Performance Testing

Performance testing means measuring and comparing traits of bee colonies such as gentleness, swarming tendency, disease resistance, wintering and productivity under the same conditions, on the same scale and according to a set schedule. In breeding, it moves breeder selection from memory to data. A well-set-up test apiary and a simple scoring sheet show within a few seasons which line is genuinely superior.

Why is performance testing needed?

Assessment based on memory is misleading. The colony that gave a lot of honey last summer is easily counted as "the best"; yet perhaps it was next to the only major forage source, or did not swarm that year. A colony inspected on a bad day, meanwhile, is unfairly labelled bad-tempered.

Testing reduces these errors: the same trait is measured on the same scale, in the same period and repeatedly. In the end you have comparable numbers between colonies.

How do you set up a test apiary?

Test results are only meaningful if the colonies are kept in similar conditions. To avoid mistaking environmental differences for genetic differences, follow these rules:

• Same apiary: The colonies being compared must be in the same place and use the same forage. Comparing different apiaries measures differences in forage, not genetics.
• Queens of the same age: The queens tested should have been reared in the same season and introduced to the colonies at similar dates.
• Equal start: The colonies are started with the same number of frames of bees and brood, in the same hive type.
• Same management: Feeding, adding boxes and varroa control are applied to all at the same time and in the same way.
• Mixed layout: Sister groups are mixed rather than placed side by side. In long straight rows, flying bees drift into neighbouring hives (drifting); varying hive orientation and colour reduces this.

Which trait is measured when?

The test period usually starts in the summer after the queen begins laying in the colony and lasts until the end of the following year's harvest; this way one wintering and one full season are observed.

What should the scoring scale look like?

There is no need for a complicated system; what matters is consistency: the same scale, if possible the same person, and similar weather conditions. Many breeding programmes use a four-step scale with 4 the best and 1 the worst.

A similar scale can be used for swarming tendency: 4 no swarm preparations at all; 3 empty or egg-containing queen cups present but not progressing; 2 swarm cells with larvae built, prevented by intervention; 1 the colony swarmed.

How do you measure varroa and disease resistance?

Resistance cannot be seen directly; it is measured indirectly. The most practical method is to count the mite load in a sample of bees taken from all colonies on the same dates:

1. From a brood frame, taking care not to take the queen, collect about 300 bees (about half a cup).
2. Shake the bees in alcohol or soapy water, or separate the mites using the icing sugar method.
3. Count the mites that come off and calculate the number of mites per 100 bees.
4. Repeat the same measurement in late summer; colonies with a low rate of increase stand out.

Counting the natural mite drop on the bottom board is also a helpful indicator, but it is strongly affected by colony strength and the amount of brood. No colony should be given separate treatment between measurements; otherwise the results cannot be compared.

The hygienic behaviour test is explained step by step on a separate page. Colonies in which brood disease is seen are noted separately from the test; if a notifiable disease is suspected, it is reported to the official veterinary service / local agriculture authority.

How do you measure colony development, wintering and productivity?

• Development: In spring and early summer, the frames covered with bees and the frames with brood are counted. It is important that the same person counts in the same way.
• Wintering: The hive is weighed in autumn and spring to calculate the stores consumed over the winter; the number of frames of bees in spring and any losses are recorded.
• Productivity: At each harvest the honey taken per hive is weighed. Honey frames left in the hive or moved to splits are also recorded.

How are test results evaluated?

An observation from a single inspection can mislead. The colony may have behaved differently that day because of the weather, a break in forage or an earlier intervention. Follow this approach in the evaluation:

1. Take the average: Calculate the average of the scores each trait received over the season.
2. Compare with the apiary average: Assess each colony's result against the average of the colonies in the same apiary; "how far above average" is a fairer question than "how many kilos".
3. Set problem colonies aside: Exclude from the test for that trait colonies that became queenless, had their queen replaced or fell ill; do not cull them as genetically poor.
4. Weight the traits: Create a simple total score that gives more weight to the traits most important to you.
5. Decide: The top few colonies are breeder candidates; the queens of those at the bottom are replaced.

Common mistakes in performance testing

• Deciding on a single measurement: A consistent picture requires observations repeated through the season.
• Treating colonies differently: Giving one colony a box early and another late disconnects the yield difference from genetics.
• Changing the scale: Changing the scoring criteria mid-season makes the earlier data unusable.
• Filling in records later: Write the score at the hive; a score remembered in the evening is often wrong.