Cost Considerations in Choosing Group Size for Group Testing in the Seed Potato Certification Program

• Published in: American journal of potato research 2010-10, Vol.87 (5), p.472-478
• Main Authors: Chiang, Kuo-Szu, Lai, Sin-Hong, Chung, Wen-Chuan, Lin, Shang-Hu, Yang, Tso-Chi
• Format: Article
• Language: English
• Subjects: Agriculture; Agronomy. Soil science and plant productions; Biological and medical sciences; Biomedical and Life Sciences; Fundamental and applied biological sciences. Psychology; Life Sciences; Plant Breeding/Biotechnology; Plant Genetics and Genomics; Plant Pathology; Plant Sciences
• ISSN: 1099-209X; 1874-9380
• DOI: 10.1007/s12230-010-9149-0

Viral diseases that can reduce yield and tuber quality are major concerns for potato growers. To produce healthy potato tubers and prevent yield reduction, a certification scheme for foundation class seed potato propagation has been established at the Seed Improvement and Propagation Station in Taiwan. The certification scheme uses an ELISA technique to index specific virus diseases such as Potato virus Y (PVY), Potato leafroll virus (PLRV), Potato virus X (PVX), Potato virus A (PVA), and Potato virus S (PVS). In order to produce foundation class potato tubers, the Dorfman group testing procedure is adopted. The Dorfman testing procedure utilizes a two-stage sampling scheme. The first stage comprises making measurements on all groups. The second stage involves no retesting for negative testing groups and exhaustive testing of all constituent individual samples of positive testing groups. In the seed potato certification scheme in Taiwan, field investigations have usually used groups of 20 plants. However, because of the limitations of available financial resources, the effect of group size on cost is an important consideration. Thus, selection of group size for design considerations is a critical issue. In order to solve the problem, we present the expected cost model as a nonlinear integer programming problem and determine the optimal group size such that the expected cost is minimized. The result demonstrates that the optimal sample size for group testing is 13 plants per group and that the cost, using this optimal group size, is not greatly different from the cost when using 20 plants per group. Thus, 20 plants per group is reasonable for the group testing procedure with regard to cost considerations. The result of this research could be useful in the execution of the seed potato certification program.