Advances in breeding for enhanced iron and zinc concentrations in common bean in eastern Africa

Abstract

Micronutrient malnutrition is one of the most serious health challenges facing vast sectors of the population of Africa, particularly resource-poor women and children. The main deficiencies include iron (Fe), zinc (Zn), and vitamin A. Plant breeding has frequently been advocated as the most sustainable strategy to provide varieties of different food crop species with enhanced micronutrient density to combat the global hidden hunger problem which affects >2 billion people. However, there are few research programmes which have implemented this approach, from concept stage to finished products, which can be widely disseminated and commercialized to create meaningful impact. The east African bean biofortification programme offers a case study of such a programme. The aim of this programme was to develop well-adapted, high-yielding, Fe- and Zn-rich bush and climbing bean cultivars and agronomic approaches that enhance expression of the high mineral trait. The objective of this review is to provide a synthesis of the progress made in the last 22 years, with a focus on genetic diversity, inheritance, bioavailability of Fe and Zn, and cooking quality, as well as to identify research gaps and suggest future directions.

This review provides a synthesis of research on developing new market-led bean varieties with 80% more iron and 60% more zinc in the last 22 years in eastern Africa.

Introduction

Micronutrient malnutrition is one of the most serious health challenges facing vast sectors of Africa’s population, particularly resource-poor women and children (Smith, 2000; Welch and Graham, 2000). The main deficiencies include iron (Fe), zinc (Zn), and vitamin A. The primary cause of these deficiencies are diets rich in energy but poor in proteins, minerals, and vitamins. This is aggravated by limited access to the more expensive animal-based products such as milk, eggs, and meat which are rich in vitamins and minerals. Data on micronutrient deficiencies in Africa for the period 1980 to early 1990s show very high prevalence rates of iron deficiency anaemia (IDA) amongst children, and non-pregnant and pregnant women (Smith, 2000). In east and central Africa, IDA prevalence rates among pregnant women varied from 6% in Ethiopia to 88% in Malawi. Among non-pregnant women, prevalence rates varied from 8% in Ethiopia to 69% in Burundi. IDA prevalence among children under 5 ranged from 14% in Malawi to 67% in Tanzania.

Zn deficiency was only recently recognized as a public health problem. In Kenya, a national micronutrient survey showed that Zn deficiency is a widespread problem that requires urgent attention (Ministry of Health, 2011). Zn is essential for normal growth, appetite, and immune function. It is an essential component of >100 enzymes involved in digestion, metabolism, and wound healing (Guzman-Maldonado, et al., 2003). Zn plays an important role in synthesis of retinal-binding protein, increasing lymphatic absorption of retinal and its intracellular and intercellular transport (Christian and West, 1998).

Intervention strategies in Africa

A three-pronged approach has been followed in alleviating the micronutrient deficiency problem in Africa. These are: supplementation of vulnerable groups with micronutrients; fortification of common foods; and dietary improvement. Mineral supplementation is effective for easy to reach vulnerable groups with access to medical facilities. This approach leaves out those hard to reach or practically unreachable at-risk groups as well as other community members not targeted to receive any kind of supplementation. In Africa, this group constitutes the majority who are located in rural communities with hardly any access to medical facilities and ill from the effects of deficiency.

Fortification of common foods has had a limited degree of success in Africa because of the underdeveloped food industry and lack of effective legislation. Fortification is effective for small affluent communities mostly in urban areas and households with a capacity to purchase fortified foods on a regular basis. This leaves out the majority urban poor and rural communities who rely on foods processed at home or in their locality.

Dietary improvement is probably the most effective and sustainable strategy for reducing micronutrient deficiencies in Africa. This approach aims to increase dietary availability, regular access, and consumption of mineral-rich foods in at-risk and micronutrient-deficient groups of populations (Zulfiqar et al., 2024a, b).

Why common bean?

Common bean offers unique opportunities for improving micronutrient nutrition because it is widely grown (>7 Mha annually in Africa), widely consumed, rich in protein (>20%), minerals, and calories, and is relatively cheap and highly marketable (Katungi et al., 2020; Huertas et al., 2022a). Common bean is probably the most important grain legume in east, central, and southern Africa. Some of the highest per capita bean consumption in the world has been reported in Africa. For example, per capita consumption is estimated at 50–60 kg year^–1 in Rwanda and 66 kg in the Kisii region in Kenya (Jaezold and Schmidt, 1983). Beans typically have four times more Fe and six times more Zn compared with maize (Mamiro et al., 2011). Moreover, common bean has been grown in Africa for >500 years since its introduction in the Sofala Coast of Mozambique by the Portuguese in the 16th century, and is culturally acceptable as a staple (Greenway, 1945; Wortmann et al., 1998). Today, beans are an important source not only of protein and minerals, but also of income for many rural and urban communities, an export crop, and a raw material for the food processing industry.

Breeding micronutrient-dense varieties

For the last two decades, considerable work has been done in eastern Africa to develop bean varieties which combine micronutrient density with important agronomic traits, cooking quality, organoleptic characteristics, and commercial grain types. This work started at the University of Nairobi, which also hosted and led the International Center for Tropical Agriculture (CIAT) regional bean breeding programme for east and central Africa from 2000 to 2010. Research on micronutrient bean varieties subsequently spread to the 10 countries [Burundi, DR Congo (north), Ethiopia, Kenya, Madagascar, Rwanda, Sudan, Tanzania, and Uganda] which constituted the East African Bean Research Network (ECABREN) supported by the Association of Strengthening Agricultural Research in East and Central Africa (ASARECA), and the 10 countries which constituted the Southern Africa Bean Research Network (SABRN). These were Angola, Malawi, DR Congo (south), Malawi, Zambia, Zimbabwe, Swaziland, Lesotho, Botswana, South Africa, and Mauritius. Activities spread to West Africa from 2007, when the West African Bean Research Network (WECABREN) was formed. The objective of this paper is to review and provide a synthesis of the progress made to develop biofortified bean varieties in the last 22 years.

Programme objective and strategy

The main objective of the regional bean biofortification programme was to develop well-adapted, high-yielding, Fe- and Zn-rich bean cultivars with resistance to major biotic and abiotic stress factors (CIAT, 2008a). A strategy was formulated to develop and disseminate micronutrient-dense varieties in east, central, southern, and later west Africa (CIAT, 2008b). This was a multistage process that involved collection, screening local and introduced germplasm for grain Fe, Zn, and protein concentrations, agronomic potential, genotype×environmental interactions through participatory selection, preliminary, intermediate, advanced, national, and regional yield trials, variety release, and seed production and dissemination (Fig. 1). A second stream of this strategy focused on transferring the high mineral trait and resistance to biotic and abiotic stress factors to commercial cultivars. Complementary studies were conducted to determine inheritance of the high mineral trait, and nutritional quality including cooking time, taste, and mineral retention, since little was known about these aspects. Progress made in these aspects is reviewed briefly in the next section.

Fig. 1.

Strategy for development and dissemination of micronutrient-dense lines in east, central, and southern Africa.

Germplasm collection

Although, previous work at CIAT had demonstrated that there was adequate variation for enhancing the Fe and Zn concentrations in common bean (Beebe et al., 2000), little was known about the variation for these nutrients in African bean germplasm, which is significant because the highlands of eastern Africa are considered a secondary centre of diversity for common bean (Broughton et al., 2003; Asfaw et al., 2009). Consequently, one of the first activities of the regional programme which started in 2004 with the support of ASARECA was to: (i) characterize the variation of grain Fe and Zn concentrations; (ii) identify potential parents for further breeding work; (iii) determine whether there were regions with high diversity for the high mineral trait, which could complement existing collection at the Genetic Resources Unit of CIAT, Colombia; and (iv) identify lines which could be fast-tracked as mineral-dense lines for cultivation by farmers in regions with severe Fe and Zn malnutrition. By December 2007, a total of 2830 accessions had been collected, partially or completely characterized for some agronomic traits, and seed increased to facilitate mineral analyses.

Mineral analysis

Mineral analysis was carried at the University of Nairobi in Kenya, CIAT (Cali, Colombia), Cornell University (USA), University of Copenhagen (Denmark), Sokoine University of Agriculture (Tanzania), and Swiss Federal Institute of Technology Zurich (ETH) to facilitate cross-laboratory comparisons (CIAT, 2008b). In Kenya, mineral analyses were performed by ashing and wet digestion techniques (Zarcinas et al., 1987). Results of germplasm screening showed that grain Fe concentration varied from 61 mg kg^–1 in M’Sole to 147 mg kg^–1 for AND 620 (CIAT, 2001). Zn concentration varied from 12 mg kg^–1 in M211 to 62 mg kg^–1 in VNB 87010. Two genotypes, Nakaja and AND 620, combined high Fe and high Zn traits. The 38 most promising lines with >70 mg kg^–1 Fe and >30 mg kg^–1 Zn were also analysed at Cornell University in 2004. Results showed that the Fe concentration varied from 56 mg kg^–1 in K131 to 109 mg kg^–1 for Roba-1. Thirty-three of the 38 lines had Fe levels above the 70 mg kg^–1 baseline, confirming results of previous analyses. Zn concentration of the 38 lines varied from 30.1 mg kg^–1 in MLB 49 89A to 44.9 mg kg^–1 in Ituri Matata. All 38 lines had Zn levels above the baseline of 30 mg kg^–1. These results corroborated previous analyses at the University of Nairobi. The results of ETH Switzerland were also comparable with those reported from Cornell University and the University of Nairobi. Thilsted (2007) reported that Fe concentration in 50 genotypes from Kenya and DR Congo varied from 55 mg kg^–1 (BRB 194) to 127.6 mg kg^–1 (Chivuzo). Zn concentration varied from 29.5 mg kg^–1 (RWR 10) to 45.7 mg kg^–1 (Chivuzo). Forty-five of the 50 genotypes had an Fe concentration above the baseline of 70 mg kg^–1. Forty-eight of the 50 lines had a Zn concentration above the baseline of 30 mg kg^–1.

Blair et al. (2007) evaluated 211 genotypes grown in eastern and southern Africa for grain Fe and Zn concentrations. The microsatellites used in the analysis efficiently separated the two genepools and identified intermediate genotypes. This study confirmed previous studies showing that genotypes from this region had ample diversity for improvement of nutritional quality characteristics such as Fe and Zn content. Blair et al. (2010) analysed another 365 bean landraces from the Great Lakes region of central Africa for grain Fe and Zn, and concluded that Central African bean varieties are a source of wide genetic diversity with variable nutritional quality that can be used in crop improvement programmes for the region.

These studies implied that considerable genetic variation existed to facilitate improvement of Fe by >90% and Zn by >60%. Although mineral density was affected by the growing conditions (soil type and soil nutrients) and methods of analyses, the results showed that most of the lines consistently showing high levels of micronutrients originated from the Great Lakes region, especially DR Congo and Rwanda (Lubobo et al., 2021; Kataliko et al., 2023, 2024). Based on these analyses, 38 lines with >70 ppm Fe and/or >25 ppm Zn were selected for further work. These lines were referred to as ‘fast-track nursery’, which later became the source of the first-generation biofortified bean varieties in Africa.

Participatory evaluation of fast-track lines

Bean research programmes in eastern and southern Africa have applied participatory breeding methods since the early 1990s to ensure genotypes advanced to release stage were adapted to local growing conditions and had as many traits preferred by farmers, local markets, and consumers as possible (Sperling et al., 1993; Kimani and Anthony, 2019). The 38 fast-track lines were distributed to collaborating countries in east, central, and southern Africa for on-farm and on-station agronomic evaluation in 2004 and 2005.The trials were conducted by the national bean programmes of Rwanda, Uganda, Tanzania, Ethiopia, Burundi, DR Congo, and Malawi. The lines were sown in observation nurseries to increase seed, followed by participatory on-station and on-farm evaluations and selection (CIAT, 2005).

Characteristics of fast-track lines

The growth habit, seed size, seed colour, and Fe and Zn concentrations of the fast-track lines are shown in Table 1. All four main growth habits of common bean are represented in this nursery. Thirty-two were bush type with either type I or type II growth habit. Five had type IV growth habit and one type III. Type IV are climbing beans which have a 3:1 yield advantage compared with bush types. They have variable seed sizes: 15 are large seeded, characteristic of the Andean genepool and the other 23 have small and medium size seeds, characteristic of the Mesoamerican genepool. They also represent the major market classes, seed sizes, and popular grain types which are commercially important in east, central, and southern Africa. For example, the large seeded red mottled grain type is the most important market class in Kenya, Uganda, and Malawi; small red and small white types are important in Ethiopia; large white types are important in Madagascar (Kimani et al., 2005a, b); brown and tan grain types are important in DR Congo; and black seeded types are popular in northern Uganda and the south-western region of Ethiopia. Growth habit, grain type, and size are key factors which can lead to the acceptance or rejection of a new variety by farmers and consumers.

Table 1.

Seed colour, seed size, and iron and zinc concentrations of fast-track bean lines (Fe and Zn data from Lunjalu, 2007).

Variety	Growth habita	Seed sizeb	Seed colourb	Genepoolc	Fe (g kg–1)	Zn (g kg–1)
AFR 708	I	Large	Red mottled	A	78.1	35.7
AND 620	I	Large	Red mottled	A	101.5	43.9
Awash Melka	I	Small	White	M	80.8	33.8
G59/1-2	IV	Large	Red	A	94.8	30.9
GLP-2	I	Large	Red mottled	A	92.3	36.0
GLP-92	I	Medium	Pinto	M	79.1	39.2
Gofta	II	Medium	Brown	M	81.9	36.3
HRS 545	II	Small	White	M	78.4	37.5
Ituri Matata	II	Large	White	A	71.3	37.9
Jesca	I	Large	Purple	A	107.1	33.7
K131	II	Small	Carioca	M	76.7	34.9
K132	I	Large	Red mottled	A	94.1	37.3
OBA-1 (NABE 1)	I	Large	Red mottled	A	74.0	31.5
Kiangara	IV	Medium	Yellow/brown	M	77.0	29.7
Kirundo	II	Medium	Yellow	M	84.9	33.7
LIB-1	IV	Medium	Yellow	M	74.5	31.7
Lingot Blanc	II	Large	White	A	83.6	37.2
Maharagi Soja	II	Small	Brown/yellow	M	90.1	34.2
Maasai Red	I	Small	Red	M	82.9	40.3
MCM 2001	I	Small	Red	M	80.7	41.0
Mexican 142	II	Small	White	M	84.5	39.2
MLB-49-89A	I	Medium	Black	M	106.9	31.2
Mwa Mafutala	II	Small	Brown	M	105.9	34.2
Nain de Kyondo	I	Small	White	M	102.2	41.2
Nakaja	II	Small	Brown	M	83.7	41.0
Nguaku Nguaku	I	Large	Brown/yellow	A	76.3	28.8
PVA 8	I	Large	Red mottled	A	87.7	31.4
Ranjonomby	II	Large	White	A	70.0	34.9
Red Wolaita	II	Small	Red	M	74.1	37.2
Roba-1	II	Small	Brown/tan	M	93.6	37.0
RWR 10	I	Large	Red	A	81.1	38.2
Selian 97	I	Large	Red	A	80.8	37.5
Simama	I	Large	Red mottled	A	90.6	37.6
Soya Fupi	II	Medium	Purple	M	96.6	27.0
TY3396-12	III	Medium	Carioca	M	104.5	33.3
VCB 81013	IV	Small	White	M	102.6	34.4
VNB 81010	IV	Medium	Black	M	97.2	43.4
Zebra	II	Medium	Cream zebra	M	73.4	37.1
Mean					75.3	31.2
LSD (05)					2.8	1.6
CV (%)					3.2

^a Growth habit: I, bush determinate; II, bush indeterminate; III, semi-climbing; and IV, climbing.

^b Large (>40 g/100 seeds); medium (26–39 g/100 seeds); small (<25 g/100 seeds).

^c A, Andean; M, Mesoamerican genepools.

Reaction of fast-track lines to diseases

Diseases are one of the most important constraints to bean productivity in Africa (Wortmann et al., 1998). Major constraints include fungal diseases such as angular leafspot (Pseudocercospora griseola), anthracnose (Colletotrichum lindemuthianum), and root-crown rots (Pythium spp and Fusarium spp); bacterial diseases, especially common bacterial blight (Xanthomonas axonopodis pv phaesoli) and halobight (Pseudomonas savastanoi pv. phaseolicoli); and viral diseases such as bean common mosaic virus (BCMV) and bean common mosaic necrotic virus (BCMNV). It is therefore critical to screen new germplasm for reaction to the prevalent diseases and identify those that are resistant or tolerant. The biofortified fast-track lines were evaluated for reaction to diseases in trials conducted in Kenya, Rwanda, Uganda, and Malawi (Chirwa et al., 2005; Wagara and Kimani, 2006). Data from these studies showed that some of the fast-track lines have good levels of resistance to major diseases but are susceptible to others. Consequently, these lines can be grown by farmers subject to epidemics of the diseases to which they are susceptible.

Nutritional evaluation of fast-track lines

Consumer acceptability of the new varieties depends on not only their agronomic potential, but also their cooking and organoleptic qualities (Kimani, 2017; Losa et al., 2022). Moreover, their nutritional value partly depends on the retention of the target nutrients after cooking and their bioavailability (Frohlich, 1995). Little was known about these attributes for the micronutrient-dense bean lines. Several studies were conducted to determine the effects of soaking on cooking time, retention of Fe, Zn, and protein, and mineral bioavailability of the 38 fast-track lines (Lunjalu, 2007; Maryange et al., 2010; Mamiro et al., 2011, 2012, 2016).

Cooking time

These studies showed highly significant differences (P<0.001) in cooking time among fast-track lines and soaking treatments (Table 2). Cooking time for unsoaked beans ranged from 106.5 min for Kirundo, to 220 min for AND 620. Ngwira and Mwangwela (2001) reported a cooking time range of 88.4–209 min in freshly harvested beans cooked in borehole water and a range of 61–132 min of the same varieties cooked in treated tap water. Soaking significantly reduces cooking time, but the magnitude varies with genotypes. On average, soaking reduced cooking time of the fast-track lines by 33%. Reduction in cooking time varied from 0 (Nain de Kyondo) to 58.6% (AND 620). Awash Melka had the shortest cooking time when soaked, while GLP 92 had the longest cooking time. Previous studies have shown that cooking time varies with the type of water used (Ngwira and Mwangwela, 2001; Kimani et al., 2017). Use of deionized water results in faster cooking time because it is free of monovalent and divalent ions. Shorter cooking time saves on fuel costs and time (Elia et al., 1997; Wiesinger et al., 2018; Wood, 2016).

Table 2.

Cooking time of fast-track biofortified bean lines (data from Kimani et al., 2007).

Genotype	Time (min)		% cooking time reduction due to soaking
	Soaked	Unsoaked
AFR 708	93.5	165.0	43.3
AND 620	91.0	220.0	58.6
Awash Melka	75.0	111.5	32.7
G59/1-2	107.5	155.0	30.7
GLP-2	92.5	161.0	42.6
GLP-92	132.5	163.5	18.9
Gofta	112.5	209.5	46.3
HRS 545	120.0	160.5	25.2
Ituri Matata	93.0	131.5	29.3
Jesca	112.5	161.0	30.1
K131	122.5	169.0	27.5
K132	82.5	141.0	41.5
Kiangara	80.0	125.0	36.0
Kirundo	87.5	106.5	17.8
LIB-1	80.0	147.5	45.8
Lingot Blanc	92.0	107.5	14.4
Maharagi Soja	123.0	162.5	24.3
Maasai Red	117.5	170.0	30.9
MCM 2001	102.0	155.0	34.2
Mexican 142	92.5	156.0	40.7
MLB-49-89A	90.0	175.0	48.6
Mwa Mafutala	77.5	140.0	44.5
Nain de Kyando	110.0	110.0	0.0
Nakaja	110.0	150.0	26.7
Nguaku Nguaku	112.5	132.5	15.1
PVA 8	107.1	157.5	32.0
Ranjonomby	87.5	120.0	27.1
Red Wolaita	117.0	175.0	33.1
Roba-1	145.0	150.0	3.3
RWR 10	115.0	139.5	17.6
Selian 97	105.0	127.5	17.7
Simama	90.0	152.5	40.9
Soya Fupi	102.5	197.5	48.1
TY3396-12	122.5	158.5	22.7
VCB 81013	110.0	190.0	42.1
VNB 81010	116.0	170.0	31.8
Zebra	85.0	147.5	36.8
a LSD0.05	7.7
CV (%)	4.2

^a LSD for comparisons between varieties; LSD for comparing treatment means is 1.8.

Mineral retention

The amount of Fe retained after cooking varied from 71.4% in VNB 81010 to 99.6% in Ituri Matata (Table 3). On average, genotypes retained 87% of the grain Fe after cooking. These values are within the range reported by other researchers. Augustin et al. (1981) reported Fe retention values of 94.5–103.2% for nine market classes of dry beans. The effects of soaking and cooking on grain Zn concentration differed with genotypes (Table 4). Zn retention ranged from 69.0% to 97.3% in cooked samples, with a mean of 87.0%. The nutrient reduction in cooked and soaked beans was attributed to leaching into soaking and cooking water. Augustin et al. (1981) reported Zn retention values of 77–105% (DW basis) for nine bean classes. However, soaking has been shown to be beneficial in lowering anti-nutritional factors in beans. Conventional soaking followed by cooking was found to reduce phytate, phenolic acids, and flatulence caused by stachyose, raffinose, and verbascose (Iyer et al., 1980; Loggerenberg, 2004; Ritho et al., 2023).

Table 3.

Iron concentration in raw, cooked, and soaked and cooked beans of the fast-track biofortified common bean genotypes and their retention values (data from Kimani et al., 2007)

Genotype	Fe, mg kg–1a			% Fe retention
	Raw	Cooked	Soaked and cooked	Cooked	Soaked and cooked
AFR 708	78.1	72.5	67.3	92.8	86.2
AND 620	101.5	79.6	72.3	78.4	71.2
Awash Melka	80.8	66.6	71.1	82.4	88.0
G59/1-2	94.8	79.5	68.9	83.9	72.7
GLP-2	92.3	82.1	70.6	88.9	76.5
GLP-92	79.1	69.4	64.6	87.7	81.7
Gofta	81.9	76.3	71.2	93.2	86.9
HRS 545	78.4	73.0	67.0	93.1	85.5
Ituri Matata	71.3	70.9	67.3	99.6	94.5
Jesca	107.1	97.1	68.6	90.7	64.0
K131	76.7	66.9	65.6	87.2	85.5
K132	94.1	71.8	70.6	76.3	75.0
Kiangara	77.0	69.7	68.8	90.5	89.3
Kirundo	84.9	66.6	67.8	78.4	79.8
LIB-1	74.5	67.4	51.8	90.4	69.4
Lingot Blanc	83.6	76.8	74.6	91.8	89.2
Maharagi Soja	90.1	70.0	61.2	77.7	67.9
Maasai Red	82.9	78.3	64.8	94.4	78.1
MCM 2001	80.7	79.3	70.3	98.3	87.1
Mexican 142	84.5	74.0	75.2	87.6	88.9
MLB-49-89A	106.9	92.6	65.6	86.6	61.4
Mwa Mafutala	105.9	83.9	76.4	79.2	72.1
Nain de Kyondo	102.2	86.0	76.4	84.1	74.7
Nakaja	83.7	74.8	59.5	89.4	71.0
Nguaku Nguaku	76.3	68.0	69.2	89.2	90.7
PVA 8	87.7	67.8	67.3	77.3	76.8
Ranjonomby	70.0	61.9	57.3	88.4	81.8
Red Wolaita	74.1	71.9	68.2	97.0	92.0
Roba-1	93.6	84.0	86.8	89.8	92.8
RWR 10	81.1	74.5	66.8	91.8	82.3
Selian 97	80.8	77.8	69.3	96.3	85.8
Simama	90.6	84.5	67.4	93.3	74.4
Soya Fupi	96.6	87.0	60.6	90.1	62.7
TY3396-12	104.5	75.2	61.7	71.9	59.1
VCB 81013	102.6	74.5	62.9	72.6	61.3
VNB 81010	97.2	69.4	70.6	71.4	72.6
Zebra	73.4	64.1	60.4	87.3	82.2
Mean	75.3	87.1	67.7	87.0	78.7

^a LSD_0.05 was 2.8 (treatments and genotypes); CV (%)=3.2.

Table 4.

Zinc concentration in raw, cooked, and soaked and cooked beans of fast-track biofortified common bean genotypes and their percent retention values (data from Kimani et al., 2007)

Genotype	Zn, mg kg–1			% Zn retention
	Raw	Cooked	Soaked and cooked	Cooked	Soaked and cooked
AFR 708	35.7	30.2	28.5	84.6	80.0
AND 620	43.9	30.3	27.7	69.0	63.0
Awash Melka	33.8	31.2	24.7	92.3	73.0
G59/1-2	30.9	27.9	21.5	90.5	69.8
GLP-2	36.0	33.5	27.0	93.0	74.9
GLP-92	39.2	32.4	24.0	82.7	61.2
Gofta	36.3	34.4	28.3	94.8	78.0
HRS 545	37.5	32.3	27.9	86.2	74.5
Ituri Matata	37.9	33.2	29.6	87.7	78.1
Jesca	33.7	29.2	21.1	86.6	62.4
K131	34.9	30.9	25.1	88.3	71.9
K132	37.3	31.9	29.4	85.6	78.9
Kiangara	29.7	27.5	25.2	92.5	84.9
Kirundo	33.7	30.4	30.4	90.0	90.2
LIB-1	31.7	29.1	22.9	91.9	72.2
Lingot Blanc	37.2	33.3	24.9	89.6	67.1
Maharagi Soja	34.2	30.4	27.2	88.8	79.3
Maasai Red	40.3	31.8	37.1	79.1	92.2
MCM 2001	41.0	36.6	27.3	89.3	66.7
Mexican 142	39.2	38.2	27.9	97.3	71.1
MLB-49-89A	31.2	25.5	24.8	81.7	79.5
Mwa Mafutala	34.2	31.2	24.8	91.1	72.4
Nain de Kyondo	41.2	33.1	25.5	80.2	61.9
Nakaja	41.0	35.7	24.9	87.0	60.8
Nguaku Nguaku	28.8	26.5	22.9	92.0	79.3
PVA 8	31.4	30.2	23.2	96.2	74.0
Ranjonomby	34.9	32.6	23.0	93.6	66.0
Red Wolaita	37.2	29.7	26.5	79.7	71.1
Roba-1	37.0	31.0	26.4	84.0	71.3
RWR 10	38.2	34.4	26.5	90.0	69.5
Selian 97	37.5	35.4	29.5	94.5	78.6
Simama	37.6	31.3	28.5	83.1	75.8
Soya Fupi	27.0	23.3	20.1	86.4	74.5
TY3396-12	33.3	27.6	22.6	82.7	68.0
VCB 81013	34.4	26.7	20.4	77.8	59.3
VNB 81010	43.4	34.8	29.3	80.3	67.5
Zebra	37.1	29.1	24.9	78.5	66.9
Mean	31.2	35.9	26.0	87.0	72.6
LSD (05)	1.6
CV (%)	4.4

Bioavailability

Bioavailability is the fraction of the ingested nutrient that is utilized for normal physiological function or storage (King et al., 2000; Welch and Graham, 2004). Bioaccessibility is the amount of ingested nutrient that is released from the food matrix and potentially available for absorption and physiological function (Huertas et al., 2022a, b; Sulaiman et al., 2021). An in vitro method was used to determine Fe and Zn bioavailability in dry beans and green shelled beans of the 38 fast-track lines because it is a rapid, relatively cheap, and convenient method of preliminary screening to identify promising plant sources before human trials (Mamiro et al., 2016; Sulaiman et al., 2021). Bioavailable Fe in raw dry beans varied from 1.1% to 6.6%, with a mean of 3.2%, compared with 1.7–6.8% and a mean of 3.4% in cooked beans. Cooking increased bioavailable Fe in nearly all cases. In general, bioavailable Fe was higher in green shelled beans (6.2%) compared with dry beans (3.3%). In green shelled beans, bioavailable Fe varied from 1.9% to 8.9% in raw beans, and from 3.9% to 16.5% in cooked beans. Cooking increased bioavailable Fe to 7.4% in green shelled beans compared with 3.4% in cooked dry beans, suggesting that it is nutritionally advantageous to consume green shelled beans. This also implies that Fe bioavailability can be enhanced by selection.

Cooking enhanced bioavailability of Zn in dry beans. Bioavailable Zn increased from 1.3% in raw dry beans to 1.6% in cooked beans. There was little difference in bioavailable Zn in cooked beans of Andean (1.7%) and Mesoamerican (1.5%) groups. Green shelled beans have more bioavailable Zn compared with dry beans of the same genotypes. Bioavailable Zn was 2.4% in raw green shelled beans compared with 1.3% in raw dry beans of the 38 genotypes. Cooking enhanced bioavailable Zn in shelled green beans. Bioavailable Zn in the 38 genotypes increased from 2.4% in raw green shelled beans to 3.8% after cooking. Bioavailable Zn in cooked green shelled beans varied from 2.5% in Red Wolaita to 6.3% in Maasai Red, suggesting that this trait can be improved through selection.

Genetics of the high mineral trait

As interest in developing micronutrient-rich varieties of major crops grew during the first decade of the 21st century, questions arose on the nature of the genetic mechanism controlling this trait. Studies reported by several authors showed that genetic variation existed for crops such as beans (Beebe et al., 2000; Welch et al., 2000), maize (Banziger and Long, 2000; Maziya-Dixon et al., 2000), wheat (Cakmak et al., 2000; Monasterio and Graham, 2000); cassava (Chavez et al., 2000; Maziya-Dixon et al., 2000), rice (Gregorio et al., 2000), and sweet potato (Hagenimana and Low, 2000). For a breeding programme, it is necessary not only to establish the presence of genetic diversity for trait(s) of interest, but also to determine their mode of inheritance. Consequently, to develop micronutrient-rich bean varieties in eastern Africa, we determined that adequate genetic diversity exists for grain Fe and Zn accumulation, as described in the previous section. Early studies in screening for variation for Fe and Zn showed that bean genotypes grown under uniform conditions in a greenhouse at Kabete tended to show lower levels of Fe and Zn than the original samples grown under diverse conditions in their country of origin, although their relative rankings remained (CIAT, 2003). This implied that mineral density was influenced by environment and showed genotype×environment interactions, and, most probably, was quantitatively inherited. This was later confirmed in studies reported by Blair et al. (2007, 2009a, b). Recombinant inbred lines (RILs) were developed from two mapping populations: Mesoamerican×Mesoamerican (G14519×G4825) and Andean×Andean genotype (G21242×G21078). About 110 Mesoamerican RILs and 100 Andean RILs were analysed for Fe and Zn concentration. Genetic maps were constructed and analysed for quantitative trait loci (QTLs) controlling mineral content using single-point regression and composite interval mapping analysis. Three Fe QTLs were identified in chromosomes 1, 9, and 10 in G21242 (Table 5). Another three QTLs in G21242 and one in G21078 could account for Zn concentration in these genotypes. Zn QTLs in G21242 were located in chromosomes 1, 2b, and 10. A Zn QTL in G21078 was located in chromosome 2. Microsatellite mapping of the two populations confirmed that Fe QTLs were located in chromosomes 1, 9, and 10 (Fig. 2). Zn QTLs were found in chromosomes 2 and 8. Three QTLs located in chromosomes 5, 8, and 10 jointly controlled Fe and Zn accumulation in bean genotypes. This could possibly explain the positive correlation between Fe and Zn concentration reported by several authors in different crops: Welch et al. (2000), Beebe et al. (2000), and Cichy et al. (2009) in beans; Maziya-Dixon et al. (2000) in cassava; Cakmak et al. (2000) in wheat; and Monasterio and Graham (2000) in rice.

Table 5.

QTLs for iron and zinc content identified in the Andean population G21242×G2178 with both simple and composite interval mapping (data from Blair et al., 2007)

QTL	Chromosome	Marker	Method	LOD	Source	Additivity	R 2
Fe1	10	BM157	CIM	3.80	G21242	4.35	14.19
	10	W1201A	CIM	3.77	G21242	4.19	12.96
	10	H1801A	CIM	2.65	G21242	3.38	8.50
	10	AI1401A	IM	3.03	G21242	4.35	14.85
	10	L0204A	IM	3.01	G21242	4.26	14.21
Fe2	1	L0401A	CIM	2.62	G21242	3.26	7.98
	1	R0402A	IM	2.88	G21242	6.55	32.40
	1	P0901A	IM	2.79	G21242	4.83	17.00
Fe3	9	M1201B	CIM	2.50	G21242	3.73	9.29
Zn1	1	P0901A	IM	4.02	G21242	2.48	24.22
	1	R0402A	IM	3.86	G21242	3.15	37.70
	1	L0401A	IM	3.74	G21242	2.30	21.42
Zn2	8	I0601A	CIM	3.94	G21078	1.75	11.97
	8	CLON638	CIM	3.73	G21078	1.72	11.73
Zn3	2b	BM156	IM	2.75	G21242	2.45	24.38
Zn4	10	W1201A	IM	2.50	G21242	1.82	13.21

IM, simple interval mapping; CIM, composite interval mapping.

Fig. 2.

Iron and zinc QTLs on the genetic maps of two common bean populations (Blair et al., 2007).

Cichy et al. (2009) found that QTLs for seed Fe and Zn co-localized on three linkage groups: B1, B6, and B11 in F_5.7 Andean RILs of AND 696 and G19833. The QTL on B1 accounted for 20–34% of the phenotypic variation for Fe, and 10–27% of the variation for Zn. The QTL on chromosome B6 explained 14–36% of the phenotypic variation for Fe and 12–39% for Zn. The QTL on chromosome B11 explained 12–23% of the variation for Fe and 9–15% for Zn. The authors concluded that seed Fe and Zn concentrations were quantitatively inherited. This has been confirmed by other studies (Guzman-Maldonado et al., 2003; Gelaw et al., 2023; Lampteya et al., 2023). However, in a previous study, Cichy et al. (2005) found that a single dominant gene controlled high seed Zn concentration in navy beans. Blair et al. (2011) found that inheritance of Fe and Zn accumulation in the Mesoamerican population was quantitative, but with much of the variation explained by a single locus that was common to both minerals. They concluded that the inheritance within the Mesoamerican genepool is simpler than that in inter-genepool crosses (Blair et al., 2011) and different from that in the Andean genepool (Cichy et al., 2009). The existence of co-localizing QTLs and correlation of Fe and Zn concentrations suggested that at least some of genes controlling both minerals are linked.

Genotype×environment interactions

Understanding the possible effects of environmental factors on mineral stability is a key issue in the development and release of micronutrient-rich bean varieties. This is important considering that beans are grown in very diverse environments from lowlands to high altitude agroecological zones (Jaetzold et al., 2010a, b). Wortmann et al. (1998) described 14 African bean production environments (AFBEs) differing in altitude, latitude, rainfall distribution, soil pH, major soil types, and day length. The first indication that micronutrient density may be influenced by environmental factors was the observation that Fe and Zn concentrations in seed samples received from several countries in east and central Africa were generally higher compared with levels obtained when these lines were grown in uniform conditions in a greenhouse at Kabete Field Station (CIAT, 2003). The differences were attributed to variation in environmental factors such as soil nutrient status, moisture availability, and pH, which are known to differ in bean-growing regions where the samples originated (Wortmann et al., 1998; Farrow and Muthoni-Andriatsitohaina, 2020). This implied that mineral concentration was determined by both genetic and environmental factors under field conditions and by genetic factors in uniform greenhouse conditions. More evidence that environmental factors influenced mineral density in common bean was obtained when 72 lines were evaluated in 40 on-farm trials in Marani and Suneka divisions of Kisii County for three seasons (CIAT, 2005). Fe concentration varied with seasons and farms, indicating that this trait is influenced by environmental factors. CIM 9314-13, a pureline variety developed in Malawi, illustrates variation in mineral concentration among the farms. Grain Fe concentration of this variety varied from 75 mg kg^–1 in Nyakundi’s farm to 107 mg kg^–1 in Bosire’s farm during the SR2003 cropping season. Zn concentration varied from 22.5 mg kg^–1 in Nyakundi farm to 32.5 mg kg^–1 in Joyce Onyingo farm in Marani, and in Kwamboka Ayieko’s farm in Suneka. Since CIM 9314-13 is pureline, the variation among farms could only be attributed to environmental or non-genetic factors. Results of this study stimulated further research in subsequent years to determine the role of macronutrients such as nitrogen (N), phosphorus (P), and potassium (K), micronutrients such as Fe and Zn, soil pH, and manures on grain Fe and Zn concentrations in common bean (Okonda, 2008; Felix, 2009). These studies showed significant location, season, treatment, and genotypic effects, and genotype×environment interactions on the grain Fe and Zn concentrations (Mutari et al., 2022). They implied that adequate levels of N, P, and K have the potential to increase bean yields, and seed Fe and Zn concentrations. However, the significant genotype×environment interactions suggest that grain mineral density should be expected to vary with locations and genotypes (Lubobo et al., 2021).

Release of the first biofortified bean varieties from a fast-track nursery

Agronomic performance is major consideration for the adoption of the lines by farmers and approval for release by national regulatory agencies. Therefore, the fast-track lines were subjected to adaptability trials in all ECABREN and SABRN countries (Chirwa et al., 2005; CIAT, 2005). Nearly all countries in east, central, and southern Africa have since released one or more varieties from this nursery (Mukankusi et al., 2018; Muthoni-Andriatsitohaina et al., 2020). For example, seven varieties were released and gazetted in Kenya. Four were bush beans with type I and II growth habit (Rosecoco Madini, Kenya Maua, Kenya Cheupe, and Kenya Almasi), and three are climbing beans with type IV growth habit (Kenya Afya, Kenya Madini, and Kenya Majano).

Welch and Graham (2004) suggested five criteria which must be met to ensure that a micronutrient-enriched variety is beneficial to people at risk of developing micronutrient malnutrition. A good micronutrient variety should: (i) increase productivity or at least maintain existing productivity levels; (ii) have a micronutrient concentration to significantly impact on human heath; (iii) have micronutrient traits which are relatively stable across various edaphic environments and climatic zones; (iv) have a high or acceptable level of mineral bioavailability; and (v) have organoleptic traits such as taste and cooking quality acceptable to consumers to ensure significant impact on nutritional health. For beans, we add that such a variety must also be marketable, because farmers grow them not only for household consumption but also for income generation. The biofortified bean varieties released from the fast-track nursery in Kenya (Fig. 3) and elsewhere meet these criteria. For example, the four bush biofortified varieties in Kenya showed outstanding agronomic performance in the national performance trials conducted independently by the regulatory agency, Kenya Plant Health Inspectorate Service (KEPHIS), at 10 locations in different agroecological zones for two seasons (CIAT, 2008a, b). In these trials, Rosecoco Madini showed an average yield advantage of 45% compared with the check varieties. The yield of this variety varied from 1150 kg ha^–1 to 2030 kg ha^–1. Kenya Maua had yield advantage of 35% over the check varieties. Its grain yield across agroecological zones varied from 1010 kg ha^–1 to 1890 kg ha^–1. Kenya Cheupe showed an exceptionally good performance, with an average yield advantage of 98% over the checks. It had a yield range of 1130–2800 kg ha^–1. Kenya Almasi had a yield advantage of 45% over the check varieties across the agroecological zones. It had a yield range of 1030–1960 kg ha^–1.

Fig. 3.

Grain types of biofortified bean varieties released in Kenya.

The three biofortified climbing bean varieties released in Kenya were Kenya Afya, Kenya Majano, and Kenya Madini (Kimani, 2016). These varieties showed significantly better yield in national performance trials conducted in seven locations over 2 years (CIAT, 2008a, b). Kenya Afya had a yield advantage of 46% over the checks. It had a mean yield of 2230 kg ha^–1 across test agroecological zones, Kenya Majano had a yield advantage of 44% and a mean yield of 2200 kg ha^–1, and Kenya Madini showed a yield advantage of 40% and a mean yield of 2150 kg ha^–1. These results indicate that all seven varieties met the productivity criteria of a biofortified variety proposed by Welch and Graham (2004).

Both in vitro and in vivo methods have been used to determine Fe and/or Zn bioavailability in common bean (King et al., 2000; Welch et al., 2000; Donangelo et al., 2003; Ariza-Nieto et al., 2007; Petry et al., 2014; Tako et al., 2015; Haas et al., 2016; Mamiro et al., 2016; Glahn et al., 2017; Wiesinger et al., 2019; Sulaiman et al., 2021). The studies provide evidence that Fe-biofortified bean varieties have a nutritional benefit, and delivered more absorbable Fe. The Fe-biofortified beans improved indicators of physical performance and Fe status, including maximum aerobic power (VO₂max), haemoglobin, serum ferritin, and total body Fe. It can be concluded that Fe-biofortified bean varieties remain a promising vehicle for increasing intake of bioavailable Fe in African populations that consume these beans.

Breeding second-generation micronutrient-dense bean varieties

The mineral concentration target of the first-generation varieties (70 mg kg^–1 Fe and 30 mg kg^–1 Zn) was largely met, and in some cases exceeded. Several varieties were released from the fast-track nursery in east, central, and southern Africa (Kimani, 2016; Mukankusi et al., 2018; Muthoni-Andriatsitohaina et al., 2020). Mineral analysis and agronomic evaluation of the >2800 bean accessions revealed that some genotypes had high mineral density but poor agronomic traits. In other cases, there were released varieties with some desired agronomic traits but deficient in others such as high mineral density, resistance to major diseases, and drought tolerance. The challenge was to combine desirable traits of these two genotypic categories and develop a new generation of more productive and nutritious bean varieties.

Breeding second-generation micronutrient-dense varieties started in 2005 at the University of Nairobi (Kimani and Warsame, 2019; Mondo et al., 2019). The objective of this programme was to combine the high mineral trait with resistance to major biotic stresses including regionally important diseases such as angular leaf spot, anthracnose, and root rots, and abiotic stresses, especially tolerance to drought, and with other farmer-preferred traits using the gamete selection procedure (Singh, 1994; Singh et al., 1998; Shahzad et al., 2021). The breeding scheme followed, and milestones achieved in the last 10 years are shown in Table 6. Several lines combining target traits have been developed through this programme (Kimani and Warsame, 2019). Superior lines were selected from populations BF01, BF07, BF16, and BF36. Eighty‐four lines had 50% more yield under stress and no‐stress conditions compared with the parental lines, suggesting transgressive segregation. The average grain Fe concentration of the best nine lines grown at Kabete and Thika varied from 91 mg kg^–1 (BF08-7-84) to 136 mg kg^–1 (BCB11-145), which was significantly higher than that of their parents and first-generation check varieties. Zn concentration of the top 10 lines varied from 35.2 mg kg^–1 (BF08-36-18) to 41.2 kg^–1 (BF08-7-74 and BF08-36-127), which was also significantly higher than their parents and first-generation check varieties (22.2–25.2 mg kg^–1). Six lines combined high Fe and Zn (BF08-13-181, BF08-1-18, BF08-7-74, BF08-16-36, BF08-36-18, and BF08-7-84), suggesting that selection for high Fe also enhanced Zn, probably because some of alleles for the two nutrients are in the same B11 linkage group (Blair et al., 2010). Moreover, the six lines showed combined resistance to angular leafspot, anthracnose, common bacterial blight, and root rots, and were high yielding (Kimani and Warsame, 2019). Their mean grain yield across locations varied from 2094 kg ha^–1 (BF08-36-18) to 2920 kg ha^–1 (BF08-1-18). They also have popular commercial grain types. Five of these lines are yellow seeded and one is red mottled. Recent studies indicate that yellow grain type is becoming popular in eastern Africa and often commands premium prices (Wiesinger et al., 2018). Data from this study show that gamete selection was effective in combining multiple traits such as high micronutrient density, resistance to diseases, drought tolerance, marketable grain types, and high yield potential as predicted by Saradadevi et al. (2021).

Table 6.

Gamete selection breeding scheme for the second-generation biofortified bean lines at the University of Nairobi (based on Warsame, 2014 and Kimani and Warsame, 2019)

Period	Generation	Milestones
2005	Parental	Population development Contrasting parents identified and characterized for mineral density, resistance to biotic and abiotic stress and other market-demanded traits
2006	Male gamete development	Multiparent male gametes developed by combining 11 commercial varieties and/or sources of resistance to angular leaf spot, anthracnose, and root rots, and tolerance to low soil fertility into single, three‐way, and double crosses
2007	Parents and crosses	Male gametes crossed to deficient commercial varieties to develop 47 F1 multiparent populations
2008	F1	Recombination and segregation in the 47 population bulks at Kabete Field Station
2009 (LRa)	F1.2	Advance population bulks at Kabete Field Station
2009 (SRa)	F1.3	Advance population bulks at Kabete Field Station to F4 generation Screen for mineral density
2010 (LR)	F1.4	Early generation selection Plant spacing changed, and single plant selection for plant vigour, angular leaf spot, anthracnose, root rots, and pod load 6612 single plants selected and kept separately
2011 (LR)	F4.5	Establish progeny rows and score for plant vigour, growth habit, reaction to diseases, duration to maturity, pod load, and seed yield 1672 progeny rows with resistance to two or more diseases, plant vigour scores of 1–6 (excellent to moderate), and good pod load selected at Kabete Field Station Undesirable populations, families discarded
2012 (LR)	F4.6	Line development and evaluation for drought tolerance 102 lines of eight populations evaluated for drought tolerance and reaction to diseases in moisture‐stressed and non-stressed conditions at Kabete Field Station and KALRO‐Thika during the 2012 long rain season Commercial varieties and parental lines included as checks Populations and lines showing susceptibility to diseases and poor agronomic potential were discarded
2013	F4.7	Multilocation testing Selected lines evaluated at four locations with diverse agroecological conditions, biotic and abiotic stresses Screen for mineral density
2014–2019	F4.8–F4.10	Seed increases and maintenance of finished lines
2020–2022	F4.11	Advanced yield trials at four locations (Kabete, Nairutia, Ol Jorok, and Loitokitok) Participatory variety trials in Gatundia (Laikipia West), Nairutia (Kieni West), Mikinduri (Tigania), Ndeiya, Ndenderu, and Tigoni(Kiambu County)

^a LR, long rain season (March–July); SR, short rain season (September–December).

Conclusions

The goal of the regional biofortification programme was to develop well-adapted, Fe- and Zn-rich bush and climbing bean varieties with high yield potential, marketable grain types, resistance to diseases and abiotic stresses, and cooking and eating characteristics acceptable to people in the target regions. Screening >2800 germplasm accession collected in 10 countries in east and central Africa showed that considerable genetic variation existed to facilitate improvement of Fe by >90% and Zn by >60%. Most of the lines consistently showing high levels of micronutrients originated from the Great Lakes region, especially DR Congo and Rwanda. Fe concentration varied from 45 mg kg^–1 to 147 mg kg^–1 and Zn concentration ranged from 12 mg kg^–1 to 62 mg kg^–1, and both are within the range reported globally. After considerable discussion with other global partners working on biofortification of common bean, a baseline of 70 mg kg^–1 Fe and/or 30 mg kg^–1 Zn was adopted. Thirty-eight lines were selected and shared with >25 countries in east, central, southern, and west Africa for further agronomic evaluation under local conditions, consumer preferences, and the eventual release of the first-generation varieties. Critical appraisal of the released first-generation bean varieties showed that they broadly met attributes considered essential for a biofortified variety. Genotypes with high Fe and Zn but deficient in agronomic traits were entered into a hybridization and selection programme. Forty-seven multiparent populations segregating for mineral density, disease resistance, drought, growth habit, and other traits were developed. Selection from these populations led to the second-generation lines with >90 mg kg^–1 Fe and >30 mg kg^–1 Zn, high yield potential, multiple disease resistance, drought tolerance, bush and climbing growth habits, and commercial grain types.

Future directions

Although considerable progress has been made in developing the biofortified bean varieties in eastern Africa in the last 23 years, more needs to be done to realize their potential in order to reduce nutritional, health, and economic challenges associated with micronutrient deficiencies and also to provide a better understanding of fundamental processes or mechanisms underlying mineral uptake and seed translocation, which might better explain the observed variability of mineral content in common bean.

Creating impact with first-generation varieties

The first-generation varieties have been released in several countries (Mukankusi et al., 2018). However, their impact is limited due to inadequate awareness, seed production, dissemination, and utilization. There is urgent need to produce certified seed, and promote and disseminate the released varieties on a large scale to create meaningful impact. Guidelines for bean seed production for smallholder producers and entrepreneurs were published by CIAT (David, 1998; David and Oliver, 2002), and economic analysis is discussed by Katungi et al. (2011).

Release and registration of second-generation varieties

The second-generation lines need to be released and registered to facilitate commercial seed production and dissemination (Kataliko et al., 2024). For countries with a requirement for formal validation by an independent regulatory agency such as Ethiopia, Kenya, Uganda, and Tanzania, these lines should be submitted for multilocation national performance trials (NPTs) and for distinctiveness, stability, and uniformity (DUS) tests.

Enhancement of bioavailability by selecting for low phytate

Breeding for low phytates is probably the most promising method of reducing the effects of phytates in common bean. Campion et al. (2009) isolated and characterized a low phytic acid mutant (lpa-280-10) with a 90% reduction of phytic acid and absence of lectins (lf) in the seeds. Using the in vitro Caco-2 cell model, they showed that the white seeded lf+lpa lines had 12 times more bioavailable Fe (Petry et al., 2013, 2014). In earlier studies, there were concerns that less complexed Fe due to low phytate might lead to negative effects such as oxidative stress due to a Fenton-type reaction on seed and plant performance (Frossad et al., 2000). However, subsequent studies showed that despite strong phytic acid reduction in the seed, the seedling emergence, seed yield, and plant growth of lpa mutant lines were not different from those of parental genotypes (Cominelli et al., 2020). The lpa mutant has been introgressed mainly into small and medium sized beans from interspecific crosses between P. coccineus and P. vulgaris L (Giuberti et al., 2019). It would be interesting to study how this mutation can influence phytate concentration and Fe bioavailability in large and medium seeded biofortified cultivars popular in east, central, and southern Africa.

Understanding the fundamental processes and mechanisms underpinning the improvement of common bean for the sustainable production of micronutrient-rich varieties

This review indicates that there is considerable variability in mineral accumulation in common bean. However, the mechanisms for mineral uptake and translocation into the seeds are not well understood. Therefore, there is a need for future research to provide better understanding of the fundamental processes or mechanisms of mineral uptake and seed translocation to better explain the variability of mineral accumulation in beans. Loading and accumulation of Fe and Zn in edible portions, especially seeds and leaves, involves uptake from the rhizosphere, xylem loading, root-to-shoot transfer, distribution to the leaves or seed-covering tissues, phloem loading for movement to seed, and finally loading into the seed (Zulfigar et al., 2024b). This process involves many genes which code for transporter proteins and enzymes, and may require active and selective transport mechanisms, which vary with species. There is a need to elucidate the specific expression patterns and roles of each gene involved in Fe transport within this intricate structure, including identifying key transporters responsible for releasing Fe into the extracellular space from transfer cells and those involved in its uptake by the embryo.

The availability of nutrients from soil to plants is influenced by factors such as soil pH, organic matter content, soil moisture and aeration, soil nutrient concentration, interaction with other mineral nutrients and beneficial microorganisms, and by the crop variety that defines the structure and function of the root system, and response to these factors. Genetic variation for tolerance to low soil P, N, and K, toxicities of Al and Mn, and adaptation to low fertility acid soils prevalent in some major bean production regions, especially in east and central Africa, and in Latin America, has been reported (Lubanga et al.,2007; Beebe et al., 2008). These adaptive traits to edaphic stress factors appear to be controlled quantitatively (Kimani et al., 2007; Kimani and Tongoona, 2008; Kimani and Derera, 2009), but the underlying mechanisms are not well known. It will be interesting to determine how the soil concentration of macroelements such as P, K, and Ca, micronutrients such as Fe and Zn, and soil pH interact and influence these fundamental processes and mechanisms.

Author contributions

The author solely wrote and edited the manuscript, and constructed the tables and figures.

Conflict of interest

The author declares no conflict of interest.

Funding

The research reviewed in this paper was supported by the Association for Strengthening Agricultural Research in East and Central Africa (ASARECA) [grant no. ASARECA/HVNSC/012/01]; Kilimo Trust (Uganda) [grant no. KT-M4P-01]; BBRC (UK) through The James Hutton Institute (Scotland) [BBSRC/GCRF grant ref: BB/T008865/1], and the National Research Fund (Kenya) [grant no. NRF/1/MMD199].

Data availability

All the data used in this review are available at the University of Nairobi repository http://erepository.uonbi.ac.ke/