5
THE NEST
He must be a dull man who can examine the
exquisite structure of a comb,
so beautifully adapted to its end, without
enthusiastic admiration.
—Charles Darwin, On the Origin of Species, 1859
The aim of this book is to review what we know about how honey bee
colonies live in the wild, so naturally it focuses mainly on the bees them-
selves. We must, however, also give close attention to the nests that the
bees build. This is because it is the qualities of a colony’s inert nest, as
much as the abilities of its lively bees, that determines how long a colony
survives, how much it reproduces, and thus how well it achieves genetic
success. We should expect, therefore, that when a colony builds a nest,
its workers follow genetically based rules of behavior that guide them to
find a good nesting site, secrete beeswax in a timely manner, skillfully
build the hexagonal- celled beeswax combs, and laboriously coat their
nest’s walls with antiseptic tree resins. We will see in this chapter that
the nest of a honey bee colony contributes hugely to its survival and
reproduction.
By looking at the nest of a wild colony as a survival tool that extends
beyond the bees’ own bodies, we will become aware that beekeepers risk
disrupting the adaptive biology of their bees by housing them in movable-
frame hives that are crowded together in apiaries. We shall see that when
Seeley.indb 99
2/21/2019 8:07:30 AM
100 Chapter 5
colonies live under a beekeeper’s supervision in wooden boxes, rather
than on their own in hollow trees, they are forced to cope with homes that
are often jumbo- sized, poorly insulated, and closely spaced. Recognizing
these alterations of the bees’ natural housing conditions, and their effects
on colony health and biological fitness, raises many questions about opti-
mal hive design and best management practices in beekeeping. It also high-
lights some fundamental conflicts of interest between the bees and their
keepers.
NATURAL NESTS IN TREES
Most of what we know about the nests of honey bee colonies living in the
wild comes from a study that I made in the mid- 1970s with my first sci-
entific mentor, Professor Roger A. Morse, then the professor of apiculture
at Cornell University. Back then, I was in my early 20s and was starting to
investigate what a swarm of honey bees seeks when it chooses its future
dwelling place. My first goal was to describe the natural nests of honey
bees, because I figured that doing so would provide clues about what con-
stitutes a dream homesite for a honey bee colony. Then, once I knew the
typical properties of their natural nesting sites—cavity volume, entrance
size, entrance height, and so forth—I would conduct experiments to de-
termine whether what I had found for each property of the bees’ natural
nests (e.g., the typical size of the entrance opening) was an expression of
the bees’ nest- site preferences or merely an indication of what was avail-
able to the bees.
Professor Morse and I decided that we would try to describe at least 20
natural nests. I knew already, from roaming in the woods around my par-
ents’ home outside Ithaca, the whereabouts of three bee trees. We located
several dozen more by running an advertisement in the local newspaper,
the Ithaca Journal. It read “bee trees wanted. Will pay $15 or 15 lb. of honey
for a tree housing a live colony of honey bees.” This worked surprisingly
well, and in less than 10 days we had a list of 36 bee trees, all of them
within 15 miles of Ithaca. Great! The next step was to work with a techni-
cian in the Entomology Department at Cornell, Herb Nelson, who had
Seeley.indb 100
2/21/2019 8:07:30 AM
The Nest 101
worked as a logger in the Maine woods and could help me fell these bee
trees safely. Most were massive, old individuals. We began by conducting
reconnaissance at each bee tree, to measure things like the height and di-
rection of the nest entrance, but mainly to see if we could get Roger
Morse’s pickup truck close to the tree. Truck access to a bee tree’s location
was essential, because to dissect a nest properly, we would need to load
the section of tree trunk housing the nest onto the truck for transport back
to the Dyce Laboratory for Honey Bee Studies at Cornell, where I could
slowly and carefully and comfortably dissect the nest. Twenty- one of the
36 bee trees allowed us a close enough approach with the truck, so we
managed to collect 21 natural nests. To do so, Herb and I felled each tree,
sawed out the portion of its trunk that contained the bees’ nest, wrestled
it into the truck’s bed, and hauled it back to the lab. There I would carefully
split open the log to expose the nest (Fig. 5.1) and start my dissection. The
other 15 bee trees were too deep in the woods to allow us to collect their
nests, but at 12 of them we managed to measure several key features—
height, area, and compass direction—of the entryways to the honey bee
homes that they harbored.
The entrances of the 33 bee- tree nests that we studied showed several
striking features. First, most (79%) consisted of just one opening (e.g., see
Figs. 1.1, 2.9, and 5.1); the others had two to five. Second, these entrance
openings were usually knotholes (56%), but sometimes they were fissures
in the tree’s trunk (32%) or gaps among its roots (12%). Third, they
tended to face in a southerly direction (23 nests) rather than a northerly
one (10 nests). Fourth, most were in the bottom third of the nest cavity
(58%) rather than the middle (18%) or top third (24%). And fifth, the
average size of the nests’ entrances was rather small, only 29 square cen-
timeters (4.5 square inches), and the most common size was just 10–20
square centimeters (1.5–3.0 square inches) (see Fig. 5.2, top). In com-
parison, the standard entrance opening of a Langstroth hive is several
times larger, approximately 75 square centimeters (almost 12 square
inches). We began to wonder, do wild colonies prefer rather small, easily
guarded entrance openings?
Seeley.indb 101
2/21/2019 8:07:30 AM
102 Chapter 5
Fig. 5.1. The first of the 21 bee- tree
nests that were dissected in 1975. Left:
The intact tree, with the knothole that
served as the colony’s nest entrance
visible in the left fork. Right: The section
of the tree housing the nest has been
brought to the laboratory and carefully
split open to reveal the nest inside. The
combs containing honey (in cells cov-
ered with yellow cappings) are in the
top, and those containing brood— eggs,
white larvae, and pupae (in cells with
brown cappings)— are below. The nest
entrance is on the left side, about two-
thirds of the way up the cavity.
Seeley.indb 102
2/21/2019 8:07:32 AM
The Nest 103
The results regarding entrance height for the b
ee- tree nests proved
especially illuminating. This is because what we learned about this nest- site
property from the 33 bee trees sampled in the mid- 1970s was contra-
dicted—and, as it turns out, corrected—by some later work. The bottom
portion of Figure 5.2 shows that the 49 entrance openings of the 33 bee-
tree nests that we examined in the mid- 1970s were mostly low on the tree.
About half were less than 1 meter (ca. 3 feet) off the ground, though there
were also others more than 5 meters (16.4 feet) up. Given these results,
I thought, OK, honey bees living wild in tree cavities generally have low
nest entrances, just as they do when they live in beekeepers’ manufactured
hives. This seemed to be a reasonable conclusion at the time, but since then
I have learned that it was dead wrong.
What corrected my thinking about the entrance heights of colonies
living in the wild is what I discovered bit by bit—between 1978 and
2011—when I made three surveys of the wild colonies living in the Arnot
Forest. Whenever I found a bee tree in one of these surveys, I did my best
to locate the entrance of the bees’ home and to measure its height, and
most times I succeeded. I measured entrance heights by either climbing
the bee tree and extending a tape measure or using a forester’s clinom-
eter. There were only two colonies possessed of entrance openings so well
hidden in the bee tree’s crown that I could not spot them, even with
powerful binoculars. Figure 5.2 shows that the 21 entrance openings of
the nests that I found in these Arnot Forest surveys were all well above
ground level. The lowest was 4 meters (13 feet) up, and 90 percent were
more than 5 meters (16.4 feet) high. Moreover, only one of these Arnot
Forest nests had more than one entrance opening, and it had just two.
Incidentally, this pattern of wild colonies usually having nest entrances
high above ground level was recently confirmed by a friend, Robin Rad-
cliffe, when he conducted a survey of the wild colonies living in a
5- square- kilometer (ca. 2- square- mile) portion of the Shindagin Hollow
State Forest that is adjacent to his farm (see chapter 2). Robin beelined
his way to five colonies, one in the wall of a hunter’s cabin just outside
the forest, and four in old- growth eastern hemlock trees inside the forest.
Seeley.indb 103
2/21/2019 8:07:32 AM
104 Chapter 5
4
Langstroth hive
Number of nests 0 0
20
40
60
80
100
Entrance area (cm2)
4
1978, 2003, & 2011
by bee hunting
0
20
mid-1970’s
by chance
8
Number of entrance openings 4
0 0
5
10
15
20
Entrance height (m)
Fig. 5.2. Top: Distribution of entrance areas for 32 bee- tree nests. Not shown is
the 204-square- centimeter (31.6-square- inch) value for one nest. Bottom: Distri-
butions of entrance opening heights for 33 bee- tree nests found by chance (in the
mid- 1970s, n = 49 openings) and for 20 bee- tree nests found by bee hunting (in
later years, n = 21 openings).
The average height of the entrances of Robin’s four bee- tree colonies is
9.5 meters (31.2 feet).
Why is there such a striking difference between the two sets of entrance-
height results shown in Figure 5.2? Looking back, I now recognize that my
mid- 1970s sample of bee- tree nests was unintentionally biased in favor of
Seeley.indb 104
2/21/2019 8:07:32 AM
The Nest 105
nests that had low entrances. Such nests, relative to ones whose entrances
are up in the leafy crowns of trees, are much more likely to be found by
chance, which is how the nests studied in the mid- 1970s were discovered
by the farmers, villagers, and forest owners who responded to our adver-
tisement in the Ithaca Journal. In contrast, when I located bee trees in the
Arnot Forest in 1978, 2002, and 2011, I avoided this sampling bias. I dis-
covered these nests not by noticing them inadvertently but by intentionally
tracking them down using the methods of a bee hunter—that is, by letting
foragers from these nests lead me to their homes wherever they might be.
Success in finding wild colonies by bee hunting is biased very little, if at
all, by nest- entrance height. The stark difference between the two distribu-
tions of entrance height presented in Figure 5.2 taught me an important
general lesson about studying the bees: always be on guard against uninten-
tional sampling biases. More specifically, the data presented in Figure 5.2
show us that wild colonies living out in the woods around Ithaca usually
reside in tree- cavity homes whose entryways are high up and therefore
well hidden from ground- dwelling animals, most importantly (as we shall
see in chapter 10) from black bears.
When Roger Morse and I shifted our attention from the exteriors of
our bee- tree nests to their interiors, we were not surprised to see that the
tree cavities occupied by the bees were generally tall and cylindrical: on
average, 156 centimeters (62 inches) tall and 23 centimeters (9 inches) in
diameter. This is consistent with the shape of tree trunks. But we were
surprised to find that most of the wild colonies were living in tree cavities
much smaller than beekeepers’ hives. The average volume of the bees’ nest
cavities (not including one very large outlier) was just 47 liters (12.4 gal-
lons), which is only slightly larger than the 42 liters (11.1 gallons) of one
deep box (or hive body) in a Langstroth hive (Fig. 5.3). This means that,
on average, a wild colony’s nest cavity has only one- quarter to one- half of
the living space of a typical beekeeper’s hive. At this point we wondered,
do wild colonies prefer rather small and snug nesting sites?
Certainly, the bees living in these smallish tree cavities were making
good use of their living space, for each colony had nearly filled its nest
Seeley.indb 105
2/21/2019 8:07:32 AM
106 Chapter 5
1
2
3
Deep hive bodies
6
4
2
No. of nests
0 0 20 40 60 80 100 120 140 400 420 440 460
Nest volume (liters)
Fig. 5.3. Distribution of nest- cavity volumes for 21 bee- tree nests.
cavity with multiple sheets of beeswax comb (just eight combs, on aver-
age). Each comb formed a wall- to- wall curtain spanning the relatively
slender cavity, but the bees had built small passageways through the combs
where they attached to a cavity’s walls and ceiling (Fig. 3.7). These open-
ings served, no doubt, to allow the bees to crawl easily from one comb to
the next. Most combs hung free along their bottom edges, leaving several
centimeters (an inch or more) of open space between the bottoms of the
combs and the nest cavity’s floor. In recently occupied cavities, which
contained
yellow or light- brown combs, the floor was usually covered with
a layer of soft, dark, rotten wood several centimeters thick. But in ones
that were filled with dark combs, and probably had been occupied for
years, the cavity’s floor was coated with a dry, hard layer of tree resins
(propolis) several millimeters (up to ca. 0.10 inch) thick, which made it
shiny and waterproof. Likewise, the ceilings and walls of these long-
occupied nest cavities had hard coatings of tree resins about 0.5 millimeter
(0.02 inch) thick (Fig. 5.4). Evidently, when swarms had moved into these
cavities, the workers had promptly chewed off the soft, rotted wood from
their inner surfaces to expose firm wood for attaching their combs, and
then had gradually coated the wall and ceiling surfaces between the combs
with tree resins. We presumed that the bees had done all this work to seal
off the innumerable cracks and crevices where molds and bacteria would
Seeley.indb 106
2/21/2019 8:07:32 AM
The Nest 107
Fig. 5.4. Close- up of the propolis layer that a honey bee colony has built by coat-
ing the inner surfaces of its nest cavity with resins. The propolis has been chipped
off in the upper- right region to show the substrate of decayed wood with its
countless cracks and crevices, an ideal environment for bacterial growth.
otherwise thrive. Usually, the bees had extended their resin work to the
surfaces just outside the entrance opening, filling the fissures in the bark
here with propolis. This smoothed the bark surfaces around this busy spot.
I suspect that this improves the flow of bee traffic at this unavoidable point
of congestion; perhaps it also helps the bees disinfect their tarsi (feet)
before they enter their home.
Seeley.indb 107
2/21/2019 8:07:33 AM
108 Chapter 5
One more noteworthy feature of the walls of the bees’ nest cavities was
their thickness and thus their insulation. On average, the length of the
passageway through the tree’s trunk that the bees had to walk when enter-
ing or leaving their nest was slightly more than 15 centimeters (6 inches),
and in one case it was 74 centimeters (29 inches)!
Eight of the 21 wild colonies whose nests we analyzed had filled their
nest cavity with combs. On average, the total area of the combs in each
nest was 1.17 square meters (12.6 square feet). This is the amount of
The Lives of Bees Page 13
