Floodplain woodland

The white-flowered form in these pictures are from the same population as in this observation:

https://www.inaturalist.org/observations/25603344

A few dozen clumps of this white form and purple form were growing mixed together, seemingly forming a single population. While both forms were similar in many respects, they also differed significantly. In fact, I imagine that each form might be interpreted as a different species if seen in isolation. And while they could possibly be separate species, I have a hard time believing that two separate species with such distinct similarities would be growing mixed together like this.

The purple form here fits the more conservative, pubescent woodland form of V. sororia pretty well, although the leaves may be slightly more acute and deltoid than typical.

The white-flowered form is harder to place. Its leaves are glabrous, except for the petioles and along the main veins abaxially, at least proximally (visible in the other observation I linked to). The peduncles are glabrous. And while the leaves are very similar in shape to the purple form, they seemed to be more consistently acute-tipped.

The white flowered plants would fit into V. sororia, perhaps V. sororia f. priceana if not for the deltoid leaves, although the total lack of purple veins on the petals isn't typical for V. sororia f. priceana. If the leaves were a little narrower, I'd say this was a white flowered form of V. missouriensis, although the pubescent petioles might be atypical.

Flower color, pubescence, and shape of the leaf tips seem to be linked in this population. If there were more variation within the forms, or less characters involved, I'd suspect this to just be a case of polymorphism. Yet there was no apparent gradation between the forms. I guess this could still be the result of mutations within a single population, but I'd also guess that this explanation would require that many genes would be involved in order to affect this many characteristics. It's unlikely that so many mutations could happen independently. So this would possibly require a major, non-lethal chromosomal mutation. Another potential explanation is genetic drift - that a number of recessive genes happened to become homozygous at the same time by chance. However, genetic drift has a stronger effect on smaller, more inbred populations. Anything in the V. sororia complex is unlikely to suffer from inbreeding in this area, where V. sororia is extremely common. Another possible explanation is a loss of function mutation in genes involved in the production of both trichomes and pigment.

Most of the potential genetic explanations I can think of rely on rare events or aren't quite applicable here. But "hybridization" would explain the variation within this population easily. Hybridization is considered to be frequent in many violet "species" with a number of "species" often considered to be varieties of the same highly polymorphic species. I presume that the purple form is one of the parents of the white form, since it fits a well known, more conservative form of V. sororia. I didn't find any plants nearby that may have acted as the other parent. I hypothesize that the other parent may have had white flowers and been entirely glabrous. V. sororia f. priceana is not uncommon in the area. But, if f. priceana were the other parent, that wouldn't be a satisfactory explanation for the entirely white petals and more acute leaves. Of course, both parents may have contributed a recessive allele for flower color or any other suite of characteristics. With that in mind, there is no reason to assume that V. missouriensis or V. affinis couldn't be the other parent. Both of these "species" are present on the site but uncommon, are mostly glabrous, and have pointed leaf tips. Of course, I may be placing too much importance on the subtle differences in leaf shape for the two forms here. In which case, V. sororia f. priceana or another one of the weedier glabrous forms of V. sororia would be the best candidate for the other parent.

The generally even number of both forms in this population is somewhat puzzling. And even if the differences here are the result of hybridization, I'd expect that there would be more gradation between the forms, unless the hybrid happened to be sterile or somehow couldn't backcross with the nearby purple form. But if this hybridization event was recent, these white forms might all be F1 hybrids. I plan to check out this population in the subsequent years to see if I can locate another putative parent or see if backcrossing occurs.