The Lives of Bees
Page 20
tioned in chapter 6 (see Fig. 6.5). Page measured the areas of capped drone
cells (pupal brood) in 13 colonies living in standard, movable- frame hives
Seeley.indb 158
2/21/2019 8:07:43 AM
Colony Reproduction 159
with two frames of drone comb per 10- frame hive body. Smith et al. mea-
sured the areas of drone comb that contained brood (eggs, larvae, and
pupae) in four colonies that were living in large observation hives filled
with natural combs that the colonies had built the previous summer. For
each study, I have converted what the authors reported—measurements
of area of occupied drone comb on various days across the summer—to
estimates of number of drone cells occupied on each sampling date. I then
calculated the total number of drones produced per colony across the
summer by calculating the total number of occupied brood cell- days per
colony throughout a summer and dividing this by the relevant develop-
ment time for drones: the 14- day capped brood period for the Page study
and the 24- day entire brood period for the Smith et al. study. The two
studies yielded similar values for the average number of drones produced
over a summer by an unmanaged colony living in a hive with a normal
amount of drone comb: 7,812 drones (Page) and 6,949 drones (Smith et
al.). The significance of these two numbers will become clearer later in the
chapter, when we use them to compare the investments that colonies make
in their male (drone) and their female (queen) means of reproduction.
Because honey bee colonies benefit from starting to rear drones in
early spring, and because they have filled their drone comb with honey in
the previous summer and fall, they often face a problem in early spring
of having many of the cells in their drone comb plugged with honey. It is
not surprising, therefore, that colonies preferentially remove honey from
their drone comb in the spring, when this comb is needed for rearing
drones, and preferentially store honey in their drone comb in late summer
and autumn, when this comb is best used for honey storage. This seasonal
shift in the use of drone comb for honey storage was demonstrated re-
cently in a study Michael L. Smith and colleagues conducted in Ithaca, in
which once a month, from April to September, they installed in the hives
of several colonies two frames of comb—one of drone comb and the
other of worker comb—whose cells they had filled with thick sugar
syrup. The two test frames installed in each colony’s hive were positioned
on opposite sides of two frames that contained brood; this ensured that
there were nurse bees near both test frames. Fourteen days later, the
Seeley.indb 159
2/21/2019 8:07:43 AM
160 Chapter 7
Fig. 7.3. Drone comb before (top) and after (bottom) being placed in a colony
for 14 days in April, when the colony was preparing to rear drones en masse. In
both images, the cells with reflections contain sugar syrup. The bottom frame
shows that the bees removed “honey” from the center of the drone comb to make
space for rearing drones. The study colony was living in Ithaca, New York.
investigators removed the two test frames from each hive and measured
in each the area of comb that had been cleared of sugar syrup (see Fig.
7.3). They found that in April and May, the average comb area cleared was
markedly larger for drone comb than for worker comb, and that in August
and September the pattern was reversed: the average comb area cleared
was noticeably smaller for drone comb than worker comb. Presumably,
Seeley.indb 160
2/21/2019 8:07:44 AM
Colony Reproduction 161
the workers in these colonies did not remove much sugar syrup from
their drone comb in late summer and autumn because they knew that
drone production was no longer the most important use of this special,
large- celled comb.
QUEEN PRODUCTION AND SWARMING
The life cycle of a honey bee colony can be regarded as beginning in the
spring when an established colony builds up its worker population and
starts rearing a batch of queens in preparation for swarming. The first step
in these preparations is the construction of queen cups along the lower
margins of the colony’s brood- nest combs. These queen cups, tiny inverted
bowls made of beeswax, form the bases of the large, ellipsoidal cells in
which queens are reared (Fig. 6.3). Next, the queen lays eggs in a dozen
or more of the queen cups, and workers feed the hatching larvae the royal
jelly that ensures their development into queens. The formation of these
new queens is remarkably rapid; only 16 days pass from when an egg is laid
to the moment an adult queen climbs from her cell. As the daughter
queens develop, changes unfold simultaneously in the physiology of the
mother queen in the colony. With each passing day, she is fed less and less
by the workers. Her egg production declines, and her abdomen, no longer
swollen with fully formed eggs, shrinks dramatically. Furthermore, the
workers begin to shake their queen, grabbing onto her one at a time with
their front legs and letting loose a volley of five or six shaking movements.
These bouts of shaking, which can eventually reach a frequency of 40 to
80 per hour, appear to force the queen to keep walking about the nest.
This exercise, together with reduced feeding, results in a 25 percent re-
duction in the queen’s body weight. Shortly after the first queen cell is
capped, the mother queen flies off in a swarm of some 10,000–20,000
workers, leaving behind only about a quarter of the colony’s population of
worker bees in the parental nest. After flying a short distance, the swarm
condenses into a beard- like cluster on a tree branch (Fig. 7.1). From here
the swarm’s scout bees explore for nest cavities, select one that is suitable,
and finally signal the swarm to break cluster and fly to the chosen home-
site. The new dwelling place is rarely less than 300 meters (ca. 1,000 feet)
Seeley.indb 161
2/21/2019 8:07:44 AM
162 Chapter 7
from the bees’ original residence, and can be 3,000 or more meters (more
than 2 miles) away.
For about eight days following the departure of their mother queen, the
workers in the parental nest are queenless, but this situation ends with the
emergence of the first daughter queen. If the colony is still greatly weak-
ened by the departure of the first swarm—what beekeepers call the prime
swarm—then the remaining workers allow the daughter queen that
emerges first to search through the nest to find her rival sister queens and
kill them, by stinging them while they are still in their cells. Usually, how-
ever, by the time the first daughter has appeared, enough young worker
bees have emerged from cells in the brood combs to restore the parent
colony’s strength. In this situation, the workers guard the remaining queen
cells against destruction by the first daughter queen, they start shaking this
queen in preparation for flight, and eventually they push her out of the nest
in an afterswarm. As
is shown in Figure 7.4, this process of afterswarming
may be repeated with another daughter queen, and when this happens the
colony is usually left weakened to the point where it cannot support fur-
ther division. At this point, if there remains more than one daughter queen
roaming the parental nest, the workers allow these queens to fight each
other until just one remains alive. It is she who, partly by luck and partly
by skill, inherits the parental nest with its rich endowments of beeswax
combs and honey stores, both of which are immensely important assets
that will give the colony living in the parental nest a high likelihood of
survival through the coming winter.
The production of afterswarms depends strongly on a colony’s residual
strength—measured in workers and especially in brood—after the prime
swarm has departed, so the number of afterswarms produced per episode
of colony reproduction varies greatly. Fortunately, the results from several
detailed studies of swarming and afterswarming by unmanaged colonies
make it possible to put probabilities on the events shown in Figure 7.4.
First, we know from long- term studies (described in the next section of
this chapter) that, on average, the annual probability of queen turnover
within an unmanaged colony living in the region of Ithaca is 0.87. There-
Seeley.indb 162
2/21/2019 8:07:44 AM
Colony Reproduction 163
Cycle begins
Colony builds Swarming
up in spring
0.87
0.13
1.00
0.70
0.60
1.00
0.81
Mother Q
Mother Q
Daughter Q1
Daughter Q2
Daughter Qn
stays at home
departs in
departs in
0.23
departs in
inherits the
(no swarming)
prime swarm afterswarm #1
afterswarm #2 original nest
survives to next spring0.12
Moves into
Col. 1
y
Col. 2
Col. 3
Col. 4
n
new site
ol
Moves into
o C
new site
Moves into
0.12
new site
Stays in
mates
old site
0.81
Queen
Fig. 7.4. Principal events in the life cycle of honey bee colonies, starting in the
spring when a colony builds up its worker population, which sets the stage for
swarming. Q = queen. Numbers along the lines denote the probabilities of the
various events (e.g., the probability of a colony swarming after building up in the
spring is 0.87).
fore, 0.87 is a good estimate of the probability that on any given year the
mother queen in a colony will leave her nest in a prime swarm and will
occupy a new nest cavity located several hundred or several thousand me-
ters away. Second, we know from painstaking studies performed by Mark
L. Winston, working in Lawrence, Kansas, and by David C. Gilley and
David R. Tarpy, working in Ithaca, a great deal about the fates of the daugh-
ter queens that are produced in unmanaged colonies after the mother
queen has departed in the prime swarm. Gilley and Tarpy, for example,
worked with colonies living in large observation hives that enabled them—
aided by a team of helpers—to monitor continuously the activities of the
daughter queens in five colonies, each of which had cast a prime swarm.
They maintained a round- the- clock surveillance of the daughter queens in
their observation hive colonies until each one had either departed or been
killed, except the one who inherited the parental nest. Taken together, the
Winston and the Gilley and Tarpy studies show us that in an unmanaged
colony that has produced a prime swarm, the probability that one of the
Seeley.indb 163
2/21/2019 8:07:44 AM
164 Chapter 7
daughter queens will leave in a first afterswarm is 0.70, the probability
that another daughter queen will leave in a second afterswarm is 0.60, and
the probability that a third daughter queen will inherit the original nest
(after killing all her rivals) is 1.00.
The process of queen production and colony foundation is completed
when all the surviving daughter queens have flown from their nests and
mated with drones from neighboring colonies (discussed later in this chap-
ter). At this point, the colonies that have moved into new nest sites have
begun building their nests, and all the colonies—including the one occu-
pying the old nest—are rearing brood to build up their populations and
are foraging intensively to build up their honey stores to prepare for win-
ter. The probability of surviving the coming winter is quite high, approxi-
mately 0.81, for the fortunate daughter queen that inherited the old nest
and its store of honey. Sadly, for her mother queen and her sister queens,
whose colonies must build new nests from scratch, the probabilities of
winter survival are much lower, often less than 0.20, for reasons that will
be discussed shortly.
One might wonder, why has natural selection favored mother queens
who leave the old nest in a prime swarm and thereby incur a low probabil-
ity of winter survival in a new nest? I think the answer is simple: by leaving
in the prime swarm rather than lingering in the old nest, a mother queen
dodges the high risk of being killed by one of her daughter queens when
they start emerging from their cells. The danger to the mother queen of
staying at home is borne out by the data on regicide committed by virgin
queens, as reported by Gilley and Tarpy, and by M. Delia Allen, who
worked in Aberdeen, Scotland, in the 1950s. These three investigators
report the fates of 44 virgin queens who were reared in six study colonies
that were living in observation hives and that swarmed. The researchers
observed that within a colony, on average, one virgin queen left in an af-
terswarm, one virgin queen inherited the original nest, and 5.3 virgin
queens died from being stung by a fellow virgin queen. Clearly, the mother
queen does well to flee the killing field of her old nest before her murder-
ous daughter queens emerge from their cells.
Seeley.indb 164
2/21/2019 8:07:44 AM
Colony Reproduction 165
HOW A POPULATION OF WILD
COLONIES IS PERSISTING
Under favorable conditions, a colony that is alive at the end of a summer
will survive the following winter and will go on to reproduce the next
summer. Colonies living on their own, however, don’t always experience
favorable conditions, and many perish over winter through starvation, dis-
ease, or failure to replace a senescing queen. If the rate of colony deaths
exceeds the rate of colony births (through swarming), then the population
of colonies in a region will shrink and may even expire. In chapter 2, we
reviewed evidence that the population of wild honey bee colonies living
in an
d around the Arnot Forest has been stable since it evolved resistance
to Varroa mites in the 1990s. Let us now examine how this population of
wild colonies can persist. We will do so by reviewing what I have learned
about the patterns of colony generation and colony loss for honey bees
living on their own in the wild places outside of Ithaca. What we know
about these matters comes from two long- term studies that I made—in
1974–1977 and in 2010–2016—on the demography of wild colonies.
Both studies comprise two avenues of investigation: one of reproduction
(swarming) by simulated wild colonies (SWCs) living in movable- frame
hives, and one of survival by wild colonies living in natural nest sites. The
hive- based work on colony reproduction involved setting up approxi-
mately 20 SWCs in separate, secluded places. Each SWC was established
by catching a natural swarm—either by collecting it from its bivouac site
or by capturing it in a bait hive—and then installing it in a 10- frame Lang-
stroth hive (Fig. 7.5). The hive contained two frames of drone comb and
eight frames of worker comb, and its entrance was reduced to a small,
natural- size opening. In short, each SWC occupied a hive that simulated a
natural cavity, except that its wooden walls were thinner and its entrance
was lower. I labeled the queen in each colony with a paint mark so that I
could detect turnovers of a colony’s queen—presumably by swarming,
primarily. I kept the colonies’ queens labeled over the years by applying
paint marks to any unlabeled queens I found during colony inspections. I
Seeley.indb 165
2/21/2019 8:07:44 AM
166 Chapter 7
Fig. 7.5. One of the hives used for housing a simulated wild colony. The blue
structure is a screen board used in getting counts of how many Varroa mites
dropped onto a sticky board in 24 hours. Each colony’s hive was permanently
equipped with a screen board to make noninvasive measurements of the colony’s
mite load.
inspected each SWC three times each summer, in early May, late July, and
late September. This meant that each colony was inspected before and after
the main swarming season for the area around Ithaca (mid- May to mid-
July) and before and after the secondary swarming season (mid- August to
mid- September). These inspections served two purposes: to check each