Agronomy (Apr 2022)
Diversity Characterization of Soybean Germplasm Seeds Using Image Analysis
Abstract
Soybean (Glycine max) is a native field crop in Northeast Asia. The National Agrobiodiversity Center (NAC) in Korea has conserved approximately 26,000 soybean germplasm and distributed them to researchers and growers. The phenotype traits of soybean were investigated during periodic multiplication. However, it is time-consuming to collect sufficient data, especially on the width and height of seeds. During the last decade, the development of phenomics efficiently assisted the analysis of high-throughput phenotyping seed morphology. This study collected and analyzed seed morphological traits of 589 germplasm (53,909 seeds) from diverse origins using a digital camera and a computer-based seed phenotyping program. Measured traits included size and shape, 100-seed weight, height, width, perimeter, area, aspect ratio (AR), solidity, circularity, and roundness. The diversity of soybean germplasm seeds was analyzed based on 8-seed morphological traits and 100-seed weight, as determined by image phenotyping and direct weighting, respectively. The data obtained from 589 soybean germplasm were divided into five clusters by k-means clustering. Orthogonal projections to latent structures discriminant analysis (OPLS-DA) were performed to compare clusters. The major differences between clusters were in the order of area, perimeter, 100-seed weight, width, and height. Based on cultivar origins, the seed size of US origin was the largest, followed by Korea and China. We classified size, shape, and color according to the International Union for the Protection of New Varieties of Plants (UPOV) guidelines. In particular, we postulated that shape could be distinguished based on the AR and roundness values as secondary parameters. High-throughput phenotyping could make a decisive contribution to resolving the phenotyping bottleneck. In addition, rapid and accurate analysis of a large number of seed phenotypes will assist breeders and enhance agricultural competitiveness.
Keywords
