Click on “Download PDF” for the PDF version or on the title for the HTML version.

If you are not an ASABE member or if your employer has not arranged for access to the full-text, Click here for options.

High Throughput Phenotyping for Fusiform Rust Disease Resistance in Loblolly Pine Using Hyperspectral Imaging

Published by the American Society of Agricultural and Biological Engineers, St. Joseph, Michigan

Citation:  2020 ASABE Annual International Virtual Meeting  2000872.(doi:10.13031/aim.202000872)
Authors:   Piyush Pandey, Kitt G. Payn, Yuzhen Lu, Austin J. Heine, Trevor D. Walker, Sierra Young
Keywords:   disease phenotyping; fusiform rust; population screening; spectral modeling; tree breeding    

Abstract. Loblolly pine is an economically important timber species in the United States, with almost 1 billion seedlings produced annually. The most significant disease affecting this species is fusiform rust disease, caused by the fungus, Cronartium quercuum f.sp. fusiforme. The breeding and deployment of disease resistant families has proven to be a successful strategy for combating rust. Testing for fusiform rust resistance in the greenhouse environment involves artificial inoculations carried out at the USDA Forest Service Resistance Screening Center in Asheville, North Carolina. Disease incidence is determined through visual inspection. However, an automated, high-throughput phenotyping method will improve both the efficiency and accuracy of the disease screening process. This study investigates the use of hyperspectral imaging for screening loblolly pine seedlings for fusiform rust resistance in the greenhouse environment. A nursery trial containing families with known in-field rust resistance data was created, and the seedlings were artificially inoculated with fungal spores. Hyperspectral images using a visible near infra-red camera were collected before inoculation, and regular scans were collected at approximately monthly intervals post-inoculation. Additionally, the disease incidence for all plants was scored with traditional methods based on visible stem galls. The hyperspectral images were segmented from the background using spectral methods, and algorithms were developed for discriminating stem pixels from needle pixels using both spectral and spatial features. Statistical discrimination models were built for classifying seedling scans into diseased and non-diseased classes. A classification model built using stem spectra was found to be more accurate compared to the model built using needle spectra.

(Download PDF)    (Export to EndNotes)