Allow me to get somewhat hypothetical.

Has anyone considered the possibility that the "marker" trait has inadvertently been selected for but in and of itself is not actually a product of the actual pied allele??