captured upon their first active visit (i.e., they
flew on their own rather than being trans-
ported by the experimenter) to a feeder to which
they were previously brought. Capture occurred
either immediately upon arrival at the feeder
or 1 min after landing and feeding. Figure 1C
shows a significant interaction between the
factors“biogenic amine”and“behavioral con-
dition”(F1,65= 16.36,P< 0.001). Whereas
serotonin levels did not vary between the two
experimental conditions (Tukey tests, not sig-
nificant), dopamine levels were higher in bees
arriving at the feeder for the first time (P<
0.001). Thus, specific expectations about prof-
itable rewards are accompanied by high dopa-
mine levels, irrespective of the amount of flight
experience.
Having shown that waggle dances tran-
siently enhance dopamine levels of dancers,
we sought to determine whether the same
effect was observable in dance followers that
had never visited the feeder. We monitored
waggle dances of marked foragers reporting
the location of a trained feeder and captured
nonmarked followers (i.e., bees with no feeder
experience), either within 5 s after they started
following the dancer (“short-term followers”) or
after 50 to 60 s (“long-term followers”). We also
captured dancers in the middle of their dances
(typically 20 to 30 s after dance start). We mea-
sured brain levels of biogenic amines and found
that the interaction between the factors“bio-
genic amine”and“behavioral category”was
not significant (fig. S1;F2,100= 0.17,P= 0.85).
Significant differences were found between
dopamine and serotonin (F1,100= 41.26,P<
0.0001) and between bee categories (F2,100=
16.38,P< 0.0001). Dopamine levels were
higher in dancers and increased in followers
from the start to the end of the dance, yet this
increase was not significant (Tukey tests). A
similar trend was found for serotonin. Thus,
dances might not activate per se a dopamine-
based wanting system in followers.
To confirm the presence of a dopamine-
based wanting system in foragers, we interferedwith dopaminergic signaling and determined
the effect of this interference on foraging.
We trained marked foragers to collect sucrose
solution from a feeder and measured their
frequency of foraging bouts and their return
time, that is, the length of time between two
consecutive visits to the feeder, as these vari-
ables are sensitive indicators of appetitive
motivation ( 9 , 10 ). Measurements were per-
formed 30 min before topical delivery on the
thorax ( 11 ) of the dopamine receptor antag-
onist fluphenazine ( 12 , 13 ), the dopamine
receptor agonist 6,7-ADTN hydrobromide
(hereafter, 6,7-ADTN) ( 14 – 18 ), or the solvent
N,N′-dimethylformamide (DMF). Another con-
trol group remained untreated (“sham”). After
20 min, the same foraging variables were mea-
sured during a second 30-min period. The
interaction between“treatment”and“mea-
surement time”was not significant for the
frequency of foraging bouts (Fig. 2A; analysis
of variance for repeated measurements;F3,82=
1.57,P= 0.20). However, a significant variation
was observed between measurement times
(before versus after:F1,82= 29.73,P< 0.0001)
and between treatments (F3,82= 3.40,P< 0.05)
as the antagonist fluphenazine decreased the
frequency of bouts (Tukey test,P< 0.01). In
the case of the return time, the interaction
between“treatment”and“measurement time”
was significant (Fig. 2B;F3,82= 7.49,P< 0.001),
and fluphenazine was again responsible for
prolonging it (P< 0.001). The lack of effect of
6,7-ADTN on both variables measured may have
been caused by a ceiling foraging motivation.
As fluphenazine might have affected motor
activity ( 19 ) rather than foraging motiva-
tion, we repeated the previous experiment
but decomposed the return time into (i) the
time spent flying from the feeder to the hive,
(ii) the time spent in the hive, and (iii) the time
spent flying from the hive to the feeder. Foraging
times were quantified in marked foragers before
and after topical exposure to DMF, fluphena-
zine, or 6,7-ADTN. An untreated sham group
was also included. Neither the time spent fly-
ing from the feeder to the hive (Fig. 2C; inter-
action between“time”and“group”factors:
F3,73= 1.92,P= 0.13) nor the time spent flying
from the hive to the feeder (Fig. 2D; inter-
action between "time" and "group" factors:
F3,73= 2.26,P= 0.09) varied before and after
treatment or between groups. On the contrary,
the time spent within the hive exhibited a
significant interaction (Fig. 2E;F3,73= 8.01,
P< 0.001). Specifically, fluphenazine in-
creased the time spent in the hive (Tukey test,
P< 0.001). Thus, blockade of dopaminergic
signaling had no effect on flight activity but
prolonged the time spent in the hive at the
expense of foraging.
As honey bee foraging is determined by
colony needs rather than by individual hunger
( 20 ), we asked whether individual appetitiveSCIENCEscience.org 29 APRIL 2022•VOL 376 ISSUE 6592 509
Fig. 1. Food expectation upon departure toward a food source and dance start increase dopamine
levels in the honey bee brain.The three panels show dopamine (DA, orange bars) and serotonin (5-HT,
green bars) levels (nanograms per brain; mean ± SEM) measured in individual brains of foragers belonging to
different behavioral categories. Lowercase letters (a and b) above the bars indicate significant differences
(Tukey tests;P< 0.05). Scatter plots show individual data distribution. (A)“Feeder arrival”corresponds to
foragers that were observed dancing within the hive and that were captured upon arrival at the feeder
reported in their dances and before they started feeding (DA:n= 11; 5-HT:n= 10).“Feeder feeding”
corresponds to the same kind of bees captured 1 min after they landed on the feeder and started feeding
(DA:n= 15; 5-HT:n= 15).“Dancing start”corresponds to bees that started dancing right after their
return to the hive from the feeder and that were captured upon dance initiation (DA:n= 15; 5-HT:n= 14).
“Dancing end”refers to the same kind of bees captured upon dance ending, ~1 min after dance start
(DA:n= 15; 5-HT:n= 15).“Dancing cessation”refers to bees that stopped dancing after food was made
unavailable for 30 min (DA:n= 15; 5-HT:n= 15). (B)“Hive departure”corresponds to foragers captured
upon departure from the hive after dancing (DA:n= 24; 5-HT:n= 24).“Feeder arrival”corresponds to
the same kind of bees captured upon arrival at the feeder and before they started feeding (DA:n= 23; 5-HT:
n= 23). (C)“First arrival”shows foraging bees arriving at the feeder in their first active flight after prior
appetitive experience at that location (DA:n= 19; 5-HT:n= 19).“First feeding”shows the same kind of
bees captured 1 min after they landed on the feeder and started feeding (DA:n= 13; 5-HT:n= 18).
RESEARCH | REPORTS