Re-creating ancient hybrid species' complex phenotypes from early-generation synthetic hybrids: three examples using wild sunflowers.

Can the complex phenotypes that characterize naturally occurring hybrid species be re-created in early-generation artificial hybrids? We address this question with three homoploid hybrid species (Helianthus anomalus, Helianthus deserticola, Helianthus paradoxus) and their ancestral parents (Helianthus annuus, Helianthus petiolaris) that are phenotypically distinct and ecologically differentiated. These species, and two synthetic hybrid populations of the ancestral parents, were characterized for morphological, physiological, and life-history traits in greenhouse studies. Among the synthetic hybrids, discriminant analysis identified a few individuals with the multitrait phenotype of the natural hybrid species: 0.7%-1.1% were H. anomalus-like, 0.5%-13% were H. deserticola-like, and only 0.4% were H. paradoxus-like. These relative frequencies mirror previous findings that genetic correlations are favorable for generating the hybrid species' phenotypes, and they correspond well with phylogeographic evidence that demonstrates multiple natural origins of H. deserticola and H. anomalus but a single origin for H. paradoxus. Even though synthetic hybrids with hybrid species phenotypes are rare, their phenotypic correlation matrices share most of the same principal components (eigenvectors), setting the stage for predictable recovery of hybrid species' phenotypes from different hybrid populations. Our results demonstrate past hybridization could have generated hybrid species-like multitrait phenotypes suitable for persistence in their respective environments in just three generations after initial hybridization.