Original discovery of feral bees co-existing with varroa
A 2007 study by Seeley looking at the honeybee population of the feral bee population of the Arnot Forest showed a stable host - parasite relationship with the varroa mite.
An interesting part of this paper is that it describes how survivor bees co-existing with varroa were taken out of the Arnot forest and inoculated with mites from another apiary. As mite growth in inoculated colonies occurred at a level consistent with control colonies, we might guess less virulent mites have evolved in the unmanaged bee population of the Arnot forest.
The paper can be downloaded here: https://hal.archives-ouvertes.fr/file/index/docid/892236/filename/hal-00892236.pdf
An audio lecture covering the research in this paper can be heard here:
Research examining what advantages feral bees may have in the battle against varroa
In years following that paper, Seeley has spent time investigating some of the factors present in feral colonies that differ from what is typical of managed colonies. Word is that experiments that involved spacing colonies out at a large distances from each other, thus reducing drift and robbing, and bees kept in smaller hives that swarm more often, thus experience a break in the brood cycle, have shown promising results in terms of reducing overall mite load.
Genetic analysis of the Arnot forest survivor bees
In 2015 Alexander S. Mikheyev, Mandy M.Y. Tin, Jatin Arora & Thomas D. Seeley published a study showing that the mite resistant bees living in the Arnot forest are genetically distinct from the bees in the nearby apiaries. They also compared genes of current bees with museum sample of Arnot forest bees from the 1970's. This comparison shows there was likely a genetic bottle neck, and Arnot forest bees have evolved distinct traits as compared to the bees living in the forest prior to varroa. It also suggests that some influx of some amount of new genetics has occurred in the population, including those associated with africanized bees.