Robert Brenchley wrote:

>I also wonder whether Amm genes were really  eliminated, given that they
>appear to have been well established by the  Victorian era. I always
suspect that what you really have over there is a  variable hybrid which
includes a good bit of Amm. 

* Well, you are right about that. In the southern US, over 20% retained
genetic markers from the original imports, here referred to A. m.
iberica/mellifera, which could include bees from Spain, France, England,
Germany, etc.

Quoted material:

> Restriction enzyme analysis of 422 feral honey bee colonies collected from
non-Africanized areas in the southern United States revealed that over 21%
of them had mitochondrial DNA (mtDNA) derived from a European race
established in North America by the 17th century, 77% of them had mtDNA
common in honey bees maintained by beekeepers and about 1% exhibited African
mtDNA.    

> The large percentage of carnica/ligustica haplotypes probably reflects the
influence of commercial colonies that have escaped since the mid and late
19th century introductions of these races. In addition, as descendents of
imported A. m. ligustica constitute the majority of current US commercial
honey bees, they probably continue to provide genetic input to feral
populations. However, the iberica/mellifera markers found in 92 of the 422
feral colonies remain from 16th and 17th century importations.   

See:

Mitoehondrial DNA evidence for the 19th century introduction of African
honey bees into the United States by N. M. Schiff and W. S. Sheppard
Experientia 49 (1993)

******************************************************
* Full guidelines for BEE-L posting are at:          *
* http://www.honeybeeworld.com/bee-l/guidelines.htm  *
******************************************************