Although there may not be sufficient differentiation among the diploids to warrant the designation of separate genomes (Staudt, 1959), cytogenetic studies have indicated that there are distinct sets of chromosomes associating in the hexaploid and octoploid species (Federova, 1946; Senanayake and Bringhurst, 1967). Lerceteau-Köhler (2003), studying segregation ratios of AFLP markers in a full-sib family, found that 92% (727 out of 789) had simplex ratios and 8% (62 out of 789) fitted a multiplex ratio. This suggests that inheritance in the octoploid strawberries is mixed, being mostly disomic but not completely.
Cytogenetic studies indicated that at least two pairs of genomes are represented in the octoploid species. When they are crossed with the diploids F. vesca or F. viridis, bivalent or multivalent numbers approaching 14 are commonly observed in pentaploid hybrids, suggesting that there is pairing between one set of diploid and octoploid chromosomes, and another set of octoploid chromosomes (Ichijima, 1926, 1930; Federova, 1946; Bringhurst and Khan, 1963; Senanayake and Bringhurst, 1967). An additional set of chromosomes is left as largely univalents, either due to non-homology with the other sets or competition with a homologous set of chromosomes from the diploid. Similar results have commonly been obtained whether F. chiloensis, F. virginiana or F. × ananassa was used as the octoploid parent, although a few studies have reported much higher numbers of bivalents (21–28) in diploid × F. × ananassa crosses (Yarnell, 1931; Ellis, 1962).
Based on the cytogenetic studies, three genome formulas were suggested for the octoploids: AAAABBCC (Federova, 1946), AAA′A′BBBB (Senanayake and Bringhurst, 1967) and AAA′A′BBB′B′ (Fig. 1.10; Bringhurst, 1990). It seems likely that species similar to F. vesca and F. viridis are in the background of all the octoploid strawberries, as chromosomes from both pair regularly with those of F. chiloensis, F. virginiana and F. × ananassa. Federova suggested that the A genome came from an ancestor of F. orientalis other than F. vesca, the B from F. nipponica and the C from F. vesca. After examining the pairing relationships of octoploids crossed with 2x and 4x F. vesca, Senanayake and Bringhurst suggested that the genomic formula should be AAA′A′BBBB, as higher bivalent numbers were observed in hexaploid than pentaploid hybrids. This indicated that the chromosome of F. vesca had at least partial homology with another set of octoploid chromosomes. Their guess was that the A genome was contributed by either F. vesca or F. viridis, and they had no idea about the origin of the A′ and B genomes. Bringhurst (1990) later suggested that the genome formula should be AAA′A′BBB′B′ to reflect his contention that the octoploids are completely diploidized with strict disomic inheritance. Noguchi et al. (1997) produced hybrids between F. iinumae and F. × ananassa that were highly fertile after chromosome doubling, suggesting that F. iinumae contributed a genome to the octoploids. Liu et al. (2016) using genomic in situ hybridization (GISH), found that octoploid-derived gametes carried seven chromosomes with hybridization affinities to F. vesca, while the remaining 21 chromosomes displayed varying affinities to F. iinumae.
Fig. 1.10. Genomic origin of the octoploid strawberry species proposed by Bringhurst (1990). Different letters represent highly divergent species, while those distinguished with a prime originated from much closer relatives.
The first phylogenetic studies utilizing the sequences of nuclear low-copy genes indicated that one of the octoploid subgenomes was closely related to F. vesca or F. mandshurica, and a second subgenome donor was F. iinumae (Rousseau-Gueutin et al., 2009; DiMeglio et al., 2014; Sargent et al., 2016), together forming a AAA′A′BBB′B′ genome structure. More recently, Tennessen et al. (2014), using sequenced regions anchored in the F. vesca map, came up with a more complex subgenome compliment of 2Av,2Bi,2B1,2B2. The Av genome was hypothesized to have come from a diploid F. vesca ancestor, the Bi from a diploid F. iinumae ancestor, and the B1, B2 ancestor from an F. iinumae-like autotetraploid. The divergence between homeologous chromosomes appeared to have been greatly augmented by interchromosomal rearrangements. Kamneva et al. (2017) using ‘haploSNPs’ found that one or more diploid ancestors, possibly related to F. viridis, F. bucharica or F. mandshurica, formed a hexaploid with F. iinumae (2Bi, 2B1, 2B2), which then introgressed with a diploid F. vesca-like species (2Av). Using sequences of 24 single-copy or low-copy nuclear genes, Yang and Davis (2017) found that at least five diploid ancestors contributed to the subgenomes of octoploids (F. vesca, F. iinumae, F. bucharica, F. viridis, and at least one additional allele contributor of unknown identity) and that the composition of the subgenomes was complex and not derived from a single ancestral source.
Edger et al. (2019) examined the species origin of the octoploid species by generating a near-complete chromosome scale assembly for the cultivated octoploid strawberry (F. × ananassa cv. Camarosa) and comparing it with 31 sequenced and de novo assembled transcriptomes of every described diploid Fragaria species, including 19,302 nuclear genes in the genome. Their analysis revealed that four species are in the ancestry of the octoploid strawberry: F. iinumae, F. nipponica, F. viridis and F. vesca ssp. bracheata. Because the range of F. viridis overlaps with that of the hexaploid F. moschata, they hypothesized that these were the evolutionary intermediates between the diploids and the wild octoploid species. This conclusion has been questioned based