Why can't it just be a genetic variation? I remember a while back there was discussion about achieving color by where in the mantle a bit of graft tissue was placed. If one mollusc can produce different colored pearls at one time (or is it called a multi-seeded batch?), I think it follows that they might well do it on one pearl. After all, we're looking right at it (proof) with the pondslimes.
Not only that, but I've shaved some bi-colored pearls after I removed the poor dye-job with bleach. These were cheap pearls, so it was just for fun.
I mostly saw peach and white combinations, and blue and peach. As I scraped I saw that the colors didn't blend like paint, or layer color like a transparency, but seemed rather to diffuse across a border.
The border shifted a bit like a spiral, but not that regular, and there wasn't much in the way of luster. I ought to just send my favorite blue/peach to one of you excellent photographer-scientists...just for science's sake.