Decision tree scoring system to guide selection for consumer preference in sweetpotato breeding trials

Area of Science:

• Agricultural Science
• Food Science
• Genetics

Background:

• Established a quantitative descriptive analysis (QDA) lexicon for the Uganda sweetpotato breeding program.
• Determined the link between QDA scores for sensory attributes and consumer preference for effective decision-making.
• Aimed to create a gender-responsive decision tree for an overall sweetpotato eating quality score to support demand-led breeding.

Purpose of the Study:

• Develop a gender-responsive decision tree for sweetpotato breeding.
• Integrate consumer preferences and sensory attributes into breeding selection.
• Facilitate demand-led targeted breeding for improved sweetpotato varieties.

Main Methods:

• Utilized quantitative descriptive analysis (QDA) data from Uganda sweetpotato clones, varieties, and landraces.
• Conducted consumer preference analysis in Kamuli and Hoima districts.
• Developed a regression decision tree model to calculate an eating quality selection index.

Main Results:

• Sensory characteristics like mealiness, sweetness, aroma, firmness, and color significantly impact consumer preference.
• 'D20', 'NASPOT 8', and 'NAROSPOT 1' were the most preferred genotypes.
• Identified threshold values for consumer acceptability of key sensory attributes and a positive correlation between color and liking.

Conclusions:

• Decision trees incorporating consumer needs and gender considerations enhance demand-led breeding effectiveness.
• The developed tool facilitates more effective varietal selection in sweetpotato breeding programs.
• Prioritizing sensory attributes important to end-users, particularly women, improves breeding outcomes.