Estimation of morphological variation in seed traits of Sophora moorcroftiana using digital image analysis

Abstract

Sophora moorcroftiana is a leguminous plant endemic to the Qinghai-Tibet Plateau. It has excellent abiotic stress tolerance and is considered an ideal species for local ecological restoration. However, the lack of genetic diversity in the seed traits of S. moorcroftiana hinders its conservation and utilization on the plateau. Therefore, in this study, genotypic variation and phenotypic correlations were estimated for nine seed traits among 15 accessions of S. moorcroftiana over two years, 2014 and 2019, respectively from 15 sample points. All traits evaluated showed significant (P< 0.05) genotypic variation. In 2014, accession mean repeatability was high for seed perimeter, length, width, and thickness, and 100-seed weight. In 2019, mean repeatability for seed perimeter and thickness, and 100-seed weight were high. The estimates of mean repeatability for seed traits across the two years ranged from 0.382 for seed length to 0.781 for seed thickness. Pattern analysis showed that 100-seed weight was significantly positively correlated with traits such as seed perimeter, length, width, and thickness, and identified populations with breeding pool potential. In the biplot, principal components 1 and 2 explained 55.22% and 26.72% of the total variation in seed traits, respectively. These accessions could produce breeding populations for recurrent selection to develop S. moorcroftiana varieties suitable for restoring the fragile ecological environment of the Qinghai-Tibet Plateau.

Keywords: Sophora moorcroftiana; digital technologies; genotypic variation; image analysis; seed traits.