Queen selection

[boca]

I am thinking about an objective method to select the best mated queens to be used in production.

Let's assume that you have 100 mated queens at the end of June and by the beginning of September you want to keep only 20 of them to go in winter.

I stress that I am not talking about breeding but simply selecting the best 20 % from a given population.

As you say in Britain "the proof of the pudding is in the eating". For a queen it is heading a full family through winter and a production season. Obviously it is not possible to give this chance to all queens to show their capabilities.

What would be your criteria to select the best and how would you measure or estimate their performance?

 

[Hivemaker.]

Obviously it is not possible to give this chance to all queens to show their capabilities.

Why not?

 

[For want of a better word]

Why not?

 

[boca]

Because in that case we don't do any selection.

Imagine you want a new car that you plan to use for ten years. On the market there are 500 different types. You cannot buy all of them and try them for 10 years. It is necessary to make this decision before the real production.
If you want one car, the answer cannot be this: buy only one car randomly and choose that one. After ten years you will know if it was a wise decision.

 

[Skyhook]

Slightly dodging the question- but could you not use a number to populate small nucs, (eg 3-frame, not apidea), which would give you a chance to assess laying rate and temperament, then combine to 1/3 or 1/4 of the number as strong colonies.

Other than that, surely all you can do is bet on bloodline.

.

 

[adamf]

I am thinking about an objective method to select the best mated queens to be used in production.

What would be your criteria to select the best and how would you measure or estimate their performance?

Hi
Your question depends on what the goals of your breeding program were. Once you had those and subsequent selection criteria, measuring it per candidate is fairly simple--use what others have already used, or come up with your own tests.

In your examply population, say for Varroa tolerance, keeping 20 would be way too loose a selection threshold. I'd keep and test 5 from 100, maybe.

Adam Finkelstein

 

[oliver90owner]

Ask Finman - he is clearly your man. You might believe him. He say 'buy in new queens'.

 

[For want of a better word]

Because in that case we don't do any selection.

Imagine you want a new car that you plan to use for ten years. On the market there are 500 different types. You cannot buy all of them and try them for 10 years. It is necessary to make this decision before the real production.
If you want one car, the answer cannot be this: buy only one car randomly and choose that one. After ten years you will know if it was a wise decision.

don't buy a Lancia!... Volvo have a reputation of longevity... I would buy a SAAB.... or one of those tracked Haglund contraptions for getting to the out apiaries!

I can see your dilemma... you will have to make do with random selection with a 5% probability of a good result!

 

[boca]

Slightly dodging the question- but could you not use a number to populate small nucs, (eg 3-frame, not apidea), which would give you a chance to assess laying rate and temperament, then combine to 1/3 or 1/4 of the number as strong colonies.

OK. Let's start with these traits.

Laying rate
In small nucs even the poor queens lay as much as the good ones. The bottleneck is not the queen but the nurse bees.
A few queens could be so poor that not even in a small nuc lay enough, maybe two out of the 100. The real test is the spring buildup which is too late in the context of this topic.

The Laying pattern however can be assessed quite early.

Temperament
This trait we cannot assess directly on the queen but only on the colony composed by the worker daughters of the queen. We introduce the new queen in a nuc where the workers are not her daughters, not even the brood. So before we make any observation we have to wait till these bees die off and all age group is descendant of the queen we are assessing.
That is a long time and the summer is short. If the nuc is nasty in July and I eliminate the queen I make a big error because the guards and foragers are not related genetically to our queen.
Even after two months when all the workers are daughter of the queen, I don't think it is straightforward to assess the temperament in a nuc. In my opinion small nucs tend to be calm even if they are predisposed to be aggressive in a strong family with honey to be defended.


Setting up three frame nucs require one frame emerging brood with bees for each nuc. One of the practical problem is taking 100 frames of emerging brood from 20 hives.
Therefore I believe the first screening has to be done directly on the queens without giving them a nuc.
For example body weight could be used to eliminate a number of them. Do you know other traits that can be measured directly on the queen without harm?

 

[Teemore]

If you were to overwinter all the queens you might reduce the size of the sample of queens to select from. My understanding of other posts that touch upon this issue is that some beekeepers overwinter everything in order to get a degree of natural selection. The size of Nuc you have available to do this will dictate whether you can do this effectively: I experimented with overwintering a mating Nuc last year and will be doing so again this year - Kieler + riser + small supply of fondant.
Would you rely upon the records relating to the mothers of the Queens as a selection measure?

 

[Scutellator]

OK. Let's start with these traits.

Laying rate
In small nucs even the poor queens lay as much as the good ones. The bottleneck is not the queen but the nurse bees.
A few queens could be so poor that not even in a small nuc lay enough, maybe two out of the 100. The real test is the spring buildup which is too late in the context of this topic.

Disagree. In 4-5 week cycle you can get pretty good idea about the layng capability of the queen. Just make sure to provide more room for build up (but this doestn't work in the late season). You have to assess not the total brood area, but the correlation between the area covered by the bees and the amount of brood or the rate of building up (so, if the queen lays eggs outside the area covered by the bees, she is gonna be a vigorous layer for sure, if some mishap doesn't happen to her). As soon as the queen start to lay you can assess (to some extent) the pollen collection, comb building, or even the temperament.
The queens/larval pheromones make a big difference. BUT there is always room for surprises.

I remove all undersized, imperfect and those who delay in emerging. I found that the queens who hatch earlier usually happens to be the best ones.

Just you need to keep closer look over them with a comprehensive for your needs record keeping and after some time of practice, the initial selection could be done in just a feeling. You could search for a certain pattern, based on your experience.

 

[dilys]

I think a lot depends on the bees in your locality and the culture of the local beekeepers. If you live in an area where everyone has his own pet race of bee then many of the queens you produce will be hybrids or will lay hybrid eggs. When this is happening laying rate will mean nothing. You can't tell if it's due to hybrid vigour or genetics.

 

[psafloyd]

As you say in Britain "the proof of the pudding is in the eating".

NO, alas most people say the proof is in the pudding, but that is because most people are stupid (present company excepted).

It's lovely to see someone using the phrase correctly.