Identification of Prospective Soybean Accessions for the Creation of a Genebank Core Collection Based on High Density DNA Marker Data

Abstract

Genebanks typically collect and conserve as many existing genetic variations as possible. However, a smaller core population that is representative of the whole collection but maintains as much diversity as possible is more practical for in-depth phenotypic characterization, genetic studies, and plant breeding. Here we describe genotype-based core collection development from USDA soybean germplasm collection, which contains 20,095 accessions that had been genotyped using 50,000 single nucleotide polymorphism DNA markers. Analysis using trimmed marker data and Core Hunter 3 software identified 382 accessions with maximum genetic distances from each other. Population structure analysis of the selected accessions indicated that the software favored admixed individuals instead of those with pure ancestor population genotypes. Allele frequencies of most markers are largely conserved in the core population, which indicates successful maintenance of alleles in the core population. Phenotypic variations for branch number and yield were largely similar to the original population, but slightly different for seed weight and maturity, where very high seed weight and high-latitude maturity groups were less prominent in the core collection. The resulting core population can be used as a base population for any institutions interested in utilizing worldwide soybean genetic diversity for breeding and genetic studies.