2019 ASHS Annual Conference

Pumpkin (Cucurbita spp.) is a major vegetable and three species (C. moschata, C. maxima, and C. pepo) are commonly cultivated worldwide. To identify genome-wide single nucleotide polymorphisms (SNPs), we collected 48 commercial F₁ pumpkin cultivars consisting of 18 C. moschata, 15 C. maxima, 7 C. pepo, and 8 interspecific hybrids (C. moschata x C. maxima). Genotyping by sequencing found a total of 202,722 SNPs across 20 chromosomes. These SNPs were pre-filtered with three selection criteria: minor allele frequency of > 5%, missing data rate of < 10%, and minimum depth of 5X. The resulting 37,869 SNPs were further filtered to develop a core set of 400 SNPs based on the polymorphism information content values of ≥ 0.3 and genome distribution. Of these core SNPs, 192 SNPs were used to genotype additional 188 accessions (94 F₁ cultivars, 50 breeding lines, 44 landraces) for validation. One hundred and seventy SNPs (88.54%) showed polymorphisms in these accessions. With the 170 SNPs, genetic variations were assessed between and within the predefined groups of 223 cultivated pumpkin accessions. Principal component analysis and hierarchical clustering found four sub-populations representing three pumpkin species and interspecific hybrids. These genetic differentiations were supported by pairwise estimates of F_st. The sub-population consisting of interspecific hybrids showed a higher level of genetic diversity relative to the other three sub-populations. Furthermore, 206 of 223 accessions (all 85 breeding lines and landraces and 121 of 138 F₁ cultivars) were distinguished using the 170 SNP markers. The SNPs identified in this study will be a useful resource to develop breeder’s tools for cultivar identification, seed purity test, and marker-assisted selection in pumpkin.