http://orcid.org/0000-0002-4859-0631
ABSTRACT
Micronutrient deficiencies in diets, including iron (Fe) and zinc (Zn), are an important public health problem across population in Ethiopia. Potato is a modest source of dietary nutrients. The purpose of this study was to assess the mineral (Fe & Zn) composition of potato germplasm in Ethiopia and grower's willingness to try new potato types. The Fe and Zn contents were assessed in peeled tubers of popular farmer varieties, and newly introduced Andean diploid group Phureja accessions and improved tetraploid clones grown in multi-location trials. Grower's preferences for the new potato types was assessed using a participatory approach. Significant variations in mineral and tuber yield traits exist among the germplasm assessed and broad-sense heritability appears to be high for most traits measured, suggesting that progress through breeding can be made among these materials. However, the slightly negative correlation of tuber yield with Fe and Zn content and user appreciation traits found in this study suggests that progress in improving crops for yield and enhanced Fe and Zn concentrations with end-user acceptable quality will be challenging. The results from this study provide baseline information on Fe and Zn composition of many popular farmer varieties in Ethiopia and end-user preference criteria for evaluating new potato varieties that would help to orient the biofortification breeding to the conditions and preference of farmers.
Introduction
Around two billion people worldwide, mostly in developing countries, suffer from micronutrient malnutrition often referred as ‘the hidden hunger’ (von Grebmer et al. Citation2014). Lack of iron (Fe) and zinc (Zn) in the diet are among the leading causes of micronutrient malnutrition, and women of fertile age and children are most vulnerable (Bouis and Welch Citation2010; Paget et al. Citation2014). Ethiopia is among the countries where the prevalence of Fe and Zn deficiencies are high (EPHI Citation2016; Melash et al. Citation2016). Iron deficiency among the preschool age children, school age children and women of reproductive age is about 30%, 20% and 16%, respectively, whereas prevalence of Zn deficiency is about 35% in preschool age children, 36% in school age children and 34% in non-pregnant women of age 15–49 in Ethiopia (EPHI Citation2016).
Crop plants are primary sources of essential dietary nutrients for many families worldwide (Miller and Welch Citation2013). Potato is among the world's most widely grown and consumed staple food crops (Devaux et al. Citation2014; Machida-Hirano Citation2015). Potato's importance as a food crop is steadily expanding over years in Ethiopia (Haverkort et al. Citation2012; Asfaw et al. Citation2015). Potato contains a range of nutritional and health-promoting compounds which are essential in the human diet (Brown Citation2005; Andre et al. Citation2014). As little as 150 g of potato can supply 6% and 3% of the daily recommended allowance of Fe and Zn for adults respectively (Woolfe Citation1987, cf. Brown et al. Citation2010). Tubers of potato plants possess variability for essential micronutrients, Fe and Zn (Andre et al. Citation2007; Burgos et al. Citation2007; Brown et al. Citation2010) that could allow substantial increases in these minerals through plant breeding to combat the micronutrient malnutrition problem (Welch Citation2002).
Enriching potato tubers with more zinc and iron concentrations through plant breeding at the International Potato Centre (CIP), with its headquarters in Peru has begun in 2004 with diploid Andean landraces and improved disease resistant tetraploid clones selected for their Fe and Zn levels with the support from the HarvestPlus Program (Paget et al. Citation2014). This effort at CIP has produced prototype diploid potatoes combining desirable agronomic features with higher levels of Fe and Zn (Asfaw et al. Citation2014; Paget et al. Citation2014). The transmission of plant traits from diploid breeding lines to the tetraploid lines has also been successfully achieved for other traits in potatoes (Watanabe Citation2015). In order to devise a suitable deployment strategy for these products and further optimize the tuber yield and quality at target set environments, user preference trait elicitation, nutrient profiling of the existing popular varieties, and farmer reflections on exposure to the new Andean landrace group Phureja and improved tetraploid potatoes were assessed in Ethiopia.
The cultivated potatoes in Ethiopia constitutes exclusively the European and North American type tetraploid Solanum tuberosum L. genotypes with a genomic constitution of 2n = 4x = 48. Its production is characterized by fairly-structured preferences on the well-acquainted cultivar types for taste, colour, productivity and time to tuber balking (Kolech et al. Citation2015, Citation2016). Number of studies in Ethiopia has documented the agronomic adaptation; reactions to major diseases and adoption of the traditional as well as the improved cultivars (KidaneMariam Citation1979; Dessalegn et al. Citation2008; Woldegiorgis et al. Citation2008; Abebe et al. Citation2013; Kolech et al. Citation2015) but none of them systematically assessed nutrient profile along with the farmer preference traits. The information on micronutrient profile of popular farmer cultivars, and farmers anticipated demand and perception on new and more nutritious potato genotypes are scanty except one study reported Fe and Zn concentration of three farmer varieties along with the 15 released and three elite clones (Abebe et al. Citation2012). With the present study, we sought to determine the Fe and Zn profile of widely grown local farmer varieties, and expose Ethiopian farmers to new diversity and learn about their preferences for incorporation of end users preference traits in the potato breeding.
Material and Methods
Assessment of Fe and Zn concentration of traditional and newly introduced potato in Ethiopia
Traditional and newly introduced potatoes field grown in different trials were assessed for Fe and Zn concentrations and tuber yield characteristics. The traditional potatoes encompassed twelve popular farmers’ and three released varieties whereas the newly introduced potatoes included 12 clones sampled by CIP's breeding program to represent the landrace group in which high Fe accessions were found and tetraploids selected for higher Fe and Zn concentrations before breeding for these traits was taken on. and show the varieties used in this study. Detail description on the traditional varieties and their collection sites is presented in Kolech et al. (Citation2016). The traditional potatoes were planted in separate trials in 2014 cropping season at two locations: Yilmana located at 11°16′N latitude and 37°28′ E longitude at an altitude of 2800 m above sea level, and Laigaint situated at 11°43′N latitude and 38°28′E longitude at an altitude of 3100 m above sea level in the north-western part of Ethiopia. The two locations represented the major potato production ecologies in the north-western Ethiopia in terms of soil type, fertility, and topography. Yilmana is represented the wet and Laigaint the dry and less dependent rainfall ecologies for the potato production in Ethiopia.
Table 1. The description of the Andean landrace group Phureja accessions and improved tetraploid clones introduced from Lima, Peru to Ethiopia for farmer exposure and adaptation trials.
Table 2. Mean trait values, heritability, variance components, coefficient of variation and probability values for testing significance of genotype, location and genotype × location effects of the traits assessed on traditional potato varieties grown at two sites in Ethiopia.
The newly introduced potato germplasm was grown in a separate experiment at two locations (Holetta and Kulumsa research station farms of the Ethiopian Institute of Agricultural Research, EIAR) in central highlands of Ethiopia. Holetta site is located at 9°03′16″ N and 38°30′17″ E at an altitude of 2400 m above sea level whereas Kulumsa at 8°01′07″ N and 39°09′32″ E at an altitude of 2200 m above sea level. This experiment was executed for two years: in 2013 at Holetta and in 2014 both at Holetta and Kulumsa sites.
The genotypes in both experiments were laid out in a randomized complete block design with three repetitions. The plot size per variety in each replication was 9 m2 in traditional potatoes experiment and 12 m2 in newly introduced potatoes experiment with between ridge and plant distance of 70 and 30 cm, respectively. Ridging (hilling) and agronomic operations for the optimum potato crop growth and development in both experiments were implemented as described by Woldegiorgis et al. (Citation2008). Late blight disease was controlled with spray of Ridomil Gold MZ 68WG fungicide at the rate of 4 kg/ha.
In all trials, harvest was performed when the plants were completely senescent where yellowing was complete and uniform and the stems were decumbent. The foliage was cut 15 days prior to harvesting to induce proper suberisation of tubers. During harvest, data on tuber yield traits were collected as per the CIP protocol for healthy tuber yield trial (https://research.cip.cgiar.org/confluence/display/GDET4RT/Protocols). At harvest, tubers were sampled from each repetition for Fe and Zn content scan on freeze-dried, milled peeled tubers with X-ray fluorescence (XRF) spectrometry. Tuber sampling and sample processing were done as per the procedure described in Porras et al. (Citation2014).
Participatory evaluation of new prototype potato varieties with farmers and researchers
The newly introduced potato varieties were further evaluated for farmer preference criteria. The participatory assessment was executed in 2013 trial planted at Holetta to capture farmer's impression on the new potato types for the tuber yield and eating quality traits of boiled potato.
Nineteen farmers (10 female and 9 male) and 5 research technicians participated in the evaluation process. The evaluators were purposively selected farmers from the surroundings of Holetta research station and potato research technicians. Evaluators were experienced potato growers who have been participating on potato research and development activities with the Holetta research station. The evaluators were brought to the Holetta station where the trial was planted to assess the clones at harvest for tuber yield traits and after cooking traits for eating quality. Upon arrival at the trial field, group discussion was held with the evaluators to list traits they use for describing potato varieties at harvest and after cooking when assessing for food quality of boiled potatoes. The trait list of the evaluators from the group discussion was written in a cardboard arranged individually for each trait. The traits were displayed to the evaluators for discussion and consensus on the final trait list. Effort was made to make each participant interact freely and speak out his or her mind in a trait elicitation exercise during the group discussion. After the participants reached a final consensus on the trait list, each evaluator was requested to identify and rank the traits he or she feels valuable in describing potato varieties in order importance without any influence from other participants. Each evaluator was given 15 grains (maize for men and beans for women) to rank the traits by depositing grains: 5 for most important, 4 grains for 2nd, 3 grains for 3rd, 2 grains for 4th and 1 grain for 5th most important trait in plastic bags placed in front of names of the traits written in a cardboard. After the trait elicitation, the participants ranked the clones for the previously identified criteria by depositing 3 grains in the best clone, 2 grains in the 2nd best, and 1 grain in the 3rd best clone's bag. The clone name codes and paper bags for depositing the grains up on the selection were placed in front of each clone in a visible way. Explanation was given to evaluators on making selection and ranking of the clones using grains provided for each of them as per the CIP participatory variety selection (PVS) protocol (https://research.cip.cgiar.org/confluence/display/GDET4RT/Protocols). After the procedure was made clear, the evaluators did the ranking by depositing the grains (maize for men and beans for women as previously for trait ranking). The evaluators also did organoleptic evaluation of clones for cooking quality traits: appearance, taste, and texture of the boiled potatoes.
Statistical analysis
The tuber yield and nutrient (Fe and Zn) content data from all the trials were subjected to analysis of variance (ANOVA) in GenStat v.16 and R statistical software packages. The data obtained in the evaluation of each trait in a trial were initially subjected to individual analysis of variance and mean separation for trait values of the genotypes was done using Least Significant Difference (LSD 5%) method. The effect of environment and genotype × environment interaction were assessed with linear mixed model considering genotypes as fixed effect and environment as random effect in Meta-R package (Alvarado et al. Citation2015). Adjusted means (Best Linear Unbiased Estimates), broad sense heritability, variance components, and coefficient of variation statistics were computed for individual as well as data combined across trials. Phenotypic correlations between the iron and zinc concentration, tuber yield traits and farmers’ qualitative preference traits were computed on combined across trials data. Farmer's qualitative scoring of tuber-yield traits at harvest and appearance of cooked potatoes were counted for ranking the clones. Frequency tabulation for farmer selection exercise was performed with Excel. Chi-square analysis was applied to test gender differences in significance of farmer criteria of importance for describing potato varieties. Logistic regression analysis was used to identify clones most preferred by farmers for taste and texture as per the procedure described in Asfaw et al. (Citation2012) using GenStat v16 software. The plots were constructed with SigmaPlot software (Systat Software, Inc., San Jose, CA).
Results
Fe and Zn profile of widely grown local farmer potato varieties in Ethiopia
The tuber yield, and Fe and Zn profile of farmer and released potato varieties grown at two locations in Ethiopia is presented in . Varietal differences in tuber Zn concentrations were significant (p < 0.02) at both locations while for the Fe content, the differences were significant at Yilmana site but not at the Laigaint. Similar to the trend for Fe, the genotypic differences for total tuber yield were significant at Laigaint but not at Yilmana. The genotypic differences for average tuber weight and tuber dry matter content were significant at both locations. Genotype by location interaction was significant for average tuber weight and Zn concentration but not for total tuber yield, dry matter and Fe concentration. This result agrees with Burgos et al. (Citation2007) for Zn but not for Fe content. Brown et al. (Citation2010) also reported genotype by environment interaction as not significant source of variation for Fe content in some North American potato trials.
The Fe content in mg per kg (ppm) dry weight (DW) ranged from 15.8 (Akime and Belete) to 20.6 (Rejim china) at Laigaint site and 16.4 (Hakime) to 24.9 (Enat Beguaro) at Yilmana site. Zn content in ppm in DW ranged from 8.0 (Hakime) to 16.3 (Gera) at Laigaint site and 10.1 (Hakime) to 23.4 (Enat Beguaro) at Yilmana site. Across location mean Fe concentration of more than 20 ppm in DW was recorded on three clones: Key deme, Enat Beguaro and Rejim China. None of the clones on average accumulated higher than 20 ppm in DW tuber Zn concentration. Enate Beguaro, Holland, Gera and Rejim China accumulated a higher Zn concentration of around 17 ppm in DW. Farmer variety Hakime had lower concentration of both Fe and Zn across locations compared to other clones evaluated in the trial. Another farmer variety, Rejim China accumulated relatively a higher and consistent Fe concentration across locations. Heritability was generally high for all traits measured except low for Fe content at one location. The nutrient profiling of popular farmer varieties in this study provided baseline information which could serve as benchmark to facilitate new Fe and Zn potato variety release decision in the country.
End-user preference trait elicitation with the newly introduced potato types
A panel of evaluators comprising farmers and research technicians participated in selection and evaluation of the new set diploid Andean and improved tetraploid potato types grown at Holetta agricultural research station in 2013. The panelists participated in group discussions to elicit important potato traits for describing potato varieties and individual scoring on performance of the clones for the identified traits. In group discussion, important criteria for evaluating and assessing the new potato varieties at harvest and organoleptic quality of the boiled potatoes after cooking were identified. The criteria identified as important for evaluating new potato varieties at harvest were tuber colour, tuber size, tuber number, tuber shape and freeness from disease and pest attack (). There were no statistically significant (P < 5%) gender differences in criteria identified for describing the potatoes varieties (chi square statistic for testing the gender difference is 2.66 with p value 0.62). However, the order of importance of the criteria coinciding between men and women panelists were freeness from damage of any pest and disease, tuber number and tuber size. Men valued tuber skin colour more than tuber shape whereas women attached more weight to tuber shape than tuber colour. For the organoleptic assessment after cooking, taste, texture and appearance of boiled potatoes were described as the most important criteria to identify best potatoes for eating quality. Appearance was described by the panelists as how the boiled potato looks when presented or served on a plate for eating. The panelists in the group discussion perceived the above listed organoleptic qualities are equally important for assessing potatoes for eating quality and hence were not ranked in order of importance likewise the criteria for the agronomic traits.
Table 3. Farmers perceived traits for assessing potato varieties at harvest and their order of importance by gender elicited based on individual trait assessment at Holetta, Ethiopia in 2013.
After the trait elicitation exercise, the panelists attempted ranking of clones for tuber yield traits at harvest and boiled tubers for eating quality after cooking. The participatory ranking of the clones based on the overall assessment for the tuber yield traits at the harvest identified most and least preferred clones by the male and female farmers, and the research technicians (A). For the male farmers, the released varieties, Belete and Gudene, and the Andean landrace diploid Phureja CIP703793 were the top three preferred clones for tuber yield traits at harvest whereas for the female farmers, the top three preferred clones were the improved tetraploid clones Belete, CIP393536.13 and Gudene. For the research technicians, the three most preferred clones were CIP397067.2, Gudene, and Belete. Out of the 15 clones exposed to the selection panelists, 9, 11 and 5 clones appeared among the preferred list for tuber yield traits by the male, and female farmers, and the research technician's, respectively. We noticed some commonality and specificity on preference for the clones between the gender groups. Three clones CIP399079.22, CIP704205, and CIP704227 were specific to the female farmers’ selection but not to the male farmers or research technicians’ selection. Five improved tetraploid clones namely Belete, Gudene, CIP397067.2, CIP393382.44 and CIP39136.13 were common selections to all the panelists. Out of 8 diploid Phureja landrace clones, three (CIP703897, CIP706828 and CIP3295) were non-preferred clones for the tuber yield traits at harvest.
Figure 1. Ranking of clones (A) at harvest based on over all the selection criteria and (B) after cooking for appearance (referring to how boiled potato looks when presented on plate). Ranking was done by depositing 3 grains in the best clone, 2 grains in the 2nd best, and 1 grain in the 3rd best clone.
Assessment of the clones for eating quality after cooking is presented in (B) and . In the organoleptic assessment excises only farmers participated. Research technicians felt they were not qualified to make the organoleptic assessment and hence were excluded from this exercise. The formally released Ethiopian variety Belete, the improved tetraploid clone CIP397067.2, and the Phureja landrace accession CIP703295 were most preferred clones for boiled tuber appearance by the male panelists. The Phureja landraces CIP703295, CIP706828, and the improved tetraploid clone CIP397067.2 were most preferred clones by female panelists for the boiled potato appearance ((B)). Considering all panelists, clones CIP703295, CIP397067.2, and Belete were the most preferred clones, and CIP395446.1, Gudene and CIP703580 were least preferred clones for appearance of the boiled tubers for consumption.
Figure 2. Comparison of clones based on logistic regression analysis of farmer's perception on taste and texture. The scale used for the assessment of clones was poor, fair or excellent for taste; and soggy, intermediate or mealy/floury for texture. Coding for clones, e2 = CIP703897; e3 = CIP706758; e4 = CIP395446.1; e5 = CIP397067.2; e6 = CIP703793; e7 = CIP706828; e8 = CIP393536.13; e9 = CIP704205; e10 = CIP703295; e11 = Gudene; e12 = CIP704227; e13 = CIP703580; e14 = CIP393382.44; e15 = CIP399079.22.
Comparison of the clones based on logistic regression analysis of farmer perceptions for taste and texture of the boiled tubers presented in . Two Phureja landrace and one improved tetraploid clones CIP704205, CIP397067.2 and CIP706828 were assessed excellent for taste by most the panelists whereas CIP704205 and CIP704227 were assessed good for texture (perceived mealy or floury in the mouth while eating). Neither of the formally-released Ethiopian varieties were preferred for taste or texture of the boiled tubers for consumption. However, the improved tetraploid clone CIP397067.2 was preferred for appearance of boiled tubers as well as taste but did not combine all favourable aspects of organoleptic quality.
The Andean landrace cultivar group Phureja accession, CIP704205 was favoured for its good taste and texture by majority of the panelists though was not high for micronutrient (Fe and Zn) contents and it was not among preferred clones for tuber yield traits in this study (, ). Despite the moderate Fe and Zn contents of this genotype, significant heritable variation for Fe and Zn contents among Phureja landraces reported by Burgos et al. (Citation2007) and Paget et al. (Citation2014) were the basis for initiating directed breeding to improve the micronutrient content of potato; and agronomic and culinary properties of this group influence the characteristics of early generation materials addressing this new breeding objective. Furthermore, Phureja cultivars are characterized by short dormancy which could enable year-round harvests in Ethiopia as a contribution to food security. From the combination of nutrient profile scan and farmer exposure trials, we learn that local traditional varieties are low in Fe and Zn levels to satisfy the dietary micronutrient needs, and farmers are open to test new types of potatoes in their farming system. Hence, the conventional approach of developing varieties that have the same appearance as those farmers are accustomed to growing may restrict the introduction and exposure of farmers to novel, attractive, nutritious and adapted germplasm.
Table 4. The best linear unbiased estimates and ANOVA statistics for tuber yield parameters, Fe and Zn contents of the fifteen potato clones evaluated at Holetta (2013 and 2014) and Kulumisa (2014), Ethiopia.
Agronomic performance and nutritional quality of the new set of potatoes in Ethiopia
presents varietal mean trait values, variance components, heritability and probability values for testing genotype, environment and the interaction effects of tuber yield and nutrient traits assessed in new set of the Andean landrace Phureja and improved tetraploid clones evaluated in Ethiopia. The genotypic difference was significant (P < 0.05) for the tuber yield and micronutrient traits assessed except for the Fe concentration. The environment (a year by location combination) effect was significant at least by 5% probability value for all the traits studied except total and marketable tuber weight per plant whereas the genotype by environment interaction effect was highly significant (P < 0.001) for all traits assessed. The mean total tuber yield ranged from 14 (CIP703897, Phureja) to 53 tons ha−1 (Belete, released check variety) whereas the marketable tuber yield ranged from 11 (CIP703897) to 37 tons ha−1 (Belete). The released variety, Belete and newly introduced Andean landrace Phurja, CIP706758 were higher in producing total as well as marketable tuber yield per plant. Four test clones CIP393382.44, CIP397067.2 and CIP399079.22 (improved, tetraploid), and CIP706828 (Phureja, diploid) produced total tuber yield of 41, 33, 31 and 27 tons ha−1, respectively. Iron concentration ranged from 16 (Gudene, released tetraploid variety) to 22 (CIP706758, Phureja, diploid) with mean value of 18 ppm in DW. The Zn concentration ranged from 11 (Gudene) to 20 (CIP703295, Phureja, diploid) with mean value of 14 ppm in DW. Three test clones CIP706758 (Phureja, diploid), CIP703893 (Phureja, diploid) and CIP393536.13 (improved, tetraploid) produced higher Fe concentration 22, 21 and 20 ppm in DW, respectively. The Phureja diploid clones CIP703295 and CIP703793 were the two higher Zn accumulators in the tubers (20 and 19 ppm, respectively) among the test clones. Heritability values for total and marketable tuber yield and Zn concentration were higher. These traits also had higher genotypic variance compared to genotype by environment interaction variance.
Relationship of tuber yield, micro-nutrient and organoleptic traits
presents pattern of correlations between tuber yield, micronutrient, and organoleptic traits in new potato types evaluated in Ethiopia. The correlations were positive and significant (P < 0.05) between Fe and Zn concentrations. The correlations between tuber yield traits showed positive trend, however, such trend appeared to be significant only between marketable tuber yield and marketable tuber weight per plant (P < 0.05), between marketable tuber weight per plant and total tuber weight per plant (P < 0.01) and between marketable tuber yield and total tuber yield (P < 0.01). Organoleptic traits (taste and texture) showed a positive association but such trend was not statistically significant at 5% probability level. The relationships between micronutrient traits (Fe and Zn concentrations at tubers) and tuber yield traits were generally negative but not significant at 5% and below probability level. The correlations of tuber Zn level with total tuber and marketable tuber yield were negative and significant at 10% probability level. The organoleptic traits (taste and texture) didn't show a well-structured and significant associations with both micronutrient and tuber yield traits.
Table 5. Phenotypic correlations between tuber yield, micronutrient and organoleptic traits in potato clones evaluated in Ethiopia.
Micronutrient profile and tuber yield among different background potatoes
presented the overall mean Fe and Zn concentrations and total tuber yield of different genetic backgrounds (i.e., Andean landrace diploid Phureja, popular farmer varieties, newly introduced improved tetraploid clones, and released tetraploid varieties) evaluated under the two experiments. The micronutrient traits (Fe and Zn levels) didn't show structured superiority among the different germplasm groups assessed. For the tuber yield potential, the tetraploids (improved as well released varieties) expressed clear superiority over the diploids and farmers varieties. Overall mean iron content (ppm in DW) ranged from 18 in released tetraploid varieties to 19 in popular farmer varieties while the mean Zn level in ppm in DW ranged from 13 in released varieties to 16 in Andean landrace diploid Phureja accession groups. The farmer's varieties were on average low in tuber yielding potential even compared with the Andean diploids landraces. Heritability for the measured traits showed a fairly good agreement between the farmer and introduced germplasm ( and ). However, heritability for Fe content was much larger in the farmer varieties than in the CIP introduced germplasm. The observed differences in trait values and heritability could be due to the differences in varietal background and/or locations the materials were tested. The trials for the two groups were separate experiments in different soil and whether hence it is hard to make a robust comparison of trait performance between the different background germplasm tested.
Figure 3. Meta comparison of mean trait performance [tuber yield (tonsha-1), Fe concentration (ppm), and Zn concentration (ppm)] among different category of potato genotypes (Andean landrace Phureja (n = 7), popular farmer varieties (n = 12), and improved newly introduced tetraploid clones (n = 5) and released varieties (n = 4)] assessed in Ethiopia under different experiments.
![Figure 3. Meta comparison of mean trait performance [tuber yield (tonsha-1), Fe concentration (ppm), and Zn concentration (ppm)] among different category of potato genotypes (Andean landrace Phureja (n = 7), popular farmer varieties (n = 12), and improved newly introduced tetraploid clones (n = 5) and released varieties (n = 4)] assessed in Ethiopia under different experiments.](/cms/asset/4c3cca52-72bc-43d0-83bc-93060df95709/sagb_a_1448885_f0003_b.gif)
Discussion
The micronutrient profiling of farmer, released, and newly introduced clones along with the farmers preference traits for potatoes in this study provided baseline information that would facilitate the release decision of future improved Fe- and Zn rich potato varieties for Ethiopia. Our result showed significant variation in mineral and tuber yield traits among the genotypes assessed and no significant gendered differences in farmer preferred potato traits (). The significant variation in the germplasm tested and associated high broad-sense heritability for most of the measured traits suggest progress in biofortification breeding could be made even with the current set of potatoes in Ethiopia. The values of mineral contents in the potato varieties assessed ranged from 16 to 25 ppm in dry weight for Fe and 8 to 23 ppm in dry weight for Zn as shown in and . These values are within the baseline range of 15–20 ppm for Fe (cf. Brown et al. Citation2010) and 11–20 ppm for Zn (White et al. Citation2009) in the potato tuber but not at high end for tuber Fe concentration. Furthermore, tuber Fe and Zn concentrations showed a negative relationship with tuber yield traits but such relationship was not statistically robust (). This means higher yielding clones in the current study were not the highest accumulator of Fe and Zn in their tubers, and those with higher Fe and Zn contents, in turn, were not the highest yielding clones. Such a trend of high yield with low mineral concentration has previously been reported in potato and other crops (cf. White et al. Citation2009). Tuber Fe and Zn concentrations didn't show a statistically robust relationship with the qualitative organoleptic traits: taste and texture as shown in . Our results generally indicate that potato varieties accumulating higher Fe and Zn concentration in their tubers were neither superior in tuber yield potential nor overwhelmingly preferred for their organoleptic properties. The distribution of micronutrient, agronomic and user preference traits across the germplasm sets evaluated suggests that neither selection alone nor breeding within any one of the sets is likely to result in micronutrient-dense varieties with high yield and preferred quality traits. Instead, population improvement that combines the complementary characteristics of selected clones from different sets of materials would likely enable the exploitation of heritable variability for new traits in tandem with the combination of favorable yield, quality and nutritional characteristics required for variety development. Breeding programmes require suitable strategies that breaks the negative linkage between target traits (Butcher Citation1994). Optimum strategies to overcome negative correlations between biofortification, productivity and user acceptability traits include assessment of component traits in mineral accumulation pathways that may relate negatively with productivity and/or acceptability, use of molecular markers that provide greater precision in selection of recombinant “correlation breaker” plants (genomic-assisted breeding), use of combined index selection, and developing then inter-mating independent populations for the negatively correlated traits are among the approaches likely to assist with complex breeding objectives. The latter approach would optimize the probability of identifying transgressive segregant “correlation breaker” individuals. A breeding effort of generating micronutrient dense potatoes with effective future applications requires a more realistic achievement closer to the grower conditions and consumers concerns. Engaging farmers or stakeholders in the selection of breeding materials is a powerful tool to orientate the biofortification effort introduced here to the needs and requirements of the end-users. The participatory assessment of new potato types in the current study didn't identify varieties consistently appreciated for both tuber yield traits and organoleptic quality, as reflected by the correlation result in . This suggested lack of materials in the current diversity that simultaneously address the growers concern for productivity and eating quality for potato in Ethiopia. A good potato variety for subsistence farming should possess the potential for high productivity and quality for consumption. Appearance, taste, and texture of boiled potatoes are perceived by farmers as equally important criteria to identify best clones for eating quality. CIP704205 (diploid Phureja) was appreciated by most of the panelists for both taste and texture, and showed moderate total tuber yield (24 tons ha−1), and Fe (17 ppm) and Zn (16 ppm) levels ( and ). Clones presenting the highest levels of Fe and Zn, all of which are from the Phureja germplasm set, (CIP703793: 21 & 19 ppm Fe & Zn; CIP703295: 19 ppm Fe and 20 ppm Zn, CIP706758 = 22 ppm Fe and 16 ppm Zn, respectively) were neither high yielder nor very highly ranked much for their organoleptic properties (taste and texture). Notable, diploid potatoes such as these Phureja landraces are generally considered to have lower yield potential than tetraploid potatoes, and they have not been subject to formal breeding for yield or uniformity expected in modern commercial varieties. Overall, our result showed the Andean landrace potatoes showed remarkable adaptation to African highlands and fairly appreciated by farmers for yield and consumption traits. The strong and positive association between tuber Fe and Zn contents and their negative or weak association with agronomic traits and organoleptic properties indicate the possibility for simultaneous improvement for high Fe and Zn in the different background for agronomic or organoleptic properties. Farmers are generally open to trying new potato types and hence the conventional approach of developing varieties that have the same appearance as those farmers are accustomed to growing may actually restrict the introduction and exposure of farmers to the novel, attractive and adapted germplasm with impact for food and nutrition security.
Acknowledgements
This work was supported by the HarvestPlus Challenge Program grant to CIP. We thank all CIP and Ethiopian Institute of Agricultural Research (EIAR), Holetta centre and nutrition laboratory staff directly or indirectly contributed to the field trials and laboratory analysis. The contribution of farmers, researchers, and technicians at Holetta and Kulumsa, Ethiopia who participated and assessed the potato clones for harvest as well as quality traits is highly appreciated.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes on contributors
AA, GW, SAK, MB designed and executed the research presented here, MB selected bred and landrace materials and obtained funding for the research, AA, GW, GM involved in user trait elicitation and participatory variety selection, DM, AB, GB, TF involved in laboratory analysis, AA conducted data analysis and wrote the first draft, and all authors edited the subsequent versions of the manuscript.
ORCID
Asrat Asfaw http://orcid.org/0000-0002-4859-0631
Additional information
Funding
Related Research Data
References
- Abebe GK, Bijman J, Pascucci S, Omta S. 2013. Adoption of improved potato cultivars in Ethiopia: the role of agricultural knowledge and innovation system and smallholder farmers’ quality assessment. Agr Sys. 122:22–32. doi: 10.1016/j.agsy.2013.07.008
- Abebe T, Wongchaochant S, Taychasinpitak T, Leelapon O. 2012. Variation of mineral concentrations among different potato varieties grown at two distinct locations in Ethiopia. Kasetsart J Nat Sci. 46:837–850.
- Alvarado G, López M, Vargas M, Pacheco Á, Rodríguez F, Burgueño J, Crossa J. 2015. “META-R (Multi Environment Trial Analysis with R for Windows) Version 5.0.” http://hdl.handle.net/11529/10201 International Maize and Wheat Improvement Center [Distributor] V17 [Version].
- Andre C, Legay S, Iammarino C, Ziebel J, Guignard C, Larondelle Y, Hausman J, Evers D, Miranda L. 2014. The potato in the human diet: a complex matrix with potential health benefits. Potato Res. 57:201–214. doi:10.1007/s11540-015-9287-3.
- Andre CM, Ghislain M, Bertin MP, Oufir M, del Rosario Herrera M, Hoffmann L, Hausman JF, Larondelle Y, Evers D. 2007. Andean potato cultivars (Solanum tuberosum L.) as a source of antioxidant and mineral micronutrients. J Agric Food Chem. 55:366–378. doi:10.1021/jf062740i.
- Asfaw A, Almekinders CJM, Blair MW, Struik PC. 2012. Participatory approach in common bean (Phaseolus vulgaris L.) breeding for drought tolerance for southern Ethiopia. Plant Breed. 131:125–134. doi: 10.1111/j.1439-0523.2011.01921.x
- Asfaw A, Bonierbale M, Khan MA. 2015. Integrative breeding strategy for making climate-smart potato varieties for sub-Saharan Africa. In: Low J, Nyongesa M, Quinn S, Parker M, editors. Potato and sweetpotato in Africa: transforming the value chains for food and nutrition security. Kenya: International Potato Centre; p. 134–142.
- Asfaw A, Woldegorgis G, Mulugeta D, Schulz S, Burgos G, Zum Felde T, Bonierbale M. 2014. Participatory evaluation of new disease resistant and Andean landrace potatoes in Ethiopia, poster presentation in European Association of Potato Research 19th Triennial conference, Brussels, Belgium.
- Bouis HE, Welch RM. 2010. Biofortification—a sustainable agricultural strategy for reducing micronutrient malnutrition in the global South. Crop Sci. 50:S20–S32. doi:10.2135/cropsci2009.09.0531.
- Brown CR. 2005. Antioxidant in potato. Am J Potato Res. 82:163–172. doi: 10.1007/BF02853654
- Brown CR, Haynes KG, Moore M, Pavek MJ, Hane DC, Love SL, Novy RG, Miller Jr JC. 2010. Stability and broad-sense heritability of mineral content in potato: iron. Am J Pot Res. 87:390–396. doi: 10.1007/s12230-010-9145-4
- Burgos G, Amoros W, Morote M, Stangoulis J, Bonierbale M. 2007. Iron and zinc concentration of native Andean potato cultivars from a human nutrition perspective. J Sci Food Agri. 87:668–675. doi: 10.1002/jsfa.2765
- Butcher PA. 1994. Genetic diversity in Melaleuca alternitolia: implication for breeding to improve production of Australian tea tree oil. PhD Thesis, Australian National University.
- Dessalegn Y, Mengistu F, Tesfaye Abebe T. 2008. Performance stability analysis of potato varieties under rainfed and irrigated potato production systems in northwestern Ethiopia. Eth J Sci Technol. 5(2):90–98.
- Devaux A, Kromann P, Ortiz O. 2014. Potatoes for sustainable global food security. Pot Res. 57:185–199. doi:10.1007/s11540-014-9265-1.
- EPHI. 2016. Ethiopian National Micronutrient Survey Report. Ethiopian Public Health Institute, Ministry of Health.
- Haverkort A, Koesveld F, Schepers H, Wijnands J, Wustman R, Zhang Z. 2012. Potato prospects for Ethiopia: on the road to value addition. Wageningen UR. http://edepot.wur.nl/244969.
- KidaneMariam H. 1979. Preliminary assessment of the responses of potato genotypes in the eastern, southern and central regions of Ethiopia. Eth J Agr Sci. 1(1):41–47.
- Kolech SA, Halseth D, Perry K, De Jong W, Tiruneh FM, Wolfe D. 2015. Identification of farmer priorities in potato production through participatory variety selection. Am J Potato Res. 92:648–661. doi:10.1007/s12230-015-9478-0.
- Kolech SA, Halseth D, Perry K, Wolfe D, Douches DS, Coombs J, De Jong W. 2016. Genetic diversity and selationship of Ethiopian potato varieties to germplasm from North America, Europe and the International Potato Center. Am J Potato Res. 93:609–619. doi:10.1007/s12230-016-9543-3.
- Machida-Hirano R. 2015. Diversity of potato genetic resources. Breed Sci. 65(1):26–40. doi:10.1270/jsbbs.65.26.
- Melash AA, Mengistu DK, Aberra DA. 2016. Linking agriculture with health through genetic and agronomic biofortification. Agri Sci. 7:295–307.
- Miller BDD, Welch RM. 2013. Food system strategies for preventing micronutrient malnutrition. Food Policy. 42:115–128. doi: 10.1016/j.foodpol.2013.06.008
- Paget M, Amoros W, Salas E, Eyzaguirre R, Alspach P, Apiolaza L, Noble A, Bonierbale M. 2014. Genetic evaluation of micronutrient traits in diploid potato from a base population of Andean landrace cultivars. Crop Sci. 54:1949–1959. doi:10.2135/cropsci2013.12.0809.
- Porras E, Burgos G, Sosa P, zum Felde T. 2014. Procedures for sampling and sample preparation of sweetpotato roots and potato tubers for mineral analysis. Lima, Peru: International Potato Center (CIP), Global Program Genetics and Crop Improvement. ISBN 978-92-9060-445-7. 13p. doi:10.4160//9789290604457.
- von Grebmer K, Saltzman A, Birol E, Wiesmann D, Prasai N, Yin S, Yohannes Y, Menon P. 2014. Global Hunger Index: Armed Conflict and the challenge of hidden hunger. Bonn: International Food Policy Research and Concern Worldwide. p. 51. doi:10.2499/9780896299580.
- Watanabe K. 2015. Potato genetics, genomics and applications. Breed Sci. 65(1):53–68. doi:10.1270|jsbbs.65.53 doi: 10.1270/jsbbs.65.53
- Welch RM. 2002. Breeding strategies for biofortified staple plant foods to reduce micronutrient malnutrition globally. J Nutr. 132(3):495S–499S. doi: 10.1093/jn/132.3.495S
- White PJ, Bradshaw JE, Finlay M, Dale B, Ramsay G, Hammod JP, Broadley WR. 2009. Relationship between mineral concentration in potato tubers. HortScience. 44:6–11.
- Woldegiorgis G, Gebre E, Lemaga B. 2008. Potato cultivar development. In: Woldegiorgis W, Gebre E, Lemaga B, editors. Root and tuber crops: the untapped resources. Addis Ababa, Ethiopia: Ethiopian Institute of Agricultural Research; p. 15–32.
- Woolfe JA. 1987. The potato in human diet. Cambridge: Cambridge University Press.