New biometric method used in Sweden

by Per Thunman, Stockholm

The pattern of the veins of the wings is specific for each breed of bees. To measure some characteristics of the pattern is an established method to distinguish breeds from each other. We have developed a computer program and a measuring method using a scanner or a digital camera to help us in the selection of the breeding material.

Background

As early as 1910 some scientists started to use biometric methods to distinguish breeds from each other.

The Russian bee scientist V. V. Alpatov performed in the years 1924 to 1958 an extensive and important study of the geographical influence of the morphology of the bees. Together with the Austrian scientist Friedrich Ruttner he also developed a method for classifying bees by using the cubital index.

The German scientist Goetze has further developed this method.

In Sweden there were some beekeepers during the 60s that were using wing index but not in a large scale.

We started to measure wings in the middle of the 80s by using a slide projector or a microscope.

The Cubital index

The cubital index, Ci, is defined as the quotient between the length of two veins, a and b, of the third Cubital cell.

This quotient differs from around 1 to above 3,
Apis mellifera mellifera having the lowest value and
Apis mellifera carnica the largest.

Discoidal shift Angle

Another property that differs between the breeds is the position of the crossing (8) of two veins in the discoidal cell compared to a line from the radial cell.

A pure mellifera should have this point to the left of the line as shown in the picture.
The discoidal shift angle, DsA, is then defined as negative.

Hantel index = "5-7" / "3-4"

Hantel index

Another property is what in Geman language is called the Hantel index (Dumb bell index).

It is defined as the distance between points 5 and 7 divided by the distance between points 3 and 4.

Measuring methods

At the first measurements we performed in the middle of the 1980s we used a slide projector and a callipers to measure the two lengths in the cubital cell.

The right fore wing from 25 - 30 bees were glued to a slide glass and projected on the wall.
We used an electronic calculator to determine the quotient and the results were noted by hand.
It was a time-consuming method.

To use a microscope and projectors are tiring. When scanners with high resolution became cheaper we acquired one.

We also developed a program for MS Excel where we entered the coordinates for eight points on the wing (as shown in the picture above).

We fixed the wings first on a strip of tape, e.g. Scotch Magic Tape 810, and then we attached the tape on a film used for OH projectors and put the film on the scanner.

Scanner requirements: The requirement for the scanner is an optical resolution of at least 1200 dpi.

Accuracy: The accuracy we get in a single wing measurement is around
for Ci +/- 0.06
for DsA +/- 0.25 degrees
for Hantelindex +/- 0.012


CooRecorder

Cybis CooRecorder program


The scanned image is opened in the CooRecorder program. We record eight points for each wing. The points are automatically recorded when we click on them with the mouse.


CBeeWing

The resulting coordinate file is then opened in the CBeeWing program. The result is two diagrams both with the discoidal shift angle as the y-axis. The left diagram has the cubital index as x-axis, the right diagram has the Hantel index as x-axis.

ure Mellifera
Mellifera queen inseminated with Mellifera drones
Hybrid colony
Mellifera queen free mated in surroundings of hive
The pictures show the results from a pure mellifera colony and a hybrid colony.

In the program we can introduce the limits for the values of cubital and Hantel index and discoidal shift angle we consider valid for a pure breed. The program calculates the fraction of bees in the sample that comply with the standard.

Selection criteria

Based on the results we have obtained from our tests with single drone insemination we have introduced the following criteria for pure mellifera:

Cubital index upper limit: 1.9
Discoidal shift angle upper limit: 0

According to Ruttner a mellifera bee could have a cubital index of 2. We still think it is better to use a lower index to really be on the safe side when selecting a colony as a base for further breeding.

Alleged Mellifera colony (from old Excel plot)

Conclusions

To breed pure bees we need more properties than colour for selection. To measure the vein pattern on the wings is a reliable method.
As computers and scanners are available for most beekeepers today we strongly recommend using them for selecting colonies for rearing.

We have seen many examples on colonies that show a typical mellifera colour but have wing properties that indicates a high degree of mixture with other breeds, see picture above.

To measure only the cubital index is not enough but combined with the discoidal shift angle it gives sufficient information for sorting out hybrids.

Even if a pure queen is mated by some pure drones and some from a different breed the offspring could be used in the breeding if the resulting queens are investigated by this method and those with wrong values are sorted out.



Copyright 2004, Cybis Elektronik & Data AB. Email: