FAQ BeeBreed and breeding values

How was the ZWS 2024 of the main Carnica population carried out?

The BVE was carried out on the basis of a proposal by DIB Zuchtbeirat on 12 February 2024, which includes the mutual use of pedigrees of the main Carnica population and the population of Austrian professional beekeepers up to the year of birth 2021, but excludes the mutual use of performance and trait data. As explained below, there are limitations in accuracy due to the lack of consideration of external tests.

The population of Austrian professional beekeepers, the Association Mésange Huppé and the Carnica breeding of Norges Birøkterlag receive separate breeding value estimates.

Why do the ACA and some DIB representatives not want the AT-2 association in the main Carnica population?

Many breeders in AT-2 are commercial beekeepers and test a relatively large number of colonies on their companies, Dr Mandl currently tests around 1000 queens per year. The commercial orientation results in differences in the organisation of performance testing, which leads to irritations in small-scale breeders.
The founding of AT-2 was accompanied by organisational upheavals in the ACA, which had both human and professional reasons. Therefore, from an objective point of view, the formation of a separate breeders' group was the appropriate solution.
Temporary issues due to fast increase in test colonies in Bienenhof Mandl were exaggarated for political impact in these upheavals.
Breeders from the DIB took sides in the ACA processes because there are long-standing cooperative relationships.

Will the ACA remain on BeeBreed?

In the ACA board resolution of 13 January 2023, the ACA announced its termination and also abandoned the path of cooperative cooperation, attempting to influence the BeeBreed platform and the breeding value estimation via letters of complaint to German authorities and legal claims. This is an illegitimate attempt to enforce the ACA's claim to be the sole representative of Carnica breeding in Austria. However, if the ACA stops these attempts and concentrates on professional cooperation again, ACA can remain on BeeBreed.

Will BeeBreed be transferred to another provider of breeding value estimates? BeeBreed was developed by the LIB from the outset, is its intellectual property and is closely linked to its research programme. It was supported by public funds and contributions from a wide range of associations with the intention of long-term implementation at the LIB. A transfer away from the LIB is neither desirable nor technically possible, as it is based on proprietary software that is not designed for use by third parties.

Has the LIB broken promises?

The subject of the commitments was that AT-2 would receive a separate breeding value estimate, and this was honoured in the 2023 and 2024 breeding value estimates. The realised solution went against the expert advice of the LIB and is based on the strategically motivated data usage restrictions of the associations. A more technically sound solution will be realisable in the future with breeding value estimates adapted to different breeding guidelines on the basis of respectful cooperation between the associations.

Does the exclusion of AT-2 breeders from the main Carnica population have a negative impact on breeding values for the DIB and ACA?

The exclusion of the 21 AT-2 breeders from the main Carnica population results in massive pedigree gaps, as they are strongly interlinked with the ACA and DIB pedigrees. This is illustrated by the following figures (as of January 2024):

106 mating station seasons of non-AT-2 mating stations were equipped with AT-2 material, of which 100 in ACA and 5 in DIB.
6893 queens from non-AT-2 breeders have a mother or mating from AT-2 breeders, including 5774 from ACA breeders and 1049 from DIB breeders.
3784 queens from AT-2 breeders have a mother or mating from breeders outside AT-2.
1196 queens from AT-2 breeders have a mother or mating in DIB.
3323 queens from AT-2 breeders have a mother or mating in the ACA
12573 queens were tested by AT-2 breeders, of which 3438 were tested by Stefan Mandl and 9135 by the 20 other breeders
942 queens from AT-2 breeders were tested by examiners outside AT-2, of which 841 were tested in the ACA and 51 in the DIB
560 queens from breeders outside AT-2 were tested by AT-2 examiners, 487 of which were from the ACA and 28 from breeders from the DIB

Gaps in pedigrees and not taking into account third-party testing are disastrous for the accuracy of breeding values.

Was the "Austrian professional beekeepers" population calculated separately from the main Carnica population in the 2023 / 2024 breeding value estimate?

Despite rumours to the contrary, the following example shows that there were separate calculations. Dr Mandl tested this queen of Fr Sekuja (Mali Drvenik). The breeding value level in the population of "Austrian professional beekeepers" is low, in particular the swarm inertia, and the securities are low at best. The mother, on the other hand, is highly valued in the main population. If there had been no separate CSE, the very high breeding values of the mother would have influenced the queen tested by Dr Mandl, and Dr Mandl's tests --- especially the swarm inertia --- would have influenced the mother. Certainties of 0.3-0.35 occur with parentless queens simply due to the large test stand. If the parents had been taken into account in the ZWS of the "Austrian professional beekeepers", then the certainty would be in the range 0.45-0.55, as for example with this queen, a full sibling of the queen above.

Why has the main Carnica population temporarily been abdandoned?

Until 2022, the basis of the main Carnica population was that the breeding work of all Carnica associations was recognised as being of equal value in principle, even though each association has its own breeding guidelines. The free usability of breeding data across association boundaries was not in doubt. The D.I.B. Breeding Advisor (F.-K. Tiesler) restricted the rights of use of the D.I.B. data on 20 January 2023. Even if this restriction was formally only related to one association (Bienenhof Mandl), the basis of a cross-association breeding value estimation has thus been destroyed. The only valid argument for not recognising international performance tests and pedigrees can be based on non-compliance with the D.I.B. breeding guidelines. However, this is also the case for all other international associations, especially as the breeding programme of the Bienenhof Mandl association follows very closely the traditional Carnica breeding based on the Apimondia protocol. The main Carnica population was reintroduced in December 2023 at the request of the associations, but without resolving the fundamental problem.

How does the integration into cross-association populations work?

With a contract with the LIB for BeeBreed support, initially only a breeding value estimate can be offered exclusively in this association. Integration into a cross-association breeding value estimation cannot be guaranteed. An indispensable prerequisite for participation is the declaration of free usability of the data.

How is the evaluation of disease susceptibility achieved?

We evaluate the occurrence of diseases in the breeding population using a methodology similar to the breeding value estimation of the other traits. The occurrence of a disease is considered as the combination of unfavourable conditions in the apiary (presence of the pathogen, generally difficult conditions for bees), genetic susceptibility, and special factors. To represent genetic susceptibility, the other factors must be factored out.

What exactly is a sibling group?

In traditional mating and insemination with drone colonies of a selected parentage, the question can still be answered clearly: a sibling group consists of colonies whose queens descend from a common mother (2a) and which were mated with a common sire colony (4a).

However, as alternative mating methods such as 1b insemination, one-drone insemination etc. are now also used, the situation becomes more complicated and a distinction must now be made between 2 types of siblings. What they all have in common is the common mother (2a), it is the paternal lineage that makes the difference. Siblings in the broader sense are now (as before) siblings who have a common 4a. In special cases, these can also be 1b-inseminated queens with different drone colonies. Siblings in the narrower sense are those that had an identical parentage situation, i.e., for example, that they were inseminated with drones from the same colony during 1b insemination.

Why the distinction? For the breeding value estimation, the siblings are always considered in the narrower sense, it reflects the actual relationship most accurately. In most BeeBreed functions (e.g. breeding values of sibling groups), siblings in the broader sense are taken into account. The sibling group in the broader sense is also shown on page 2 of the licence. There is a special feature in the breeding certificate: in the case of tolerance mating stations with several 4a, colonies are shown that have visited the same mating station in addition to the same mother - i.e. siblings in the narrower sense, but which do not correspond exactly to siblings in the broader sense because there is no common 4a.

Sister queens that were mated at different mating stations, mating station passes or mating station cohorts that use drone colonies of the same father colony (4a) also count as siblings in the broader sense but not in the narrower sense.

When will the new breeding values be published?

On 15 February of each year for most breeding populations, i.e. the main Carnica, Mellifera and Ligustica populations. If this day falls on the weekend, they are published on the previous weekday.

When do I have to enter my breeding data by?

If possible, by 20 December to give the association's administrator/chairman enough time to review and approve the data. We will start the first analysis steps on 10 January. The final data deadline is around 12 February, but this data is then not subject to the thorough review that can take place if the data is entered on time.

What can I do to improve my breeding values?

First of all, realistic breeding values are more important than high breeding values. breeding values. Because only breeding values that reflect the true genetic value of the bees bees are a useful tool for breeding.

So I would change the question to "What can I do to ensure that my breeding values more realistic and better?". For the quality of the bees is not improved if the breeding values are high and unrealistic - this will take its toll in the next generations.

Of course, the D.I.B. sets lower limits of the breeding value for its breeding value for its grading classes. For good reason, because of the development of the overall population, it is important that only above-average queens are bred. But: the minimum requirements of the D.I.B. are extremely modest. The grading class B requires only a performance index of 100, which by definition is half of all breeding queens of the last 5 years. Because of the selection, the current generation has above-average breeding values, the majority of the current generation is worthy of breeding! In addition that the breeding values within a sibling group benefit from good inherent performance. This means that in almost every sibling group contains a queen worthy of breeding.

Now to the measures that lead to realistic and better breeding values:

External testing. External testing is the key to realistic breeding values. How can the true value of a breeding line be assessed if it has never been compared with others? This means always testing queens from other breeders and giving your own queens to other breeders for testing. Occasionally also with distant breeders (other national associations, other countries) for external testing.
Do not plan test stands too small. The breeding values are determined by the performance differences on the test stand, and stands that are too small stands provide little information. See also our recommendation.
Continue to check the poor colonies and enter the data in BeeBreed enter the data. They are very important for the comparison!
Pay attention to the correctness of the data entered in BeeBreed, and check whether the chairman has authorised it.
Avoid excessive inbreeding when selecting the breeding centre.
Decide on good breeding values when selecting, but at the same time but do not neglect your own beekeeping expertise.
Look critically at very high breeding values, they are often the result of extreme special situations and are therefore not realistic.

Of course, it is an option to give up your own bee breeding line and build a new breeding line on "successful" other queens build up. But regionally adapted bees are, on the one hand, a cultural cultural asset, and can also be more robust in the local climate. in the local climate. It is also possible that they will stand out in the comparative performance test with top breeding value bees from other regions. other regions. This is why such external testing is so important, before considering abandoning a local breeding line. It can be useful to cross bees classified as high quality (e.g. for Varroa resistance), e.g. by visiting an apiary. a hive. This should also be done carefully and be accompanied by comparative performance testing.

Why was the breeding value of my queen last year still ..., but this year only ...? ?

First of all: The breeding values of the previous breeding value assessment, which are noted on the breeding licence or, if applicable, the breeder's price, have not become invalid as a result of a new assessment. They have been estimated on the basis of the information available at the last breeding value estimation. Breeding values change to a greater or lesser extent from estimate to estimate. Some of the reasons are obvious, but others are due to the complex interplay of information. This means that a breeding value cannot be fully explained. It is an estimate of an abstract quantity, the hereditary quality of a trait. An important reason why the breeding values of older queens decrease is due to normalisation. The 100 is related to the average of the last 5 years. And since most breeders select the best queens, new, improved queens are added every year are added every year and poorer ones are dropped from the average. Another reason why the evaluation changes in the next ZWS is newly added information. Particularly strong influence is exerted by tested offspring. The breeding values of the current crop are most affected by changes most affected by changes, as these breeding values are estimated with the greatest uncertainty. Only when own offspring have been tested the breeding values are estimated with a high degree of certainty. Here the breeding values breeding values no longer change significantly. The highest and lowest rated queens are usually affected by breeding value revaluations. lowest rated queens are affected. In the 2020 ZWS, however, there are also model changes that have led to major changes. We regularly check whether the published breeding values correspond to the later determined quality of the offspring and therefore make adjustments. This year, we had especially focussed on the breeding values catapulted upwards by an catapulted by an outstanding own performance, which later turned out to be unjustified. turned out to be unjustified.

Does the breeding value apply to the colony, the queen or the workers?

In principle, yes, strictly speaking it applies to the mated queen, i.e. the queen together with the sperm from the mating. But at the same time it applies to the average of all workers. Since all the workers in the colony contain genetic material from the queen and the sperm of the mated drones, there is no genetic difference between the mated queen and her colony.

The queen herself, i.e. without the sperm, is genetically different from the workers. This is why she also has her own, different breeding value. However, this breeding value is not shown in BeeBreed because it has practically no significance. This is because the sperm from the mating is retained for the entire life of the bee.

But what about a virgin queen? As long as she has not been mated and her colony is tested, her genes and breeding value are unknown. This is why the breeding value of a virgin queen corresponds exactly to the breeding value of her mother colony. Genetically, she is initially equal to the workers of this colony.

Why does the breeding value of an ancestor colony decrease from breeding value estimate to breeding value estimate?

The breeding values of past vintages become lower from breeding value estimate to breeding value estimate, and this is intentional.

The breeding value 100 is the average of the last 5 birth cohorts. Even if descendants are tested, the evaluation can change significantly.

The rating can change significantly, especially if it has not been verified by external testing.

Why do breeds with the same rating for gentleness have different breeding values for gentleness at the same level?

This is due to the genetic correlation, especially to the comb fit. The genes that influence the individual traits are not independent of each other. Taking genetic correlations into account is a great advantage because the fluctuations caused by random events are buffered and the breeding values become more reliable as a result.

How is the overall breeding value calculated if no Varroa traits have been tested, i.e. without a Varroa index

? The overall breeding value is still calculated with the weighting factors 15% for honey performance, gentleness, comb fit and swarming tendency and 40% for Varroa index. A Varroa index is determined even if no Varroa characteristics have been tested, it is just not shown (according to the D.I.B.) because it is very uncertain. It is virtually an average breeding value for colonies without measured varroa characteristics.

Why is the total breeding value not the average of the individual breeding values - usually higher?

In principle, you could also take the (weighted) average of all individual breeding values as the overall breeding value. But then it would no longer fulfil the extremely useful criterion:

Breeding value	better than ... of all queens
100	50%
105	69%
110	84%
115	93%
120	98%

correspond. Therefore, all total breeding values are normalised once again. Queens that are very good in all traits (example: 110 for all traits) are rarer in the overall population than the breeding value 110 for a single trait. In this example, normalisation raises the overall breeding value to 113 (Carnica breeding value estimate 2018). This makes perfect sense, because the combination of 5 good breeding values is more outstanding than a good breeding value for a single trait. The extent to which the good overall breeding values are increased varies from breeding population to breeding population and can also change over the years. It depends on how often good and bad breeding values coincide.

4a mating - 1b mating, what should I enter there?

In most normal cases - do not enter anything! When mating on a mating station, nothing needs to be entered - it is basically a 4a mating. There are very special cases (mostly artificial insemination) that can be recorded here in detail. These are

one-drone inseminations: here 1b mating, drones: 1 is entered - drone colonies should remain free
inseminations with drones directly from a registered and lead-tested colony: 1b mating and the number of drones estimated to be used for an insemination
for normal insemination with drone colonies 4a mating.
In Norway, for example, there is a special type of mating centre where breeding colonies are set up directly as drone colonies - 1b mating and the colonies set up are registered as several 4a colonies.

Number of drones, how should I know?

With natural mating, the number is uncontrollable, we assume an average number. Even with artificial inseminations, the number of drone colonies does not have to be recorded, we also assume an average number in this case. Only for special inseminations such as 1-drone insemination, heavily mixed semen from many drones, a number should be entered.

Number of drone colonies, how should I know?

The number of drone colonies is also recorded at a registered apiary and does not need to be entered again in the data record. However, if the apiary is not (yet) registered, this is an opportunity to disclose the number of drone colonies. The number of drone colonies should also be entered for artificial inseminations, as it can vary from insemination to insemination, even in a fixed insemination centre. If nothing is entered, we assume an average number.

How can I use your database to find the gentlest possible line or its breeder?

For example, proceed as follows (for Germany):

Select menu "Breeding values,"
Population: "Carnica main population"
Breeding value results for selected queens
Year
Click on "licensed"
Delete weighting factors for honey, comb fit, swarming tendency, varroa index
Sort by total breeding value
Click "Next"
Click on a queen code -> Pedigree browser
Contact the breeder / inspector

Explanation:

Population: Carnica main population: Of the breeds available in BeeBreed, only these are readily available in Germany. Mellifera would have to be imported from Switzerland or Norway, further breeding in Germany will Buckfast is currently not represented in BeeBreed.
Birth year 2022: this is the vintage of queens that have been tested 2023 and are licensed 2024
click on "licensed": these are the queens from which the breeders sell daughter queens
Delete weighting factors for honey, comb fit, swarming tendency, varroa index: the total breeding value then becomes equal to the breeding value gentleness

Note that the breeding value is only one of the criteria for a suitable queen. Regional adaptation, the other characteristics, inbreeding levels are others.

Most breeders only sell queens by pre-order before the current season, some breeders do not sell at all. In any case, it makes sense to get in touch with the breeder, who can also say something about the individual characteristics beyond the breeding values.