The Lives of Bees
Page 32
nuclear DNA of the old and new bees living in the woods near Ithaca.
First, it revealed that since 1977 there has been some introgression of
genes from Africa (see Fig. 1.4), presumably via queens and colonies
moved by beekeepers from Florida to the area around Ithaca. This work
also revealed some clues about the mechanisms by which the colonies of
the modern bees can tolerate infestations of Varroa destructor. These are
signatures of selection—significant changes in the bees’ genes—scattered
among 634 sites across the honey bee genome, about half of which seem
to relate to the bees’ development. Evidently, the colonies that are surviv-
ing now in the wild possess resistance mechanisms that are based, at least
in part, on changes in the development programs of their members. Con-
sistent with the findings of changes in genes for development, Sasha found
that the worker bees collected in 2011 are markedly smaller in body
size—that is, in head width and distance between the bases of the wings—
than those collected in 1977, but just how this contributes to resistance to
Varroa remains a mystery.
Starting in 2015, one of my PhD students, David T. Peck, began to in-
vestigate the mechanisms of mite- resistance possessed by the worker bees
in the colonies living in the Arnot Forest. There are two general possibili-
ties: attack the female mites during the phoretic phase, when they are cling-
ing to adult bees, or do so during the reproductive phase, when they are
sealed in brood cells. When mites are in the phoretic phase, worker bees
can groom mites from themselves or nest mates, and then damage them
by chewing off the mites’ legs, antennae, and the structure called the
dorsal plate. When mites are in the reproductive phase, worker bees can
Seeley.indb 257
2/21/2019 8:08:00 AM
258 Chapter 10
disrupt the mites’ reproduction either by uncapping mite- infested cells and
removing the pupae inside— Varroa- sensitive hygienic (VSH) behavior—
or by simply uncapping and recapping the cells of pupating bees.
David has tested the Arnot Forest bees with respect to both general
ways for worker bees to suppress the population of Varroa mites in a col-
ony. He began by capturing swarms of honey bees in the forest using bait
hives that were hung out of reach of black bears (see Fig. 2.14) and then
moving the swarm- founded colonies into hives in an isolated apiary out-
side of the Arnot Forest (for safety from bears). He then tested the Arnot
Forest bees for the grooming/chewing responses that kill the mites in the
phoretic phase and for the disruption responses (VSH behavior and cell
uncapping and recapping) that kill them in the reproductive phase. He has
found that the workers in his colonies of Arnot Forest bees, relative to
workers in control colonies headed by queens of lines not selected for
Varroa resistance, have both a stronger grooming/chewing response and a
stronger VSH response. Also, he has found that some of his Arnot Forest
bee colonies have a high percentage (over 40%) of brood cells that had
been uncapped and recapped, which we are learning is another way that
the bees can disrupt the mites’ reproduction. What I think is the most
important insight gained from David’s work is that the worker bees in wild
colonies living in the Arnot Forest possess multiple behavioral mechanisms
for suppressing the mite populations in their colonies. In short, they are
deploying a diverse set of behavioral resistance weapons against Varroa
destructor, not just a single silver bullet. Bee breeders, take note.
LIVING WITH COLONIES FAR APART VS. CLOSE TOGETHER
When we humans switched from hunting wild colonies to keeping man-
aged colonies, we imposed on the colonies under our supervision a fun-
damental change in their ecology: an enormous reduction in colony spac-
ing. We have seen already that colonies in the wild usually live spaced far
apart, often with a kilometer (0.6 mile) or more of woodland between
their bee- tree homes (see Figs. 2.6, 2.12, and 7.11). Colonies managed by
beekeepers, however, almost always live crowded together. The clustering
Seeley.indb 258
2/21/2019 8:08:00 AM
Colony Defense 259
of colonies in apiaries is certainly beneficial for beekeepers because it
makes beekeeping practical, but it is not altogether beneficial for the bees.
Relative to wild colonies, managed colonies experience greater competi-
tion for forage, a higher risk of having their honey stolen, and more prob-
lems with reproduction, such as when a young queen coming home from
a mating flight enters the wrong hive and is killed by worker bees standing
guard against intruders.
I suspect, though, that the greatest harm that we inflict on honey bee
colonies by forcing them to live jam- packed in apiaries has come through
the evolution of highly virulent strains of their parasites and pathogens. We
cause this by facilitating the spread of disease agents between unrelated
colonies (so- called horizontal transmission). This mode of disease trans-
mission favors virulent strains, because horizontally transmitted parasites
and pathogens do not need their hosts to stay healthy. Instead, natural se-
lection favors the strains that reproduce rapidly in a host, for although this
damages the host, it boosts the parasite’s, or pathogen’s, odds of being
transmitted to another host. This way of life works well for parasites and
pathogens that can easily spread to other potential hosts. Among the many
parasites and pathogens of honey bees, there are three that are often trans-
mitted horizontally (between unrelated colonies), and all three have highly
virulent strains: Varroa destructor, American foulbrood, and the deformed
wing virus. It is not surprising that these are the agents of bee disease most
hated by beekeepers.
Horizontal transmission of honey bee diseases can occur when bee-
keepers move combs bearing infected bees and brood between colonies
and when bees rob honey from a colony that has been weakened by dis-
ease. I suspect, though, that the most common mechanism of disease trans-
mission between unrelated colonies within an apiary is drifting—adult
bees returning to the wrong hive by accident. The frequency of this mis-
take depends on how the hives are arranged in the apiary, and it can be
greatly reduced by increasing their spacing, painting them different colors,
and having them face different directions. A typical arrangement of hives,
however, is lined up in a straight row, spaced about 1 meter (ca. 3 feet)
Seeley.indb 259
2/21/2019 8:08:00 AM
260 Chapter 10
apart, painted the same color, and facing the same direction. In this situa-
tion, it is common for 40 percent or more of the bees to drift from their
natal colony to a neighboring colony. This much drifting means that patho-
gens and parasites spread easily between unrelated colonies.
When I started thinking about the implications for disease ecology of
the enormous difference in colony spacing between wild and manag
ed
colonies, I searched widely for articles by biologists and beekeepers
about the health effects of clustering colonies in apiaries, but I found es-
sentially nothing. So, I decided to perform an experiment that I hoped
would shed light on how colony spacing affects the challenge that honey
bee colonies face in avoiding infections with pathogens and parasites from
their neighbors.
I began, in June 2011, by setting up two groups of 12 small colonies in
one of Cornell University’s designated natural areas. The one I used is an
expanse of flat, brushy, abandoned farmland near a large beaver pond
north of Ithaca (Fig. 10.6). In one group, the colonies were clustered in an
apiary in which the hives were arranged in a row, with less than 1 meter
(ca. 3 feet) between neighboring colonies (Fig. 10.7). In the other group,
which was a few hundred meters from the first group, the colonies were
dispersed among clearings in and around a long field, so on average each
colony was 34 meters (110 feet) from its nearest neighbor.
In both groups, each colony occupied a hive that consisted of two full-
depth Langstroth hive bodies. To measure the levels of bee drift among
hives, I installed in 10 of the 12 colonies in each group a Golden Italian
queen. This is a queen that is homozygous for a mutant form (called
Cordovan) of the gene that provides instructions for making an enzyme
(tyrosine) involved in the synthesis of melanin, the substance that gives
honey bees their normal black or leathery- brown color. By installing a
Golden Italian queen in 10 of the 12 colonies, I insured that all the drones
produced in these 10 colonies would have a bright yellow coloration.
Meanwhile, I installed in the other two colonies in each group a Carniolan
queen, which ensured that all the drones produced in these two colonies
had a dark brown, or even black, coloration. In each group, the two colo-
nies with a Carniolan queen were placed in the center of their group.
Seeley.indb 260
2/21/2019 8:08:00 AM
Beaver pond
Mixed trees and shrubs
(beaver dam)
Apiary
0
50
100 m
Fig. 10.6. Map of the study site used in investigating the effects of crowding colo-
nies in apiaries on the mixing of bees and the spread of diseases among colonies.
The 12 hives of the colonies crowded in the apiary are indicated by the black bar
beneath the word “Apiary” on the left. The 12 hives of colonies dispersed around
the nearby field are indicated by black squares surrounded by circles.
Fig. 10.7. The 12 crowded colonies arranged in the apiary. The two hives with
bricks standing upright on top are the ones that had Carniolan queens and pro-
duced only dark drones.
Seeley.indb 261
2/21/2019 8:08:00 AM
262 Chapter 10
Every colony’s queen was labeled with a paint dot, so I could detect turn-
overs in queens (usually from swarming) when I made monthly inspections
of the colonies from May to October. The study colonies received no
treatments to control the Varroa mites throughout the two- year period of
the experiment, June 2011 to May 2013. The colonies started out as two-
frame nucleus colonies, so they all spent the summer of 2011 building up
strength, but they did not swarm. They all went into the winter of 2011–
2012 in good health.
Then, over the summer of 2012, I observed several striking differences
between the lives of the crowded and the dispersed colonies. First, when
7 out of 12 colonies in each group swarmed, the crowded colonies had
poorer success than the dispersed colonies in getting requeened after
swarming: only two successes out of seven vs. five out of seven, respec-
tively. I attribute most of this difference to the young queens in the
crowded group entering the wrong hive when they came home from
their mating flights; twice I found a dead queen lying outside the hive
entrance of a non- swarming colony in the crowded group. Second, there
was vastly more drifting of drones among the crowded colonies than
among the dispersed colonies. In counts made in September 2011 and
April 2012—before any colonies had swarmed, so when each had its
original queen, either Golden Italian or Carniolan—I found that 46 per-
cent (in September) and 56 percent (in April) of the drones flying into
the two black- bee (Carniolan) colonies in the crowded group were bright
yellow (Cordovan) drones! Counts made at the same time at the two
black- bee colonies in the dispersed group were much lower, just 1 per-
cent and 3 percent. It is likely that the difference between the two groups
in their levels of intercolony mixing of drones applies also to intercolony
mixing of workers. Third, late in the summer of 2012, when the colonies
were finishing their second summer, there were conspicuous surges in the
Varroa levels in the two crowded colonies that had swarmed and re-
queened but not in the five dispersed colonies that had swarmed and re-
queened. All of these seven colonies had low mite counts through June
and July, presumably an effect of their swarming in June, but this healthy,
Seeley.indb 262
2/21/2019 8:08:01 AM
Colony Defense 263
low- mite condition persisted into August in only the five dispersed colo-
nies. I cannot say for sure why the mite counts surged in August in the
crowded colonies, but I suspect that foragers from these two colonies
unintentionally brought home lots of mites while robbing honey from the
nearby colonies that were collapsing. This experiment ended in May 2013,
and my final inspection revealed that none of the 12 crowded colonies
were alive but that 5 of the 12 dispersed colonies were still alive. Indeed,
they were thriving.
This was a small- scale experiment, and it was performed just one time
and in just one place, so we cannot draw broad and rock- solid conclusions
from it, but I feel that it is a valuable step toward a better understanding
of how the challenge of colony defense differs between colonies living far
apart in the wild vs. close together under management.
LIVING IN SMALL VS. LARGE NEST CAVITIES
Packing colonies together in apiaries is not the only way that beekeepers
enlarge the problems of colony defense for managed colonies relative to
wild colonies. They also do so by keeping their colonies in hives that pro-
vide vastly more living space than the bees usually have in nature. We saw
in chapter 5 that wild colonies living in the woods around Ithaca occupy
tree cavities with volumes that are usually in the range of 30–60 liters (ca.
8–16 gallons), whereas most managed colonies in the United States are
housed (over summer) in hives with volumes of 120–160 liters (ca. 32–42
gallons) (see Fig. 5.3). Beekeepers give their colonies such roomy homes
so the bees have plenty of space for their honey stores. Roughly speaking,
giving a colony an additional 100 or so liters (ca. 26 gallons) of nesting
space enables it to store up an additional 50 kilograms (ca. 100 pounds) of
honey. Beekeepers also like to give their colonies roomy homes so that
their bees are less likely to become crowded and produce swarms. We have
seen in chapter 7 that when a colony casts a prime (first) swarm, it sheds
nearly 75 percent of its workforce (see Fig. 7.9), which is good for the bees
as they strive for colony reproduction but is disastrous for the beekeeper
who wants his or her bees to focus on honey production.
Seeley.indb 263
2/21/2019 8:08:01 AM
264 Chapter 10
Biologists have long recognized that housing honey bee colonies in large
hives makes perfect sense for the beekeeper but is utter nonsense for the
bees, as it hampers their reproduction. The arrival of Varroa destructor has
complicated this already awkward situation, because a big, strong colony
is likely to produce a big, strong crop of deadly Varroa mites, not just a
mammoth crop of honey. This has been demonstrated in a two- year ex-
periment that I conducted with two students, J. Carter Loftus and Michael
L. Smith, in which we compared two groups of 12 colonies that lived in
two apiaries that were spaced only 60 meters (200 feet) apart. In one
group, each colony occupied a small (42- liter/11.1- gallon) hive that con-
sisted of one full- depth, 10- frame Langstroth hive body. These were our
small- hive colonies; they simulated wild colonies of honey bees. In the
other group, our large- hive colonies, each colony occupied a large
(168- liter/44.4- gallon) hive and was managed in the usual ways that help
maximize a colony’s honey production. Each was given two full- depth hive
bodies for a brood chamber plus another two full- depth hive bodies (honey
supers) for honey storage. We also inspected these large- hive colonies bi-
monthly for queen cells—a sign of a colony preparing to swarm—and
removed all that we found, as does any beekeeper who desires a good
honey crop from his or her bees.
We set up these 24 colonies in May 2012 and then tracked them until
May 2014. Once a month, from May to October in 2012 and 2013, we took
measurements of each colony’s brood and adult bee populations, mite in-
festation level, presence of disease, and signs of swarming (queen turn-
over). We also recorded the annual honey production and survival of these
24 colonies over this two- year period. Every colony started out (in May