The Use of Hierarchical Sampling in the Surveillance Program for Plum pox virus Incidence in the United States

Plum pox virus (PPV) was first reported in North America in late 1999. As part of the eradication effort mounted in response to this finding, a Plum Pox Virus National Surveillance Program has been implemented in the United States. The survey method adopted for this program is a form of "group testing" that utilizes relationships between disease incidence at two levels in a spatial hierarchy, known as hierarchical sampling. Leaf material is collected from groups of four trees and tested by enzyme-linked immunosorbent assay (ELISA) for PPV infection without distinguishing between the individual trees in a group. PPV incidence at the scale of the individual tree is then calculated by means of a formula. Incidence data were collected during an initial survey conducted by the Pennsylvania Department of Agriculture from nine PPV-infected orchards in the fall of 1999 and from an additional 11 PPV-infected orchards during the spring and summer of 2000. The pattern of PPV-infected trees at the within-group scale was weakly aggregated. As a result, an effective sample size of v = 3.2, rather than the nominal group size of four, should be used in the formula to calculate an estimate of incidence at the scale of the individual tree from observations of incidence collected at the group scale.