Bees in the Forest, Still 29
out and watch closely to see in which direction they fly. At this point you
wait, anxiously, for some of them to return to your little feeding station.
Usually some do, and if the plants in bloom at the time are providing only
meager forage, then your first customers will be excited by your syrup-
filled comb and will recruit hive mates to help them exploit its treasure.
After an hour or so, the bees will be familiar with the flight route between
your feeder and their home, and many will fly a direct course—a bee-
line—back to their nest. At this point, you determine the direction to
their home by measuring their vanishing bearings with a magnetic com-
pass, and you estimate the distance to their home by measuring the round-
trip times of a half dozen bees that you have labeled with paint marks. If
your bees need only two to three minutes to fly home, unload, and fly back
to you, then you know that their nest is only about 100 meters (330 feet)
away, but if they are gone for six or seven minutes, then it is probably
about 1 kilometer (0.6 miles) away. Now you will want to move your
whole operation down the beeline. To do so, you trap in your bee box as
many of the bees as possible, then you move your gear 100–200 meters
(ca. 300–600 feet) down the beeline to another clearing, and here you
release your bees. Now you again note the bees’ vanishing bearings, to
check that you are moving in the right direction, and you again record
their round- trip times, to update your estimate of the distance. By pa-
tiently making a series of moves down the beeline, you will find your way
to the stand of trees in which the bees reside, then to the one tree that is
their dwelling place, and ultimately to the knothole or fissure that is the
entryway to their home. Discovering it is always a huge thrill!
Kirk and I began our survey of the wild colonies in the Arnot Forest by
driving to a small clearing near its center and searching there for flowers
being visited by honey bees. We had difficulty finding even one honey bee
in this spot, but eventually Kirk spied a bee on a multiflora rose ( Rosa
multiflora) in full bloom, and he managed to capture her in his bee box. He
then slid into the bee box, without releasing the bee, a small square of
beeswax comb filled with anise- scented sugar syrup. The bee loaded up on
this bait and upon release flew off to the east, which revealed to us the
Seeley.indb 29
2/21/2019 8:07:15 AM
30 Chapter 2
Fig. 2.5. Crowd of forager bees recruited to help exploit a square of comb filled
with sugar syrup, at the start of a hunt for a wild colony’s home.
general direction to her home. Nine minutes and 20 seconds later, she
arrived back at our feeding station and landed on the comb for a refill. As
she calmly drank in more of our syrup, Kirk applied a dot of green paint
to her abdomen, so we could identify her. Green- abdomen became a
steady visitor, and after an hour or so she had recruited several dozen nest
mates to help exploit the amazing food source we were offering (Fig. 2.5).
Meanwhile, Kirk and I had labeled about 10 of the recruits and had mea-
sured the bearings of their beelines home: nearly due east (see the red line
in Fig. 2.6). We then began moving our feeding station, together with the
bees, in a series of steps down their flight path home. Each move took
more than an hour and brought us only about 100–200 meters (330–660
Seeley.indb 30
2/21/2019 8:07:16 AM
Bees in the Forest, Still 31
feet) closer to the bees’ residence, but we were determined to find their
bee- tree home, so we persisted. By carefully noting at each stopping place
the vanishing bearings of the departing bees, and moving ahead in this
direction, we managed to zero in on the entryway of their residence: a
knothole about 6 meters (20 feet) up in an eastern hemlock tree some 800
meters (0.5 miles) east of our starting point.
Because we had difficulty finding honey bees on flowers in the Arnot
Forest in early July, Kirk and I postponed further bee hunting until late
August. We knew that by then the endless stands of goldenrod plants lining
the roadsides and filling the clearings in the forest would be in bloom, at-
tracting droves of foraging bees. This should make it easy to find the honey
bee foragers needed to establish beelines leading to the other colonies
living in this forest.
When I returned to the Arnot Forest on 26 August 1978, to begin sev-
eral weeks of intense bee hunting, I did indeed find seas of goldenrod
plants (mostly Solidago canadensis) in bloom, and they were alive with
honey bees bobbing on their brilliant yellow inflorescences. Beautiful! Fig-
uring that I could not survey the whole Arnot Forest in the three weeks I
had available, I decided to focus my search on the forest’s southern and
western sectors. I chose these places because there is a narrow, flat-
bottomed valley below the forest’s southern and western boundaries (see
Fig. 2.6), and within it there were abandoned pastures and a derelict rail-
road track brimming with patches of goldenrod. It was delightfully easy
to find foraging honey bees here. Working in this valley, I also found it
remarkably easy to get readings of the bees’ vanishing bearings, because
after they lifted off from my feeding station, laden with heavy payloads of
thick syrup, they would fly slowly as they struggled up the steep hillsides
en route to their homes high in the woods. Figure 2.6 shows that these
homeward flights by the bees guided me to nine more wild colonies, eight
within the Arnot Forest and one just outside its western boundary. And
one more thing delighted me about this bee hunting: it hadn’t led to any
beekeepers’ colonies! It was becoming clear that there were only wild
colonies living in and around the Arnot Forest.
Seeley.indb 31
2/21/2019 8:07:16 AM
1400
Newfield State Forest
18
1
0
7
0
1300
00
2
1600
1500
1800
1900
1500 3
1800
1700
1950
1600
Recknagel
1800
1700
Hill
TOMPKINS CO
SCHUYLER CO 1800
1800
Ban
1700
field
4
Cre 1 e 5
1950
k 0
1700
0
1500
170
1600
1800
0
1
140
1
0
600
1300
1600
Irish
6
Hill
5
1950
1900
1300
7
1800
McClary Road
1700
10
1600
1800
8
9
12
17
0
1400
00
0
1700
SCHUYLER CO CHEMUNG CO
1600
Jackson Creek
Cayuta
1200
1400
Creek
11
1
0
20
140
1500
0
0
0
1200
1300
1200
railroad
0.5 km
1
waterways
0.5 mile
trails
bee
buildings
1650
contour interval
tree
roads
50 feet
Fig. 2.6. Map of the Arnot Forest showing the locations of the 10 bee trees found
there in 1978. The site of each bee tree is marked by the base of a bee- tree sym-
bol. Red line denotes the path of the author’s first bee hunt.
Seeley.indb 32
2/21/2019 8:07:16 AM
Bees in the Forest, Still 33
I knew, of course, that the nine bee- tree colonies that Kirk and I found
in this forest were not all the wild colonies living in it. After all, no bee-
lines had been followed from flower patches in the northern and eastern
regions of the forest, so about half of the Arnot Forest was terra incognita.
Moreover, I could not even be confident that I had located all the colonies
in the southern and western parts of the forest. I concluded, therefore,
that the nine colonies that had been found were at most about one half of
all the colonies residing in this forest; hence there were 18 or more colo-
nies living in this 17- square- kilometer (6.6- square- mile) forest. So, I fig-
ured that the density of wild colonies living in the Arnot Forest in Septem-
ber 1978 was at least one colony per square kilometer, hence 2.5 or more
per square mile.
HOW ABUNDANT ARE WILD COLONIES
OF HONEY BEES ELSEWHERE?
Building on the 1978 study of the density of honey bee colonies living
within the Arnot Forest, other biologists have investigated this matter at
various sites in North America, Europe, and Australia. The first of these
additional studies was led by Roger A. Morse, the entomology professor
at Cornell University who generously let me start working in his honey
bee laboratory when I was still a high school student back in 1969. He and
a team of seven graduate students conducted their study in the spring of
1990, in the small port city of Oswego, on Lake Ontario in northern New
York State. Their investigation was triggered by the discovery of a colony
of Africanized honey bees—a hybrid between European subspecies and
the African subspecies A. m. scutellata (discussed below)—nesting in a ship-
ment of pipes from Brazil. The presence of these exotic honey bees raised
concerns that Africanized bees, and the fearsome ectoparasitic mite ( Varroa
destructor) that these bees could carry, might have been introduced to
North America, so attempts were made to locate all the honey bee colo-
nies living near the port so they could be checked for Africanized bees and
Varroa mites. Newspaper and radio advertisements were run offering a $35
reward for information on honey bee colonies living in the semicircular
Seeley.indb 33
2/21/2019 8:07:16 AM
34 Chapter 2
area within 1.6 kilometers (1 mile) of the port. Eleven wild colonies living
in trees and buildings, and one managed colony residing in a backyard
beehive, were found. This work revealed that in this small city, the density
of the wild colonies was 2.7 colonies per square kilometer (7 colonies per
square mile), much higher than what Kirk and I had found in the woods of
the Arnot Forest. Fortunately, no Africanized honey bees or Varroa destruc-
tor mites were found.
A still higher density of wild colonies was found in a remarkable study
conducted by a team of biologists led by M. Alice Pinto at Texas A&M
University in 1991–2001. This group worked in the Welder Wildlife Ref-
uge, a 31.2-square- kilometer (12.2-square- mile) nature preserve in south-
ern Texas. Their aim was to track the “Africanization” of a population of
wild honey bees living in the southern United States, and they did so by
sampling the colonies living in this wildlife refuge before, during, and after
the arrival of Africanized honey bees from Mexico. Africanized honey bees
are derived from a founder population of an African subspecies, A. m. scutel-
lata, that was introduced to Brazil from South Africa in 1956. The purpose
of this introduction was to crossbreed a tropical- evolved African subspe-
cies with several temperate- evolved European subspecies already in Brazil
to create a honey bee well suited to tropical conditions. However, several
colonies of A. m. scutellata escaped from the quarantine apiary, thrived in
the Brazilian climate, and spawned strong populations of wild colonies of
this subspecies throughout the American tropics.
The vegetation in the Welder Wildlife Refuge is a mix of open grassland,
chaparral brushland, scattered mesquite trees ( Prosopis spp.), and groves
of live oaks ( Quercus virginiana) (Fig. 2.7). Several times a year, for 11 years
straight, a team of biologists from Texas A&M University searched a
6.25- square- kilometer (2.4- square- mile) study area within the refuge for
wild colonies of honey bees, and they collected samples of worker bees
from each colony they found. Nesting cavities were abundant in the wood-
land areas; nearly all (85%) of the colonies were found in cavities in oak
trees. When the mitochondrial DNA of these bees was analyzed to deter-
mine their maternal ancestry, it became clear that for the first three years
Seeley.indb 34
2/21/2019 8:07:16 AM
Bees in the Forest, Still 35
Fig. 2.7. Researcher in a grove of live
oaks ( Quercus virginiana) within the
Welder Wildlife Refuge. He has just
sampled workers from the colony of
Africanized honey bees nesting inside
the tree directly behind him. The
nest’s entrance is visible just above the
top of the insect net.
of the study (1991–1993) the queen bees living in the study area had been
mainly descendants of several European subspecies: 68 percent A. m. ligus-
tica and A. m. carnica (both from southern Europe), 26 percent A. m. mel-
lifera (from northern Europe), and 6 percent A. m. lamarckii (from northern
Africa). Over the next several years, however, the queen bees living here
became primarily descendants of the southern African subspecies, A. m.
scutellata. And what did the surveys of the colonies living in the study area
reveal about the colony density during the first four years, when the popu-
lation was dominated by European honey bee colonies? They showed that
the density of the wild colonies in this mixture of grassland, brushland,
and woodland habitats was remarkably high: 9–10 colonies per square
Seeley.indb 35
2/21/2019 8:07:17 AM
36 Chapter 2
In Europe, where Apis mellifera is a native species, three teams of re-
searchers have investigated the abundance of wild colonies. One team
in Poland, led by Andrzej Oleksa at the Kazimierz Wielki University in
Bydgoszcz, studied a population of wild colonies living in the lowlands of
northern Poland, just south of the Baltic Sea. The landscape here is domi-
nated (68%) by agricultural areas—cultivated fields, meadows, and or-
chards—and the rest is covered mostly (27%) by forests. The population
of honey bees living in this region of Poland still consists primarily of A. m.
mellifera, the native dark honey bee of northern Europe.
Andrzej Oleksa and his colleagues focused on assessing the occurrence
of wild colonies in rural avenues—linear stands of old trees along coun-
tryside roads, an example of which is shown in Figure 2.8. These research-
ers inspected 15,115 large trees in 201 avenues that were carefully chosen
to provide uniform coverage of their 15,000- square- kilometer (6,000-
square- mile) study area. In total, they searched along 142 kilometers (88
miles) of avenue fragments and found 45 colonies of honey bees, which
indicated a density of 0.32 colonies per kilometer (0.51 colonies per mile)
of avenue. Knowing the density of avenues on the landscape, they esti-
mated the overall density of wild colonies living in avenue trees at 0.10
colonies per square kilometer (0.26 colonies per square mile). These re-
searchers point out that their estimate is surely an underestimate of the
total abundance of wild colonies, because it did not take account of wild
colonies present in the woodlands, which cover 27 percent of their study
area. They also note that some colonies nesting high in the avenue trees
could have been overlooked. Nevertheless, their estimate of wild- colony
density is valuable, for it reveals that rural avenues are serving as a refuge
for wild colonies. Moreover, it shows that wild colonies of honey bees still
exist in a place where the natural environment has been largely replaced
with agriculture and where beekeeping is extremely popular. Beekeepers
maintain some 4.4 colonies per square kilometer (11.4 per square mile)
in the region of Poland where this study was conducted.
Just to the west of Poland, in Germany, Robin Moritz and colleagues at
the University of Halle have investigated the abundance of honey bee colo-
The Lives of Bees Page 5
