Evaluation of Sampling and Testing Efficiencies of Plum pox virus Eradication Programs in Pennsylvania and Ontario.

Abstract

Plum pox virus (PPV) was first detected in the United States in Pennsylvania in 1999 and in Ontario, Canada in 2000. Following a 10-year survey and eradication program, PPV was officially declared eradicated in Pennsylvania in 2009. Although incidence of PPV was greatly reduced from 2000 to 2008, PPV remains present in Ontario. The objective of this study was to compare how the Pennsylvania and Ontario PPV eradication programs affected the probability of detecting PPV at the leaf, scaffold, tree, and Prunus orchard block scales. A simulation model was developed to evaluate the sampling and testing efficiency of the two programs. At the tree scale, the Pennsylvania sampling and detection protocols had a detection efficiency of 71.8% compared with 40.5% for the Ontario program. Several components in the Pennsylvania and Ontario PPV eradication programs affected PPV detection efficiency. A stratified (by tree scaffold) random sampling design did not improve PPV detection efficiency in either program, compared with a simple random sampling design to select leaves for enzyme-linked immunosorbent assay (ELISA) testing. Detection efficiency for both programs increased with sample size but gains in detection efficiency diminished as sample size increased. There was good agreement (between the commercial ELISA kit used in Pennsylvania and the kit used in Ontario) at the leaf and scaffold scales but not the tree scale. Based on simulation modeling, the Pennsylvania PPV eradication program required that >2 PPV-positive trees must be present within a Prunus block to achieve a 95% probability of correctly detecting PPV at the block scale, whereas the Ontario program required >5 PPV-positive trees within a block to achieve 95% probability of detection. The results from this study have important implications with regard to the efficiency of the two eradication programs to detect PPV-positive trees.