Genetic studies on iron and zinc concentrations in common bean (Phaseolus vulgaris L.) in Ghana

Iron and zinc deficiencies cause high health risk to young children and expectant mothers in sub Saharan Africa. The development of biofortified common bean (Phaseolus vulgaris L.) varieties could address the acute micronutrient deficiencies with associated improvement in the nutrition and health of women, children and adults. The objective of this study was to determine the mode of gene action and genetic advance in iron and zinc levels in common bean. Field experiment was carried out using six generations of two populations made of crosses between pairs of low iron, low zinc and high iron, moderate zinc genotypes (Cal 96 ˣ RWR 2154; MCR-ISD-672 ˣ RWR 2154). Each generation (P1, P2, F1, F2, BC1P1 and BC1P2) was evaluated on the field in a randomized complete block design with three replications. Generation mean analysis were performed for each trait measured in each of the crosses while iron and zinc levels were quantified by x-ray fluorescence. The study showed that both additive and non-additive gene effects were important in determining the expression of high iron and zinc levels. Iron concentration in the common bean seeds ranged from 60.68 to 101.66 ppm while zinc levels ranged from 25.87 to 34.04 ppm. Broad sense heritability estimates of iron and zinc were high in the two crosses (62–82% for Fe and 60–74% for Zn) while narrow sense heritability ranged from low to high (53–75% for Fe and 21–46% for Zn). Heritability and genetic gain were used as selection criteria for iron and zinc, and it was concluded that doing so would be beneficial for future improvement.


Introduction
Common bean (Phaseolus vulgaris L.) known as kidney bean, dry bean, navy bean, snap bean, haricot bean or French bean is considered as one of the most important protein and mineral source in developing countries [1][2][3]. Seed iron concentration in bean varieties varied from 40 to 120 ppm and zinc concentration from 20 to 52 ppm [4]. Given the importance of beans as a source of inexpensive protein and minerals, the crop plays an important role in reducing poverty and improving food security [5]. Although, world production trend of common bean is scanty, it is estimated that 12 million tons is produced annually. The production of common bean in Africa is estimated to be around 5.9 million tons of which is mainly centered in the Eastern and Central regions [6]. In Ghana, the crop was recently introduced from Uganda following evaluation of introduced lines, and official release and commercialization of adapted varieties in 2016.
Over 3 billion people worldwide suffer from malnutrition due to iron and zinc deficiencies [7][8][9][10]. Iron deficiency in children and adolescence adversely affect physical growth, mental development and learning capacity. In adults, iron deficiency causes anaemia which reduces the capacity to do physical labour. Iron deficiency also increases the risk of women dying during delivery or in the post-delivery period. Zinc deficiency has been linked to stunting in children [7,8,11]. In Ghana, sixty-six percent (66%) of children between 6 and 59 months and forty-two (42%) of women are anaemic [12]. Research conducted in Ghana shows that micronutrient deficiency is a major contributor to child morbidity and mortality. In women, anaemia has resulted in complications of pregnancy among them being poor foetal development and death [12]. This situation has created a serious problem in Ghana, hence the need to improve micronutrient content in common bean cultivars would contribute in addressing malnutrition. The development and adoption of biofortified beans will go a long way to address the acute micronutrient deficiencies with concomitant improvement in the nutrition and health for women, children and men in Ghana.
The magnitude of genetic variability within breeding populations has important bearing on crop improvement [3]. Generation mean analysis provides estimates of epistatic effects information on the relative importance of average effects of the genes (additive effects), dominance deviations and effects due to non-allelic genetic interactions to guide breeding and selection for quantitative traits [13]. The success of genetic improvement depends on sufficient genetic variability and heritability of the trait under consideration. Therefore, the objectives of the current study were to determine the mode of gene action, the extent of genetic variability, heritability and genetic gain for iron and zinc contents of the two common bean crosses.

Experimental materials
Three (3) genotypes ( Fig. 1 and Table 1) received from the Alliance Bioversity International and International Center for Tropical Agriculture (CIAT)-Uganda were used as parental materials. MCR-ISD-672 is small seeded, high yielding but low in iron and zinc. RWR 2154 is large seeded, early maturing, high but low in iron and zinc contents respectively. RWR 2154 is a released material in Rwanda as a bushy plant type but climbs profusely in Ghana. The low Fe/Zn parents included, CAL 96, a red mottled bean developed and released in Uganda. It is highly susceptible to all major diseases and is used as a local disease and high yield check [14]. Two single crosses were made using the three parents; cross 1 (CAL 96 ˣ RWR 2154) and cross 2 (MCR-ISD-672 ˣ RWR 2154). The F 1 seeds generated from the parents were planted to generate F 2 population through selfing and the backcross populations (BC 1 P 1 and BC 1 P 2 ) were simultaneously generated through crossing F 1 hybrids back to their respective parents (P 1 and P 2 ). The experimental materials consisted of six generations for each cross include parents (P 1 and P 2 ), the first and second filial generations (F 1 and F 2 ) and backcrosses with the donor or recurrent parent (BC 1 P 1 and BC 1 P 2 ) ( Table 2).

Site description
Crosses to generate six basic generations and field experiment were carried out at Fumesua Crops Research Institute (CRI) of the Fumesua has a bimodal rainfall distribution with an average total annual rainfall of about 1,727.2 mm. The major rainfall season is from March to July while the minor rainfall season is from August to November. The mean annual relative humidity is 95% in the morning and 61% at noon. The mean minimum and maximum temperatures are 21 • C and 31 • C, respectively. The soil at the experimental site belongs to the Asuansi series and is classified as Ferric Acrisol.

Experimental design
Randomized complete block design (RCBD) with three replications was used for the field evaluation. Six treatments (P 1 , P 2 , F 1 , F 2 , BC 1 P 1 and BC 1 P 2 ) for each cross were used. Each replication had 14 plants in two rows for each of the parents (P 1 and P 2 ) and F 1 populations, 28 plants in four rows for backcross populations, and 70 plants in ten rows for the F 2 populations. Plant spacing of 50 × 30 cm, 2 m plot length with 1 m between plots and 2 m between replications were used in the experimental design. Seeds were planted at a density of one per hole.
Standard agronomic procedures were followed for proper plant development. Triple Super Phosphate (TSP 46% P 2 O 5 ) was applied as side placement fertilization for each plant at a recommended rate of 100 kg P 2 O 5 /ha. The trial was monitored for insect pest control with Lambda Super 2.5 EC (0.4-0.8 L/ha) and Dimethoate 40 EC (0.2-0.3 L/ha) at pre-flowering and post-flowering stages, respectively.

Iron and zinc quantification
Matured pods were allowed to dry before harvesting and threshed by hand to avoid soil contamination. Clean, dirt free and dry pods were harvested before the main harvest. The harvested pods were hand threshing and seeds were subsequently dried to a moisture content of 12-14%, out of which 40 seeds were counted from each plant. The seeds were cleaned with dry cotton cloth, packed and shipped to Rwanda Agriculture Board (RAB) Rubona Research Station for rapid screening of iron (Fe) and zinc (Zn) content in common bean seeds. For quantification of iron and zinc, X-ray fluorescence (XRF) method was used for analyzing and determining Fe and Zn in the grains [15,16]. Briefly, the samples were thoroughly washed and cleaned with distilled water and clean cloth to avoid aluminium contamination. The samples were then oven dried at 60 • C for 12 h and then ground using Sunbeam Conical Burr Mill EM0480 Grinder. The ground samples were transferred into small sample cups on a tray positioned in the machine's tray. Iron and zinc content in the seeds was determined using X-Ray Fluorescence Spectrometer (XRF) by analyzing each sample for 100 s with spinning of sample cup to determine iron and zinc contents and record intensities of X-rays emitted from the generator. The screen tray rotates to place the samples being measured at the top and the results of the analysis displayed automatically when all the samples on the tray have been measured.

Data analyses
Analysis of variance (ANOVA) and mean separation of the six generations of each cross combination were done using GenStat version 12 software [17]. Treatment means were separated using Tukeys Honesty Significant Difference (HSD) at 5% level of significance test.

Generation mean analysis
Generation mean analysis was carried out with R package for determination of gene action for high iron and zinc. The observed means of the six generations and their standard errors were used to estimate the mid-parent (m), additive (a) and dominance (d) gene effects using the joint scaling test of [18]. The adequacy of the simple additive-dominance model (mean, additive and dominance effects) was determined by χ 2 test. Individual scaling tests (A, B, C and D) of [19,20] were employed to test their fitness to the additive-dominance model. The significance of any of these four scales indicated the presence of epistasis. In case of the inadequacy of the three-parameter genetic model and significance of the scaling tests, six-parameter genetic model suggested by Ref. [21]  Broad-sense and narrow-sense heritability were estimated by Ref. [22] as follows: The components of variation in six populations were calculated by the formulae of F 2 variance were obtained by the following formula of [18] as: Estimates of heritability were classified according to low (0-30%), moderate (31-60%) and high (>60%) [23].

Analysis of variance
The analysis of variance showed highly significant (p < 0.001) variation among the generations in iron and zinc for crosses 1 and 2 (Table 3). This indicates that selection for improvement of iron and zinc would be effective. Although there was significant genetic diversity among the populations, lower coefficient of variation (CV) recorded showed the reliability of the results from the XRF method of quantification.

Treatment mean comparisons
Mean values and standard errors (SE) for iron and zinc for the two crosses are presented in Table 4. The mean values of iron were higher in progenies of all the two crosses than zinc. This may be due to high variability observed in iron content than zinc content. Parents in two of the crosses (CAL 96 ˣ RWR 2154 and MCR-ISD-672 ˣ RWR 2154) showed significant differences (P ≤ 0.05) in iron and zinc, indicating that there were significant genetic variations among the parents for these micronutrients. The results also imply that generation mean analysis could be carried out for the two crosses. Contrasting parents are prerequisite to accurate estimation of the number of genes and genetic parameters for the traits [25] The F 1 hybrids in the cross CAL 96 ˣ RWR 2154 recorded significantly (P ≤ 0.05) higher iron content (101.66 ppm) than its parents. Further, the mean zinc concentration of F 1 hybrids was significantly different from the parent with lower zinc content (P 1 ) and not different from the parent with higher zinc content (P 2 ). The mean iron and zinc contents of hybrids in cross MCR-ISD-672 ˣ RWR 2154 were significantly different from P 1 and not different from P 2 . The results suggest dominance in the direction of the parent with higher iron and zinc contents (RWR 2154). Thus, genes for high iron and zinc contents were transmitted from the male parent (RWR 2154) to its offspring.

Gene action of iron and zinc
Test for epistasis was done before estimation of genetic variation to help in deciding the method of analysis for the components of variation. The results of simple scaling test for iron and zinc for the two crosses are presented in Table 5. The results showed that in Cal 96 ˣ RWR 2154 cross, scale A and D were significant for iron concentration. While in MCR-ISD-672 ˣ RWR 2154 cross, scale C was significant for iron concentration. For the zinc concentration, scale B was significant while scale D was highly significant in MCR-ISD-672 ˣ RWR 2154 cross. The significant of any of the scales indicated the presence of epistasis.
The joint scaling test (additive-dominance model) showed the three parameters (m, [a] and [d]) to be significant for all the traits studied (Table 5). This was an indication that additive and dominance genetic effects contributed significantly to the inheritance of seed iron and zinc in both crosses. However, χ 2 test values (p < 0.05) were significant for iron in the two crosses and zinc in cross MCR-ISD-672 ˣ RWR 2154, indicating that the additive-dominance model was not adequate in explaining the mode of gene action governing high iron and zinc in the crosses. Thus, epistatic interactions were involved in the genetic control of the traits. However, the adequacy of the additive-dominance model for zinc in cross CAL96 ˣ RWR 2154 suggests the absence of non-allelic interaction in the inheritance of the trait. The results showed that both additive and dominance gene effects were implicated in the inheritance of high zinc in cross CAL96 ˣ RWR 2154, indicating ineffective selection for the trait in early segregating generations. Estimates of genetic effects from the six-parameter model indicated that mean and additive effects for both traits in cross CAL96 ˣ RWR 2154 and MCR-ISD-672 ˣ RWR 2154 were significant ( Table 5). The significance of the additive gene effect for traits in the crosses suggests the potential for further improvement of iron and zinc through selection from early segregating generations.  [18]. In most cases, the magnitude of dominance [d] gene effects appeared to be higher than additive gene effects for both traits in both crosses. The large positive dominance effect relative to additive effect suggests that the parents are genetically different since dominance is due solely to the heterozygosity of those genes for which the parents differ [26]. Again, the positive and highly significant dominance effect [d] increased iron content in CAL 96 ˣ RWR 2154 cross. The results suggested that iron and zinc contents in common bean may be controlled by both additive and non-additive genes. Similar findings were also reported by Refs. [15,27,28].

Heritability and genetic advance
Heritability (1-8) and genetic advance as percent of mean for iron and zinc concentrations of two common bean crosses are presented in Table 6. Broad-sense heritability (1) estimates were high for iron (62%) and zinc (74%) in Cal 96 ˣ RWR 2154 cross. However, moderate to low narrow-sense of heritability (2) estimates were recorded for both iron (53%) and zinc (21%) in this same cross, respectively. Broad-sense and narrow sense heritability estimates of similar magnitude were observed by Ref. [27] for iron concentrations. Again, their narrow-sense heritability estimates were also similar in magnitude as reported by Ref. [29]. In Table 4 Mean and standard error of iron and zinc concentration in common beans measured by x-ray fluorescence (XRF) in parents (P 1 and P 2 ) and their F 1 , F 2 , BC 1 P 1 (F 1 ˣ P 1 ) and BC 1 P 2 (F 1 ˣ P 2 ). The same letters in the same column indicate no significant differences at the 5% level based on Tukeys Honesty Significant Difference (HSD).

Table 5
Estimates of additive, dominance and epistatic effects (and their standard errors) from the joint scale test for iron and zinc concentration on common beans measured by x-ray fluorescence (XRF) in parents (P 1 and P 2 ) and their F 1 , F 2 , BC 1 P 1 (F 1 ˣ P 1 ) and BC 1 P 2 (F 1 ˣ P 2 ) in two crosses grown in CRI research station. *, **, *** indicate terms are significant at P < 0.05, P < 0.01 and P < 0.001, respectively. m = mid-parent effect, a = additive effect, d = dominance effect, aa = additive x additive effect, ad = additive x dominance effect and dd = dominance x dominance effect, ns = non-significant. MCR-ISD-672 ˣ RWR 2154 cross, high to moderate broad-sense heritability estimates for iron (82%) and zinc (60%) were recorded. Again, similar trend was obtained in narrow-sense heritability. High to moderate broad-sense heritability estimates recorded for iron and zinc is an indication of lesser environmental effects on the traits. High to moderate narrow-sense heritability estimates observed for iron in both crosses and zinc in MCR-ISD-672 ˣ RWR 2154 cross indicate that early selection for improvement of the traits would be effective. However, low narrow-sense heritability recorded for zinc in Cal 96 ˣ RWR 2154 cross is an indication that selection for improvement of the trait may not be effective in early segregating generations. A relative comparison of heritability and expected genetic advance will give an idea about the nature of gene action governing a particular trait [30]. Genetic advance as percent of mean (GAM) (10) was moderate to high for iron and zinc in both crosses. The results of this study showed moderate to high broad-sense heritability and GAM for iron and zinc in both crosses, suggesting that the traits are governed by additive genes and selection for improvement of the traits would be rewarding. Heritability combined with genetic advance (9) is a more reliable index for selection of traits [31]; hence, the selection from segregating populations in these crosses would be based on the interpretations given above.

Conclusion
Genetic variation was observed in crosses CAL 96 ˣ RWR2154 and MCR-ISD-672 ˣ RWR 2154 for iron and zinc content. Additive and non-additive gene effects were both important in the expression high iron and zinc concentrations. Moderate to high broad sense heritability and variable narrow sense heritability were observed for iron and zinc in the two crosses, suggesting that the variation observed among the parents and progenies was largely due to additive gene effects. Thus, selection for improvement of iron and zinc in early segregating generations would be effective. The results of the present study generally showed iron and zinc superiority of the F 1 hybrids over the mid-and better parent in common bean.

Author contribution statement
Maxwell Lamptey: Conceived and designed the experiments, Performed the experiments and Wrote this paper. Richard Adu Amoah: Analyzed and interpreted the data and Contributed reagents, materials, analysis tools or data. Francis Osei Amoako-Andoh, Louis Butare, Kwabena Asare Bediako, Isaac Tawiah and Stephen Yeboah: Analyzed and interpreted the data and Contributed reagents, materials, analysis tools or data.
James Yaw Asibuo and Hans Adu-Dapaah: Conceived and designed the experiments and Wrote the paper.

Data availability statement
Data will be made available on request.

Additional information
No additional information is available for this paper.

Declaration of competing interest
The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.