Variability in seed traits has been reported in different species, populations or individuals (see, e.g., Susko and Lovett-Doust, 2000; Lopez et al., 2008; Cervantes and Martín Gómez, 2019). Differences in seed (germination) traits among populations (i.e., within species) have been considered an effective evolutionary strategy that influence on species resilience by increasing the chances of reproduction under heterogeneous environmental conditions, thus minimizing the extinction risk (Venable and Brown, 1988). Hence, understanding interpopulation differences might help us addressing issues related to the evolution of variability patterns regarding plant regeneration from seeds. Seed physical traits (e.g., seed size/mass, shape, color, seed coat thickness, permeability, etc.), for instance, are directly or indirectly involved in regulating germination responses and many seed functions such as dispersal, type and level of dormancy, formation of the soil seed bank, seedling establishment, survival, and competitive ability (Eriksson, 1999; Larios et al., 2014; Larson and Funk, 2016; Saatkamp et al., 2019). Therefore, these traits have been related to plant fitness and species persistence across ecosystems (Atis et al., 2011; Xu et al., 2014; Bhatt et al., 2016, 2021a; Bu et al., 2016). Evolutionary history, environmental context, genetic constraints, and plasticity are responsible for variation in seed traits at both intra- and inter-species levels (Violle et al., 2012; Ge et al., 2020).

Within species, interpopulation variation is vital to distinguish the ecological and evolutionary drivers shaping regeneration traits (Moreira et al., 2012). Interpopulation variability of ecological traits generally occur either due to (i) differences in environmental conditions among the different populations (Nicotra et al., 2010; Cochrane et al., 2014) or (ii) hereditary characteristics of populations (Cheptou et al., 2008; Cochrane et al., 2015). In addition, other factors can be responsible for shaping the interpopulation variability in seed traits, including local climate, habitat characteristics, inter- and intra-species competition levels, predation, resource provisioning, and epigenetics (Kleunen et al., 2001; Tautenhahn et al., 2008; Saatkamp et al., 2019). The variability in seed germination within a species, when seeds are collected from different populations, has been linked to spatial variability due to climatic and edaphic parameters or a differentiated gene pool (Elnaggar et al., 2019; Bhatt et al., 2021b,c). Most of the studies so far focused on testing interpopulation variation regarding seed mass or size and their effect on seed germination (Xu et al., 2016; Kołodziejek, 2017; Moya et al., 2017; Bhatt et al., 2019). However, few studies investigated interpopulation variation in other physical traits, like seed shape and color and their role in regulating seed dormancy and germination (but see Zhang and Maun, 1990; Mölken et al., 2005; Atis et al., 2011; Bhatt et al., 2016).

Generally, seed germination requirements are synchronized with the population’s environmental condition to increase the possibility of seedling survival and establishment (Baskin and Baskin, 2014). Studies have shown that interpopulation variation in seed dormancy and germination is broadly correlated with abiotic factors, such as altitude, latitude, temperature, soil moisture, and habitat type (Wagmann et al., 2012; Fernández-Pascual et al., 2013; Cochrane et al., 2014; Pendleton and Pendleton, 2014). Because different populations among different localities may experience different natural selection pressure, they may lead to local adaptation over time. For instance, higher interpopulation variability in germination responses has been reported in populations living in unpredictable environmental conditions than those living in more stable (i.e., less variable) conditions (Meyer and Allen, 1999). Therefore, the same species growing in different populations can respond differently to the same environmental signals based on the selective pressures acting on the population (Oduor et al., 2016; Amini et al., 2017).

Given that interpopulation variation usually reflects the environmental adaptation of species (Kyle and Leishman, 2009), the present study focused on interpopulation variation in seed physical (mass, size, shape, color, etc.) and physiological (germination responses) traits of five Polygonaceae species. Germination at species level has been previously investigated for this family, but mostly using seeds from a single population, as reported for seeds of Polygonum lapathifolium (Timson, 1965; Staniforth and Cavers, 1979), Reynoutria japonica (Mariko et al., 1993), and Rumex obtusifolius (Hand et al., 1982; Van Assche and Vanlerberghe, 1989; Benvenuti et al., 2001). In the present study, we hypothesized that different populations of species growing in different microclimates might also differ in their seed morphology and germination strategy. To test this hypothesis, we examined the following questions (i) is there a difference in seed physical traits (size, mass, shape, and color) according to the seed population? (ii) are there any light and temperature mediated mechanisms affecting seed physiological (germination) traits under laboratory conditions? (iii) is there a different seed germination response to light and temperature among the seed populations within each study species?

Germination responses can be difficult to predict over time and space, as they are affected by several environmental factors such as temperature, light, and soil moisture (Baskin and Baskin, 2014; Bhatt et al., 2019, 2020, 2021a). Identification of the environmental factors that are responsible for maintaining phenotypic variation in nature is essential to understand the evolutionary mechanisms of plants (Larios et al., 2014; Manzano-Piedras et al., 2014). Existence of interpopulation variability in seed traits is commonly found, and important for enhancing the ability of species to persist against climatic variability (Cochrane et al., 2015). In this study, we found a relatively high interpopulation variation in seed morphology and germination response (physical and physiological traits) of five Polygonaceae, although these species belong to the same family and were collected from geographically close locations (located approximately 4–32 km from each other). Despite the multiple seed dispersal vectors of the study species and relatively close distance between populations, it is important to note that the collected populations occur in a fragmented landscape, which may limit their gene flux from one population to others.

Intraspecific (interpopulation) trait variation could be due to environmental, genetic differences, or a combination of both genetic and environmental factors. Understanding the role of seed physical traits and their relationship with seed germination could thus be important for predicting which seed physical traits are more influential in affecting germination. However, until now, most of the studies mainly focused on the role of seed mass/size in germination responses (Wu and Du, 2007; Münzbergová and Plačková, 2010; Wu et al., 2011; Alstad et al., 2018; Chen et al., 2018). Despite differences between species, seeds of all studied species can be classified as small-sized, with a fresh mass varying from 0.45 to 4.44 mg. Small-seeded species have been recognized to be light-dependent (positively photoblastic) in different vegetation types worldwide, including temperate grasslands (Milberg et al., 2000), arid zones (Flores et al., 2011), and tropical rainforests (Aud and Ferraz, 2012). Such physical trait-germination relatedness may help to explain the germination patterns we found.

Only seeds of P. lapathifolium and R. japonica had significant interpopulation variability in seed mass and seed length. The smallest seed size found in the Rj1 population proportioned a low percentage of germination regardless of light and temperature treatments. However, other seed traits, such as seed width, seed height, and seed shape index, varied significantly among populations of most studied species. Differences in the environmental conditions among populations can be responsible for seed trait variability, as suggested to vegetative and floral plant traits (Kuppler et al., 2020). But in our study, one must take in account that seeds of the different species were collected from populations growing geographically in close to moderate distance localities (4–32 km). Therefore, we argue that spatial heterogeneity in edaphic factors (i.e., nutrients) or differences in the gene pool among populations can contribute to determining the interpopulation variability, especially in geographically distance populations. In addition to subtle contrasts in environmental gradients, different populations are highly likely to be subjected to different selection pressures under natural conditions, leading to the evolution of interpopulation variability in seed traits as an adaptive response that allows the species to survive under spatially heterogeneous landscapes (Ge et al., 2020).

The exception was found for R. trisetifer seeds, with no difference in seed traits among the studied populations. Additionally, both Rumex species had the least light-dependent germination responses, despite their small seed size and round seed shape. Round and small seeds, due to their compact structure, should be more easily incorporated into soil seed banks and therefore subjected to dark conditions or minimal light penetration in the upper layer of the soil (Benvenuti, 2007; Saatkamp et al., 2014). Funes et al. (1999) reported that compact seeds tend to persist for longer periods ungerminated in the soil. In contrast, in our case, the seeds of R. trisetifer and R. obtusifolius had the lowest shape index but high germination under both light and dark conditions, under optimal temperature. This indicates that their seeds would be promptly recruited rather form persistent seed banks. Factors other than seed physical traits may influence germination responses, such as adaptive responses and phylogenetic relatedness, regardless of population variability. Moreover, interactions of temperature and light have been pointed out as important drivers of the capacity of seeds to germinate depending on the species and specific temperature requirements (see Pons, 1992).

Seeds of all the five species were able to absorb water, indicating they do not show physical dormancy. Seeds of only two taxa (P. lapathifolium var. salicifolium and R. obtusifolius) had a significant variability in seed coat permeability (water imbibition) among populations. This variability in water imbibition among populations could be related to variation in coat hardness (thickness) due to spatial heterogeneity in habitat conditions, genetic background, and their interactions, which may be responsible for interpopulation variability in seed coat permeability as reported in other species (Jaganathan et al., 2019; Bhatt et al., 2021b). Bearing non-dormant seeds, all studied species were able to germinate immediately after dispersal, although the percentage of germination varied between or among populations and the tested temperature conditions. Fast germination could be related to the availability of sufficient precipitation throughout the year due to subtropical moist monsoonal climate in the seed collection sites (Kang et al., 2017). This contrasts with seasonal water restrictions, which are generally considered one of the main limitations for germination in most climatic conditions.

Temperature is one of the main environmental factors that regulate the germination, especially where availability of water is not a constraint for germination (Fenner and Thompson, 2005). Regarding thermal requirements, in the light, optimal germination ranges mainly varied from 10/20°C to 25/35°C, with significant decrease in germination percentage at both cooler and warmer extremes. However, some species, such as Polygonum spp., showed a marked variation in the optimal temperature requirement for germination depending on the population. Seeds of P. lapathifolium var. salicifolium from Pls2, for example, germinated equally well at the warmest tested temperatures, reaching >80% germination at 25/35°C and 35/40°C. Conversely, seeds from Pls1 attained maximum germination (56%) at a milder temperature regime (20/30°C). Seeds of P. lapathifolium also showed similar trends: The Pl2 population attained 85% germination under the warmest conditions (35/40°C), while Pl1 and Pl3 (both populations collected at elevation >1,000 m) attained maximum germination of 72–89% at temperatures 20/30°C and 25/35°C. Producing seeds with varying germination cues (temperature) may enhance fitness of these species in future climate change scenarios and reduce the risks of synchronizing recruitment in case of unfavorable conditions (Santo et al., 2015; Bhatt et al., 2020, 2021b).

An even higher interpopulation variability in maximum germination percentage was observed for R. japonica seeds—from 29% in Rj1 to 89% in Rj3, both at 10/20°C. Although Rj3 also germinated well (66%) at 25/35°C. This relatively wide temperature niche width (especially in Rj3) indicates such seeds (populations collected at elevation from 500 to 1,200 m) can better recruit when other environmental factors controlling seed germination (light &/or moisture) may be available beyond the normal window of seed germination season. Collection sites of R. japonica and P. lapathifolium populations had remarkable differences in altitude, which may be responsible for microclimatic contrasts. We argue that habitat heterogeneity and climatic variability contribute more toward determining the interpopulation variability rather than the geographic distance. Additionally, differences in germination requirements among populations of P. lapathifolium var. salicifolium could be attributed to the variation in competition, soil moisture, and mineral nutrition (Burton et al., 2000; Allison, 2002; Swain et al., 2006). Spatio-temporal heterogeneity in climatic conditions and edaphic factors among the population are reported to be the main driver for interpopulation variability (Kuppler et al., 2020), which ultimately lead to the interpopulation variability in seed morphology and germination traits (Baskin and Baskin, 2014; Hudson et al., 2015; Bhatt et al., 2021b).

Rumex species had less variation in germination percentage among the tested populations within their respective optimal ranges. Seeds of R. trisetifer germinated equally well at 10/20°C and 20/30°C (≥96%), while R. obtusifolius germinated equally well at 10/20°C, 20/30°C, and 25/35°C (89%). The relatively broad thermal germination niche indicates that seeds of R. obtusifolius can germinate throughout the year, although the germination percentage was significantly reduced during winter (December to February). Similarly, Polygonum spp. can germinate outside their normal germination window, when the temperature ranges between 20 and 35°C. These results further support that seed recruitment in our study species is widely spread throughout the growing season, and this will increase their long-term fitness by spreading the risk of germination failure. Besides this, these species have higher possibility to establish in a broad environmental condition (Carlucci et al., 2015) and will be less affected by climate change as compared to those have specific temperature requirement for germination. For example, R. trisetifer and R. japonica have a specific temperature requirement and thus their germination may be restricted to the specific time (i.e., between March to June or between September to November for R. trisetifer, and between March to April or between October to November for R. japonica).

Overall, seed germination was significantly inhibited at the lowest temperature regime (except for Ro2), although different species showed interpopulation variability in terms of germination percentage. Avoiding germination in winter (December to February), when the temperature is around 5/10°C, could be a survival strategy for these species, as the chances of seedling successful establishment would be extremely low due to cold and frost. Low-temperature conditions may induce physiological dormancy, while the cold stratification period also helps seeds to synchronize germination with a favorable season, as reported in Polygonum and Rumex spp. (Totterdell and Roberts, 1979; Bouwmeester and Karssen, 1992; Batlla and Benech-Arnold, 2005). However, at the highest temperature (35/40°C), one population of P. lapathifolium var. salicifolium (Pls2) and a population of P. lapathifolium (Pl2), both from the lowlands, germinated well (>80%). For all other species/populations, germination percentage was significantly reduced at the warmest condition. These results indicate that tolerance to high and low temperatures during the germination stage can vary between/among populations, depending on species (Cochrane, 2016). Therefore, it could be imperative to test the interpopulation variability under a wider range of temperature, in order to model regeneration behavior regarding threshold models (Bradford, 2002).

Here, we found general patterns of seed and germination traits common to all five Polygonaceae, such as small seed size and predominant light-dependent germination (except Rumex). However, we also identified particularities of the studied populations. The optimal temperature requirements for germination varied according to the population of the seeds, mostly for P. lapathifolium and R. japonica, seeds collected from comparatively longer altitudinal gradients. Furthermore, interpopulation variability among the studied species could be linked to maternal effects (Luzuriaga et al., 2006; Kagaya et al., 2011; Bhatt et al., 2019, 2021b). Heterogeneity in maternal habitat conditions where populations grow is likely to be the main factor contributing to interpopulation trait variability. Phylogenetically related species have been reported to share similar traits in a general sense (Losos, 2008). For example, some studies suggest that related species share similarities regarding seed traits, dormancy, and germination characteristics (Moles et al., 2005; Norden et al., 2009; Carta et al., 2016). Other studies argue that seed traits, dormancy, and germination characteristics are regulated by environmental cues rather than the phylogeny (Vandelook et al., 2008; Arana et al., 2016; Fang et al., 2017). Therefore, it is important to compare the physical and physiological traits of closely related species to better understand which factors (phylogeny and/or environment) are responsible for influencing regeneration strategies, aiding to improve our predictive capacity to comprehend species distribution related to trait variability in changing environments.

Although seeds of the Polygonaceae shared similar physical traits, generally displaying small-sized seeds and light-dependent germination patterns. We found great variation in germination requirements between and/or among populations. For instance, optimal temperature regimes may change according to seed population, mostly in altitudinal gradients, as we show for P. lapathifolium and R. japonica seeds. Seeds of Rumex spp., on the other hand, seemed to be less affected by light availability, with similar responses regardless of the collection sites. Variability in regeneration traits can be crucial to comprehending the species’ ability to colonize different areas, aiding in the identification of tolerant/sensitive populations affected by environmental changes.

The original contributions presented in the study are included in the article/Supplementary material, further inquiries can be directed to the corresponding author.

AB conceived and designed the experiments. AB and XC performed the experiments. AB and SP analyzed the data. AB, SP, and LD wrote the manuscript. All authors approve of this submission.

AB thanks to the Talents Program of Jiangxi Province (PR China) (Grants jxsq2020104003) for funding.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.