The Lives of Bees
Page 16
side. Next, the raised edges of these cells are sculpted into thin lines to
form the bases of each cell’s six walls, with the adjoining walls laid out at
an angle of 120 degrees. This gives each cell a hexagonal cross section from
the start. As additional bits of wax are deposited, the bases of additional
cells begin to take shape at the appropriate distances from the preexisting
cells, and the walls of the first cells are raised by adding rough particles of
wax to the top of each wall and then shaving each one down on both sides
to form a thin, smooth, plane of wax in the middle (Fig. 5.10, middle). The
cutaway wax is then piled up, together with some fresh wax, on the top of
the wall, and the process is repeated. So, bit by bit, a wonderfully thin wall
of wax grows steadily outward, always crowned by a broad coping.
Throughout this process of comb construction there runs a theme of
economy in the use of the energetically expensive beeswax. The most
conspicuous expression of this frugality is the cell shape in the combs of
honey bees: a right hexagonal prism capped on the inner end by a trihedral
pyramid. Because the cells of honey bee nests were originally circular in
Seeley.indb 124
2/21/2019 8:07:36 AM
The Nest 125
0
10 mm
0
0.5 mm
Fig. 5.10. Left: Illustration of the economy achieved by building hexagonal rather
than circular cells in wax. Middle: Cross section through a comb showing the way
bees precisely plane down cell walls during construction to minimize cost of
comb building. Right: Cross sections of the outer margins of cell walls built by
normal bees (top) and by bees from which the six distal segments of their anten-
nae were amputated (bottom). Normally, loose bits of wax are packed together
at the tip of the cell wall while the rest of the wall is a single, thin layer of wax.
Bees with antennal operations built costly, triple- layered cell walls.
cross section, as they still are in the nests of all other bees, one can think
of the comb in a honey bee nest as an assemblage of identical cylinders
compressed into hexagonal prisms. The perimeter of a hexagon of a given
area is 5 percent longer than the perimeter of a circle with the same area,
but because each hexagonal cell in a comb shares its walls with other cells,
whereas circular cells do not, building the cell walls in a comb with hex-
agonal cells requires only about 52 percent of the wax needed to build
those in a comb with circular cells. For example, the average wall- to- wall
distance of the worker cells in the nests of the wild colonies that I dissected
was 5.20 millimeters (0.20 inches). A hexagon of this size has an area of
Seeley.indb 125
2/21/2019 8:07:36 AM
126 Chapter 5
23.40 square millimeters (0.04 square inches) and a perimeter of 18.01
millimeters (0.71 inch). A circle with the same area has a perimeter of only
17.15 millimeters (0.67 inch). But because each cell wall in a hexagonal-
cell comb is shared by two cells, the effective perimeter per hexagonal cell
is just 18.01 millimeters divided by two, which equals 9.00 millimeters
(0.35 inch), and 9.00 millimeters divided by 17.15 millimeters equals
0.52, hence the figure of 52 percent mentioned above. A comb of circular
cells also requires wax to fill the spaces between the cell walls (Fig. 5.10,
left), so the volume of wax required to build a hexagonal- cell comb with
a given number of cells is, all things considered, less than half that required
to build a circular- cell comb with the same number of cells.
Several other features of the comb construction process, beside the
hexagonal cell design, also contribute to economy in wax use by honey bee
colonies. One is the skill of worker bees in shaving down the wax parti-
tions between cells, which leaves the walls of the cells only 0.073 millime-
ter (0.003 inch) thick. An experimental investigation of the sensory abili-
ties of worker honey bees has revealed that a worker’s antennae play a
critical role in the process of sensing cell- wall thickness. When the six
distal segments of both antennae of several hundred bees were amputated,
and their comb building was studied, it was found that their building prac-
tices were grossly disrupted. Some cells were built with holes gnawed in
their walls, while other cells were built with walls twice as thick as normal
(Fig. 5.10, right). Because the temperature and composition of the bees’
building material (beeswax) are constant, and because the shape of the
cells in their combs is uniform, it is likely that a worker bee can judge the
thickness of a cell wall by pressing on it with her mandibles and noting the
elastic resiliency of the substrate with her antennae.
Bees achieve further economy in wax production by constantly recy-
cling old wax. Whenever a bee emerges as an adult, the fragments of the
cap of her brood cell are carefully bitten off, by the emerging bee or
nearby nurse bees, and then are stuck to the cell’s rim for reuse later.
Likewise, queen cells—the special, large, peanut- shaped cells in which
queen bees are reared—are built from bits of wax cut from adjoining
Seeley.indb 126
2/21/2019 8:07:37 AM
The Nest 127
worker cells, and once vacated they are torn down to free wax for other
purposes.
Perhaps the most important way that a honey bee colony achieves econ-
omy in wax use is by carefully timing its comb building, limiting it to pe-
riods when the benefits of comb construction outweigh the costs of wax
production. Consider first the situation of a swarm that has just moved
into an empty tree hollow. Here the benefits of comb building are huge,
for comb is vital to the colony’s future. A fledgling colony cannot begin to
rear brood or store food until it has built some combs, so it makes sense
that it invests in a burst of comb building. This phenomenon was described
recently by one of my PhD students, Michael L. Smith, who monitored
the lives of several honey bee colonies from the moment they occupied
their nest cavities to when they died. Each colony started out as an artifi-
cial swarm of average size—some 12,000 worker bees plus a queen—and
was installed in a large, glass- walled observation hive that provided a 38-
liter (10- gallon) nest cavity. Each swarm was fed sugar syrup ad libitum
before it was installed in its hive, so its members were stuffed with energy-
rich food when they moved into their new homesite, as is usually the case
for natural swarms. For the next 20 months, the three colonies were left
undisturbed, except that once a week, at night, the insulation boards cov-
ering the glass walls of each hive were removed to census the colony’s
worker population and drone population; trace the area of its worker
comb and drone comb; and record which comb cells contained brood,
pollen, honey, or nothing at all.
Figure 5.11 shows how each colony quickly built, over the first three
weeks, 2,000–4,000 square centimeters (ca. 300–600 square inches) of
comb surface
. This much comb comprises only 8,500 to 17,000 worker
cells—less than 20 percent of the number found in a completed nest—but
it was enough for each colony to begin rearing worker bees within a few
days of moving into its new home. Doing so was critically important,
because the development period of a worker honey bee is 21 days, which
means that for the first three weeks after moving into its new home, a
colony’s population shrinks steadily, as old workers are dying off and
Seeley.indb 127
2/21/2019 8:07:37 AM
128 Chapter 5
worker
drone
12
Colony 1
8
) 4
2
0
s of cm 12 Colony 2
8
4
0
Fig. 5.11. Top: Histories of comb
building in three unmanaged col-
Comb area (1000’ 12
Colony 3
onies, each of which began as a
8
12,000- bee swarm in July 2012.
4
Magenta line, worker comb; blue
0
line, drone comb. Each colony had
J A S O N D J F M A M J J A S
just two pulses of comb building in
2012
2013
the first year: one when it moved
15
Colony 1
into its empty hive and one in late
10
4
August and early September, when
5
2 Number of drones (1000’
goldenrod ( Solidago spp.) plants
s) 0
0
were producing copious nectar.
15
Colony 2
Comb areas shown represent both
10
4
sides of each comb. Bottom: Popu-
5
2
lation dynamics in the same three
0
0
colonies. Magenta line, workers;
blue line, drones. The casting of a
15
Colony 3
s)
swarm (or an afterswarm) is marked
Number of workers (1000’ 10
4
by an asterisk (or two). Arrow in
5
2
colony 2 plot marks when colony
0
0
J A S O N D J F M A M J J A S
was given a queen to replace its
2012
2013
original queen, which had died.
young ones are not yet emerging to replenish the workforce. Indeed, Fig-
ure 5.11 shows that by the end of the first three weeks, each colony’s
worker population had fallen to just 20–50 percent of its original level,
and that even by summer’s end none of the colonies had regained its origi-
nal workforce size.
Seeley.indb 128
2/21/2019 8:07:37 AM
The Nest 129
After its initial burst of comb building to initiate brood rearing and food
storage, a colony faces a dilemma with respect to further comb building.
On the one hand, building more comb can help a colony to efficiently
exploit unpredictable nectar flows by giving the colony spare storage
space, which helps keep the nectar- receiver bees from becoming mired in
long searches for empty cells in which to deposit the fresh nectar. On the
other hand, building more comb can shrink a colony’s honey stores through
the costly production of wax. Back in the 1990s, another one of my PhD
students, Stephen C. Pratt, decided to investigate how the bees deal with
this dilemma. Using a technique called stochastic dynamic programming,
he started his work by modeling the situation that a colony faces in decid-
ing when, and how much, to invest in comb building. His goal was to de-
termine the optimal timing of additional comb construction during a
colony’s first summer based on the availability of nectar to be collected in
the field, the amount of nectar already stored in the comb (as honey), and
the amount of comb already present in the nest. Stephen’s modeling work
revealed that an established colony will achieve nearly optimal timing of
comb building by limiting the construction of additional comb to times
when two requirements are met: 1) the colony has filled its comb above a
threshold level, and 2) the colony is busy collecting more nectar.
Stephen tested this prediction by performing experiments in which he
monitored the comb building by colonies living in three- frame observation
hives in which one frame was filled with brood, one frame was partially
filled with honey, and one frame was empty (to provide space for comb
building). He moved each study colony (one at a time) to a location with-
out natural nectar sources, so that he could control its rate of nectar intake
by adjusting the availability of “nectar” (sugar water) from a feeder. Figure
5.12 shows the results of one experiment that tested the prediction that a
colony that is collecting nectar starts to build additional comb only after a
threshold level of comb fullness has been reached. In phase 1, the colony
had a high rate of nectar intake, but its comb fullness was kept at a low
level. In phase 2, the high rate of nectar intake continued, and now the
colony could fill its storage comb with honey. Finally, in phase 3, the colony
Seeley.indb 129
2/21/2019 8:07:37 AM
130 Chapter 5
) 100
1
2
3
80
Cells filled (% 60
Fig. 5.12. Results of an experiment
that tested the role of comb fullness
) 300
in a colony’s decision to start build-
ing additional comb. Phase 1: Col-
200
ony had a high rate of “nectar” in-
take (collection of a 65% sucrose
100
solution from a feeder), but its
Nectar intake (ml
honey- storage comb was kept at a
50
low level of fullness. Phase 2: The
2 )
“nectar” intake continued, but now
the bees were permitted to fill their
30
storage comb with honey. Phase 3:
Colony was returned to the condi-
Comb built (cm 10
tions of phase 1. The search time
120
variable plotted at the bottom is the
90
time taken by a returning nectar
60
forager to find a food- storer bee
time (s)
willing to receive her nectar load.
Mean search
30
The three mean search times indi-
1
3
5
7
9
11
cated by dark bars are significantly
Day
higher than the others.
was returned to the conditions in phase 1. As predicted, the colony built
no comb in phase 1 but then started to do so in phase 2 once its comb full-
ness had reached a threshold level (about 80%). In phase 3, however, the
colony did not cease construction even though the fullness of its storage
comb had been reduced (by Stephen)
to a low level. This experiment shows
that colonies will start building comb before they have filled all the existing
storage comb in their nest. I suspect that the bees follow this strategy—
which entails some risk of premature comb construction—to be sure that
they have adequate storage space in the event of a large honey flow. This
Seeley.indb 130
2/21/2019 8:07:37 AM
The Nest 131
experiment also shows that once a colony has begun building comb, it will
continue doing this so long as it enjoys a strong influx of nectar and thus a
good supply of the fuel for comb building. Stephen also performed experi-
ments in which colonies were deprived of a high rate of nectar intake, but
their storage comb had a high level of fullness (85% or more). He found
that these colonies did not build comb. Taken together, these two experi-
ments show that Stephen’s modeling analysis was correct; a colony invests
in additional comb building only when it has reached a threshold level of
comb fullness and it is experiencing a strong influx of nectar.
How do worker bees monitor both the comb fullness and the nectar
intake rate of their colony so that they know when to start comb building
at the right time? One possibility is that when a colony’s storage combs get
close to being full, the nectar receivers—the bees that unload the return-
ing nectar foragers—experience increasing difficulty in finding cells for
storing the fresh nectar, and they respond to this coincidence of high nec-
tar influx and increasing comb fullness by starting to secrete wax and build
comb. In support of this hypothesis is our knowledge that the nectar re-
ceivers and the comb builders are both middle- aged bees, about 10–20
days old, which means that nectar receivers are the right age to become
comb builders. Further support of this hypothesis comes from the experi-
ment presented in Figure 5.12. In performing this experiment, Stephen
noticed that the average search time of a nectar forager—the time she spent
searching in the hive for a bee who would receive her nectar load—rose
markedly when the colony’s comb construction started up. This conspicu-
ous rise in the search times of the nectar foragers may have been caused
by many of the colony’s middle- aged bees switching from nectar receiving
to comb building. However, when Stephen applied paint marks to 30–40
percent of the nectar receivers in a colony, and then induced this colony
to start building comb, he found that the marked bees made up less than