Introduction
Available information indicates that seeds of several wetland species can tolerate flooding and remain viable even after long periods of submergence (Parolin et al., Reference Parolin, Ferreira and Junk2003; Lucas et al., Reference Lucas, Mekdece, Nascimento, Holanda, Braga, Dias, Sousa, Rosa and Suemitsu2012; de Melo et al., Reference de Melo, Franco, Silva, Piedade and Ferreira2015; Poschlod and Rosbakh, Reference Poschlod and Rosbakh2018). Some of them can germinate or even form seedlings while they are submerged under water (de Oliveira Wittmann et al., Reference de Oliveira Wittmann, Piedade, Parolin and Wittmann2007; Ferreira et al., Reference Ferreira, Piedade, Junk and Parolin2007; Phartyal et al., Reference Phartyal, Rosbakh and Poschlod2018; Valdez et al., Reference Valdez, Hartig, Fennel and Poschlod2019), whereas others may not tolerate flooding at all (de Melo et al., Reference de Melo, Franco, Silva, Piedade and Ferreira2015; Valdez et al., Reference Valdez, Hartig, Fennel and Poschlod2019). However, to our knowledge, there is very little information available (Phartyal et al., Reference Phartyal, Rosbakh, Ritz and Poschlod2020; Rosbakh et al., Reference Rosbakh, Phartyal and Poschlod2020) on how seed germination of wetland species such as mudflat species that remain viable under long flooded (hypoxic) conditions respond to non-flooding (aerobic) conditions immediately after water drawdown.
Regenerative traits like seed dormancy, germination timing and seedling establishment (Poschlod et al., Reference Poschlod, Abedi, Bartelheimer, Drobnik, Rosbakh, Saatkamp and van der Maarel2013; Saatkamp et al., Reference Saatkamp, Cochrane, Commander, Guja, Jimenez-Alfaro, Larson and Walck2019; Phartyal et al., Reference Phartyal, Rosbakh, Ritz and Poschlod2020; Rosbakh et al., Reference Rosbakh, Phartyal and Poschlod2020) are extremely sensitive to abiotic factors, including temperature, light, moisture and aerobic/hypoxic conditions, that represent the major bottleneck to species recruitment (Fenner and Thompson, Reference Fenner and Thompson2005). A slight mismatch between regenerative traits and abiotic factors may negatively affect the synchronization of seed germination with the most favourable conditions not only for seedling development but also for juvenile plant recruitment (Saatkamp et al., Reference Saatkamp, Cochrane, Commander, Guja, Jimenez-Alfaro, Larson and Walck2019). Thus, the present study investigated how does prior exposure to flooding impacts the seed germination process of wetland (mudflat) species.
In temperate regions, the wetland habitats along the margin of rivers, streams, backwaters and oxbow lakes are often flooded during the winter season and the period of water drawdown coincides with the growing summer season (Voigtlander and Poppe, Reference Voigtlander, Poppe and Dodge1989). These recently exposed nutrient-rich muddy sediments support quick growth and dominance of annual herbs (Webb et al., Reference Webb, Dennis, Bates and Snyder1988; Baskin et al., Reference Baskin, Baskin and Chester1993, Reference Baskin, Baskin and Chester2004; Phartyal et al., Reference Phartyal, Rosbakh, Ritz and Poschlod2020). This kind of mudflat habitats are temporal, unpredictable and unstable in nature, often existing only for a few weeks at every part of the hydroperiod gradient, provided they are drained, open and free of litter. In some extreme cases, they may remain under flooded (hypoxic) conditions for decades or even centuries (Poschlod and Rosbakh, Reference Poschlod and Rosbakh2018). This unpredictable environment exposes seeds of mudflat species to a quick transition from hypoxic to aerobic conditions and alters their regeneration pattern either by promoting or inhibiting the seed germination process (Baskin et al., Reference Baskin, Baskin and Chester1994; Böckelmann et al., Reference Böckelmann, Tremetsberger, Šumberová, Grausgruber and Bernhardt2017; Phartyal et al., Reference Phartyal, Rosbakh, Ritz and Poschlod2020; Rosbakh et al., Reference Rosbakh, Phartyal and Poschlod2020).
There is contrasting evidence on how flooding affects plant regeneration in wetlands. On the one hand, flooding has been shown to facilitate seed dormancy break and enhance germination in seeds of Cyperus erythrorhizos, C. flavicomus (Baskin et al., Reference Baskin, Baskin and Chester1993), Rotala ramosior (Baskin et al., Reference Baskin, Baskin and Chester2002a), Schoenoplectus purshianus (Baskin et al., Reference Baskin, Baskin and Chester2000a), Scirpus juncoides, S. mucronatus, S. smithii and S. wallichii (Watanabe and Miyahara, Reference Watanabe and Miyahara1989). On the other hand, flooding was shown to induce secondary dormancy and suppressed germination in seeds of Bidens tripartita (Benvenuti and Macchia, Reference Benvenuti and Macchia1997), Echinochloa crus-galli (Honěk and Martinkova, Reference Honěk and Martinkova1992; Holguín et al., Reference Holguín, Crepy, Striker and Mollard2020), Lobelia dortmanna (Farmer and Spence, Reference Farmer and Spence1987) and Vallisneria americana (Jarvis and Moore, Reference Jarvis and Moore2008). As for mudflat plants, the seeds of many species have been demonstrated to come out of dormancy in a higher percentage under flooded (hypoxic) than under non-flooded (aerobic) conditions (Baskin et al., Reference Baskin, Milberg, Andersson and Baskin2002b). In temperate regions, where mudflat habitats are flooded during the unfavourable (winter) season and remain non-flooded during the favourable (summer) season, dormancy will be broken while seeds are still flooded. However, in seeds of the summer annual Leucospora multifidi, dormancy release is initiated in late autumn during non-flooded conditions and continues even during winter flooding. The non-dormant seeds germinated in summer but, if seeds were flooded prior to the onset of dormancy release, flooding prevented dormancy break in seeds of this species (Baskin et al., Reference Baskin, Baskin and Chester1994). Therefore, this adaptation can result in much higher germination percentages and rates and a more synchronized germination when the water recedes (Baskin and Baskin, Reference Baskin and Baskin2014). Another adaptation of seeds of mudflat species is that high summer temperatures do not cause flooded seeds to re-enter dormancy (Baskin et al., Reference Baskin, Baskin and Chester1993). Thus, regardless of when waters recede during the growing season (from early spring to early autumn), seeds remain non-dormant and can germinate (Baskin and Baskin, Reference Baskin and Baskin2014) if exposed to an appropriate temperature regime.
Considering the quick transition from flooding to non-flooding, we hypothesized that there should be a close integration of physiological responses of germinating seeds with a change in oxygen and temperature regimes. We were specifically interested to know whether or not hypoxic conditions interfere with the induction/release of seed dormancy in mudflat species exposed to different temperature regimes. Flooding and non-flooding conditions at an appropriate temperature regime (season) are supposed to play a crucial role in triggering germination at a time most suitable for seedling growth or to induce dormancy to avoid an unfavourable time for seedling survival. For example, seeds of the winter annual grass (Alopecurus carolinianus) flooded in autumn did not enter secondary dormancy during exposure to low winter temperatures in Kentucky (USA), but those buried in non-flooded (moist) soil and exposed to the same temperatures did so. Thus, it is indicated that hypoxic conditions prevented the induction of dormancy in non-dormant seeds (Baskin et al., Reference Baskin, Baskin and Chester2000b). Taking all these factors into account, in the present study, we specifically asked the following questions: (1) Does prior exposure of seeds of mudflat species to hypoxic conditions help them germinate in high percentages, rates (speed), and in a more synchronized manner when shifted to aerobic conditions? (2) Does exposure of dormant seeds to hypoxic conditions substitute for requirements of a cold stratification treatment to break dormancy, if any? (3) Does prior exposure of non-dormant seeds to hypoxic conditions induce secondary dormancy or suppress germination?
Material and methods
Study species
We selected five mudflat species: Hypericum humifusum L. (Hypericaceae); Limosella aquatica L. (Scrophulariaceae); Lythrum hyssopifolia L. (Lythraceae); Peplis portula L. (Lythraceae) and Samolus valerandi L. (Primulaceae), typically inhibiting mudflat habitats of Central Europe. H. humifusum primarily belongs to moist sandy grasslands/trampled communities along mudflats, L. aquatica, L. hyssopifolia and P. portula occur only in mudflats, whereas S. valerandi mainly occurs in salt marshes along with mudflat habitats (Oberdorfer, Reference Oberdorfer2001). In our previous research, seeds of L. hyssopifolia, P. portula and S. valerandi germinated moderately well in both aerobic and hypoxic conditions irrespective of temperature fluctuation with full illumination and were categorized as ‘flood-resistant mudflat species’, whereas the seeds of L. aquatica germinated well only under fully illuminated aerobic conditions at fluctuating temperatures and were categorized as ‘true mudflat species’. In contrast, the seeds of H. humifusum had a high germination percentage and rate only under aerobic conditions at both constant and fluctuating temperatures with and without light and were categorized as ‘facultative mudflat species’ (Phartyal et al., Reference Phartyal, Rosbakh, Ritz and Poschlod2020).
Fully ripened fruits of these species were harvested in several natural populations from randomly chosen individuals (>50 individuals per species) growing at a step-away distance from each other and thoroughly mixed. After collection, seeds were separated from the fruits by hand, air-dried for several days, cleaned and kept dry in a cold room at 4°C until the beginning of the germination experiments (Baskin and Baskin, Reference Baskin and Baskin2014). Seeds of all study species were non-dormant except for L. hyssopifolia that possess physiological dormancy and require cold moist stratification to overcome dormancy (Phartyal et al., Reference Phartyal, Rosbakh, Ritz and Poschlod2020).
Seed germination experiments
To determine optimum conditions for high percentages, rates (speed) and synchronization of germination, seeds of each species were incubated under different combinations of temperature and oxygen. In each incubation treatment, five replicates of 20 seeds were placed on two layers of moist filter paper in a Petri dish. All the experiments were conducted in germination chambers (RUMED GmbH, Laatzen, Germany) at one constant (22/22°C) and three diurnal fluctuating temperatures (14/6, 22/14 and 30/22°C – representing climate of cool spring, mild and warm summer seasons in temperate mudflat habitats) in 14/10 h of the light/dark regime.
To simulate hypoxic conditions, Petri dishes supplied with vents to provide consistent gas exchange were placed in a desiccator and their air volume was substituted with pure nitrogen. We repeated this procedure each time when germination was scored. Dishes to which this treatment was not applied are referred to as aerobic. The number of seeds germinated was scored every third day for 6 weeks. Germination was defined as the protrusion of a radicle through the seed coat. As highlighted above, mudflat habitats can experience flooding to non-flooding conditions in a very short period of time, which exposes seeds a rapid transition from hypoxic to aerobic conditions during the germination process. Since seeds of none of the study species germinated under strict hypoxic conditions they were, therefore, shifted to aerobic conditions at a similar temperature and light/dark conditions for another 6 weeks (see Table 1 for treatment details), to determine how prior exposure to hypoxic conditions affects seed germination.
Table 1. Outline of the seed incubation treatments
Statistical analysis
Seed germination percentages, mean germination time (MGT) and germination synchronization index (Z) were calculated using GerminaR package for R (Lozano-Isla et al., Reference Lozano-Isla, Benites-Alfaro and Pompelli2019). A lower value of MGT reflects faster rates (speed) of seed germination in a particular incubation treatment. In the case of the synchronization index, a Z-value close to one indicates that germination of all seeds occurs at the same time (more synchronized germination), while a Z-value close to zero indicates that seed germination of at least two seeds occurred at a different time (less synchronized germination). To infer statistically significant differences among the incubation treatments, Fisher's exact test was used. Seed germination percentages at any given treatments were displayed using mean values and binomial confidence intervals. All statistical analyses were carried out using R 3.4.3 (R core development team, 2020).
Results
None of the seeds of any study species germinated under strict hypoxic conditions at any temperature regimes, except those of L. hyssopifolia, within 6 weeks of incubation. Only 5% seeds of L. hyssopifolia germinated under strict hypoxic conditions at the warm (30/22°C) temperature regime. However, prior exposure to hypoxic conditions significantly (P < 0.05) stimulated seed germination percentage, MGT and germination synchrony of three study species (L. aquatica, P. portula and S. valerandi) at almost all temperature regimes when seeds were moved to aerobic conditions (Figs 1–3).
Fig. 1. Bar plots illustrating median and quartiles of seed germination percentage of mudflat species incubated under a wide range of environmental conditions. Letters represent subsets with significant (P < 0.005) differences.
Fig. 2. Box plots illustrating median, quartiles and outliers (o) of MGT of mudflat species incubated under a wide range of environmental conditions. Letters represent subsets with significant (P < 0.05) differences.
Fig. 3. Box plots illustrating median, quartiles and outliers (o) of germination synchronization index of mudflat species incubated under a wide range of environmental conditions. Letters represent subsets with significant (P < 0.05) differences.
On the contrary, germination percentage, MGT and germination synchrony of H. humifusum seeds were significantly (P < 0.05) suppressed by prior exposure to hypoxic conditions at most of the tested temperature except at the cool temperature (14/6°C) regime (Figs 1–3). The seeds germinated equally well (>90%) in both strict aerobic conditions and when moved from hypoxic to aerobic conditions at the cool temperature regime.
In contrast to seeds of the other four study species, those of L. hyssopifolia either failed to germinate or had a low (0–37) germination percentage when incubated only under strict aerobic conditions. Whereas a prior exposure to hypoxic conditions significantly (P < 0.05) stimulated seed germination (58–93%), MGT and germination synchrony at all test temperature regimes (Figs 1–3).
Prior exposure to hypoxic conditions induced dormancy in non-dormant seeds of H. humifusum except at low cool (14/6°C) temperature, whereas it acted as a dormancy-breaking treatment for dormant seeds of L. hyssopifolia.
Overall, majority of the germination parameters (germination percentage, MGT and synchrony index) performed significantly (P < 0.05) well at moderately warm fluctuating (22/14°C) temperatures followed by constant (22/22°C) temperatures in comparison to warm (33/22°C) and cool (14/6°C) temperature regimes, which were found to be comparatively sub-optimal for seed germination in most of the study species (Figs 1–3).
Discussion
Seed germination of none of the study species displayed any positive response only to strict hypoxic (except for L. hyssopifolia) conditions but generally displayed a positive response to strict aerobic conditions. This confirms our previous findings that the majority of mudflat species need aerobic conditions to germinate (Phartyal et al., Reference Phartyal, Rosbakh, Ritz and Poschlod2020; Rosbakh et al., Reference Rosbakh, Phartyal and Poschlod2020). However, the overall trends indicate that prior exposure of seeds to hypoxic conditions stimulates germination in the majority of our study species, thus confirming our hypothesis (Table 2).
Table 2. A summary of seed response to prior exposure to hypoxic environments. Values in column 3 and 4 represent the overall mean of four incubation temperature regimes
Non-dormant seeds of H. humifusum germinated in a moderately broad range of environments including dark except under hypoxic conditions (Phartyal et al., Reference Phartyal, Rosbakh, Ritz and Poschlod2020). Seeds of this species had a higher percentage, faster rate and more synchronized germination only in aerobic conditions in all tested temperature regimes. However, seeds responded poorly when moved from hypoxic to aerobic conditions at warmer temperatures (22/14, 22/22 and 30/22°C) but not a low temperature (14/6°C). Thus, prior exposure of seeds to hypoxic conditions induced secondary dormancy except at low temperature, suitable for cold stratification. This also confirms earlier reports that flooding (low oxygen level) may induce secondary dormancy in seeds of several wetland species (Farmer and Spence, Reference Farmer and Spence1987; Honěk and Martinkova, Reference Honěk and Martinkova1992; Benvenuti and Macchia, Reference Benvenuti and Macchia1997; Nishihiro et al., Reference Nishihiro, Araki, Fujiwara and Washitani2004a,Reference Nishihiro, Miyawaki, Fujiwara and Washitanib; Jarvis and Moore, Reference Jarvis and Moore2008; Holguín et al., Reference Holguín, Crepy, Striker and Mollard2020). Secondary dormancy may be eventually broken by cold stratification. This is an unusual behaviour of a species of an ephemeral habitat such as mudflats that might be explained by the main occurrence of H. humifusum in moist sandy grasslands or trampled sites with a comparatively low probability of flooding (Oberdorfer, Reference Oberdorfer2001). Thus, our results confirm the classification of this species by Phartyal et al. (Reference Phartyal, Rosbakh, Ritz and Poschlod2020) as a ‘facultative mudflat species’ that demonstrated a broad germination niche width (Phartyal et al., Reference Phartyal, Rosbakh, Ritz and Poschlod2020) as the species occurs not only in the mudflats.
Physiologically dormant seeds of L. hyssopifolia germinated over a broad range of environmental conditions, including hypoxic conditions after cold stratification (Phartyal et al., Reference Phartyal, Rosbakh, Ritz and Poschlod2020). However, seeds of this species either failed to germinate or germinated poorly under aerobic conditions at all tested temperatures. However, if seeds were exposed to hypoxic conditions then moved to aerobic conditions, they had a higher percentage, rate and synchronization of germination even without a cold stratification treatment. A high proportion of seeds of this species also germinated under hypoxic conditions at low temperature (4°C) during cold stratification (S. Phartyal, personal observation). Thus, a high percentage of seeds of several mudflat species, including L. hyssopifolia, classified as a ‘flooded mudflat species’ by Phartyal et al. (Reference Phartyal, Rosbakh, Ritz and Poschlod2020) come out of dormancy after they have experienced flooded (hypoxic) conditions (Watanabe and Miyahara, Reference Watanabe and Miyahara1989; Baskin et al., Reference Baskin, Baskin and Chester1993, Reference Baskin, Baskin and Chester2000a, Reference Baskin, Baskin and Chester2002a,Reference Baskin, Milberg, Andersson and Baskinb). This finding is in agreement with the results from previous studies that lack of oxygen can sometimes break dormancy even in terrestrial species, especially if their seeds require cold stratification to overcome dormancy (Come et al., Reference Come, Corbineau, Soudain, Jackson, Davies and Lambers1991).
In contrast to the response of H. humifusum and L. hyssopifolia seeds to hypoxic conditions, non-dormant seeds of L. aquatica germinated only in a narrow range of environments of light, aerobic and fluctuating temperatures (Phartyal et al., Reference Phartyal, Rosbakh, Ritz and Poschlod2020). Seeds had a higher percentage, rate and synchronization of germination when moved from hypoxic to aerobic conditions than if they were exposed directly to aerobic conditions. These results justify the classification of L. aquatica as ‘true mudflat species’ (Phartyal et al., Reference Phartyal, Rosbakh, Ritz and Poschlod2020) and its phytosociological classification into ‘Nano-Cyperetea’ (Oberdorfer, Reference Oberdorfer2001). A similar effect of prior exposure to hypoxic conditions was demonstrated in seeds of P. portula and S. valerandi. This reveals that a flooding pretreatment may help seeds to germinate as fast as possible immediately after water drawdown, which would allow seedlings to take full advantage of the short growing period in an unpredictable and unstable environment, which is the case of mudflat habitats. Prior exposure of seeds to flooding events was also reported to promote germination in other wetland species like Sphaeranthus indicus (Shetty, Reference Shetty1967) and Panicum laxum (Cole, Reference Cole1977). In a recent study on dormant seeds of Echinochloa crus-galli, Peralta Ogorek et al. (Reference Peralta Ogorek, Striker and Mollard2019) reported that hypoxic conditions weaken primary dormancy as well as hindered induction of secondary dormancy at warm temperatures. They also reported that dormancy-breaking signals are overridden during hypoxic conditions, which help to prevent seed germination when submerged in water.
The majority of mudflat species produce non-dormant seeds and show a narrow germination niche width (Phartyal et al., Reference Phartyal, Rosbakh, Ritz and Poschlod2020) as compared to the germination niche of other wetlands species (Rosbakh et al., Reference Rosbakh, Phartyal and Poschlod2020). Thus, low seedling mortality and optimize establishment may be promoted if seed germination is only triggered by high oxygen concentrations that are typical for exposed soils or sediment surfaces after water drawdown (Coops and van der Velde, Reference Coops and van der Velde1995). We conclude that in view of the very short favourable growing period for mudflat species the promotive effect of flooding (hypoxic condition) on percentage, speed and synchronization of germination when the floods ends is adaptive in the ephemeral mudflat habitat. This is another crucial role hypoxic environment play, apart from seeds to persist in submerged sediments (Poschlod and Rosbakh, Reference Poschlod and Rosbakh2018), to control germination traits of mudflat species.
Acknowledgements
We thank the students, who helped with the seed collection and germination experiment. S.S.P. was supported by the Alexander von Humboldt Foundation, Germany, for Humboldt Experienced Researcher Fellowship. The comments of two anonymous reviewers helped to improve the earlier version of the manuscript.
