Steve Sheppard is one who knows as much as anyone about the genetic basis
for honey bee identification. He coauthored a paper entitled:
"Identification of African-Derived Bees in the Americas: A Survey of
Methods" (2000) WALTER S. SHEPPARD AND DEBORAH R. SMITH. I didn't feel like
counting all the different methods but the paper runs 18 pages, so you get
the picture.

What is interesting to me is 

 > Currently, 26 subspecies of A. mellifera are formally recognized, based
primarily on morphometric characters. The subspecies concept, however,
relies on knowledge of the geographic partitioning of the overall diversity
within a species. Some authors have argued that the subspecies is inherently
subjective and thus its use and validity in systematics is questionable. The
merits of this debate are beyond the scope of this article -- yet they are
suggestive of the difficulties that have arisen in discussions of the
introduced populations of honey bees in the New World.

In other words, if the experts can't agree on what a subspecies is (formerly
"race") -- then we are forgiven for not be able to get any of it straight

pb

****************************************************
* General Information About BEE-L is available at: *
* http://www.honeybeeworld.com/bee-l/default.htm   *
****************************************************