SNPs markers can distinguish sections, species, some provenances and hybrid composition of Eucalyptus

Abstract

Eucalyptus is an economically important genus comprising more than 700 species in different subgenera and sections. Forest breeding programs have been exploiting the complementary characteristics of species to develop interspecific hybrids. However, species identification by morphological characteristics is challenging. This study evaluated the ability of SNP markers to distinguish sections, species, and provenances within the main Eucalyptus subgenera. We also identify the species compositions of known hybrids derived from controlled crosses. Samples belonging to 15 Eucalyptus species of subgenus Symphomyrtus (E. argophloia, E. brassiana, E. camaldulensis, E. tereticornis, E. grandis, E. longirostrata, E. pellita, E. robusta, E. saligna, E. urophylla, E. benthamii, E. dunnii, E. globulus, E. nitens, E. viminalis), three additional species of other two subgenera (E. deglupta, E. cloeziana, E. pilularis) and 44 interspecific hybrids were used in the study. SNP genotyping was performed using the 72K Eucalyptus Axiom array. After quality control data filtering (call rate> 95%, MAF> 0.05 and LD pruning), the dataset contained 22,230 SNPs genotyped in 528 samples. StructureSelector of Faststructure results indicated K=20 as the best clustering. This model could correctly separate all Eucalyptus subgenera and sections, assigned each species to a different cluster (except for E. pilularis and E. cloeziana) but was able to distinguish only one of the 17 provenances. Faststructure results could predict the ancestral composition of 70% the hybrids. Of all hybrids tested, 28 Faststructure correctly identified all the species known to be involved in their pedigree, 43 at least 50% of crossed parental and one had no parental species identified. Because unknown history and local intermating of some of the Eucalyptus species and provenances introductions in Brazil we cannot rule out the possibility that the presumed parental species were misclassified or were not pure species. In any case, SNP marker data can correctly discriminate subgenera, sections and species.