An Apiary Guide to Swarm Control

2nd Edition
Wally Shaw

Wally Shaw 1
 Published and funded by the Welsh Beekeepers’ Association
2020
www.wbka.com

Printed by : Stephens & George,

Dowlais, Merthyr Tydfil. CF48 3TD

2 An Apiary Guide to Swarm Control 2nd Edition

 Table of contents
Introduction

PART 1 – PRE-EMPTIVE SWARM CONTROL 5

1.1 Comb Management 7

1.2 Box Management 10

1.3 Brood Relocation 15

1.4 Splitting Colonies 18

PART 2 – RE-ACTIVE SWARM CONTROL 21

2.1 Colony has not yet Swarmed 23

2.2 Colony has already issued the prime 31

swarm but has not cast swarm

2.3 Colony has emerged queen cells and 32

may (or may not) have is sued as cast swarm

2.4 Colony appears to be queen-less (it has 33

no brood) and beekeeper has no idea what
happened and when

2.5 Late Season Swarming 34

Post-script 36

APPENDIX 1. Further Discussion of the Pagden Method 38

APPENDIX 2. Fault Finding 39

APPENDIX 3. Recommended Logistics for the Second 41

manipulation

Wally Shaw 3
Introduction

The beekeeper needs to understand that swarming is simply re-

production. All the complex, integrated behaviour that occurs within a
honey bee colony has evolved simply to improve its chance of success-
ful reproduction. Most beekeeper’s primary interest is just one aspect of
this behaviour, namely the collection of nectar and the storage of hon-
ey. In order to produce the maximum amount of harvestable honey, the
beekeeper seeks to create large colonies but also to prevent them from
achieving their natural destiny through swarming – so there is an inher-
ent conflict involved. The only way of addressing this is through swarm
control.

In their book entitled `Bait Hives for Honey Bees`, Seeley and
Morse state that, `Mature colonies have a natural urge to swarm each
year unless weakened by disease or mismanagement`. So perhaps
we should not be surprised or regard it as dysfunctional when colonies
swarm. Most beekeeping books understate swarming but its control is vi-
tal if good honey crops are to be consistently obtained. There is no doubt
that swarm control is simultaneously the most important and most dif-
ficult aspect of colony management.

The Booklet (or Field Guide)

As the name suggests, this booklet is primarily about practical

swarm control; what to do when you are out there at the hive-side faced
with making a decision about management to try and control the swarm-
ing impulse. It contains a series of diagrams that can be used as a guide
to management and can even be taken out into the apiary for reference if
required. Because no two hives are ever the same, the diagrams need to be
applied flexibly as they show generalised examples but it is the underlying
principle that matters – how the management interacts with the natural
behaviour of the colony. The accompanying text has been kept to a mini-
mum with (hopefully) just enough to explain the manipulations involved
and their function.

4 An Apiary Guide to Swarm Control 2nd Edition

 For the sake of simplicity, all the diagrams assume the use of a
Modified National hive with 12 self-spacing (Hoffman) frames. However,
the same management techniques can be applied to all types of mov-
able frame hive whatever their size, shape, number and spacing of frames.
The frames shown in the diagrams are colour coded according to their
contents; frames containing brood are coloured red, honey and pollen
yellow, drawn frames with no contents black, dummy boards blue and
un-drawn frames (containing a sheet of foundation) are shown as a thin
black line. In the real world the frames within a hive usually have a mix-
ture of contents so the colour coding refers to the primary contents (the
dominant characteristic) of the frame.

Most diagrams show a standard hive configuration of a brood and

a half, ie. one deep and one shallow box beneath the queen excluder (the
brood area), which seems to be about the right amount of space for near-
native (locally adapted) bees in Wales. Other hive configurations (eg. sin-
gle deep, double deep or 12x 14) require appropriate modification of the
methods described in this booklet and some options may be limited or
even impossible.

Types of Swarm Control

From the point of view of practical management, swarm control

can be divided into two distinct parts with a clear (biological) threshold
between them which is when the colony starts queen cells.

1) Pre-emptive swarm control – the type of management that

can be used before queen cells are present in the hive (to try
to prevent their initiation).

2) Re-active swarm control – the type of management that

can be used when queen cells are produced (to prevent the
issue of swarms).

Some beekeepers seem to have misunderstood the clear distinc-

tion between these two types of management practices and applied pre-

Wally Shaw 5
emptive management after queen cells have been started in the hope that
it would make them go away. It is a vain hope I am afraid and there is very
little chance this will cause the colony to change direction.

PART 1 – PRE-EMPTIVE SWARM CONTROL

The management activities involved in pre-emptive swarm con-

trol are multi-purpose and not just about preventing queen cells being
started. They double-up as good beekeeping practice which aims at:-

• Promoting a large colony capable of collecting a large crop

of honey.

• Systematic renewal of brood combs (particularly important

for disease prevention).

• Queen replacement and making increase (when required).

Unfortunately, producing a large colony and maintaining it over

the period during which a nectar flow may occur (3 months or more) can
create the very problem the beekeeper is trying to avoid. This is because
the larger the colony and the longer it is in that condition the more likely
it is to swarm. Good pre-emptive management always delays swarming
but may not prevent it happening eventually. Late season swarming is
particularly frustrating because it can compromise the potential for hon-
ey production during the main flow in July (or later). Two ways of dealing
with this particular problem can be found at the end of the booklet (Sec-
tion 2.5).

Triggers for Swarming

The triggers for swarming – the means by which the colony recog-
nises when it is a good time to swarm and is most likely to be successful
– are multi-factorial and a mixture of internal and external conditions.

6 An Apiary Guide to Swarm Control 2nd Edition

Internal (within the hive)

• Size of the colony, space for the queen to lay, brood nest
congestion, brood nest maturity and (possibly) the age of the
queen.

• Space for nectar processing and honey storage.

• Production and/or distribution of queen substance (thought

to be the main mechanism).

External

• Time in the season - the swarming urge is at a peak in May

and June (when it is most likely to be successful) and declines
thereafter.

• Weather – an underrated factor as interludes of poor weath-

er (with little flying time) often precipitate swarming.

The beekeeper can to some extent control internal conditions

through management of the hive but can do nothing about the external
factors. It follows that pre-emptive swarm control is mostly about man-
agement of the brood area.

The main management techniques by which the beekeeper can control

hive internal conditions are:-

1. Comb management
2. Box management
3. Brood relocation
4. Spitting colonies

1.1 Comb Management

The aim here is to ensure that as many as possible of the frames

Wally Shaw 7
below the queen excluder (the brood frames) are actually used for the
production of brood. During the main season, stores of honey and pol-
len in the brood area should be kept to a minimum. Quite early on in
the season, when the colony has not yet attained its full potential size, it
is the bees’ instinct to create a ceiling of capped honey. These (close-to-
hand) stores are a form of insurance against adverse conditions which the
bees are reluctant to uncap and make cells available for the queen to lay.
When this ceiling is in place the only way the brood nest can be extended
is in a downward direction, so the aim is to have brood in contact with
the queen excluder over as much of its area as possible. The bees will still
create a honey ceiling but it will be in the first super – which is where the
beekeeper wants it.

Comb management may involve moving existing frames within

the brood area in order to provide space for the queen to lay but the main
activity is removing old or defective frames and getting new frames of
foundation drawn. It is good beekeeping practice to replace brood frames
on roughly a 3 year cycle, so that means an average of 3-4 frames/year. The
use of foundation is thought to be an additional disincentive to swarming
by simulating brood nest immaturity and diverting bees to the task of wax
making.

However, in order to be successful (get frames drawn quickly) and

not damage existing brood by chilling, the introduction of foundation
must be done at the right time and in the right position in the hive. Early
in the season, in order to maintain brood nest integrity, foundation must
be introduced on the edge of the brood nest - so that it becomes the next
frame to be drawn if the brood nest is to expand. Later, when the colony
is crammed with bees, foundation can be interleaved with brood frames.
Never put foundation next to the hive wall because in this position it will
only be drawn as a last resort – and then usually badly.

Figure 1 illustrates where foundation should be placed early in

the season (left) and later in the season (right) in both deep and shal-
low brood boxes. Using a two-box system (ie. brood and half or double
brood) the placement of foundation in the upper box is much less critical
8 An Apiary Guide to Swarm Control 2nd Edition
because of the warmth coming up from below. Single box users (standard
deep or extra deep 12x14) should follow the procedure as shown for the
lower box.

When introducing foundation to a hive the beekeeper should un-

derstand that the colony MUST have an immediate need for more comb,
either to extend the brood nest or for storage, because a colony does not
do ‘speculative’ comb building’. There MUST also be a nectar flow (or the
beekeeper must provide one by feeding) because bees do not use stored
honey for wax-making. For further information see also WBKA booklet
on ‘Comb Management’.

An alternative (less radical) approach to increasing the number

of frames available for the queen to lay in the half brood is to move in-
dividual frames up into the first super When there is a good nectar flow
early in the season the colony will often use any available cells in the half
brood to store nectar which, particularly when it has been capped, be-
comes a permanent fixture and will prevent their use for brood for the

Wally Shaw 9
rest of the season. This can be rectified by moving such frames up into
the first super and replacing them with frames of foundation inter-leaved
with frames that contain brood (see Figure 2).
If the honey super has only recently been added and is not yet in full use,
ie. it contains few bees, nectar bearing frames from the half-brood can be
placed in the middle of the box to ‘seed’ it. This will encourage the bees to
move up and make greater use of it.
Frames that contain the remains of winter stores, which may be
substantially sugar syrup, should be removed altogether and used to feed
other colonies in need. This is to prevent contamination of honey crop with
syrup. Having been walked on by the bees for 6-7 months, such frames can
be easily identified by having dark-coloured (dirty) cappings.

1.2 Box Management

This is only applicable to beekeepers who use a two-box system;
either brood and a half or double brood. As with comb management, the
aim is to have as much as possible of the brood nest in contact with the
queen excluder thus avoiding a honey ceiling.

10 An Apiary Guide to Swarm Control 2nd Edition

Hive on Brood and a Half

The need for box management depends crucially on the position

in the hive that the brood nest develops at the beginning of the season. If
the nest is high then nothing needs to be done because the queen is free
to lay down and use as much of the available space as she needs (Fig-
ure 3a). A nest developing in the middle of the hive will, as the season
progresses, result in a honey ceiling in the upper box and this may re-
strict the ultimate size of the brood nest. In this situation the boxes can be
swapped – placing the shallow brood box beneath the deep (Figure 3b).
When the brood nest starts low then the upper box will quickly become
a honey store (a super) and the brood nest size will be restricted. In this
case the best option is to place the existing shallow brood above the queen
excluder (make sure the queen is not in it) and introduce a new shallow
brood box containing drawn frames beneath the deep brood (Figure 3c).
If drawn comb is not available then placing a box of foundation under the
existing brood nest is the last place it will get drawn. In fact, it is unlikely

Wally Shaw 11
12 An Apiary Guide to Swarm Control 2nd Edition
to be drawn before the colony has taken the decision to start queen cells.
The only option is to put the box of foundation on top of the deep brood
box and below the queen excluder where it will readily be drawn. Then,
before it is has accumulated a significant amount of sealed honey, to
move it to bottom position. In a good nectar flow this could happen quite
quickly and the beekeeper needs to catch the moment.

If at the end of the season, there is a shallow box at the bottom of the
hive it is advisable to move it up on the top of the deep brood box before
feeding. Due to contraction of the brood nest, at this time of the year it
will usually contain little or no brood and probably no stores. On top of
the deep brood it will be preferentially used to store feed in a position that
will be readily accessible to the winter cluster.

Contrary to what is said in most beekeeping books, the best place

to create space for the queen to lay is below the existing brood nest. A
box of drawn comb placed on top of the brood nest will almost inevita-
bly be treated like a super and substantially filled with honey. Although
in a good nectar flow the ‘receiver’ bees often dump incoming nectar in
empty cells below the brood nest this is only a temporary measure and it
will quickly be transferred further up the hive. Capped honey below the
brood nest will also be removed when time allows and this is the basis of
the box re-location shown in example Figure 3b.

Caution – If the hive floor has a traditional depth of 21-22mm

(solid or open-mesh) it will usually be necessary to remove brace-comb
from the bottom bars of the box immediately above the floor - otherwise
it will not fit in its new position and numerous bees will be squashed.
Frames must be shaken free of bees before attempting this process. Using
floors with the correct depth for open-mesh floors (9mm for bottom bee-
space hives and 15mm for top bee-space hives) the frames should have
little or no brace-comb and will not need any trimming.

Other Hive Configurations

In the case of a hive on double deep configuration, controlling

Wally Shaw 13
brood nest position is simply a matter of moving as many frames of
brood into the top box as possible (taking care to create a sensible nest
shape that the nurse bees can cover efficiently) and moving frames of
stores or empty comb down. Stores beneath the brood nest will quickly
be removed and space will become available for the queen to lay. Because
double brood provides more than enough space for most colonies the
situation is less critical. However, the colony is likely to end the season
with a substantial amount of honey beneath the queen excluder – which
is good or bad depending on your point of view (ie. how much you want
to feed).

Using a single deep brood box hive configuration (which gives

less space than the potential of most queens), the position of the brood
nest is less likely to present problems. Here it is just a matter of ensuring
that, as far as possible, all the frames pull their weight (are available for
the queen to lay on). To achieve this aim the frames should contain a
minimum amount of honey and pollen stores. Well managed single deep

14 An Apiary Guide to Swarm Control 2nd Edition

box hives tend to have more pollen stored in the first super simply be-
cause there is nowhere else to put it and, by default, it becomes a brood-
less extension of the brood nest.

Extra deep box hives (eg. 12x14) can have a problem with brood
nest position and this is not quite so easy to remedy. Figure 4 shows how
this can be done by introducing frames to the middle of the box – but
there must a good head of bees when this is attempted. Depending on
what is available, these can either be frames of foundation, empty drawn
frames or existing lateral frames (from the outside of the box) with any
honey stores scored using an uncapping fork. The colony will want to re-
store brood nest integrity and these frames will quickly become part of a
larger, spherical brood nest.

Supering

Also part of box management is adequate supering of a colony.

It is essential to provide plenty of space for the processing of nectar and
the storage of honey. It should be remembered that fresh nectar has 2-4
times the volume of the honey that it becomes and extra space should be
provided for the bees to conduct the drying process. Supering needs to be
kept one jump ahead of the need for storage but without creating so much
volume that heat loss becomes a problem. When the top super is full of
bees (wall-to-wall), regardless of the fact that it may be only part-full of
honey, is the time to add the next super. However, no amount of supering
will act as a substitute for poor brood nest management. Weather is the
problem that the beekeeper can do nothing about and during prolonged
adverse conditions the bees will inevitably move down from the supers
and crowd the brood area and this will often trigger swarming.

1.3 Brood Relocation

This is one of the oldest tricks in the book and usually goes under
the name of the Demaree Method, dating back to 1892 – a method that is
probably under-used in modern beekeeping. There are many variations
of the method but the basic principle is the removal of frames of brood

Wally Shaw 15
from the bottom of the hive and relocating them in a new box at the top
of the hive – above the supers. The removed frames are replaced by empty
drawn frames (if not available foundation can be used) thus giving extra
space for the queen to lay. Brood at the top of the hive attracts nurse bees
to move up to cover it and this serves to reduce congestion at the bottom
of the hive. The combination of new laying space for the queen and a re-
duction in congestion in the brood area inhibits the impulse to swarm.

The method was originally designed for hives on double (or triple) deep
brood boxes and when brood has emerged from the frames put to the top
the plan is to return them to the bottom of the hive in exchange for ones
containing more recent brood (a frame circulation system).

Figure 5a shows a classical Demaree with two deep brood boxes.

Figure 5b shows how the method can be applied to a hive on a single deep
brood box. The manipulations shown in 5b can easily be adapted to brood
and a half configuration – the shallow brood stays at the bottom of the
hive, either above or below the deep brood box depending on the position

16 An Apiary Guide to Swarm Control 2nd Edition

of the brood nest (see Section 1.2 above).

Running a hive on a single deep box, it is doubtful if the beekeeper

would ever want to re-locate the entire brood nest to the top of the hive as

this would seriously unbalance the colony. Typically only 4-8 frames are
moved at any one time, resulting an incomplete Demaree box as shown
in Figure 5b. An incomplete set of frames should be flanked by a dummy
board on either side (colour coded blue). Further frames of brood can be
moved up at a later date if required so it may eventually become a com-
plete box. A part-filled Demaree box is no problem but the beekeeper
needs to be aware that in good nectar flow wild comb may be built in any
unoccupied space at the sides of the box.
The Demaree Method is quite an effective method of pre-emptive
swarm control but it does have some downsides:-

The first is that the bees covering the brood at the top of the hive
may be far enough removed from the queen that they regard themselves

Wally Shaw 17
as queen-less and start emergency queen cells. The greater the spatial sep-
aration (the number of supers) the more likely this is to happen. After 5-7
days it is necessary to carefully examine the top box for queen cells and
destroy them.

The second is that frame re-cycling is more difficult than appears

at first sight because as soon as the brood has emerged the vacated cells
are quickly filled with nectar and later with capped honey. This means
that the top box normally has to be left in place for the rest of the season
and removed as part of the honey harvest. However, providing the frames
are fairly new this should not adversely affect honey quality.

1.4 Splitting Colonies

The previous methods of pre-emptive swarm control have kept

the colony in one piece. Splitting is different and (potentially) creates
a second colony. However, splitting is the most powerful and reliable
method of pre-emptive control and has a long history of use in beekeep-
ing. Splitting colonies is also a method of making increase which is cov-
ered in more detail in the WBKA booklet, ‘Simple Methods of Making
Increase’.

It is often said that splitting a colony is the enemy of a good

honey crop. However, splitting is always better than losing a swarm - un-
less you can guarantee catching it! The effect on the honey crop depends
crucially on the timing of the split. If it is done at the right time, eg. di-
rectly after the spring flow, it can result in an enhanced yield. Under the
right circumstances the two resultant colonies can produce more than
the original (one) colony - even if it did not eventually swarm. A con-
trolled split is always better than an artificial swarm because it enables
the beekeeper to create a better age-class distribution of worker bees in
both parts of the split.

When splitting a colony to provide pre-emptive swarm control

the following considerations should be taken into account:-

18 An Apiary Guide to Swarm Control 2nd Edition

Wally Shaw 19
 • The split should be sufficiently radical to provide swarm
control for the rest of the season.

• Both sides of the split should be viable, ie. adequate bees,

brood and stores.

• The timing should be right for the colony, ie. its state of
development.

• The timing should be right for potential nectar flows, ie. al-
lowing time for the colonies to re-build.

No firm guidance (prescription) is possible and the details of

splitting depend on the judgement of the beekeeper. Figures 6a (using a
new hive stand) and Figure 6b (using a split board - see Figure 12) are
examples which illustrate the principle. In both Figures the blue box and
the frames within are new.

Recombining Splits

If a split has been made properly (see guidelines above) there is

about a 90%+ probability that the queen-less part of the split will raise
a new laying queen. The beekeeper now has two colonies instead of one
and this may not be the desired outcome. Simple logic dictates that you
cannot go on indefinitely doubling the number of your colonies so the
solution is recombination (uniting colonies) but how and when? Early in
the season recombination is not usually a sensible option as it will only
create a large colony that is likely to undo all your good work and swarm.
Later in the season, when the swarming impulse is on the decline, it is
possible to recombine to produce a ‘super-colony’ for the main honey
flow or for taking to the heather.

There are three main options for the new colony (with a new
queen) which will currently be living either on a new stand or a split
board.

20 An Apiary Guide to Swarm Control 2nd Edition

 • Giving it permanent independence, ie creating a new Long-
term colony.

• Recombining later in the season to produce a ‘super-colony’

(with a choice of queens to be made by the beekeeper).

• Giving it temporary independence, ie. supering it, keeping

until the end of the season and then recombining (also with a
choice of queens).

The word ‘recombination’ implies that the new colony has to be

united with the original colony. However, much greater flexibility is pos-
sible and it is often the preferred option to unite it with a colony that has
an old or unsatisfactory queen as a simple method of re-queening. Arti-
ficial swarming also creates two colonies from one and there is further
discussion about how to deal with this situation in Part 2 (Aftermath of
Artificial Swarming).

PART 2 – RE-ACTIVE SWARM CONTROL

Re-active swarm control starts where pre-emptive swarm control

fails - when queen cells with contents are found in the hive. When this
happens there is no known (reliable) method by which the swarming pro-
cess can be turned off and the beekeeper needs to accept the fact that
unless something is done to prevent it, that colony will inevitably swarm.
Destroying queen cells only delays swarming and may make the situation
worse, ie. the colony swarms with barely-started second generation queen
cells and takes a higher proportion of the bees with it. Worse still is de-
stroying queen cells not realising that the colony has already swarmed. If
swarming occurred more then 4-5 days earlier there will be no eggs and
young larvae present from which emergency queen cells can be made and
the colony will be rendered queen-less.

Determining the Stage in the Swarming Process

On finding queen cells in a colony the first thing to do is to deter-

Wally Shaw 21
mine what stage in the swarming process the colony has reached. Howev-
er, before jumping to conclusions, the beekeeper needs to be sure that the
queen cells are actually swarm cells and not supersedure or emergency
queen cells. This is usually obvious by their position and number but, if in
doubt, consult WBKA booklet, ‘There are queen cells in my hive – what
should I do?’

There are four main stages in the swarming process, each with a
different solution:-

1. Colony has queen cells but has not yet swarmed.

2. Colony has issued the prime swarm but has not yet cast
swarm.

3. Colony has emerged queen cells and may (or may not)
have issued a cast swarm.

4. There is evidence that the colony has swarmed but it cur-

rently appears to be queen-less (ie. it has no brood of any
age), and beekeeper has no idea what happened and when.

How to Investigate

Full details of how to do this are given in the WBKA booklet which
contains a diagnostic tree with 12 steps, each consisting of the method of
Investigation and the recommended Remedial Action.

The beekeeper should do ABSOLUTELY NOTHING until a

careful investigation has been completed. Fortunately the key informa-
tion persists in the hive (on the frames) until at least 3 weeks after the
prime swarm has departed – the last time the queen can have laid some
eggs.

22 An Apiary Guide to Swarm Control 2nd Edition

2.1 Colony has not yet Swarmed

How do you know? – the evidence in order of reliability:-

• Queen seen

• Recently laid eggs (standing on end).

• Eggs leaning over at about 45° were laid 24-48 hour previ-
ously and those lying flat 48 hours plus and about to hatch.

• Be aware that a late afternoon inspection may reveal stand-

ing –up eggs but a swarm could have departed a few hours
earlier

• No missing bees (colony the size you expect)*

• Maturity of queen cells and recent weather – has it been

conducive to swarming?

*NB – No missing bees is to be expected if the queen has had her wings
clipped - the colony can have swarmed and there will be little or no loss of
bees. This can also occur naturally if the queen is for some reason unable
to fly – so beware of this not uncommon possibility.

The beekeeper really needs to see the queen or recently laid eggs
early in the day to be absolutely sure. If the colony is deemed not to have
swarmed the remedy is to make an artificial swarm.

Artificial Swarming

Having established that the colony has not yet swarmed the (only)
solution is artificial swarming. Most beekeeping books describe what is
called the Pagden Method, first published in 1868. The basic principle of
this method is that the parent colony, comprising the brood and queen
cells, is separated from the artificial swarm, comprising the queen and

Wally Shaw 23
the flying bees. The artificial swarm remains in the same location and the
parent colony is moved to a new one (nearby). Our experience over about
15 years of trying to make this method work showed that it had an unac-
ceptably high failure rate – probably in excess of 50%. By ‘failure’ what
is meant is that the artificial swarm does not reliably lose the impulse to
swarm and can resume its original plan (to swarm) at more or less any
time during the next 3 weeks. Further discussion of the Pagden Method
can be found in Appendix 1

The method described in this booklet has a virtually 100% success

rate and is called Snelgrove II (modified). The fundamental difference be-
tween this and the Pagden method is that the queen does not go to the
artificial swarm but remains in the parent colony along with the brood
and queen cells. Many beekeepers, brought-up on the Pagden Method,
find this counter-intuitive and do not trust having the queen with brood
and queen cells. The simple solution is to try it and then you will know.

As the name suggests, the basic method originates with L.E.

Snelgrove, so his name has been retained out of respect for a very distin-
guished beekeeper. The modified version is subtlty different by providing

24 An Apiary Guide to Swarm Control 2nd Edition

the artificial swarm with the means of making emergency queen cells.
This is thought to be the secret of its success by switching off (or effec-
tively suppressing) the swarming impulse. Despite the name it does not
require the use of a Snelgrove board (or a split board see Figure 12). It
can also be done with the parent colony moved to a separate stand but use
of a split board does have many advantages (see discussion below).

Artificial swarming by the Snelgrove II (mod) Method is a two-stage op-

eration which is illustrated in Figures 7a and b. using a new hive stand
(not a split board in sight!). For clarity the brood area of the parent colo-
ny is in blue boxes (with a blue background) and the artificial swarm in
a green box (with a green background).

In the initial manipulation (Figure 7a) all the brood, including

queen and queen cells, are moved on their existing floor to a new hive
stand within the apiary - reasonably close is convenient but more than
3 feet apart is essential. This is now referred to as the Parent Colony.
A new floor is placed on the vacated hive stand and a new box with 10
frames of preferably drawn comb (but a mixture of drawn and founda-
tion will suffice) is placed on it. This will become the Artificial Swarm.
Two bee-free frames of brood are now removed from the parent colony
and transferred to the middle of the artificial swarm. These two frames
must contain eggs and young larvae from which emergency queen cells
can be made but must not have any queen cells (or the queen) on them.
The missing frames in the parent colony should be replaced using drawn
frames if possible.

Both hives are now re-built and most (or all) of the supers are
normally given to the artificial swarm where most of the current foraging
force will reside. However, remembering that the parent colony is initially
going to lose all its flying bees, the quantity of available stores needs to
be taken into consideration. If these are thought to be insufficient to last
until a new foraging force has developed or there is unlikely to be a sig-
nificant nectar flow in the near future, the parent colony should be given a
queen excluder and one (or possibly more) of the supers. In some seasons
it may even have to be fed.

Wally Shaw 25
 The first thing that happens is that the parent colony on the adja-
cent stand will lose its flying bees back to the artificial swarm. Amongst
these will be the bees that are running the swarming process. For this
reason, the parent colony entirely loses the impulse to swarm and in due
course the queen cells will be torn down and the queen will resume laying.
In the queen-less artificial swarm the bees will start emergency queen cells –
and this appears to be vital to the loss of the swarming impulse.

The second manipulation should take place 9-10 days later (12
days is the absolute safe limit). Timing is important because all the queen
cells in the parent colony should be torn down by this time. Usually this
happens quite quickly but it does depend on the maturity of the queen

cells, ie. it is not done until the virgin queens they contain are 3-4 days
from emergence at which time the colony becomes fully aware of their
presence. However, if the queen cells were only just started at the time of
the split it may take a day or two longer. More importantly, the second
manipulation must be done before any virgin queens can emerge into the
artificial swarm. The calculation goes like this - if an emergency queen
cell is based on a 1 day old larva (day 4 from when the egg was laid) the
26 An Apiary Guide to Swarm Control 2nd Edition
first queen could emerge sometime on day 12. From this it can be seen
that day 12 is pushing the limit and it is better not to take the risk.

The second manipulation starts with checking that the queen cells
in the parent colony have been torn down and the queen has resumed
laying. If this has not happened it means that there is no queen in the
parent colony. The most likely cause for this is that the beekeeper failed
to notice that the colony had already swarmed (sometimes it is difficult
to be sure). This is no cause for alarm because the parent colony has
queen cells from which it can re-queen itself (or it may by this time
contain a virgin queen) and the artificial swarm has emergency queen cells
that are approaching maturity.
The parent colony will re-queen without swarming (no manage-
ment required) but, the artificial swarm may need steps to be taken to
prevent cast swarming (see Section 2.2 below). There are other less likely
explanations for the queen cells in the parent colony not having been torn
down and the beekeeper should refer to the Fault Finder in Appendix
2 to understand the situation and what to do next. The next step is to
examine the artificial swarm where the two frames with eggs and young
brood on them should now have emergency queen cells on them. If this
is not the case the queen (or a queen) must be present somewhere in the
artificial swarm. Again the beekeeper should refer to Appendix 2.
The frames with queen cells should now be removed and then the
-
queen can be transferred (repatriated) to the artificial swarm. Details of
the best way to do this (smoothly and safely) can also be found in Appendix 3.
The question now is what to do with the two frames with emergen-
cy queen cells on them? If at the time the colony was artificially swarmed
there were plenty of younger bees in the supers (and this is usually the
case with a colony that has set up to swarm) then the occupants of the
emergency queen cells will have been well-fed and contain fully devel-
oped virgin queens. If the beekeeper is confident this is the case, the two
frames with emergency queen cells can be placed in the parent colony to
provide it with a new queen. If there is any doubt about the quality of the
cells then it is better to destroy them and let the parent colony start from
scratch and make its own. It will have available suitable brood produced

Wally Shaw 27
by the queen who had resumed laying. Also there will be plenty of nurse
bees from the still emerging brood to feed them well. The latter is the
safer option but the former saves 9-10 days on re-queening of the parent
colony. Which option to take requires a judgement call to be made by the
beekeeper.

Starting with an identical hive that has set up to swarm, Figures

8a and b show the equivalent process using a spilt board (colour coding
of boxes the same as Figures 7a and b). The overall process is just the same
but repatriation of the queen in the second manipulation requires slightly
different logistics which are described in Appendix 3.

WARNING – If when using a split board, the artificial swarm and

the parent colony are (vertically) too close together there is a small chance
that the (queen-less) bees in the artificial swarm will find that their ‘be-
loved’ queen has just moved ‘up-stairs’. If they do find her – and they can
initially be seen following a scent trail up the side of the hive to the en-
trance on the split board – the parent colony will inevitably issue a swarm

28 An Apiary Guide to Swarm Control 2nd Edition

in the next few days. If the bees in the artificial swarm have found the
queen the only remedy is to immediately move the parent colony to a new
position in the apiary so that contact is broken.

The first reaction of a colony on losing its queen is to frantically

look for her. If she cannot be found they will then start the emergency
queen cells. Obviously if the vertical separation is small, eg. one super’s
worth, they are more likely to find her. Two supers is safer and three is
secure. Good pre-emptive swarm control usually means that the colony
will be well-developed at the time it sets up to swarm and will have three
(or more) supers on it but, if this is not the case, then extra supers should
be added to enable greater separation to be given. The alternative is to
use the two stand version (Figures 7a and b), where the artificial swarm
(more than 3 feet away) will be unable to find where the queen has gone.

Advantages of Snelgrove II (modified)

Wally Shaw 29
 Apart from its virtually 100% success rate, one of the main advan-
tages of this method of artificial swarming is that for the first manipula-
tion the queen does not have to be found. Even for an experienced bee-
keeper this is often not easy when the colony is crammed with bees and
the queen is being constantly harassed by the workers to slim her down.
She could be anywhere and probably running as fast as she can. When in
the second manipulation she does have to be found there will be fewer
bees and she will have settled back to the day job (laying eggs) – so finding
her should be much easier.

Got the timing wrong?

If, for some reason, the beekeeper gets the time of the second ma-
nipulation wrong, and the emergency queen cells have already released
virgins into the artificial swarm, it is not the end of the world. However, it
is unlikely that all the virgin queens can be found and the old queen safely
repatriated and it is better to allow the artificial swarm to re-queen natu-
rally and leave the old queen where she is in the parent colony. The latter
can now be given a queen excluder and supers and run as the main-honey
producing unit.

However, there is conflicting testimony from users as to what hap-

pens to the artificial swarm which has multiple queen cells. Does it be-
have in a normal way to a colony that has made emergency queen cells
and select a new queen without swarming? Or finding itself with multiple
queens, does it proceed to issue a swarm (or swarms)? Or are both out-
comes possible? At the present time this is not known and a safe way of
dealing with this uncertainty is given in Section 2.3 (below).

Aftermath of Artificial Swarming

If successful (ie. the new queen has mated and started laying),
when the process of artificial swarming is complete there will be two colo-
nies instead of one and this may or may not be what the beekeeper wants.

• If an increase in the number of colonies is not required

30 An Apiary Guide to Swarm Control 2nd Edition
 then re-uniting is the answer. This is easier when using a split
board than it is with the new colony on a separate hive stand.
Re-uniting should not (normally) be attempted until the par-
ent colony has developed a new laying queen and she has
proven herself to be a good’un. Using a split board, a two-tier
colony (one on top of the other each with its own supers) can
be retained until the end of the season when a decision as to
its future can be made. Note - this is not a two-queen system
because the two colonies are independent and have different
entrances and separate supers. When uniting two colonies it
is preferable for the beekeeper to choose the queen because
letting the colonies do it for themselves can result in serious
conflict. Also I find it difficult to believe the old adage that
‘the bees will always choose the best queen’. From the bee-
keeper’s point of view the new queen is usually preferable.

• If extra colonies are required then, with the new colony on

a separate stand, this has already been accomplished. If a split
board has been used then moving the new colony to a differ-
ent apiary to give it independence is preferable as there will
be no loss of flying bees. If a new colony on a split board is to
be kept in the same apiary care must be taken to avoid exces-
sive loss of flying bees. It is recommended that independence-
day should be delayed for a minimum of 5-6 weeks after the
new queen has started to lay – which, in reality, may be pretty
much the end of the season.

2.2 Colony has already issued the prime swarm but has not cast swarm

It is assumed that the beekeeper has already determined that the

prime swarm has departed with the old queen. It is then just a matter
of carefully checking that no queen cells have emerged. The age of the
youngest brood gives some clues, eg. if there are no eggs but some newly
hatched larvae then the swarm occurred 3-4 days ago and, under normal
circumstances, the queen cells are about 4 days from emergence.

Wally Shaw 31
 In this situation, unless something is done to prevent it, the colony
is likely to issue at least one cast swam. The timing of this swarm will be
2-4 days after the first virgin queen cell has emerged. There are two meth-
ods of preventing this:-

1) Thinning the queen cells to just one, selecting a cell of

good size in a well-protected position. If there are still eggs
or young larvae present it is best to delay this operation un-
til they are past the stage when these could be used to make
emergency queen cells. Alternatively the existing queen cells
can be thinned now and the hive checked for new queen cells
in a few days’ time.

2) The queen cells can be left intact until the estimated date
for their emergence – a few queen cells can be investigated
to check on their state of development and likely due-date.
What to do next is described in the next section (2.3).

2.3 Colony has emerged queen cells and may (or may not) have issued
a cast swarm

There is no easy way of knowing whether the colony has already

cast swarmed and no way of knowing for sure if the colony contains an
emerged virgin queen (or queens). Most of the remaining queen cells will
contain queens that are waiting to emerge but are being prevented from
doing so by ‘warder bees’ posted on the cells. Using the tip of a knife the
beekeeper should carefully open several cells and let the virgin queens
walk out into the colony. When you think you have done enough of this
(ie. given the bees plenty to choose from), ALL remaining queen cells
must be destroyed. It does not seem to matter how many virgin queens
you release in this way, the colony will proceed to select one (by whatever
means) and make no attempt to swarm. This is mission accomplished -
the bees have selected the new queen not you. Further information can
be found in the WBKA booklet, ‘There are queen cells in my hive – what
should I do?’

32 An Apiary Guide to Swarm Control 2nd Edition

2.4 Colony appears to be queen-less (it has no brood) and beekeeper
has no idea what happened and when

There may be a queen in the colony who is just about to start laying but,
without actually finding her, you cannot tell. The behaviour of the colony
(the bees seem calm) and the presence of laying arcs (cells prepared for
a queen to lay) imply that all is well with the colony but neither of these
signs is completely reliable. The best thing to do in this situation is to
insert a test frame (taken from another hive) containing eggs and young
larvae. If the colony is queen-less emergency queen cells will be made on
this frame but if they think they have a queen the donor brood will be
raised in the normal way. Even this is not 100% reliable because in rare
cases the colony may contain a non-laying queen and no further progress
can be made (ie. the colony cannot be re-queened) until she has been
found and removed. Further information on this situation can be found
in the WBKA booklet, ‘There are queen cells in my hive – what should I
do?’

Wally Shaw 33
2.5 Late Season Swarming

This can be one of the unfortunate by-products of pre-emptive

swarm control. Typically a large colony that has been kept together with
no attempt at swarming until late-June or early-July will suddenly develop
queen cells. This could still be controlled by means of artificial swarming
(as described above) but splitting the colony just when the main nectar
flow is about to start is definitely the last thing the beekeeper wants. One
method of dealing with this situation is by simply removing the queen.
You do not need to do anything drastic like kill her (you will probably
want to repatriate her later), you merely put her aside to tick-over with a
few workers in support. We call this practice putting the queen in ‘pur-
dah’; it is not a perfect solution but it works well enough.

As soon as queen cells are found the queen should be removed

along with a small amount of brood and some worker bees and either
installed in a nuc box or a shallow brood box which is placed on a split
board on top of the hive. There must be sufficient bees to support her and
cover any brood (remember all the flying bees will return to the bottom of
the hive). There also needs to be enough stores because foraging activity
will be limited. It is also a time of year when robbing may occur so there
needs to be enough bees and a small entrance to prevent this happening.

What to do with the queen cells?

The best option is to destroy the queen cells but first ensuring that
there are eggs and young larvae present from which the colony can start
emergency queen cells. The colony will probably be switched into emer-
gency re-queening mode and when these cells are mature (about 12 days
later) they will choose a new queen and not swarm. If you want to be sure
that swarming does not occur then follow the method of releasing virgin
queens as described in Section 2.3 (above).

BE AWARE - queens mated later in the season (from July on)

are known to be less reliable (less likely to survive their first winter) than
those mated earlier (May and June). It is thought this is due to the in-
34 An Apiary Guide to Swarm Control 2nd Edition
creased probability of the queen mating with a drone(s) that is carrying
deformed wing virus (DWV) and becoming infected herself – in other
words a sexually transmitted disease (STD). For this reason 9-10 days
later it is better to destroy the emergency queen cells and repatriate the
old queen, as in the second manipulation for the Snelgrove II method
described above.

During this enforced queen-less period the colony will continue

to forage well (but probably not quite as well as it would with an incum-
bent laying queen) and there will be little loss of honey crop.

A new method of dealing with a late season swarming colony

Since the first edition of this booklet was written we have been
experimenting with methods of dealing with late swarming and so far
(the last 3 years with more than 20 replications) the method shown in
Figure 10 (the Demaree Method) has been completely reliable. As shown

A new method of dealing with a late season swarming colony

Wally Shaw 35
in Figure 10, the colony is in the same condition as it would be for the
second manipulation of Snelgrove II (mod) – the queen cells in the parent
colony have been torn down and the queen has resumed laying. There are
emergency queen cells in the artificial swarm. Initially the manipulation
proceeds the same way but this time destroying the emergency queen
cells before the queen is repatriated to the artificial swarm.

The next move makes use of the multiple doors on the Snelgrove
board but a single entrance split board can simply be turned so the en-
trance is in the required new position. First the mesh from the middle
of the board is removed so the parent colony and artificial swarm can
mingle. The door on the Snelgrove from which the bees have been fly-
ing is closed and door at the front of the hive (top or bottom or both) is
opened (or the single entrance board is turned to this position). The bees
that fly from the parent colony will soon adapt to the new orientation (a
90°change makes for an easier transition). After 24 hours the board should
be removed altogether and the bees will quickly adapt to use the entrance
at the bottom of the hive. NB. It is essential that the bees in the parent
colony are forced to descend to the box where the queen now is in or-
der to fly. If this does not happen they will think they are queen–less
and start emergency queen cells at the top of the hive.

The desired result has now been achieved and, after limited sepa-
ration (9-10 days), the whole colony (and particularly the foraging force)
is back together as one unit. Brood in the Demaree boxes will emerge
and, depending on the time and the nectar flow, the frames will be used
to store honey which can become part of the honey harvest (or later used
as food frames for the bottom brood area).

Post-script

The above methods of pre-emptive and re-active swarm control

should provide the beekeeper with a comprehensive package of man-
agement techniques by which the swarming impulse of honey bee colo-
nies can be kept under some measure of control. The qualification ‘some
measure’ is used because the beekeeper who claims to have achieved
36 An Apiary Guide to Swarm Control 2nd Edition
complete control is probably ‘not of this world’ or is being economical
with the truth. Colonies are individuals and the most common type to
evade swarm control are those that start queen cells and swarm long
before the first one is sealed, thereby evading detection by routine inspec-
tion (a Houdini colony). Even with a 6-7 day inspection interval during
the swarming season it is impossible to prevent this sort of thing from
occasionally happening – except by clipping the queen’s wings. Queen
clipping is not really a method of swarm control, it just extends the safe
inspection period to about 14 days. As a management practice it has pros
and cons and often elicits strong opinions. A full discussion of queen clip-
ping is beyond the scope of this booklet.

To bring this booklet through the full circle, I return to what was
said at the end of the second paragraph of the Introduction; ‘there is no
doubt that swarm control is simultaneously the most important and
most difficult aspect of colony management’. We can but try!

Wally Shaw 37
APPENDIX 1. Further Discussion of the Pagden Method

Because this is the most widely recommended method of artificial

swarming which can be found in most beekeeping books (usually with
little explanation), it may be worth some further discussion.

The theory behind the Pagden Method is that the artificial swarm
it creates resembles a natural swarm, ‘thinks’ it has swarmed and will now
settle down to life in the slow lane. The reality is that the artificial swarm is
nothing like a natural swarm because it has an entirely different age struc-
ture, being dominated by older bees, whereas a natural swarm is domi-
nated by young bees. The subsequent performance of an artificial swarm
compared with that of a natural swarm clearly underlines the difference.

The probable explanation for the unreliability of the Pagden

Method is because the artificial swarm created by the beekeeper has a
structure that does not occur in nature. This means that the colony has
no evolved (hardwired) solution in its behaviour repertoire that it can
engage in order to achieve the best outcome (for itself). The only pos-
sible options are either to start some new queen cells and persist with the
swarming impulse or abandon the idea, settle down and re-build a more
normal age structure as quickly as possible to ensure survival. Users of
the method obviously expect/hope that the artificial swarm will adopt the
second option. The Heddon modification does not influence the outcome
and probably only makes the situation worst by transferring more bees to
the artificial swarm, thus making the potential swarm larger. The decision
(to swarm or not to swarm) seems to be finely balanced and there is no
known way the beekeeper can influence this.

The method described in the main text (Snelgrove II

(modified), differs by putting the artificial swarm into a condition which
it does understand and does have a clear-cut response to ensure its sur-
vival, ie. it finds itself queen-less but with the means to make a replace-
ment (emergency) queen. Being forced to make emergency queen cells
puts the colony into emergency re-queening mode and this switches off
(or suppresses) the swarming impulse.
38 An Apiary Guide to Swarm Control 2nd Edition
APPENDIX 2. Fault Finding

Feedback since 1st edition - A few people who have used the
Snelgrove II (mod.) Method have reported that it has not worked for
them. With one exception (details below) they have been unable to pro-
vide sufficient details to determine the reason for the failure. Over the
years we have had one failure in well over 100 replications but are confi-
dent we understand the cause (details below). However, there are other
*possibilities and the following is a Fault Finding check list. *NB. These
are just logical possibilities and we do not know for sure whether anybody
has ever made any of these mistakes.

1) As noted in the main text, the second manipulation (9-10

days later) starts with checking that the queen cells in the par-
ent colony have been torn down and the queen has resumed
laying. If the queen cells have not been torn down and there
are no eggs to be found then a mistake was made in the initial
manipulation, ie. the queen is not in the parent colony and
she has either been lost altogether or is still somewhere in the
artificial swarm.

2) The next step is to check the artificial swarm it see if emer-

gency queen cells have been produced on the two donor
frames. If there are queen cells then the missing queen cannot
be in the artificial swarm. As already noted, the most likely
cause is that the colony had already swarmed but there are
other possibilities - the queen had been killed, dropped on
the ground or has flown off.

3) If there are no queen cells then the queen is still residing

somewhere in the artificial swarm. Another possibility is that
the frames did not contain brood young enough for emer-
gency queen cell to made.

4) If there are eggs and young larvae in brood box contain-

ing the artificial swarm then you have inadvertently created

Wally Shaw 39
 a Pagden type artificial swarm. In which case it will probably
either have worked or not worked by now (9-10 days later).

a) If it has swarmed then there should be no

queen present and the age of the youngest brood
will tell you when this happened.

b) If it has not swarmed (but still could do so)

there will be a queen and recently laid eggs but
you will need to check for the start of queen cells
for about another 10 days.

c) In both cases (a and b) to prevent further

swarming you need to thin queen cells or emerge
virgin queens as recommended in Section 2.3 in
the main text.

5) If there are no queen cells in the artificial swarm there is

one final possibility and this is that the queen somehow found
her way into supers - but this must have happened recently or
there would have been no queen cells in the brood box. If the
queen is in the supers the best option is probably to transfer
her to the parent colony and hope for the best.

Known Causes of Failure

Our failure – at the time the artificial swarm was made the colony
was almost certainly on the verge of swarming. Either there was insuffi-
cient time for the flying bees to return to the artificial swarm or the weath-
er prevented this happening. With the flying bees still present, the parent
colony retained the urge to swarm and did so at the first opportunity.

Other known failure – being nervous about having queen cells

and the queen together in the parent colony, one beekeeper destroyed all
the queen cells himself instead of allowing the bees to do this in their own
time. This altered the behaviour of the parent colony which swarmed at
40 An Apiary Guide to Swarm Control 2nd Edition
some time during the next few days.

APPENDIX 3. Recommended Logistics for the Second manipulation

If the queen cells in the parent colony have been torn down and the
queen has resumed laying there will automatically be emergency queen
cells in the artificial swarm. This is a sign that the first manipulation has
gone according to plan. The main aim of the second manipulation is to
find the queen in the parent colony and transfer to the artificial swarm.

If a split board has been used, the artificial swarm cannot be ac-
cessed until the parent colony (on its split board) has been removed from
the top of the hive. It is best if the queen is found and put in a safe place
before this is done. If the hive is high (and it usually is by now) it can be
difficult to remove the top colony single-handed and it is useful to have
an assistant. Having a spare roof or box available is a useful place to tem-
porarily put the parent colony still on its split board.

Wally Shaw 41
 It is always best to find the queen before there has been too much
disturbance. In order to do this operation smoothly, the frame she is on
should be removed and carefully placed in a holding box where she will
be safe. A nuc. box (with its entrance blocked) can be used but we have a
box that holds a single frame specifically made for any manipulation that
requires the queen to be put in a safe place (see Figure 11).

It is best to repatriate the queen on the frame where she is found

surrounded by her own nest mates. She can be introduced on her own,
and we have done this with no problem a few times. Other people have
expressed concern about this practice (even reported that the queen had
been killed) so, to avoid that possibility, this is a safer way to ensure her
acceptance by bees from which she has been separated for 9-10 days.*
Using a split board is inherently safer because the common hive smell is
retained.

*By this time the emergency queen cells (that have just been re-
moved from the artificial swarm) will have reached the stage in their de-
velopment where the bees are fully aware of their presence and have start-
ed to show a keen interest in them. The colony should be expecting the
emergence of a virgin queen in the near future but seems to quite happily
accept that instead they suddenly get the gift of a mated and laying queen
(a miracle?).

If the artificial swarm and the parent colony are on different stands
Stage 2 is a relatively easy operation because both hives can be open at the
same time. Having found and secured the queen, the next step is to re-
move the two frames with queen cells from the artificial swarm (supers
removed first obviously). The frame with the queen on should then be
placed in the gap thus created. Another frame from the parent colony,
complete with its bees, can then be inserted next to it to fill the second
space thus providing more ‘friendly faces’ nearby - and also a useful injec-
tion of nurse age bees.

It is then just a matter of deciding what to with the emergency

42 An Apiary Guide to Swarm Control 2nd Edition

queen cells; whether to destroy them and let the parent colony re-queen
itself from scratch or transfer them intact to the parent colony to provide
a new queen 9-10 days earlier (see discussion page 26).

The last task is to re-assemble the colonies complete with supers

and adding extra supers if required. Bear in mind that the colony on the
split board has yet to re-queen and the hive should be disturbed as little
as possible until there is evidence that the new queen has started to lay.
However, if access to the bottom part of the hive is for some reason re-
quired it should be done outside potential mating hours, ie. before 10am
or after 5pm.

Wally Shaw 43