All,

At present I treat philodice, eriphyle and vitabunda as distinct species.

This is mostly based on two papers:

Pollock, David D., Ward B. Watt, Vanessa K. Rashbrook and Erika V. Iyengar

1998 Molecular phylogeny for Colias butterflies and their relatives (Lepidoptera: Pieridae). Annals of the Entomological Society of America 91(5): 524-531, 4 figs., 3 tbls. {Sep 1998}

Wheat, Christopher West and Ward B. Watt

2008 A mitochondrial-DNA-based phylogeny for some evolutionary-genetic model species of Colias butterflies (Lepidoptera, Pieridae). Molecular Phylogenetics and Evolution 47(3): 893-902, 5 figs., 6 tbls. {Jun, online 16 Mar 2008}

Some quotes from the last:

This group of ‘‘semispecies” (sensu Mayr, 1963) comprises C. vitabunda of the Northwest, C. philodice of the East and Midwest, C. eriphyle of the Intermountain West, and C. eurytheme which resides in the southern third of the continent and migrates north each summer. Members of the complex hybridize where they are sympatric; while morphological identity and parts of genomic identity remain distinct in such cases, other parts of the nuclear genome form a ‘‘syngameon” (Grant, 1981; Wang, 2005; Wang and Porter, in preparation). In our mtDNA data set, one C. vitabunda haplotype (CVI2) is closest to C. philodice (CPH542) and the other (CVI1) is closest to C. eurytheme (CEU9, CEU 543) and C. eriphyle (CER1), the latter being sister to California (CEU9) C. eurytheme

Placement of C. eriphyle as a subspecies of C. philodice (e.g. Layberry et al., 1998, and citations therein) is not supported by our results.C. vitabunda is basal in the lowland complex, and the several semispecies seem best referred to by simple binomial names. Genetic divergences are as expected from this phylogeny: CEU and CER are very close at 0.0006, the CVI haplotypes differ by 0.0036, and CPH542 differs from CEU and CER by 0.0057.

Systematics

mtDNA haplotypes reveal phylogenetic differences which were not evident from morphology. For example, earlier views of the lowland species complex are not supported here: C. philodice groups with one haplotype of basal C. vitabunda, while C. eriphyle, often called a ‘‘subspecies” of C. philodice (Klots, 1951; Layberry et al., 1998), is sister to (or derived from) western C. eurytheme, and these plus Midwest C. eurytheme group with the other C. vitabunda haplotype. More sampling and additional gene sequences are needed to clarify fully the biogeography and systematics of this group. Some aspects of these results are illuminated by data on hybridization, or lack of it, among the taxa involved. Within the lowland complex, C. eurytheme and C. philodice hybridize in the eastern US (Gerould, 1946; Remington, 1954; Wang and Porter, 2004; Wang, 2005), and so do C. eurytheme and C. eriphyle in the Southwest (Taylor, 1972). In contrast, C. eurytheme and C. meadii, in distinct subclades and with divergences of 0.009–0.010 in our data, do not hybridize in the wild, and, in the laboratory, can be forced to mate across taxa only rarely and then lay sterile eggs, although each taxon is easily mated inter se and reared in the lab with high fertility (e.g. Aé, 1959). Hybridization has been proposed, without evidence, as a cause of poorly understood morphological variation in other Colias (e.g. Hovanitz, 1962), but it now seems clear that this was wrong (e.g. Layberry et al., 1998).

JPPelham