Cost and accuracy of advanced breeding trial designs in apple

• Published in: Horticulture research 2016-03, Vol.3 (1), p.16008-16008, Article 16008
• Main Authors: Harshman, Julia M, Evans, Kate M, Hardner, Craig M
• Format: Article
• Language: English
• Subjects: 4014/159; 631/449/711; Agriculture; Biomedical and Life Sciences; Ecology; Life Sciences; Plant Breeding/Biotechnology; Plant Genetics and Genomics; Plant Sciences
• ISSN: 2052-7276; 2052-7276
• DOI: 10.1038/hortres.2016.8

Trialing advanced candidates in tree fruit crops is expensive due to the long-term nature of the planting and labor-intensive evaluations required to make selection decisions. How closely the trait evaluations approximate the true trait value needs balancing with the cost of the program. Designs of field trials of advanced apple candidates in which reduced number of locations, the number of years and the number of harvests per year were modeled to investigate the effect on the cost and accuracy in an operational breeding program. The aim was to find designs that would allow evaluation of the most additional candidates while sacrificing the least accuracy. Critical percentage difference, response to selection, and correlated response were used to examine changes in accuracy of trait evaluations. For the quality traits evaluated, accuracy and response to selection were not substantially reduced for most trial designs. Risk management influences the decision to change trial design, and some designs had greater risk associated with them. Balancing cost and accuracy with risk yields valuable insight into advanced breeding trial design. The methods outlined in this analysis would be well suited to other horticultural crop breeding programs. Crop breeding: Apple trials on trial An investigation using apples suggests that field trials could be optimized to reduce the cost of crop breeding without sacrificing accuracy. Field-testing potential new cultivars is expensive, particularly for long-lived crops such as fruit trees. An international team led by Julia Harshman at Washington State University modelled the impact of reducing the number of locations, years and harvests examined on cost and accuracy in apple trials. Their aim was to find cheaper trial designs that still allowed accurate assessments of yield, fruit quality and storage potential. They found that as fewer variables were included savings increased but accuracy was lost. The best designs reduced the number of locations and harvests tested, but maintained the number of years over which they were evaluated. The method demonstrated here should translate to a cultivar of horticultural crops.