Populations of F. virginiana vary from being completely dioecious to trioecious. In western populations of F. virginiana, all three sexes are found in similar proportions, whereas in eastern populations, only hermaphrodites and females are observed, again in relatively equal proportions (Staudt, 1968; Luby and Stahler, 1993). Levels of fertility in hermaphrodites are highly variable, with large ranges of fertility being found in natural populations from pure males to pure hermaphrodites (Hancock and Bringhurst, 1979a, 1980; Stahler, 1990). In general, females are more fertile than hermaphrodites (Dale et al., 1992). Environment probably plays a role in the fertility of hermaphrodites (Stahler, 1990; Dale et al., 1992), but most of the control is genetic. The level of fertility in hermaphrodites appears to be polygenic as large ranges of fertility are generated in crosses of native males with cultivated hermaphrodites (Scott et al., 1962) and fruit set in native hermaphrodites can be greatly enhanced through recurrent selection (Luby et al., 1995).
F. virginiana has undergone considerable genetic differentiation like the other octoploid species, F. chiloensis. Staudt (1962) described four subspecies: (i) ssp. glauca (Wats.) Staudt – southern Rocky Mountains to north-west Canada and central Alaska; (ii) ssp. platypetala (Rydb.) Staudt – Rocky Mountains from Wyoming to Colorado, west to Sierra Nevada and Cascades Mountains; (iii) ssp. grayana (E. Vilmorin ex Gay) Staudt – meadows and forest margins from New York to Alabama, Louisiana and Texas; and (iv) ssp. virginiana Duch. – meadows and forest margins from eastern USA, Newfoundland and Yukon Territory. Staudt felt that ssp. virginiana was more common in the north and ssp. grayana more common in the south.
The western subspecies have yielded considerable taxonomic debate. Darrow (1966) considered the western types of F. virginiana to be a separate species F. ovalis, but Staudt (1962) did not give species rank to this material because of the lack of barriers to hybridization and the intermediacy of its characters. The separation of ssp. platypetala and ssp. glauca has also been questioned. Welsh et al. (1987) suggested that the two subspecies completely intergrade and probably should be referred to as a single taxon var. glauca. Others have followed this designation (Scoggan, 1978) or have not attempted to recognize infraspecific taxa (Dorn, 1984). What was formerly recognized as a distinct species, F. multicipita, was recently reduced to F. virginiana ssp. glauca f. multicipita when it was discovered that its unique morphology was caused by a phytoplasma (Catling, 1995).
Whereas F. virginiana ssp. virginiana and ssp. glauca appear to be largely distinct across their range, strawberries in the Black Hills and eastern front ranges of the Rocky Mountains may be introgressive swarms between ssp. glauca and ssp. virginiana (Luby et al., 1992; Hokanson et al., 1993; Sakin et al., 1997). In a mutivariate analysis of F. virginiana and F. chiloensis populations across the northern USA, Harrison et al. (1997b) found that Black Hill populations are intermediate between collections of eastern ssp. virginiana and western ssp. glauca for morphological traits, and when analysed with RAPD markers, the Black Hill populations were part of a large cluster group incorporating both ssp. virginiana and ssp. glauca (Fig. 1.9). The Black Hill populations probably represent relicts of the late Pleiocene when the Great Plains were mainly a boreal forest that provided a continuous habitat for hybridization between eastern and western forms of F. virginiana.
Fig. 1.9. Multivariate analysis of F. virginiana and F. chiloensis populations from across the USA. The factor scores are for (a) the first three principal components of the means of morphological variables; and (b) RAPD frequencies. Groups are labelled according to the Fragaria taxonomy of Staudt (1962). (From Harrison et al., 1997b.)
Natural hybrids of F. virginiana × F. chiloensis exist where native habitats of the two species overlap in British Columbia, Washington, Oregon and Northern California. Staudt (1999) has designated these hybrids as F. × ananassa ssp. cunefolia (Nutt ex Howell). The F. chiloensis populations noted on woodland-meadow sites by Jensen and Hancock (1981) probably represent stabilized hybrid derivatives of this cross. Hybrid populations of F. chiloensis × F. virginiana are common from Vancouver Island along the coast to Fort Bragg, California, and Staudt (1989) has indicated that ‘the