Our recommendation is to plan at least 12 testing colonies from at least 3 sister groups.
Thus, most breeders should plan for a single testing apiary. Beside your own breed, include colonies from at least to other breeders.
Only breeders with at least 25 test colonies may consider to set up several testing apiarys.
Breeders with fewer than 8 testing colonies should consider to set up a common testing apiary with a neighbouring breeder.
However, data of apiaries not satisfying this condition should still be entered and released. They are important for completeness of pedigrees and recordkeeping. They just do not contribute information to the breeding values, they are now derived from parent colonies and other relatives.
The performance test of poorly performing colonies should still be continued. The data is essential to value the performance of the good colonies.
The data of deceased colonies is important for the same reason. Please also enter disease data and properties as possible.
Colonies without a performance test should be recorded in BeeBreed if they are ancestors of (tested) colonies in BeeBreed (as mother colony or mating partner). They are important for the completeness of the pedigree, thus, for the accuracy of relationships and inbreeding. They must be entered as tested and released to be considered. Empty fields in the perfomance/properties section clarify the status of the dataset.
Breeding by breeding values has higher requirements than breeding just by absolute perfomrance*. The reason is simple: it is only performance differences that matters. The performance as such is ignored**.
Testing only a single sister group at an apiary may help to find the best queen in the group, the value of the whole sister group will not be assessed. Foreign testing is the key to realistic breeding values.
The recommendation of 12 colonies considers a fault tolerance. If the performance tests of 9 colonies are completed this is still informative for the breeding values.
**Why is performance as such disregarded? For instance, honey yield depends critically on the weather, the nectar availability, and the care of the beekeeper. But breeding values only represent the genes. A genetically high-grade colony in bad condition will yield only little honey - but still more than a genetially worse colony. Oppositely, a genetically low-grade colony in great conditions will yield a lot of honey - but the genetially better even more.