shoot emergence exceeding this temporal threshold strongly indicates epicotyl dormancy, necessitating specific dormancy-breaking treatments for successful shoot system development. The germination process in epicotyl dormant seeds involves initial radicle emergence followed by subsequent shoot development, each phase characterized by distinct thermophysiological requirements and temporal dynamics. This physiological complexity is further modulated by the underlying dormancy classification. Seeds exhibiting PD with Subclass 2 Epicotyl can be further differentiated into nondeep (Level 1) and deep (Level 2). These seeds demonstrate successful radicle emergence but exhibit temporally delayed shoot development, requiring specific environmental conditions such as cold or warm stratification regimes or elevated temperature exposure for complete germination. In contrast, MPD represents a more intricate category, encompassing nondeep epicotyl, deep epicotyl, and deep simple double, each characterized by distinct physiological and morphological barriers to germination. Thus, our proposal that monitoring seed germination until the emergence of the first leaves provides a systematic approach for categorizing the diverse mechanisms governing seed dormancy, will provide more nuanced understanding with significant implications for both theoretical seed biology and practical horticultural applications.
Acknowledgements
GKJ acknowledges the financial support by National Science Foundation China (NSFC) with grant number 32001119 to carry out studies on seed dormancy.
Open access
