Age-Related Foraging Behaviour in Honey Bees Under Artificial Conditions

Original article

Age-related foraging behaviour in honey bees under artificial conditions

MH Pham-Delègue M Le Métayer, P Douault, C Masson

INRA-CNRS (URA 1190), Laboratoire de Neurobiologie Comparée des Invertébrés, 91440 Bures-sur-Yvette, France (Received 31 May 1989; accepted 24 January 1990)

Summary — The distribution of foraging among 4 worker groups of different ages was observed in a flight room. Foraging was mostly performed by the oldest workers, who came to forage first and were the most numerous. Workers from the other groups visited the feeding station infrequently until the second half of the observation period, then increased their visits. Regardless of their age group, each individual coming to forage performed a constant average level of activity, despite some vari- ability in the individual scores of visits. These experiments stressed a strong intra-age group interac- tion among a population of potential foragers, and allowed us to record data on both group and individual strategies. In conjunction with studies performed under natural conditions, our experiments conducted under artificial conditions may contribute to the analysis of some basic parameters of foraging behaviour.

Apis mellifera (ligustica x caucasica) x mellifera / foraging / age effect / division of labour / flight room

INTRODUCTION worker population were removed, it was well demonstrated that workers could shift In the normal ontogeny of honey bee from one task to another (Rösch, 1930; Winston tasks, younger workers are devoted to Lindauer, 1961; Kolmes, 1985; hive duties and older workers perform for- and Fergusson, 1985). These efforts, as aging (Rösch, 1925, 1930; Ribbands, well as that of Seeley (1982) who tested 1952; Lindauer, 1953; Sekiguchi and Sak- predictions about the design of the age agami, 1966; Michener, 1974; Seeley, polyethism schedule for labour inside the 1982; Winston and Punnett, 1982; hive, were concerned with the shift of Kolmes, 1985). However, it is known that workers duties. Little time was devoted to there is considerable flexibility in the ages analyzing the division of labour among at which workers shift from hive to field du- workers all performing the task of foraging, ties, depending on internal colony require- except Seeley (1983) who explored divi- ments or external conditions. Through ex- sion of labour between scouts and recruits periments where specific groups of the in honey bee foraging, to determine the

* Correspondence and reprints ecological significance of the dance lan- The foraging behaviour was observed on a guage. feeding device (Pham-Delegue and Masson, 1985). This device provided sources of sugar In this paper, we aim to detail the inter- solution associated with a scent. A 50% sucrose group or inter-individual distribution of the solution, known to be highly acceptable to hon- foraging task among a population of ey bees (Waller, 1972), was given as a reward. known age foragers. We have measured The associated scent was geraniol 1 % in paraf- fin oil. Six scented feeders were distributed on a the distribution among groups of foragers slowly rotating device (1/3 rpm). Such a device of closely related ages to assess the de- attracts a high number of foragers which can be gree of flexibility of such a task under con- fed simultaneously. trolled conditions. The experimental device was positioned at 1.50 m from the hive entrance every day between 2 and 4 pm. To maintain the motivation to MATERIALS AND METHODS search for food, at times other than the 2-4 pm period, the honey bees were fed at the same lo- cation with a 50% sucrose solution and a dish of The experiments were conducted under artificial pollen at all times. No additional food was given conditions to control the constitution of the ex- in the hive. perimental colony and to standardize the obser- The device was follow- vation The were carried presented every day procedure. experiments the introduction of the last but the out in a room of 12 m3 vol, as described ing age group, flight by data below with the first Douault (1978). The light intensity was 300 lux reported begin sponta- neous on the device. the observa- with a 12 hr 12 hr The landing Thus, day - night photoperiod. tions started 31 d after the introduction of the was 24 ± 2 °C with 55% relative temperature first of workers in the hive and the The air in the room was renewed group experi- humidity. reg- mental ended 38 d later. ularly (5 times per h). period The observation was as follows: Experiments were conducted with a colony procedure within the each age- of 3-strain hybrid Apis mellifera (ligustica x cau- experimental period, marked bee was collected at its first on casica) x mellifera which produce high honey landing a feeder and labelled with a num- yields under natural conditions (Fresnaye et al, individually 1974; Comuet and Fresnaye, 1979), and which bered tag according to its rank of landing. Every 15 min, the and of the is known to exhibit a high level of foraging under identity (colour number) artificial conditions (Pham-Delegue et al, 1984). foragers on the device was recorded, up to 8 re- cordings per observation day. Experiments were The colony was first constituted of unknown carried out daily, with several days of disruption, winter worker and later age bees, replaced by leading to 22 d of observations. Mortality was workers: a old with a age-marked 2-year queen noted every observation day by collecting dead cluster of 150 worker bees of unknown age bees on the floor of the flight room. were first introduced in the experimental hive (21 x 21.5 x 27 cm), set with 5 empty combs and 1 comb filled with food; these escort bees RESULTS were later removed when arriving at the feeding device. Four samples of new born worker bees were successively introduced, leading to a the the % cu- range of 4 groups of workers differing in age by During experimental period 14 d from the first group introduced to the pre- mulative mortality per group according to ceeding one: group I (400 individuals), group II age (ie number of dead workers, including (310 individuals) introduced 3 d later, group III foragers and non-foragers, versus the ini- (440 individuals) introduced 5 d later than group tial number of workers x 100), increased II, IV introduced 6 d later group (800 individuals) of the all than group III. Each age group was identified by similarly; regardless group, points a different coloured spot painted on the thorax fitted with a curve (y = 0.71 exp (0.07 x)), of the workers. with a highly significant level of correlation (r = 0.97,82 df). Until 35 d old, mortality spectively, for the oldest foragers from was of 0-5%; between 35-70 d old, mor- groups I, II, III, IV. The statistical compari- tality increased linearly from 5-65%. son of the proportions, within the common age-range when the 4 groups were Since the mortality was approximately present (from 33-57 day old), showed that equivalent in every group, it did not affect the foraging activity was significantly high- the foraging activity differentially among er in group I than in the other groups (2- each this was group. Thus, activity report- tailed, signed ranks Mann-Whitney test, ed as the of new daily percentage foragers P < 0.01); the activity in group II was lower number of new is cumulated (the foragers than in group IV (P < 0.05). The highest observation versus the initial every day) proportion of foragers from group I number of workers (eg, 400 in group I), as reached the value of 30% workers dedicat- a function of age (fig 1). The age when the ed to foraging at the end of the experi- first member of a its first group performed ment; bees from the 3 other groups were foraging trip was, respectively, 32, 33, 28, involved in foraging with maximum values 18 day old for groups I, II, III and IV. Work- of 11-13%; group II being the least repre- ers from group IV was recorded simultane- sented. At the end of the experiment, ously; workers from groups II and III ap- groups I and II, whose foragers were the peared at the third observation day. At the oldest, reached a steady state, while the end of the experiment, workers were still number of foragers was still increasing foraging at an age of 71, 68, 63, 57 day re- among groups III and IV. Consistent with the fact that foragers 80 (17th observation day), 106 (17th ob- from group I were the most numerous, servation day), respectively, for groups I, they were the most active until the 10th II, III, IV. observation day (performing more than 75% of the visits), and were then progres- Considering the individual activity ac- sively replaced by the other groups, to a fi- cording to age group, by reporting the nal level of 16.3% visits in group I, 14.2% number of visits per observation day as a in group II, 25.5% in group III, 43% in function of particular individuals performing group IV, at the 22nd observation day. The these visits (fig 2), no difference appeared highest number of visits was 245 (6th ob- between the groups. Globally, the number servation day), 63 (16th observation day), of visits was positively correlated with the number of individuals (r = 0.93, 86 df, P < servation days when each individual came 0.001; y = 4.74 x - 1.17). Thus, each indi- to forage (fig 3). Over 22 observation days, vidual performed the same average num- 19 foragers were observed only once and ber of 4.74 visits per observation day, did not come on the following observation whatever group it came from. However, a days. Groups of 6-12 workers came to for- maximum value of 38 individuals from age on 2-6 d of observations. Fewer forag- group I were foraging simultaneously, ers were observed over longer periods. while a maximum of 15 foragers appeared Thus, the faithfulness to the food source from group II; values from group III (20 in- was relatively weak for group I. Moreover, divudals) and IV (30 individuals) were in- the number of visits exhibited by foragers termediate. Also the highest number of for- of different groups was identical. These re- agers appeared earlier in group I (7th sults could explain the fact that, despite versus 16th observation day in the other the larger number of foragers in group I, groups). the other groups reached the same level of To focus on the individual strategy of activity at the end of the experiment. foraging, we detailed the behaviour of 84 individually-marked foragers from group I. The faithfulness to the food source was DISCUSSION evaluated by recording the number of ob- These experiments were conducted under artificial conditions, which allowed us to work throughout the year and to control the environment. These conditions did not strongly affect the life span of the workers. Mortality was slightly higher than under natural conditions (65% after 2 months versus 50% for winter bees, according to Fu- kuda and Sekiguchi, 1966), probably due to the intense flight activity artificially main- tained in cage conditions. Moreover, the foraging age closely matched the values mentioned under natural conditions; vari- ous studies have shown that the means for first foraging flights range from 18-43 d (re- viewed by Michener, 1974; Winston and Punnett, 1982; Kolmes, 1985). However, it must be noted that in our experiment, the youngest workers which came to forage (18 d old) were coming from the largest group (group IV). This may result from ei- ther a higher inter-individual stimulation of being in a large group, or a higher prob- ability of at least 1 bee leaving in large groups. During the first half of the observation period, foraging was mostly performed by workers from the oldest age group, which Lindauer, 1961; Kolmes, 1985; Winston is consistent with many studies showing and Fergusson, 1985). Manipulations clos- that tendency to forage is a positive func- er to the normal variations that could occur tion of age (Rösch, 1925, 1930; Ribbands, during the annual cycle of a honey bee col- 1952; Lindauer, 1953; Sekiguchi and ony, such as those conducted by Kolmes Sakagami, 1966; Michener, 1974; Seeley, and Winston (1988), might lead to a differ- 1982; Winston and Punnet, 1982; Kolmes, ent task distribution, even under artificial 1985). However, it is surprising that this conditions. activity is restricted to a particular age pool In conclusion, although these experi- (group I), when the other groups of quite ments were carried out under artificial con- similar least had age (at group II) already ditions and do not allow us to directly re- appeared and may have participated in late the observations to normal outdoor this task more actively. This may be due foraging conditions, they do allow the sep- to a particular physiological fitness or to arate control of parameters affecting the a higher motivation "to go" or "to stay" foraging task. Here, using controlled age (Nunez, 1970) within this group of work- groups, it appeared that age effect is of ers. It is known that foraging is per- major importance since bees seemed to in- formed of workers by specific groups teract more with bees of the same age among a population of potential foragers (within group I) than with bees of other von Wen- (Ribbands, 1954; Frisch, 1967; ages (group I - group II). The flexibility in ner et al, 1967), and here it may be sug- the temporal division of labour schedule, that this role-subdivision relies on a gested shown under natural conditions by remov- this a in- sharp age effect; implies strong ing groups of workers more or less in the stimulation and a tra-age group foraging range of variations normally experienced inhibition. Further ob- possible inter-group by a colony, could be tested in the flight servations of the communication process- room by constituting colonies of entirely es inside the hive are to out required point known age workers. Moreover, the colony the degree of intra- or inter-group relations requirements (linked to the brood amount and the of workers to mem- ability identify and the food available inside the hive), as bers of their own group. well as the parameters of the feeding Even though there is a differential con- source (eg, the quantity and/or quality of tribution of each age group to forage ac- the food, the way it is distributed, the olfac- cording to the number of foragers, it ap- tory or visual cues associated to it, etc) pears that whenever an individual worker could be easily controlled under artificial from any group comes to the feeding sta- conditions. Also, the observation of inter- tion, it exhibits a constant average level of individual communication processes inside activity, expressed as the number of visits the hive could be facilitated. Therefore, per observation day. However, when parallel to studies under natural conditions, considering the level of activity as the we plan to develop experiments under con- faithfulness to the food source, individual trolled conditions, such as those described strategies differed strongly; numerous above, to contribute to the analysis of pa- individuals foraging on 1 day only. These rameters of the foraging process. individuals might account for a group of "lazy" bees, who fulfilled an important func- ACKNOWLEDGMENTS tion in the hive, according to Lindauer (1953). These individuals might also shift The authors would like to thank Dr JL Deneu- to hive duties, as described after extreme bourg (Université Libre, Bruxelles) for fruitful dis- demographic manipulations (Rösch, 1930; cussions on data interpretation. Résumé — Étude du comportement de Apis mellifica (ligustica x caucasica) x butinage en fonction de l’âge chez mellifica / butinage / effet âge / division l’abeille domestique en conditions artifi- du travail / chambre de vol cielles. La distribution des tâches en fonction de l’âge chez les ouvrières d’abeilles est bien connue, l’activité de butinage ap- Zusammenfassung — Altersbezogenes paraissant chez les ouvrières les plus Sammelverhalten bei Honigbienen und âgées. Cependant peu d’études ont porté künstlichen Bedingungen. Die Arbeitstei- sur la division de l’activité de butinage au lung bei den Arbeitsbienen in Abhängigkeit sein d’une population d’ouvrières en âge von ihrem Alter ist gut bekannt; die de butiner. Afin d’analyser cet aspect, nous Sammeltätigkeiten treten bei den ältesten avons observé en cage de vol, le compor- Bienen auf. Es gibt jedoch noch wenige tement de butinage d’ouvrières apparte- Untersuchungen, die sich mit der Tei- nant à 4 classes d’âge (classes I à IV com- lung der Sammeltätigkeit innerhalb einer prises dans une période de 2 semaines). Arbeiterinnen-Population im Sammelalter Les butineuses sont repérées individuelle- beschäftigen. Um diesen Aspekt zu analy- ment sur une source alimentaire odorante sieren, haben wir das Sammelverhalten (solution de saccharose associée à du gé- von Arbeitsbienen in einem Flugkäfig beo- raniol) distribuée ad libitum durant des pé- bachtet, die zu vier Altersklassen gehörten riodes d’observation quotidiennes de 2 h, (Klasse I-IV, innerhalb einer Periode von pendant 22 jours. Les butineuses de la zwei Wochen). Die Sammlerinnen wurden classe I sont les mieux représentées sur la individuell an einer mit Duft versehenen source (30% des ouvrières de cette classe Futterstelle (Saccharoselösung mit Gerani- butinent, contre moins de 13% dans les ol) gezeichnet, die während der Versuchs- autres classes) (fig 1), et elles effectuent le dauer von 22 Tagen zur täglichen Beob- pourcentage de visites le plus élevé durant achtungszeit von 2 h ad libitum versorgt la première moitié de la période d’observa- war. Die Sammlerinnen der Klasse I waren tion, avant d’être relayées par les autres an der Futterstelle am besten vertreten classes. A l’échelle individuelle, indépen- (30% der Arbeiterinnen dieser Klasse sam- damment de la classe d’âge, les buti- melten, gegenüber weniger als 13% bei neuses réalisent le même nombre de vi- den übrigen Klassen, Abb 1), und sie hiel- sites à la source (4,74 visites par individu ten während der ersten Hälfte der Beo- par jour d’observation) (fig 2). Toutefois, bachtungsperiode den höchsten Anteil an une grande variabilité apparaît dans la fi- Besuchen, bevor sie zu gleichen Teilen délité des individus à la source alimen- von den anderen Klassen abgelöst wurden taire : sur 84 butineuses observées durant (Abb 2). Bei Bewertung jeder Biene für 22 jours, 19 sont présentes 1 seul jour, 6 à sich, unabhängig von der Altersklasse, 12 butinent de 2 à 6 jours, les autres pro- führte jede Sammlerin dieselbe Zahl von longeant leurs visites de 6 à 18 jours (fig Besuchen der Futterstelle durch (4,74 Be- 3). Ces résultats soulignent l’intérêt d’expé- suche/Individuum/Beobachtungstag; Abb rimentations conduites en conditions artifi- 3). Es ist jedoch eine große Variabilität in cielles qui, menées en parallèle à des obs- der Stetigkeit des Besuchs an der Futter- ervations en conditions naturelles, stelle durch die einzelnen Bienen festzus- devraient permettre de mieux définir les li- tellen: Von 84 Sammlerinnen, die während mites de la flexibilité dans la division du der 22 Tage beobachtet wurden, zeigten travail chez l’abeille. sich 19 nur an einem einzigen Tag, 6 bis 12 sammelten an 2 bis 6 Tagen, während Nunez JA (1970) The relationship between sug- die übrigen ihre Besuche auf 6 bis 18 ar flow and foraging and recruiting behaviour ausdehnten Diese of honey bees (Apis mellifera L). Anim Behav Tage (Abb 4). Ergeb- 18, 527-538 nisse unterstreichen die Bedeutung von MH, Masson C die unter künstlichen Be- Pham-Delegue (1985) Analyse Experimenten, par conditionnement associatif du mécanis- dingungen durchgeführt werden, weil sie me de la reconnaissance de sources alimen- es besser gestatten, zusammen mit Beob- taires par l’abeille. Bull Soc Entomol France achtungen unter natürlichen Bedingungen, 90, 1216-1223 die Grenzen der Flexibilität in der Arbeit- Pham-Delegue MH, Masson C, Douault P steilung der Bienen zu bestimmen. (1984) Etude comparée effectuée au laboratoire, des aptitudes au butinage d’abeilles de race mellifica et in- mellifera x x Apis ligustica d’hybrides Apis (ligustica caucasica) terraciaux mellifica x caucasi- mellifera / / Altersein- Apis (ligustica Sammeltätigkeit ca x mellifica). Apidologie 15, 33-42 fluß / / Arbeitsteilung Flugraum Ribbands CR (1952) Division of labour in the honeybee community. Proc R Soc Lond Ser B 140, 32-42 REFERENCES Ribbands CR (1954) Communication among honeybees. I. The response of crop-attached bees to the scent of their Proc R Soc Cornuet JM, Fresnaye J (1979) Production de crop. miel chez des hybrides interraciaux Lond Ser A 29, 10-12 d’abeilles (Apis mellifica L) lors de généra- Rösch GA (1925) Untersuchungen über die Ar- tions successives de rétrocroisements sur la beitsteilung im Bienenstaat, I. Z Vergl Physiol race locale. Apidologie 10, 3-15 2, 571-631 Douault P (1978) L’élevage expérimental des Rösch GA (1930) Untersuchungen über die Ar- abeilles en milieu clos. Bull Tech OPI- Apic im II. Z DA 5, 15-22 beitsteilung Beinenstaat, Vergl Physi- ol 12, 1-71 Fresnaye J, Lavie P, Boesiger E (1974) La vari- abilité de la production de miel de l’abeille de Seeley TD (1982) Adaptative significance of the race noire (Apis mellifica mellifica) et chez age polyethism schedule in honeybee colo- quelques hybrides interraciaux. Apidologie 5, nies. Behav Ecol Sociobiol 11, 287-293 1-20 Seeley TD (1983) Division of labor between Frisch von K (1967) The Dance Language and scouts and recruits in honey bee foraging. Orientation of Bees. Cambridge (MA), Har- Behav Ecol Sociobiol 12, 253-259 vard University Press, 566 K, SF Structure of K Sekiguchi Sakagami (1966) Fukuda H, Sekiguchi (1966) Seasonal change and related in of the bee worker in foraging population problems honey longevity Sappo- the with considerations on the di- ro, North with notes on some factors honeybee, Japan, vision of labor in bee colonies. Hokkaido affecting the life span. Jpn J Ecol 16, 206- Rep 212 Natl Agric Exp Stn 69, 1-65 Kolmes SA (1985) A quantitative study of the di- Waller GD (1972) Evaluating responses of hon- vision of labour among worker honey bees. Z eybees to sugar solutions using an artificial Tierpsychol 68, 287-302 flower feeder. Ann Entomol Soc Am 65, 857- Kolmes SA, Winston ML (1988) Division of la- 861 bour among worker honey bees in demo- Wenner AM, Wells PH, Rohlf FJ (1967) An anal- graphically manipulated colonies. Ins Soc 35 ysis of the waggle dance and recruitment in (3), 262-270 the honey bees. Physiol Zool 40, 317-344 Lindauer M Division of labour in the hon- (1953) Winston ML, Punnet EN Factors deter- ey bee colony. Bee World 34, 63-73; 85-90 (1982) division of labor in Lindauer M Communication So- mining temporal honey- (1961) Among bees. Can J Zool 60, 2947-2952 cial Bees. Cambridge (MA), Harvard Univer- sity Press, 243 Winston ML, Fergusson LA (1985) The effect of Michener CD (1974) The Social Behavior of the worker loss on temporal caste structure in Bees. Cambridge (MA), Harvard University colonies of honeybee (Apis mellifica L). Can Press, 404 J Zool 63, 777-780