Sampling protocols and risk of error significance in molecular detection tests for fruit trees certification


  • Kummert, J. , Malice, M. , Marbot, S. , Lepoivre, P. , Steyer, S. & Oger, R. (2004). Sampling protocols and risk of error significance in molecular detection tests for fruit trees certification. Acta Horticulturae, (No.657),
Type Journal Article
Year 2004
Title Sampling protocols and risk of error significance in molecular detection tests for fruit trees certification
Journal Acta Horticulturae
Recnumber 3
Issue No.657
Endnote Keywords analytical methods|apples|certification|crown|detection|frequency distribution|fruit trees|leaves|plant diseases|plant pathogens|plant tissues|polymerase chain reaction|probability|risk|sampling|techniques|trees|Foveavirus|Trichovirus|
Abstract In the frame of certification, the sampling and detection protocols satisfy the quality standards by considering the probability to erroneously declare that an infected plant is virus-free (risk beta ). In this context, the control of the risk beta essentially lies on two factors: the adequacy of the detection technique towards the variability of the targeted virus and the representativity of the sample. The significance of this second factor has been addressed by the study of the detection of Apple stem pitting virus (ASPV) and Apple chlorotic leaf spot virus (ACLSV) from crude extracts of infected apple trees tissues by real time RT-PCR. Eight leaves harvested at different dates during the growing period were separately tested. For each date and each virus, the frequency distribution of the number of positive leaves was established and the probability distribution deduced. The risk of non-detection (risk beta ) is represented by the theoretical probability to find no positive leaf on the eight leaves tested. This risk was generally below 0.1%, with a maximum value of 0.8% at one date for ACLSV. For the protocol used, the detection of ACLSV and ASPV seems thus reliable during the period concerned. However, the distribution of the virus in the crown of a tree is not a static phenomenon and the risk of non-detection depends on internal and external factors; it notably varies according to the number of leaves tested per tree and the multiplication rate of the virus in the tree. For the same number of leaves tested, the risk of non-detection increases if the fraction of infected leaves decreases.
Notes Cited Reference Count: 3 ref. Journal article, Conference paper English
Author address Unite de Phytopathologie, Faculte Universitaire des Sciences Agronomiques, B-5030 Gembloux, Belgium.
Fichier
Lien ://20053058633
Authors Kummert, J., Malice, M., Marbot, S., Lepoivre, P., Steyer, S., Oger, R.