411928.vcf

The carrot germplasm collection is a fantastic source of novel, beneficial alleles that can be used to improve carrot breeding lines. Diversity panels have been used for genomic studies to identify loci related to flavor, disease resistance, and carrot color. Many previously studied carrot traits have major loci with a large effect on the phenotype and can be scored into discrete phenotypic categories. However, most plant (and carrot) phenotypes are quantitatively inherited and environmentally influenced. In most cases an informative phenotype is predicated on accurate measurements from an appropriate assay and robust experimental design with adequate replication to estimate the phenotypic mean and environmental influence. Efforts to score quantitative traits in the carrot germplasm collection are challenging because of intra-accession variation within carrot germplasm resulting in large differences in carrot phenotypes and genotypes within the same accession. This heterogeneity poses significant challenges, including limiting experiment size, because each accession will need multiple plants genotyped. Additionally, it is more difficult to differentiate variation resulting from error, environmental effects or true intra-accession variation. In these analyses we review the extent of genetic diversity within carrot accessions, demonstrate the risk of using a single plant to represent the genotype of an accession, and describe plans to use pooled genotyping and/or focus on germplasm with lower intra-accession genetic variation. Continued efforts to characterize genetic variation within and between carrot accessions will provide carrot researchers methods to more effectively use the germplasm collection to identify beneficial alleles and map associated loci.

Genotypic data was generated from lots of 25 carrot seeds extracted with the Macherey Nagel NucleoSpin 96 Plant II extraction kit. These pooled DNA samples were genotyped using genotyping-by-sequencing (GBS) at the University of Wisconsin Biotech Center, essentially as described by Ellison et al. (2018). SNPs were called using the Tassel 5.2.39 pipeline using the Carrot Genome Assembly DH1 v3.0 as the reference genome. Taxa with low sequencing coverage (less than 8,000,000 reads) were removed. Variants were filtered with average depth ≥ 10, minor allele frequency ≥ 2.5%, and percent missing ≤ 30%.