Breeding farmer and consumer preferred sweetpotatoes using accelerated breeding scheme and mother–baby trials

Increased sweetpotato utilization has become an important breeding objective recently, with much emphasis on the development of non-sweet sweetpotatoes for income and food security in Ghana. The objective of this study was to evaluate 26 elite nonsweet and less sweet sweetpotato genotypes with regard to their release as commercial varieties using mother–baby trial. The 26 sweetpotato genotypes were tested multilocational on-farm across five ecozones from 2016 to 2017. These genotypes were selected from accelerated breeding scheme carried out from 2010 to 2013. There were no year-by-ecozone-by-genotype and year-byecozone interactions. However, ecozone-by-genotype interaction was significant for storage root dry matter, beta-carotene, iron and zinc content. This implies that the relative performance of the genotypes for storage root yield was stable across locations and years. Genotypic differences were found for all the traits and indicated that selection of superior genotypes across ecozone was possible. Storage root yield ranged from 7 t/ha to 39 t/ha, while dry matter content ranged from 34% to 46%. The storage root cooking quality preference was comparable with farmers’ check. Ten superior genotypes were identified for release as commercial varieties based on their staple-preferred taste, higher storage root yield, higher dry matter content, earliness, resistance to the sweetpotato virus, sweetpotato weevil and Alcidodes.


Introduction
Sweetpotato (Ipomoea batatas L. (Lam)) belongs to the botanical family Convolvulaceae (Thottappilly 2009) and its among the few crop plants of major economic importance in the family use for food globally (Eich 2008), which may be due to the Agrobacterium infection which occurred in its evolution (Kyndta et al. 2015). The potential of sweetpotato in food security and global well-  Baafi 2014). The decision to adopt a new cultivar is complexly related to field and yield performance as well as consumer taste acceptability (Sugri et al. 2012). Consumer preference is critical in determining the suitability of sweetpotato to any locality (Tomlins et al. 2004; Kwach et al. 2010). It is reported that some cultivars were not adopted because of lack of sufficient consideration of farmers' and consumers' preference (Toomey 1999 (Missah and Kissiedu 1994). Orange-fleshed sweetpotatoes were introduced to combat vitamin A deficiency at relatively cheaper cost but they have low dry matter content (Baafi 2014). High dry matter is one of the important attributes that affects consumer preference in most of sub-Saharan Africa (Tumwegamire et al. 2004). Development of end-user preferred sweetpotatoes has become key objective in sweetpotato breeding in Ghana (Baafi et al. 2016c) as higher yield is important in crop breeding (Rausul et al. 2002).
Successful development and release of staple-type sweetpotatoes requires accelerated breeding scheme (ABS) (Grüneberg et al. 2004) and mother-baby trial approach. The advantage of ABS is that each botanical seed of sweetpotato is a potential variety, and once the seeds rapidly multiply, multilocational field testing, which allows faster selection of promising varieties, takes place. A key part of on-farm trials is to conduct experiment on farmers' fields under farmers' conditions (John 1997). This creates opportunities for farmers to participate in the evaluation of varieties under their production environments. However, in larger breeding programmes, where the output of ABS results in a larger number of promising varieties, mother-baby trial approach, which allows quantitative data from researcher managed mother trials to be systematically crosschecked with farmer-managed baby trials with similar themes (Kamanga et al. 2001), is recommended (Mutsaers et al. 1997;Fielding and Riley 1998).
A key requirement and the final step in the development and release of improved crop varieties in Ghana involves at least two seasons, multilocational onfarm evaluation. The objective of this study was to evaluate 26 elite non-sweet and less sweet sweetpotato varieties developed through ABS on-farm with regard to their release as commercial varieties using mother-baby trial.

Materials and methods
The breeding work began with a survey aimed at identifying constraints and breeding priorities that will . Twenty-six elite F 1 s selected were tested multilocational on-farm in 2016 and 2017 using mother-baby trial approach. The 26 genotypes were divided into five groups, each subset having five genotypes (except group 2, which had six; Table 1). The trials were established in the major sweetpotato growing areas in the five ecozones of Ghana ( Table 2). Six farmers were selected at each ecozone in collaboration with the Ministry of Food and Agriculture staff. Five farmers were given a subset each for planting (baby trial). The sixth farmer planted all the 26 genotypes

Data collection
Twenty plants were harvested per plot for data collection. Storage roots considered were as reported by Ekanayake et al. (1990). The physicochemical traits determined were betacarotene, total sugars, starch, iron, and zinc content using the near-infrared reflectance spectroscopy (NIRS) (Tumwegamire et al. 2011). Dry matter content was calculated as the ratio of the weight of the dry sample expressed as a percentage of the weight of the fresh sample. In addition, the incidence and severity of diseases and pests (sweetpotato virus disease, sweetpotato weevil and Alcidodes) were scored on a scale of 1-5, where 1no disease/damage; 2minimum; 3average; 4high; and 5all plants affected. Incidence indicates the percentage of plants affected by disease or pest. At harvest of the mother trials, field days were organized for farmers to assess the vegetative part and the storage root yields as well as the cooking quality of the genotypes compared with their best-bet variety.

Data analysis
Data for 18 out of the 26 genotypes were analysed due to missing information alongside farmers' variety. The analysis excluded data on AGRA SP 02, AGRA SP 03, AGRA SP 10, AGRA SP 15, AGRA SP 18, AGRA SP 21, AGRA SP 22 and AGRA SP 26. The data were analysed using split-split plot design (YEAR = main plot; ECOZONE = sub-plot; GENOTYPE = sub-sub-plot). The data on the sensory evaluation were presented graphically.

Results
There were no year-by-ecozone-by-genotype interaction (Y × E × G) and year-by-ecozone interaction (Y × E) for  Table 5). In all, 13 genotypes had comparable dry matter content as the farmers check across ecozones over two years ( Table 5). All the genotypes were resistant to

sweetpotato virus disease, sweetpotato weevil and
Alcidodes. Cooking quality preference of the genotypes was comparable to the farmers' check ( Figure 1). Betacarotene content of the genotypes across ecozones over two years ranged from 0.73 mg/100 g DW (AGRA SP 11) to 28.46 mg/100 g DW (AGRA SP 20). Their iron and zinc values were 1.36-2.24 mg/100 g DW and 0.67-1.35 mg/ 100 g DW. These values were given by AGRA SP 24 and AGRA SP 16. The highest (18.12%) and the lowest (10.94%) total sugar content were given by AGRA SP 20 and AGRA SP 06, respectively, while AGRA SP 04 and AGRA SP 16 gave the highest (79.49% DW) and the lowest (67.73% DW) starch content, respectively ( Table 6).

Discussion
Mother-baby trial approach helped the farmers to gain experience with a few of the sweetpotato genotypes and rigorously assess them. Its use in the evaluation of crop varieties has been reported (Muungani et al. 2007;Ndhlela et al. 2007). The use of ABS in sweetpotato breeding has also been reported ).
Significant G × E for storage root dry matter, betacarotene, iron, and zinc content indicates that the sweetpotato genotypes varied for these traits relative to the different environments. Significant G × E for storage root dry matter and beta-carotene content has been reported (Chiona 2009;Oduro 2013). G × E interaction is important in evaluating genotype adaptation, selecting parents and developing genotypes with improved endproduct quality (Ames et al. 1999), and may complicate selection for such traits (Rosielle and Hamblin 1981; Falconer and Mackay 1996; Martin 2000; Ebdon and Gauch 2002;Gauch 2006). This is because progress from selection is realized only when the genotypic effects can be separated from the environmental effects (Miller et al. 1958). However, beta-carotene could be an exemption because of the orange-flesh colour associated with it (Gruneberg et al. 2015). The non-existence of G × E for storage root yield suggests that progress from selection for storage root yield can be realized (Mohammed et al. 2012; Nwangburuka and Denton 2012).
Significant differences observed among the sweetpotato genotypes for the traits indicate that superior genotypes can be identified and selected. The storage root yield of 11 of the sweetpotato genotypes tested was either higher or comparable to the farmers' best-bet. This indicates that farmers will adopt these genotypes along with their other preferred attributes.
Significant Suitability of a variety depends on the characteristics a farmer is looking for and includes sensory characteristics (Ndolo et al. 2001), and also diseases and pest tolerance. Of the 18 sweetpotato genotypes presented in the results, 11 were preferred as the farmers' best-bet when cooked. Stakeholders prefer sweetpotatoes with high storage root dry matter because that suits their food preparation preferences. Cooking causes changes in physical, sensory and chemical characteristics of the final product (Vitrac et al. 2000;Fontes et al. 2011). Low dry matter varieties lose mealiness when cooked, affecting textural characteristic preference. They also absorb more oil when fried, which is not economical to the processors and not healthy to the consumers.
Sugar content of the sweetpotato genotypes was comparable to those reported (Grüneberg et al. 2009b). The 11 non-sweet and less sweet genotypes selected during sensory test make them the staple-type sweetpotatoes preferred by Ghanaians. This is because sweetpotato genotypes that are non-sweet and less sweet allow daily consumption (Lebot 2010).
Sweetpotato has a considerable amount of genetic variation for beta-carotene All the genotypes were resistant to sweetpotato virus disease, sweetpotato weevil and Alcidodes, which are the major disease and pests attacking sweetpotato. This indicates that the superior genotypes when released as commercial varieties will be preferred by farmers. Based on the cooking quality preference, storage root yield, dry matter content, taste and resistance to major diseases and pests relative to farmers' best-bet, 10 genotypes AGRA SP 04, AGRA SP 05, AGRA SP 06 and AGRA SP 12 (bland-staple taste); AGRA SP 07, AGRA SP 09 and AGRA SP 13 (less sweet-staple taste); and AGRA SP 23, AGRA SP 19 and AGRA SP 20 (less-sweet orangeflesh) were recommended for release as commercial varieties to farmers. Four of these genotypes, AGRA SP 07, AGRA SP 09, AGRA SP 13 and AGRA SP 20, were officially released by the National Seed Council of Ghana as commercial varieties in June 2019 after recommendation for their release by the National Varietal Release and Registration Committee in 2018. Their respective varietal names are CRI-Vern Gracen, CRI-AGRA SP09, CRI-AGRA SP13 and CRI-Kofi Annan.