Vol. IOOA,No. 3, PP. 585-594, 1991
Camp. Biochem. Physiol.
Printed in Great Britain
0300-9629/91$3.00+ 0.00 1991Pergamon Press plc
INDUCTION OF BALLING IN WORKER HONEYBEES (AX!3 MELLIFER, L.) BY “STRESS” PHEROMONE FROM KOSCHEWNIKOW GLANDS OF QUEEN BEES: BEHAVIOURAL, STRUCTURAL AND CHEMICAL STUDY YAACOV LENSKY,* PIERRE CASSER,~$ SHMUEL ROSA* and DIDIER GRANDPERRIN~ lTriwaks Bee Research Center, Hebrew University, Faculty of Agriculture, 76100 Rehovot, Israel; and SUniversitb P. et M. Curie, Laboratoire d’I?volution, 105, Boulevard Raspail, 75006 Paris, France. Telephone: 42-22-65-50 (Received 7 January 1991) Abstract-l. The Koschewnikow glands (KG) of honeybee queens are composed of type III glandular units; secretions are emitted from ducts that open onto the entire surface of the intersegmental membrane. 2. Secretory activity is characterized by the appearance of dense granules containing mainly glycoproteins, stained by periodic acid-thiocarbohydrazide-silver proteinate, extracted by pronase. In one-yearold mated queens, the gland degenerates. Secretory products originating from mitochondria are likely to be only carriers of pheromonal compounds. 3. Topical treatment of worker bees (“pseudoqueens”) with EtOH extracts of queen Koschewnikow glands induced typical queen balling behaviour in workers of a bee colony. 4. Twenty-eight compounds including acids, alcohols, alkanes and alkenes (C,H,,-C,,H,,) were characterized by gas liquid chromatography-mass spectrometry in queen KG extract. None of them is present in worker alarm pheromone which is secreted from worker KG.
The society of the honeybee is basically monogynous. Attempts to achieve cohabitation of multiple queens of identical or different ages in the same brood-nest, without their being confined to a cage or a compartment were unsuccessful due to mutual aggressive
behaviour. Polygynous societies headed by multiple queens can be established when queens are separated by a mechanical barrier, such as a queen excluder (Farrar, 1953; Loubet de I’Hoste, 1959; Holzberstein, 1955; Wafa, 1956; Wallrebenstein, 1955). However, in the absence of a physical barrier in the brood nest, queen bees launch fatal battles until only one survives. Virgin queens are more aggressive than mated ones (Darchen and Lensky, 1963a). Independent of the mutual aggressiveness of queens, worker bees also play a major role in the elimination of supernumerary queens in experimental or feral colonies. Multiple queens are tolerated in a bee colony only during the swarming period (Darchen and Lensky, 1962, 1963b; Lensky et al., 1970). Even following the severing of mandibular tips and stingers of the queens to prevent mutual aggression (Darchen, 1960) and their successful introduction into a single brood chamber of a bee colony, all but one were balled by workers and eliminated 4 to 6 weeks later (Lensky et al., 1970). It is generally believed that pheromonal secretions from queen mandibular glands affect the aggressive behaviour of workers towards queens (Gary, 1962; Yadava and Smith, 1971a). However, volatile tcorrespondence and reprint requests should he addressed to: P. Cassier.
compounds originating from the sting apparatus or its vicinity are likely to be involved in the mutual recognition of queens, as well as in evoking aggressive behaviour of workers towards queens (“balling”). Worker Koschewnikow glands (WKG) secrete alarm pheromone (Grandperrin and Cassier, 1983; Mauchamp and Grandperrin, 1982), but Koschewnikow glands of queens may have a different function. Although Butler and Simpson (1965) attributed to the secretion of queen Koschewnikow glands an olfactory non-specific attractive effect on workers, their function has not yet been established. The aggressive behaviour of workers towards queens is well documented: balled queens frequently opened their sting chambers, protruding their stingers and sting sheath. Balling workers were attracted to the sting sheath and licked it (Robinson, 1984). Also, a queen’s abdomen is more attractive to workers than the head, presumably due to the secretion of tergal and Koschewnikow glands (De-Hazan et al., 1989a). It seems that some pheromonal secretions of queen bees may elicit: (a) aggressive behaviour between two queens, following mutual detection, which in general results in the death of one of them, and; (b) balling behaviour of worker bees towards either an introduced, or one of multiple queens in the nest, except during swarming season. The balling behaviour of workers has been explained by a “stress pheromone” hypothesis: a disturbed queen produces a “stress pheromone” that stimulates an attack on her (balling) and her own death (Yadava and Smith, 1971b,c). However, the glandular origin and chemical composition of the “stress pheromone” have not yet been established.
The purpose of this research is: (1) to describe the fine structure of Koschewnikow glands from queens of different ages; (2) to document the reaction of workers to extracts from queen (QKG) and worker Koschewnikow glands (WKG), and to induce in workers balling behaviour of a ‘“pseudoqueen” treated with QKG. To avoid the effect of possible pheromonal secretions from queen mandibular, tarsal (Cassier et al., 1991) and/or tergal glands (cf. De-Hazan et al., 1989a; De-Hazan et al., 1991), we used worker bees as “pseudoqueens”, following their treatment with QKG extracts; (3) to analyse the composition of queen Ko~he~ikow glands (QKG) by gas liquid chromatography-mass spectrometry. MATERIALS
Italian honeybee (Apis melriferaL. var. ligustica)queens and workers were used for structural, behavioural and chemical studies. Queens were reared using a grafting method (Laidlaw and Eckert, 1962), and they emerged
inside cages. They were marked with paint on their thorax and were kept inside the emerging cages in bee colonies until 4-5 days-old after which their Koschewnikow glands (QKG) were dissected. For behaviourai and chemical studies, QKG were removed under a stereoscopic microscope as follows: the queen’s stinger was grasped with forceps, pulled forward and the entire sting apparatus removed and pinned. The spiracuiar plate was cut off with iris scissors, and the Koschewnikow glands were removed. Fourteen glands (seven queens) were immersed in 10 ~1 cold absolute ethanol (EtOH) in conical vials (Pierce, Rockford, IL) and then crushed with a fine entomological pin. The samples were kept at - 18°C until used for bioassays. WKG were removed from workers using the same procedure. For light microscopy, queen abdomens were immersed in Bouin’s fixative, dehydrated in ethyl and butyi alcohol series, and impregnated and embedded in paraffin wax at 58°C. A series of cross sections (6 pm) were stained by Masson’s (var Goidner) Trichrome (Martoja and Martoja, 1967). For-lipid detection, stings were fixed in Baker’s fluid (Martoia and Martoia, 1967): glands were removed and sectioned after double embedding in gelose-paraffin wax. A series of cross-sections were stained in an isopropanoi Sudan-Black B saturated solution (Gabe, 1968). For electron microscopy, stings were removed and fixed for 20 min. in cold (4°C) fixative (1% giutaraidehyde in 0.1 M sodium cacodyiate buffer at pH 7.4 with 8% sucrose) and rinsed in the same buffer. After 1% osmic acid fixation and dehvdration, glands were embedded in Epon 812Araldite.- Semi-thin-sections (0.5-l pm) were stained with 1% toiuidine bue in 1% sodium borate solution. Ultra-thin sections collected on copper grids were stained by uranyl acetate and contrast enhanced by lead citrate (Reynolds, 1963). One, 5 and 9day-old virgin queens and 6, 12 and 18 month-old mated queens were used in this study. For ~tr~tructurai cytochemist~ of carbohydrates, ultrathin sections collected on gold grids were treated by periodic acid-thiocarbohydraxide-silver proteinate (Thiery, 1967); the controls were not treated with thiocarbohydrazide or periodic acid. Localization of proteins was carried out using pronase with or without periodic acid oxidation (Monneron and Bernhard, 1966). Grids were observed with Philips EM 300 or EM 201 at 80KV. For scanning electron microcopy of the gland, queens were kept at - 10°C for 10min; their stings were then removed and fixed for 20min in cold (4°C) fixative (1% glutaraldehyde in 0.1 M sodium cacodyiate buffer at pH 7.4 with 8% sucrose) and rinsed in the same buffer before dehydration in ethyl alcohol. After coating
(gold/palladium), observations were made with either a Cameca 07 or a Jeoi 840 A scanning electron microscope. Observation of worker reaction to “pseudoqueens”: A bee colony with about 2000 workers, headed by a 10 monthold laying queen, was established in a glass observation hive containing one full-depth and one shallow frame. The comb contained broods at all stages of development, honey and pollen. For the observation period, the hive was placed in the shade, and the prevailing temperature was approximately 25°C. To examine the effect of QKG and WKG on balling, we used workers that were removed from the observation hive colony instead of queens. These workers were used either treated with QKG or as control “pseudoqueens”, as described below, and after being introduced to their own colony the reaction of their sister workers was observed. All “pseudoqueens” were marked with a white spot (Tipp-Ex, Frankfurt, Germany) on their thorax. Pseudoqueens Two microlitre EtOH extract of QKG from 4-5 day old queens equivalent to 2.8 glands, were applied to the abdomen of each pseudoqueen, using a calibrated glass capillary (Brand, Wertheim, Germany). Controls included: (a) solvent control: 2~1 EtGH were applied to the abdomen of each worker bee, (b) WKG control: 2~1 EtOH extract of Koschewnikow glands removed from workers, equivalent to four glands, were applied to each worker bee. Koschewnikow glands were removed, as described above, from guardian workers that were captured on the alighting boards of beehives. Twenty glands were coBected from 10 workers and transferred into 10 ~1 absolute EtOH in conical vials. For the experiment, workers were removed with an aspirator from the observation hive, marked and topically treated with one of the extracts. Immediately thereafter they were introduced through the flight hole of the hive. Continuous observations were made following the intr~uction of the treated or control “pseudoqueens”. Only one test per day was carried out in the observation hive. Chemical analysis Gas chromatography-mass spectrometry (GC-MS). Four to 5 day-old queens were dissected (see above) to remove Ko~hewnikow glands. Ten glands obtained from five queens were transferred into a 2OOgl conical vial (Pierce, Beifonte, IL) containing 20 pl absolute EtOH. The glands were crushed with a fine glass capillary, the vial centrifuged at 10,000g for IO min, and the clear supematant was stored at - 18°C until used. One to 2 ~1 samples of QKG EtOH extracts were injected and separated on SE-54 fused capillary columns (i.d. 0.23 mm, 50m long, Chrompack, Middleburg, The Netherlands). The column temperature was maintained for I min at 60°C and then programmed from 60 to 220°C at S”C/min. GC-MS analysis were performed on a Finnigan, model 4500 GC-Mass spectrometer connected with Incas Data System and with NBS Mass spectra Library. Ionization was performed under Electron Impact (70 eV). RESULTS
Organization and structure of Koschewnikow glan& of the queen honeybee Localization and anatomy. (Figs 1 and 2) Koschewnikow glands are part of the sting apparatus of the females of most social Hymenoptera Apidae (Koschewnikow, 1899), Vespidae (Altenkirch, 1962) and primitive Formicidae (Robertson, 1968). In the queen honeybee, each gland is located in the 7th abdominal segment. The gland appears in young queens as a wide cellular mass (600 x 300 x 60 pm)
“Stress” pheromone of queen bees
Fig. 1. Scanning electron microscopy. Lateral view of the sting apparatus of a queen bee. The spiracular plate ($) partly obscures the Koschewnikow gland (-) x 48. appressed against the quadrate plate. In old ovipositing queens the thickness of the gland decreases to 30 pm or less. Dorsal to this gland are motor sting muscles and the tracheal sac. The glands are closely inserted between the quadrate plate and the spiracular plate and organised along an intersegmental membrane joining these cuticular plates. The glands are composed of a single type of secretory unit, type III as described by Noirot and Quennedey (1974). Each unit is composed of a large egg-shaped glandular cell provided with a cellular reservoir where the secretory product is extruded and then drained off by a duct elaborated by a narrow duct cell. In the old ovipositing queen, the glandular cells are separated by wide spaces indicating degeneration; the cuticular ducts are the last vestiges of the secretory units. The duct cells appear as bundles and are located at the level of the flat epithelial cells of the intersegmental membrane where the ducts open at random and the secretions are emitted. These secretions then flow into the sting chamber and reach the setaceous membrane. Tracheal cells and their tracheae are numerous among the secretory units, especially in the peripheral part of the gland. No nerve endings are seen in close association with the glandular units. Ultrastructure (Figs 3-6) Virgin queens (I- to 9-day-old) and young mated queens (3-month-old). Glandular cells, in the basal
part along the spiracular plate, rest on a fibrillar basement membrane (75-100 nm thick). Their plasma membranes form long (2.5 pm) and narrow invaginations where the basement membrane extends. In the apical region, the cells are shortened in a crypt-like reservoir (cellular reservoir: 10 pm long) with numerous, long microvilli (l-10 pm) arranged in small groups. Above the crypt, the glandular cell is closely attached to a duct cell by thin duct cell
Fig. 2. Transverse section of the 7th abdominal segment of a queen honeybee. Schematic drawing. 1. 7th tergum; 2. hind gut; 3. sting chamber; 4. stylet of the sting; 5. setaceous membrane; 6. epithelium; 7. spiracular plate; 8. quadrate plate; 9. oblong plate; 10. Koschewnikow gland; 11. 7th sternum; 12. fat body; 13. intersegmental membrane.
expansions with septate desmosomes. The large basal nucleus (8-10 pm) of the glandular cell inludes 4-10 small nucleoli (1 pm diameter) and numerous clusters of chromatin (300 nm). In the cytoplasm, dictyosomes of the Golgi apparatus are rare. Rough endoplasmic reticulum cisternae, rich in ribosomes and arranged in bunches, are well-developed close to the plasma membrane and the nucleus. Abundant mitochondria, short (0.75 pm) and poorly crested (2-3 crests), are mostly located among R.E.R. cisternae. In the glandular cells of 3-, 5- or 9-day-old virgin queen bees, the cytoplasm contains a heterogenous population of mitochondria: normal ones of reducedsize are present together with some hypertrophied ones with rare, weavy cristae and osmiophilic bodies
Fig. 3. Glandular cell of the Koschewnikow gland of a 3 month-old queen bee. N: nucleus with several nucleoli. R: reservoir of the cell with microvilli and transverse sections of the duct (d). The cytoplasm shows a heterogenous population of mitochondria (M) and several multivesicular bodies (MVB). x 3300. Fig. 4. Longitudinal section of the end apparatus. The cuticular duct (d) is only composed of two epicuticular layers: a fenestrate external epicuticle (+) and internal epicuticle composed of intricate filaments (*). m = Microvilli; M = giant mitochondria; head-arrow = osmiophilic material. x 25,600.
embedded in a granular matrix. The double envelope of these involutive mitochondria progressively disappears.
In the 3 month-old-mated queen bee, the number of bodies originating from mitochondria increases. These contain lipid droplets, myeloid structures, a
“Stress” pheromone of queen bees
Fig. 5. Koschewnikow gland of a 3 month-old queen bee. Cytoplasm of the glandular cell contains numerous involutive mitochondria (M) x 5600. MVB = Multivesicular body; N = nucleus; RER = Rough endoplasmic reticulum. Fig. 6. Bodies originating from mitochondria in a glandular cell from the Koschewnikow gland of a 3 month-old queen bee. Bodies show a positive Thiery’s reaction (oxidation, TCH: 1 h, Ag: 30 min). x 26,000. Fig. 7. As above, treated (1 hr) with pronase, after periodic acid oxidation. x 54,000. dense matrix containing glycoproteins (positive Thiery’s reaction) and pronase extractable proteins. Similar material can be observed in the duct, but the exact mechanisms of cellular extrusion remain unknown.
Duct ceils. The reservoir-like
cell is narrow of the plasm
crypt of the glandu tlar obturated by a long (lo-70pm long) smd (2.5 x 6.4 pm) duct cell. It widens at the le:vel osmiophilic nucleus (2.5 x 6.4 pm). The cy‘tois clear and scant, containing only a 1few
YAACOVLENSicY er al.
ribosomes and mitochondria. The cuticuiar duct is composed of only two epicuticular layers: an external epicuticle (25 nm, osmiophilic) and internal epicuticle (loo-120 mn, clear). In the crypt, the wall of the endapparatus is differentiated into a fenestrate external epicuticle and an internal epicuticle composed of intricate hlaments. One-year-old ovipositing queen. (Figs 7-11) The degeneration process affects only the secretory units whose number is progressively reduced; the first degenerating cells are the duct cells. Along the intersegmental membrane, the epithelial cells are unaffected. In the duct cells, nuclear involution is characterized by pycnotic chromatin and cleared nucleoplasm. Inside the cytoplasm, large spherical or polygonal lipoidic inclusion (up to 2.5 pm diameter) are filled with heterogeneous lipoprotein material. In the nucleus of the glandular cells, pycnosis shows unknown intranuclear material (up to lOpm length) in which alternate parallel strips of dense (25 nm wide) and clear material (6 nm wide) provide a pseudo-crystalline texture. Mitochondria are dilated and contain numerous crests. Degeneration results from cellular lysis. The lysis of the cell membrane releases the cell’s content into the intercellular spaces, which are then occupied by small amounts of cytoplasm, mitochondria, lipoidic inclusions and flat pseudo-crystalline formations (up to 12 pm long). These crystals are mostly located in extracellular positions but some were observed inside degenerating glandular cells. Responses of worker bees to queepr and worker Koschewnikow gland extracts Reaction of workers to PQ covered with either EtOH extracts of queen or worker Koschewnikow glancis. (a) PQ covered with QKG extract. Each introduced worker pseudoqueen was immediately surrounded by hive bees, who formed a dense ball around her, consisting of about 15-35 workers. The “ball” around the PQ persisted for 5 to 10 min, after which about two workers remained. Approximately 4min later, they abandoned the PQ (see Fig. 12). Workers who participated in balling of the PQ displayed aggressive behaviour, such as grasping and biting of the wings and hind legs, as well as pulling. We also observed non-aggressive behaviour of workers participating in the ball, such as antennating and licking. Of 10 PQ, seven were balled; two were aggressively attacked following their introduction, but there was not any dense ball formation of workers; and one PQ succeeded in escaping from the hive and flew off after being attacked by workers. None of the nine PQ that were balled died. When the balling was terminated, the PQ could move around, but in most cases their wings were damaged (Table 1). (6) PQ covered with WKG extract. Immediately after introduction, about 6-8 workers approached the pseudoqueen, pushed her and arched their abdomen as during stinging, while other only antennated her and then abandoned her. The hive workers did not display their aggressive behaviour continuously: they would approach the PQ and then retreat. This specific behaviour lasted for about I .0-l 5 min.
There were six repetitions of this observation, continuing for 4min. Successive observations place at 5 min intervals (Table 1).
Reaction of workers to control ‘tpseudoqueens” (PQJ (a) PQ marked with white paint (Tipp-Ex) only. The introduced PQ did not arouse any attention inside the hive. The observation was continued for 2min, and there were three repetitions using three different “pseudoqueens”. (b) PQ marked with white paint and covered with 2111EtUff. Introduced PQ were inspected and antennated by about five workers for about 30 sec. Each observation was continued for 3 min, and there were five repetitions. Gas liquid chro~togr~hy~~s
Ethanolic extracts of QKG were analysed by GC-MS electron impact and the 28 compounds that were characterized are listed in Table 2, according to their molecular weights, which ranged from 112 to 604. The Table shows that the compounds (Cs H,,-C,, H,,) included acids, alcohols, alkanes and alkenes. None of these compounds is present in workers’ alarm pheromone, which is produced in workers’ Koschewnikow glands (Mauchamp and Grandperrin, 1982). DISCUSSION
The Koschewnikow glands of the queen bee. show the same disposition and general organi~~on as those of worker bees (Grandperrin and Cassier, 1983). These paired organs are inserted on the sting apparatus between quadrate and spiracular plates. They are exclusively composed of type III glandular units, which predominate in the epidermal glands of Insects (Noirot and Quennedey, 1974). Nevertheless, the Koschewnikow glands of queen bees differ from similar glands of workers bees by their imaginal ontogeny and their pheromonal function. Whereas in workers, the secretory product is composed of protein, lipoprotein and lipid substances, in the queen bee the Koschewnikow gland accumulates granules with ~y~protein-~ch components. The secretory granules are derived from degenerating mitochondria and the successive steps of their ontogeny have been observed. At the end of this process, the normal organization of mitochondria disappears (cristae, double envelope) and the granules surrounded by a single membrane contain lipid droplets, myeloid structures, and a dense matrix containing glycoproteins and proteins. Similar material can be observed in the ducts, but the exact mechanism of cellular extrusion remains unknown. Such a secretory evolution of mitochondria is not uncommon; it has been described in the mandibular glands of queen bees (De Hazan et al., 1989b) and of worker bees (Vallet, 1988; Vallet et al., 1991) and in the ovaries of Molluscs and Amphibia during vitellogenesis. The secretory products observed at the ultrastructural level in the Koschewnikow glands of worker and queen bees are likely to be only carriers of the volatile pheromonal components. The presence
Fig. 8. Pseudocrystalline Fig. 9. Nucleus
in a degenerative cell. x 25,000.
cell. G. C. = Normal
(N) of a glandular cell in the Koschewnikow gland of an 18 month-old the presence of periodically arranged nuclear inclusions. x 32,000.
Fig. 10. 18 Month-old
Fig. 11. Degenerating
Lysis and nuclear pycnosis cells. x 11,200.
duct cell with lipoidic
of duct cells (D.C.).
queen bee. Note G.C. = glandular
Table 2. Chemical composition of Koschewnikow
ghmds of 4&yold virain aueens determined bv GC-MS fElectron Imnact)
Fig. 12. Balling intensity and duration of pseudoqueens by their “sister” workers. Each pseudoqueen was covered with 2~1 of Queen Koshewnikow Glands EtOH extract, equivalent to 2.8 glands.
of proteins and/or glycoproteins in the majority of epidermal glands involved in pheromone secretion is now a well-established fact (cf. Noirot and Quennedey, 1974). The Koschewnikow glands of summer worker bees are permanent organs; in contrast these glands degenerate in one-year-old mated queens similar to the fate of the venom gland (Fyg, 1956), but the factor(s) inducing this process is (are) unknown. To determine whether the degenerative process is correlated to ageing of individuals, it will be necessary to determine the fate of Koschewnikow glands of long-living winter workers in a temperate climate. The progressive degenerations concern duct and glandular cells where nuclei show a precocious pycnosis. The cytoplasms are then invaded by large lipidic inclusions, and pseudocrystalline formations are released in the intercellular spaces. Such proteinic inclusions are also observed in degenerative cells of the dog epididymis (Gouranton et al., 1977) and in the midgut of Tenebrio molitor larvae (Gouranton, 1969). We demonstrate by topically treating worker bees with QKG a specific balling behaviour toward “pseudoqueen” workers similar to that displayed towards queens. Since the workers used as pseudoqueens were removed from and introduced into their own colonies, we eliminated the effect of a foreign colony odour, which could have been responsible for aggressive behaviour towards an introduced worker or queen.
1,1,3-Trimethyl cyclopentane 5,5-Dðyl-2-hexene 3,3-Dimethyl-hexane Gctenal Methyl cyclcdecane p-menthane-9-01 4,5-Dimethyl-nonane 2-Propyl-1-heptanol 4,6,8-Trimethyl-I-nonene Nonanoic acid Decanoic acid 1,12-Tridecadiene 1.1 I -Dodecadiene Cyclohexyl hexenol Ethyl decanoate 2-Methyl-I-dodecanol Hexadecane Ethyl dodecanoate Dodecyl acetate 6-Cyclohexyl undecane Hexadecanoic acid Ethyl tetradecanoate Methylester 2-methyl hexadecanoate 2-(Hexadccyl oxy)-ethanol 2,6,10,1 S-Tetramethyl-heptadecane 1-Dotriacontanol 1,7-Pentatriacontene 3,5,24-Trimethyl-tetracontane
112 112 118 126 154 156 156 158 168 172 172 180 180 180 200 200 226 228 228 238 256 256 284 286 296 466 490 604
The aggressive reaction of hive workers toward a worker covered with WKG extract was similar to that of guardian bees towards an intruding foreign worker during which alarm pheromone is released. However, this reaction did not resemble the behavioral pattern of “balling” of a queen bee. We observed several violent and non-violent activities towards the pseudoqueens treated with QKG extracts, as described in the case of balled queens (Yadava and Smith, 1971~; Robinson, 1984). The onset of balling of a pseudoqueen covered with QKG extract would occur almost immediately following her introduction into the observation hive, whereas it took about 8.5 min for the behaviour to begin toward a foreign queen, as reported by Robinson (1984). Moreover, the relatively short balling duration of pseudoqueens (7 min), as compared to that of an intact queen (58 min, Robinson, 1984) may be due to the rapid evaporation rate of EtOH extracts from the body surface of a worker versus continued secretion from the gland and its release from the setaceous membrane in the queen bee. One may presume that QKG secretions play an important role in the mutual detection of two queens
Table I. Response and behavioural patterns of workers inside a glass observation hive to a “pseudoqueen” worker (PQ) covered with extracts of queens’ (QKG) or workers’ Koschewnikow glands (WKG). Both PQ were marked with Tipp-Ex white spot (T) on the thorax Treatment
Behaviour Behavioural patterns observed General agitation of workers Antennating Aggressiveness: biting wings, legs and other body parts Balling Duration of balling, min No. of workers participating in balling Behaviour of workers participating in balling
Treated PQ+T+2pI QKG
Treated PQ+T+ZpI WKG
Control PQ+T+Zfll EtOH
7-10 35 Pushing, towing, holding wings and legs, biting wings and other
“Stress” pheromone of queen bees
and the subsequent duels between them. When their sting chambers were sealed off and/or antennae were covered with wax, queens were unable to detect each other and did not fight (Lensky et al., 1970). In addition, the attraction of workers to queens can be modified by sealing off their tergal glands and sting chamber with paraffin (De-Hazan et al., 1989a). The pheromonal secretion of the QKG which is released on the setaceous membrane did not elicit any behavioural pattern that is characteristic of worker alarm pheromone. It assumes the function of a “suicide” pheromone and thus seems to be instrumental in the maintenance of a monogynous status of each bee colony: a supernumerary queen or a pseudoqueen covered with the QKG extract is balled and may be eliminated. The chemical composition of pheromonal components of the WKG and QKG is entirely different. We have not detected any of the alarm pheromone components of worker bees (Both et al., 1962; Pickett et al., 1982; Mauchamp et Grandperrin, 1982; Free et al., 1988) in QKG extracts. The GC-MS characterization of QKG EtOH extracts reveals that none of the 28 compounds, including acids, alcohols, alkanes and alkenes (C$H,,C,,H,,), has been reported as a component of worker bee alarm pheromone (Free et al., 1988). However, we found three of the QKG compounds in other pheromonal secretions of the queen bee, as follows: p-menthane-9-01 is also present in the mandibular glands of 3-5 day-old queens; hexadecanoic acid is also in the secretions of mandibular, tergal and tarsal glands and 1,7_pentatriocontene is found in mandibular and in tarsal gland secretions of 6-24 month-old queens (Cassier et al., 1991; De-Hazan, 1986; Finkel, 1983; Hyams, 1988). Behavioural responses of queens and workers to each of the characterized components of QKG may provide us with further information of their pheromonal function. We do not preclude the possibility that other queen bee pheromones, such as those originating from the mandibular (cf. Gary, 1962; Yadava and Smith, 197la) and tergal glands (De-Hazan et al., 1989a; De Hazan et al., 1991) act synergistically with QKG secretions. This report raises many further questions, such as: what triggers the secretion of QKG and/or the exposure of the setaceous membrane? Is the QKG pheromone also released by a well-performing, laying-queen who is surrounded and groomed by a court of workers in a normal colony? Which of the components of QKG elicit the pheromonal signal? Are there any synergistic effects of mandibular glands secretions? We are currently studying these and other related topics. Acknowlednemenfs-This
research was carried out within the framework of an agreement for scientific collaboration between the Universite Pierre et Marie Curie (Paris VI). Paris and the Hebrew University of Jerusalem; and was partially supported by the Triwaks Foundation, the E. D. Bergmann Research Fund and the Central Research Fund of the HUJ. We thank Miss Hani Gal for her excellent assistance in the preparation of this manuscript, Dr Karin Zupko, Hebrew University, Rehovot, Dr Catherine Lange
and Dr Jean-Claude Cherton, Universite Pierre et Marie Curie, Paris, for their critical remarks and helpful suggestions, Dr Arie Tishbee and Mr Roger Schlesinger, Dept of Organic Chemistry, Weizmann Institute of Science, Rehovot, for GC-MS analyses.
Altenkirch G. (1962) Untersuchungen ilber die Morphologie der abdominalen Hautdriisen einkeinuscher Apiden. Zool. Beit. 7, 161-238. Both R., Shearer D. A. and Stone B. C. (1962) Identification of isoamyl acetate as an active component in the sting pheromone of the honeybee. Nature (Land.) lM, 1018-1020. Butler C. G. and Simpson J. (1965) Pheromones of the honeybee (Apis melltfira L.). An olfactory pheromone from the Koschewnikow gland of the queen. Ved. Pr. Wyz. Us. Vcel. Dole 4, 33-36.
Cassier P., Finkel A. and Lensky Y. (1991) The chemical composition of tarsal gland secretions of honeybee, Apis mell$era (L.) queens, workers and drones. (Submitted). Darchen R. (1960) L’ablation du dard des reines et des ouvriBres d’Apis_ mel&ica. C. R. Acad. Sci., Paris 250, 934-936.
Darchen R. and Lensky Y. (1962) Les so&es polygynes de Reines (Apb me/iifica var. ligustica). C. R. Acad. Sci., Paris 255, 2300-2302.
Darchen R. and Lensky Y. (1963a) Etude preliminaire des facteurs favor&ant la creation des sociitts polygynes d-Apis mell$ca var. ligustica. Ann. Abeille 6, 69-13.
De-Hazan M. (1986) The attractiveness, composition and structure of mandibular glands of the queen honeybee (Apis meilifera L.). MSc. Thesis. Hebrew University of Jerusalem. De-Hazan M., Lensky Y. and Cassier P. (1989a) Effects of queen honeybee (Apb melrifea L.) aging on her attractiveness towards workers. Comp. Biochem. Physiol. 93A, 717-183.
De-Hazan M., Hyams J., Lensky Y. and Cassier P. (1989b) Ultrastructure and ontogeny of the mandibular glands of the queen bee, Apis me/&ra L. (Hymenoptera, Apidae). Int. J. Morph. Embryol. 18, 311-320.
De-Hazan M., Lensky Y. and Cassier P. (1991) The chemical composition of the pheromonal secretion from the tergal glands of the queen bee (Apis meiiifera L.). (Submitted). Farrar C. L. (1953) Two queen colony management. Bee World 34, 189-194.
Finkel A. (1983) Biological effects and chemical characterization of the foot-print substance in the honeybee (Apis melrifera L.). M.Sc. Thesis. Hebrew University of Jerusalem. Free J. B., Ferguson A. W. and Simpkins J. R. (1988) Honeybee responses to chemical components from the worker sting apparatus and mandibular glands in field tests. J. Apic. Res. 28, 7-21. Fyg W. (1956) Merkmale zur Alterbeistimmung von Bienenkoninginnen. Schweiz. Bienen-Ztg. 10, 428-432. Gabe M. (1968) Techniques histologiques. Masson, Paris. Gary N. E. (1962) Antagonistic reactions of worker honeybees to mandibular gland contents of the queen bee. Bee World 42, 14-17.
Gouranton J. (1969) Observations cytochimiques et ultrastructurales sur les cristaux intranucleaires de l’intestin moyen de la larve de Tenebrio molitor. C. R. Acad. Sci. Paris 268D, 2948-2951. Gouranton J., Folliot R. and Thomas D. (1977) Mise en evidence de cristaux proteiques dans le cytoplasme et le noyau des cellules Cpithdliales de la queue de l’epididyme chez le Chien. Biol. Cell. 19, 31a.
Grandperrin D. and Cassier P. (1983) Anatomy and ultrastructure of the Koschewnikow’s gland of the honeybee, Apis melltfera L. (Hymenoptera: Apidae). ht. J. Insect Morphol. & Embryol. 12, 25-42. Holzberstein J. W. Jr (1955) Some whys and hows at two queen management. Glean. Bee Cult. 83, 344347. Hyams Y. (1988) Characterization of pheromonal components of the tarsal and mandibular glands of the queen honeybee (Apis mellifera L.). MSc. Thesis. Hebrew University of Jerusalem. Koschewnikow G. A. (1899) Zur Kenntnis der Hautdriisen der Apidae und Vespidae. Anat. Am. 15, 519-528. Laidlaw H. H. and Eckert J. E. (1962) Queen Rearing. Dadant and Sons, Hamilton, IL. Lensky Y., Darchen R. and Levy R. (1970) L’aggressivite des reines entre elles et des ouvrieres vis-a-vis de la creation des socittes polygynes d’dpis mellifica L. Rev. Comp. Animal. 4, 50-62. Loubet de 1’Hoste E. (1959) La biruche. Laubet de I’Hoste, Toulouse. Martoja R. and Martoja M. (1967) Initiation aux techniques de I’Histologie animale. Masson, Paris. Mauchamp B. and Grandperrin D. (1982) Chromatographie en phase gazeuse des composes volatiles des glandes a pheromones des Abeilles: methode d’analyse directe. Apidologie 13, 29-37. Monneron A. and Bernhard W. (1966) Action de certaines enzymes sur les tissus inclus en Epon. J. Mcrosc. (Paris) 5, 697-714. Noirot C. and Quennedey A. (1974) Fine structure of insect epidermal glands. Ann. Rev. Enromol. 19, 61-80. Pickett J. A., Williams I. H. and Martin A. P. (1982) (2)-II-Eicoson-l-01, an important new pheromonal component from the sting of the honeybee, Apis melltfera L. (Hymenoptera, Apidae). J. Chem. Ecol. 8, 163-175. Reynolds E. S. (1963) The use of lead citrate at high pH as an electron opaque stain in microscopy. J. Cell Biol. 17, 208-217.
Robertson P. L. (1968) A morphological and functional study of the venom apparatus in representatives of some major groups of Hymenoptera. Austr. J. Zool. 16, 133-166. Robinson G. E. (1984) Worker and queen honeybee behavior during foreign queen introduction. Insectes Sot. 31, 254-263. Thiery J. P. (1967) Mise en evidence des polysaccharides sur coupes fines en microscopic electronique. J. Microsc. (Paris) 6, 987-1018. Vallet A. (1988) Evolution de la glande mandibulaire des ouvrieres d’Apis melltfica en relation avec l’agressivite: analyse morphologique, chimique et tthologique. M&noire DEA, Universite Pierre et Marie Curie (Paris VI). Vallet A., Cassier P. and Lensky Y. (1991) Ontogeny of the fine structure of the mandibular glands of the honeybee (Apis melltfera L.) workers and the pheromonal activity of 2-heptanone. J. Insect. Physiol. (In press). Wafa A. K. (1956) Two queen colonies for a plentiful yield of honey, safe wintering, means of propagation and swarming control. Bull. Fat. Agr. Cairo Univ., 98, 22. Wallrebenstein W. (1955) Mein Beitrag zum Mehrmutterverfahren. 17th Int. Beekeep. Congr. Yadava R. R. S. and Smith M. V. (1971a) Aggressive behavior of Apis mellifera L. workers towards introduced queens II. Role of mandibular gland contents of the queen in releasing aggressive behavior. Can J. Zool. 49, 1179-1183. Yadava R. R. S. and Smith M. V. (1971b) Aggressive behavior of Apis mellifera L. workers towards introduced queens III. Relationship between the attractiveness of the queen and worker aggression. Can. J. Zool. 49, 1359-1362. Yadava R. R. S. and Smith M. V. (1971~) Aggressive behavior of Apis mellifera L. workers towards introduced queens I. Behavioral mechanisms involved in the release of worker aggression. Behaviour 39, 21 I-236.