Improved method for assessing incidence of <em>Citrus tristeza</em> virus in large scale monitoring

• Main Authors: Anna Maria D'ONGHIA, Franco SANTORO, Yaseen ALNAASAN, Stefania GUALANO, Franco VALENTINI, Khaled DIELOUAH, Benedetto FIGORITO
• Format: Article
• Language: English
• Published: Firenze University Press 2015-04-01
• Series: Phytopathologia Mediterranea
• Subjects: sampling; DTBIA; DAS-ELISA; automatic tree extraction; standard error
• Online Access: https://oajournals.fupress.net/index.php/pm/article/view/5601
• ISSN: 0031-9465; 1593-2095

Citrus tristeza virus (CTV) induces typical quick decline when citrus trees are grafted onto the sour orange. The virus is efficiently vectored by different aphid species. Large scale CTV monitoring and eradication requires a precise sampling and testing method for assessing virus incidence. The sampling procedure described by Hughes and Gottwald (1998) was adapted in the official monitoring for the mandatory control of CTV in the Apulia Region of Southern Italy. However, instead of the hierarchical sampling (HS) of grouped samples tested by Enzyme Linked Immunosorbent Assay (ELISA), a systematic sampling of individual trees tested by Direct Tissue Blot Immunoassay (DTBIA) was applied by the Mediterranean Agronomic Institute of Bari (MIAB-S). In this study, the actual CTV incidence assessed in the eight citrus commercial groves by DTBIA was used for evaluating the efficacy and precision of MAIB-S and HS methods. The simulation of different sampling scenarios using advanced spatial (automatic tree extraction) and statistical (Standard Error, Root Mean Square Error) analyses showed that in all groves the infection rates were always underestimated using the HS method, while most infection rates were overestimated and closer to the actual CTV incidence when the MAIB-S method was performed. For the MAIB-S method, no effect of high temperature in the detection of virus incidence occurred during July monitoring. Unlike MAIB-S, noticeable reduction in the precision of CTV incidence estimation was observed with HS.