Genotypic and Environmental Impacts on Vicine and Convicine Concentrations in Faba Beans

Abstract

High concentrations of vicine and convicine (v-c) in faba beans can trigger favism in susceptible humans, posing a significant barrier to the broader adoption of faba beans as a food source. While plant breeding and various post-harvest processing methods have been adopted to reduce v-c levels, there is limited understanding of how agronomic practices may assist in reducing v-c levels. This study investigated the effect of sowing time (TOS), soil type, and genotype on v-c levels in faba beans. Twelve faba bean genotypes were evaluated across multiple field sites by applying two sowing times and two diverse soil types. The v-c content was quantified using established chromatographic techniques. Genotypes were identified as the most major factor affecting v-c levels, with significant variation observed in mean vicine and convicine contents. Sowing time also had a significant impact (p < 0.01), with lower v-c levels observed in TOS 1 compared to TOS 2. This reduction may be due to a longer plant development period and extended seed desiccation in TOS 1. Soil conditions, likely linked to nutritional factors, significantly influenced vicine concentrations (p < 0.05) but did not influence convicine levels (p > 0.05). These findings highlight the importance of agronomy practices, such as optimal sowing time, soil nutrition, and moisture management, in minimizing v-c levels; the most effective strategy remains the development of low v-c genotypes combined with farming practices that naturally suppress v-c accumulation.

1. Introduction

Vicia faba L. (Fabaceae) (commonly known as faba bean) is a cool season annual food legume crop grown as a part of crop rotation and is often grown within a cereal cropping rotation, contributing to nitrogen soil nutrition and reducing soil-borne diseases [1,2]. The worldwide production of dry faba bean is gradually increasing; in 2000, it was 3.8 Mt and, in 2020, it was 5.7 Mt (an increase of 50% in 20 years) [3]. In 2020, for dry faba beans, China was the top producer (1.7 Mt), Australia was the top exporter (0.4 Mt), and Egypt was the main importer (0.3 Mt) [3].

Faba bean, a rich source of protein and starch, is a significant pulse and vegetable crop especially for China and Mediterranean countries [4]. It is consumed in many forms fresh, dry (roasting is a traditional way of consuming faba beans in Indian subcontinent), or processed (e.g., falafel, a common dish in Mediterranean and Middle Eastern countries). In recent years, there has been an increase in popularity of plant-based protein products [5]. Due to high protein content, faba bean could be a potential source of sustainable plant-based protein [6]. Even though faba beans are protein-rich highly nutritious grains, they contain a range of anti-nutritional factors, which restricts their usage amongst certain human populations. One of the main anti-nutrient factors in faba bean are vicine and convicine (v-c) due to their potential to cause favism [7].

Favism (hemolytic anemia) is a severe reaction caused by the consumption of faba bean by people with glucose-6-phospate dehydrogenase deficiency (G6PD). G6PD deficiency is a lifelong genetic disorder and is one of the most prevalent forms of enzyme deficiency effecting more than 400 million people worldwide [8]. G6PDH deficiency impairs the ability of red blood cells to form NADPH, which leads to lower production of reduced glutathione, which, in turn, decreases the detoxification of peroxides and leaves red blood cells undefended against oxidative stress (e.g., the metabolites of vicine and convicine) and leads to hemolysis due to oxidative damage (Figure 1) [9].

Post-harvest methods for processing of faba beans have been investigated as a measure of reducing v-c contents, e.g., fermentation [10], various soaking methods [11], and other cooking methods, including roasting and boiling [12]. Soaking and cooking is one routine means of processing prior consumption and both steps help in reducing the v-c content [11,13]. Furthermore, cooking also inactivates the β-glucosidase enzyme in the seeds; this enzyme breaks down v-c into the potentially toxic metabolites, divicine and isouramil [9].

Vicine was initially isolated from the seeds of common vetch (Vicia sativa L.) in the late 19th century, and its presence was subsequently identified in faba bean in 1914 [14]. Pitz et al. [15] reported the presence of v-c in faba bean and common vetch (Vicia sativa L.), with reduced levels observed in Narbon vetch (Vicia narbonensis L.). However, the study did not detect v-c in other grain legumes such as chickpea (Cicer arietinum L.), soybean (Glycine max (L.) Merr.), lentil (Lens culinaris Medic.), field pea (Pisum sativum L.), and navy bean (Phaseolus vulgaris L.).

The v-c are heat-stable and their complete removal with processing methods is challenging [16]; thus, developing faba beans with low v-c content through targeted plant breeding or the use of genomic approaches are the future challenges for plant breeders. The 13 Gb faba bean genome [17] is relatively large compared to other legume crops (e.g., 0.7 Gb for chickpea [18] and ~4 Gb for lentil [19]) and it has recently been fully sequenced [20]. Furthermore, in recent years, several successful genomic approaches have been made in the mitigation of anti-nutrient component and enhanced seed quality traits of faba bean. Khazaei et al. [21] developed a high-throughput Kompetitive Allele Specific PCR (KASP) marker for low v-c concentration in faba bean. The low v-c faba beans have been reported to have vicine and convicine contents (0.29 mg g⁻¹ and 0.01 mg g⁻¹, respectively) significantly lower than the levels found in mainstream faba bean lines [22]. Nevertheless, the impact of growing practices or agronomy on the v-c levels on grain yield and disease resistance is yet to be fully investigated.

Crop management systems such as optimized sowing times and adaptation to soil conditions may be an effective tool in mitigating anti-nutritional compounds in faba bean. However, there is a limited knowledge on how agronomic practices can be applied to minimize the concentration of v-c. This paper reports the effect of time of sowing, soil type, and genotype on v-c in faba beans.

The high v-c content is one factor restricting the use of faba bean either in its raw form (food or feed) or as a protein concentrate (since the water-soluble v-c tend to co-extract with protein) [6]. As such any means of reducing the v-c content in grains would support increasing the production and consumption of faba beans [23]. This research investigated an agronomic treatment to mitigate v-c across 12 genotypes, considering variations in sowing times and two distinct soil types.

2. Materials and Methods

2.1. Field Trial and Genotypes

Faba bean samples were obtained from Southern Pulse Agronomy Victoria field trials grown within two diverse environments at Curyo, Victoria (35.85° S, 142.79° E), and Nhill, Victoria (36.33° S, 141.65° E), in 2021. The faba bean samples include 12 genotypes consisting of seven commercial varieties (Farah, PBA Amberley, PBA Bendoc, PBA Marne, PBA Nasma, PBA Samira, and PBA Zahra) and five pre-released breeding lines (AF12025, AF14075, AF14092, AF15278, and AF15283) from Australia. The genotypes are all spring cultivars and are characterized as having v-c content on a lower scale of commercial cultivars. The Curyo experimental site had two sowing times (TOS 1 and TOS 2) and four replications (n = 4). At Nhill site the trial was on two diverse soil types: (a) “Duplex soil”—a duplex of soil types with slightly acidic sandy loam topsoil and hard-setting alkaline clay subsoil in a swale and (b) “Sandy soil”—a hard-setting neutral to alkaline sand situated on a sand dune within the same location. The two soil types were approximately 200 m apart and, therefore, weather conditions were similar. Mean monthly averages for total rainfall, minimum temperature, maximum temperature, and global solar exposure for both sites are shown in

Table 1. Monthly rainfall, mean maximum temperature, mean minimum temperature, and mean global solar exposure for faba bean growing season at two trial sites, Curyo and Nhill, in 2021.

Trial Site	Observation	Apr.	May	Jun.	Jul.	Aug.	Sep.	Oct.	Nov.	Dec.
Curyo 2021	Total rainfall—mm	1.4	26.4	34.6	45.2	31	47.2	42.6	54.6	3.4
	Mean min. temperature—°C	7.2	5.1	5.6	3.4	4.4	5.8	6.4	10.4	12.7
	Mean max. temperature—°C	23.7	19.6	15.9	15.2	17.8	20.7	22.9	25.2	31.6
	Mean global solar exposure—MJ m−2	13.5	9.9	8.1	9.3	12.5	16.4	20.2	21.1	27.6
Nhill 2021	Total rainfall—mm	19	15.2	12	58.6	17.8	36.8	57	7.6	0
	Mean min. temperature—°C	8.1	6.3	6.2	4.5	4.9	5.9	6.6	9.2	11.2
	Mean max. temperature—°C	22.5	18.3	14.9	13.4	16.6	18.3	20.4	24.2	29.5
	Mean global solar exposure—MJ m−2	12.8	9.7	7.8	9	12.4	15	19.2	21	27.1

Note. Meteorological data for the trial sites were obtained from the Bureau of Meteorology (http://www.bom.gov.au/climate/data/ (accessed on 3 March 2023)).

. The soil nutrient composition at the Nhill site for the Duplex and Sand trials measured at 0–10 cm, 10–20 cm, and 20–40 cm depth are shown in

Table 2. Soil data for Nhill-Duplex soil and Nhill-Sandy soil trial sites at three depths.

	Depth	NH4+	NO3−	P	K	S	Cu	Fe	Mn	Zn	B	EC	pH	Soil
	(cm)	mg kg−1	mg kg−1	mg kg−1	mg kg−1	mg kg−1	mg kg−1	mg kg−1	mg kg−1	mg kg−1	mg kg−1	(dS m−1)	(CaCl2)	Moisture (%)
Nhill- Duplex soil	0–10	4.3	6.8	18.0	223	16.6	0.8	45.0	18.6	1.3	1.6	0.2	6.3	5.8
	10–20	2.3	2.8	5.3	258	8.3	0.8	35.9	4.4	0.3	3.1	0.2	6.9	15.6
	20–40	2.0	1.5	3.5	277	15.2	1.1	35.1	2.9	0.2	7.9	0.4	7.9	21.0
Nhill- Sandy soil	0–10	4.8	1.0	25.5	271	19.9	0.4	20.9	3.9	1.2	1.1	0.1	7.3	10.2
	10–20	2.5	1.5	8.0	244	9.8	0.4	26.1	1.2	0.3	2.2	0.2	7.5	16.6
	20–40	1.8	3.3	5.5	288	13.4	0.5	25.6	0.7	0.2	3.8	0.3	8.0	23.8

Note. The soil samples from trials were analyzed by a commercial laboratory (CSBP Soil and Plant Analysis Laboratory, Perth, Western Australia).

.

At both trial sites, seed was sown at a depth of 5 cm to achieve a plant density of 20 plants per m⁻² with Group F granular rhizobia (Nodulator ^®; BASF Australia Southbank, Victoria, Australia). Fertilizer (Mono Ammonium Phosphate–Nitrogen 9.2%, Phosphorus 20.2%, Sulphur 2.7%, and Zinc 2.5%) was applied in a band 3 cm below seed at a rate of 60 kg ha⁻¹ at the time of sowing. The TOS 1 and TOS 2 trials in Curyo were sown on 29 April 2021 and 1 June 2021, respectively. Both trials in Nhill were sown on 6 May 2021. The trials in Curyo and Nhill were harvested on 11 November 2021 and 6 December 2021, respectively. In addition to the commercially released varieties and five breeding lines included in the field trials, three of the breeding lines (18YB, 18ABJ, and 18ACB) developed by the faba bean breeding program to have inherently low in v-c content were analyzed for v-c contents.

2.2. Calculation of Heat Sum

The daily heat sum for the trial sites were calculated as follows:

Daily heat sum = (T_max + T_min)/2 − T_base

(1)

where T_max is the daily maximum temperature, T_min is the daily minimum temperature, and T_base is the base temperature.

For the calculation of thermal time (daily heat sum) in faba bean, a T_base of 0 °C was applied [24]. Temperatures below this threshold are considered non-contributory to phenological development and are excluded from cumulative daily heat sum. To obtain the accumulated heat sum for the podding period, the heat sum from the “flowering date” to “podding date” was pooled. Similarly, accumulated daily solar exposure (ASE) data measured as MJ m⁻² for the podding period were pooled from the “flowering date” to “podding date”. The dates of 50% flowering were taken as “flowering date” and dates of 50% podding were taken as “podding date”.

2.3. Protein, Moisture, and 100 Seed Weight Analysis

The protein and moisture content of samples were determined using near-infrared reflectance (NIR) spectroscopy (Foss XDS Rapid Content Analyser, FOSS Pacific Pty Ltd., Hilleroed, Denmark) of the harvested seeds. The 100 seed weight was determined by subsampling 100 randomly selected intact seeds from the harvested seeds and obtaining their weight.

2.4. Vicine and Convicine Analysis

The vicine and convicine (v-c) contents in faba bean samples were analyzed as per the method described in Elkins et al. [25]. The analytical method is as summarized as follows.

2.4.1. Sample Preparation

The whole-grain faba bean samples were ground with a hammer-type cyclone mill (Laboratory Mill 3100, Perten Instruments, Huddinge, Sweden) fitted with 0.8 mm sieve. The ground samples (10 mg) were extracted with 1 mL of 80% methanol and vortexed for 1 min and subsequently sonicated for 5 min. The mixture was centrifuged at 17,950 rcf for 5 min and the supernatant was transferred into a labeled tube, and the remnant v-c in the residue were extracted again with fresh 1 mL of 80% methanol. The supernatant was pooled, diluted 1:20, and analyzed for v-c contents.

2.4.2. Chromatographic Analysis

Liquid chromatographic analyses were undertaken using an ultra-high-performance liquid chromatography system (Thermo Fisher Scientific, Bremen, Germany) coupled to a Q Exactive™ Plus Hybrid Quadrupole-Orbitrap™ Mass Spectrometer (Thermo Fisher Scientific, Bremen, Germany). All MS data was acquired in positive electrospray ionization mode over a mass range of 80–1200 m/z. The column used for analysis was Synergi Polar-RP (150 mm × 2 mm, 4 μm) (Phenomenex, Lane Cove, New South Wales, Australia). The mobile phases used were 0.1% formic acid in milliQ water (A) and 0.1% formic acid in acetonitrile (B). The solvent gradient was an isocratic flow of 20% A and 80% B at 0.15 mL min⁻¹ for 4 min followed by 3 min at 100% B to clean the column and a further 3 min at initial condition for column re-equilibration with the column temperature at 40 °C. The chromatographic data were acquired using Xcalibur version 4.5.474.0 and processed using Tracefinder 5.1 Build 110 (Thermo Fisher Scientific, San Jose, CA, USA).

2.5. Statistical Analysis

For statistical analysis of the data, one-way ANOVA using Minitab^® (version 21.1, Minitab, State College, PA, USA) was carried out on multiple field replicates to investigate the effect of TOS and soil type on the vicine and convicine contents in faba bean samples. For ease of interpretation of the results and clarity, two factors (TOS and soil type) were individually analyzed with one-way ANOVA. The Tukey pairwise comparison was used for determining mean differences between treatments. The Pearson correlation was used to determine correlation between v-c content and various abiotic factors. Principal Component Analysis (PCA) was conducted on the various measured parameters.

3. Results and Discussion

Although the sowing dates for the TOS trials were four weeks apart, it was observed that physiological differences in plant development were less pronounced, as can be seen by number of days after sowing to 50% flowering and 50% podding (

Table 3. Number of days from sowing to 50% flowering and 50% podding for genotypes within trials.

Genotype	Treatment—Time of Sowing				Treatment—Soil Type
	TOS 1		TOS 2		Duplex Soil		Sandy Soil
	50% Flowering	50% Podding	50% Flowering	50% Podding	50% Flowering	50% Podding	50% Flowering	50% Podding
Farah	103	147	90	115	114	147	116	141
PBA Amberley	113	150	97	121	120	150	119	143
PBA Bendoc	112	151	95	120	118	138	118	143
PBA Marne	102	148	89	116	118	145	117	147
PBA Nasma	94	138	87	110	111	137	109	139
PBA Samira	111	151	95	118	119	139	118	143
PBA Zahra	113	152	96	119	120	148	121	147
AF12025	91	134	84	109	113	141	107	147
AF14075	109	149	95	118	120	147	119	139
AF14092	101	151	89	117	109	138	110	134
AF15278	112	149	95	119	118	147	118	137
AF15283	111	150	97	118	116	141	117	134

). These findings are consistent with the observations of McDonald, Adisalwanto, and Knight [26]. This study found that, within the typical range of sowing dates for faba beans in southern Australia, there was minimal difference in days to flowering due to delayed sowing, with developmental progression being predominantly regulated by thermal and photoperiodic responses. Effects of TOS 1 and TOS 2 may have also impacted on root development due to decreasing available soil moisture; however, these were not quantified in the field notes. For the trial at Nhill, physiological development was similar for duplex soil and sandy soil, as indicated by a comparable number of days from sowing for the 50% flowering and 50% podding (Table 3). Thus, any measured differences are most likely associated with the soil environment and soil moisture levels (Table 1).

3.1. Effect of Genotype

In this study genotype was the largest source of variation for v-c contents. Amongst the 12 genotypes analyzed the mean vicine content ranged from 4.23 to 5.40 mg g⁻¹ DB and the mean convicine ranged from 2.26 to 3.29 mg g⁻¹ DB (Figure 2). These results are similar to studies by Mayer Labba, Frøkiær, and Sandberg [27].

3.2. Effect of Time of Sowing (TOS)

Optimizing time of sowing (TOS) is a common agronomic practice to increase grain yield and improve quality [28,29]. The mean moisture content for the samples from TOS 1 and TOS 2 were 100.7 mg g⁻¹ and 100.4 mg g⁻¹, respectively. The TOS had a significant effect (p < 0.01) on the concentration of both vicine and convicine (Figure 3 and Figure 4). The samples from TOS 1 had a mean vicine content of 4.24 mg g⁻¹ DB and ranged from 3.27 to 4.83 mg g⁻¹ DB, whereas the samples from TOS 2 had a significantly higher mean vicine content of 5.19 mg g⁻¹ DB and ranged from 4.74 to 5.65 mg g⁻¹ DB. Similarly, the samples from TOS 1 had a mean convicine content of 2.42 mg g⁻¹ DB and ranged from 2.00 to 3.14 mg g⁻¹ DB, whereas the samples from TOS 2 had a significantly higher mean convicine content of 2.85 mg g⁻¹ DB and ranged from 2.32 to 3.40 mg g⁻¹ DB.

Harvest index is defined as yield of a crop versus the total biomass produced and is the result of combined interaction between genotype, environment, and crop management [30]. The mean harvest index for TOS 1 and TOS 2 is 0.48 and 0.46, respectively, and there was no significant effect of TOS on the harvest index (p > 0.05). Therefore, the measured differences in v-c content in TOS 1 and TOS 2 are thought to arise from temperature effects during pod-filling prior or additional flowering and podding later in the season.

The synthesis pathway of v-c in faba beans remains unclear; according to Björnsdotter et al. [7], v-c is suggested to originate from purine metabolism, being generated in the seed coat and subsequently transported to developing embryos. Cardador-Martínez et al. [12] noted that mid-harvest and late-harvest samples exhibited significantly lower v-c levels compared to early-harvest samples, suggesting that an extended plant life cycle results in reduced v-c content. In a study by Lattanzio, Bianco, and Lafiandra [31], an increase in v-c levels was observed during seed development and ripening, while a decrease occurred during the drying stage. Therefore, in our study, the observed lower v-c levels in TOS 1 compared to TOS 2 could potentially be attributed to an extended plant life cycle and a prolonged seed drying stage.

3.3. Effect of Soil Type

Soil with ample macro- and micro-nutrients maintains an optimal rate of crop growth and enhances grain quality [32], whereas nutrient deficient soil adversely affects grain quality [33]. The soil in the Nhill-sand trial site had a lower concentration of nitrate, copper, iron, and manganese contents compared to the Nhill-flat trial site. Faba bean has a well-developed taproot, which bears a profusion of fibrous roots in the top 30 cm of soil [34] and, hence, quality of topsoil may influence crop growth and seed quality.

The mean moisture content for the samples from duplex soil and sandy soil were 92.8 mg g⁻¹ and 94.2 mg g⁻¹, respectively. There was a significant effect of soil type on the concentration of vicine (p < 0.05) but not for convicine (p > 0.05) (Figure 5 and Figure 6). The mean vicine contents for duplex soil and sandy soils were 5.07 mg g⁻¹ DB and 5.36 mg g⁻¹ DB, respectively. The convicine contents for duplex soil and sandy soil were 2.70 mg g⁻¹ DB and 2.82 mg g⁻¹ DB, respectively. Compared to the effect of TOS, the effect of soil type on v-c contents of the analyzed samples was less distinct.

3.4. Solar Exposure During Podding and Its Impact on V-C Contents

The accumulated heat sum (AHS) and accumulated solar exposure (ASE) during the podding stage, mean 100 seed weight, and total v-c content are summarized in

Table 4. The accumulated heat sum (AHS) and solar exposure (ASE) during podding stage of various faba.

Trial	AHS During Podding	ASE During Podding (MJ m−2)	Mean Yield (t ha−1)	Mean 100 Seed Weight (g)	Mean Protein (mg g−1)	Sum of v-c (mg g−1 DB)
Curyo–TOS 1	326 A	647 A	2.63 A	68.1 A	281 a	6.6 A
Curyo–TOS 2	200 B	427 B	2.29 B	56.5 B	270 b	8.0 B
Nhill–Duplex soil	177 B	416 B	3.10 C	61.9 B	276 ab	7.8 B
Nhill–Sandy soil	172 B	396 B	3.53 D	58.4 B	281 a	8.2 B

Note. The means in the same column with different letters are significantly different (Tukey pairwise comparison, capital letter for p < 0.01 and small letter for p < 0.05).

. It was found that both AHS and ASE were inversely correlated with mean v-c contents in faba bean trials (r = −0.65 and r = −0.61, respectively) and positively correlated with 100 grain weight (r = 0.53 and r = 0.50, respectively). The negative correlation of AHS and v-c content is also demonstrated in the PCA loading plot of various measure parameters (Figure 7). Confalone, Lizaso, Ruiz-Nogueira, López-Cedrón, and Sau [35] similarly observed that field temperature and photoperiod had a large impact on faba bean growth, grain yield, and its component (e.g., seed weight, pod number, seeds per pod, and harvest index). Vicine and convicine are synthesized during the seed filling stage with the highest concentration in the early stages of seed development and decreases rapidly when a seed reaches physiological maturity [14]. Furthermore, Cardador-Martínez et al. [12] also found that faba beans with longer seed development had significantly lower v-c contents when compared with faba beans that had a shorter development. Thus, the observed inverse correlation of total v-c content with grain weight (r = −0.60) could be due to improved grain development and potentially could also be due to dilution effect because of seed growth.

3.5. Genotype by Treatment Interactions

The vicine and convicine contents differ significantly amongst the genotypes and treatments. Among all the samples analyzed (12 genotypes and 4 treatments), AF14092 grown at Curyo-TOS 1 had the lowest mean vicine content (3.27 mg g⁻¹ DB) and PBA Amberley samples from Nhill-Sand had the highest vicine content (5.87 mg g⁻¹ DB) (Figure 8). Similarly, Farah samples from “Curyo–TOS 1” had the lowest convicine content (2.00 mg g⁻¹ DB) and AF14075 samples from “Curyo–TOS 2” had the highest convicine contents (3.40 mg g⁻¹ DB). These data suggest that, by selection of the genotype and treatment, the v-c levels in the grains could potentially be reduced by ~40%. Overall, genotype and environment (TOS and soil type) had the most significant effect on v-c content (p < 0.01) but there was no interaction or additive effect due to genotype by environment. Similar to our findings, Pulkkinen et al. [36] also reported significant genotypic variation in v-c content in their studies, whereas environmental variation (growing years) was significant but comparatively minor.

3.6. A Study of Low V-C Lines

Due to the limited availability of seed of the low v-c lines, this study was unable to incorporate low v-c lines in the TOS and soil-type trials. Nevertheless, for a comparative analysis of v-c levels, three ultra-low v-c breeding lines were sourced from the University of Adelaide faba bean breeding program grown in a separate trial. These lines were then examined for their v-c contents, revealing vicine contents of 0.19 mg g⁻¹, 0.24 mg g⁻¹, and 0.26 mg g⁻¹ and convicine contents of 0.01 mg g⁻¹, 0.03 mg g⁻¹, and 0.01 mg g⁻¹, respectively. The v-c contents in low v-c samples are over 10 times lower than those observed in the normal v-c lines analyzed in this study. These findings align with the reported v-c levels in low v-c samples, as documented by Björnsdotter et al. [7]. In their in vivo study, Gallo et al. [37] observed that individuals with G6PD deficiency did not exhibit any indications of red blood cell damage or intravascular hemolysis even after consuming 500 g of raw low v-c faba bean seeds per 70 kg of body weight (equivalent to 5–10 times the amount of an average faba bean meal). The study by Debnath, Rai, Tyagi, Majumder, and Meetei [38] demonstrated that, in faba bean samples consisting of indigenous, exotic, landraces, and two commercial varieties, reduced vicine content was generally associated with reduced grain yields. The study concluded that complete elimination or substantial reduction in vicine could compromise yield potential unless an improved grain-yield selection strategy was employed when developing low-vicine varieties. The broader commercial availability and adoption of low-vicine faba bean varieties, which offer comparable yield and disease resistance to normal v-c varieties, could potentially expand faba bean consumption. This expansion may open the possibility of incorporating faba bean products into the diets of more than 400 million G6PD-deficient individuals.

4. Conclusions

As the global demand for alternative protein sources increase, faba beans are recognized as a sustainable option for plant-based protein. However, the presence of elevated concentrations of vicine and convicine (v-c) could substantially limit broader adoption. To address this, breeding programs are now prioritizing the development of new varieties with reduced concentrations of v-c. As with all plant metabolites, the synthesis of v-c is influenced by abiotic factors during seed development, which alter the metabolic pathways and affect the accumulation of both beneficial and anti-nutritional compounds.

This study demonstrated that agronomic factors—particularly the time of sowing—can significantly impact v-c concentration in faba beans. While post-harvest processing can reduce the v-c levels to an extent, the most effective strategy in the near term may cultivate low v-c lines under environmental conditions that minimize the accumulation of v-c concentration in the grain.

Until ultra-low or zero v-c varieties become widely available, optimizing both genotype selection and agronomic conditions remains essential for ensuring the safety and quality of faba bean as a food and feed source.