> I keep referring to Tom Seeley's wonderful book, The Wisdom of the Hive,

Agreed.
Don't wait for the paperback, there may never be one.

> Tom [Seeley] points out that he, and other researchers, have found that
> within a hive there are usually several castes.  These have the same mother,
> but different fathers.  What is fascinating is that bees within a caste seem to
> somehow recognize each other, and will sometimes act in a different manner
> than bees in another caste!

Yes, I read that.

> This is true with swarming.

But where did Seeley say that this applied to SWARMING?  I pulled
out my copy of WotH, and I could find no mention of swarming
and "relatedness".

> When bees swarm, most of those in the swarm will be of the same caste,
> and most of those left behind will be of different castes!

Even if Seeley said it somewhere, I still don't get the line above.
It makes no sense to me.

If "most of those who swarm will have the same drone "father"
(since that is how you use the term "caste"), then swarms could
NOT consist of roughly 1/2 the colony, as is commonly agreed,
but could only be a much smaller fraction of the colony.

Let's assume that:

1)  A queen mates with 10 drones
2)  A swarm consists of roughly 1/2 the colony

Reasonable starting points, so let's go:

a)  Each worker gets 1/2 its genes from the queen, and 1/2
     its genes from one of the 10 drones.

b)  The queen cannot segregate her sperm "by drone", so we
     can assume the "drone component" of each worker to be a
     random selection, or something close to random.

c)  So, there are up to 10 groups (daughters of each drone) in the hive.
     While all workers are daughters of the same queen, each has a
     1-in-10 chance of being a full sister, and a 9-in-10 chance of being
     a half sister to any other worker bee.

d)  Now, if swarms were a small fraction of a colony population,
     I might see a connection between drones as "fathers" and
     who goes and who stays.

But swarms are a significant fraction of a hive (half? slightly more?)
and must be daughters of more than one drone.  The group that
swarms cannot possibly be even mostly daughters of the same
drone.

So how could "which drone" makes any difference?
At minimum, the daughters of more than one drone
swarm with the queen, AND the daughters of more
than one drone "stay".

This is nothing more than simple math.

If anyone wants to get into the relatedness of the drones that just
happened to mate with the queen, this would be getting VERY
silly and start to approach odds found only in state lotteries.

Real mating yards can be assumed to have multiple
drone-producing hives.  Anything else would not be
a decent genetic pool.  It would be a genetic puddle.

> I recall that Tom also said "more research is needed"...

Agreed.
I hope it is done by someone who can do some fractions.   :)

It is my personal pet theory (unverified by anyone as yet) that
multiple queen hives are more common that anyone may
suspect, which brings a whole new factor into the "swarming
puzzle".  I wish that everyone would inspect recently-swarmed
hive closely for eggs, and report any eggs found the day of (or the
day after) a swarm.  Eggs directly after a swarm would prove that
the swarming hive was a dual-queen hive before it swarmed.
(Now does my question about queens [plural] make sense?)

        jim