Babendreier D, Kalberer N, Romeis J, Fluri P, Bigler F. Pollen consumption in honey bee larvae: a step forward in the risk assessment of transgenic plants. Apidologie

Abstract

In order to assess the potential impacts of transgenic plants on larvae of the honey bee, Apis mellifera, information on pollen consumption is needed. We here report on experiments that were conducted with small bee colonies kept in field cages (8 × 14 m) containing only flowering maize plants as protein source. Fully grown worker bee larvae were found to contain between 1720 and 2310 maize pollen grains in their gut before defecation, corresponding to 1.52–2.04 mg of pollen consumed per larva. On average, 74.5% of pollen grains were completely digested while 23.3% were partially digested and 2.2% remained undigested. Our data indicate that the contribution of the protein by directly feeding larvae with pollen is less than 5% in relation to the total amount of protein necessary for complete larval development. We suggest that our measurement for pollen consumption should be taken into account when establishing dose regimes to assess the risk that transgenic plants pose for honey bee larvae.

Full text

293
Apidologie 35 (2004) 293–300
© INRA/DIB-AGIB/ EDP Sciences, 2004
DOI: 10.1051/apido:2004016
Original article
Pollen consumption in honey bee larvae: a step forward
in the risk assessment of transgenic plants
Dirk BABENDREIERa*, Nicole KALBERERa, Jörg ROMEISa, Peter FLURIb,
Franz BIGLERa
aAgroscope, FAL Reckenholz, Swiss Federal Research Station for Agroecology and Agriculture,
Reckenholzstr. 191, 8046 Zürich, Switzerland
bAgroscope, FAL Liebefeld-Posieux, Swiss Federal Dairy Research Station, Swiss Bee Research Centre,
Schwarzenburgstr. 161, 3003 Bern, Switzerland
(Received 14 April 2003; revised 6 June 2003; accepted 11 August 2003)
Abstract – In order to assess the potential impacts of transgenic plants on larvae of the honey bee, Apis
mellifera, information on pollen consumption is needed. We here report on experiments that were conducted
with small bee colonies kept in field cages (8 × 14 m) containing only flowering maize plants as protein
source. Fully grown worker bee larvae were found to contain between 1720 and 2310 maize pollen grains
in their gut before defecation, corresponding to 1.52–2.04 mg of pollen consumed per larva. On average,
74.5% of pollen grains were completely digested while 23.3% were partially digested and 2.2% remained
undigested. Our data indicate that the contribution of the protein by directly feeding larvae with pollen is
less than 5% in relation to the total amount of protein necessary for complete larval development. We
suggest that our measurement for pollen consumption should be taken into account when establishing dose
regimes to assess the risk that transgenic plants pose for honey bee larvae.
Apis mellifera / pollen consumption / transgenic plant / agrochemical / maize / risk assessment
1. INTRODUCTION
Honey bees are of great ecological and eco-
nomic importance as pollinators of many crop
and wild plants (Free, 1993). Therefore, novel
plant protection methods have to be evaluated
for potential effects on this group of beneficial
insects. The studies conducted to evaluate the
effects of pesticides have recently been reviewed
by Devillers and Pham-Delègue (2002) while
Malone and Pham-Delègue (2001) have sum-
marised studies carried out to evaluate the effects
of transgenic plants. Bees in general may be
exposed to transgene products or agrochemicals
via pollen or nectar. Pollen is the main source of
protein for bees (Crailsheim, 1990) while nectar
is known to contain mainly sugars and rather low
concentrations of proteins (Baker and Baker,
1986; Carter et al., 1999). As the toxins produced
by transgenic plants are proteins which have not
yet been detected in nectar (Malone and Pham-
Delègue, 2001), potential risks for honey bees
due to transgene products will most li kely be due
to pollen feeding. Regarding agrochemicals,
potential risks may occur due to systemic insec-
ticides and to a lesser extent, to insecticides
sprayed onto flowering crops (Kubik et al., 1999;
Villa et al., 2000; Tasei, 2001). Potential risks for
larvae may not only occur by pollen feeding but
also by nectar feeding because agrochemicals
present in nectar may be delivered to larvae
through the contents from the honey sac (Davis
and Shuel, 1988). In the present paper, we focus
on pollen consumption of honey bee larvae aim-
ing to examine the effects of exposure to trans-
gene products. Nevertheless, we would like to
* Corresponding author: dirk.babendreier@fal.admin.ch

294 D. Babendreier et al.
stress the relevance of this contribution for risk
assessment of agrochemicals that are present in
pollen.
Numerous studies have been carried out indi-
cating that pollen from a wide range of pl ant spe-
cies is collected during each bee foraging season
(Alfonsus, 1933; Wille et al., 1985a, b). How-
ever, a study carried out in the low lands of
Northern Switzerland showed that only five
plant species including maize accounted for
51% of the total pollen collected annually (Wille
and Wille, 1983; Wille et al., 1985a). Therefore,
honey bees in some regions may well be
exposed to the pollen of transgenic lepidopteran
resistant maize containing genes from the soil
bacterium Bacillus thuringiensis Berliner (Bt-
maize) if this crop is deployed on a large scale.
In order to assess the risks that transgenic plants
pose on non-target organisms, both the exposure
and the hazard have to be considered. Regarding
honey bees, the hazard is basically determined
by the toxicity of the transgene product while the
exposure is dependent on the amount of pollen
that adults and larvae ingest. Bioassays have
been carried out to test potential effects of trans-
genic plants or transgene products on adult
honey bees (Malone and Pham-Delègue, 2001)
but few studies have considered the larval stages
(Malone et al., 2002; Brodsgaard et al., 2003).
In addition, studies including the larvae lacked
a detailed knowledge on how much pollen is
ingested during development.
During the first 3–4 days of larval develop-
ment, honey bee larvae are fed with jelly pro-
duced in the hypopharyngeal glands of adult
worker bees (Crailsheim, 1990). The composi-
tion of this jelly depends on whether the larva
is being raised to become a queen, a worker or
a drone, but is almost free of pollen. During
subsequent development, larvae are fed with
modified jelly that is less rich in protein but
rich in sugars (honey) and contains some pol-
len (Planta, 1888; Kunert and Crailsheim,
1988; Malone et al., 2002). However, the
amount of pollen that honey bee larvae are
exposed to has never been investigated in
detail. The only study on honey bees dealing
with that question experimentally was carried
out by Simpson (1955) who measured the
amount of pollen in the guts of larvae that have
fed on a diverse mixture of plants.
We here aimed to quantify the amount of
pollen that honey bee larvae ingest during
development by exposing them exclusively to
maize pollen under near-field conditions. This
bioassay makes use of the fact that honey bee
larvae do not defecate until development of the
larva is finished. Such a measurement for pol-
len consumption will allow for a more realistic
estimation of the exposure of honey bee larvae
to transgene products expressed in pollen and
insecticide residues. It should furthermore be
used to determine meaningful amounts of tox-
ins in future feeding experiments to assess
potential risks of genetically modified plants
or agrochemicals to honey bee larvae.
2. MATERIALS AND METHODS
2.1. Field cages
Large field cages (8 × 14 m, height in the centre
3.5 m, mesh size 1 mm) were set up at the Swiss
Federal Research Station for Agroecology and
Agriculture in Zurich, Switzerland. Eight rows of
maize (variety Monumental) with 80 cm space
between rows were sown resulting in a total of
about 800 plants per cage. Maize was sown at three
different times in biweekly intervals beginning on
May 8 in order to set up three replicates over time.
Fertilisation and weed control were done in agree-
ment with standard agricultural practices in Swit-
zerland. At the beginning of flowering, the tent was
installed and two colonies of Apis mellifera L. were
introduced per cage as soon as 5–10% of the maize
anthers were open. From the total of six bee colo-
nies, only five yielded brood and were subjected to
further analysis.
2.2. Bee colonies
Bee hives were set up in the field cages close to
the soil on small posts surrounded by paraffin to
protect them against ants. They were also protected
against rain and partly against direct sunlight by a
brick (25 × 40 cm) put on top of the hive. Each hive
contained one queen and about 1000 worker bees of
varying age that had ad libitum access to a protein-
free sugar patty and water.
The hives were composed of three combs of 10 ×
10 cm embedded into a Styrofoam box (Apidea).
At the start of the experiment, one comb was empty
while the second one contained predominantly open
and capped brood. The third comb contained eggs <
48 h old laid by the queen used in the experimental
hive. Larvae from this comb hatched shortly after
introducing the hives into the field cage. Larvae were
sampled for experimental purposes from this comb

Pollen consumption in honey bee larvae 295
only. At the beginning of the experiment, all combs
were checked for pollen stores and the few cells that
contained pollen were filled with liquid bee wax to
prevent the bees from gaining access to this pollen.
2.3. Sampling of bee larvae
When larvae were fully grown, i.e. five to six
days after the hives had been introduced into the
cages, the third comb of each colony was checked
for capped cells. If such cells were found, the comb
was temporarily transferred to the laboratory and all
larvae that were partly or completely capped, but
had not yet started cocoon formation and defeca-
tion, were removed, weighed on a microbalance
(Mettler Toledo M × 5, d = 1 µg; ± 2 µg) and stored
at 4 °C in 1.5 mL Eppendorf tubes. Afterward, the
comb was placed back into the hive. This procedure
was repeated three times a day until about 30 larvae
were obtained per colony. In addition, we found
drone larvae in one of the hives (colony 5) which
were also sampled the same way. The experiment
lasted for a maximum of 8 days during which time
bees had abundant access to pollen. This was also
indicated by numerous cells filled with maize pol-
len in the course of the experiment. Fresh weights of
all worker bee larvae were compared using single
factor ANOVA among the five colonies. Means
were subsequently separated using the Tukey HSD
test.
2.4. Counting of pollen
Larvae were dissected and each complete gut was
frozen at –18 °C in 200 µL 0.5 M sugar water. For
pollen counts, each thawed gut was homogenised
with a pestle and a10 µl aliquot of the homogenate
transferred onto a counting device and covered with
a coverslip (18 × 18 mm). The number of pollen
grains in the 10 µl sample was determined by micro-
scopic examination (magnification 50 ×). Complete
pollen grains and fragments larger than half of a pol-
len grain were counted (cf. Crailsheim et al., 1992).
Three times 10 µL aliquots were taken from each
sample and counted. The mean was multiplied by 20
to estimate the number of pollen grains in a single
larval gut. Between 26 and 35 larvae were analysed
per colony.
In order to estimate the weight of a single pollen
grain, six samples of fresh pollen pooled from ten
maize plants each (variety Monumental) were col-
lected by hand from a field that bordered the field
cages. These were weighed immediately and then
all pollen grains (300–720 pollen grains per sample)
were counted under a microscope at 50 × magnifi-
cation. Dividing the weight of the samples by the
number of grains we found the weight (± SE) of a
single maize pollen grain to be 882 ± 2.2 ng.
Mean weight of pollen grains in the guts of larvae
among the three different field cages set up on con-
secutive dates were compared using single factor
ANOVA. Since no effects of the cages were found
(F2,147 = 1.24, P = 0.293), a single factor ANOVA
was conducted on the mean weight of pollen grains
in the guts of larvae among the five colonies. Regres-
sion analysis was carried out to reveal a potential
relationship between larval weight and the weight of
pollen grains in the gut.
2.5. Digestion of pollen
To estimate the proportion of pollen grains that
were digested by the larvae, 10 µl of each gut sample
were checked under a transmission microscope at
100 × magnification. For each sample, 50 randomly
selected pollen grains were classified into one of
three groups: fully digested (at most an estimated
content of 10%), shrunken, i.e. all grains that con-
tained some material (see Crailsheim et al., 1992)
and undigested (no detectable difference to fresh pol-
len). To illustrate our classification, a few typical
images are shown in Figure 1.
3. RESULTS
The mean weight of one fully grown honey
bee larvae varied from 132 to 155 mg and sig-
nificant differences in the weight of worker bee
larvae among the five colonies were detected
(F4,145 = 15.22, P < 0.001, Fig. 2A). Between
1720 and 2310 maize pollen grains were found
in the guts of worker honey bee larvae. Based
on the weight of a single pollen grain, we cal-
culated the amount of pollen consumed by one
honey bee worker larvae to be between 1.52 ±
0.108 and 2.04 ± 0.104 mg (Fig. 2B) with sig-
nificant differences among the five bee colo-
nies (F4,145 = 3.44, P = 0.010). Significantly
more pollen grains were found in larval guts of
colony 1 compared to colony 5 (Tukeys HSD
test, P = 0.021).
There was no relationship between the
number of pollen grains consumed and larval
weight if whole colonies were studied (regres-
sion analysis, F1,3 = 0.34, P = 0.60). We found
significant relationships between the number of
pollen grains consumed and the weight of indi-
vidual larvae in colonies 2, 3, 4 but not in col-
onies 1 and 5. As shown in Table I, the coeffi-
cients of determination (R2) were generally low
with a maximum of 0.22.

296 D. Babendreier et al.
The degree of digestion of maize pollen by
honey bee larvae was in the same range for the
five colonies investigated (Fig. 3). On average,
74.5% of pollen grains were completely digested
while 23.3% were partially digested and 2.2%
remained undigested.
4. DISCUSSION
The aim of this study was to assess the total
amount of pollen consumption by honey bee
larvae in colonies restricted to foraging on
maize plants and to estimate the contribution
of pollen relative to the total amount of protein
larvae receive during development. We have
shown that worker honey bee larvae under our
experimental conditions were fed with a total
amount of 1.5 to 2 mg maize pollen during
their complete development. In the present
study, drone larvae were 39% heavier than
worker bee larvae and their gut contained 36%
more pollen, suggesting that drone larvae are
fed with similar quantities of pollen compared
to worker larvae. However, because the drone
brood is normally not produced during August
and those we measured may have been under-
weight, these results should be regarded as
preliminary.
As the colonies in our experiments were
allowed to collect pollen only from maize
plants, it can not be ruled out that different
Figure 1. Sample images showing the different
degrees of digestion of maize pollen grains in the
guts of honey bee larvae and the classification we
used in this study: (A) fully digested pollen, (B)
shrunken pollen and (C) undigested pollen.
Images were made at 100 × magnification under
a transmission microscope, the scale bar repre-
sents 100 µm.

Pollen consumption in honey bee larvae 297
quantities of pollen may be fed to honey bee
larvae under field conditions where bees may
forage on a variety of plant species. Such dif-
ferences may relate to the quality (N-content,
amino-acid composition) as well as to the size
of different pollen grains. Most pollen grains
are smaller than pollen from maize (Stanley
and Linskens, 1974, p. 28), thus small larvae
may not be able to uptake the relatively large
maize pollen grains but may easily ingest
smaller ones. If such an effect would occur,
pollen consumption of bees fed only larger
pollen grains, such as maize, may be underes-
timated.
Our study has further shown that by far the
most pollen grains are digested which is in gen-
eral agreement with other studies on honey bee
larvae (Simpson, 1955) and adults (Schmidt
and Buchmann 1985; Crailsheim et al., 1992).
This suggests that larvae were able to utilise
most of the protein that maize pollen contained.
In addition to pollen, however, honey bee lar-
vae are fed with worker jelly which is provided
by nurse bees. Although the relative protein
contributions of pollen and worker jelly are not
exactly known as yet, several studies have
shown that the latter is a substantial source of
protein and needed for the protein balance of
the larvae (see Haydak, 1970 for a review). If
one is aiming to assess potential risks of trans-
gene products for honey bee larvae, informa-
tion is needed on whether the worker jelly con-
tains any toxins expressed by the transgenic
plant. This appears to be unlikely because such
proteins would have to pass the insects guts and
become incorporated into the hypopharyngeal
glands. Even for insecticides that are much
smaller molecules, only traces were found in
the hypopharyngeal glands of worker bees
when fed the compound in a sugar solution
(Davis and Shuel, 1988). Consequently, direct
feeding on pollen may be the only significant
source of transgene products for honey bee
larvae.
Earlier estimations revealed that about 160 mg
of pollen is necessary to rear a honey bee
(Alfonsus, 1933; Wille et al., 1985b). How-
ever, this estimate includes all the protein that
Tab le I. Results of a regression analysis to test for
a relationship between the weight of the larvae and
the number of pollen consumed for each of the
colonies.
R2Fdf P
Colony 1 0.046 1.59 1, 33 0.217
Colony 2 0.201 6.04 1, 24 0.022
Colony 3 0.216 8.82 1, 32 0.0056
Colony 4 0.176 5.55 1, 26 0.026
Colony 5 0.00032 0.008 1, 25 0.929
Figure 2. (A) Mean weight of fully grown larvae
(+ SE) and (B) mean consumption of maize pollen
(+ SE) per honey bee larva for the five bee colonies
investigated (N = 26–35 larvae). Different letters
indicate significant differences among colonies
(Tukeys HSD test, P < 0.05).
Figure 3. Percentage of digested maize pollen
grains in the guts of honey bee larvae for the five
bee colonies (N = 26–35 larvae).

298 D. Babendreier et al.
adults need for their own turn over and signif-
icant amounts of protein is needed for adults
during both summer and winter (Schmidt
and Buchmann, 1985; Crailsheim, 1990). We
believe that the amount of protein, and thus
pollen, needed for development of honey bee
larvae can be best estimated from the total
amount of nitrogen a larva contains after fin-
ishing development. Data by Haydak (1959)
and Imdorf et al. (1998) show that a single pupa
of A. mellifera contains 1.87 mg and 1.85 mg
nitrogen, respectively. Since defecation has
occurred already at this stage, a loss of nitro-
gen of 20% has to be taken into account
(Schmidt and Buchmann, 1985). Based on the
conversion factor of nitrogen to protein of 6.25
(Maynard and Loosli, 1969) and the protein
content of maize pollen of 16.7% (variety
Monumental, unpublished data), we calculate
that approximately 86 mg maize pollen is nec-
essary to rear a single honey bee larva. A slightly
lower amount of pollen is needed if the calcu-
lation is based on the average protein content
of pollen of 20% (Imdorf et al., 1998). Our
data indicate that the contribution of the pro-
tein by directly feeding larvae with pollen is
less than 5% in relation to the total amount of
protein necessary for complete larval develop-
ment. Thus pollen constitutes only a minor
part of the protein supply of honey bee larvae.
This is very much in agreement with findings
of Simpson (1955) who estimated from a large
sample of dissected honey bee larvae that on
average 0.078 mg N (corresponding to 2.2 mg
pollen) was present in the guts of larvae which
were fed with pollen from different plants.
Focussing on a small sub-sample where bees
collected virtually only red clover pollen,
Simpson (1955) calculated that 5.4 mg pollen
were present in larval guts which still is in
qualitative agreement with our findings. Our
data and the results of Simpson (1955) may
indicate that a low pollen consumption rate is
a general feature in honey bee larvae. The data
furthermore indicate that the larval stages of
the honey bee are less exposed to transgene
products than the adults as the latter uptake
large quantities of pollen not only for their
own requirements but also for the rearing of
the brood.
From the measurements obtained in the
present study, it is possible to calculate the
amount of e.g. Bt-toxin that a larva would ingest
during development if fed solely with pollen
from a transgenic plant. For example, pollen
from Bt-maize ‘Event 176’ contains a relatively
high amount of the Bt-toxin (5–11 µg Cry1Ab/g
fresh weight pollen) as the cry1Ab gene is
expressed under a green leaf tissue- and a pol-
len-specific promoter (Fearing et al., 1997).
Based on this it can be calculated that a honey
bee larva would ingest a maximum of 10–22 ng
of Cry1Ab when fed exclusively on Bt-maize
pollen (Event 176). This amount is not toxic for
honey bees according to published information
(Malone and Pham-Delègue, 2001). Other com-
mercialised Bt-maize varieties (Event Bt11 and
Event MON 810) contain even less of the trans-
gene product in the pollen as it is expressed
under a constitutive promoter (CaMV 35S) that
has little activity in the pollen. Thus, ingestion
of Bt-toxin is reduced by nearly two orders of
magnitude if larvae are fed with pollen of the
latter two varieties (Sears et al., 2001).
However, we here like to stress that the same
calculation is not restricted to Bt-maize but could
be applied for any potential toxin present in pol-
len. The few studies carried out so far on poten-
tial risks of transgenic plants on honey bee larvae
have assumed that considerably higher amounts
of pollen are ingested by larvae. As our data indi-
cate a very low pollen consumption for larvae,
exposure to toxins produced by transgenic plants
has most probably been overestimated so far
(e.g. Malone et al., 2002; Brødsgaard et al.,
2003). Similarly, exposure to agrochemicals via
pollen has been overestimated in the few studies
dealing with potential risks for bee larval stages
(Villa et al., 2000). We suggest that our meas-
urement for pollen consumption should be taken
into account to establish more realistic dose
regimes in further risk assessment studies of
honey bee larvae.
ACKNOWLEDGEMENTS
We like to thank Hans Ueli Thomas (Zurich)
who provided the honey bees. We also thank Syn-
genta for the large field cages that we were allowed
to use for this study. The comments of Louise
Malone, Henrik Brødsgaard, Jean-Noël Tasei and
two anonymous reviewers on earlier versions of the
manuscript were very much appreciated. This study
was funded by the Swiss Agency for the Environ-
ment, Forests and Landscapes, project number
810.3189.004.

Pollen consumption in honey bee larvae 299
Résumé – La consommation de pollen par les
larves d’abeilles : une étape vers l’évaluation des
risques présentés par les plantes transgéniques.
En tant que pollinisateurs de nombreuses plantes
cultivées et sauvages, les abeilles domestiques
(Apis mellifera L.) ont un rôle très important du
point de vue écologique et économique. Les nou-
velles méthodes de protection des plantes, que ce
soit l’utilisation de pesticides ou de plantes transgé-
niques, doivent en conséquence être testées quant à
d’éventuels effets négatifs sur ces insectes auxiliai-
res. Puisque les abeilles ne consomment les pro-
duits de transgènes que via le pollen, des données
sur la quantité de pollen consommé par les larves et
par les adultes sont d’une extrême importance pour
l’évaluation des risques. De telles données man-
quant en particulier pour les larves, nous avons
mesuré la quantité de pollen qui est consommée par
des larves d’abeilles durant tout leur cycle de déve-
loppement. Pour cela les larves ont été exposées en
conditions semi-naturelles à du pollen de maïs uni-
quement.
On a semé du maïs (variété Monumental) sous trois
grandes cages et obtenu environ 800 pieds par cage.
Dès le début de la floraison du maïs, deux petites
ruchettes d’environ 1000 abeilles ont été apportées
sous chaque cage. Les abeilles ont été nourries ad libi-
tum avec une pâte sans protéines et de l’eau. Les
rayons ne contenaient aucune réserve de pollen afin
que les abeilles ne se nourrissent que de pollen de
maïs. Il était important pour cette expérience que les
larves ne puissent déféquer qu’après leur développe-
ment complet et donc que l’on puisse prélever des lar-
ves pleinement développées qui renfermaient encore
dans leur intestin la quantité de pollen consommé.
Dans chacune des cinq colonies qui ont élevé du cou-
vain, on a prélevé entre 26 et 35 larves dans les cel-
lules fraîchement operculées, puis on les a pesées et
disséquées. Les grains de pollen ont tous été comptés
et répartis en trois classes : entièrement digérés (au
maximum une teneur de 10 %), partiellement digérés
ou non digérés (aucun différence reconnaissable avec
le pollen frais, voir Fig. 1). Pour chaque colonie on
a trouvé en moyenne entre 1720 et 2310 grains de
pollen de maïs par intestin de larve d’abeille. Cela
correspond à une quantité de pollen comprise entre
1,52 ± 0,108 et 2,04 ± 0,104 mg (Fig. 2B). En
moyenne 74,5 % des grains de pollen étaient totale-
ment digérés, 23,3 % l’étaient partiellement et 2,2 %
pas du tout. Aucune relation n’a pu être établie entre
le poids des larves et le nombre de grains de pollen
dans l’intestin (Tab. I).
Nos résultats montrent que la contribution du pollen
à l’alimentation des larves en protéines durant tout
leur cycle de développement ne représente qu’une
petite partie (moins de 5 %). Ils montrent en outre
que jusqu’à présent l’exposition des larves d’abeilles
aux toxines produites par les plantes transgéniques
ou aux pesticides a été surévaluée. Nos données sur
la consommation de pollen devraient être utilisées
comme base pour de futures études de risques pour
les larves d’abeilles, afin de garantir que des quan-
tités réalistes de toxines sont utilisées dans les expé-
riences de nourrissement.
Apis mellifera / consommation de pollen / plante
transgénique / pesticide / maïs / estimation des
risques
Zusammenfassung – Pollenaufnahme von Larven
der Honigbiene: ein wichtiger Aspekt für die
Risikoanalyse von transgenen Pflanzen. Als Be-
stäuber von vielen Kultur- und Wildpflanzen sind
Honigbienen (Apis mellifera L.) sowohl ökologisch
als auch ökonomisch von grosser Bedeutung. Neue
Pflanzenschutzmassnahmen müssen daher auf mögli-
che schädliche Nebenwirkungen für Bienen getestet
werden. Dies betrifft sowohl Pestizide als auch trans-
gene Pflanzen. Da Bienen die transgenen Produkte nur
über den Pollen aufnehmen, sind Daten zur aufgenom-
menen Menge an Pollen sowohl für die Larven als
auch für die adulten Bienen äusserst wichtig für eine
Risikoanalyse. Da solche Daten insbesondere für die
Larven fehlen, haben wir in dieser Studie die Pollen-
menge quantifiziert, die von einzelnen Honigbienen-
larven während der gesamten Entwicklungsdauer
aufgenommen und verwertet wird. Hierzu wurden die
Larven unter Halbfreilandbedingungen ausschliess-
lich gegenüber Maispollen exponiert.
In grossen Feldkäfigen (n = 3) wurde Mais gesät
(Sorte Monumental), wobei insgesamt etwa 800
Pflanzen pro Käfig wuchsen. Sobald die Maispflan-
zen zu blühen begannen, wurden je zwei kleine Bie-
nenvölker mit ca. 1000 Bienen in den Käfigen eta-
bliert. Die Bienen wurden ad libitum mit einem
Protein-freien Futterteig sowie Wasser versorgt, hat-
ten jedoch keine Pollenvorräte in den Waben. Dies
stellte sicher, dass die Larven nur mit Maispollen
gefüttert wurden. Wichtig für diesen Versuch war
die Tatsache, dass die Larven der Honigbiene ihren
Kot erst nach vollendeter Entwicklung abgeben und
daher ausgewachsene Larven gesammelt werden
konnten, die noch die insgesamt aufgenommene
Pollenmenge im Darm enthielten. Für jedes der fünf
Völker, die Bruterfolg hatten, wurden zwischen 26
und 35 Larven aus den frisch verdeckelten Zellen
entfernt, gewogen und seziert. Alle Pollenkörner
wurden gezählt und in drei Klassen eingeteilt: voll-
kommen verdaut (maximal 10 % Inhalt), teilweise
verdaut oder unverdaut (kein erkennbarer Unter-
schied zu frischem Pollen, siehe Abb. 1). Gemittelt
über die Daten aus je einem Volk wurden zwischen
1720 und 2310 Maispollenkörner im Darm der ein-
zelnen Honigbienenlarven gefunden. Dies ent-
spricht zwischen 1,52 ± 0,108 und 2,04 ± 0,104 mg
Pollen (Abb. 2B). Durchschnittlich waren 74,5 % der
Pollenkörner vollkommen verdaut, während 23,3 %
teilweise und 2,2 % unverdaut waren. Zwischen dem
Larvengewicht und der Anzahl Pollenkörner im
Darm wurde keine Beziehung gefunden (Tab. I).
Unsere Daten zeigen, dass der Beitrag des Pollens an
der Proteinversorgung der Larven während der
gesamten Entwicklungsdauer nur einen kleinen Teil

300 D. Babendreier et al.
ausmacht (weniger als 5 %). Sie zeigen weiter, dass
bisher die Exposition der Honigbienenlarven gegenü-
ber den von transgenen Pflanzen produzierten Toxi-
nen oder auch Pestiziden überschätzt wurde. Unsere
Daten zum Pollenkonsum sollten als Grundlage für
weitere Risikostudien zu den Larven der Honigbie-
nen herangezogen werden um zu gewährleisten, dass
realistischere Toxinmengen in Fütterungsversuchen
verwendet werden.
Apis mellifera / Pollenaufnahme / transgene
Pflanzen / Pestizide / Mais / Risikoanalyse
REFERENCES
Alfonsus E.C. (1933) Zum Pollenverbrauch des
Bienenvolkes, Arch. Bienenkd. 14, 220–223.
Baker H.G., Baker I. (1986) The occurrence and
significance of amino-acids in floral nectar, Plant
Syst. Evol. 151, 175–186.
Brødsgaard H.F., Brødsgaard C.J., Hansen H., Lövei
G.L. (2003) Environmental risk assessment of
transgene products using honey bee (Apis mellif-
era) larvae, Apidologie 34, 139–145.
Carter C., Graham R.A., Thornburg R.W. (1999)
Nectarin I is a novel, soluble germin-like protein
expressed in the nectar of Nicotiana sp., Plant
Mol. Biol. 41, 207–216.
Crailsheim K. (1990) The protein balance of the
honey bee worker, Apidologie 21, 417–429.
Crailsheim K., Schneider L.H.W., Hrassnigg N.,
Bühlmann G., Brosch U., Gmeinbauer R.,
Schöffmann B. (1992) Pollen consumption and
utilisation in worker honeybees: dependence on
individual age and function, J. Insect Physiol. 38,
409–419.
Davis A.R., Shuel R.W. (1988) Distribution of C-14-
labeled Carbofuran and Dimethoate in royal jelly,
queen larvae and nurse honeybees, Apidologie
19, 37–50.
Devillers J., Pham Delègue M. (2002) Honey bees:
Estimating the environmental impact of chemi-
cals, Taylor and Francis Inc., New York, 332 p.
Fearing P.L., Brown D., Vlachos D., Meghji M.,
Privalle L. (1997) Quantitative analysis of
CryIA(b) expression in Bt maize plants, tissues,
and silage and stability of expression over
successive generations, Mol. Breed. 3, 169–176.
Free J.B. (1993) Insect pollination of crops, Academic
Press, London, UK.
Haydak M.H. (1959) Changes with age in weight and
nitrogen content of honeybees, Bee World 40,
225–229.
Haydak M.H. (1970) Honey bee nutrition, Annu. Rev.
Entomol. 15, 143–156.
Imdorf A., Rickli M., Kilchenmann V., Bogdanov S.,
Wille H. (1998) Nitrogen and mineral constitu-
ents of honey bee worker brood during pollen
shortage, Apidologie 29, 315–325.
Kubik M., Nowacki J., Pidek A., Warakomska Z.,
Michalczuk L., Goszczynski W. (1999) Pesticide
residues in bee products collected from cherry
trees protected during blooming period with
contact and systemic fungicides, Apidologie 30,
521–532.
Kunert K., Crailsheim K. (1988) Seasonal changes in
carbohydrate, lipid and protein-content in emerg-
ing worker honeybees and their mortality, J. Apic.
Res. 27, 13–21.
Malone L.A., Pham-Delègue M.H. (2001) Effects of
transgene products on honey bees (Apis mellifera)
and bumblebees (Bombus sp.), Apidologie 32,
287–304.
Malone L.A., Tregidga E.L., Todd J.H., Burgess
E.P.J., Philip B.A., Markwick N.P., Poulton J.,
Christeller J.T., Lester M.T., Gatehouse H.S.
(2002) Effects of ingestion of a biotin-binding
protein on adult and larval honey bees, Apidolo-
gie 33, 447–458.
Maynard A.B., Loosli J.K. (1969) Animal Nutrition,
McGraw-Hill, New York, 613 p.
Planta A. v. (1888) Über den Futtersaft von Bienen, Z.
Physiol. Chem. 12, 327–354.
Schmidt J.O., Buchmann S.L. (1985) Pollen digestion
and nitrogen-utilization by Apis-mellifera L.
(Hymenoptera, Apidae), Comp. Biochem. Phys-
iol. 82A, 499–503.
Sears M.K., Hellmich R.L., Stanley-Horn D.E.,
Oberhauser K.S., Pleasants J.M., Mattila H.R.,
Siegfried B.D., Dively G.P. (2001) Impact of Bt
Corn Pollen on Monarch Butterfly Populations: a
Risk Assessment, Proc. Natl. Acad. Sci. 98,
11937–11942.
Simpson J. (1955) The significance of the presence of
pollen in the food of worker larvae of the
honeybee, Q. J. Microsc. Sci. 96, 117–120.
Stanley R.G., Linskens H.F. (1974) Pollen: Biology, Bio-
chemistry, Management, Springer-Verlag, Berlin.
Tasei J.N. (2001) Effects of insect growth regulators
on honey bees and non-Apis bees. A review,
Apidologie 32, 527–545.
Villa S., Vighi M., Finizio A., Serini G.B. (2000) Risk
assessment for honeybees from pesticide-
exposed pollen, Ecotoxicology 9, 287–297.
Wille M., Wille H. (1983) Vergleichende pollenana-
lytische Untersuchungen des Rückbehaltens in
Pollenfallen, Ermittlungsjahre 1981 und 1982,
Schweiz. Bienenztg 5, 253–268.
Wille H., Wille M., Kilchenmann V., Imdorf A.,
Bühlmann G. (1985a) Pollenernte und Massen-
wechsel von drei Apis mellifera Völkern auf dem-
selben Bienenstand in zwei aufeinanderfolgenden
Jahren, Rev. Suisse Zool. 92, 897–914.
Wille H., Imdorf A., Bühlmann G., Kilchenmann V.,
Wille M. (1985b) Beziehung zwischen Pollenein-
trag, Brutaufzucht und mittlerer Lebenserwartung
der Arbeiterinnen in Bienenvölkern (Apis mellif-
era L.), Mitt. Schweiz. Entomol. Ges. 58, 205–
214.