Mating Expenditures Reduced Via Female Sex Pheromone Modulation

Ó Springer-Verlag 1999

Behav Ecol Sociobiol (1999) 45: 95±106

ORIGINAL ARTICLE

Manfred Ayasse á Wolf Engels á Gunter LuÈbke Timo Taghizadeh á Wittko Francke

Mating expenditures reduced via female sex pheromone modulation in the primitively eusocial halictine bee, Lasioglossum (Evylaeus) malachurum (Hymenoptera: Halictidae) Received: 30 April 1998 / Accepted after revision: 24 July 1998

Abstract The present paper reports on behavioral experiments and gas chromatographic analysis of chemical communication in the mating biology of the primitively eusocial sweat bee Lasioglossum (Evylaeus) malachurum. In a dual-choice experiment, a female made odorless was signi®cantly less attractive than an untreated one. Attraction in L. (Evylaeus) malachurum is therefore mediated by a female-produced sex pheromone. Further bioassays showed that unmated gynes are more attractive to males than mated ones. Males are able to differentiate between the two groups of females as little as 3 h after mating. Biotests with di€erent samples obtained from attractive gynes showed surface extracts to be most attractive. Behavioral tests with synthetic copies of the compounds identi®ed as cuticular constituents of virgin gynes were highly attractive to males; the volatile bouquets consisting of n-alkanes, n-alkenes and isopentenyl esters of unsaturated fatty acids were the most attractive samples. Isopentenyl esters of unsaturated fatty acids were the key compounds in inducing male inspections as well as stimulating pounces and copulatory attempts. Virgin and nesting gynes di€ered clearly in the relative and absolute amounts of the volatiles on the cuticle. The total amount of volatiles was signi®cantly higher in virgin gynes and decreased in breeding queens. Hydrocarbons were the dominant group of

M. Ayasse (&) Institute of Zoology, Department of Evolutionary Biology University of Vienna, Althanstrasse 14, A-1090 Vienna, Austria e-mail: [email protected] Tel.: +43-1-313361398, Fax: +43-1-31336778 W. Engels Institute of Zoology, Department of Developmental Biology University of TuÈbingen, Auf der Morgenstelle 28 D-72076 TuÈbingen, Germany G. LuÈbke, T. Taghizadeh, W. Francke Institute of Organic Chemistry and Biochemistry University of Hamburg, Martin-Luther-King-Platz 6 D-20146 Hamburg, Germany

compounds in both groups of females. The relative amounts of the wax-type ester, hexadecyl oleate, isopentenyl esters, and a hitherto unidenti®ed steroid were higher in attractive virgin gynes, while the relative proportions of hydrocarbons and lactones dominated in nesting queens. The site of sex pheromone production in attractive young L. (Evylaeus) malachurum gynes remains unknown. Head glands or Dufour's gland secretions may be involved. Another possible source of the `active principle' found among the cuticular lipids could be glandular cells of the epidermis. The signi®cance of modulation of female sex pheromone composition is discussed in terms of a reduction in mating expenditures. Key words Lasioglossum (Evylaeus) malachurum á Sweat bee á Sex pheromone modulation á Mating behavior á Chemical analyses

Introduction Bees have evolved a remarkable variety of mating strategies (Eickwort and Ginsberg 1980). Males may search for mates at resources like nests or ¯owers (Alcock 1978; Alcock et al. 1978). In social bees, the males may form leks called drone congregations that function as rendezvous to which gynes but not workers are attracted (Engels 1988). A speci®c situation is seen in those ground-nesting bees that can be found in aggregations of sometimes thousands of nests (Michener 1990). In the sweat bee, Lasioglossum (Evylaeus) malachurum, large numbers of reproductives emerge during a short mating season (Knerer 1992). On warm summer days, July through September in central Europe (Westrich 1989), the nest aggregation area then becomes a lek (Thornhill and Alcock 1983) with many bees ¯ying around close to the ground. The daily operational sex ratio is strongly male biased (M. Ayasse, unpublished data), as in other bees (Wcislo 1992). Most females are mated gynes and foraging workers and only

96

few are receptive virgins. Consequently, there is strong competition among males for receptive females. Within these ¯ying aggregations, males are able to recognize conspeci®c gynes (Barrows 1975a). However, only the sparse receptive gynes provoke excitement and are approached. Among these, males prefer females to which they are not related (Smith and Ayasse 1987). Copulations are rarely seen and in L. (Evylaeus) malachurum, monandrous mating seems to be the rule (Ayasse 1987; M. Ayasse and B.H. Smith, unpublished data), although polyandry may occur (R.J. Paxton, unpublished data). Males can evidently di€erentiate gynes (Ayasse 1990). In L. (Dialyctus) zephyrum this has been interpreted as the consequence of an anti-aphrodisiac label printed by the copulating male on the female, thus making her unattractive for subsequent wooers (Kukuk 1985). Experiments on mating behavior, however, failed to con®rm this assumption (Barrows et al. 1975; Wicslo 1987), although all these observations indicate that males discriminate mates. Proximate requirements are releaser signals functioning as discriminators and enabling the males to select promising mates. The relative importance of visual and olfactory cues in this sweat bee mating system remains an open question. Previous studies on L. (Evylaeus) malachurum queens have analyzed the content and ontogenetic patterns of the volatile composition of Dufour's gland secretions (Ayasse et al. 1990a). However, this gland turned out not to be the main source of the female sex pheromone (Ayasse et al. 1993). The question where the female sex pheromone components are synthesized thus remained open. Here we present evidence on the production site and chemical composition of the signal emitted by receptive virgins. Analytical data were substantiated by ®eld bioassays. Modulation of the sex pheromone composition apparently minimizes both molestation of non-receptive gynes by males and ``useless'' mating attempts. This considerably reduces mating expenditures in the numerous encounters. In halictine bees, these mating economics may ultimately compensate the constraints imposed by gregarious nesting and by the mating strategy described above. The present paper reports on behavioral experiments and gas chromatographic analysis of the volatile signal used for chemical communication in the mating biology of the primitively eusocial sweat bee L. (Evylaeus) malachurum. The sex pheromone of the species was identi®ed by behavioral tests and chemical analysis. The odor bouquets of unmated (attractive) gynes [de®ned following Michener (1990) as females that will found nests or have done so, but do not have worker associates] and breeding (unattractive ) queens (females that have worker associates) were identi®ed and compared. The objective was to answer the following questions. Do the males recognize females' mating status? Which parts of the body emit the sex pheromone and which glands produce it? An e€ort was also made to investigate the possible function of different chemical classes of compounds in attracting males and stimulating copulatory behavior.

Methods Natural history L. (Evylaeus) malachurum is a primitively eusocial sweat bee common throughout the west Palearctic region (Knerer 1992). Nesting aggregations with up to several thousand nests can be found in bare soil sometimes covered with scattered vegetation. Seasonal activity starts in early spring (March±April) when nests are founded by single overwintered gynes. The ®rst brood consists only of workers (Sakagami 1974), although, depending on the climatic region, a few males can occasionally be found (Knerer 1992). In the last brood of the year (in TuÈbingen the second and in Vienna the third brood), ®rst males and several days later female reproductives emerge. Mating takes place at the nesting site, where large numbers of males patrol just above the nest entrances. Only young mated gynes overwinter and start nesting in the following year. Mating experiments were conducted from July to early September 1985±1993. Bees were collected at two nesting sites close to TuÈbingen (south Germany) and one near MoÈdling (Austria). Attractiveness of odorless and untreated gynes The ®rst experiment was a control to compare the importance of female odor versus visual cues in male attraction. Two to 3 days before the experiment was performed, emerging gynes were collected from the nest entrance early in the morning. Each gyne was placed in a cooled, clean plastic cup (Eppendorf) and killed by freezing. Half the females were extracted in pentane (Uvasol, Merck) for at least 24 h and subsequently heated at 60 °C for 8 h. This group of females was almost free of volatiles and termed ``odorless.'' The other half was stored in the freezer and termed ``untreated.'' The vials containing both groups of females were brought to the ®eld in a container with ice. Before each test, an odorless and an untreated gyne were removed from the plastic cups and individually ®xed on insect pins. In a dual-choice experiment lasting 3 min, the number of contacts of patrolling males with the two females, pinned 1.5 cm above the ground and 5 cm apart, was counted. The behavioral event termed contact could either be a brief touch or a copulatory attempt. Each pair of gynes was tested only once.

Attractiveness of mated and unmated young gynes In a second experiment we investigated whether males are able to discriminate between unmated and mated young gynes. In the ®eld, nests were excavated to obtain gyne pupae. In the laboratory, single pupae were transferred to 1.5 ml plastic cups (Eppendorf), which were closed with a piece of moist ®lter paper. The cups were stored in the dark at room temperature (21 °C), and every day a few drops of water were added to the ®lter paper until gynes emerged from the pupae. For each dual-choice experiment, pairs of sister bees were used. Three days after emergence, one of the two sister bees mated in a ¯ight cage (40 ´ 40 ´ 40 cm) with males collected in the ®eld. The unmated gynes were ¯ying in a cage without males. Three hours after mating, both sister bees (one unmated, one mated) were frozen and stored in the freezer at ±20 °C until they were used in the ®eld tests on the following day, where handling and testing were as described for the choice experiment using odorless and untreated bees.

Mating experiments with odor extracts and synthetic compounds Mating experiments similar to those described by Smith and Ayasse (1987) were performed to prove the role of volatiles in attracting males and stimulating copulatory behavior. Natural

97 extracts and blends of synthetic copies of the volatiles identi®ed in unmated young gynes were applied on pinned odorless gynes (dummy) treated before each test as described below. In each experiment, a dummy bee was impregnated with 10 ll of a pentane solution containing one female equivalent of the test solution. Before the dummy bee was ®xed 1.5 cm above the ground, the solvent was allowed to evaporate for 1 min. To avoid habituation, the tests were alternately carried out at two or three different spots within the male patrolling area; these sites were at least 10 m apart. Earlier marking experiments showed that the males do not move freely throughout the nesting aggregation (Smith and Ayasse 1987) and, therefore, di€erent males were present at the three testing spots. Each odorless bee was used only once per day; it was then extracted with pentane (Uvasol) and dried in an oven for at least 8 h at 60 °C prior to renewed use. Within a testing period of 3 min, three di€erent male reactions were recorded (after Kullenberg 1973): inspection ± a patrolling male approaches to within 5 cm of the dummy bee; pouncing ± a male contacts the dummy bee with its head or antennae; copulation ± a male mounts a dummy bee and tries to copulate. To combine the tests carried out on di€erent days and at di€erent localities where the number of patrolling males was not the same, the male reactions were standardized by measuring their ¯ight activity. A 1-m-long wire was mounted at the experimental area. The number of males crossing the wire in 3 min was counted at least every 30 min. For each 3-min test, individual reactions of the males were divided by the mean male ¯ight activity and multiplied by 100. Origin of the female sex pheromone In a ®rst experiment, attempts were made to localize the source of the sex pheromone of young and attractive gynes (collected 1 day before the experiments). Only females that were attractive to males were used. This was checked by placing one female together with three to ®ve males in a small plastic tube (10 cm high, diameter 5 cm). If a male tried to copulate within the next 30 s, the female was killed by freezing and used for odor extractions as described below. We checked for the emptiness of the spermatheca by dissections and found all of the attractive females to be unmated. The following samples were used in ®eld experiments: (1) Surface extracts. Single frozen females were extracted for 15 s in 1 ml diethyl ether (Merck). Subsequently, the diethyl ether was evaporated in a glass capillary in a water bath (40 °C) and 10 ll of pentane was added. (2) Extracts of whole females. The gynes were placed in 0.5 ml pentane (Uvasol) and stored at room temperature (21 °C) for 24 h, before they were removed from the solvent. (3) Extracts of the Dufour's glands. Individual Dufour's glands were extracted in 100 ll pentane (Uvasol) for 24 h. (4) Extracts of the head. After decapitation, heads were extracted in 100 ll pentane (Uvasol) for 24 h. Further extracts of whole females and heads were obtained from attractive young gynes after rinsing with diethyl ether before the females or heads were extracted in pentane. Prior to transporting the test solutions to the ®eld, they were concentrated and adjusted to 10 ll/female equivalent in a water bath at 40 °C. Biotest with synthetic copies of the compounds identi®ed in surface extracts To examine the possible behavior-releasing e€ect of compounds identi®ed in the surface extracts of virgin young gynes in attracting males and releasing copulatory behavior, synthetic copies of the volatiles were used in another series of tests. The biological activity of a total of 33 compounds was tested. Mixtures of synthetic copies were prepared according to the results of GC-MS analysis of surface extracts (Ayasse 1991). The relative and absolute amounts of the compounds in all test solutions were checked by GC. The total mixture contained n-alkanes, n-alkenes, saturated and unsaturated

macrocyclic lactones, ethyl esters, as well as isopentenyl esters of saturated and unsaturated fatty acids. A list of all synthetic compounds and mixtures is provided in Table 1. As control, 10 ll pentane was used (Uvasol). In the ®eld, test solutions were added onto odorless ``dummy bees.'' All tests were performed in the same manner as the tests with gland extracts or surface extracts. Chemical analysis The surface extracts of virgin young gynes and old breeding queens were analyzed on a Carlo Erba Fractovap 2450 gas chromatograph, equipped with a DB5 capillary column (30 m ´ 0.32 mm internal diamter), operating at 120 °C for 30 s, followed by programming to 280 °C at 4 °/min. For quantitative analyses, n-octacosane was added as an internal standard. Response factors were determined by single-level calibration. Structure elucidation of individual compounds was based on GCMS (VG70/250 SE instrument, Vacuum Generators, Manchester, UK, linked to a HP 5890; gas chromatographic condition as mentioned above) and comparison with mass spectra reported in the literature (Mc La€erty and Stau€er 1989), and comparison of gas chromatographic retention times (co-injection) with those of authentic reference samples. Double-bond positions in unsaturated compounds were assigned according to Buser et al. (1983) and Dunkelblum et al. (1985). Determination of the stereochemistry of double bonds was achieved by comparison of retention times using corresponding reference samples including DMDS derivatives. The erythro- and threo-aducts can be well separated by gas chromatography. Statistics Di€erences in reactions recorded in the choice experiments for unmated and mated females and odorless and untreated females were examined with a Wilcoxon signed-ranks test (Sokal and Rohlf 1995). Di€erences in numbers of inspections, pounces and copulatory events of all testing solutions in comparison to the control (pentane blank) were examined with a Mann-Whitney U-test (Sokal and Rohlf 1995). The relative and absolute amounts of single compounds in virgin gynes and nesting queens were compared by a Mann-Whitney U-test (Sokal and Rohlf 1995).

Results L. (Evlaeus) malachurum gyne odors as sexual releasers In the ®rst choice experiment, where the attractiveness of an odorless female was compared with that of an untreated one, the latter received signi®cantly more copulation attempts during the 3-min testing period (Fig. 1; P < 0.01, Wilcoxon signed-ranks test). As the attractiveness of the females was highest during the 1st minute and decreased in the 2nd and 3rd minute, some kind of learning or habituation might have taken place. The results demonstrate that attraction in L. (Evylaeus) malachurum is mediated by a female-produced sex pheromone. Attractiveness of unmated and mated young gynes Unmated and mated young gynes also di€ered in their attractiveness to the males. Unmated gynes received a signi®cantly higher number of pounces and copulatory

98 Table 1 Synthetic copies of the compounds identi®ed in the surface extracts of attractive young (Evylaeus) malachurum gynes. Between August 1989 and 1993 the following volatile mixtures were tested in the ®eld: 1 total mixture 1989, 2 total mixture 1990, 3 alkanes, 4 alkenes, 5 saturated and unsaturated lactones, 6 saturated lactones, 7 ethyl esters and isopentenyl esters of saturated fatty acids, 8 isopentenyl esters of saturated and unsaturated fatty acids, 9 isopentenyl esters of unsaturated fatty acids. Relative and absolute amounts of single compounds in the mixtures were the same as found in gas chromatographic analyses of surface extracts of virgin queens (Ayasse 1991)

Compounds

Amount (ng)

Mixtures of synthetic compounds 1

2

3

3.27 5.73 2.78 157.26 190.83 165.89 124.60 81.98 2.76 2.32 34.86 99.66 1.95 126.19

x x x x x x x x x x x x x x

x x x x x x x x x x x x x x

x x x x x x x x

Lactones 18-Octadecanolide 20-Eicosanolide 22-Docosanolide 18-Octadec-(Z)-9/-(Z)-11-enolide 20-Eicos-(Z)-11/-(Z)-13-enolide

28.55 16.44 27.71 13.82 26.78

x x x

x x x x x

Isopentenyl esters 3-Methyl-3-butenyl 3-Methyl-2-butenyl 3-Methyl-3-butenyl 3-Methyl-3-butenyl 3-Methyl-2-butenyl 3-Methyl-3-butenyl 3-Methyl-2-butenyl 3-Methyl-3-butenyl 3-Methyl-3-butenyl 3-Methyl-3-butenyl 3-Methyl-3-butenyl

48.10 5.54 16.92 23.11 1.10 66.17 38.38 4.86 12.59 2.92 0.43

x x x x x x x

7.26 5.12 4.62

x x x

Alkanes and alkenes Heptadecane Nonadecane Heneicosane Tricosane Pentacosane Heptacosane Nonacosane Triacontane (Z)-7-Heneicosene (Z)-7-Tricosene (Z)-9-Pentacosene (Z)-9-Heptacosene (Z)-7-Heptacosene (Z)-9-Nonacosene

Ethyl Ethyl Ethyl Ethyl

octadecanoate octadecanoate eicosanoate docosanoate docosanoate tetracosanoate tetracosanoate linoleate oleate linolenoate eicos-11-enoate

esters octadecanoate eicosanoate docosanoate

attempts compared to mated ones (Fig. 2, P < 0.05, Wilcoxon signed-ranks test). Males are already able to discriminate between both groups of females 3 h after mating. Corresponding to the results of the ®rst experiment, the number of male reactions decreased in the 2nd and 3rd minute after the females were o€ered to the males. Source of the female sex pheromone Biotests with di€erent samples obtained from attractive gynes showed the surface extracts (SE) to be the most attractive principle (Fig. 3, Table 2). Fewer male reactions were recorded when dummy bees were impregnated with extracts of whole females (WAE) or heads (HE). Extracts of Dufour's glands (DGE) and also those produced from heads after surface extraction (HE a. se.) elicited the lowest number of male reactions. A dummy bee impregnated solely with pentane (control) was signi®cantly less attractive than a dummy bee to which one of the volatile mixtures of attractive young gynes was

4

5

6

7

8

9

x x x x x

x x x

x x x x x x x x x x x

x x x x x x x

x x x x x x x x x x x

x x x x

x x x

x x x

x x x x x x

added. The attractiveness of odorless bees to the males may be due to visual cues. Only approaches or short contacts with the antennae of the male were observed, never copulatory attempts. The surface extracts were found to be the most attractive sample and yielded the same number of male reactions as frozen virgin gynes (Ayasse 1987; Smith and Ayasse 1987): the male reactions here were therefore set at 100%. Referring to this, more than 70% of the males were attracted to a dummy bee impregnated with Dufour's gland, head or whole-body extracts. Surface extracts, whole-animal and head extracts elicited a signi®cantly higher number of pouncing reactions than the pentane blank (Table 2). The result was almost the same when the number of copulatory attempts were considered: surface extracts applied to odorless dummy bees released signi®cantly more copulatory attempts than the control (pentane blank, Fig. 3, Table 2). Since among the other samples tested in the ®eld, only extracts from whole animals and from the cuticle surface stimulated signi®cantly more copulatory attempts than the control, males are obviously stimulated to mate

99

with a female by the compounds covering the cuticular surface. The number of pouncing events was the same in experiments with whole female extracts after extraction of the cuticular surface (WAE a. se.) and head extracts (HE); furthermore, the copulatory events recorded with whole female extracts (WAE a. se.) and Dufour's gland secretions were also the same (Fig. 3). Therefore, volatiles of the head seem to be important in attracting males from a longer distance and eliciting pouncing events, while Dufour's gland secretions might stimulate copulatory attempts. As more copulatory events were recorded using whole female extracts (WAE) than Dufour's gland secretions, other glands or glandular cells seem to be important as well. Behavior experiments with synthetic compounds Synthetic copies of the compounds identi®ed on the cuticular surface of virgin gynes were highly attractive to males (Table 3). Mating experiments showed the total mixtures and the volatile bouquets consisting of alkanes, alkenes and isopentenyl esters of unsaturated fatty acids to be the most attractive samples (Fig. 4, Table 3). In tests with all these mixtures, the number of inspecting males was signi®cantly higher than in the control. Hydrocarbons, isopentenyl esters of saturated and unsaturated fatty acids, and the total mixtures of synthetic volatiles yielded male inspections at the 80% level compared to the surface extracts. Isopentenyl esters of unsaturated fatty acids were found to be the most important compounds in stimulating pouncing events (50% compared with the surface extracts). Less stimulating but still more reactive than the control were the total mixture of volatiles tested in 1990 and isopentenyl

Fig. 1 Comparison of the mean male response to untreated and odorless L. (Evylaeus) malachurum gynes over a 3-min testing period. In a total of 16 tests, over three successive 1-min intervals, odorless queens were signi®cantly less attractive than the untreated ones (Wilcoxon signed-ranks test, P < 0.01) (vertical bars SE of the mean)

esters of saturated and unsaturated fatty acids (25% compared with the surface extracts). Among all synthetic mixtures, isopentenyl esters of unsaturated fatty acids stimulated the highest number of copulatory attempts, although only four of nine identi®ed compounds were used. All other mixtures containing isopentenyl esters of unsaturated fatty acids triggered copulatory attempts, as did a mixture of saturated and unsaturated lactones. A striking result was the lower number of reactions when other compounds were included beside the isopentenyl esters of unsaturated and saturated fatty acids. This may be caused by a potentially higher rate of error when mixing a high number of compounds. Synthetics released only 10% of the copulatory attempts induced by dummy bees impregnated with cuticular surface washings. Chemical analysis Overall, 131 compounds were identi®ed in the surface extracts [for a complete list of compounds see Ayasse (1991) and Taghizadeh (1996)]. The mixture was made of straight-chain saturated and unsaturated hydrocarbons with 11±33 carbons (alkanes, alkenes, alkadienes), methyl branched hydrocarbons, carboxylic acids, alcohols, various types of esters, saturated and unsaturated lactones, and compounds belonging to other chemical classes. Virgin and nesting gynes di€ered clearly in the relative and absolute amounts of the volatiles identi®ed on the cuticular surface (Table 4, Fig. 5). Hydrocarbons and a long-chain wax-type ester hexadecyl oleate represent the main compounds in both groups of females. In breeding queens, hydrocarbons formed the dominant group of compounds. The total amount of volatiles was

Fig. 2 Comparison of the mean male response to unmated (virgin) and mated young L. (Evylaeus) malachurum gynes. In 14 dualchoice experiments, mated queens were signi®cantly less attractive to the males than unmated ones as early as 3 h after copulation (Wilcoxon signed-ranks test, P < 0.05) (vertical bars SE of the mean)

100

Discussion Are chemical cues important in mate recognition by L. (Evylaeus) malachurum males? In L. (Evylaeus) malachurum, queens nest in dense aggregations and remain receptive for only a short time after emergence (StoÈckhert 1923). Therefore, there is a strong selective pressure for a male to recognize a virgin female before the huge number of other conspeci®c males. In fact, virgin young gynes, willing to mate, released a higher intensity of the typical search ¯ights in the males, even before they appeared at the nest entrance (M. Ayasse, unpublished observations). A typical chemotactically mediated searching behavior can be observed. Males assembled closely around nesting areas, where virgin gynes soon appeared. Some even inspected entrances of nests containing receptive gynes. Other males on the ground started to clean their antennae with the front legs, a behavior indicating sexual stimulation by the gyne odor. To minimize the time and energy required to resist male mating attempts, already mated queens may release unambiguous signals of non-receptivity (Alcock et al. 1978). Unlike related species such as L. (Dialictus) zephyrum (Barrows 1975a; Wcislo 1987), L. (Evylaeus) malachurum males were never observed pouncing on black spots, stones, other insects and worker females while they perceive the odor of a virgin female. This con®rms the relevance of pheromonal interactions. Direct evidence came from our ®eld bioassay with dead, pinned queens made odorless by solvent extraction: such dummies were widely ignored by the males, in contrast to untreated gynes, which were objects of frequent copulation attempts. In L. (Evylaeus) malachurum, olfactory cues clearly provide information for both males and female sexuals; this minimizes both the disturbance of already mated gynes and time-consuming mating attempts by males.

Fig. 3 Attractiveness of di€erent odor samples of virgin young gynes to the males. Within a testing period of 3 min, three di€erent male reactions were recorded (after Kullenberg 1973): inspection a patrolling male approached to within 5 cm of the dummy bee; pouncing a male contacted the dummy bee with its head or antennae; copulation a male mounted a dummy bee and tried to copulate. To combine the tests carried out at di€erent days and localities where the number of patrolling males was not the same, male reactions were standardized by measuring the ¯ight activity. A 1-m-long wire was mounted at the experimental sites. The number of males crossing the wire over a 3-min period was counted at least every 30 min. For each 3-min test, individual reactions of the males were divided by the mean male ¯ight activity and multiplied by 100. Since surface extracts elicited the same number of male reactions as did living virgin queens, the mean coecient of attractiveness of the three registered male reactions was set at 100% (statistics in Table 2). All investigated odor samples (for abbreviations see Table 2) were more attractive to the males than the control (pentane blank)

signi®cantly higher in virgin gynes and decreased in breeding queens (virgin: x ˆ 8.55 lg, breeding: x ˆ 3.36 lg; Mann-Whitney U-test P < 0.05). The relative amounts of hexadecyl oleate, isopentenyl esters and a steroid were higher in attractive virgin gynes, while the relative proportions of hydrocarbons and lactones were higher in nesting ones.

Table 2 Attractiveness of di€erent extracts obtained from unmated, attractive young gynes for L. (Evylaeus) malachurum males in the ®eld. (Coecient of attractiveness number of male reactions per 3 min ´ 100/¯ight activity, x mean, Sx SE of mean) Test solution (n)

Coecient of attractiveness Inspection

Surface extracts (SE) (37) Whole-animal extract (WAE) (20) WAE after surface extraction (WAE a. se.) (23) Dufour's gland extract (DGE) (22) Head extract (HE) (7) HE after surface extraction (HE a. se.) (18) Control (pentane blank) (58)

Pouncing

Copulation

x

Sx

x

Sx

x

Sx

83.33 83.11 80.90 72.82 80.78 61.17 37.59

5.51** 11.56** 10.52** 7.90** 8.38** 8.32* 3.11

20.79 9.67 5.65 2.49 5.61 5.18 1.98

5.28** 1.96** 1.88 0.61 3.48* 1.64* 0.40

2.98 1.28 0.09 0.08 0.00 0.00 0.00

1.21** 0.70** 0.09 0.08 0.00 0.00 0.00

*P < 0.05; **P < 0.001 relative to control in Mann-Whitney U-test

101 Table 3 Attractiveness of mixtures (see Table 1) of synthetic copies of the compounds identi®ed in surface extracts of attractive young L. (Evylaeus) malachurum gynes. (Coecient of attractiveness Test solution (n)

number of male reactions per 3 min ´ 100/¯ight activity, x mean, Sx SE of mean). Mixtures of isopentenyl esters of unsaturated fatty acids were found to be most attractive to the males Coecient of attractiveness Inspection

(1) Total mixture tested 1989 (22) (2) Total mixture tested 1990 (20) (3) Alkanes (23) (4) Alkenes (20) (5) Saturated and unsaturated lactones (20) (6) Saturated lactones (20) (7) Isopentenyl esters of saturated fatty acids+ethyl esters (22) (8) Isopentenyl esters of saturated and unsaturated fatty acids (19) (9) Isopentenyl esters of unsaturated fatty acids (19) (10) Control (pentane blank) (58)

Pouncing

Copulation

x

Sx

x

Sx

x

Sx

78.96 69.92 72.10 72.93 57.29 58.44 48.83 64.83 77.97 37.59

9.22** 5.11** 8.59** 13.07** 4.65** 6.03** 5.03 5.74** 8.71** 3.11

1.99 4.99 2.82 3.88 3.02 4.85 3.03 5.53 10.69 1.98

0.63 0.90** 0.79 1.66 0.93 1.50* 0.70 1.55** 2.14** 0.40

0.00 0.11 0.00 0.00 0.10 0.00 0.00 0.05 0.27 0.00

0.00 0.11 0.00 0.00 0.10 0.00 0.00 0.05 0.18* 0.00

*P < 0.05; **P < 0.001 relative to control in Mann-Whitney U-test

Fig. 4 The biological activity of mixtures of synthetic copies of surface extracts. The relative and absolute amounts of the compounds in all test solutions were prepared according to the results of GC-MS analysis of surface extracts (Table 4). The total mixture contains alkanes, alkenes, saturated and unsaturated macrocyclic lactones, ethyl esters, as well as isopentenyl esters of saturated and unsaturated fatty acids. A list of all synthetic compounds and mixtures is shown in Table 1. For a control, 10 ll of pentane blank (Uvasol, Merck) was used. In the ®eld, test solutions were added onto odorless ``dummy bees.'' All tests were performed in accordance with the tests with females and gland extracts or surface extracts. The odor mixture of isopentenyl esters of unsaturated fatty acids was most attractive for males and released copulatory attempts (statistics in Table 3)

Are chasing males capable of discerning receptive young gynes? Male competition for receptive females is evidently severe in lek situations and can be observed in many bee species (see Alcock et al. 1978; Thornhill and Alcock 1983); this should create selection for abilities to rapidly habituate to cues associated with unreceptive females. In fact, as early as 3 h after copulation took place, L. (Evylaeus) malachurum males were able to discriminate unmated and mated young gynes (Fig. 2) by odor cues. An explanation for the lower attractiveness of

mated young gynes may be an anti-aphrodisiac that males apply on a female during copulation. Other males possibly perceive this anti-aphrodisiac and avoid wasting time copulating with an already mated and therefore unreceptive female. At the same time, the marking male could increase its ®tness, since all of the diploid eggs a female produces would be fertilized by his sperm. Based on behavioral experiments in L. (Dialictus) zephrum, Kukuk (1985) proposed this so-called anti-aphrodisiac hypothesis, which was, however, not veri®ed by subsequent studies with the same species (Wicslo 1987, 1992). An alternative explanation for the lower attractiveness of young gynes after mating could be a post-copulatory change in the female's pheromonal signal, induced by mating. Several arguments support this assumption. First, our ®ndings show that in young L. (Evylaeus) malachurum gynes, the relative proportions and absolute amounts of volatiles of extracts of the Dufour's gland, of the head and also of surface extracts changed after mating (Ayasse 1991; Ayasse et al. 1991, 1993). Furthermore, the total amount of volatiles on the cuticular surface decreases. Therefore, we can exclude an anti-aphrodisiac on the cuticular surface of recently mated young gynes. For instance, here we present evidence that isopentenyl esters, which were found to be the most attractive compounds of virgin gynes, almost disappeared within a few hours after mating (Fig. 5). Differences in volatile constituents of virgin and mated females have been shown in several bee species by behavior experiments (Wcislo 1992; Engels et al. 1993; Dutzler and Ayasse 1996) and by chemical analyses (Engels and Engels 1988; Ayasse et al. 1990a,b ; Engels et al. 1993, 1997). A mating-induced termination of female sex pheromone production has been shown in cockroaches (Schal et al. 1997), moth (Raina 1997), lepidoptera (McNeil et al. 1997), and in other insects. In Heliothis zea, the male accessory glands have been implicated as the source of a peptide that terminates pheromone production in mated females. This peptide is transferred to the female at the time of mating (Raina

102 Table 4 Relative amount (%) of volatile compounds identi®ed in surface extracts of virgin (n ˆ 7) and nesting (n ˆ 12) L. (Evylaeus) malachurum queens (x mean, Sx SE of mean)

Compound

Relative amount (%) Unmated x

Nesting Sx

x

Sx

Alkanes Heptadecane Nonadecane Heneicosane Tricosane Pentacosane Heptacosane Nonacosane Triacontane

0.10 0.06 0.13 6.46 6.79 6.47 5.24 3.42

0.06 0.02 0.03 1.51 1.29 1.09 0.94 1.02

0.00* 0.01* 0.37* 19.87** 4.74* 15.16* 12.11* 1.46

0.00 0.01 0.11 1.55 0.50 1.37 1.60 0.46

Alkenes (Z)-7-Heneicosene (Z)-9-Tricosene (Z)-7-Tricosene (Z)-9-Pentacosene (Z)-7-Pentacosene (Z)-9-Heptacosene (Z)-7-Heptacosene (Z)-9-Nonacosene

0.12 0.00 0.13 1.50 0.00 3.31 0.10 6.07

0.08 0.00 0.09 0.85 0.00 0.84 0.07 1.69

0.00 0.15* 0.36* 4.82* 0.25** 5.65* 0.96** 6.56

0.00 0.03 0.07 1.08 0.06 0.59 0.08 0.61

Lactones 18-Octadecanolide 20-Eicosanolide 22-Docosanolide 24-Tetracosanolide 18-Octadec-(Z)-9/(Z)-11-enolidea 20-Eicos-(Z)-11/(Z)-13-enolidea 22-Docos-(Z)-13/(Z)-15-enolidea 24-Tetracos-(Z)-15/(Z)-17-enolidea

1.36 0.86 0.87 0.00 0.61 1.22 1.56 0.07

0.66 0.33 0.46 0.00 0.17 0.28 0.43 0.04

1.58 5.21* 3.75** 0.35* 0.27 1.92 1.82 0.08

0.49 1.90 1.07 0.13 0.11 0.49 0.33 0.03

Isopentenyl esters 3-Methyl-3-butenyl 3-Methyl-2-butenyl 3-Methyl-3-butenyl 3-Methyl-3-butenyl 3-Methyl-2-butenyl 3-Methyl-3-butenyl 3-Methyl-2-butenyl 3-Methyl-3-butenyl 3-Methyl-3-butenyl 3-Methyl-3-butenyl 3-Methyl-3-butenyl 3-Methyl-3-butenyl 3-Methyl-3-butenyl

2.29 0.23 0.78 1.11 0.05 2.39 1.38 0.24 0.63 0.19 0.16 0.02 0.12

1.22 0.11 0.27 0.34 0.05 0.67 0.49 0.12 0.42 0.10 0.09 0.02 0.07

0.09* 0.02 0.09** 0.03** 0.05 0.00** 0.20* 0.07 0.01* 0.02 0.06 0.00 0.00

0.05 0.01 0.05 0.02 0.03 0.00 0.06 0.03 0.01 0.02 0.02 0.00 0.00

esters octadecanoate eicosanoate docosanoate oleate linoleate eicosenoateb docosenoateb

0.42 0.26 0.18 0.35 1.16 0.17 0.54

0.11 0.08 0.05 0.08 0.22 0.09 0.11

0.28 0.21 0.40 0.01** 0.03** 0.54* 0.46

0.04 0.02 0.04 0.01 0.01 0.10 0.05

Carboxylic acids Hexadecanoic acid Hexadecenoic acidb 9-Octadecenoic acid Eicosenoic acidb Docosenoic acidb Tetracosanoic acid

0.50 0.00 2.87 0.42 0.18 0.53

0.50 0.00 1.09 0.23 0.08 0.09

0.38 0.10 0.01* 0.72 2.55** 1.58**

0.12 0.07 0.01 0.20 0.34 0.27

Other compounds Unknown (M=326) Unknown (M=354) Steroid Hexadecyl oleate

0.00 0.00 9.33 27.43

0.00 0.00 2.64 11.39

2.55 2.31* 2.55 0.39**

0.79 0.84 0.34 0.17

Ethyl Ethyl Ethyl Ethyl Ethyl Ethyl Ethyl Ethyl

octadecanoate octadecanoate eicosanoate docosanoate docosanoate tetracosanoate tetracosanoate linoleate oleate linolenoate eicosadienoateb eicosenoatec eicosatrienoateb

*P < 0.05; **P < 0.001 in Mann-Whitney U-test Unsaturated lactones are mixtures of two positional isomers b Position of the double bond has not been determined c Mixture of two positional isomers (11, 13) a

103

Fig. 5 Comparison of the relative amounts of n-alkanes (1), n-alkenes (2), a wax-type ester (hexadecyl oleate, 3), saturated macrocyclic lactones (4), unsaturated macrocyclic lactones (5), isopentenyl esters of saturated (6) and unsaturated (7) fatty acids, ethyl esters saturated (8) and unsaturated (9), fatty acids (10) and a steroid compound (11) in surface extracts of young virgin and old nesting L. (Evylaeus) malachurum queens (Mann-Whitney U-test: *P < 0.05; **P < 0.001)

1989). Whether a similar mechanism is responsible for the post-copulatory change of the pheromonal signal in L. (Evylaeus) malachurum remains open. Where is the female sex pheromone located and where is it produced? In halictid bees, the Dufour's gland takes up a large part of the abdomen. Because of its large size, many investigations on pheromonal communication have focused exclusively on the contents of this gland (Dueld et al. 1984). In L. (Dialictus) zephrum, Smith et al. (1985) found that macrocyclic lactones of the Dufour's gland are used as recognition signals and sex pheromones. However, their experiments failed to compare the attractiveness of volatiles from the Dufour's gland and other possible sources like the head and the cuticle surface. Furthermore, only short contacts of the males and no attempts to copulate were recorded. In our experiments, extracts of the Dufour's gland and the head were less attractive than the volatiles on the cuticular surface, which released the highest intensity of male mating behavior (Fig. 3). In L. (Dialictus) zephyrum, compounds on the cuticular surface were also important in attracting males (Barrows 1975b; Wisclo 1987). Since dummies impregnated with cuticular extracts of virgin gynes released the same number of male reactions as did frozen receptive gynes (Ayasse 1987) and, on the other hand, the attractiveness of dead and active Lasioglossum gynes are similar (Barrows 1975b; Smith 1983), the female sex pheromone in sweat bees is

evidently located on the cuticular surface, and probably distributed all over the body. The production site of the sex pheromone in L. (Evylaeus) malachurum gynes remains unknown. Secretions from cephalic glands and/or the Dufour's gland may be involved. Another possible source of the `active principle' on the cuticle could be glandular cells of the epidermis below tergites 3±5 or from the sternite 6 described by Altenkirch (1962) in several species of the genus Halictus. In Eucera palestinae, secretions of tergal glands released mating behavior in the males (Shimron and Hefetz 1985). The volatile bouquet on the cuticle could be a mixture of compounds produced in several glands. Our ®ndings that all test solutions released signi®cantly more male reactions than the control support this hypothesis. The speci®city of the volatile bouquet on the cuticle of attractive young gynes may be a result of combining the secretions from di€erent glands, and a quick change of the bouquet after mating might be e€ected by ceasing production or terminating secretion of compounds. Which cuticular volatiles of a gyne attract males most e€ectively? The present behavior experiments with synthetic cuticular volatiles demonstrated a contribution of isopentenyl esters to the female sex pheromone in L. (Evylaeus) malachurum. When a test mixture contained such compounds, the males not only inspected dummy bees but also tried to copulate. Although isopentenyl esters have also been found in other Halictinae (Dueld et al. 1981, 1984; Smith et al. 1985; Hefetz et al. 1986) and in Nomiinae (Dueld et al. 1982), their potential role in chemocommunication remained unknown. In L. (Evylaeus) malachurum, the prevalence of isopentenyl esters is a caste-speci®c trait of gynes, and the Dufour's gland bouquet di€ers in young gynes and young workers (Ayasse et al. 1993). A more complete copy of the virgin gyne pheromone containing more of the identi®ed compounds might result in male reaction intensities similar to those released by a live bee, because the behavioral reactions released by a complete mixture of compounds is often higher than the sum of the behavioral events triggered by single components (Silverstein 1984). Synergetically active compounds may be inactive when o€ered individually. Aliphatic hydrocarbons were most abundant in the surface extracts of virgin gynes and breeding queens. Cuticular hydrocarbons have been shown to be involved in nest and nestmate recognition in termites (Howard et al. 1982; Bagneres et al. 1990), ants (BonavitaCourgourdon et al. 1987; Soroker et al. 1995), wasps (Lorenzi et al. 1997; Singer and Espelie 1997) and honey bees (Getz and Page 1991; Page et al. 1991; Breed and Stiller 1992), all highly eusocial insect groups. In the primitively eusocial species L. (Evylaeus) malachurum we

104

found a similar function. The hydrocarbon pattern was characteristic in females belonging to di€erent castes or functional groups (Ayasse et al. 1991). Hydrocarbons were also involved in the colony recognition cues used by homing bees to identify their nest entrance (Ayasse 1990, 1991) and were used by kleptoparasitic bees to recognize host nests (Sick et al. 1994). Until now hydrocarbons could not be identi®ed as sex pheromone components in bees (Dueld et al. 1984; Vander Meer et al. 1998). In L. (Evylaeus) malachurum, n-alkanes and n-alkenes attracted males to the odor source, and synthetic mixtures of hydrocarbons released 80% of male reactions compared with the complete volatile bouquet on the cuticle. However, copulatory attempts were released by other compounds like isopentenyl esters of unsaturated fatty acids and unsaturated macrocyclic lactones, which were identi®ed on the cuticle of attractive young gynes in small amounts only. Therefore, the main role of hydrocarbons might be to provide information on conspeci®c identity as well as on e.g., nestmates, individual age, and caste, as was shown in several social insect species. Although wax-type esters have also been identi®ed in bumble bees (Ayasse et al. 1995; Hefetz et al. 1996), honey bees (Katzav-Gozansky et al. 1997), and solitary bees (Sick et al. 1994), almost no communicative function has been described. Hexadecyl oleate was the main compound identi®ed on the cuticular surface of virgin gynes and also in males (Ayasse 1991, 1994). Interestingly, in ®eld experiments, a synthetic sample of this compound was not attractive to males (M. Ayasse, unpublished data). When cuticular extracts of an attractive gyne were o€ered simultaneously with hexadecyl oleate in the same concentrations as found in males, it even inhibited copulatory attempts (Ayasse 1994). Headspace samples of mating bees showed this compound to be released by the bees during copulation (Ayasse 1994).Therefore hexadecyl oleate seems to have a communicative function in signaling to conspeci®c males the mating status of a female during copulation. Macrocyclic lactones are abundant in the Dufour's glands of halictine bees (Cane 1981; Johansson et al. 1982; Smith et al. 1985; Hefetz 1987, 1990; Hefetz and Graur 1988). In L. (Evylaeus) malachurum, Ayasse et al. (1993) showed that these compounds are not produced at high rates throughout life. Instead, lactone production is clearly correlated with the use of macrocyclic lactones to line the walls of the brood cells and the nest entrance tube. The age-speci®c patterns of volatile compounds in queens revealed that macrocyclic lactones were the main components only in breeding queens with large ovaries and in workers (Ayasse et al. 1990a, 1993). A study of con¯ict over nest ownership among L. (Evylaeus) malachurum queens during the solitary phase of the nest cycle demonstrated that lactones in addition to lining the walls of brood cells can signal size (Smith and Weller 1989). Whether lactones are used in dominance interactions of nestmates is presently under investigation (M. Ayasse, unpublished data).

Composition and modulation of the female sex pheromone in L. (Evylaeus) malachurum: how mating expenditures are economized According to our experimental data, the female sex pheromone in L. (Evylaeus) malachurum is a complex mixture of chemical compounds signaling di€erent messages over long-range, short-range, and touch-range distances. The long-range message is probably the label of a conspeci®c queen, di€ering from the worker odor (Ayasse 1991; Ayasse et al. 1993). The short- and touchrange message, which releases pouncing and copulation, is evidently deposited on the abdominal cuticle and may be produced by several glands. Isopentenyl esters are the main constituents of this ®nal mating signal. Shortly after copulation (3 h or even less), the inseminated young gyne drastically reduces the emission of the isopentenyl esters in particular; this modulation of female sex pheromone composition enables a young gyne to de®nitively inform the males about her receptive status. Mating strategies always evolved in directions which increased ®tness (Emlen and Oring 1977). The speci®c lek situation in L. (Evylaeus) malachurum requires an e€ective economization of the mating tactics in both sexes. Competition among the males should create selection that proximately rewards males capable of precisely recognizing promising mates, and should ultimately lead to a mating behavior where males spend no time unsuccessfully chasing already mated queens. Selective pressure on the female side ultimately operates in the direction of a pheromonal gestalt which, except during a rather short period of receptiveness, always and perfectly repels the mate-searching males; the proximate result is an undisturbed female after mating. In fact, mated L. (Evylaeus) malachurum gynes are recognized but not chased by the males. Therefore, the females do not lose the time and energy they require to provide themselves with pollen and nectar necessary to survive the winter diapause, which lasts for up to 7 months (Knerer 1992). Acknowledgements The authors wish to thank Jan TengoÈ and Hannes Paulus for valuable comments on the manuscript. This work was supported by the Deutsche Forschungsgemeinschaft (En 89/11, Fr 507/8) and the FWF Austria (P09773-BIO).

References Alcock J (1978) Notes on male mate-locating behavior in some bees and wasps of Arizona. Pan-Pac Entomol 54:215±225 Alcock J, Barrows EM, Gord G, Hubbard LJ, Kirkendall L, Pyle DW, Ponder TL, Zalom FG (1978) The ecology and evolution of male reproductive behaviour in the bees and wasps. J Linn Soc Lond Zool 64:293±326 Altenkirch G (1962) Untersuchungen uÈber die Morphologie der abdominalen HautdruÈsen einheimischer Apiden (Insecta, Hymenoptera). Zool Beitr 7:161±238 Ayasse M (1987) Wie beein¯ussen Weibchen-Pheromone die Partnerwahl bei maÈnnlichen Furchenbienen? Paarungs-

105 strategien bei Lasioglossum malachurum (Hymenoptera: Halictidae). Diplomarbeit, TuÈbingen University Ayasse M (1990) Odor based interindividual and nest recognition in the sweat bee Lasioglossum malachurum (Hymenoptera: Halictidae). In: Veeresh GK, Mallik B, Viraktamath CA (eds): Social insects and the environment. Oxford & IBH, New Delhi, pp 511±512 Ayasse M (1991) Chemische Kommunikation bei der primitiv eusozialen Furchenbiene Lasioglossum malachurum (Hymenoptera: Halictidae): Ontogenese kastenspezi®scher Duftsto€bouquets, Paarungsbiologie und Nesterkennung. PhD thesis, TuÈbingen University Ayasse M (1994) The meaning of male odor in the mating biology of sweat bees (Hymenoptera: Halictidae). In: Lenoir A, Arnold G, Lepage M (eds): Les insectes sociaux. Publications Universite Paris Nord, Paris Ayasse M, Engels W, Hefetz A, LuÈbke G, Francke W (1990a) Ontogenetic patterns in amounts and proportions of Dufour's gland volatile secretions in virgin gynes and nesting queens of Lasioglossum malachurum (Hymenoptera: Halictidae). Z Naturforsch 45c:709±714 Ayasse M, Leys R, Pamilo P, TengoÈ J (1990b) Kinship in communally nesting Andrena (Hymenoptera: Andrenidae) bees is indicated by composition of Dufour's gland secretions. Biochem Syst Ecol 18:453±460 Ayasse M, LuÈbke G, Francke W (1991) Alters- und kastenspezi®sche Duftsto€muster bei der Fortp¯anzung der Furchenbiene Lasioglossum malachurum (Hymenoptera: Halictidae). Verh Dtsch Zool Ges 84:386±387 Ayasse M, Engels W, Hefetz A, TengoÈ J, LuÈbke G, Francke W (1993) Ontogenetic patterns of volatiles identi®ed in Dufour's gland extracts from queens and workers of the primitively eusocial halictine bee, Lasioglossum malachurum (Hymenoptera: Halictidae). Insect Soc 40:1±18 Ayasse M, Marlovits TC, TengoÈ J, Taghizadeh T, Francke W (1995) Are there pheromonal dominance signals in the bumblebee Bombus hypnorum (Apidae). Apidologie 26:163±180 Bagneres GA, Clement JL, Blum MS, Severson RF, Joulie C, Lange C (1990) Cuticular hydrocarbons and defensive compounds of Reticulitermes ¯avipes (Kollar) and R. santonenis (Feytaud): polymorphism and chemotaxonomy. J Chem Ecol 16:3213±3244 Barrows EM (1975a) Individually distinctive odors in an invertebrate. Behav Biol 15:57±64 Barrows EM (1975b) Mating behavior in halictine bees (Hymenoptera: Halictidae). III. Copulatory behavior and olfactory communication. Insect Soc 3:307±332 Barrows EM, Bell WJ, Michener CD (1975) Individual odor differences and their social functions in insects. Proc Natl Acad Sci USA 72:2824±2828 Bonavita-Cougourdan A, Clement JL, Lange C (1987) Nestmate recognition: the role of cuticular hydrocarbons in the ant Camponotus vagus Scop. J Entomol Sci 22:1±10 Breed MD, Stiller TM (1992) Honey bee, Apis mellifera, nestmate discrimination: hydrocarbon e€ects and the evolutionary implications of comb choice. Anim Behav 43:875±883 Buser H-R, Arn H, Guerin P, Rauscher S (1983) Determination of double bond position in mono-unsaturated acetates by mass spectrometry of dimethyl disul®de adducts. Anal Chem 55:818± 822 Cane JH (1981) Dufour's gland secretion in the cell linings of bees (Hymenoptera: Apoidea). J Chem Ecol 7:403±410 Dueld RM, Fernandes A, Lamb C, Wheeler JW, Eickwort GC (1981) Macrocyclic lactones and isopentenyl esters in the Dufour's gland secretion of halictine bees. J Chem Ecol 7:319±331 Dueld RM, LaBerge WE, Cane JH, Wheeler JW (1982) Exocrine secretions of bees. IV. Macrocyclic lactones and isopentenyl esters in the Dufour's gland secretions of Nomia bees (Hymenoptera: Halictidae). J Chem Ecol 8:534±543 Dueld RM, Wheeler JW, Eickwort GC (1984) Sociochemicals of bees. In: Bell WJ, Carde RT (eds): Chemical ecology of insects. Chapman & Hall, London, pp 387±428

Dunkelblum E, Tan SH, Silk PJ (1985) Double-bond location in monounsaturated fatty acids by dimethyl disul®de. Derivatization and mass spectrometry: application to analysis of fatty acids in pheromone glands of four Lepidoptera. J Chem Ecol 11:265±277 Dutzler G, Ayasse M (1996) The function of female sex pheromones in mate selection of Osmia rufa (Megachilidae) bees. Proc 20th Int Congr Entomol, Firenze, p 376 Eickwort GC, Ginsberg HS (1980) Foraging and mating behavior in Apoidea. Annu Rev Entomol 25:421±446 Emlen ST, Oring LW (1977) Ecology, sexual selection, and the evolution of mating systems. Science 197:215±223 Engels W (1988) Fortp¯anzungsstrategien bei Bienen. Verh Dtsch Zool Ges 81:155±167 Engels E, Engels W (1988) Age-dependent queen attractiveness for drones and mating in the stingless bee Scaptotrigona postica. J Apicult Res 27:3±8 Engels E, Engels W, LuÈbke G, SchroÈder W, Francke W (1993) Agerelated patterns of volatile cephalic consituents in queens of the neotropical stingless bee Scaptotrigona postica Latr (Hymenoptera, Apidae). Apidologie 24:539±548 Engels W, Engels E, Francke W (1997) Ontogeny of cephalic volatile patterns in queens and mating biology of the neotropical stingless bee, Scaptotrigona postica. Invert Reprod Dev 30:251±256 Getz WM, Page RE Jr (1991) Chemosensory kin-communication systems and kin recognition in honey bees. Ethology 87:298± 315 Hefetz A (1987) The role of Dufour's gland secretions in bees. Physiol Entomol 12:243±253 Hefetz A (1990) Individual badges and speci®c messages in multicomponent pheromones of bees (Hymenoptera: Apidae). Entomol Gen 15:103±113 Hefetz A, Graur D (1988) The signi®cance of multicomponent pheromones in denoting speci®c compositions. Biochem Syst Ecol 16:557±566 Hefetz A, BergstroÈm G, TengoÈ J (1986) Species, individual and kin speci®c blends in Dufour's gland secretions of halictine bees ± chemical evidence. J Chem Ecol 12:197±208 Hefetz A, Thagizadeh T, Francke W (1996) The exocrinology of the queen bumble bee Bombus terrestris (Hymneoptera: Apidae, Bombini). Z Naturforsch 51c:409±422 Howard RW, McDaniel CA, Nelson DR, Blomquist GJ, Gelbaum LT, Zalkow LH (1982) Cuticular hydrocarbons of Reticulitermes virginicus (Banks) and their role as potential speciesand caste-recognition cues. J Chem Ecol 8:1227±1239 Johansson I, Svensson BG, TengoÈ J, BergstroÈm G (1982) Systematic relationship of halictine bees based on the pattern of macrocyclic lactones in the Dufour's gland secretion. Insect Biochem 12:161±170 Katzav-Gozansky T, Soroker V, Hefetz A (1997) Plasticity of caste-speci®c Dufour's gland secretion in the honey bee (Apis mellifera L.). Naturwissenschaften 84:238±241 Knerer G (1992) The biology and social behaviour of Evylaeus malachurus (K.) (Hymenoptera; Halictidae) in di€erent climatic regions of Europe. Zool Jahrb Syst 119:261±290 Kukuk PF (1985) Evidence for an antiaphrodisiac in the sweat bee Lasioglossum (Dialictus) zephyrum. Science 227:656±657 Kullenberg B (1973) Field experiments with chemical sexual attractants on aculeate hymenoptera males II. Zoon Suppl 1:31±44 Lorenzi MC, Bagneres AG, Clement JL, Turrlazzi S (1997) Polistes biglumis bimaculatus epicuticular hydrocarbons and nestmate recognition (Hymenoptera, Vespidae) Insect Soc 44:123±138 Mc La€erty FW, Stau€er DB (1989) The Wiley NBS registry of mass spectral data, Whiley Interscience, New York McNeil JN, Delisle J, Cusson M (1997) Regulation of pheromone production in Lepidoptera: the need for an ecological perspective. In: Carde RT, Minks AK (eds): Insect pheromone research ± new directions. Chapman & Hall, New York, pp 31±41 Michener CD (1990) Reproduction and castes in social halictine bees. In: Engels W (ed): Social insects. Springer, Berlin Heidelberg New York, pp 77±121

106 Page RE Jr, Metcalf RA, Metcalf RL, Erickson EHJ, Lampman RL (1991) Extractable hydrocarbons and kin recognition in honeybee (Apis mellifera L.). J Chem Ecol 17:745±756 Raina AK (1989) Male-induced termination of sex pheromone production and receptivity in mated females of Heliothis zea. J Insect Physiol 35: 821±826 Raina AK (1997) Control of pheromone production in moth. In: Carde RT, Minks AK (eds): Insect pheromone research ± new directions. Chapman & Hall, New York, pp 21±30 Sakagami SF (1974) Sozialstruktur und Polymorphismus bei Furchen- und Schmalbienen. In: Schmidt GH (eds): Sozialpolymorphismus bei Insekten. Wissenschaftliche Verlagsgesellschaft, Stuttgart, pp 257±293 Schal C, Liang D, Blomquist GL (1997) Neural and endocrine control of pheromone production and release in cockroaches. In: Carde RT, Minks AK (eds): Insect pheromone research ± new directions. Chapman & Hall, New York, pp 3±20 Shimron O, Hefetz A (1985) Mating behavior and sex attraction of Eucera palestinae Friese (Hymenoptera: Anthophoridae). J Kans Entomol Soc 58:526±531 Sick M, Ayasse M, TengoÈ J, Engels W, LuÈbke G, Francke W (1994) Host-parasite relationship in six species of Sphecodes bees and their halictid hosts: nest intrusion, intranidal behavior, and Dufour's gland volatiles (Hymenoptera: Halictidae). Insect Behav 7:101±117 Silverstein RM (1984) Chemistry of insect communication. In: Lewis T (eds): Insect communication. Academic Press, London, pp 105±121 Singer TL, Espelie KE (1997) Exposure to nest paper hydrocarbons is important for nest recognition by a social wasp, Polistes metricus Say (Hymenoptera, Vespidae), Insect Soc 44:189±302 Smith BH (1983) Recognition of female kin by male bees through olfactory signals. Proc Natl Acad Sci USA 80:4551±4553

Smith BH, Ayasse M (1987) Kin-based male mating preferences in two species of halictine bees. Behav Ecol Sociobiol 20:313± 318 Smith BH Weller C (1989) Social competition among gynes in halictine bees: the in¯uence of bee size and pheromones on behavior. J Insect Behav 3:397±411 Smith BH, Carlson RG, Frazier J (1985) Identi®cation and bioassay of macrocyclic lactone sex pheromones of halictine bee Lasioglossum zephyrum. J Chem Ecol 11:1447±1456 Sokal RR, Rohlf F (1995) Biometry, 3rd edn. Freeman, New York Soroker V, Vienne C, Hefetz A (1995) Hydrocarbon dynamics within and between nestmates in Cataglyphis niger (Hymenoptera: Formicidae). J Chem Ecol 21:365±378 StoÈckhert E (1923) UÈber Entwicklung und Lebensweise der Bienengattung Halictus Latr. und ihrer Schmarotzer. Konowia 2:48±64, 164±165, 216±247 Thagizadeh T (1996) Untersuchungen ¯uÈchtiger Inhaltssto€e von Hymenopteren. PhD thesis, Hamburg University Thornhill R, Alcock J (1983) The evolution of insect mating systems. Harvard University Press, Cambridge, Mass Vander Meer RK, Breed MD, Espelie KE, Winston ML (1998) Pheromone communication in social insects: ants, wasps, bees, and termites. Westview, Boulder, Colo Wcislo WT (1987) The role of learning in the mating biology of a sweat bee Lasioglossum zephyrum (Hymenoptera: Halictidae). Behav Ecol Sociobiol 20:179±185 Wcislo WT (1992) Attraction and learning in mate-®nding by solitary bees, Lasioglossum (Dialictus) ®guerisi Wcislo and Nomia triangulifera Vachal (Hymenoptera: Halictidae) Behav Ecol Sociobiol 31:139±148 Westrich P (1989) Die Wildbienen Baden-WuÈrttembergs I. Ulmer, Stuttgart

Communicated by R.F.A. Moritz