Introduction
Northern wild rice (NWR; Zizania palustris L.), also referred to as wild rice or manoomin, is an annual, outcrossing, aquatic species in the Poaceae family, native to North America (Hayes et al., Reference Hayes, Stucker and Wandrey1989). While NWR has been harvested from lakes and rivers in the Great Lakes region for centuries, commercial cultivation began in 1950 and breeding efforts in the 1960s (Oelke and Porter, Reference Oelke, Porter, Wrigley, Corke, Seetharaman and Faubion2016). Today, cultivated NWR is primarily grown in Minnesota and California. Breeding objectives have primarily focused on domestication traits such as shattering resistance and uniform plant maturity (Grombacher et al., Reference Grombacher, Porter and Everett1997). However, due to the unique seed characteristics of this high-value, small commodity crop, adequate storage protocols to maintain seed viability and vigour are essential to breeding efforts (Kennard et al., Reference Kennard, Phillips, Porter, Grombacher and Phillips1999; Porter, Reference Porter, Greene, Williams, Khoury, Kantar and Marek2019).
Within the Poaceae family, 98.6% of species are classified as orthodox or desiccation tolerant, while only 0.8 and 0.6% are considered recalcitrant and intermediate, respectively (Dickie and Pritchard, Reference Dickie, Pritchard, Black and Pritchard2002). NWR is generally considered recalcitrant (Pence, Reference Pence and Bajaj1995). However, some researchers oppose the idea that desiccation tolerance is strictly categorical, instead favouring a sliding-scale view (Pammenter and Berjak, Reference Pammenter and Berjak1999). The degree of desiccation tolerance in NWR is affected by storage temperature, rate of drying and degree of metabolic activity (Probert and Longley, Reference Probert and Longley1989; Kovach and Bradford, Reference Kovach and Bradford1992a; Berjak and Pammenter, Reference Berjak and Pammenter2008). Kovach and Bradford (Reference Kovach and Bradford1992a) found that NWR seeds could survive desiccation under specific drying conditions with embryonic moisture contents as low as 6%. Similar phenomena have been observed in coffee (Coffea arabica L.) and papaya (Carica papaya L.) (Ellis, Reference Ellis1991), where the term intermediate seed storage behaviour has been used.
Seed dormancy is the absence of germination under normally favourable environmental conditions, such as water, oxygen, temperature and light (Hilhorst, Reference Hilhorst1995; Kermode, Reference Kermode2011). Seed dormancy is only found in a small percentage of species with recalcitrant and intermediate storage behaviour and a clear understanding of the physiology underlying the cessation of dormancy in these seeds is lacking (Tweddle et al., Reference Tweddle, Dickie, Baskin and Baskin2003). NWR has a dormancy period of at least 3–6 months (Cardwell et al., Reference Cardwell, Oelke and Elliott1978), which can be affected by temperature, seed pericarp stability and growth hormones, such as abscisic (ABA) and gibberllic (GA) acids (Grombacher et al., Reference Grombacher, Porter and Everett1997). Optimal conditions for germination change over the course of dormancy. Diurnal temperature fluctuations are most important at earlier stages and higher mean daily temperatures at later stages (Atkins et al., Reference Atkins, Thomas and Stewart1987). While studies have provided important information regarding the stratification of NWR seed, specific parameters for the temperature and duration of stratification have not been well established.
Generally, NWR seeds are stored on water between 1 and 3°C (Grombacher et al., Reference Grombacher, Porter and Everett1997; Oelke and Porter, Reference Oelke, Porter, Wrigley, Corke, Seetharaman and Faubion2016). However, wet, submerged storage is not a viable long-term storage option for NWR, in part due to seed deterioration caused by microbial growth (Berjak and Pammenter, Reference Berjak and Pammenter2008) and the high germination potential of seed once stratified at these cool temperatures (Simpson, Reference Simpson1966). Orthodox germplasm of various kinds can be preserved at very low temperatures, with little to no resulting damage (Bajaj, Reference Bajaj1995). In species with recalcitrant seeds, it can be much more challenging to utilize freezing storage due to the potential formation of intracellular ice crystals, particularly within the embryo tissue (Pammenter et al., Reference Pammenter, Berjak and Herendeen2014). However, several studies have evaluated the potential of storing NWR in freezing temperatures with varying success rates (Duvel, Reference Duvel1906; Fyles, Reference Fyles1920; Moyle and Krueger, Reference Moyle and Krueger1964; Simpson, Reference Simpson1966; Oelke and Standwood, Reference Oelke and Standwood1988). Additionally, it has been successfully demonstrated that NWR can be stored at freezing temperatures (≤−10°C), but that a cold storage period (<10°C) is still needed to break seed dormancy (Kovach and Bradford, Reference Kovach and Bradford1992a). Therefore, it may be possible to extend NWR seed viability in storage at freezing temperatures by maintaining seed dormancy.
It has been well documented that seed maturity and viability can be significantly affected by the maternal environment, specifically temperature, photoperiod, nutrient availability and soil conditions (Luzuriaga et al., Reference Luzuriaga, Escudero and Pérez-García2006; Postma and Ågren, Reference cPostma and Ågren2015; Penfield and MacGregor, Reference Penfield and MacGregor2016; Li et al., Reference Li, Song, Yao, Chai, Simpson, Li and Nan2017). The most recent studies evaluating storage of NWR seed at freezing temperatures (Kovach and Bradford, Reference Kovach and Bradford1992a, Reference Kovach and Bradfordb) were conducted with California-grown seed, which is anecdotally known to be more physiologically mature than Minnesota-grown seed due to longer growing seasons (personal communication). Growing regions in California, specifically the Sacramento River Valley (16°C average annual temperature, elevation 91 m), and Northern Minnesota (4.63°C average annual temperature, elevation 392 m) are vastly different. Therefore, in this study, alternatives to standard submerged storage conditions were evaluated for NWR seed harvested from Minnesota paddies. In particular, an emphasis was placed on non-submerged seed storage in freezing conditions, a method which has not been evaluated sufficiently in NWR. Germination tests were conducted to estimate the viability of seeds following three experiments designed to test varying storage conditions and their effect on dormancy. Multiple varieties and experimental breeding lines adapted to Minnesota growing conditions were evaluated to assess potential genetic variability in the response to storage conditions, providing a novel aspect to this work in NWR.
Materials and methods
Plant materials and storage condition
A total of ten genotypes representing the range of phenotypic diversity present in the University of Minnesota (UMN) cultivated NWR breeding programme were selected for this study (Table 1). Genotypes were open pollinated and seeds were harvested in September 2012 and 2013 from research plots at the UMN North Central Research and Outreach Center (NCROC), in Grand Rapids, MN. For all experiments, seeds were stored in sealed heavy plastic bags after harvest and processing. For non-submerged, freezing storage (NSFS), the seed was placed in bags without water and stored at −3.5°C in a Crown Tonka freezer (Plymouth, MN) in the dark. The freezer was maintained at approximately 80% relative humidity for the duration of the experiments. For submerged, cold storage (SCS), the seed was placed in bags with water and stored at 3°C in a Vollrath cooler (Sheboygan, WI). Temperature and relative humidity were monitored using a Hobo temperature/relative humidity data logger (Onset, Bourne, MA) in the freezer.
Table 1. Eleven genotypes of cultivated NWR used in NSFS and SCS experiments along with their initial SMC prior to storage
Seed moisture content
In 2012 and 2013, seed moisture content (SMC) for all genotypes was assessed following harvest and processing (Table 1). A 5-g green sample of each genotype was dried for 2 days in a drying oven set to 140°C. Samples were then re-weighed to determine the dry weight. SMC was calculated on a fresh-weight basis using the following equation:
$${\rm SMC} = \left({\displaystyle{{{\rm green}\,{\rm weight}-{\rm dry}\,{\rm weight}} \over {{\rm green}\,{\rm weight}}}} \right)\times 100$$NSFS experiment
In 2012, seeds were stored in NSFS conditions for between 12 and 26 weeks. Starting at 12 weeks, germination tests were carried out once every 2 weeks for a total of eight germination tests. Due to a lack of seed germination during testing, the same seeds from the NSFS tests were placed in SCS conditions for an additional 2 weeks, in an attempt to stratify the seed. A second round of germination testing was then performed. Three replications of 50 seeds per genotype were evaluated.
SCS stratification experiment
Once it was established that Minnesota-grown NWR could tolerate NSFS conditions, an experiment was conducted in 2013 to evaluate the optimal SCS period to maximize germination in NWR following freezing storage. Subsequent to an initial 12-week period in NSFS conditions, samples were moved and stored in SCS for a range of 2–14 weeks. Three replications of 50 seeds per genotype were evaluated. Germination tests were conducted every 2 weeks for a total of seven tests. Additionally, the rate at which each genotype reached 50 and 75% of its maximum germination percentage (MGP) (G50 and G75, respectively) was calculated using the MASS package in Rstudio (Venebles and Ripley, Reference Venebles and Ripley2002). Significance within G50 and G75 values was determined using calculations of the 95% confidence interval.
Storage and stratification timing experiment
Next, an experiment to evaluate the differences in the germination of seed stored at varying lengths of time in NSFS and SCS was conducted in 2013. Five genotypes were selected for this experiment (Table 1). Seeds were stored for a total of 29 weeks, representing the time between harvest and planting, in five different NSFS, SCS stratification treatment combinations as follows: (1) 29 weeks NSFS and 0 weeks SCS, (2) 25 weeks NSFS followed by 4 weeks SCS, (3) 22 weeks NSFS followed by 7 weeks SCS, (4) 17 weeks NSFS followed by 12 weeks SCS and (5) 0 weeks NSFS and 29 weeks SCS. Four replications of 50 seeds were evaluated for each genotype by storage treatment combination. Following treatments, germination testing was conducted.
Germination testing
Before germination testing, genotypes were removed from NSFS or SCS conditions. For seeds in NSFS, each bag was filled with enough water to cover the seeds (~100 ml dH2O). For seeds in SCS, water from each bag was drained and replaced. For germination testing, seed bags were placed in an I-35L growth chamber (Percival, Perry, IA) set to 17 h days at 17.5°C and 7 h nights at 8.5°C, for 14 days. Temperature was monitored using a hobo data logger (Onset, Bourne, MA). The per cent germination (%G) was then calculated. Germination was defined as coleoptile emergence equal to the length of the seed.
Data analysis
Data analysis was conducted using the Rstudio software, version 1.1.423 (Rstudio-team, 2016). Prior to analysis, the data were assessed in Rstudio using standard residual diagnostic plots, including residuals versus fitted, normal Q-Q, scale-location and residuals versus leverage. An analysis of variance (ANOVA) was used to assess differences in %G between treatments, genotypes and storage time points. The breakdown of differences was then analysed using a post hoc Tukey LSD test, calculated using the Agricolae package in Rstudio (Mendiburu, Reference Mendiburu2019). For the NSFS experiment, NSFS and SCS treatments were first analysed together to compare the germination responses between the two treatments followed by the evaluation of SCS treatments alone. The correlation between SMC and germination values for all experiments was assessed using Pearson correlation coefficients.
Results and discussion
NSFS experiment
Initial testing Demonstrated that the %G of seeds stored in NSFS conditions alone was very low (0–7%) across the 12–26-week testing period (Fig. 1a). No significant differences between genotypes within and among weeks were identified (data not shown). Significant differences in %G were identified between the testing of seeds initially stored in NSFS and the same seeds following SCS (storage conditions), the time in NSFS (storage weeks), genotype and all the interactions of these variables (Table 2). Specifically, %G increased following the short SCS stratification period at 3°C and continued to improve the longer the seeds were stored in NSFS conditions (Fig. 1a). These results confirmed that storage in water at cold temperatures is necessary for NWR stratification following NSFS conditions, confirming Kovach and Bradford's (Reference Kovach and Bradford1992a) results. More interestingly, the increase of %G across the 12–26-week testing period indicated that while NSFS conditions did not break seed dormancy, physiological changes were still occurring within the NWR seed over time and appear to be related to stratification processes. Because the relative humidity in NSFS was ~80%, and the temperature just slightly below freezing, it is possible that some respiration occurred during this period, potentially explaining the continued physiological changes in the seed. Recalcitrant seeds are known to be metabolically active following shedding from the mother plant, and some studies have even shown an increase in germination following partial drying in other recalcitrant species (Pammenter and Berjak, Reference Pammenter and Berjak1999; Walters et al., Reference Walters, Pammenter, Berjak and Crane2001; Bai et al., Reference Bai, Yang, Tian, Chen, Shi, Yang and Hu2011; Umarani et al., Reference Umarani, Aadhavan and Faisal2015). The metabolic activity of MN NWR seed may be slowed but not halted by the parameters used in this study (Pammenter and Berjak, Reference Pammenter and Berjak1999). It has been suggested that recalcitrant seed should be stored at the lowest non-damaging temperature, which this study and previous indicate is lower than −3.5°C (Kovach and Bradford, Reference Kovach and Bradford1992a; Berjak and Pammenter, Reference Berjak and Pammenter2008). More research is needed to determine if NWR seed dormancy is truly maintained during longer periods of NSFS and, therefore, is a viable option for prolonging seed viability in storage.
Fig. 1. (a) Average per cent germination (%G) across ten cultivated NWR genotypes after 12–26 weeks in NSFS before and after 2 weeks of stratification in SCS. (b) Per cent germination (%G) for individual genotypes after SCS stratification, post-NSFS throughout weeks. Maximum %G across all weeks is included for each genotype.
Table 2. Analysis of variance (ANOVA) for three storage and stratification experiments conducted in cultivated NWR
Abbreviations: d.f., degrees of freedom; Sum Sq, sums of squares; Mean Sq, mean squares.
Pre-storage SMC was not significantly correlated with %G in this experiment (data not shown) indicating that the SMC of NWR used in this study was not associated with the maintenance of seed viability during storage. While previous studies have identified a relationship between SMC and viability in NWR (Probert and Longley, Reference Probert and Longley1989), the limited range of SMC variability in this study may account for the lack of a detectable relationship (Table 1). Often in recalcitrant seeds, a loss of viability is associated with moisture loss, in part due to structural damage often associated with the drying rate (Berjak and Pammenter, Reference Berjak and Pammenter2008). While SMC data were not collected post-storage, the significant increase in %G following SCS indicates that seed viability or a reduction in SMC was not determining factors of the low %G of seeds in NSFS conditions. Additionally, the high relative humidity in the NSFS freezer suggests that the SMC most likely would not have been reduced during NSFS. However, more testing is needed to confirm this hypothesis. Additionally, variation in the response of other recalcitrant species to non-submerged conditions has been reported (Farrant et al., Reference Farrant, Pammenter and Berjak1989; Probert and Longley, Reference Probert and Longley1989; Berjak and Pammenter, Reference Berjak and Pammenter2008), which has drawn skepticism regarding the usefulness of defining a critical water content for the viability of recalcitrant seeds (Berjak and Pammenter, Reference Berjak and Pammenter2001).
The short SCS period following the initial NSFS period resulted in differences in %G among genotypes both within and across weeks (Table 2). Within weeks, variability between genotypes increased the longer the seeds were in storage. Across weeks, the most significant genotypic differences were between the first and last testing dates, or 12 and 26 weeks in NSFS, respectively. The pattern of %G across weeks varied by genotype, with peaks and dips occurring at differing time points (Fig. 1b). The genotypes 14S-C5 and K2EF-C11 had the highest %G across all weeks, with the exception of low K2EF-C11 germination in week 8 of testing. The genotypes Franklin and PM3E-C19 had the lowest %G across weeks. These differences indicate genotypic variability within NWR's response to NSFS conditions. For example, 14S-C5 is derived from a hybrid between Zizania aquatica L., which lacks dormancy, and Z. palustris. As such, the genotype may have retained some reduced dormancy, explaining its tendency to germinate earlier and more vigorously than other varieties. Barron and K2EF-C11 were selected for early maturity, possibly allowing these genotypes to reach physiological maturity more quickly during the growing season. The likelihood that genetic variability contributed to higher %G indicates mapping and heritability studies to dissect the genetic mechanisms underlying specific traits related to NSFS in NWR may be possible.
SCS stratification experiment
Following an initial 12-week storage period in NSFS, seeds were placed in SCS, with germination tests occurring every 2 weeks, for a total of 14 weeks, to determine the optimal duration of SCS between growing seasons. Significant differences were found between genotypes, duration of SCS (storage weeks) and the interaction between the two variables (Table 2). The %G of all genotypes trended higher as the weeks of SCS increased; however, there was genotypic variation in the rate of increase across weeks (Fig. 2a). Overall, the largest increase in %G occurred between 4 and 6 weeks of storage. The variation between genotypes was also greatest during these periods. Averaged across weeks of SCS, 14S-C5 had the highest %G (64%), followed closely by Barron (60%). Franklin had the lowest %G overall at 41%. The G50 values for all genotypes ranged from 4.6 to 8.3 weeks in SCS storage and G75 from 8.2 to 11 weeks in SCS storage (Fig. 2b). While no significant difference in the G50 or G75 values between genotypes was identified, K2EF-C11 reached its G50 value most quickly, consistent with its classification as an early maturing variety. The lack of significant differences may indicate that while there is variation in the rate of germination between genotypes, there is also variation within genotypes, as a result of heterozygosity, that may overshadow intergenotypic differences. The required period of SCS stratification shown in this study was shorter than the 17–26-week dormancy period typically seen for NWR. These results are consistent with those found in the NSFS experiment, suggesting that physiological changes are taking place in NSFS, which may be related to stratification processes.
Fig. 2. Per cent germination (%G) of ten cultivated NWR genotypes (a) plotted over 2–14 weeks of stratification in SCS following 12 weeks in NSFS. (b) Maximum germination potential represented as G50 and G75 for each genotype across 14 weeks of stratification testing in SCS. Error bars are 95% confidence interval.
Pre-storage SMC was not found to be significantly correlated with %G (data not shown). This result is perhaps unsurprising as seeds were maintained at an SMC no lower than 20%, which is within the average range required for the maintenance of recalcitrant seed viability (Roberts, Reference Roberts1973; Umarani et al., Reference Umarani, Aadhavan and Faisal2015).
Storage and stratification timing experiment
This experiment was designed to determine an optimal NSFS:SCS timing combination to maximize the germination potential of NWR, on an annual basis, between growing seasons. Seeds in this experiment were stored for a total of 29 weeks, split between varying times in NSFS and SCS conditions. Significant differences in %G between genotypes, storage timing ratios and the interaction between the two variables were identified (Table 2). However, the storage timing ratios accounted for the majority of variation identified in this experiment. Averaged across all genotypes, a general trend emerged where %G increased as the number of weeks in NSFS decreased and time in SCS increased (0.5%G at 29 DSFS: 0 SCS weeks, 78.5% at 0 DSFS: 29 SCS weeks). However, the differences in timing plateaued after 7 weeks in SCS, with no significant differences in %G between 7 and 29 weeks of SCS (Fig. 3; 75.6%G at 22 DSFS: 7 SCS weeks, 80.6%G at 0 DSFS: 29 SCS weeks). Similar to results in the NSFS experiment, these results demonstrate that dormancy could be overcome after NSFS with a short period in SCS. Therefore, stratification-associated physiological processes may be occurring during NSFS at the temperature and relative humidity parameters used in these studies. In Spartina patens, a salt march-inhabiting perennial grass, germination peaked at 8 weeks of wet pre-chilling and no significant increase in germination was identified thereafter (Plyler and Proseus, Reference Plyler and Proseus1996).
Fig. 3. Per cent germination (%G) of five cultivated NWR genotypes after five different ratios of weeks in dry, freezing storage (NSFS) and weeks in SCS. Significance across different storage ratios was determined by Tukey's LSD test.
Experiments in Papaver rhoeas and Papaver dubium similarly showed maximum germination after ~6 weeks of wet pre-chilling (Golmohammadzadeh et al., Reference Golmohammadzadeh, Zaefarian and Rezvani2015). In Fagus sylvatica, the release of dormancy under wet pre-chilling conditions was shown to be associated with the decline of the glycine-rich protein GRPF1, which is upregulated by ABA (Nicolás et al., Reference Nicolás, Rodríguez, Poulsen, Eriksen and Nicolás1997; Mortensen et al., Reference Mortensen, Rodríguez, Nicolás, Eriksen and Nicolás2004). Because ABA controls dozens of genes, further research on the proteome of dormant and non-dormant NWR seed may elucidate similar proteins involved in the release of dormancy during SCS (Eggers et al., Reference Eggers, Erdey, Pammenter, Berjak, Adkins, Ashmore and Navie2007).
When averaged across timing treatments, 14PD-C5 and 14S-C5 had the highest %G (55.3 and 55.0%), while PBM-C18 had the lowest (45.9%) (Fig. 3). The differences between genotypes, averaged across weeks of storage, were small but statistically significant (Table 2). However, Tukey's LSD test was unable to detect differences between genotypes (data not shown). There were differences in the interaction between genotype and timing ratios, but differences in genotypic performance could not be identified until SCS was reduced to 4 weeks. Interestingly, while 14PD-C5 and 14S-C5 had the highest overall %G, when stored solely in NSFS conditions, they performed equally with PBM-C18, which had the lowest overall %G. Like 14S-C5, 14PD-C5 was also derived from a hybrid of Z. aquatica × Z. palustris, which although not tested at this time, may have a reduced dormancy period. The incrementally higher %G of these early or less dormant varieties may be an indication that selection could be made for improved post-storage germination.
Conclusion
This study demonstrated the feasibility of storing MN grown NWR seeds in non-submerged, frozen storage for up to 29 weeks, offering new storage possibilities for researchers and conservationists. While long-term storage in NSFS may slow seed deterioration, the results of these studies indicated that physiological changes were occurring in NWR seed during NSFS. Therefore, the maintenance of NWR seed dormancy and viability during longer-term NSFS, as well as at lower storage temperatures and relative humidity than used in these studies, still need to be evaluated. While not measured in this study, microbial growth was limited but still present in NSFS conditions by comparison to SCS conditions and could be further evaluated. For storage of NWR seeds between growing seasons, we recommend at least 7 weeks of SCS to fully imbibe the seeds and break dormancy, to maximize germination potential following NSFS. Additionally, variation among genotypes indicates that selection could be implemented to improve the germination potential and storability of NWR.
Acknowledgements
We acknowledge Erv Oelke and Christina Walters for their earlier work on Zizania seed storage that laid important groundwork and inspired our continuing efforts. Thanks to Andy David and the Forest Biology project at NCROC for providing cold room space.
Financial Support
This work was supported by the Minnesota Cultivated Wild Rice Council (R.P. grant numbers 34434, 40630) and the University of Minnesota's Rapid Agricultural Response Fund (R.P. grant number AES00RR193).
