Estimating sensitivity and specificity of current approaches to detect Phytophthora Agathis in NZ kauri - Stage 1
E. Vallee, G. Jones, N. Cogger
To design a sampling protocol to collect the data needed to assess the diagnostic sensitivity and specificity of aerial inspection and laboratory test of soil sampling conducted in the same way as ground-truthing to diagnose kauri dieback. Knowing the sensitivity and specificity will inform the Kauri dieback surveillance program, in particular in the design of a protocol to classify an area as free of the pathogen. However, the design of a survey to detect freedom from disease is outside of the scope of this project.
Report recommended sampling a location that had an area showing high prevalence and an area showing low prevalence with a random selection of at least 800 sampling units. Sampling includes, aerial surveillance, ground truthing and laboratory diagnostics of samples collected.
A preliminary assessment was also calculated to determine the diagnostic sensitivity and specificity of remote sensing for the detection of kauri trees or dead trees. LiDAR and aerial images have a better sensitivity than hyperspectral at the expense of specificity. Sensitivity and specificity for the identification of kauri trees with clinical signs of stress could not be directly assessed. Further analysis should be done once the remote sensing research is complete.