Deciphering the Genetic Diversity, Population Structure, and Stability of Perilla (Perilla frutescens L.) Landraces of Northeastern Himalayan Region
摘要
Perilla, a versatile crop with rich nutraceuticals, is an important potential crop for food and nutritional security. Keeping this in view, a set of 42 Perilla germplasms from the northeastern Hill Region of India was assessed based on agro-morphology, nutritional traits, and molecular aspects. Enormous variations for grain yield per plant (6.35 to 40.13 g) vis-à-vis quality traits, namely, fiber (13.50 to 27.63%), fat content (18.75 to 46.58%), protein (13.55 to 28.61%), and carbohydrate (11.86 to 32.10%) contents, were observed. The morphological clustering grouped the 42 genotypes into three major clusters. The principal component analysis revealed 58.08% of variations accounted for the first three principal components. The number of inflorescences, yield of primary and secondary branches, biomass, and various nutritional traits displayed significant positive correlations with grain yield per plant. In the path coefficient analysis, the total number and length of inflorescences and the yield of primary and secondary branches per plant showed direct positive effects. Genotype plus genotype by environment (GGE) stability analysis identified NL-23 (40.13 g) and MN-16 (35.34 g) as the most promising genotypes, while the multi-trait genotype ideotype distance index (MGIDI) stability analysis identified MN-13 and MN-15 as superior genotypes. Molecular diversity based on 20 polymorphic SSR markers grouped the entire 42 accessions into three major clusters with 17, 12, and 13 genotypes each, respectively. The mean number of alleles per locus of 2.25, major allele frequency (0.49), number of effective alleles per locus (1.6), expected heterozygosity (0.35), Shannon’s information index (0.54), and gene diversity values of 0.59 were detected among the 42 genotypes. The PIC values varied from 0.35 (KWPE 48) to 0.63 (KWPE 53) with a mean of 0.51 per locus. The model-based genetic population structure detected a sharp peak of ΔK detected at K = 4, and the entire genotypes were grouped into four sub-populations. The analysis of molecular variance (AMOVA) revealed a 1.20% variation among the population and an 89.80% variation accounted for within the population. The principal coordinated analysis (PCoA) cumulatively explained about 63.03% of the total variations by the first two coordinates. The average inbreeding coefficient of subpopulations relative to the total population (FST) detected was moderate (0.102). The pairwise FST matrix revealed wide genetic distance between Arunachal Pradesh with Nagaland and Manipur germplasms. The information on the genetic variations and promising genotypes with stable yield performance unraveled in the study would be of immense help for future breeding of perilla for yield augmentation and quality trait improvement.