I wrote up the following a few years ago [added note: it was not for iNat but for a different discussion elsewhere] to address this question of what we do taxonomically with wild vs domestic derivatives in mammals, but it could apply to other organisms as well. I still think our approach to naming is rather messy but I don’t have a good answer.
The recognition of subspecies has declined in recent years and, as we know, many still-recognized mammal subspecies have never been properly evaluated genetically. But, as far as I know, the subspecies concept as defined by Mayr and Ashlock (1991, Principles of Systematic Biology) still holds and has not been significantly reinterpreted:
“ … an aggregate of phenotypically similar populations of a species inhabiting a geographic subdivision of the range of that species and differing taxonomically from other populations of that species.”
Which to me raises a problem with using a subspecies designation to denote the domesticated form. Almost any domesticated mammal is the product of many generations of selective breeding and is at least potentially derived from various source populations and multiple distinct wild lineages. The subspecies in these cases is not denoting a geographically-defined natural variant of a more widespread species, it is describing a human-selected form (or forms) of the wild type, possibly derived from various geographically-separated sources that might not yet be identified. In some cases, the history of a domesticated form might never be fully unraveled as now-extinct wild forms may have been involved.
The domesticated form is a different animal, but still capable of interbreeding with its wild counterpart, so it potentially can be considered as conspecific. But it doesn’t fit the definition of subspecies as has been recognized for all other mammals. So what is it?
For practical reasons, I think the domesticated form should be recognized as a separate nominal species from its wild counterpart since the species designation captures the unique ancestry (which in most cases is not fully known) of that animal and doesn’t presuppose any equivalency with any natural subspecies in the wild species. Thus: Equus caballus (domesticated horse) and Equus ferus (wild horse; tarpan); Capra hircus (domesticated goat) and Capra aegagrus (wild goat; bezoar); Felis catus (domesticated cat) and Felis silvestris (wild cat); etc.
Yes, it runs counter to the biological species concept, but if you view this arrangement through the phylogenetic species concept, the approach makes more sense. It’s evident that a domesticated form, whatever its history, is on a separate evolutionary trajectory (thanks to human intervention) compared to the wild form (or forms) from which it was derived. And the species designation – moreso than the subspecies – would make clear when you’re talking about the domesticated animal vs. wild.
Using this arrangement, it follows that a feral population derived from the domesticated form would still be recognized by its domesticated species name. Rather than feral domesticated populations of Equus ferus or E. ferus caballus here in western North America, we have feral E. caballus.