Wing Network morphology came up in the "Why Treatment Free" thread.

I prepared a number of wing scans to look at this typing approach.

Ten wings from a hive with "Ukrainian queen" genes.

http://imageshack.us/a/img42/4636/avh6.jpg

The "cubital ratio" (which is the ratio of segment lengths on the middle window) is 3.06 s.d. 0.23, which is at the extreme upper end of the Apis mellifera range.

The "DrawWing.org" reference wing for Apis melllifera carnica

http://imageshack.us/a/img546/9269/2mo1.jpg

The DrawWing image has a cubital ratio of 2.73

A comments, the Discoidal angle -- used as a primary axis for detection of AMM depends on the precise location of two points along an wide obtuse curve and the precise location of a single nearby broadly intersecting point to extend a very much longer perpendicular. This value is thus highly error prone since both small shifts in the baseline angle and small shifts in the intersection of the erected perpendicular would move the value.

Modern papers use a 18 point network with 3 axis principal component analysis of the relationships among the larger network. I haven't yet imported this much expanded data set or tried the PCA.

An Italian x Feral hive has wings, a sample (n=10) yielded a "Cubital Index" of 2.24 (SD 0.23):

http://imageshack.us/a/img822/8199/bcrc.jpg