Seeley.indb 36
2/21/2019 8:07:17 AM
Bees in the Forest, Still 37
Fig. 2.8. Rural avenue in northern Poland lined with a mixture of European ash
( Fraxinus excelsior) and Norway maple ( Acer platanoides) trees.
nies living in natural areas. This group worked in three widely spaced sites
spread north to south within Germany. Two sites were in national parks:
the 318- square- kilometer (123- square- mile) Müritz National Park, in the
Müritz lake region north of Berlin, and the 25- square- kilometer (10- square-
mile) Harz National Park, in the forested Harz Mountains in central Ger-
many. The third site was a rural spot in Bavaria (southern Germany) west
of Munich. These researchers did not adopt the direct, brute- force ap-
proach of exhaustively searching for the colonies living in their three study
sites. They took instead an indirect approach based on genetic analyses.
Specifically, they had approximately 10 virgin queen honey bees conduct
their mating flights within each study site, and then they analyzed the
genes of each queen’s worker offspring to determine how many colonies
had produced the drones that mated with each queen. The beauty of this
Seeley.indb 37
2/21/2019 8:07:17 AM
38 Chapter 2
approach is that it made use of the astonishing promiscuity of queen honey
bees to get a strong sampling of the drones in the study area; each queen
mates with 10–15 males, so the worker offspring of the 10 queens pro-
vided a sampling of 100–150 drones from the colonies living in the area
where these queens mated. The genetic analysis revealed that there were
24–32 colonies producing drones in the three locations. If we assume that
both queens and drones fly 900 meters (0.56 miles), on average, to reach
a mating site, then the area over which these 24–32 colonies were dis-
persed was that of a circle with a radius of 1.8 kilometers (1.1 miles) and
an area of 10.2 square kilometers (3.9 square miles). Therefore, the esti-
mated average density of colonies for the three study locations was 2.4 to
3.2 colonies per square kilometer (6.2 to 8.2 colonies per square mile).
It must be noted, however, that in Germany beekeepers sometimes can
(and do) place their colonies inside the country’s national parks, so the
estimates of colony density reported in this study represent the combined
densities of managed and wild colonies.
Recently, another team of biologists in Germany, Patrick L. Kohl and
Benjamin Rutschmann, graduate students at the University of Würzburg,
have made surveys of wild colonies in largely undisturbed European beech
( Fagus sylvatica) forests in central and southwestern Germany: the 160-
square- kilometer (62- square- mile) Hainrich forest in Thuringia, and sev-
eral forest clusters in the 850- square- kilometer (328- square- mile) bio-
sphere reserve in the Swabian Alb mountain range in Baden- Württemberg.
The Hainrich forest site is off- limits to beekeepers seeking locations for
their hives. In this forest, the researchers used the same method that Kirk
Visscher and I used in the Arnot Forest (beelining), but in the forest clus-
ters in the Swabian Alb region they inspected 98 known nest cavities of the
black woodpecker ( Dryocopus martius). This is the largest woodpecker in
northern Europe, and it excavates nest cavities that are spacious enough
(20- plus liters/5.3- plus gallons) to be suitable nesting sites for honey
bees. In the Hainrich forest, the team found or inferred the locations of
nine wild colonies, from which it calculated an estimate of the wild- colony
Seeley.indb 38
2/21/2019 8:07:17 AM
Bees in the Forest, Still 39
density here: 0.13 colonies per square kilometer (0.34 colonies per square
mile). In the forests in the Swabian Alb Biosphere Reserve, they inspected
98 beech trees with old woodpecker nest cavities and found seven occu-
pied by honey bees. Knowing the density of the woodpecker- cavity trees
in the region, they calculated an estimate of the wild- colony density in the
Swabian Alb forest patches of at least 0.11 colonies per square kilometer
(0.28 colonies per square mile). Of course, both figures are minimum
estimates of the actual colony density, because the investigators could not
be certain they had found all the colonies living within their search areas.
Biologists in Australia, where European honey bees have lived as an
introduced species since 1822, have also used national parks to measure
the density of honey bee colonies living in undisturbed habitats. This team,
led by Benjamin Oldroyd of the University of Sydney, worked in the
755- square- kilometer (292- square- mile) Barrington Tops National Park,
the 83- square- kilometer (32- square- mile) Weddin Mountains National
Park, and the immense, 3,570- square- kilometer (1,378- square- mile) Wy-
perfeld National Park. All three parks are in southeastern Australia, but
their vegetation types range from subtropical rain forest to semiarid Euca-
lyptus woodland. Like the German investigators, the Australian biologists
used an indirect, genetic approach that produced estimates of the density
of the wild colonies at their three study sites. Here again, the goal was to
determine how many colonies produced the drones present in a given
location. The Australian researchers, however, estimated this based on the
genetics of drones captured in drone traps suspended from helium- filled
weather balloons in drone congregation areas (see chapter 7), rather than
the genetics of workers sired by drones residing within a study area. They
captured drones in two drone congregation areas in each national park,
and from the genetics of these bees they calculated estimates of colony
density that ranged from 0.4 to 1.5 colonies per square kilometer (1.0 to
3.9 colonies per square mile). Beekeeping is banned within the large na-
tional parks used by these Australian investigators, so their results are not
influenced by the presence of managed colonies.
Seeley.indb 39
2/21/2019 8:07:17 AM
40 Chapter 2
Fig. 2.9. Left: Eastern hemlock ( Tsuga ca-
nadensis) bee tree in the Shindagin Hollow
State Forest. Right: Close- up of the nest
entrance. Photo taken in late November,
when the colony was quiet inside its nest.
Seeley.indb 40
2/21/2019 8:07:19 AM
Bees in the Forest, Still 41
Table 2.1. Estimates of the density of wild colonies of European honey bees
Study location
Colonies/sq. km
Colonies/sq. mile
Arnot Forest, NY (USA)
1.0+
2.5+
Oswego, NY (USA)
2.7
6.9
Welder Wildlife Refuge, TX (USA)
9.3– 10.1
23.8– 25.8
Shindagin Hollow State Forest (USA)
1.0
2.5
Agricultural land (Poland)
0.1+
0.26+
Two national parks and a forest (Germany)
2.4 - 3.2
6.2– 8.2
Hainrich Forest and Swa
bian Alb (Germany)
0.1+
0.3+
Three national parks (Australia)
0.4– 1.5
1.0– 3.9
The most recent study of the density of wild colonies in undisturbed
forest habitat comes from a survey conducted in 2017 by Robin Radcliffe,
who thoroughly searched—by means of bee hunting—a 5- square- kilometer
(1.9-square- mile) portion of the Shindagin Hollow State Forest in New
York State. This is a 21- square- kilometer (8-square- mile) area of old-
growth forest southeast of Ithaca and about 30 kilometers (20 miles) east
of the Arnot Forest (Fig. 2.9). Robin located five colonies, which yields an
estimate of one colony per square kilometer (2.5 colonies per square
mile), virtually identical to what Kirk Visscher and I found in the Arnot
Forest in 1978 and what I have found there in later surveys conducted in
2002 (see below, this chapter) and in 2011 (to be discussed in chapter 10).
In review: What have we learned about the abundance of wild colonies
of European honey bees? We see in Table 2.1, which summarizes the find-
ings of the studies just described, that the density of colonies living in natu-
ral areas varies greatly, but that generally it is estimated to be somewhere
in the range of one to three colonies per square kilometer (2.6–7.8 colo-
nies per square mile). Evidently, the density of wild colonies is rather low,
which means that on average the honey bees’ nests are widely spaced. In
the Arnot Forest, for example, the average distance between each colony
and its nearest known neighbor is 0.87 kilometers (0.54 miles). Clearly,
wild colonies living in natural settings are generally spaced far more widely
Seeley.indb 41
2/21/2019 8:07:19 AM
42 Chapter 2
than managed colonies living in apiaries. We will see in chapter 10 that this
difference in colony spacing can have enormous consequences for the
health of the bees.
HAVE THE WILD COLONIES BEEN KILLED
OFF BY VARROA DESTRUCTOR?
Varroa destructor is a small but dangerous mite that parasitizes honey bees.
A full- grown female of this species is no bigger than the head of a pin, so
it is just a small fry compared to a worker bee (Fig. 2.10). Nevertheless, a
heavy infestation of these tiny, reddish- brown, disk- shaped creatures can
kill a colony of honey bees. What makes these mites so deadly is that they
feed on the fat body (energy storage) tissue in immature bees (larvae and
pupae), and this feeding spreads viruses that can weaken the developing
bees and, still worse, can induce crippling physical defects such as shrunken
abdomens and shriveled wings. The latter is symptomatic of high levels of
the deformed wing virus, probably the most damaging of all the pathogens
for which these mites are vectors. The deadliness of V. destructor is amplified
by the rapid development of individuals; the mites grow from egg to adult
in less than a week. This means that even if the population of Varroa mites
infesting a colony starts out small in spring, it can, through the magic of
exponential growth, explode by late summer. And when a colony has a
high mite load, the worker bees it produces start their lives with such high
virus titers that they are too sick to work. Eventually, the colony’s pop ula-
tion collapses for lack of sufficient gains of vigorous young bees to offset
the losses of older bees dying from predation, accidents they suffer while
working outside the hive, and the accelerated senescence that comes from
being parasitized by Varroa mites.
Varroa mites originally lived only in the mainland regions of eastern
Asia. There they still have a stable host- parasite relationship with their
orig inal host, the eastern hive honey bee, Apis cerana, whose range is all of
Asia east of the vast deserts in Iran and south of the towering mountains
of central Asia. Unfortunately, these mites made a successful host shift
from Apis cerana to the western hive honey bee, Apis mellifera, and since
Seeley.indb 42
2/21/2019 8:07:19 AM
Bees in the Forest, Still 43
Fig. 2.10. Adult female of the mite Varroa destructor on a worker honey bee.
doing so they have spread, and today their distribution is nearly world-
wide, like that of their new host, A. mellifera. This host shift occurred some-
time in the early 1900s, after beekeepers had moved colonies of A. mellifera
from west ern Russia and Ukraine to Primorsky, the easternmost region of
Russia. This region is far outside the native range of A. mellifera but is within
that of A. cerana. Once the range of A. mellifera overlapped that of A. cerana, the mite began infesting colonies of A. mellifera. This perhaps began when
colonies of Apis mellifera robbed honey from colonies of Apis cerana, but it
may have been unwittingly fostered when beekeepers tried to strengthen
their colonies of A. mellifera by giving them brood (larvae and pupae) from
colonies of A. cerana. Russian beekeepers then spread V. destructor to Europe
in the 1950s or 1960s when they shipped queens of A. mellifera, carrying
the mites, from the Primorsky region to western parts of the Soviet Union.
Varroa mites were first reported in A. mellifera colonies living in Europe in
Seeley.indb 43
2/21/2019 8:07:19 AM
44 Chapter 2
1967 in Bulgaria, in 1971 in Germany, and in 1975 in Romania. These
mites also spread to northern Africa when hundreds of honey bee colonies
were sent from Romania and Bulgaria to Tunisia and Libya in 1975 and
1976, as parts of foreign aid programs. Evidently, V. destructor reached
South America when Japanese beekeepers moved colonies of A. mellifera
carrying V. destructor (acquired in Japan) into Paraguay in 1971. Somehow
these mites also reached Brazil in 1972.
Varroa destructor entered North America relatively recently and via two
routes. The first was probably into Florida, where it is likely that in the
mid- 1980s either a beekeeper smuggled in some Brazilian queen bees that
were carrying the mites or a swarm of Africanized honey bees (infested
with Varroa mites) came off a cargo ship. Records from the Bureau of Plant
and Apiary Inspection for the state of Florida show that between 1983 and
1989 swarms of Africanized honey bees were found on eight ships—usu-
ally in shipping containers—that had arrived from Central or South Amer-
ica. These records indicate that in at least one ship, the “bees were varroa
infested.” The second pathway was in Texas, where in the early 1990s
swarms of Africanized honey bees infested with Varroa mites flew north
across the U.S.–Mexico border.
My first encounter with a Varroa mite came in June 1994 at my labora-
tory in Ithaca. I remember the moment vividly: I was labeling young
worker bees with paint marks in preparation for an experiment when, to
my amazement, I spied a Varroa mite scuttling across the thorax of the bee
I was about to label with a dot of yellow paint. Sighting this mite was dis-
maying, for I knew that V. destructor is a deadly parasite, but I hoped that it
marked just the arrival of these mites at my laborat
ory. In late August,
however, I observed a chilling scene that made it clear the mites had ar-
rived well before 1994 and that my laboratory’s colonies were now heavily
infested: dozens and dozens of worker bees, many with shriveled wings,
were crawling feebly through the grass in front of my hives. Clearly, many
of the worker bees in my colonies were ill, but still I hoped for the best.
This seemed like a realistic hope at the time, for when I inspected the 19
colonies in my laboratory’s apiary a month later, in September 1994, I
Seeley.indb 44
2/21/2019 8:07:19 AM
Bees in the Forest, Still 45
found them all well populated with worker bees and possessing large
honey stores, which suggested that all were in good shape for surviving
the coming winter. Several months later, though, I learned how badly Var-
roa mites can rob honey bees of their health. By April 1995, only two of
the 19 colonies remained alive. The experience of 89 percent colony mor-
tality over the winter of 1994–1995 was a chilling lesson on the fearful
virulence of the Varroa mite.
There seemed only one way to help the bees: treat them with miticides.
I started doing so in the summer of 1995, using fluvalinate (Apistan), and
now, some 20 years later, my students and I treat many of the colonies that
we keep in our apiaries with mite- killing chemicals, usually formic acid or
thymol- based medications. We do so because for many of our experiments
we need colonies that are not stressed by heavy infestations of Varroa
destructor.
Back in the mid- 1990s, when my students and I were learning how to
control the Varroa mites infesting the managed colonies living in our apiar-
ies, I was also fretting about what was happening to the wild colonies living
in the woods around Ithaca. Certainly, nobody was treating them with the
life- saving miticides, so I wondered: Were they dying out? Or were they
already all gone? That they had been wiped out seemed possible, maybe
even likely. And even if a few were still alive, I feared that any such survivor
colonies might be merely the issue of a beekeeper’s miticide- treated col-
ony that had cast a swarm. If so, then those still alive were probably just
short- lived colonies doomed by the mites spreading the deadly viruses.
The Lives of Bees Page 6
