Effects of a Bacillus thuringiensis toxin, two Bacillus thuringiensis biopesticide formulations, and a soybean trypsin inhibitor on honey bee (Apis mellifera L.) survival and food consumption


Malone, L.A.; Burgess, E.P.J.; Stefanovic, D.

Apidologie 30(6): 465-473

1999


Newly emerged adult honey bees, Apis mellifera L., were fed with a pollen-based food containing various additives: purified and activated Cry I Ba 8-endotoxin, from Bacillus thuringiensis Bt4412 (Bt) (1, 0.25 and 0.025 % w/w), Bt biopesticide preparations, Dipel 2X (1 and 0.25 %) and Foray 48B (0.25 %), and Kunitz soybean trypsin inhibitor (SBTI) (1, 0.5 or 0.05 %). The bees received these foods for 7 days and were then given control food without additives for the rest of their lives. Bee survival time was unaffected, and the food was consumed at the same rate as control food for all treatments, except I % Dipel, where both survival and food consumption were significantly reduced. A second experiment showed that bees completely deprived of the pollen-based food also had poorer survival than those fed with the control food. Adult bees are unlikely to be harmed by transgenic plants expressing CrylBa or SBTI, or by Bt hiopesticides that are used as recommended.

Apidologie
30
(1999)
465-473
465
Inra/DIB/AGIB/Elsevier,
Paris
Original
article
Effects
of
a
Bacillus
thuringiensis
toxin,
two
Bacillus
thuringiensis
biopesticide
formulations,
and
a
soybean
trypsin
inhibitor
on
honey
bee
(Apis
mellifera
L.)
survival
and
food
consumption
Louise
Anne
Malone*,
Elisabeth
Phyllis
June
Burgess,
Dragana
Stefanovic
Horticulture
and
Food
Research
Institute
of
New
Zealand
Limited,
Mt
Albert
Research
Centre,
Private Bag
92169,
Auckland,
New
Zealand
(Received
13
April
1999;
revised
5
August
1999;
accepted
18
August
1999)
Abstract
Newly
emerged
adult
honey
bees,
Apis
mellifera
L.,
were
fed
with
a
pollen
-based
food
containing
various
additives:
purified
and
activated
Cry
I
Ba
8-endotoxin,
from
Bacillus
thuringien-
sis
Bt4412
(Bt)
(1,
0.25
and
0.025
%
w/w),
Bt
biopesticide
preparations,
Dipel
2X
(1
and
0.25
%)
and
Foray
48B
(0.25
%),
and
Kunitz
soybean
trypsin
inhibitor
(SBTI)
(1,
0.5
or
0.05
%).
The
bees
received
these
foods
for
7
days
and
were
then
given
control
food
without
additives
for
the
rest
of
their
lives.
Bee
survival
time
was
unaffected,
and
the
food
was
consumed
at
the
same
rate
as
control
food
for
all
treatments,
except
I
%
Dipel,
where
both
survival
and
food
consumption
were
significantly
reduced.
A
second
experiment
showed
that
bees
completely
deprived
of
the
pollen
-based
food
also
had
poorer
survival
than
those
fed
with
the
control
food.
Adult
bees
are
unlikely
to
be
harmed
by
trans
-
genic
plants
expressing
CrylBa
or
SBTI,
or
by
Bt
hiopesticides
that
are
used
as
recommended.
Inra/DIB/AGIB/Elsevier,
Paris
honey
bee
/
Bacillus
thuringiensis
/
CrylBa
toxin
/
Kunitz
soybean
trypsin
inhibitor
/
pest
-resistant
transgenic
plants
1.
INTRODUCTION
In
response
to
concerns
about
the
envi-
ronmental
and
human
health
effects
of
chemical
pesticides,
as
well
as
the
evolu-
tion
of
pesticide
-resistant
insect
biotypes,
an
increasing
number
of
crop
plants
are
being
genetically
modified
to
make
them
resistant
to
pest
attack.
Honey
bees
polli-
nate
many
of
these
crops,
for
example,
*
Correspondence
and
reprints
E-mail:
LMalone@HORT.CRI.NZ
466
L.A.
Malone
et
al.
clover
(Trifolium
repens),
sunflower
(Helion-
thus
annuus),
sweet
potato
(Ipomoea
batatas),
strawberry
(Fragaria
x
ananassa),
apple
(Malus
domestica),
oilseed
rape
(Bras-
sica
napus),
and
cotton
(Gossypium
spp.)
[11,
14].
The
success
of
new
transgenic
cul-
tivars
will
depend
in
part
on
their
safety
for
pollinating
insects.
Pollinating
bees
could
be
affected
by
pest
-
resistant
transgenic
plants
either
directly
or
indirectly.
Direct
effects
may
arise
upon
ingestion
of
pollen
expressing
or
carrying
the
pest
-resistance
gene.
Pollen
is
about
24
%
protein
[22]
and
thus
represents
a
likely
site
for
transgene
expression.
CrylAc
toxin
of
Bacillus
thuringiensis
(Bt)
has
been
expressed
as
0.6
lug
per
gram
of
fresh
weight
(0.24
%
of
total
protein)
of
transgenic
cotton
pollen
[13].
However,
neither
serine
nor
cysteine
proteinase
inhibitors
could
be
detected in
the
pollen
of
transgenic
oilseed
rape
plants
containing
the
genes
for
these
inhibitors
[20].
Nectar
is
unlikely
to
con-
tain
gene
products
as
it
is
virtually
pure
car-
bohydrate
and
usually
contains
only
a
few
amino
acids
[1].
There
are
no
published
records
of
gene
products
being
found
in
the
nectar
of
transgenic
plants.
Indirect
effects
may
arise
either
via
inadvertent
changes
in
phenotype
resulting
from
the
position
of
the
new
gene
in
the
plant
genome
(insertional
mutagenesis)
or
via
pleiotropic
effects,
whereby
the
expression
of
the
new
gene
alters
a
biochemical
pathway
with
pheno-
typic
consequences.
Pleiotropic
effects
would
be
expected
to
occur
in
every
line
of
a
transformed
crop
plant,
whereas
deleteri-
ous
insertional
mutagenesis
effects
may
be
avoided
by
selecting
lines
that
do
not
have
the
undesirable
change
in
phenotype.
Reduc-
tions
in
nectar
volume
or
concentration,
or
changes
in
fl
ower
morphology
are
exam-
ples
of
phenotypic
changes
which
could
indirectly
affect
bees
[21].
White
clover
is
an
important
forage
crop
and
a
significant
nectar
source
for
honey
production
[18].
Successful
pollination
of
clover
is
also
important
in
locations
where
clover
dies
during
winter
and
adequate
seed
reservoirs
in
the
soil
must
be
maintained
[26].
Two
gene
products
are
candidates
for
incorporation
into
transgenic
white
clover.
The
Bt
8-endotoxin
CrylBa
has
been
shown
in
laboratory
experiments
to
be
effective
against
the
porina
caterpillar,
Wiseana
spp.
(Lepidoptera:
Hepialidae)
[23],
and
the
pro-
teinase
inhibitor,
SBTI
(Kunitz
soybean
trypsin
inhibitor),
has
been
shown
to
be
effective
against
the
black
field
cricket,
Teleogryllus
commodus
(Orthoptera:
Gryl-
lidae)
[4],
both
of
which
are
pests
of
white
clover
in
New
Zealand.
The
present
study
examines
the
rate
of
pollen
-food
consumption
and
survival
of
newly
emerged
adult
honey
bees
after
feed-
ing
on
two
different
pest
-resistance
gene
products
and
two
commercial
Bt
formula-
tions.
Pollen
is
a
necessary
protein
source
for
newly
emerged
adult
bees
to
complete
their
development,
and
of
all
bee
life
stages,
these
consume
the
greatest
quantities
of
this
food
[31].
Thus,
transgenic
pollen
is
likely
to
have
a
greater
impact
on
newly
emerged
adult
bees
than
on
larvae
or
older
adults.
2.
MATERIALS
AND
METHODS
Stocks
of
Italian
race
bees
were
obtained
from
our
apiary
at
Mt
Albert
Research
Centre,
Auck-
land.
Brood
frames
containing
capped
cells
were
brought
into
the
laboratory
and
the
cappings
gen-
tly
removed
using
forceps.
Adult
bees
that
were
ready
to
emerge
were
gently
pulled
from
their
cells
and
held
in
groups
of
about
50
at
32
°C
in
darkness
until
enough
had
been
collected
for
the
experiment.
This
uncapping
procedure
was
employed
as
a
precaution
against
the
bees
becom-
ing
infected
with
Nosema
apis
Zander
via
by
ingestion
of
spore
-contaminated
wax.
All
bees
used
in
the
experiments
were
less
than
12
h
old.
Two
experiments
were
carried
out.
The
first
consisted
of
nine
different
treatments
(three
rates
each
of
CrylBa,
Bt
biopesticides,
or
SBTI)
and
a
control.
The
second
consisted
of
one
treatment
(bees
completely
deprived
of
pollen
-food)
and
a
control.
This
experiment
was
undertaken
to
provide
some
information
on
the
survival
of
bees
starved
of
protein,
with
a
view
to
assessing
Effects
of
Bt
toxins
and
SBTI
on
bees
467
whether
effects
observed
in
the
first
experiment
could
be
explained
as
a
simple
avoidance
of
the
pollen
-food,
rather
than
direct
toxicity
of
the
additives.
In
the
first
experiment,
bees
were
assigned
randomly
to
wooden
cages
(9
x
8
x
7
cm)
with
mesh
on
two
sides,
40
bees
per
cage.
Thirty
cages
in
total
were
set
up:
three
blocks
x
nine
treat-
ments
and
one
control.
Each
cage
was
fitted
with
two
gravity
feeders,
one
containing
water
and
the
other
sugar
syrup
(60
%
w/v
sucrose
solu-
tion),
which
were
replenished
as
necessary
dur-
ing
the
experiment.
Sufficient
pollen
-food
was
prepared
(0.33
parts
pollen,
0.08
parts
sodium
caseinate,
0.16
parts
brewer's
yeast,
and
0.43
parts
sucrose
mixed
with
water
to
a
paste)
to
supply
the
total
num-
ber
of
bees
in
each
treatment
group
for
about
8
days.
The
pollen
used
was
bee
-collected
from
unknown
floral
sources
and
stored
at
—20
°C.
For
each
cage,
about
3
g
of
pollen
-food
with
the
appropriate
treatment
additive
(described
below)
was
placed
in
a
plastic
receptacle.
This
was
weighed
at
the
beginning
of
the
experiment,
at
12-h
intervals
for
5
days,
daily
for
a
further
4
days,
and
then
every
2
or
3
days
until
all
the
bees
had
died.
The
numbers
of
surviving
bees
in
each
cage
were
recorded
and
dead
bees
removed
at
these
times
also.
On
day
7,
each
pollen
-food
receptacle
was
weighed,
removed,
and
replaced
with
a
new,
weighed
receptacle
containing
fresh
pollen
-food
without
any
additive.
For
the
first
experiment,
CrylBa
and
SBTI
(Sigma,
St.
-Louis,
MO)
were
each
mixed
thor-
oughly
into
pollen
-food
at
three
concentrations.
These
were
chosen
to
represent
an
unrealistically
high
concentration,
a
high
but
realistic
concen-
tration
(equivalent
to
the
highest
expression
level
that
might
be
expected
to
be
effective
in
a
trans
-
genic
plant
with
that
gene)
and
a
realistic
low
concentration
(equivalent
to
a
low,
but
still
effec-
tive,
plant
expression
level).
For
CrylBa,
these
were
1,
0.25
and
0.025
%
w/w
in
pollen
-food
(equivalent
to
4,
1
and
0.1
%
of
total
protein),
based
on
bioassay
results
with
Bt
-cotton
and
Tri-
choplusia
ni,
Spodoptera
exigua,
Helicoverpa
virescens,
and
Helicoverpa
zea
[2,
19]
and
Bt
expression
levels
in
cotton
pollen
[13].
SBTI
was
used
at
1,
0.5
and
0.05
%
w/w
in
pollen
-
food
(equivalent
to
4,
2
and
0.2
%
of
total
protein),
based
on
bioassay
results
with
SBTI-
tobacco
and
Spodoptera
litura
(E.P.J.
Burgess,
unpublished
data).
Activated
CrylBa
toxin
was
obtained
from
a
large-scale
fermentation
of
B.
thuringiensis
Bt4412,
purified
and
cleaved
according
to
the
method
described
by
Simpson
et
al.
[23].
Activated
toxin
was
used
as
this
most
closely
resembles
the
form
in
which
CrylBa
will
be
expressed
in
transgenic
clover
plants.
Two
biopesticides,
Dipel
2X
(Abbott,
North
Chicago,
IL)
and
Foray
48B
(Novo
Nordisk,
Danbury,
CT),
were
added
to
pollen
-food
at
0.25
%
w/w
of
active
ingredient
in
pollen
-food
(equivalent
to
1
%
of
total
protein).
This
concentration
was
chosen
as
it
approximates
the
minimum
LD„
for
a
pesticide
which
is
'virtually
non-toxic'
to
honey
bees
as
defined
by
Crane
and
Walker
[10],
and
it
also
allows
comparison
with
the
treatments
where
the
gene
products
were
delivered
as
1
%
of
total
protein.
A
further
Dipel
2X
treatment
deliv-
ered
an
unrealistically
high
dose
(1
w/w
of
active
ingredient
or
4
%
of
total
protein)
to
allow
for
direct
comparison
with
the
high
-concentration
gene
product
treatments.
A
second
experiment
assessed
the
survival
of
bees
starved
of
protein.
It
was
set
up
in
simi-
lar
fashion,
with
eight
cages
in
total:
one
con-
trol
with
pollen
-food
without
additive
and
one
protein
-starvation
treatment
with
no
pollen
-food
x
four
blocks.
Each
cage
had
water
and
syrup
provided,
and
was
checked
at
regular
intervals.
A
survival
curve,
in
which
the
percentage
of
bees
remaining
alive
in
each
cage
was
plotted
against
time
in
days
from
the
beginning
of
the
experiment,
was
generated
for
each
cage
of
bees.
Mantel-Haenzel
(log
-rank)
tests
[15]
were
carried
out
to
compare
Kaplan
Meier
estimates
of
sur-
vival
distribution,
S(t),
for
bees
receiving
each
treatment.
Food
consumption
(mg
per
bee
per
12
h)
was
calculated
and,
as
the
data
had
a
skewed
distribution,
was
transformed
by
log
(value
+
0.05).
Mean
transformed
food
con-
sumption
values
for
each
treatment
at
each
time
point
were
compared
by
analysis
of
variance.
3.
RESULTS
AND
DISCUSSION
In
the
first
experiment
only
one
treat-
ment,
Dipel
at
1
%
of
active
ingredient,
resulted
in
significantly
poorer
bee
survival
than
the
other
treatments
(log
-rank
test,
P
<
0.001)
(figure
lb).
The
bees
in
the
sec-
ond
experiment
that
were
starved
of
protein
also
had
significantly
poorer
survival
than
their
controls
(log
-rank
test,
P<
0.001)
(fig-
ure
1d).
468
L.A.
Malone
et
al.
Percent
Alive
a)
>
<
C
U
(a)
- - -
1%
Cry1Ba
100
------
0.25%
Cry1
Ba
0.025%
Cry1Ba
80
-
Control
60
40
20
10
20
30
40
50
60
70
80
Days
(b)
100
- - -
1%
Dipel
80
------
0.25%
Dipel
0.25%
Foray
-
Control
60
20
Percent
Alive
Percent
Alive
100
80
60
40
20
100
80
6
4
20
(c)
-
1%
SE1TI
0.5%
SBTI
0.05%
SBTI
-
Control
10
20
30
40
50
60
70
80
Days
(d)
Control
Starved
10
20
30
40
50
60
70
80
0
10
20
30
40
50
60
70
80
Days
Days
Figure
1.
Survival
of
adult
bees
fed
for
7
days
with
pollen
-food
containing
a)
1,
0.25
or
0.025
%
w/w
CrylBa
Bt
toxin;
b)
1
or
0.25
%
Dipel
or
0.25
%
Foray
(Bt
biopesticides);
or
c)
1,
0.5
or
0.05
%
SBTI
proteinase
inhibitor.
Survival
of
control
bees
fed
pollen
-food
with
no
additives
is
shown
in
each
graph
(a,
b,
c).
The
survival
of
bees
deprived
completely
of
pollen
-food
is
compared
with
a
second
control
group
(d).
Food
consumption
rates
were
similar
for
all
bees
except
those
receiving
1
%
Dipel
(figure
2b).
These
bees
consumed
their
food
at
a
significantly
lower
rate
than
bees
in
the
other
treatments
between
days
3
and
6
(ANOVA,
at
day
3,
morning,
P
=
0.003;
at
day
4,
morning,
P
<
0.001;
at
day
5,
morn-
ing,
P
<
0.001;
and
at
day
6,
P
<
0.001).
Later
in
the
experiment,
when
all
bees
were
receiving
control
food,
these
bees
consumed
significantly
more
than
the
others
(ANOVA,
at
day
21,
P
=
0.022;
at
day
26,
P
=
0.002;
at
day
28,
P
=
0.009;
at
day
30,
P
=
0.035;
and
at
day
33,
P
=
0.004).
Thus,
bees
were
not
harmed
by
ingest-
ing
the
gene
products
tested
here,
even
at
concentrations
much
higher
than
the
expected
expression
levels
in
transgenic
plants.
Neither
were
bees
harmed
by
real-
istic
concentrations
of
two
commercial
Bt
formulations.
The
mortality
and
reduced
food
consumption
observed
among
bees
receiving
an
unrealistically
high
concentra-
tion
of
Dipel
2X
serves
to
demonstrate
that
the
methods
used
in
these
experiments
were
appropriate
for
demonstrating
toxic
effects
of
the
additives.
The
safety
of
commercial
Bt
formula-
tions
for
honey
bees
was
previously
estab-
lished
[3,
7-9,
12, 17,
27].
Only
prepara-
tions
containing
exotoxin,
in
addition
to
the
endotoxins
encoded
by
cry
genes,
have
been
shown
to
have
any
harmful
effects
on
bees
[16,
28].
Additionally,
Bt
is
used
for
con-
trol
of
a
lepidopteran
pest
of
bee
hives,
the
greater
wax
moth,
Galleria
mellonella
[6,
Effects
of
Bt
toxins
and
SBTI
on
bees
469
25, 29,
30],
further
suggesting
a
lack
of
sus-
ceptibility
of
honey
bees
to
Bt.
In
the
present
study,
only
1
%
Dipel
(which
does
not
contain
any
exotoxins)
Consumption
Rate
(mg/bee/12h)
7
6
-
5-
4
-
3
2
0
7
\
-
6
0)
O
5
0)
4
ry
to
3
CC
0
0)
O
C
2
0
(a)
1%
Cry1Ba
0.25%
Cry1
Ba
0.025%
Cry1Ba
Control
LSD
(0.05)
5
10
15
20
25
30
35
Days
(b)
I
,
- - -
1%
Dipel
-•----
0.25%
Dipel
0.25%
Foray
Control
LSD
(0.05)
t
.0
8
C
--(15
-
7-
o
CD
6-
F
(0
4
c0
CC
3-
c
0
2
E
5
1
I
11
4
10
15
20
25
30
35
Days
(c)
- -
-
1%
SBTI
---
0.5%
SBTI
0.05%
SBTI
-
Control
LSD
(0.05)
10
15
20
25
30
35
Days
Figure
2.
Rates
(mg/bee/12
h)
at
which
adult
bees
consumed
pollen
-food
containing
a)
1,
0.25
or
0.025
w/w
Cry
1Ba
Bt
toxin;
b)
1
or
0.25
%
Dipel
or
0.25
%
Foray
(Bt
biopesticides);
or
c)
1,
0.5
or
0.05
%
SBTI
proteinase
inhibitor.
The
rate
of
consumption
by
control
bees
of
pollen
-food
without
additives
is
shown
in
each
graph
(a,
b,
c).
After
7
days
bees
in
all
treatments
were
transferred
to
control
food
(arrows).
resulted
in
reduced
food
consumption
and
significant
bee
mortality.
Although
a
statis-
tical
comparison
cannot
be
made
between
results
from
the
two
different
experiments,
it
is
interesting
to
note
that
the
median
longevity
of
bees
fed
1
%
Dipel
(12
days
with
95
%
confidence
interval
of
8-16
days)
was
lower
than
that
of
bees
that
were
starved
of
protein
(21
days
with
95
%
confidence
interval
of
21-22
days).
This
suggests
that
their
mortality
was
not
simply
a
result
of
the
bees
being
repelled
by
the
Dipel
in
the
diet,
but
that
it
also
had
some
toxic
effect.
However,
we
did
not
ascertain
whether
this
effect
was due
to
the
Bt
spores
and
toxins
in
the
preparation
or
to
one
of
the
'inert'
ingre-
dients
in
the
mixture.
One
important
difference
between
Bt-transgenic
plants
and
Bt
biopesticides
is
that
the
transgenic
plant
will
express
only
a
single
Bt
toxin,
or
a
well-defined
combi-
nation
of
toxins,
whereas
a
biopesticide
may
contain
several
toxins
in
unknown
propor-
tions,
as
well
as
spores
and
vegetative
stages.
Furthermore,
Bt-transgenic
plants
will
express
only
the
soluble
and
cleaved
form
of
the
toxin,
rather
than
the
full-length
and
crystalline
forms
found
in
commercial
Bt
preparations.
There
are
few
published
stud-
ies
describing
honey
bee
tests
with
purified
Bt
toxins
that
represent
single
cry
gene
prod-
ucts.
Our
results
agree
with
those
of
Sims
[24],
who
used
full-length
purified
CrylAc
toxin
and
found that
the
mortality
(24
%)
of
adult
bees
fed
20
µg-mL
-1
of
this
toxin
in
syrup
for
7
days
did
not
differ
significantly
from
that
of
control
bees
fed
either
heat
-
attenuated
toxin
(22
%)
or
no
toxin
(25
%).
These
mortality
figures
are
higher
than
those
recorded
in
our
study
after
7
days.
This
may
have
been
because
the
bees
in
Sims'
study
were
kept
at
low
temperatures
(22-26
°C)
and
were
not
supplied
with
any
pollen
-based
or
other
protein
food.
Consumption
of
syrup
was
not
recorded
here
or
in
previously
published
studies,
but
if
we
assume
that
caged
bees
consume
0.032
mL
of
syrup
per
bee
per
day
(L.A.M.,
470
L.A.
Malone
et
al.
unpublished
data),
then
the
bees
in
Sims'
study
received 0.64
pig
of
full-length
CrylAc
protein
per
day.
This
represents
a
lower
dose
of
Cry
protein
than
any
received
by
bees
in
our
own
study.
Averaged
over
their
lifetimes,
our
bees
received
0.95,
11
or
40µg
of
cleaved
CrylBa
per
day
(equiv-
alent
to
0.1,
1
or
4
%
of
total
protein,
respec-
tively).
CrylBa
expression
levels
that
will
result
in
effective
pest
control
on
transgenic
white
clover
have
not
yet
been
established.
However,
Bt
-cotton
plants
expressing
CrylAb
or
CrylAc
as
0.05-0.1
%
of
total
soluble
plant
protein
have
been
shown
to
effectively
control
the
pest
insects,
T.
ni
and
S.
exigua
[19].
If
these
levels
are
typical
of
Bt-transgenic
plants,
then
we
can
assume
that pest
-resistant
white
clover
expressing
CrylBa
will
not
harm
adult
bees.
The
SBTI
results
recorded
here
suggest
that
this
gene
product
will
also
be
safe
for
adult
honey
bees,
as
none
of
the
treatments
reduced
longevity
or
food
consumption
sig-
nificantly.
This
is
in
agreement
with
an
ear-
lier
study
[5]
in
which
adult
bees
were
sup-
plied
ad
libitum
with
sugar
syrup
to
which
SBTI
had
been
added
at
various
concentra-
tions.
To
compare
results
from
the
two
stud-
ies,
the
lifetime
doses
of
SBTI
consumed
in
each
may
be
estimated.
In
the
present
study,
this
can
be
determined
by
multiplying
the
mean
amount
of
pollen
food
consumed
over
the
first
7
days
of
the
experiment
by
the
concentration
of
SBTI
in
the
food.
This
gives
three
treatments
delivering
total
doses
of
866,
451
and
36
[tg
of
SBTI
(1,
0.5
and
0.05
%
treatments,
respectively).
None
of
these
treatments
caused
significant
bee
mor-
tality.
In
the
earlier
study
[5],
assuming
that
bees
consume 0.032
mL
of
syrup
per
day,
the
lifetime
doses
of
SBTI
can
be
estimated
by
multiplying
the
median
lifetimes
of
the
bees
in
each
treatment
by
0.032
(mL•day
-
')
and
by
the
concentration
of
SBTI
in
the
syrup
(mg•mL
-
').
This
gives
three
high
doses,
1.92,
2.24
and
0.77
mg,
which
caused
significant
bee
mortality
and
two
low
doses,
122
and
12
lig,
which
did
not.
Thus,
it
appears
that
SBTI
will
be
safe
for
adult
bees
provided
the
lifetime
dose
received
by
each
bee
is
less
than
a
'threshold'
dose
some-
where
between
700
and
900
jig
of
SBTI.
Given
that
nectar
would
be
unlikely
to
con-
tain
SBTI,
pollen
of
a
transgenic
plant
would
have
to
be
expressing
SBTI
as
4
%
of
total
protein
or
more
to
have
the
possibility
of
adversely
affecting
bee
longevity.
It
seems
unlikely
that
transgenic
SBTI-plants
will
express
SBTI
at
4
%
or
more
of
total
protein
because
transgenic
SBTI-tobacco
plants
have
been
shown
to
reduce
the
growth
of
the
pest
S.
'num
when
expressing
SBTI
at
0.2-0.4
%
of
total
protein
and
to
kill
this
insect
at
0.4-1
%
(E.P.J.
Burgess,
unpub-
lished
data).
In
conclusion,
it
is
unlikely
that
trans
-
genic
white
clover
expressing
either
CrylBa
or
SBTI
for
protection
against
pest
attack
will
be
toxic
to
adult
honey
bees.
However,
further
investigation
is
required
to
establish
whether
the
same
lack
of
toxicity
would
be
observed
in
bees
kept
under
field
conditions.
Furthermore,
sub
-lethal
effects,
particularly
on
foraging
behaviour,
were
not
studied
here
but
could
have
significant
effects
on
colony
survival.
Inadvertent
changes
in
clover
phe-
notype
that
may
indirectly
affect
bees,
such
as
changes
in
flower
morphology,
nectar
volume
or
nectar
concentration,
may
be
avoided
by
careful
testing
and
selection
of
the
transgenic
lines
prior
to
field
release.
ACKNOWLEDGEMENTS
We
thank
Robert
Simpson
(Horticulture
and
Food
Research
Institute
of
New
Zealand
Ltd)
for
help
in
preparation
and
activation
of
CrylBa
and
for
comments
on
this
manuscript,
and
Anne
Gunson
(Horticulture
and
Food
Research
Institute
of
New
Zealand
Ltd)
and
Chris
Triggs
(Univer-
sity
of
Auckland,
New
Zealand)
for
statistical
advice
and
analyses.
This
work
was
funded
under
sub
-contract
to
the
New
Zealand
Pastoral
Agri-
culture
Research
Institute
Ltd
by
the
New
Zealand
Public
Good
Science
Fund,
programme
C10639.
Effects
of
Bt
toxins
and
SBTI
on
bees
471
Résumé
Effets
d'une
toxine
de
Bacillus
thuringiensis,
de
deux
formulations
de
biopesticide
a
base
de
Bacillus
thurin-
giensis
et
d'un
inhibiteur
de
trypsine
soja
sur
la
survie
et
la
prise
alimentaire
de
l'abeille
mellifere
(Apis
mellifera
L.).
Des
abeilles
fraichement
&loses
ont
ete
enca-
gees
(40/cage)
et
ont
recu
un
sirop
de
sucre
(60
%
masse/volume)
et
de
]'eau
ad
libitum
et
une
nourriture
a
base
de
pollen
et
de
divers
additifs.
Les
additifs
utilises
etaient
les
suivants
:
la
8-endotoxine
Cry
1Ba
puri-
fit&
et
activee,
venant
de
Bacillus
thurin-
giensis
Bt4412
(Bt)
(1,
0,25
et
0,025
%
masse/masse),
deux
preparations
biopesti-
cides
de
Bt
:
Dipel
2X
(1
et
0,25
%)
et
Foray
48B
(0,25
%)
et
]'inhibiteur
de
trypsine
soja
de
Kunitz
(SBTI)
(1,
0,5
ou
0,05
%).
Les
concentrations
de
CrylBa
et
de
SBTI
ont
ete
choisies
de
facon
a
correspondre
a
trois
niveaux
d'expression
dans
les
plantes
trans-
geniques
:
un
niveau
eleve
non
realiste
et
deux
concentrations
que
l'on
peut
s'atten-
dre
a
trouver
dans
les
plantes
exprimant
ces
proteines
a
des
niveaux
efficaces
contre
les
insectes
depredateurs.
Les
concentrations
en
Cry
1Ba
equivalent
a
des
niveaux
d'expression
dans
la
plante
de
4,
1,
et
0,1
%
de
la
proteine
totale
et
celles
de
SBTI
a
4,
2
et
0,2
%
de
la
proteine
totale.
Le
traitement
au
Dipel
a
I
%
equivaut
a
4
%
de
la
pro-
teine
totale
et
se
situe
bien
au-dessus
des
niveaux
de
pulverisation
recommandes
pour
ce
biopesticide.
Les
traitements
au
Dipel
et
au
Foray
a
0,25
%
equivalent
a
1
%
de
la
proteine
totale
et
s'approchent
de
la
DL
50
pour
un
pesticide
qui
est
«
virtuellement
non
toxique
»
pour
les
abeilles.
Trois
blocs
de
neuf
traitements
et
un
temoin
ont
ete
mis
en
place
(30
cages
au
total).
Les
abeilles
ont
ete
maintenues
en
etuve
a
32
°C.
Elles
ont
recu
les
aliments
durant
Sept
jours,
puis
la
nourriture
temoin
sans
additifs
pendant
le
reste
de
leur
vie.
La
mortalite
des
abeilles
dans
les
cages
a
ete
verifiee
et
les
recipients
de nourriture
a
base
de
pollen
ont
ete
peses
toutes
les
12
h
durant
cinq
jours,
puis
chaque
jour
durant
les
quatre
jours
suivant
et
trois
fois
par
semaine
jusqu'
a
ce
que
les
abeilles
meurent.
La
survie
des
abeilles
n'
a
pas
ete
affect&
par
les
traitements
et
la
nourriture
avec
additif
a
ete
consommee
au
meme
rythme
que
la
nourriture
temoin
pour
tour
les
traitements,
sauf
celui
au
Dipel
a
1
%
(figures
lb
et
2b).
Une
seconde
experience
a
montre
que
les
abeilles
totalement
privees
de
nourriture
a
base
de
pollen
survivaient
moms
Iongtemps
que
celles
ayant
recu
la
nourriture
temoin
(figure
Id).
Nous
concluons
qu'il
est
peu
probable
que
les
plantes
transgeniques
exprimant
CrylBa,
SBTI
ou
les
biopesticides
utilises
aux
doses
recommandees
presentent
une
toxicite
directe
pour
les
abeilles
adultes.
Neanmoins
it
serait
souhaitable
de
proceder
a
d'autres
etudes
sur
des
abeilles
en
conditions
natu-
relies
et
d'examiner
les
effets
sub-letaux.
Inra/DIB/AGIB/Elsevier,
Paris
Apis
mellifera
/
Bacillus
thuringiensis
l
toxine
CrylBa
/
inhibiteur
de
trypsine
soja
Kunitz
/
plante
transgenique
/
toxicite
Zusammenfassung
Wirkung
eines
Gifts
von
Bacillus
thuringiensis,
Wirkung
zweier
Formulierungen
von
Biopestizi-
den
auf
der
Basis
von
B.
thuringiensis
sowie
eines
Trypsinhemmers
aus
Soja-
bohnen.
Frisch
geschliipfte
Honigbienen
(Apis
mellifera
L.)
wurden
in
kleine
Kafige
iiberfiihrt
(40
Bienen
pro
Kafig),
die
Zucker-
wasser
(60
%
mN)
und
Wasser
ad
libitum
enthielten.
Zusatzlich
wurde
ein
auf
Pollen
basierendes
Futter
mit
unterschiedlichen
Zusatzen
geboten.
Als
Zusiitze
wurden
gege-
ben:
a)
gereinigtes
und
aktiviertes
Cry
1Ba
S-Endotoxin
vom
Bacillus
thuringiensis
Bt4412
(Bt)
(1,
0,25
und
0,025
%
w.w),
b)
Praparate
von
Bt
Biopestiziden,
Dipel
2X
(1
und
0,25
%)
und
Foray
48B
(0,25
%),
und
c)
Kunitz
Sojabohnen
Trypsin
Hem-
mer
(SBTI)
(1,
0,5
oder
0,05
%).
Die Kon-
zentrationen
von
CrylBa
und
SBTI
wurden
entsprechend
3
Starken
der
Expression
in
transgenen
Pflanzen
gewahlt:
ein
unreali-
stisch
hohes
Niveau
und
2
Konzentratio-
nen,
wie
sie
in
Pflanzen
erwartet
werden
472
L.A.
Malone
et
al.
konnten,
wenn
sie
gegen
Insektenbefall
wirksam
sein
sollen.
Die
CrylBa
Konzen-
trationen
entsprechen
den
Expressionsni-
veaus
von
Pflanzen
mit
4,
1
und
0.1
%,
die
von
SBTI
entsprechen
4,
2
und
0,2
%
des
Gesamtproteins.
Die
1
%
Dipel
Behandlung
entspricht
4
%
des
Gesamtproteins
und
liegt
weit
Ober
der
empfohlenen
Spriihmenge
fur
dieses
Biopestizid.
Die
0,25
%
Dipel
und
Foray
Behandlungen
entsprechen
1
%
des
Gesamtproteins
und
liegen
nahe
dem
Mini-
mum
der
LD
50
von
Pestiziden
,
die
fur
Honigbienen
als
praktisch
ungiftig
gelten.
Drei
Blocke
dieser
9
Behandlungen
und
eine
Kontrolle
wurden
durchgefuhrt
(insgesamt
30
Kafige).
Die
Bienen
wurden
im
Brut-
schrank
bei
32
°C
gehalten.
Sie
wurden
7
Tage
lang
mit
den
Zusatzen
geftittert,
danach
erhielten
sie
fur
den
Rest
ihres
Lebens
Kontrollfutter
ohne
Zusatz.
In
den
Kafigen
wurde
der
Totenfall
kontrolliert.
Die
Bellalter
mit
dem
Pollenfutter
wurden
5
Tage
lang
in
12stiindigem
Abstand,
wei-
tere
4
Tage
taglich
und
danach
3
mal
pro
Woche
gewogen,
bis
alle
Bienen
tot
waren.
Die
Uberlebensdauer
der
Bienen
blieb
unbe-
einflusst,
und
das
Futter
wurde
bei
alien
Behandlungen
in
gleichen
Mengen
wie
bei
den
Kontrollen
aufgenommen,
mit
Aus-
nahme
von
1
%
Dipel,
bei
dem
beides,
Uberlebensrate
und
Futterverbrauch,
signi-
fikant
reduziert
waren
(Abbildung
lb
und
2b).
Ein
2.
Versuch
zeigte,
dass
Bienen
ohne
Pollenfutter
schlechter
tiberlebten
als
die
mit
Kontrollfutter
(Abbildung
Id).
Wir
schliePen
daraus,
dass
transgene
Pflanzen,
die
CrylBa
oder
SBTI
erzeugen,
oder
auch
die
Bt
Biopestizide,
die
nach
Vorschrift
angewendet
werden,
wahrscheinlich
keine
direkte
giftige
Wirkung
auf
adulte
Bienen
haben.
Trotzdem
waren
weitere
Studien
mit
Bienen
unter
Feldbedingungen,
unter
Ein-
schluss
von
Prtifungen
von
sublethalen
Wir-
kungen,
wiinschenswert.
©
Inra/DIB/AGIB/
Elsevier,
Paris
Honigbienen
/
Bacillus
thuringiensis
I
CrylBa
Gift
/
Kunitz
Sojabohnen
Trypsinhemmer
/
krankheitsresistente
transgene
Pflanzen
[I1
REFERENCES
Baker
H.G.,
Baker
I.,
Intraspecific
constancy
of
fl
oral
nectar
amino
acid
complements,
Bot.
Gaz.
138
(1977)
183-191.
12]
Benedict
J.H.,
Sachs
E.S.,
Altman
D.W.,
Deaton
W.R.,
Kohel
R.J.,
Ring
D.R.,
Berbe-
rich
S.A.,
Field
performance
of
cottons
express-
ing
transgenic
CrylA
insecticidal
proteins
for
resistance
to
Heliothis
virescens
and
Helicov-
erpa
zea
(Lepidoptera:
Noctuidae),
J.
Econ.
Ento-
mol.
89
(1
996)
230-238.
131
Buckner
C.H.,
Gochnauer
T.A.,
McLeod
B.B.,
Prebble
M.L.,
The
impact
of
aerial
spraying
of
insecticides
on
bees,
in:
Aerial
Control
of
Forest
Insects
in
Canada,
Department
of
the
Environ-
ment,
Ottawa,
Canada,
1975,
pp.
276-279.
[41
Burgess
E.P.J.,
Main
C.A.,
Stevens
P.S.,
Chris
-
teller
J.T.,
Gatehouse
A.M.R.,
Laing
W.A.,
Effects
of
protease
inhibitor
concentration
and
combinations
on
the
survival,
growth
and
gut
enzyme
activities
of
the
black
field
cricket,
Teleogrvllus
commodus.
J.
Insect
Physiol.
40
(1994)
803-81
I.
Burgess
E.P.J.,
Malone
L.A.,
Christeller
J.T.,
Effects
of
two
proteinase
inhibitors
on
the
diges-
tive
enzymes
and
survival
of
honey
bees
(Apis
mellifera),
J.
Insect
Physiol.
42
(1996)
823-828.
Cantwell
G.E..
Shieh
T.R.,
Certan
a
new
bac-
terial
insecticide
against
the
greater
wax
moth,
Galleria
mellonella
L.,
Am.
Bee
J.
121
(1981)
424-426,430-431.
[7]
Cantwell
G.E.,
Knox
D.A.,
Lehnert
T.,
Michael
A.S.,
Mortality
of
the
honey
bee,
Apis
mellifera,
in
colonies
treated
with
certain
bio-
logical
insecticides,
J.
Invertebr.
Pathol.
8
(1966)
228-233.
[81
Cantwell
G.E.,
Lehnert
T.,
Fowler
J.,
Are
bio-
logical
insecticides
harmful
to
the
honey
bee?,
Am.
Bee
J.
1
12
(1972)
255-258.
[91
Celli
G.,
Conditions
for
the
survival
of
the
honey
bee
in
present
agricultural
systems,
Ann.
Accad.
Naz.
Agric.
Bologna
94
(1974)
395-411.
1
101
Crane
E.,
Walker
P.,
The
Impact
of
Pest
Man-
agement
on
Bees
and
Pollination,
Tropical
Development
and
Research
Institute,
London,
1983.
1111
Crane
E.,
Walker
P.,
Pollination
Directory
for
World
Crops,
International
Bee
Research
Asso-
ciation,
London,
1984.
[12]
Georgevitis
R.P.,
Comparison
of
results
in
the
control
of
Thaumetopoea
pityocampa
Schiff.
with
Dimilin,
Decis,
Bactospeine
and
Thuricide
HP,
Anakoinoseis
Idrumaton
Dasikon
Ereuon
7
(1979)
7-34
(in
Greek).
[131
Greenplate
J.,
Response
to
reports
of
early
dam-
age
in
1996
commercial
Bt
transgenic
cotton
(BollgardTM)
plantings,
Soc.
Invertebr.
Pathol.
Newslett.
29
(1997)
15-18.
151
[61
Effects
of
Bt
toxins
and
SBTI
on
bees
473
1
141
James
C.,
Krattiger
A.F.,
Global
Review
of
the
Field
Testing
and
Commercialization
of
Trans
-
genic
Plants:
1986
to
1995:
The
First
Decade
of
Crop
Biotechnology,
ISAAA
Briefs
No.
I
,
ISAAA,
Ithaca,
New
York,
1996.
1
151
Kalbtleisch
J.D.,
Prentice
R.L.,
Statistical
Anal-
ysis
of
Failure
Time Data,
John
Wiley,
New
York,
1980.
1
161
Kazakova
S.B.,
Dzhunusov
K.K.,
Protsenko
A.1.,
The
effect
of
Bitoxibacill
in
-202
on
certain
orchard
insects
in
the
Issyk-Kul'
depression,
Entomol.
Issled.
Kirgizia
10
(1975)
49-51
(in
Russian).
1
171
Krieg
A.,
Hassan
S.,
Pinsdorf
W.,
Wirkungsver-
gleich
der
Varietat
israelensis
mit
anderen
Vari-
etaten
des
Bacillus
thuringiensis
an
Nicht-Zielor-
gani
mem
der
Ordnung
Hymenoptera:
Tricho-
gramma
cacoeciae
and
Apis
mellifera,
Anz.
Schadlingskd.,
Pflanzenschutz,
Umweltschutz
53
(1980)
81-83.
1
181
Matheson,
A.,
Practical
Beekeeping
in
New
Zealand,
GP
Publications,
New
Zealand,
1997.
1
191
Perlak
F.J.,
Dcaton
R.W.,
Armstrong
T.A.,
Fuchs
R.L.;
Sims
S.R.,
Greenplate
J.T.,
Fis-
chhoff,
D.A.,
Insect
resistant
cotton
plants,
Biotechnology
8
(1990)
939-943.
1201
Pham-Delegue
M.H.,
Jouanin
L.,
L'abeille
et
les
plantes
transgeniques,
Rev.
Fr.
Apic.
574
(1997)
250-251.
121
1
Picard-Nizou
A.L.,
Pham-Delegue
M.H.,
Ker-
guelen
V.,
Douault
P.,
Marilleau
R.,
Olsen
L.,
Grison
R.,
Toppan
A.,
Masson
C.,
Foraging
behaviour
of
honey
bees
(Apis
mellifera
L.)
on
transgenic
oilseed
rape
(Brassica
napus
L.
var.
oleifera),
Transgenic
Res.
4
(1995)
270-276.
1221
Schmidt
J.0.,
Buchmann
S.L.,
Other
products
'of
the
hive,
in:
Graham
J.M.
(Ed.),
The
Hive
and
the
Honey
Bee,
Dadant
and
Sons,
Hamilton,
Illi-
nois,
1992,
pp.
927-988.
1231
Simpson
R.M.,
Burgess
E.P.J.,
Markwick
N.P.,
Bacillus
thuringiensis
8-endotoxin
binding
sites
in
two
Lepidoptera,
Wiseana
spp.
and
Epiphymv
postvittana,
J.
In
vertebr.
Pathol.
70
(1997)
136-142.
1241
Sims
S.R.,
Bacillus
thuringiensis
var.
kurstaki
[CryIA(c)I
protein
expressed
in
transgenic
cot-
ton:
Effects
on
beneficial
and
other
non
-target
insects,
Southwestern
Entomol.
20
(1995)
493-500.
[251
Tutkun
A.,
Inci
A.,
Yilmaz
B.,
Bal
arisi
(Apis
mellifera
L.)
kovanlarinda
zarar
yapan
Bala-
mumu
guvesi
(Galleria
mellonella
L.)
larvalar-
ina
Bacillus
thuringiensis
preparatlarinin
etkisi
uzerinde
arastirmalar,
in:
Turkiye
I.
Unto-
moloji
Kongresi
Bildirileri,
13-16
Ekim
1987,
Ege
Universitesi,
Bornova,
1987,
pp.
585-594.
1261
van
Eaton
C.,
How
much
is
honey
bee
pollina-
tion
worth?,
N.
Z.
Beekeeper
1
(1994)
23.
1271
Vandenberg
J.D.,
Safety
of
four
ento-
mopathogens
for
caged
adult
honey
bees
(Hymenoptera:
Apidae),
J.
Econ.
Entomol.
83
(1990)
755-759.
1281
Vandenberg
J.D.,
Shimanuki
H.,
Two
commer-
cial
preparations
of
the
beta
exotoxin
of
Bacillus
thuringiensis
influence
the
mortality
of
caged
adult
honey
bees,
Apis
mellifera
(Hymenoptera:
Apidae),
Environ.
Entomol.
15
(1986)
166-169.
1291
Vandenberg
J.D.,
Shimanuki
H.,
Viability
of
Bacillus
thuringiensis
and
its
efficacy
for
larvae
of
the
greater
wax
moth
(Lepidoptera:
Pyrali-
dae)
following
storage
of
treated
combs,
J.
Econ.
Entomol.
83
(1990)
760-765.
1301
Verma
S.K.,
Studies
on
the
control
of
greater
wax
moth,
Galleria
mellonella
L.
in
Apis
cer-
ana
F.
colonies
with
the
biological
insecticide,
Dipel,
Indian
Bee
J.
57
(1995)
121-123.
131
1
Winston
M.L.,
The
Biology
of
the
Honey
Bee,
Harvard
University
Press,
Cambridge,
Mas-
sachusetts,
1987.