1958), F. nipponica (Lilienfeld, 1933), F. orientalis (Federova, 1946), and with difficulty with F. vesca (Mangelsdorf and East, 1927).
In the older studies on crossibility, F. orientalis and F. viridis were shown to cross relatively easily and the resulting hybrids were partially fertile (Federova, 1946). Hybrids of F. orientalis and 2x F. vesca are much more difficult to make, although Staudt (1952) was able to produce a fertile hexaploid hybrid between them, and Bors and Sullivan (1998) found the cross of 4x F. vesca and F. orientalis to be relatively easy to make. Schiemann (1937) described plants that looked like F. orientalis that were derived from a pentaploid F2 population of F. vesca × F. moschata.
The diploid–tetraploid relationships previously proposed in the ‘China’ clade by Staudt (2008) based on geographical and morphological similarities (F. pentaphylla > F. tibetica, F. chinensis > F. gracilis/F. corymbosa, F. nubicola > F. moupinensis) has not been supported by recent molecular phylogenetic studies (Kamneva et al., 2017; Yang and Davis, 2017). It appears that F. corymbosa, F. gracilis and F. moupinensis all share F. pentaphylla and F. chinensis as parent, and F. tibetica is most likely derived from F. pentaphylla and F. nubicola.
Dominance of subgenomes in octoploid strawberries
The octoploid genome sequence provided by Edger et al. (2019) along with gene expression data show that the last species to enter into the octoploid species, F. vesca, is the dominant subgenome ‘with significantly greater gene content, gene expression abundance, and biased exchanges between homoeologous chromosomes, as compared with the other subgenomes’. Their data fits the ‘subgenome-dominance theory’ which predicts that genome-wide expression disparity can arise when the merged genomes differ in their transposable element (TE) complement and in their level of TE-mediated repression of gene expression (Bertioli, 2019). Edger and his group found that F. vesca has 20% fewer transposable elements, has retained 20% more genes and has generally higher gene expression. The metabolic pathways that give rise to strawberry flavour, colour and fragrance are largely controlled by this dominant subgenome.
Origin of the North American octoploid strawberries
The phylogenetic study of Edger et al. (2019) combined with the geographic distributions of extant species supports a North American origin for the octoploid strawberry with F. vesca ssp. bracheata being the last diploid to be added in the formation of the ancestral octoploid strawberry. It is likely that F. chiloensis and F. virginiana are extreme forms of the same biological species, which emerged in Beringia and subsequently evolved differential adaptations to coastal and mountain habitats of North America. The north-eastern Asia distribution of F. iturupensis suggests that it may have been part of the genomic pool that originated in north-east Asia before spreading across the Bering Strait to north-western North America. The rise of the octoploid clade is estimated to have occurred 0.37–2.05 million years ago (Njuguna et al., 2013).
While F. chiloensis and F. virginiana are completely interfertile, there are significant morphological distinctions between them. F. chiloensis has thick, dark-green, coriaceous leaves and large achenes, whereas F. virginiana has thin, bluish-green leaves and smaller achenes (Staudt, 1999). Substantial separation of F. virginiana and F. chiloensis has been observed in cluster analysis of morphological characters, simple sequence repeats and RAPDs (Fig. 1.9; Harrison et al., 1997b; Hokanson et al., 2006). Yang and Davis in their phylogenetic study also found substantial divergence, finding a number of well-supported clades comprised of sequences exclusively from F. chiloensis and F. virginiana.
The origin of Hawaiian and Chilean F. chiloensis is also obscure, but presumably they were introduced from North America via bird migrations. Dillenberger et al. (2018) found F. chiloensis from Oregon and Northern California to be sisters to a Chilean sample of F. chiloensis, but only a handful of F. chiloensis clones were analysed. It is likely that multiple introductions were made into Chile as the habitats of South American F. chiloensis have an extensive range from beaches and headlands to montane forests at 1900 m elevation (Darrow, 1966; Cameron et al., 1991, 1993). Glaucous forms of octoploids may also have been introduced, as at high elevations in Chile leaf colour and thickness resemble F. virginiana ssp. glauca (Cameron et al., 1993).
Origin of decaploid species
Because of the current sympatry between the octoploid F. virginiana ssp. platypetala, diploid F. vesca ssp. bracteata and the decaploid F. cascadensis in the Cascade Mountains of Oregon, it was originally hypothesized that F. cascadensis originated from hybrid polyploid speciation between the two octoploid and diploid congeners (Hummer, 2012). However, Wei et al. (2017a) discovered its origin was more likely due to an ancient hybrid speciation in Beringia. In a phylogenetic analysis of linkage-mapped chromosomes from targeted sequences, they found that the additional subgenome of F. cascadensis was derived from a F. iinumae-like diploid progenitor rather than F. vesca ssp. bracteata. The F. cascadensis found in the Oregon Cascade Mountains may be a remnant of a more widespread species that survived in a Willamette Valley refugium during the last glacial maximum (Dillenberger, 2018). Whether F. cascadensis and the other decaploid F. iturupensis have a common ancestor or evolved independently is unknown. They could have both evolved in Beringia and then moved southwards in different directions along the Pacific Ocean and became isolated (Wei et al., 2017a).
Ahmadi H. and Bringhurst R.S. (1991) Genetics of sex expression in Fragaria species. American Journal of Botany 78, 504–514.
Ahokas H. (1995) Fragaria virginiana and F. vesca on the southern coast of Finland. Annales Botanica Fennici 32, 29–33.
Arulsekar S. and Bringhurst R.S. (1981)
