UC Davis

Multispectral and hyperspectral imagery allow non-destructive and high-throughput monitoring of plant physiological traits. This study compares five different remote sensing approaches to identify Quantitative Trait Loci (QTLs) in a genetic mapping population resulting from the crossing of two Riesling and Cabernet sauvignon, two of economically important wine grape (Vitis vinifera) cultivars. Measurements were taken at different distances and resolutions from the grapevines using cameras or sensors measuring surface reflectance mounted in an aircraft, unmanned aerial vehicle (UAV), ground vehicle (ATV), hand-held, and a hyperspectral camera mounted on a tripod. Different reflectance bands and vegetative indices were analyzed to determine the most viable approaches for large scale grapevine phenotyping for breeding. Results indicate that remote sensing techniques can be used as preliminary screening to quantify QTL related to in field plant performance. Twenty QTLs were identified on chromosome 2, 6, 11, 14, 17 and 18. High resolution imagery obtained with a camera mounted on a UAV was the most efficient approach; using the UAV we could measure a wider range of variability within the population for the calculated vegetation indices, which permitted the identification of a higher number of QTLs than using any of the other platforms tested.