Effects of Acarapis woodi on overwintered colonies of honey bees (Hymenoptera: Apidae) in New York


Otis, G.W.; Scott Dupree, C.D.

Journal of Economic Entomology 85(1): 40-46

1992


Colonies of honey bees, Apis mellifera L., infested with Acarapis woodi (Rennie) were studied during the four winters of 1985-1989 in New York state. Samples of bees were obtained from colonies on several dates from fall to spring to determine mite prevalence and mite load scores. Mite infestations were much heavier than those reported elsewhere in North America. Over the two winters for which adequate data were available (1987-1988 and 1988-1989), colonies with heavy mite infestations had significantly greater mortality. Spring brood areas were negatively correlated with mite prevalence and mite load scores. However, the strength of these correlations varied depending on the month and the year. These results indicate that tracheal mites have a substantial negative effect on colonies of honey bees in New York.

Effects
of
Acarapis
woodi
on
Overwintered
Colonies
of
Honey
Bees
(Hymenoptera:
Apidae)
in
New
York
GARD
W.
OTIS
AND
CYNTHIA
D.
SCOTT-DUPREE
Department
of
Environmental
Biology,
University
of
Guelph,
Guelph,
Ontario
NIG
2W1,
Canada
J.
Econ.
Entomol.
85(1):
40-46
(1992)
ABSTRACT
Colonies
of
honey
bees,
Apis
mellifera
L.,
infested
with
Acarapis
woodi
(Rennie)
were
studied
during
the
four
winters
of
1985-1989
in
New
York
state.
Samples
of
bees
were
obtained
from
colonies
on
several
dates
from
fall
to
spring
to
determine
mite
prevalence
and
mite
load
scores.
Mite
infestations
were
much
heavier
than
those
reported
elsewhere
in
North
America.
Over
the
two
winters
for
which
adequate
data
were
available
(1987-1988
and
1988-1989),
colonies
with
heavy
mite
infestations
had
significantly
greater
mortality.
Spring
brood
areas
were
negatively
correlated
with
mite
prevalence
and
mite
load
scores.
However,
the
strength
of
these
correlations
varied
depending
on
the
month
and
the
year.
These
results
indicate
that
tracheal
mites
have
a
substantial
negative
effect
on
colonies
of
honey
bees
in
New
York.
KEY
WORDS
Insecta,
Apis
mellifera,
tracheal
mites,
colony
mortality
Acarapis
woodi
(Rennie)
is
a
parasite
of
the
tra-
chea
of
the
honey
bee,
Apis
mellifera
L.
The
adults
of
these
mites
enter
the
trachea
of
young
bees,
where
they
feed
on
host
haemolymph
and
reproduce.
Among
other
effects,
they
have
been
reported
to
reduce
the
longevity
of
adult
bees
(Bailey
1958,
Bailey
&
Lee
1959,
Giordani
1962,
Maki
et
al.
1986)
and
to
increase
colony
mortality
during
the
winter
(Bailey
1958,
1961;
Bailey
&
Lee
1959;
Furgala
et
al.
1989;
Komeili
&
Am-
brose
1990;
Royce
&
Rossignol
1990).
Kjer
et
al.
(1989)
provided
a
recent
summary
of
the
biology
of
tracheal
mites
and
their
effects
on
honey
bee
colonies.
In
1984,
tracheal
mites
were
discovered
in
the
United
States
for
the
first
time.
They
have
sub-
sequently
been
recorded
in
>42
states
and
eight
provinces
of
Canada.
There
has
been
disagree-
ment
over
the
potential
seriousness
of
the
tra-
cheal
mite
as
a
pest
of
honey
bees
in
North
America.
This
uncertainty
arises
from
conflicting
information
on
the
degree
of
susceptibility
of
North
American
bees
to
the
mites
(Adam
1962,
1968;
Bailey
1965,
1967,
1985;
Bailey
&
Perry
1982)
and
from
a
lack
of
data
on
the
economic
effects
of
tracheal
mites
in
North
America.
Much
of
the
more
recent
research
on
this
pest
was
conducted
by
Bailey
and
his
associates
in
En-
gland
in
the
1950s
and
1960s,
at
least
35
yr
after
A.
woodi
is
known
to
have
become
established
there.
Their
conclusion
that
tracheal
mites
are
rarely
abundant
enough
to
have
much
effect
on
managed
bee
colonies
may
not
be
applicable
to
the
North
American
setting,
where
the
mite
ap-
pears
to
have
been
recently
introduced
(Eckert
1961,
Shimanuki
et
al.
1983,
Shimanuki
&
Knox
1989,
Furgala
et
al.
1989),
and
the
bees
may
not
have
evolved
resistance
mechanisms
(Adam
1987,
Taber
1988).
Recent
data
suggest
that
tra-
cheal
mites
are
associated
with
economic
losses
to
beekeepers
in
North
America
(Eischen
1987,
Eischen
et
al.
1989,
Furgala
et
al.
1989).
Our
objectives
were
to
determine
the
inci-
dence
of
mites
in
infested
colonies
in
commer-
cial
beekeeping
operations
in
New
York,
to
quantify
the
relationship
of
mite
prevalence
and
mite-load
scores
to
colony
mortality
and
spring
brood
areas,
and
to
test
for
association
between
the
incidence
of
tracheal
mites
and
infection
with
the
protozoan,
Nosema
apis
Zander.
Materials
and
Methods
Apiaries
with
honey
bee
colonies
known
to
be
infested
with
tracheal
mites
were
studied
in
New
York
State
during
four
winters.
In
no
case
was
it
known
how
long
the
apiaries
had
been
infested
with
mites.
In
1985-1986,
we
worked
in
two
apiaries
with
27
and
12
hives
respectively,
located
near
LeRoy,
Genesee
County,
N.Y.
The
following
year,
two
nearby
apiaries,
each
with
20
colonies
and
belonging
to
the
same
beekeeper,
were
studied.
The
colonies
were
neither
equal-
ized
in
strength
nor
requeened
because
we
wanted
to
study
colonies
that
were
under
routine
fall
management
as
performed
by
the
beekeeper.
Colonies
had
two
or
three
hive
bodies,
with
top
entrances
or
holes
drilled
in
the
upper
hive
body
and
sometimes
with
empty
hive
bodies
with
0022-0493/92/0040-0046$02.00/0
©
1992
Entomological
Society
of
America
February
1992
OTIS
&
SCOTT-DUPREE:
TRACHEAL
MITES
&
OVERWINTERED
HONEY
BEES
41
frames
on
top
of
the
hives
above
the
inner
cov-
ers.
In
preparation
for
overwintering,
colonies
were
fed
sugar
syrup
(1:1
sugar/water)
contain-
ing
sodium
sulfathiazole
to
control
American
foulbrood
disease;
they
were
not
fed
fumagillin.
In
both
years,
several
colonies
had
insuf
f
icient
food
reserves
to
survive
the
winter.
All
live
col-
onies
were
fed
granulated
sucrose
on
the
inner
hive
cover
on
17
February
1986
or
10
February
1987,
respectively,
to
prevent
starvation.
Samples
of
adult
bees
were
collected
from
study
colonies
on
17
October,
26
November,
2
January,
17
February,
15
March,
28
April,
and
6
June
1985-1986
and
on
20
June,
16
September,
28
October,
8
December,
29
January,
17
March,
and
5
May
1986-1987.
Bees
were
collected
from
the
tops
of
frames,
at
top
hive
entrances,
or
di-
rectly
from
the
outer
edge
of
the
bee
cluster.
(Results
by
Robinson
et
al.
[1986]
indicated
that
sampling
location
within
the
hive
does
not
influ-
ence
the
prevalence
of
mites
in
the
sample.)
Col-
ony
mortality
was
determined
during
sampling.
It
was
not
possible
to
sample
bees
without
some
minor
disturbance
to
the
colonies.
During
the
winter
of
1987-1988,
we
worked
with
colonies
belonging
to
a
different
beekeeper
north
of
Ithaca,
N.Y.,
between
Cayuga
and
Oneida
Lakes.
All
colonies
in
four
apiaries
known
to
be
infested
with
mites
were
studied.
The
apiaries
contained
21, 22,
19,
and
5
colonies.
Colonies
were
fed
sugar
syrup
(1:1)
containing
fumagillin
from
open
barrels
placed
in
the
apiary
in
early
October.
Most
were
overwintered
in
two
or
three
brood
chambers
with
substantial
stores
of
honey.
Extender
patties
containing
oxytetra-
cycline
hydrochloride
(Terramycin
25,
Pfizer
Pharmaceutical
Company,
Mississauga,
Ont.)
were
applied
in
April.
No
other
drugs
were
ad-
ministered.
Bees
were
sampled
from
inner
cov-
ers
or
outer
frames
on 9
October,
24
November,
and
12
March
1987-1988.
As
part
of
a
comparative
study
of
different
miti-
cides
in
a
fall
treatment
program,
four
additional
apiaries
containing
110
colonies
belonging
to
one
beekeeper
were
studied
in
Orleans,
Monroe,
and
Genesee
counties
in
1988-1989.
Samples
were
taken
from
inner
covers
or
outer
frames
on
8
October,
9
November,
and
21
April.
The
miti-
cides
had
no
effect
on
controlling
mite
popula-
tions
(unpublished
data);
consequently,
we
re-
port
colony
mortality
as
well
as
initial
mite
prevalence
values.
Samples
usually
contained
>100
bees;
few
samples
contained
<80
and
none
had
<60
bees.
Bees
were
placed
immediately
into
190-ml
plas-
tic
jars
containing
70%
ethanol.
The
presence
of
mites
was
determined
for
—100
bees
in
each
sample
unless
the
sample
contained
fewer
bees.
Thoracic
disks
were
cut
such
that
they
contained
the
main
tracheal
trunks.
Groups
of
disks
were
incubated
in
5%
KOH
for
24
h
in
a
warm
(43°C)
oven
(Delfinado-Baker
1984).
Each
disk
was
inspected
individually
with
a
dissecting
mi-
croscope
at
40x
or
63x
for
the
presence
or
ab-
sence
of
mites.
If
there
was
any
doubt
concern-
ing
the
presence
of
mites,
the
tracheae
were
removed,
mounted
on
a
microscope
slide,
and
viewed
at
40x
or
80x
under
a
compound
micro-
scope.
In
addition,
for
the
first
3
yr,
samples
taken
in
March
were
analyzed
for
N.
apis
using
the
technique
described
by
Cantwell
(1970).
For
all
years,
we
report
mite
prevalence,
which
is
simply
the
percentage
of
mite-infested
bees
in
the
sample
(Margolis
et
al.
1982,
"prev-
alence
(w)"
of
Eischen
1987,
Eischen
et
al.
1989).
In
the
second
and
third
winters,
approxi-
mate
mite
infestation
was
ranked
for
each
tra-
chea
examined.
Rankings
were:
0,
no
mites;
1,
<10
mites
of
all
stages;
2,
11-20
mites;
3,
21-30
mites;
4,
31-40
mites;
5,
>40
mites.
The
sum
of
the
ranks
for
the
two
tracheae
of
a
bee
provided
a
mite
load
score.
The
sum
of
these
scores
for
all
bees
in
a
sample
divided
by
the
number
of
bees
in
the
sample
gave
the
mean
mite
load
score
("parasite
load
score"
of
Eischen
1987,
Eischen
et
al.
1989).
Spring
brood
areas
of
the
surviving
colonies
were
estimated
on
28
April
1986,
5
May
1987,
and
26
April
1988.
A
frame
divided
into
six
sec-
tions
of
equal
area
was
placed
in
front
of
each
comb
containing
brood.
Only
one
side
of
each
comb
was
measured
because
brood
patterns
on
the
two
sides
of
a
comb
are
usually
similar.
The
percentage
of
each
section
containing
sealed
worker
brood
was
estimated
to
the
nearest
5%.
These
percentages
were
converted
to
areas,
dou-
bled
to
account
for
the
two
sides
of
the
combs,
and
summed
to
yield
total
brood
area
of
the
col-
ony.
Frequency
distributions
of
mite
prevalence
values
of
colonies
sampled
in
October
of
each
year
of
the
study
were
produced
for
comparison.
A
contingency
table
was
constructed
to
compare
mortality
of
colonies having
low
mite
infestation
with
those
of
moderate
and
high
infestation.
The
mortality
data
were
further
analyzed
with
a
lo-
gistic
regression
analysis
(SAS
general
linear
models
procedure)
(SAS
Institute
1985).
This
sta-
tistical
technique
uses
the
binomial
data
of
col-
ony
death
or
survival
over
the
range
of
observed
mite
prevalence
values
to
calculate
predicted
values
of
the probability
of
mortality
at
any
given
autumn
prevalence
of
mites.
Values
of
mite
prev-
alence
were
transformed
(arcsine
transformation)
before
correlations
were
calculated
between
mite
prevalence
and
spring
brood
areas.
A
Spear-
man
rank
correlation
analysis
(Steel
&
Torrie
1980)
was
conducted
on
Nosema
spore
counts
and
mite
prevalence
of
bees
from
the
same
15
March
1986
samples
to
test
for
an
association
between
infestation
with
these
two
organisms.
42
PERC
E
N
TA
GE
O
F
COLO
N
IES
SU
RV
EYED
JOURNAL
OF
ECONOMIC
ENTOMOLOGY
Vol.
85,
no.
1
60
60
1.0
50
1985
50
1986
40
40
30
39
Colonies
30
40
Colonies
0.8
20
20
Ts
10
10
°
2
0
6
60
60
50
40
1987
50
40
1988
0.4
c
30
66
Colonies
30
110
Colonies
o
c
20
20
o
0.2
6
,
0
10
10
po
scA
C5
4'
°
1ry
o
Ns"'
C.
e
qp
TRACHEAL
MITE
PREVALENCE
Fig.
1.
Frequency
distribution
of
October
mite
prevalence
values
(i.e.,
percentage
of
mite-infested
bees)
for
4
yr
of
study
in
New
York.
Results
Frequency
distributions
of
mite
prevalence
values
of
colonies
sampled
in
October
of
each
year
of
study
are
presented
in
Fig.
1.
The
per-
centage
of
colonies
with
>20%
mite
prevalence
was
43.6%
(n
=
39)
in
1985,
47.5%
(n
=
40)
in
1986,
30.3%
(n
=
66)
in
1987,
and
43.6%
(n
=
110)
in
1988;
the
overall
average
was
40.8%
(n
=
255).
This
compares
with
an
average
of
10.4%
(range
in
yearly
average,
0.7-21.5%)
reported
by
Bailey
(1961)
from
1955
to
1960
in
England.
Al-
most
all
colonies
sampled
in
New
York
were
infested
with
mites;
only
three
colonies
in
1987-
1988
and
five
colonies
in
1988-1989
had
no
mites
detected
in
three
100-bee
samples.
Colony
mortality
was
related
to
mite
infesta-
tions.
In
1987-1988,
lightly
infested
colonies
(0-
20%
mite
prevalence,
n
=
46)
in
October
expe-
rienced
9.1%
mortality
during
the
subsequent
winter.
Among
moderately
infested
colonies
(21-60%
mite
prevalence,
n
=
12)
there
was
16.7%
mortality,
and
among
heavily
infested
col-
onies
(>60%
mite
prevalence,
n
=
8),
75%
died.
Mortality
among
more
heavily
infested
colonies
was
significantly
greater
(G
test
with
Yates
cor-
rection,
G
=
5.36,
P
<
0.025).
In
1988-1989,
the
mortality
of
moderately
infested
(33.3%,
n
=
33)
and
heavily
infested
(86.6%,
n
=
15)
colonies
was
even
greater
than
in
the
preceding
year.
The
combined
mortality
among
moderately
and
heav-
ily
infested
colonies
was
significantly
greater
(G
test
with
Yates
correction,
G
=
23.24,
P
<
0.001)
than
the
mortality
among
lightly
infested
colo-
nies
(8.1%,
n
=
62).
For
comparison,
Bailey
(1961)
recorded
values
similar
to
ours
for
mortal-
ity
of
colonies
with
light
infestation
(9.7%,
n
=
893),
moderate
infestations
(32.0%,
n
=
78),
and
heavy
infestations
(82.9%,
n
=
35).
The
mortality
data
for
1988-1989
have
been
further
analyzed
with
a
logistic
regression
anal-
ysis
to
calculate
predicted
values
of
the
proba-
20
40
60
80
100
Mite
Prevalence
(%,
October
1988)
Fig.
2.
Results
of
logistic
regression
analysis
on
1988-1989
data
of
colony
mortality
and
mite
preva-
lence
values.
Solid
dots
represent
the
predicted
prob-
ability
of
mortality
as
calculated
from
the
regression
equation;
open
circles
indicate
95%
confidence
limits
calculated
for
each
predicted
mortality
value.
bility
of
mortality
at
any
given
autumn
preva-
lence
of
mites
(Fig.
2).
The
regression
equation,
In
(p/[1
=
—3.094
+
0.6478x,
where
p
is
the
probability
of
mortality
and
x
is
mite
prevalence,
is
highly
significant
(P
<
0.001).
Inclusion
of
an
x
2
term
failed
to
improve
the
fit
of
the
regression
line
to
the
data
(P
=
1.0).
In
the
first
2
yr
of
the
study,
the
more
heavily
infested
colonies
experienced
greater
mortality
(in
1985-1986,
10/19
lightly
infested
colonies
and
13/17
moderately
and
heavily
infested
colo-
nies
died;
in
1986-1987,
12/20
lightly
infested
colonies
and
17/19
moderately
and
heavily
in-
fested
colonies
died).
However,
these
differ-
ences
were
not
significant
because
of
the
high
overall
mortality
(presumably
the
result
of
poor
colony
management)
and
the
relatively
small
numbers
of
colonies
studied.
There
was
a
negative
correlation
between
mite
infestation
level
and
spring
brood
area.
For
each
of
the
first
3
yr
of
the
study,
mean
prevalence
was
negatively
correlated
with
log-transformed
spring
brood
area
at
some
time
during
fall
or
winter
(Table
1).
These
data
suggest
that
those
colonies
with
high
mite
infestations
that
did
sur-
vive
usually
had
reduced
brood
areas.
However,
the
pattern
from
year
to
year
was
variable.
In
1985-1986,
the
highest
correlation
occurred
in
March,
whereas
in
1986-1987,
the
highest
cor-
relation
occurred
in
late
October.
In
the
final
year,
1987-1988,
the
correlations
for
fall
and
early
spring
were
almost
identical.
Data
from
late
November—early
December
samples
are
de-
picted
in
Fig.
3.
The
additional
information
ob-
tained
from
mite
load
scores
did
not
substantially
improve
the
values
of
the
correlation
coeffi-
cients.
We
tested
for
an
association
between
infesta-
tion
with
tracheal
mites
and
N.
apis
infection.
A
January
1992
OTIS
&
SCOTT-
DUPREE:
TRACHEAL
MITES
&
OVERWINTERED
HONEY
BEES
43
log
Broo
d
Area
Table
1.
brood
area
Correlations
of
mite
prevalence
and
spring
Year
Date"
Mite
prevalence,
Mite
load
score,
1985-1986
17
Oct.
26
Nov.
-0.168ns
-0.50ns
2
Jan.
-0.564*
17
Feb.
-0.605"
15
Mar.
-0.723*
28
Apr.
-0.594*
1986-1987
16
Sept.
-0.587ns
-0.519ns
28
Oct.
-0.718* -0.684*
8
Dec.
-0.707*
-0.648*
29
Jan.
-0.614ns
-0.627ns
17
Mar.
-0.632*
-0.517ns
5
May
-0.427ns
-0.444ns
1987-1988
9
Oct.
-0.578***
-0.662***
24
Nov.
-0.552***
-0.668***
12
Mar.
-0.572***
-0.627***
Spring
brood
areas
were
measured
on
28
April
1986,
5
May
1987,
and
26
April
1988,
and
were
log-transformed
before
correlations
were
calculated.
ns,
not
significant;
",
P
<
0.05;
P
<
0.001.
"Date
samples
were
taken
for
mite
analyses.
The
data
for
italicized
dates
are
presented
graphically
in
Fig.
3.
b
Linear
correlations
of
mite
prevalence
(arcsine
square
root
of
percent
bees
infested)
and
spring
brood
areas.
`Linear
correlations
of
mite
load
scores
(based
on
rankings
of
number
of
mites
per
bee)
and
spring
brood
areas.
Spearman
rank
correlation
analysis
(Steel
&
Tor-
rie
1980)
was
conducted
on
Nosema
spore
counts
and
mite
prevalence
of
bees
from
the
same
15
March
1986
samples.
The
correlation
coefficient
was
negative
and
not
significant
(r,
=
-0.321;
0.10
>
P
>
0.05).
Nosema
was
found
to
be
neg-
ligible
in
1987
and
1988.
Spore
counts
were
not
taken
in
1989.
Discussion
In
this
study
we
have
documented
relatively
high
levels
of
mite
infestation
and
a
negative
relationship
between
mite
prevalence
and
both
winter
survival
of
colonies
and
the
strength
of
surviving
colonies
in
spring.
Mite
prevalence
was
much
higher
than
reported
by
Bailey
(1961)
in
England
during
1955-1960.
In
comparison
with
Bailey's
(1961)
findings,
we
consistently
found
that
three
to
four
times
more
colonies
in
New
York
had
mite
prevalence
values
>20%
regardless
of
location,
management
system,
and
queen
stocks.
Other
researchers
have
reported
high
mite
prevalence
in
Mexico
(Eischen
1987),
Minnesota
(Furgala
et
al.
1989),
Arizona
(Waller
&
Hines
1990),
and
elsewhere
in
the
United
States.
In
contrast,
tracheal
mites
are
currently
present
at
low
levels
throughout
most
of
Europe
and
only
occasionally
become
common
enough
to
affect
colony
performance
(Otis
1990).
The
most
likely
explanation
for
the
striking
differ-
ence
between
mite
prevalence
values
in
North
America
and
Europe
is
that
North
American
bees
lack
resistance
to
tracheal
mites.
This
has
also
been
suggested
by
Adam
(1962,
1968)
and
Taber
(1988).
Studies
that
have
been
conducted
to
ad-
dress
this
question
have
suggested
that
there
was
no
difference
in
susceptibility
between
British
and
North
American
stocks
of
bees
(Bailey
1965,
1967;
Gary
et
al.
1990).
However,
a
recent
re-
evaluation
of
Bailey's
experiments
reached
the
opposite
conclusion
(Kjer
et
al.
1989).
Additional
studies
comparing
European
and
North
Ameri-
can
stocks
of
bees
are
needed
to
settle
these
contradictory
conclusions.
Other
explanations
for
the
difference
in
mite
prevalence
values
are
possible.
It
has
long
been
known
that
climate
and
foraging
conditions
in-
fluence
the
incidence
of
tracheal
mites
(Bailey
1985).
However,
this
does
not
appear
to
be
the
cause
of
the
differences
that
have
been
observed.
According
to
Bailey
(1985),
the
generally
stron-
ger
honey
flows
in
North
America
should
result
in
a
lower
incidence
of
tracheal
mites,
whereas
we
have
observed
a
higher
mite
incidence.
An-
other
possibility
is
that
the
mites
in
North
Amer-
ica
are
different
than
those
in
Europe.
There
are
no
data
to
address
this
question
at
present.
In
all
4 yr
of
our
study,
heavily
infested
colo-
nies
experienced
greater
winter
mortality.
In
the
latter
2
yr
in
which
colony
management
was
not
1985-86
r
=
-
0.506
ns
=
14)
10
20
30
40
50
4.5
1986-87
r
=
-0.707
*
4
=
10)
3.5
3
2.5
2
70
eo
10
20
30
40
50
60
4
-
1987-88
r
=
-
0.552***
3.5
3
..
.0
.
(n
=
53)
2.5
-
'•
2
-
1.5
1
10
20
30
40
50
60
70
arcsin
V
Mite
Prevalence
Fig.
3.
Relationship
between
early
winter
mite
prevalence
and
spring
brood
areas.
Mite
prevalence
was
based
on
samples
taken
on
26
November
1985,
8
December
1986,
or
24
November
1987.
Correlations
of
spring
brood
areas
and
mite
prevalence
values
from
all
months
are
presented
in
Table
1.
4
-
3.5
-
3-
2.5
2
.
44
JOURNAL
OF
ECONOMIC
ENTOMOLOGY
Vol.
85,
no.
1
a
major
factor
affecting
mortality,
the
probability
of
colony
death
was
significantly
related
to
mite
prevalence.
Interestingly,
our
data
on
the
rela-
tionship
between
mite
prevalence
and
percent-
age
mortality
are
very
similar
to
those
of
Bailey
(1961).
In
contrast,
in
warmer
climates,
tracheal
mites
infrequently
cause
death
of
colonies
(Eis-
chen
1987;
H.
Cromroy,
personal
communica-
tion)
and
have
little
effect
on
individual
bees
(Gary
&
Page
1989).
Where
winters
are
longer
and
more
severe
than
in
New
York
(e.g.,
Minne-
sota),
mortality
at
any
given
mite
prevalence
is
greater
than
we
recorded
(Furgala
et
al.
1989).
This
trend
of
greater
effect
of
tracheal
mites
in
colder
climates
is
probably
related
to
the
longer
broodless
period
and
the
continued
mite
repro-
duction
in
their
host
bees
throughout
most
of
the
winter
(Otis
et
al.
1988).
Winter
bees
in
New
York
must
live
at
least
5
mo,
and
the
direct
effect
of
the
parasitic
mites
over
an
extended
period
of
time
and
perhaps
the
indirect
effect
of
pathogens
that
may
be
introduced
into
the
bees
by
feeding
mites
apparently
cause
the
death
of
the
colony
before
brood
production
can
produce
enough
young
bees
in
the
spring.
It
is
through
this
effect
on
wintering
colonies
that
tracheal
mites
are
likely
to
have
their
most
significant
effect
on
North
American
beekeeping.
The
physical
symptoms
of
mite-infested
colo-
nies
deserve
mention.
There
were
no
obvious
signs
of
mite
infestation
in
the
bees
themselves.
Generally
the
colonies
appeared
normal
as
well,
but
we
had
the
impression
that
the
winter
cluster
was
not
as
tight
in
heavily
infested
colonies.
Moreover,
on
warm,
late-winter
days,
large
num-
bers
of
bees
could
sometimes
be
seen
crawling
away
from
heavily
infested
hives
and
not
return-
ing.
Similar
behavior
has
been
observed
by
Kil-
lion
&
Lindenfelser
(1988)
and
Thoenes
&
Buch-
mann
(1990).
By
early
spring,
these
colonies
frequently
had
ample
honey
and
pollen
stores
but
only
a
handful
of
(or
no)
bees
alive
in
the
hive.
The
often
rapid
reduction
in
numbers
of
bees
led
to
chilled
brood
and
small
clusters
of
dead
bees
on
the
combs,
very
different
from
the
tight
cluster
of
bees
between
frames
and
in
cells
when
colonies
starve
in
late
winter.
These
are
symptoms
that
beekeepers
now
associate
with
tracheal
mite
infestations.
Mite
infestations
are
highly
variable
over
time.
Within
a
colony,
mite
prevalence
can
increase
or
decrease
dramatically
over
a
period
of
a
few
weeks
(at
the
same
time
that
neighboring
colo-
nies
are
undergoing
changes
in
prevalence
in
the
opposite
direction)
(Otis
et
al.
1988).
Rapid
de-
clines
in
mite
infestation
typically
occur
in
late
spring
as
the
number
of
newly
emerged
bees
is
rapidly
increasing
and
old
infested
bees
are
dy-
ing
(see
summary
in
Otis
et
al.
1988),
but
other
changes
in
prevalence
during
fall
and
winter
re-
main
unexplained.
This
unexplained
variation
confounds
detailed
analysis
of
the
changes
in
mite
populations.
It
may
also
have
contributed
to
the
relatively
weak
correlations
of
mite
preva-
lence
and
mite
load
scores
with
spring
brood
areas.
A
recent
report
by
Hyser
(1986)
suggested
that
there
was
a
positive
correlation
between
Nosema
infection
and
tracheal
mite
prevalence.
In
our
study,
there
was
a
weak
negative
correlation
be-
tween
these
two
variables.
Lozano
de
Haces
et
al.
(1989)
detected
no
statistical
association
be-
tween
the
presence
of
Nosema
and
A.
woodi
in
northeastern
Mexico.
Wille
(1966)
found
very
lit-
tle
Nosema
in
colonies
infested
with
tracheal
mites,
although
other
indications
of
pathogenic
organisms
such
as
bacterial
septicaemia
and
amoeba
cysts
were
present.
From
the
informa-
tion
that
is
available,
it
seems
that
Nosema
is
not
consistently
associated
with
tracheal
mites,
if
at
all.
If
further
research
indicates
a
significant
re-
lationship
between
mites
and
N.
apis,
it
might
be
usefully
expanded
by
investigating
correlations
with
other
variables
(e.g.,
microclimatic
differ-
ences
between
apiary
sites).
Several
studies
have
concluded
that
A.
woodi
are
not
significant
pests
of
honey
bees
in
Europe
(Bailey
1961,
1964;
Otis
1990;
Wille
et
al.
1987).
These
results
strongly
contrast
with
recent
data
on
the
effects
of
tracheal
mites
on
overwintering
honey
bee
colonies
in
North
America
(Eischen
1987;
Eischen
et
al.
1989;
Furgala
et
al.
1989;
Waller
&
Hines
1990,
present
study).
The
major
difference
appears
to
be
due
to
the
widespread
infestation
of
virtually
all
colonies,
often
at
high
levels
of
mite
prevalence,
in
fall
and
winter
in
North
America.
Research
directed
at
understand-
ing
the
host—parasite
interaction,
especially
mite
and
bee
population
dynamics
and
mechanisms
of
resistance
to
mites,
are
needed
to
enhance
the
effectiveness
of
control
strategies.
Acknowledgment
Doug
McRory
and
Roger
Morse
provided
the
initial
encouragement
for
this
study.
The
numerous
people
who
assisted
with
collection
and
analysis
of
samples
of
bees
include
Dave
Boyes,
Judy
Bath,
Barbara
Da-
wicke,
Rob
Dupree,
Gordon
Grant,
Paul
Kelly,
Barb
Kormos,
Doug
McRory,
Rhonda
Nauta,
Charlie
Parker,
Deb
Randall,
Mary
Catherine
Rillett,
Karen
Watkins,
Otillie
Welsh,
and
Tom
Young.
Dan
Ryan
conducted
the
logistic
regression
analysis.
The
participants
in
the
Honey
Bee
Tracheal
Mite
Scientific Symposium,
St.
Paul,
Minn.,
July
1986,
generously
shared
their
ideas
and
data
on
tracheal
mites.
Robert
Mungari
and
Paul
Cappy
(New
York
State
Department
of
Agriculture)
and
Richard
Nowogrodzki
(Cornell
University)
helped
us
locate
beekeepers
with
mite-infested
colonies.
The
manuscript
was
improved
by
comments of
R.
Nowo-
grodzki,
Judy
Wearing-Wilde,
and
two
anonymous
re-
viewers.
Finally,
without
the
cooperation
of
Art
Brew,
Jon
Ryan,
and
Jim
and
Ed
Doan,
who
provided
the
bee
colonies
and
permission
to
sample
from
them
through-
out
the
winter,
this
project
would
not
have
been
pos-
sible.
Funding
was
provided by
the
Ontario
Ministry
of
February
1992
OTIS
&
SCOTT-DUPREE:
TRACHEAL
MITES
&
OVERWINTERED
HONEY
BEES
45
Agriculture
and
Food
(Special
Research
Grant
and
Contract
with
the
University
of
Guelph),
the
National
Science
and
Engineering
Research
Council
of
Canada,
and
Agriculture
Canada
(Department
of
Supply
and
Services).
References
Cited
Adam,
Bro.
1962.
Experiences
with
acarine
disease.
Bienenvater
83:
35-38.
1968.
"Isle
of
Wight"
or
acarine
disease:
its
histor-
ical
and
practical
aspects.
Bee
World
49:
6-18.
1987.
Breeding
the
honey
bee.
Northern
Bee
Books,
Mythomroyd,
Hebden
Bridge,
England.
Bailey,
L.
1958.
The
epidemiology
of
the
infesta-
tion
of
the
honeybee,
Apis
mellifera
L.,
by
the
mite
Acarapis
tvoodi
Rennie
and
the
mortality
of
in-
fested
bees.
Parasitology
48:
493-506.
1961.
The
natural
incidence
of
Acarapis
woodi
(Rennie)
and
the
winter
mortality
of
honeybee
col-
onies.
Bee
World
42:
96-100.
1964.
The
"Isle
of
Wight
disease":
the
origin
and
significance
of
the
myth.
Bee
World
45:
32-37.
1965.
The
effects
of
Acarapis
woodi
on
honeybees
from
North
America.
J.
Apic.
Res.
4:
105-108.
1967.
The
incidence
of
Acarapis
woodi
in
North
American
strains
of
honeybee
in
Britain.
J.
Apic.
Res.
6:
99-103.
1985.
Acarapis
tvoodi:
a
modern
appraisal.
Bee
World
66:
99-104.
Bailey,
L.
&
D.
C.
Lee.
1959.
The
effect
of
infesta-
tion
with
Acarapis
tvoodi
(Rennie)
on
the
mortality
of
honey
bees.
J.
Insect
Pathol.
1:
15-24.
Bailey,
L.
&
J.
N.
Perry.
1982.
The
diminished
in-
cidence
of
Acarapis
woodi
(Rennie)
(Acari:
Tarson-
emidae)
in
honey
bees,
Apis
mellifera
L.
(Hy-
menoptera:
Apidae),
in
Britain.
Bull.
Entomol.
Res.
72:
655-662.
Cantwell,
G.
E.
1970.
Standard
methods
for
count-
ing
nosema
spores.
Am.
Bee
J.
110:
222-223.
Delfinado-Baker,
M.
1984.
Acarapis
woodi
in
the
United
States.
Am.
Bee
J.
124:
805-806.
Eckert,
J.
E.
1961.
Acarapis
mites
of
the
honey
bee,
Apis
mellifera
L.
J.
Insect
Pathol.
3:
409-425.
Eischen,
F.
A.
1987.
Overwintering
performance
of
honey
bee
colonies
heavily
infested
with
Acarapis
tvoodi
(Rennie).
Apidologie
18:
293-304.
Eischen,
F.
A.,
D.
Cardoso-Tamez,
W.
T.
Wilson
&
A.
Dietz.
1989.
Honey
production
of
honey
bee
col-
onies
infested
with
Acarapis
woodi
(Rennie).
Api-
dologie
20:
1-8.
Furgala,
B.,
S.
Duff,
S.
Aboulfaraj,
D.
Ragsdale
&
R.
Hyser.
1989.
Some
effects
of
the
honey
bee
tra-
cheal
mite
(Acarapis
woodi
Rennie)
on
non-migra-
tory,
wintering
honey
bee
(Apis
mellifera
L.)
colo-
nies
in
east
central
Minnesota.
Am.
Bee
J.
129:
195-197.
Gary,
N.
E.
&
R.
E.
Page,
Jr.
1989.
Tracheal
mite
(Acari:
Tarsonemidae)
infestation
effects
on
forag-
ing
and
survivorship
of
honey
bees
(Hymenoptera:
Apidae).
J.
Econ.
Entomol.
82:
734-739.
Gary,
N.
E.,
R.
E.
Page,
Jr.,
R.
A.
Morse,
C.
E.
Hend-
erson,
M.
E.
Nasr
&
K.
Lorenzen.
1990.
Compar-
ative
resistance
of
honey
bees
(Apis
mellifera
L.)
from
Great
Britain
and
United
States
to
infestation
by
tracheal
mites
(Acarapis
woodi).
Am.
Bee
J.
130:
667-669.
Giordani,
G.
1962.
Richerche
di
laboratorio
su
Ac-
arapis
woodi
Rennie
agente
dell'
acariosi
delle
api
(Apis
mellifera
L.).
Apic.
Ital.
29:
4-7,27-31
(Apic.
Abstr.
168/65).
Hyser,
D.
1986.
Tracheal
mite,
nosema,
and
win-
tered
honeybee
colonies
in
Minnesota,
pp.
32-37.
In
Proceedings
of
the
honey
bee
tracheal
mite
(Ac-
arapis
woodi,
R.)
scientific
symposium,
July
8-9,
1986,
Saint
Paul,
Minn.
American
Association
of
Professional
Apiculturalists
and
Canadian
Associa-
tion
of
Professional
Apiculturalists.
Killion,
E.
E.
&
L.
A.
Lindenfelser.
1988.
Observa-
tions
on
the
honey-bee
tracheal
mite
in
Illinois,
pp.
518-525.
In
G.
R.
Needham,
R.
E.
Page,
Jr.,
M.
Delfinado-Baker
&
C.
E.
Bowman
[eds.],
African-
ized
honey
bees
and
bee
mites.
Wiley,
Toronto.
Komeili,
A.
B.
&
J.
T.
Ambrose.
1990.
Biology,
ecol-
ogy
and
damage
of
tracheal
mites
on
honey
bees
(Apis
mellifera).
Am.
Bee
J.
130:
193-199.
Kjer,
K.
M.,
D.
W.
Ragsdale
&
B.
Furgala.
1989.
A
retrospective
and
prospective
overview
of
the
honey
bee
tracheal
mite,
Acarapis
woodi
R.
Am.
Bee
J.
129:
25-28,112-115.
Lozano
de
Haces,
L.,
W.
L.
Rubink,
W.
T.
Wilson
&
M.
Guillen-M.
1989.
Nosema
and
honey
bee
tra-
cheal
mite
interaction
in
swarms
from
northeastern
Mexico.
Am.
Bee
J.
129:
818.
Maki,
D.
L.,
W.
T.
Wilson
&
R.
L.
Cox.
1986.
Infes-
tation
by
Acarapis
tvoodi
and
its
effect
on
honeybee
longevity
in
laboratory
cage
studies.
Am.
Bee
J.
126:
832.
Margolis,
L.,
G.
W.
Esch,
J.
C.
Holmes,
A.
M.
Kuris
&
G.
A.
Schad.
1982.
The
use
of
ecological
terms
in
parasitology
(Report
of
an
ad
hoc
committee
of
the
American
Society
of
Parasitologists).
J.
Parasitol.
68:
131-133.
Otis,
G.
W.
1990.
Results
of
a
survey
on
the
eco-
nomic
impact
of
tracheal
mites.
Am.
Bee
J.
130:
28-31.
Otis,
G.
W.,
J.
B.
Bath,
D.
L.
Randall
&
G.
M.
Grant.
1988.
Studies
of
the
honey
bee
tracheal
mite
(Ac-
arapis
woodi)
(Acari:
Tarsonemidae)
during
winter.
Can.
J.
Zool.
66:
2122-2127.
Robinson,
F.
A.,
K.
L.
Thel,
R.
C.
Littell
&
S.
B.
Linda.
1986.
Sampling
apiaries
for
honey
bee
tracheal
mite
(Acarapis
woodi
Rennie):
effects
of
bee
age
and
colony
infestation.
Am.
Bee
J.
126:
193-195.
Royce,
L.
A.
&
P.
A.
Rossignol.
1990.
Honeybee
mortality
due
to
tracheal
mite
parasitism.
Apidolo-
gie
21:
147-151.
SAS
Institute.
1985.
SAS
user's
guide:
statistics.
SAS
Institute,
Cary,
N.C.
Shimanuki,
H.
&
D.
A.
Knox.
1989.
Tracheal
mite
surveys.
Am.
Bee
J.
129:
671-672.
Shimanuki,
H.,
D.
A.
Knox,
M.
Delfinado-Baker
&
J.
Lima.
1983.
National
honey
bee
mite
survey
(1).
Apidologie
14:
329-332.
Steel,
R.
G.
&
J.
H.
Torrie.
1980.
Principles
and
procedures
of
statistics,
a
biometrical
approach,
2nd
ed.
McGraw-Hill,
New
York.
Taber,
S.
1988.
Recent
research
on
the
internal
par-
asitic
mite.
Am.
Bee
J.
128:
740-741.
Thoenes,
S.
C.
&
S.
L.
Buchmann.
1990.
Tracheal
mite
induced
colony
mortality
monitored
by
an
electronic
scale.
Am.
Bee
J.
130:
816.
Waller,
G.
D.
&
L.
H.
Hines.
1990.
A
search
for
tracheal
mite
resistance
in
Arizona
honey
bees.
Am.
Bee
J.
130:
818.
46
JOURNAL
OF
ECONOMIC
ENTOMOLOGY
Vol.
85,
no.
1
Wile,
H.
1966.
Die
Milbe
(Acarapis
woodi)
in
Zusammenhang
mit
anderen
ICrankheitserregern.
Sudwestdeutsche
Imker
18:
322-324.
(Apic.
Abstr.
137/67).
Wille,
H.,
G.
Geiger
&
A.
Muff.
1987.
Einfluss
der
Milbe
Acarapis
woodi
auf
den
Massenwechsel
von
Bienvolkeni.
Forschungsanstalt
fur
Milchwirt-
schaft,
Mitteilungen
der
Sektion
Bienen,
No.
1,
Mai
1987.
Received
for
publication
22
January
1990;
accepted
5
August
1991.