Common buckwheat (Fagopyrum esculentum
Moench) is a valuable plant of Polygonaceae
family, grown mainly for human food due to favorable chemical composition of its fruits (commonly termed seeds), especially substantial content of lysine and other amino acids. Seeds are also gluten free [1
]. Nectar of common buckwheat is a source of valued honey, while grain and straw are used as animal feed. The presence of two morphs of flower, Pin and Thrum, is a manifestation of heteromorphic self-incompatibility (heterostyly). Both types of flowers exhibit differences in pistil and stamen lengths [2
]. Fertilization occurs only following cross-pollination between two different types of flowers.
Factors such as low resistance to excessive water, lodging, and pre-harvest sprouting occurring before flowering affect buckwheat yield, which is low and variable [3
Plant yield is also affected by the short life span (1–2 days) of individual flowers and disturbances in female but not in male line development. The number of degenerated ovule sacs in plants is high and ranges from 10% to 30%, depending on the strain and cultivar [4
]. Additionally, buckwheat flowering is sensitive to many environmental stresses, including frost, heat, and drought occurring in the spring and summer despite stronger vegetative growth at 30 °C than at 20 °C [5
]. A drastic reduction in the number of properly developed embryo sacs was clear in open flowers at 30 °C in cultivars ‘Panda’ and ‘Korona’ (= strain PA15) [5
]. Similarly, sensitivity to the thermal stress was shown by Slawinska and Obendorf [1
]; plants grown at 18 °C had by 40% increased seed set. They set seeds over a longer period and produced by 40% more dry mass per seed than plants grown at 25 °C. Several authors [7
] described ultrastructural changes in reproductive cells under thermal stress and suggested that premature synergid collapse may prevent a pollen tube from reaching the embryo sac. It was recently shown that radiation restriction resulting from plant growth could have increased floret mortality and thereby decrease the number of achenes (fruits) per raceme (type of inflorescence in common buckwheat) [10
]. High temperature and other environmental factors cause premature flower and immature seed fall. A considerable increase in abscisic acid in open flowers ready for fertilization under thermal stress may serve as a signal inducing flower senescence observed in the next few days [6
]. The results of a project carried out during 2014–2016 (‘Increasing the vitality and degree of pollination of buckwheat flowers in order to obtain a higher seed yield’), concerning flowering and yielding biology of Polish buckwheat accessions, indicate that the more flowers a plant produces, the greater their abortion, and the number of flowers negatively correlates with the number of mature seeds. Therefore, increasing the number of flowers per plant may not improve the yield It is, however, opposite to a selection index, which was constructed based on seven agro-morphological traits measurable in a single plant, showing that seed yield is positively correlated with the number of flower clusters in common buckwheat [11
The frequency of aborted embryos in controlled conditions after hand-pollination is rather low (<10%) [12
] or moderate (9.6–19.1%) [2
], depending on the genotype, thus it seems that selective embryo abortion resulting from better or worse embryo fitness is not the case of low yield in common buckwheat. Halbrecq et al. [13
] and Cawoy et al. [2
] suggested that embryo abortion in common buckwheat is fixed by an internal mechanism at a relatively constant level and is not the result of insufficient nutrient supply from leaves. In contrast, studies in four Polish cultivars based on observation of embryo development with free access of pollinators showed that more, up to 28%, of embryos could have degenerated, some of them showing a typical hallmark of starvation [4
]. This is in accordance with earlier observations of Inoue and Hagiwara [14
] and Inoue et al. [15
], who proposed that the percentages of flower fertilization and seed set are influenced by the degree of competition for nutrients between source and sink organs. Buckwheat blooms throughout the whole growing season, which may result in a strong competition for assimilates between the already set seeds and flowers still in production. To resolve these inconsistencies in the above-mentioned results, we analyzed various reproductive and yield parameters and the hormone profile of the flowers of cv. ‘Panda’ and ‘Korona’ of F. esculentum
under in vitro and in planta conditions. Our aim was to investigate whether the seed limitation is associated with a strong sink restriction and linked to internal flower structure and fertilization. We hypothesized that plants with a shorter flowering period may give a higher seed yield.
Due to the challenges of common buckwheat cultivation described in the Introduction, especially regarding its low yield, breeding and genetic studies have been carried out for decades (for review, see Matsui and Yasui [16
]). Genomic Selection in Mass Selection Breeding program for common buckwheat is a powerful program enhancing buckwheat yield by almost 21% [11
]. However, classical breeding treatments such as the one described in this study shed a light on the mechanisms involved in seed production of F. esculentum
. Key genes involved in seed development are already recognized, and they are genes responsible for Ca2+
signal transduction pathway, hormone signal transduction pathways, and coding transcription factors (TFs), as well as starch biosynthesis-related genes [17
]. Regarding seed size, AP2 and bZIP transcription factors, BR-signal, and ABA are considered the most important regulators [18
Our in vitro experiment showed that depriving flowers of nutrients leads to deterioration of their quality and to the abortion of ovule sacs, therefore we decided to start reducing the number of inflorescences in common buckwheat. We were also inspired for the further studies by the never-ending flower overproduction of common buckwheat throughout the whole season.
In our in planta experiment, we partly followed the suggestions of Guglielmini et al. [10
], not knowing their outcome while conducting our study in 2019. They recommended to determine the causes of reduction in the number of achenes per raceme (= fruits per inflorescence) during the critical period (the period when grain number is determined and it is crucial to obtain higher yields), as they showed that radiation restriction and subsequent assimilate limitation could increase floret mortality and thereby the number of achenes per raceme.
Yabe et al. [11
], who used 92FE1-F4, a population produced by bulk crossing of ‘Tempest’, ‘Kitawasesoba’, ‘Natsusoba’, and ‘Shinanonatsusoba’ cultivars, showed that the number of clusters positively correlated with seed yield. We did not find such a correlation in any of the investigated cultivars. On the contrary, we demonstrated a positive correlation between flower production and the abortion of flowers and fruits.
‘Korona’ plants after removal of 50% of flowers showed significantly lower number of degenerated embryo sacs and higher number of mature seeds, higher efficiency of seed setting than the control plants and plants from the other variants. Since we detected more IAA in these flowers, it is possible that the auxin supports development of embryos, which could be important for higher seed yield in these plants. Although only 3% of degenerated embryo sacs were found, the percentage of aborted flowers did not drop. This may be due to limited pollination or impaired embryo development, which in cv. ‘Panda’ flowers were at 9% and 13% depending on the type of flower (Pin or Thrum) [4
]. Despite higher percentage of abnormally formed embryo sacs in cv. ‘Korona’, higher effectiveness of seed setting, calculated as the number of seeds divided by the number of flowers, was found after removing 75% of flowers vs. 50% of flowers. ‘Panda’ plants with 75% flowers removed also showed a much higher percentage of degenerated embryo sacs than the plants treated differently, but the seed setting efficiency was similar to plants with 50% of flowers removed. It is possible that this effect was caused by a higher concentration of gibberellins and that these hormones are more effective than auxin in keeping the embryos alive, similarly as in Arabidopsis thaliana
Salicylic acid plays a crucial role in flowering and luring insects, similarly to jasmonic acid. Moreover, both hormones are involved in defense responses of plants in the event of a pathogen attack [20
]. Jasmonic acid is also necessary for the formation of an ovulum, and its absence is characteristic of sterile flowers [21
]. Such hormones as gibberellins, brassinosteroids, and abscisic acid are of great importance during flowering. It is usually believed that ABA, as an antagonist of gibberellins, inhibits flowering [22
]. It was therefore unexpected that the flowers that set seeds the most efficiently accumulated higher levels of this hormone than those of plants with a greater degree of embryo sac degeneration. GA to ABA ratio in plants from which 50% of flowers were removed was much lower than in the control plants. These results mean that the proportion between gibberellins and abscisic acid is more important than absolute concentrations of these hormones. On the other hand, in the flowers of control plants, this ratio was much higher, proving a significant predominance of gibberellins over ABA, and at the same time these plants were characterized by a greater degree of ovule sac degeneration. These findings thus indicate that a direct reason for embryo sac abnormal development is probably independent of these two classes of hormones.
Salicylic acid controls, similar to JA, the process of attracting insects during the flowering phase. However, while JA confers flowers their attractive fragrance, SA can increase flower temperature to release volatile compounds [23
]. Benzoic acid (BA) is a precursor of salicylic acid, so it was not surprising that BA concentration in flowers was high and translated directly into high content of SA. Our results indicate that the two cultivars studied also differed in terms of producing the hormones responsible for luring insects. ‘Panda’ flowers accumulated much greater amounts of salicylic acid, while ‘Korona’ ones produced mainly jasmonic acid. It should also be remembered that majority of hormones stimulate flower formation in the vegetative phase, so the presence of hormones in flowers may not give a complete picture of the role of individual hormones in embryological development.
Taylor and Obendorf [12
] argued that poor seed yielding in common buckwheat results from problems with ovule development and fertilization. They also claimed that the lack of fertilization is influenced by variable viability and quality of microspores. Our research shows that pollen viability in both studied genotypes is high, and even enhanced by flower removal in cv. ‘Panda’. This research further confirmed our long-term observations of a strong positive correlation between the number of flowers and their abortion, which means that the more flowers a plant produces, the more of them are rejected. According to ourresults, flower abortion is primarily influenced by plant genotype, and this barrier would be very difficult to break. In our opinion, buckwheat plants have a certain limit of seeds, and, above this limit, the plant will not allocate its ‘resources’ (assimilates) to their filling. A very high negative correlation between flower abortion and efficiency of seed setting was found, which proves that a decrease in assimilate competition significantly increases seed setting. As we noticed, the plant seems to be reaching a certain limit of flower production, and it is difficult to change this limit. Halbrecq et al. [13
] reported that the abortion mainly affects flowers from the upper floors, that is, those arising in the later stage of flowering, in relation to the flowers located lower, i.e., previously produced. Sugawara [24
] and Asako et al. [25
] found that flowers formed earlier are more likely to set seeds than flowers formed later. Our previous unpublished research showed that the nectar is richer in sugars in earlier flowers than in later ones. This factor likely has a significant impact on the flight of pollinating insects, which directly translates into seed setting. In addition, Taylor and Obendorf [12
] reported that the embryo sacs are better developed in the earlier formed flowers. Our observations indicate that flower abortion occurs throughout the flowering period and its degree depends on many factors. In buckwheat, the period of vegetative development overlaps with the period of generative organ formation. The formation of vegetative organs coincides with intense flowering and seed formation, which results in strong competition for assimilates. According to Halbrecq et al. [13
], when plants reach their maximum vegetative development, competition for assimilates decreases. Assimilates can therefore be located mainly in the generative organs. These authors claimed that more flowers and seeds are produced during this time, and the degree of abortion decreases along with competition for assimilates. Our research did not confirm these conclusions. We proved that the increased number of flowers did not correlate with the number of seeds, because at the same time the abortion of flowers increased. In previous studies, we observed a certain percentage of ‘starved’ embryos, so the competition for assimilates between the set seeds and the still emerging flowers continued [4
]. The present experiments confirmed a negative correlation between the number of empty seeds and mass of a single seed.
Halbrecq et al. [13
] performed an experiment involving defoliation (partial and total) and partial removal of buckwheat inflorescences from the main shoot in order to modify the availability of assimilates and reduce competition between seeds. In all cases, regardless of the procedure, they observed a drastic reduction in the number of grains in relation to the number of flowers formed, with very low seed yield, around 20–30%. The authors found that the critical seed setting stage occurs shortly after flowering and is not affected by a change in donor-acceptor relationship due to defoliation or removal of part of the inflorescences. The drastic limitation of competition between inflorescences and seeds had only a negligible effect on the final grain yield per plant, which indicated a strong compensation by the remaining grains. In our research, we showed that only the removal of 50% of flowers allowed for increasing the seed yield. In the remaining cases, removing flowers reduced the yield or did not change it as compared with the control. Despite the fact that the percentage of empty seeds did not significantly correlate with the number of whole seeds, it significantly negatively correlated with the weight of individual mature seed. These findings clearly indicate that the setting of seeds, although not filled later on, reduced the mass of mature seeds. It can therefore be assumed that flowering and fertilization are the phases critical for seed yield. Flower overproduction and embryo formation is a process that exhausts the plant reserves. Failure to fill all the seeds will no longer compensate for these losses. Forming empty seeds is therefore, apart from flower abortion, another form of crop regulation when the plant produced too many flowers.
It is worth underlining that plants cultivated as single-shoots finished their blooming period earlier and their seeds matured also earlier than those of the other plants. The study results confirm our hypothesis that plants with a shorter flowering period (self-finishing) may achieve higher seed yield. Considering that the number of flowers, their abortion, and the percent of defective embryos are controlled genetically, breeders are faced with a challenging task of producing new genotypes with amended traits. Other difficulties involve strong self-incompatibility and impossibility of inbreeding. Given these limitations, mutations seem to be the only way to obtain new forms of common buckwheat, and we will explore this approach in the years to come.