Specialized Metabolites Accumulation Pattern in Buckwheat Is Strongly Influenced by Accession Choice and Co-Existing Weeds

Screening suitable allelopathic crops and crop genotypes that are competitive with weeds can be a sustainable weed control strategy to reduce the massive use of herbicides. In this study, three accessions of common buckwheat Fagopyrum esculentum Moench. (Gema, Kora, and Eva) and one of Tartary buckwheat Fagopyrum tataricum Gaertn. (PI481671) were screened against the germination and growth of the herbicide-resistant weeds Lolium rigidum Gaud. and Portulaca oleracea L. The chemical profile of the four buckwheat accessions was characterised in their shoots, roots, and root exudates in order to know more about their ability to sustainably manage weeds and the relation of this ability with the polyphenol accumulation and exudation from buckwheat plants. Our results show that different buckwheat genotypes may have different capacities to produce and exude several types of specialized metabolites, which lead to a wide range of allelopathic and defence functions in the agroecosystem to sustainably manage the growing weeds in their vicinity. The ability of the different buckwheat accessions to suppress weeds was accession-dependent without differences between species, as the common (Eva, Gema, and Kora) and Tartary (PI481671) accessions did not show any species-dependent pattern in their ability to control the germination and growth of the target weeds. Finally, Gema appeared to be the most promising accession to be evaluated in organic farming due to its capacity to sustainably control target weeds while stimulating the root growth of buckwheat plants.


Introduction
Pests and weeds are probably the primary biotic limitation that farmers face when attempting to increase the yield of their crops. Therefore, research, development, and the use of chemical products exclusively produced for the control of weeds/pests in the field has increasingly grown in the last decades. However, these synthetic chemical products, which have been extensively applied in the field, are dangerous to terrestrial environment and to human health [1], and increase weed resistance [2]. Several synthetic herbicides, including glyphosate, exhibit significant potential for soil adsorption and cannot move around freely in the environment. The effects of herbicides on human health depend on the concentration, length, and frequency of exposure, and often lead to cytotoxic and DNA damage and carcinogenicity [3]. Reducing their use is an increasing necessity in order for more sustainable food production. Table 1. Allelopathic impact of different buckwheat genotypes (Gema, Kora, Eva, and PI481671) on the germination (Germ: number of seeds as % of the control), fresh plant weight (PW; g given as % of the control) (i.e., weed species grown alone), and shoot and root lengths (cm given as % of the control) in monocot (Lolium rigidum) and dicot (Portulaca oleracea) weeds. Grey cells indicate significant inhibition, while bold numbers indicate significant stimulations compared with the control. The asterisk indicates statistical significant differences compared to the control at p < 0.05.
Regarding the impact of the presence of buckwheat accessions on weed growth and development, we obtained a species dependent behaviour again, with L. rigidum as the most sensitive weed species to buckwheat co-cultivation ( Table 1). The most inhibited parameter was the fresh plant weight for L. rigidum in the presence of Kora (73%), PI481761 (68%), and Eva (52%), while the most stimulated parameter was the shoot length of P. oleracea in the presence of PI481761 (186%), Gema (158%) and Eva (146%). L. rigidum was especially sensitive to Eva, as shoot length significantly increased (149%) while reducing fresh plant weight (52%), which resulted in longer but weaker plants. Something similar was found for P. oleracea in front of Gema accession, which significantly increased both, shoot (158%) and root (166%) length while maintaining unaltered the total weight of the plant (107%).
Regarding the germination and growth parameters of the different buckwheat accessions in front of the two tested weeds (Table 2), our results show that the shoot and root lengths of Eva were significantly inhibited in front of P. oleracea, as well as the root length in front of L. rigidum. In fact, Eva was the only accession where the shoot and root length were statistically inhibited in the presence of the weeds, while Gema and PI481671 also showed significant stimulation of root length when co-cultured with P. oleracea. PI481671 showed a stimulation of leaf weight in front of P. oleracea when compared with the buckwheat plants growing alone. Finally, the root weight was only inhibited in Kora in the presence of the monocot weed L. rigidum. From the parameters measured in the different accessions, we could see also that common buckwheat accessions (Eva, Gema, and Kora) grew much more than Tartary buckwheat (PI481671), which showed shoot length values that were two times lower than common buckwheat and root lengths values that were even three times lower than Eva and Kora and two times lower than Gema. These differences were even stronger in the case of leaf weight, where Tartary buckwheat values were ten times under common buckwheat accessions. Table 2. Mean values and standard deviation of growth parameters for different buckwheat accessions (Gema, Kora, Eva, and PI481671) when growing alone or in co-culture with the monocot weed Lolium rigidum (LR) or the dicot weed Portulaca oleracea (PO). The parameters measured are plant and root weight (g), and shoot and root lengths (cm). Grey cells indicate significant inhibition, while bold numbers indicate significant stimulations compared with the control.

Polyphenols Profile in Different Buckwheat Accessions
The identification and quantification of different polyphenols (phenolic acids and flavonoids), such as 4-chlorobenzoic acid ( (Table 3).
From all of the polyphenols measured, DA, P-CA, LU, 4-HA, OR, and VIT were the most commonly accumulated compounds, especially in the roots, for most of the four buckwheat accessions in the presence of both weeds, L. rigidum and P. oleracea, while DA, LU, 4-HA, CAT, ECAT, and OR were the most altered compounds in the roots, shoots, and root exudates of the different buckwheat accessions (Table 3). Although in the presence of L. rigidum buckwheat tended to accumulate the compounds indistinctly in roots or shoots, depending of the buckwheat accession, there was a clear trend in the four buckwheat accessions to accumulate the compounds in the roots when co-grown with P. oleracea, as DA, P-CA, LU, OR, and VIT were significantly accumulated in the roots in three of the four buckwheat accessions tested when P. oleracea was present in the medium. On the contrary, L. rigidum was the weed that more root exudates induced from the buckwheat accessions tested, especially for Eva (Table 3). Finally, the Tartary buckwheat accession (PI481671) showed a generally reduced root exudation of polyphenols in comparison with the three common buckwheat accessions (Gema, Kora, and Eva), which reacted more to the presence of weeds exuding more or less phenolic acids and flavonoids, but also exuding more polyphenols when growing alone. Table 3. Significant increases and decreases (p ≤ 0.05) for different polyphenols (PP) in the shoots, roots, or root exudates of the buckwheat accessions Gema (G), Kora (K), Eva (E), and PI481671 (PI) after co-culture with L. rigidum (LR) or P. oleracea (PO) compared to the control (each accession growing alone). The PP detected were 4-chlorobenzoic acid (4-CBA), vanillic acid (VA), dihydroxybenzoic acid (DA), ferulic acid (FA), p-coumaric acid (P-CA), syringic acid (SA), luteolin (LU), syringaldehyde (SY), protocatechuic acid (PTA), quercetin (QE), rutin (RU), vanillin (VN), salicylic acid (SAA), 4hydroxyacetophenone (4-HA), catechin (CAT), epicatechin (ECAT), orientin (OR), vitexin (VIT), phthalic acid (PA), m-toluic acid (M-TA), and hypericin (HYP). During the comparison of the polyphenol content of shoots, roots, and root exudates of the four buckwheat varieties (Eva, Gema, Kora, and PI481671), the generated score plot of the PCA ( Figure 1A) revealed a complex accession distribution pattern that, in the largest distance, was not mainly accession-dependent, but organ-dependent as the root exudates showed strong dissimilarity with shoots and roots, especially for Gema, Eva, and PI491671. The supervised PLS-DA analysis ( Figure 1B) confirmed the separation, previously observed with the PCA. The separation was achieved by virtue of the first two principal components, which explained a total variance of 60.2%. Component 2 explained the highest variance (40.5%), while component 1 explained 19.7% of the total variance. The hierarchical cluster showed similarity in the root and shoot polyphenol contents of each accession, being the root and shoot contents of the four accessions grouped together in the same cluster of the dendrogram, but each species constituting and independent subclusters of the tree ( Figure 1C). Surprisingly, this division was not species dependent, as there was more distance among the common accessions Eva and Gema with Kora than among the common accessions Gema and Eva with the Tartary accession PI481671. Catechin (CAT); protocatechuic acid (PTA); rutin (RU); vanillin (VN); m-toluic acid (M-TA); 4-hydroxyacetophenone (4-HA); orientin (OR); luteolin (LU); quercetin (QE); phthalic acid (PA); 4-chlorobenzoic acid (4-CBA); salicylic acid (SAA); syringaldehyde (SY).  In contrast, the root exudates of Gema, Eva, and PI481671 were grouped in a totally independent cluster, showing its dissimilarity from the rest of the samples and indicating a characteristic behaviour of exudation in buckwheat. Moreover, although the content of root exudates of common and Tartary accessions was grouped in the same cluster of the dendrogram, Gema and Eva were grouped in a different subcluster than PI481671, suggesting a species-dependent dissimilarity among accessions at this level of distance. The compounds that had stronger weight in this classification were LU, QE, 4-HA, M-TA, SAA, CAT, and PTA (over 1.0 VIP score). LU, QE, and 4-HA were characterized to be increased in Gema, Kora, and PI481671 shoots, roots, and root exudates in comparison with Eva, while M-TA was increased in Eva shoots, roots, and root exudates in comparison with Gema, Kora, and PI481671 ( Figure 1D). When Eva was co-cultured with L. rigidum, the concentration of most of the polyphenols was significantly increased in the shoots, roots, and especially root exudates, following a co-culture of 7 days (Table 4). Table 4. Quantification of polyphenols (PP) from shoots, roots, and root exudates from buckwheat accession Eva, alone (control) or in association with L. rigidum (LR) or P. oleracea (PO). Quantities are expressed in µg kg −1 dry weight (shoots and roots) and µg L −1 (root exudates). Means followed by different lowercase letters within a column show significant differences among treatments (p ≤ 0.05). Bold numbers indicate a significant increase when compared with the control, while grey cells indicate a significant decrease.  Different specialized metabolites such as VA, DA, FA, P-CA, LU, SY, PTA, QE, VN, and 4-HA were significantly increased in the shoot tissues after the co-culture of Eva with L. rigidum (Table 4). Even VA and VN were found in high concentrations after co-growth with annual ryegrass, although could not be detected in Eva shoots when growing alone.

Sample
In contrast, there was a clear trend to reduce most of the detected polyphenols (DA, FA, P-CA, PA, LU, PTA, QE, SAA, CAT, ECAT, OR, VIT, and RU) in Eva shoot tissues when co-cultured with P. oleracea, especially LU, ECAT, and RU, which were reduced for more than 57 times, 47 times, and 31 times, respectively, followed by OR and PA, with reductions of more than 12 times. There was also a significant reduction in PTA (6.6 times), and QE (4 times). In contrast, a significant increase was only obtained in 4-CBA, VA, and 4-HA levels in the shoot tissues of Eva when co-cultured with P. oleracea (Table 4).
In root tissues, the significant increase in the content of phenolic acids and flavonoids, such as P-CA, LU, SY, PTA, 4-HA, CAT, ECAT, OR, and VIT when Eva was co-cultured with L. rigidum was coherent with the increase observed in the shoots and root exudates of this buckwheat accession in the presence of annual ryegrass. Especially interesting were the results of SY, ECAT, and VIT that increased their contents by more than 12 times, and of LU, which could not be detected in the control, although it appeared in Eva when co-growing with L. rigidum. In Eva roots co-cultured with P. oleracea, there was also an increase in VA, P-CA, LU, SY, PTA, CAT, ECAT, OR, and VIT following the pattern previously described for Gema and Kora. Surprisingly, LU was found to be increased after co-culture with both weeds, but was absent in the root tissues of the control plants of Eva. As a result, Eva root tissues accumulated in general significantly strong levels of polyphenols (Table 4).
In contrast, Eva root exudates were strongly increased only in the presence of L. rigidum, with 18 significantly increased exuded compounds (VA, DA, FA, P-CA, PA, SA, LU, SY, PTA, QE, VN, RU, SAA, 4-HA, CAT, ECAT, OR, and VIT) out of 20 compounds detected (Table 4). In fact, nine of the polyphenols significantly increased after L. rigidum co-culture did not practically appear in the control samples (when Eva was growing alone), which could not be detected in the analyses. Especially interesting is the exudation of PTA (2836-times), ECAT (1034-times), FA (511-times), DA (477-times), RU (467-times), CAT (346-times), VA (345-times), QE (304-times), OR (216-times), LU (200-times), VIT (81.6-times), SA (47-times), and VN (22.6-times). In contrast, in general, root exudates significantly decreased following the co-cultivation of Eva with P. oleracea and even the polyphenols DA, P-CA, and SY practically disappeared from the medium, while the content was only significantly increased for the polyphenols PTA, 4-HA, and CAT, which could not be detected in the root exudates when the Eva buckwheat plants were growing alone (Table 4).

Chemical Profile of GEMA Polyphenols
As shown in Table 5, the distribution of polyphenols in the shoots, roots, and root exudates of Gema was different depending on whether this accession was grown with L. rigidum or with P. oleracea.
Compared with the control, where Gema plants were growing alone, only the concentration of DA was significantly increased in shoot tissues when Gema was co-cultured with L. rigidum (Table 5), while the level of PTA, SAA, and 4-HA significantly increased in the shoot tissues when Gema was co-cultured with P. oleracea. Although the increases in PTA and SAA were not especially relevant, 4-HA increased by 24 times its concentration in the shoot tissues in the presence of P. oleracea. In addition, there were more polyphenols inhibited in the shoots when Gema was co-cultured with P. oleracea (DA, LU, SY, VN, ECAT, OR, and VIT) than with L. rigidum (VN, CAT, and ECAT). Curiously, the concentration of the polyphenol VN almost disappeared in the shoot tissues when Gema was co-cultured with either weed species. Similar to the shoot analyses, Gema increased significantly more polyphenols in root tissues when co-cultured with P. oleracea (DA, P-CA, LU, 4-HA, RU, OR, and VIT) than with L. rigidum, where only 4-HA and RU showed a significantly increased content. In fact, a significantly strong accumulation in the amounts of DA (56-times), LU (98-times), OR (5-times), VIT (4-times), and RU (3-times) was found in the roots of Gema when co-cultured with P. oleracea (Table 5). Table 5. Quantification of polyphenols (PP) from shoots, roots, and root exudates from buckwheat accession Gema, alone (control) or in association with L. rigidum (LR) or P. oleracea (PO). Quantities are expressed in µg kg −1 dry weight (shoots and roots) and µg L −1 (root exudates). Means followed by different lowercase letters within a column show significant differences among treatments (p ≤ 0.05). Bold numbers indicate a significant increase when compared with the control, while grey cells indicate a significant decrease.

Sample
Treatm  Regarding the presence of polyphenols in root exudates, the levels of many of the specialised metabolites decreased significantly when Gema plants were grown in co-culture with P. oleracea (4-CBA, DA, FA, SY, RU, CAT, and ECAT) and some of them (DA, FA, SY, RU, CAT, and ECAT) practically disappeared in Gema root exudates following co-growth with common purslane. In fact, only one polyphenol (QE) was slightly more exuded in front of common purslane than when Gema plants were grown alone. In contrast, none of the polyphenols was decreased in the presence of annual ryegrass while CAT and OR significantly increased their exudation to the medium with increases of more than five times for CAT and more than three times for OR (Table 5).

Chemical Profile of KORA Polyphenols
When Kora was co-cultured with L. rigidum (Table 6), there was a general decrease in polyphenols in the roots (DA, FA, P-CA, LU, QE, RU, OR, and VIT) and root exudates (DA, FA, P-CA, SA, LU, QE, 4-HA, RU, CAT, ECAT, and VIT) that was not found for common purslane, which showed the most inhibited polyphenols in the shoots, with DA, FA, P-CA, SA, M-TA, LU, 4-HA, CAT, and ECAT practically disappearing from the shoot tissues, and QE, OR, RU, and VIT also significantly reduced in the presence of this weed compared with the control. As a result, no polyphenols were found to be accumulated in the shoots of Kora after co-growth with P. oleracea. Table 6. Quantification of polyphenols (PP) from shoots, roots, and root exudates from buckwheat accession Kora, alone (control) or in association with L. rigidum (LR) or P. oleracea (PO). Quantities are expressed in µg kg −1 dry weight (shoots and roots) and µg L −1 (root exudates). Means followed by different lowercase letters within a column show significant differences among treatments (p ≤ 0.05). Bold numbers indicate a significant increase when compared with the control, while grey cells indicate a significant decrease.  In contrast, once more, different polyphenols were found to be significantly accumulated in the presence of P. oleracea in root tissues, as M-TA (919-times), 4-HA (17-times), DA (16-times), LU (10-times), VIT (7-times), and OR (5.6-times), while most of the polyphenols, were strongly reduced after co-cultivation with L. rigidum, as already explained above.
This reduction in the polyphenols in root tissues of annual ryegrass was consistent with the general reduction of polyphenols found in Kora root exudates when co-grown with L. rigidum, with many of them, such as DA, FA, P-CA, SA, LU, 4-HA, CAT, and ECAT, practically disappearing from the root exudates following co-growth with L. rigidum. In contrast, Kora strongly exuded a higher amount of several phenolic acids and flavonoids in the presence of P. oleracea, such as M-TA (321-times), 4-HA (145-times), DA (10-times), FA (5.6-times), OR (5.4-times), and P-CA (3.2-times). Finally, OR was the only polyphenol more exuded when Kora was co-cultivated with L. rigidum compared with the control (Table 6).

Chemical Profile of PI481671 Polyphenols
In general, the root and shoot tissues of PI481671 were more sensitive to the presence of P. oleracea than L. rigidum in the medium (Table 7). In fact, while only three polyphenols were altered in PI481671 shoot tissues in the presence of annual ryegrass (increases in M-TA and QE, and a decrease in PA), the content of seven polyphenols was altered in the presence of common purslane (increases in DA, FA, SAA, 4-HA, and ECAT, and decreases in VA and PA). This pattern was even more obvious in the root tissues, where the content of four polyphenols was altered in the presence of annual ryegrass, while the content of up to 14 polyphenols was altered in the presence of common purslane. In fact, after co-growth with P. oleracea, the root tissues of PI481671 accumulated strong amounts of polyphenols, showing a significant increase in the concentration of different specialized metabolites, such as VA, DA, FA, P-CA, SA, VN, SAA, 4-HA, RU, and ECAT. Especially interesting are the strong increases of ECAT (200-times), RU (26-times), and SAA (23-times). In contrast, the results only showed a significant increase in one polyphenol, OR (278-times), in the root tissues after the co-cultivation of PI481671 with L. rigidum (Table 7). Table 7. Quantification of polyphenols (PP) from the shoots, roots, and root exudates from buckwheat accession PI481671, alone (control) or in association with L. rigidum (LR) or P. oleracea (PO). Quantities are expressed in µg kg −1 dry weight (shoots and roots) and µg L −1 (root exudates). Means followed by different lowercase letters within a column show significant differences among treatments (p ≤ 0.05). Bold numbers indicate a significant increase when compared with the control while grey cells indicate a significant decrease.  Although there was a general lower root exudation in the Tartary buckwheat accession PI481671 when compared with the common buckwheat accessions (Kora, Gema, and Eva), there was a significant increase in the root exudates of DA, SA, QE, and ECAT following the co-cultivation of PI481671 with L. rigidum, indicating that the levels of polyphenols in the root exudates had a propensity to rise. In contrast, only VIT increased in the root exudates after co-cultivation with P. oleracea.

Multivariate Analyses of Polyphenols' Profile of Shoots, Roots, and Root Exudates of the Four Buckwheat Accessions
When independently comparing the polyphenol profile of the shoots (Figure 2A), roots ( Figure 2B) and root exudates ( Figure 2C) of Gema, Eva, Kora, and PI481671, the results of the PLS-DA analysis further exacerbated the separation among accessions for both shoots and roots (Figure 2A left, Figure 2B (Figure 2B left). The same was found for the shoot polyphenol profile of the accessions Gema, PI481671, and Eva (Figure 2A left). In contrast, no clear groups were found for root exudates, where the polyphenol profile of the plants of the different accessions growing alone or in co-culture with L. rigidum or P. oleracea were completely overlapped for the four accessions tested in this study (Figure 2C left).
The variable importance in projection (VIP) scores (built on the polyphenols with a VIP score higher than 1.0) revealed that PTA and M-TA were the only polyphenols with VIP scores higher than 1.0 that were found in all of the analyses (shoots, roots, and root exudates), while CAT and RU were also common for the shoots and roots. In particular, the compounds with a higher VIP score in the shoots were, in order of importance, CAT,  Table 8). Table 8. Variable importance of projection (VIP) features for the groups from the PLS-DA analysis for roots, shoots, and root exudates of the three common buckwheat accessions Eva (E), Gema (G), and Kora (K), and the Tartary accession PI481671 (PI) growing alone or in co-culture with L. rigidum or P. oleracea. The compounds included in the table are those compounds with a VIP score higher than 1.0 for each type of sample (shoots, roots, and root exudates). Bold numbers indicate compounds with important VIP score in the shoots, roots, and root exudates, while italic numbers indicate compounds with significant VIP score in the shoots and roots.

Discussion
Crop species with allelopathic activity are known as good options for reducing weed damage in sustainable agroecosystems [39]. Although the allelochemical potential of buckwheat crops to manage weeds in the field has not been deeply studied up until now, different papers suggest the presence of bioactive compounds on their extracts and residues that can control the development of different weeds [23,26]. Moreover, different genotypes may produce specialized metabolites differently, opening a wide variety of allelopathic potentials and, consequently, suppressive effects on weeds [40,41]. For this reason, studies such as those carried out in this work, evaluating different varieties of buckwheat that, by themselves, as a crop (alive plants), can control the presence of weeds in their environment, are highly indispensable.
The current study provides further evidence that phenolic compound synthesis, distribution, and exudation vary among buckwheat accessions, and that these compounds play a role in the interference among plants of crops and weeds. Phenolic compounds are specialized metabolites that can behave as phytotoxic when exuded into the medium, affecting the growth of neighbouring plants, as reported for several phenolic acids [42]. Meanwhile, the identification of allelochemical substances and their particular mode of action and interference in different physiological processes is required to make use of the allelopathic capabilities of crop plants in weed control [43]. Additionally, previous studies revealed that production of phenolic compounds differ greatly between buckwheat accessions in the different tissues [40,44].
Phenolic compounds, and especially flavonoids, have been reported for decades as strong antioxidant compounds, behaving as protectors into the plant metabolism against any external biotic or abiotic damage to which the plant can be exposed [45,46]. In fact, flavonoids are the most reported specialized metabolites in the plant defence system [47], so they may play an important role in plant−plant competition other than allelopathy, by accumulating in the different organs of the plant (leaves, roots, stems, seeds, etc.), and make plants more resistant and resilient against external attacks [48].
Moreover, as reported by Uddin et al. [41], different buckwheat cultivars can show different contents of phenolic compounds, and even the same cultivar can show organ-related differences in the phenolic composition. Studying three common buckwheat cultivars (Suwon1, Suwon 2, and Suwon 12), they found that Suwon 1 had the highest levels of catechin and epicatechin, while the greatest amount of 4-hydroxybenzoic acid, chlorogenic acid, and 4-hydroxy-3-methoxybenzoic acid was present in the cultivar Suwon 2. In this context, the Suwon 2 cultivar dominated over the other two cultivars, with the highest phenolic compound content in the stem, flowers, and roots of common buckwheat. Something similar was found in this study, as when comparing the four accessions, the hierarchical cluster showed more dissimilarities between the samples, grouping shoots, and roots separately from the root exudates, than between species, as common and Tartary accessions were grouped together in the same branch of the dendrogram, although each accession was separated from the others in the sub-branches of the tree.
When analysing the response of the different buckwheat accessions to the weeds, our results showed that the most relevant polyphenolic compounds were DA, LU, 4-HA, CAT, ECAT, and OR, as were the compounds whose chemical profile changed more in the roots or shoots and root exudates of the different buckwheat accessions along the study.
4-HA was discovered in buckwheat root exudates and in soil extracts following buckwheat cultivation, in addition to flavonoids and phenolic acids [26]. On the other hand, CAT was found to be highly phytotoxic against Arabidopsis thaliana (L.) Heynh. and Festuca idahoensis Elmer [49]. According to previous research, the phytotoxicity of CAT on the root cell tissues of A. thaliana is caused by the cytoplasm condensing due to the rapid induction of reactive oxygen species, which is followed by an increase in Ca 2+ and acidification of the cytoplasm, resulting in cell death [50]. In our results, a strong significant increase in the root exudates (346-fold higher than the control) of CAT was observed in buckwheat variety Eva when co-cultured with L. rigidum. Golisz et al. [25] also established the effective concentrations for lettuce to be in this range (0.4 mM). These findings suggest that CAT is highly phytotoxic, but less selective against different weeds. In a similar way, Serniak [51] showed, in a comparative allelochemical study, that ECAT exerted strong phytotoxic effects on radish seedling growth. Moreover, ECAT significantly decreased the growth of Lepidium sativum L. [52], and radish root growth was also inhibited in vivo as the result of the phytotoxic activity of ECAT [51].
Because of the quick evolution of resistance in L. rigidum and P. oleracea, comprehensive weed management measures including crop allelopathic varieties are required to slow down this rapid evolution and promote sustainable control [53]. Determination of the mechanism(s) associated with weed suppression is essential to determine if the use of crop varieties for allelopathic and competitive weed suppression in cereal and pseudo-cereal crops is going to provide sustainable solutions for weed management and to overcome resistance problems in weeds.
In this study, the variety Gema showed the strongest crop competitive ability against mono and dicot weeds compared with the other buckwheat accessions of Eva, Kora, and PI481671. This common buckwheat accession greatly inhibited the germination and root length of the monocot L. rigidum in more than 70% when compared with the control (i.e., L. rigidum growing alone). Strong effects of Gema were also observed on the dicot weed P. oleracea, where shoot and root lengths were stimulated, while no increases in fresh plant weight could be detected, resulting in longer, but much weaker, shoots and roots. Gema accumulated more DA, FA, P-CA, LU, 4-HA, OR, RU, and VIT in the roots when co-cultured with P. oleracea, while QE was the only polyphenol significantly more exuded to the medium after co-growing with this dicot weed. Previous works have reported that buckwheat varieties can accumulate polyphenols in the roots and shoots as a means of defence or protection [27,54]. Our results indicate that strong competition may be taking place between Gema and P. oleracea, and that P. oleracea might be trying to colonize more space (via longer roots) at the cost of making its roots weaker. The pressure that Gema has on P. oleracea can be related to the significant increase in root length of Gema in the presence of this weed, which would be competing with P. oleracea by colonizing the medium. In this context, flavonoid accumulation in the roots might be protecting Gema in front of this dicot weed. Root exudation represents a carbon cost to the plant [55]; therefore, the reduced root exudation of polyphenols could save energy that the buckwheat plant could use for defence or protection against P. oleracea. Several researchers have reported the antioxidant properties of flavonoids from different buckwheat varieties [45,56]. In this sense, the significant increase in some polyphenols in the roots and shoots of Gema plants could be protecting them from the damage induced by the presence of this dicot weed.
Allelochemical plants, such as buckwheat, have distinct mechanisms for inducing the phytotoxic effects on monocot and dicot weeds, so that biological action on the target weed differs from one weed to another [24,26]. In this sense, the behaviour of Gema with the monocot weed L. rigidum was totally different than with P. oleracea. In fact, no alterations in leaf and root weight or shoot and root length were observed in Gema plants when co-growing with L. rigidum. On the contrary, the germination and root length of L. rigidum were strongly inhibited by Gema, with 80% and 70% inhibitions, respectively. Gema increased the exudation of CAT by more than five times and OR by more than three times in the presence of L. rigidum, which could be enough to inhibit the germination and growth of L. rigidum, as there is no relevant accumulation of polyphenols on the roots or shoots of buckwheat plants and neither growth parameters of Gema plants are affected in front of this monocot weed, which suggests that L. rigidum does not represent a threat to Gema plants. The phytotoxic activity of CAT and OR would be enough for Gema to handle L. rigidum development [22].
The next accession with a strong capacity to sustainably control weeds was the Eva variety. This common buckwheat accession greatly inhibited the germination and fresh plant weight of the monocot L. rigidum, while it strongly stimulated the shoot length of this weed, which resulted in longer but weaker plants. Something similar was observed on the dicot P. oleracea, where Eva stimulated the shoot length but did not increase the plant weight. We revealed that this competitive ability of Eva was related to its robust root exudation of different polyphenols (phenolic acids and flavonoids), such as VA, DA, FA, P-CA, PA, SA, LU, SY, PTA, QE, VN, 4-HA, RU, CAT, ECAT, OR, and VIT, when co-cultured with both weeds; this was especially true for L. rigidum, which induced a significant increase in the exuded content for 16 polyphenols out of the 19 analysed. This increased the production and exudation of different polyphenols, especially PTA (2836-times), ECAT (1034-times) FA (511-times), DA (477-times), RU (467-times), CAT (346-times), VA (345-times), QE (305times), OR (215-times), and LU (200-times) by Eva compared with the control, which would ensure the success in the inhibition of the germination and development of L. rigidum and in the induction of increased weakness in P. oleracea. Our results demonstrated that when buckwheat recognizes the presence of the weeds it subsequently changes its root exudation profile to impede their growth. These results are consistent with those found by Gfeller et al. [57] for buckwheat in the presence of redroot pigweed. Moreover, although previous studies [20,54] have suggested the accumulation and exudation of RU as the responsible allelochemical molecule to inhibit the growth of different weeds, our results showed that there are a plethora of compounds participating in this phenomenon, and that there are other polyphenols, such as PTA, ECAT, or FA, that can playing an even more strong allelochemical role than RU. In this sense, in recent research, Krumsri et al. [58] evaluated the phytotoxic potential of Dalbergia cochinchinensis Pierre ex Laness. and found that PTA, the most exuded compound by Eva roots, caused growth inhibition on Echinochloa crus-galli (L.) P. Beauv. and L. sativum at low concentrations. At concentrations greater than 10 mM, ECAT, the second most exuded compound by Eva, significantly decreased the growth of L. sativum [52]. Radish root growth was also inhibited in vivo because of the ECAT phytotoxic activity [51]. In another study, Hussain and Reigosa [43] evaluated the effects of FA and DA on the photosynthesis of Rumex acetosa L., and found that both compounds behaved as potent inhibitors of photosynthetic traits, leading to weaker plants. The strong increase in the root exudation by living plants indicates that Eva molecules attacked the herbicideresistant weeds (L. rigidum and P. oleracea), inhibiting the germination of L. rigidum and hindering the development of both monocot and dicot weeds. The exudation of several flavonoids (QE, VN, 4-HA, CAT, ECAT, OR, RU, and VIT) demonstrates that the defence strategy of Eva is alive and working closely with the attacking phenomena to obtain access to the available resources (space and light) for its growth and development. Moreover, polyphenols can also play a role of defence and protection in the plant, as previously demonstrated by several authors [27,54]. In this context, Eva also significantly increased the production of several polyphenols in the shoots (VA, DA, FA, P-CA, LU, SY, PTA, QE, VN, and 4-HA) and roots (P-CA, LU, SY, PTA, 4-HA, CAT, ECAT, OR, and VIT) after growing with L. rigidum, and in root tissues (VA, P-CA, LU, SY, PTA, CAT, ECAT, OR, and VIT) after growing with P. oleracea. Especially interesting were the polyphenols P-CA, LU, SY, PTA, CAT, ECAT, OR, and VIT, which were found to increase in the root tissues in the presence of both weeds in very high concentrations.
Kora was the common buckwheat accession that affected the development of L. rigidum and P. oleracea less in the present study, although it showed a strong effect on the germination of both monocot and dicot weeds. Although Kora did not exude phenolic compounds and flavonoids after growth with L. rigidum, the accumulation in the roots and shoots of different polyphenols, such as M-TA, 4-HA, and OR could improve the antioxidant activity in buckwheat plants, providing an advantage in plant−plant competition [59].
In contrast, Kora exuded significant amounts of DA, FA, P-CA, M-TA, 4-HA, and OR after growth with P. oleracea, which were not only related to the strongest decrease in P.
oleracea germination for the four tested accessions, but also to the chemical control of P. oleracea by Kora, where the weed seedlings could normally grow, without affecting the growth and development of Kora plants.
The Tartary buckwheat accession PI481671 followed a similar chemical profile and buckwheat plant development to the common buckwheat accession Gema, although this accession did not affect the weeds in a similar pattern. PI481671 stimulated the weed total biomass and shoot length of P. oleracea while inhibiting the total weight of L. rigidum. This could be explained by the results previously found by Sijahović et al. [38], who demonstrated that buckwheat−weed interactions are dependent on the type of weed present in the neighbours, resulting in changes in the exudation behaviour of buckwheat plants. The root tissues of PI481671 indeed accumulated considerable levels of polyphenols after co-growing with P. oleracea, displaying a marked rise in the content of many specialized metabolites, including VA, DA, FA, P-CA, SA, VN, SAA, RU, 4-HA, and ECAT. The significant increases in ECAT (200 times) and SAA (23 times), a well-known defence compound, are particularly fascinating. Competitive genotypes can better access light, nutrients, and water resources in limited space, thus suppressing the growth and reproduction of nearby weed species [37]. Although the Tartary buckwheat (PI481671) showed generally less root exudation than the common buckwheat accessions, PI481671 increased, as Eva, the exudation of DA, SA, and QE, by several folds after co-culture with L. rigidum. In a recent study, Šćepanović et al. [60] showed that strong doses of VA, DHA, and P-CA, as well as the phenolic acid mixture, inhibited the early growth of Ambrosia artemisiifolia L.
As shown by the multivariate analyses, in this study, no differences were found among species (F. esculentum and F. tataricum), but among the analysed samples. In fact, the roots and shoots were grouped separately from the root exudates. When having a look at the compounds with a higher VIP score (weight) for the comparison of the different samples, the polyphenols PTA and M-TA were common to all of the analyses (shoots, roots, and root exudates), while CAT and RU were also common for the shoots and roots. Most of these compounds (i.e., CAT, RU, and PTA) have been shown to have antioxidant properties [61][62][63] against different stress factors, which could be one main reason for their accumulation in the roots and shoots of common and Tartary buckwheat accessions, giving an advantage to buckwheat plants in front of the surrounding weeds.
Our findings show that different buckwheat accessions have varying capacities to release or accumulate specific metabolites in the presence of surrounding weeds, as well as varying capacities to manage those weeds sustainably. Additionally, each accession exhibited varied the inhibitory capacities and chemical profiles against monocot or dicot weeds, depending on the type of weed in their vicinity [20,25]. These results are consistent with those of Kalinova [30], who found varietal differences for the inhibition of lettuce by three different buckwheat varieties and related these differences to allelochemical action of buckwheat by measuring the production of the known allelochemical compounds ECAT and RU.
The findings of this study indicate that the buckwheat accessions that most significantly impacted the growth of the tested weeds were those with the highest production of allelopathic compounds and their exudation into the rhizosphere. The inhibitory effect on weed germination and growth could be caused by the allelochemicals that were exuded to the medium, because in the current experiments, there was no direct physical contact between the roots or shoots of the buckwheat and weeds. This was particularly true for the variety Eva, which demonstrated a high potential for controlling monocot weed L. rigidum trough root exudates. However, the superior competitive ability that the accumulation of polyphenols in shoots and roots provided buckwheat plants in front of weeds could be driving the negative impact of the tested buckwheat accessions on the two target weeds, even though these weeds showed resistance against different herbicides.
The present results highlight the necessity to screen different buckwheat accessions to find the better ones to be used in organic agriculture, due to the variation in the synthesis, distribution, and exudation of polyphenols, which can provide a different allelopathic or competitive ability to different accessions.

Germination and Growth Bioassays
Four different buckwheat accessions (Gema, Kora, Eva, and PI481671), previously pre-selected for their potential to be used in organic farming (in the frame of the EU project ECOBREED), were screened for their allelopathic potential against germination and seedling growth of common purslane (Portulaca oleracea L.) and annual ryegrass (Lolium rigidum Gaud.) using perlite as an inert substrate. Eva is a commercial variety from Kmetijski inštitut Slovenije/Agricultural Institute of Slovenia (KIS), Slovenia. Gema, Kora, and PI481671 were provided by the Czech Gene Bank-Crop Research Institute of Prague (Czech Republic). Seeds of common purslane and annual ryegrass were obtained commercially from "Semillas Cantueso" (Cantueso Natural Seeds, Cordoba, Spain) and Herbiseed (Herbiseed Twyford, Berkshire, UK), respectively. Surface-sterilized seeds of each buckwheat accession were grown alone in individual plastic trays (32 × 20 × 6 cm) filled with a 5 cm deep layer of perlite (500 g/tray), watered with distilled water, and kept in a growth chamber with a day and night temperature of 20 • C and 12/12 h light/dark photoperiod. The light was supplied by cool white fluorescent tubes and irradiance was maintained at 275 µmol m −2 s −1 . The buckwheat seeds were sown and left to germinate for 5 days before transferring to treatment plastic trays.
Ten buckwheat seedlings per plastic tray and three plastic trays per treatment were used in these experiments for each common and Tartary buckwheat accession. At day 1, similar seedlings for each buckwheat accession were selected and sown in a plastic tray with perlite and were watered each other day. The seedlings were placed in three rows on one half of the tray and placed in a controlled environment growth cabinet with a daily photoperiod of 12L:12D and continuous temperature of 20 • C. After 10 days of growth for the buckwheat accessions, L. rigidum or P. oleracea seeds were added to the other half of the tray. The arrangement was such that the allelochemicals produced and released by the buckwheat seedlings could diffuse throughout the perlite medium to influence weed germination and growth, but no physical contact was allowed among the roots or shoots of buckwheat seedlings and weeds.
After growing together (buckwheat and weeds) under the same conditions for 7 days, the germination rate, total weight, shoot and root length, and plant height were measured in the two target weeds, L. rigidum and P. oleracea.

Plant Harvest and Metabolite Extraction
After harvesting the shoots and roots of the buckwheat seedlings and collecting the root exudates from each treatment tray, the samples were stored at −80 • C until extraction. The samples from each buckwheat accession were separately processed for the identification and quantification of phenolic compounds and flavonoids from the shoots, roots, and root exudates. The plant tissues (shoots and roots) were lyophilized, ground into powder with a mortar and pestle after the addition of liquid nitrogen, and macerated with 9 mL of HCl (1 mM). Afterwards, the whole solution was transferred into a vial, sonicated for 15 min (Branson SINIFIER 250; microTip limit, output 3), and centrifuged at 20,000 rpm at 10 • C for 15 min (SORVALL RC 5B Plus, Du Pont). The supernatant was collected and extracted three times with diethyl ether (DE). The aqueous layers were discarded, and the corresponding organic layers were combined. The organic phases were evaporated in a multivapor (P-12; Buchi, Switzerland) with 12 simultaneous evaporating positions. The multivapor (P-12) comprised of a vacuum pump (V-700), vacuum controller (V-850), rotavapor (R-210), heating bath (B-491), and recirculating chiller (F-105). The temperature of the recirculating chiller was set to −10 • C. The organic layers (DE) were placed in 15 mL plastic tubes and attached to the multivapor to evaporate the organic solvent under reduced pressure (456 mbar for DE) at 35 • C. Because of the chilling temperature (−10 • C) of recirculating chiller, and the temperature of the heating bath (35 • C), the organic solvent was evaporated and condensed in the attached crystal balloon. The final volume of the residual solution was approximately 1 mL and this solution was further dried with N 2 . Methanol was used to dissolve the residual powder and was injected for the LC-MS analysis.
The perlite-based nutrient-free water-growth medium was collected and adjusted to pH 3.0 with 0.06 M HCl. Then, 25 mL of the root exudated water was extracted three times with 25 mL DE. Further treatment of the root exudated water samples was identical to the preparation of shoots or roots extraction as described above.

Identification of Phenolic Compounds by LC-MS
Shoots, roots, and root exudates were extracted from each buckwheat accession using diethyl ether. Specialized metabolites (phenolic acids and flavonoids) were separated on ultra-high performance liquid chromatography coupled to a quadrupole time-of-flight high-definition mass spectrometry detector (UHPLC-qTOF-MS, Thermo Fisher Scientific Inc., Madrid, Spain) according to Wu et al. [64] with small slight amendments according to Hussain et al. [65].
High Performance Liquid Chromatography−Mass Spectrometry (HPLC-MS, 1260 Series, Agilent, Santa Clara, CA, USA) was performed using a system consisted of compact mass detector equipment (TRIPLE QUAD 3500; AB SCIEX INSTRUMENTS, AB Sciex Pte. Ltd., Framingham, MA, USA). Polyphenols were separated with a C18 column (PHE-NOMENEX LUNA, 150 mm × 2 mm and 3 µm, Phenomenex, Inc., Torrance, CA, USA) using different chromatographic conditions depending of the compounds. Hypericin was separated at a flow rate of 400 µL min −1 with a column temperature of 40 • C, an injection volume of 10 µL, and the column was equilibrated for 6 min between runs. The isocratic elution used was a mixture of two solvents: A, consisting of 5 mM ammonium acetate and 0.1% acetic acid in water, and B, consisting of acetonitrile. The isocratic conditions were 25% A and 75% B for 10 min. The other phenolic acids were separated at a flow rate of 300 µL min −1 , the column temperature was 40 • C, the injection volume was 10 µL, and the column was equilibrated for 6 min between runs. The gradient elution used was a mixture of two solvents: A, consisting of 0.1% formic acid in water, and B, consisting of 0.1% formic acid in acetonitrile. Initial conditions (98% A and 2% B) were held for 4 min before ramping to 20% B at 7 min and 90% B at 14 min. Initial conditions were recovered at 15 min and held until 21 min. The instrument parameters were as follows: curtain gas (CUR), 25 psi; collision gas (CAD), 7 psi: ion spray voltage (IS), −4500 V; temperature (TEM), 400 • C; ion source gas 1 (GS1), 55 psi; ion source gas 2 (GS2), 55 psi; interface heater, on.
The quantification of the concentration of the compounds was obtained from calibration curves that related the detector's response to the pure analyte's concentration of those compounds identified in the chemical analyses.

Data Analyses
The experiments were carried out using a completely randomized design with three replications (each replication was a bulk of 10 buckwheat plants and 10 weed plants). IBM SPSS software (SPSS Inc., Chicago, IL, USA, version 22.0) was used to analyse the data. To detect outliers, an exploratory data analysis was performed. The Kolmogorov−Smirnov test was used to check for deviation from normality, and the Levene test was used to check for homogeneity. Depending on the homoscedasticity of the samples, one-way ANOVA or Kruskal−Wallis tests were performed for germination and seedling growth data to establish the significant effect (p ≤ 0.05) of the treatments (different accessions). The results are presented in the tables as the percentage of increase or decrease when compared with the control. Different letters represent significant differences in treatment. Polyphenol (phenolic compounds and flavonoids) data were analysed through analysis of variance and the Duncan multiple range test was performed to establish the significant effect (p ≤ 0.05) within the treatments (alone, co-cultured with L. rigidum, or co-cultured with P. oleracea).
The identified polyphenols were analysed using the Metaboanalyst 5.0 software. The missing values were replaced with half of the minimum value found, and then data were Log 10 transformed and Pareto scaled. Data were then analysed through unsupervised principal component analysis (PCA), to visualize group discrimination, and through the supervised partial least square discriminant analysis (PLS-DA). Feature selection with the highest discriminatory power was based on their variable importance in projection (VIP) score > 1.0. To avoid overfitting, the PLS-DA model was validated using Q2 as a performance measure, the 10-fold cross-validation and setting in the permutation test a permutation number of 20 (see tables reported in Supplementary Table S1).

Conclusions
We conclude that the selection, evaluation, and development of buckwheat accessions with an increased competitive ability and strong allelopathic potential will be a good option for sustainable weed management. Our results confirm the ability of different buckwheat accessions to suppress monocot and dicot weeds, and this ability clearly appears to be accession dependent. In this regard, Gema appears to be the accession that should be used for growing in organic agriculture due to its capacity to sustainably regulate the germination and growth of the monocot weed L. rigidum and the dicot weed P. oleracea, while stimulating the root growth of buckwheat plants. Meanwhile, all four buckwheat accessions showed varying degrees of allelochemical production and release to control weeds through affecting multiple processes, such as germination, growth, and weed biomass. We conclude that different buckwheat genotypes may have different capacities to produce and exude several types of specialized metabolites, which lead to a wide range of allelopathic and defence functions in the agroecosystem to sustainably manage the growing weeds in their vicinity.