, Volume 94, Issue 3, pp 218–222

Floral odor learning within the hive affects honeybees’ foraging decisions


  • Andrés Arenas
    • Departamento de Biodiversidad y Biología Experimental, IFIBYNE-CONICET, Facultad de Ciencias Exactas y NaturalesUniversidad de Buenos Aires
  • Vanesa M. Fernández
    • Departamento de Biodiversidad y Biología Experimental, IFIBYNE-CONICET, Facultad de Ciencias Exactas y NaturalesUniversidad de Buenos Aires
    • Departamento de Biodiversidad y Biología Experimental, IFIBYNE-CONICET, Facultad de Ciencias Exactas y NaturalesUniversidad de Buenos Aires
Short Communication

DOI: 10.1007/s00114-006-0176-0

Cite this article as:
Arenas, A., Fernández, V.M. & Farina, W.M. Naturwissenschaften (2007) 94: 218. doi:10.1007/s00114-006-0176-0


Honeybees learn odor cues quickly and efficiently when visiting rewarding flowers. Memorization of these cues facilitates the localization and recognition of food sources during foraging flights. Bees can also use information gained inside the hive during social interactions with successful foragers. An important information cue that can be learned during these interactions is food odor. However, little is known about how floral odors learned in the hive affect later decisions of foragers in the field. We studied the effect of food scent on foraging preferences when this learning is acquired directly inside the hive. By using in-hive feeders that were removed 24 h before the test, we showed that foragers use the odor information acquired during a 3-day stimulation period with a scented solution during a food-choice situation outside the nest. This bias in food preference is maintained even 24 h after the replacement of all the hive combs. Thus, without being previously collected outside by foragers, food odors learned within the hive can be used during short-range foraging flights. Moreover, correct landings at a dual-choice device after replacing the storing combs suggests that long-term memories formed within the colony can be retrieved while bees search for food in the field.


Apis melliferaAssociative learningOlfactory experiencesInformation transferForaging choice


In honeybees, odors play an important role in food discovery and choice. It is known that bees direct their search for food according to innate search images, but also to previous experiences acquired either in the field or within the social environment (Ribbands 1955; von Frisch 1967). In this sense, olfactory memories formed during foraging flights can be retrieved within the hive by the presentation of the food odor, i.e., experienced foragers resume nectar collection at their feeding places when the conditioned stimulus (odor) is released inside the hive (Ribbands 1954; Johnson and Wenner 1966; Jakobsen et al. 1995; Reinhard et al. 2004). Also, the intake of scented food from foragers can lead to bias foraging preferences of their nest mates in the field (e.g., for honeybees, von Frisch 1923; Wenner et al. 1969; for bumblebees, Dornhaus and Chittka 1999). During recruitment to scented food sources, hive mates can learn the scent from the body of successful incoming foragers (von Frisch 1967) and from the food samples passed through mouth-to-mouth trophallactic contacts (Farina et al. 2006). Furthermore, it has recently been shown that the circulation of odor present in liquid food affects not only foragers but also younger preforagers (Grüter et al. 2006). In the last two studies, foragers were directly tested in the laboratory using the proboscis extension response (PER) assay (Bitterman et al. 1983). The PER procedure was also used to train bees that later improved their flight orientation performance towards the learned odor, suggesting a beneficial role of olfactory information during navigation in the vicinity of food sources (Chaffiol et al. 2005).

Given the ability of bees to associate odors with food during social interactions and the behavioral consequences of olfactory learning, the question arises of whether prior appetitive olfactory experiences within the colony provide enough information to forager mates to subsequently guide them during food search to new floral patches. Previous reports have shown that the efficiency of using a scented solution inside the hive to stimulate foraging at particular crops is highly variable (von Frisch 1967; Free 1969). Therefore, a quantitative study analyzing the influence of food-odor stimuli offered inside the hive, as well as its long-term effect on foraging choices, is required.

In the present work, we tested whether foraging decisions of foragers in the field might be affected by previous appetitive olfactory experiences established inside the hive. In order to test this, we presented a scented sugar solution using in-hive feeders and tested whether foragers prefer a food source having the same scent to a food source having an unfamiliar scent under two different experimental situations. In the first experiment, the in-hive feeder was removed from the colony 24 h before testing; in the second, the frame-feeder was removed 24 h before testing and all the hive combs were replaced.

Materials and methods

The entire training and testing period lasted 5 days. During the first 3 days, two possible food sources were presented: (1) an in-hive feeder constantly placed within the hives containing a scented sugar solution and (2) during 30-min training periods, a training feeder containing an unscented sugar solution presented 6 m away within the flight chamber. Except for the 30-min training periods when the training feeder was presented, the flight-chamber remained open so that foraging could also take place in the surrounding environment. For the last 2 days the in-hive feeder was removed, leaving only the occasionally presented training feeder or the outside world. On the last day, 10-min testing periods were presented after the 30-min training periods. During the testing period, foraging preferences towards two different odors were recorded: the odor offered inside the hive and a novel odor. The experiment was repeated using linalool (LIO) (hives 1) and phenylacetaldehyde (PHE) (hives 2) as olfactory cues paired with a sugar solution inside the hive (Table 1). In the first couple of experimental hives (hives 1A and 2A), the in-hive feeders were removed but the scented food remained stored in the combs. In a second group of hives (hives 1B and 2B), the in-hive feeders were removed and all the combs were replaced by combs from untreated commercial hives. In our study we compared the number of bees landing at the two testing feeders placed 6 m from the hive. Testing feeders offering the sugar solution were scented with two different odors. Bees foraging at testing feeders were observed and captured to analyze their food preference. The odors used were natural flower odors (Knudsen et al. 1993) and were obtained from Sigma-Aldrich, Steinheim, Germany.
Table 1

Functional groups, Chemical structures, carbon-chain lengths and vapor pressures of the odors used in the experiment

Functional group



Carbon_chain length

Vapor Pressure (mm Hg; 25°C)









The compounds are general floral odorants (after Knudsen et al. 1993)

Study site and animals

Four colonies of European honeybees (Apis mellifera ligustica) housed in the apiary of the experimental field at the School of Exact and Natural Sciences of the University of Buenos Aires (34°32′S, 58°26′W) were used in this study. The first series of the experiment (hives with nonremoved honey stores) took place from February to March 2005, and the second series (hives with removed honey stores) during the same months of 2006. Experimental colonies were reduced to four-frame Langstroth hives (about 5,000 individuals) and kept with only limited honey stored to facilitate the acceptance of artificial stimulation.

To avoid interference from other food sources and bees from other colonies during data acquisition, the experiment was conducted in a flight chamber consisting of a wooden structure with a transparent polyethylene rectangular mesh (6 × 3 × 2 m) hanging inside. Hives were moved from the apiary and fixed on one side of the flight chamber. A lateral frame was exchanged for a frame-feeder with 1.5 l of sugar solution in each hive.

Experimental procedure

The scented sugar solution was obtained by mixing 50 μl pure odorant per liter of 1.8 M sucrose solution. Olfactory experience was established by placing a frame-feeder with scented sugar solution in the hives for three consecutive days. After this period, the in-hive feeder was removed and replaced with an empty frame. All the combs of the hives 1B and 2B were replaced by others from an ordinary commercial hive of our apiary. As described above, the stimulation was repeated using LIO for the hives 1 and PHE for the hives 2.

Bees were trained to forage at an artificial ad libitum feeder, consisting of an acrylic dish (6.5 cm high, 4 cm diameter) inverted over a Plexiglas plate 10 mm thick, 5.5 cm diameter, and with 16 grooves (1 × 1 × 10 mm) cut in a radial arrangement on one side. Bees were trained to forage at a feeder on the opposite side of the flying cage (located 6 m from the hive). The feeder offered 1.8 M sucrose solution and was presented before we started with the testing phase in order to familiarize the bees with the food-source characteristics (shape, size, color, and location). Bees were trained four times per day for five consecutive days, and each training event lasted 30 min. A few minutes before each test, the training feeder was removed and two similar feeders were placed equidistantly to the hive (6 m) and 1.3-m from each other. These feeders offered an unscented sugar solution and had different scents coming from two glass Petri dishes (1 cm high, 15 cm diameter) containing a filter paper circle (55 mm diameter) soaked with a pure odor (50 μl pure odorant), which were placed below each feeder. The bees flying around the feeding area were able to make a choice between the two scented feeders. During the tests, the number of landings at each feeder was counted and the bees were captured. The presence of (unscented) sugar solution in the test feeders facilitated the capture. The tests lasted 10 min and were performed four times during the second day after the stimulation. The 10-min period was enough to obtain a sufficient amount of data without forming a negative association. Training events of 30 min were performed between test periods to maintain levels of motivation and an elevated number of bees searching for the food source as well.

The position of the feeders (left or right with reference to the entrance of the hive) was chosen at random. Landings were compared between olfactory stimulation (LIO and PHE). The flight chamber was closed only during training to the chamber-feeder and testing events, allowing normal activity during the rest of the day. While the flight chamber remained open, the feeders were removed in order to avoid visits of bees from other colonies.


In order to compare behavioral responses among treatments, Chi-squared goodness of fit and contingency tables (χ2) were applied on the number of landing bees (Zar 1999).


For hives with nonremoved honey stores, we observed a clear preference for visiting the treatment-odor feeder in comparison to a novel-odor feeder in both colonies (for hive 1A, Chi2LIO = 9.39, p = 0.002182, df = 1, N = 72; for hive 2A, Chi2PHE = 6.38, p = 0.01154, df = 1, N = 106, Chi-square goodness of fit, Fig. 1a). Interestingly, there was no difference between the two colonies in the number of landings at the feeder offering the odor treatment (Chi2 = 0.63, p = 0.4278, df = 1, N = 178, Chi-square contingency table). Hives with removed honey stores also presented a strong preference to the odor treatment (for hive 1B, Chi2LIO = 11.2676, p = 0.000789, df = 1, N = 142; for hive 2B, Chi2PHE = 63.5347, p = 0.0000, df = 1, N = 148, Chi-square goodness of fit, Fig. 1b). For this series we found differences between the two colonies in the number of landings at the feeder offering the odor treatment (Chi2 = 9.21, p = 0.0024, df = 1, N = 178, Chi-square contingency table). For hives with removed honey stores, similar response patterns were found for two odors belonging to different chemical groups (LIO, a terpene, and PHE, an aldehyde) and having different vapor pressures (Table 1) but not for hives with the honey stores removed.
Fig. 1

Percentages of landings performed by bees at each chamber-testing feeder 24 h after the removal of the in-hive feeder. Hives 1 received sugar solution scented with LIO while hives 2 received sugar solution scented with PHE. a The food stored in hives A was maintained in honeycombs after removing the in-hive feeder. b The honey stored in hives B was removed and their combs replaced. The number of bees landing on each chamber-testing feeder is shown at the bottom of each bar. Landings were analyzed within treatments using χ2-test (triple asterisksp < 0.001, double asterisksp < 0.01, single asteriskp < 0.05; see “Results” for details)


In agreement with previous reports (Farina et al. 2005, 2006; Gil and De Marco 2006; Grüter et al. 2006), the present results indicate that honeybees can form appetitive olfactory memories within the hive and show moreover that these in-hive memories can be retrieved later in a floral-choice situation biasing foragers’ decisions. In order to form these appetitive memories, honeybees do not require the presence of food stores to prime the retrieval of a previously learned olfactory memory.

Feeding scented sugar solution, either in or outside the honeybee hives, has been used in early attempts to direct honeybees to certain crops (for review, see von Frisch 1967). In this sense, the method of offering scented-sugar solution inside the hive has been reported to be highly variable, and only results obtained by Free (1969) have shown a very brief operant response of foragers (30 min) which collected sugar solutions from different scented feeders, after exchanging the stored food between two hives. Thus, the results obtained in this study differ from those of Free because we observed a very clear operant response within a foraging context during the whole experimental period (24 h after the stimulation), and even after removing the scented-food stores.

Many potential foragers would be involved in collecting sugar solution from our frame feeder and would presumably learn the odor then or during its circulation to the storing combs. At least for the situation in which the food stores were not removed, we can speculate that the recorded foragers have learned the odor from the food stores or from food donors in a relatively short period before leaving the hive. In this case, the operant responses measured were a consequence of transferring a short-term appetitive association formed in the hive to an external foraging context response. However, the results obtained after removing all the combs suggest that individual bees would still retain in-hive olfactory memory over a day even without continued priming by scented honey stores. This indicates that the olfactory information acquired within the hive could be transferred at least to an early long-term memory in the field (Menzel 1999).

Since both experimental series, with and without scented-food stores, were performed during different years, we avoided comparing both experiments. However, the higher responses found for bees from hives without scented-food stores could be explained by differences in the dynamic of consolidation processes of both experiments: in one case, the continued presentation of the contingency odor-reward until the test; in the other, a 1-day period with the absence of multiple olfactory experiences within a short period of time that would undergo a sequential transfer to consolidate long-term memories (Menzel 1999).

Inside the hive, naive foragers could have learned the association between the odor and reward either during ingestion of the sugar solution from the in-hive feeder or by receiving food via trophallaxis. This odor information has guided food choice within the flight chamber. It is important to mention that the honeybees from the experimental hive were allowed to forage at natural food sources (the flight chamber was opened) during almost the entire time of the experiment, which implies that other appetitive memories could have been formed during these periods. In comparison with the artificial (in-hive) stimulation, the acquisition of this alternative information could be much more efficient when it involves recruitment mechanisms displayed by successful foragers.

Floral odors and food-source profitability represent essential information that can be transferred by incoming foragers during recruitment (von Frisch 1967; Seeley 1995). Recruits can learn the association between odors and reward inside the hive (Farina et al. 2005; Grüter et al. 2006) and use this information to modify the orientation performance during searching flights (von Frisch 1923, 1967). The present study provides additional information about the role of food scent circulating among nest mates, demonstrating that olfactory memories acquired inside the hive could later modify odor-mediated decisions during foraging tasks.

It is worth mentioning that honeybees do not use exclusively olfactory cues to enhance their individual and collective foraging efficiency in the field (von Frisch 1967). In this study, specific-odor memories formed inside the hive may help foragers to identify rewarding flowers when they fly in the close range of the feeding site. Other sensory modalities such as visual information would help bees to confirm navigational memories that can be transmitted during dancing (Riley et al. 2005), providing recruits with far-range information about the profitable source discovered.

Because floral scents are species-specific, the transfer of olfactory information represents the most primitive component for the communication during recruitment in social bees. As it happens in bumblebees, the recruitment system is specific for a flower species (Dornhaus and Chittka 1999). The successful forager brings home the odor of the newly discovered food source, but it does not transfer food samples to nest mates. The forager unloads the collected food directly into the honey pots, which might be later inspected by colony mates. In this case, the role of the scented-food stores is essential to acquire information to find a profitable source. The communication of bumblebees seems to be the evolutionary origin of most complex communication systems in eusocial bees. The odors learned inside the honeybee hives could help foragers to find a particular food source in the surrounding area of the hive, even when this information is not accompanied by spatial information.

The propagation of floral scents involving a large number of individuals might have important implications in our understanding of the colony level responses to the environment. For social insects, the acquisition of global information about food may help a colony increase foraging success in a changing environment. Our findings emphasize an important role played by the food scent itself for olfactory conditioning in a social environment and show a long-term effect in foraging preferences. Food stores may thus be an important source of olfactory information about previous and present resource sources.


We are indebted to C. Grüter and R. Josens for their valuable comments and suggestions on an early version of this manuscript. We thank also C. Grüter for helping us with the English and G. Ramirez and H. Verna for technical assistance. This study was supported by funds from Agencia Nacional de Promoción Científica y Tecnológica (01-12319), Consejo Nacional de Investigaciones Científicas y Técnicas (02049), and the University of Buenos Aires (X 036) to WMF. We declare that our experiments comply with the current laws of the country in which they were performed.

Copyright information

© Springer-Verlag 2006