Image-based Phenotyping and Genetic Analysis of Potato Skin Set and Color

Abstract

Image-based phenotyping offers new opportunities for fast, objective, and reliable measurement for breeding and genetics research. In the current study, image analysis was used to quantify potato skin color and skin set, which are critical for the marketability of new varieties. A set of 15 red potato varieties and advanced breeding lines was evaluated over two years at a single location, with two harvest times in the second year. After mechanical harvest and grading, 7-8 representative tubers per plot were photographed, and the photos were analyzed with ImageJ to measure skinning (as % surface area) and skin color using the Hue, Chroma and Lightness (HCL) representation. The plot-based heritability was consistently high (> 0.77) across traits and environments; the genetic correlation between environments was also high, ranging from 0.81 to 0.98. Significant increases in Lightness and Chroma, as well as a decrease in skinning, were observed at the late compared to early harvest, while the opposite trends for color were observed after six weeks of storage. The three color traits were unexpectedly collinear in this study, with the first principal component explaining 86% of the variation. This result may reflect the physiology of red color in potato, but the highly selected nature of the 15 genotypes may also be a factor. Image-based phenotyping offers new opportunities to advance genetic gain and understanding for tuber appearance traits that have been difficult to precisely measure in the past.