On optimal nectar foraging by some tropical (: ) Dw Roubik, D Yanega, M Aluja S, Sl Buchmann, Dw Inouye

To cite this version:

Dw Roubik, D Yanega, M Aluja S, Sl Buchmann, Dw Inouye. On optimal nectar foraging by some tropical bees (Hymenoptera: Apidae). Apidologie, Springer Verlag, 1995, 26 (3), pp.197-211. ￿hal- 00891259￿

HAL Id: hal-00891259 https://hal.archives-ouvertes.fr/hal-00891259 Submitted on 1 Jan 1995

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non, lished or not. The documents may come from émanant des établissements d’enseignement et de teaching and research institutions in France or recherche français ou étrangers, des laboratoires abroad, or from public or private research centers. publics ou privés. Original article

On optimal nectar foraging by some tropical bees (Hymenoptera: Apidae)

DW Roubik D Yanega M Aluja S 1 SL Buchmann DW Inouye

1 Smithsonian Tropical Research Institute, Balboa, Panama; 2 Illinois Natural History Survey, 607, E Peabody Drive, Champaign, IL 61820; 3 Carl Hayden Research Center, 2000, East Allen Road, Tucson, AZ 85719; 4 Department of Zoology, University of Maryland, College Park, MD 20742-4415, USA

(Received 8 November 1994; accepted 27 February 1995)

Summary — Sugar concentration in nectar foraged by 13 Euglossini, 16 Meliponini and 8 Centridini (Apidae) was monitored in floristically rich habitats. For 6 Meliponini, sugar solution profitability by imbibement rate was compared to nectar choice (figs 1-4, tables I-IV). In addition, foraging assays tested meliponine response to 10 amino acid solutions (table V, fig 5). Optimal nectars were 35-65% sugar among bees, and apparently their modal forage. Bees also accepted suboptimal nectar, or were ’rate maximizers’. Two species had modes well below the means, indicating facultative selection of dilute nectars. For pollination, heterogeneity and optimal nectar sweetness are likely key mechanisms caus- ing forager fidelity. Mean sugar contents were 38, 44 and 48 for Euglossini, Meliponini and Centri- dini, respectively; modes averaged 3-4% higher. All bees used nectars of 30-45% sugar; some also used nectars of 10-15 or 65-70% sugar; others did not use nectar of over 60% sugar. Amino acid solu- tions of 35-80 mM concentration did not generally affect attractiveness, but Melipona avoided glu- tamic acid, glycine, serine, alanine and proline, responding comparably to 20-40% sugar solutions con- trasted with 50% concentration. nectar composition / optimal foraging / pollination

INTRODUCTION Buchmann, 1984). Considering the wide range in nectar concentration, total sugar flower, and nectar composition among An intringuing question was posed by Eick- per we know little about cer- wort and Ginsberg (1980): is energy alone angiosperms, why tain flowers are chosen an appropriate measure of fitness for for- by foraging bees, aging bees? Bees often do not receive opti- or even what resources are normally avail- mal sugar reward from flowers (Roubik and able. Bees provide a means of sampling

* Mailing address: Unit 0948, APO AA 34002-0948, USA nectar on a community-wide basis. Because and Inouye, 1993; Endress, 1994). Here we nectar recently stored within the bee’s crop present data on nectar selection by 37 bee is not dehydrated, a sample taken from a species of the long-tongued family Apidae forager shows mean sugar concentration in (Roig-Alsina and Michener, 1993). These flowers. Whereas nectars used by varied include Euglossini, whose tongue lengths classes of pollinators, such as birds, bats reach 4.5 cm (D Roubik, unpublished data), and , seem to show differences in a variety of Meliponini (see also Roubik and concentration, divisions between ’large and Buchmann, 1984; Roubik et al, 1986) and small’ or ’long-tongued and short-tongued’ Centris (Centridini). The bees ranged from bees have rareely been assessed (Baker, permanently social to solitary, and from very 1978; Baker et al, 1983). Larger bees evi- small to very large (ca 8-1 000 mg, table dently use high-sucrose nectar, while I). We measured profitability directly for sev- smaller bees take nectars of all kinds eral meliponines by assaying caloric intake (sucrose-poor, balanced fructose/glu- rate at sucrose solutions, and for certain cose/sucrose solutions, sucrose-rich, and euglossines (Kato et al, 1992) or sucrose-dominant; Southwick et al, 1981). meliponines by choice studies with artificial However, comparative data on sugar con- feeders. We also made tests of forager centration preference are needed for both response to 10 nectar amino acids, using solitary and social species. The economy sucrose solutions (Inouye and Waller, 1984). of daily metabolic demands may have a greater effect on solitary, non-food-storing species because they cannot exchange MATERIALS AND METHODS dilute nectar for more profitable ’fuel’ at the home nor do make from base, they honey Following the methods of Roubik and Buchmann studies of sev- watery nectar. Nonetheless, (1984); we studied nectar harvest through the eral temperate and tropical Apis, and 4 day by capturing bees and extracting their crop species of Melipona show these social for- contents from mouthparts, using either 20 or 50 μl agers profit most from nectar having calibrated microcapillary tubes, while pressing bees Less than 1% of 45-60% or slightly higher sugar (Roubik dorsoventrally. Meliponini and none of the other bees carried of less and Buchmann, 1984; Roubik, 1994). Sim- liquid than 5% TDS (total dissolved solids), which may Harder discussed influence of ilarly, (1986) have been water. These samples were not proboscis and body size, lapping rate, and included. Euglossine males, which readily fly to corolla tube dimensions for temperate-zone chemical attractants (Roubik and Ackerman, Bombus, making a theoretical prediction 1987), were used in place of their relatively rare that the bees prefer nectar of 50-65% sugar. females. Meliponine foragers were sampled as returned to their nests. female Are significant nectar feeding constraints they Stinging bees, Centris, were killed quickly in cyanide collecting or size, and have imposed by tongue body jars, after which nectar remained in the crop but to flowering plants responded evolutionarily could be expelled. The Centris were collected at nectar quality sought by different bees? Malpighiaceae, and thus their nectar came from Tropical habitats, whose flora show the other flowers. Sugar content in solutions was range in sugar concentration comparable determined with an Atago hand-held refractometer Co, corrected for to temperate areas, are ideal for study of (National Japan) temperature. Our study included nectar that may have been such questions. Not only are many kinds of rich in amino acids, thus sugar content in nectar lowland condi- flowers and bees present, would be lower than that recorded on the refracto- tions remove constraints imposed on smaller meter, but likely by no more than a few percent or non-thermoregulating bees by cold (Inouye et al, 1980). weather, wind and cloud cover (Eickwort Study sites were primarily forested or rela- and Ginsberg, 1980; Roubik, 1989; Kearns tively undisturbed lowlands. Nectar collections

were made both during the wet and dry seasons followed Inouye and Waller (1984). Solutions of at the Panama sites, and during the wet season 35-80 mM amino acid concentrations were pre- at all other sites. Studies took place from 1976 pared for the 25 or 50% solutions, respectively, of to 1992, as detailed below. Meliponines were valine, isoleucine, leucine, glutamic acid, threo- studied at Barro Colorado Island, Parque Nacional nine, alanine, glycine, serine, arginine and proline. Soberania, Pipeline Road, and Curundu These are not necessarily essential amino acids (Panama); San Bernadino, Quintana Roo (Mex- for bees but all are found in floral nectar (Baker et ico), and near Kourou (French Guyana). Sites al, 1983). Solutions were presented on randomly used for Centris were northeastern Puerto Rico placed artificial feeders in a grid of 6. One control (coastal habitat), and the Centro Ecológico of and 5 sucrose/amino acid solutions were on sat- Hermosillo (Mexico). Euglossines were studied urated yellow or light blue sponges, 4 mm thick, at Parque Nacional Soberania, Nusagandi, Cerro 6 cm square, within whitish plastic dishes. Grids Campana, Montaña de Chepo, Herrera Province, were rotated 180° each 5 min to prevent position and Howard Air Base (Panama); Parque Nacional effects. The number of bees on each feeder was Braulio Carrillo (Costa Rica); Parque Nacional counted each 5 min during 1 h, after bees had Amacayacu (Colombia); Degrad Saramaca, been trained to 50% sugar solutions on the pre- Kourou (French Guyana) and Felipe Carrillo vious day. Sponges and feeders were washed Puerto, Quintana Roo (Mexico). Ambient tem- in boiling water after each replicate, and no peratures during the day were 27-34°C at lowland sponge was used more than 4 times. The grid sites, while at Montaña de Chepo and Cerro Cam- was also used with only sucrose solutions of 10, pana (Panama) temperatures were approximately 20, 25, 30, 40 and 50%, to compare directly the 7°C cooler. response to sugar with response to amino acids Assays of imbibement rate at sugar solutions in the experimental setting. From 1 to 4 colonies also followed Roubik and Buchmann (1984), with of each of the 8 species was used, either on Barro some modifications. Artificial nectar droplets were Colorado Island or at Curundu, Panama. The presented to bees that climbed or alighted directly studies were made in 1981. Replication was 1-9 by slight depressions in either paraffin or plastic times for the pure sucrose solutions, and 3-26 discs of gray or whitish color. Discs held either times for the sucrose/amino acid solutions (see fig 20 or 10 μl droplets, used for bees of different 5 below). sizes. For Scaptotrigona, Trigona and Taxonomic studies resulted in name changes Cephalotrigona, we measured the time to com- and some new data are incorporated here (D pletely consume a droplet of 10 μl for 15 individ- Roubik, unpublished results, see also Roubik, ual bees at 15, 30, 45 and 60% sucrose solu- 1992): T amalthea Olivier is the correct name for tions. For Melipona beecheii, 30 bees were scored Panamanian and Central American bees formerly at 25, 35, 45, 55 and 65% sugar. The ambient called T silvestriana Vachal (holotype in Paris, temperature was 26-32°C. In both sets of assays, National Museum of Natural History), the senior some bees were scored more than but once, synonym of T truculenta Almeida. T muzoensis probably not more than 3 times. This procedure Schwarz is the correct name for bees from and produced ’pseudoreplication’ (Hurlbert White, Panama formerly called T pallens or T pallida. T 1993) but does not affect our statistical results. necrophaga Carmago & Roubik, from Panama The content of sucrose solutions on a vol- sugar and Costa Rica, was originally referred to as T ume basis was calculated from a table given in hypogea Silvestri in Panama studies. M panam- Kearns and Inouye (1993: p 172). This allowed ica Cockerell is the correct name for M ’fasciata’ energy intake to be quantified as calories con- Latreille from Panama, from which Schwarz sumed per second (where 1 mg sucrose provides (1932) unfortunately designated the lectotype for 3.938 which seemed to the calories), preferable a Mexican Melipona of another species. Sub- alternative notation, joules per second (Watts). species of M panamica include M p fuscopilosa Either 1 or 2 colonies were used of each melipo- Moure & Kerr, M p indecisa Cockerell, M p costari- nine for which studies were made in both species, censis Cockerell and M p melanopleura Cock- the dry and wet seasons of 1983 and 1990. erell (see Roubik, 1992, Camargo, 1994), and Foraging response to amino thus its geographic range extends at least from acids was compared to control solutions of both Brazil to Nicaragua. M micheneri Schwarz was 25 and 50% sugar concentration, and generally formerly called M marginata micheneri. RESULTS to 48%, with individual extremes of 12-59%, while a greater range of means, 20-61% (extremes 5-67%) was found for Nectar harvest in nature meliponines (figs 2 and 4). Centris ranged from 31 to 67% (fig 3). Associations between concentrations and bee were Nectar statistics in table I and figures 1-4 sugar group highly significant (contingency table test, χ2 give the mean, mode and ranges of con- = 634, P < 0.0001, table II). However, nec- centrations used by 3 bee tribes, and are tar concentrations differed temporally only compared to time of day. Nectar gathered by for Meliponini (χ2 = 214, P < 0.0001), while Meliponini (16 species) averaged 44%, they were stable through the day for Euglossini (13 species) averaged 37%, and Euglossini (χ2 = 0.26, P = 0.88) and Centri- Centridini (2 groups) had a mean of 48% dini (χ2 = 0.17, P = 0.92, table III and fig 4). sugar. Euglossine means ranged from 22 Although bee tribes showed nectar dif- ferences, stingless bees of varied size and genera harvested the highest mean sugar content. M favosa Fabricius collected nectar averaging 61 %, T necrophaga 55%, and Nannotrigona testaceicornis Cresson 53% sugar, compared with Centris spp averaging 52%. However, the samples of Centris species were limited. Both Centris and M beecheii Bennett foraged nectar of 67% sugar, while the highest sugar content taken by Euglossini was 59% (table I). Mode nec- tar sugars were generally greater than means, particularly for Meliponini and Euglossini, although T fulviventris Guérin and M panamica displayed modes lower than means (table I).

Profitability of nectar having tended to be at relatively high concentra- different sugar concentrations tions for the larger bees. However, samples were relatively small, and thus variation was Artificial nectars were imbibed at rates that not significant (Anova and Tukey’s highly increased with bee body size (tables I and significant difference test) for T muzoensis IV). The greatest rate of caloric intake at 30-45% sucrose solutions, T fulviventris muzoensis, preferred solutions of at least 50% sugar in this experimental setting, with an exception in T corvina Cockerell.

Response to amino acids

All 8 Meliponini used amino acid/sucrose solutions in largely the same way as solu- tions of pure sucrose (fig 5, table V). Results from 50 and 25% sugar solutions were com- bined for analysis, due to similar foraging at amino acid solutions and the sugar con- trols. Although glutamic acid often depressed colony foraging (fig 5), statisti- at 30-60% solutions, Scaptotrigona luteipen- cal analysis of individual replicates by Anova nis Friese at 45-60% solutions, nor M showed significant effect only for M fuligi- beecheii foraging 35-45% sugar solutions. nosa Lep (table V). This species also for- M beecheii attained highest caloric reward at aged substantially less at solutions con- 65% solutions, while Cephalotrigona capi- taining serine, glycine, alanine, and arginine tata Cockerell and S luteipennis did so at (table V). M panamica also avoided proline. 45-60% solutions. T muzoensis obtained For the significant differences (Anova), the the greatest reward from 30-45% solutions. amino acids were less in foragers to an Figure 5 provides another view of profitability extent comparable to that of a 0.2-0.6 from feeders presented simultaneously. The decrease in sugar content between pure figure suggests most species, including T sucrose solutions (fig 5). DISCUSSION foraged nectar with less than 31 % sugar. Flowers and floral morphology used by each bee consist of used Considering nectar quality for 37 species, group primarily species little the other 2 1989). tropical bees specialize on nectar of 17-63% by (Roubik, had extended sugar. Excluding 5 species using the most Euglossines tongue lengths mentum base to of of dilute nectars, the modes were no lower (from tip glossa) 12-38 while those of the bees than 39%. These figures are modes of arith- mm, stingless were 2-6 and those of Centris were metic means, ie the sugar concentration in mm, 11-14 mm. Nectar of flowers with tubu- bee foraging loads. Therefore, ranges of long, acceptable nectar sweetness for individu- lar corollas used by euglossines tends to als are hidden, but likely expressed in the have a lower sugar concentration, remaining extremes among bees. Euglossine species relatively constant through the day, in used less concentrated nectars and so marked contrast to smaller more open flow- ranges averaged 22-48%, compared to ers, including those of short, fused corollas 20-61% for Meliponini, while no Centridini or exposed to sunlight (Corbet et al, 1979;

Baker et al, 1983). These characteristics trend in foraging behavior is justified, because are indicated by consistency in daily nectar arithmetic means may have no actual rep- concentration for euglossines, its changing resentation among individual cases. How- daily values for Meliponini, and lower sugar ever, mode and mean averages usually fell concentrations in euglossine nectars (tables within ranges predicted by imbibement I, III, fig 4). Nectar in tropical dry forests may experiments (tables I, IV). All species prof- range from 5-80% in sugar concentration, ited least at the dilute solutions of 15 or 25% where bees may visit flowers in dry season sugar, although for T muzoensis, 15 and 60% averaging 46% sugar (Baker, 1978; Baker et solutions were equally profitable (table IV). al, 1983). If there are ’nectar-niches’ deter- Mode and mean averages of nectar differed mined by sweetness, these may occur in slightly less for Euglossini (modes averaged the range of >60% sugar and <30% sugar, 1.06 times the mean) than for Meliponini which were rare among large bees for (modes were 1.07 times the means), while Euglossini and Centridini, respectively. combined species of Centris showed com- Meliponini seem to use all kinds of nectar, parable mean and mode (table I). with individual extremes of 5 and 67% sugar. Mode average nectar selection may be We assume that mean and mode nectar near the optimum, and bees differing pri- concentration used individual bees are by marily in size may have predictable nectar similar, eg, foraging loads were not bimodal, selection differences, but further qualifica- with 30% derived from floral sugar primarily tions are needed. Despite potential behav- nectars of 15 and 45%. ioral differences between social and soli- Do bees forage for optimal nectar con- tary bees, and the problems that small bees centration? Our data indicate that they do. or those with extremely long proboscides Whether they were social or solitary, large or may have when imbibing very viscous nec- small, often had no apparent relevance to tar, small singless bees such as Nannotrig- the end result. We predict that a sizeable ona occasionally specialized on nectar of proportion of flowers must have nectar of the same quality as large bees like Centris optimal sweetness for pollinating bees, of Melipona (table I). For Euglossini, choice regardless of flower nectar volume. The 6 experiments (Kato et al, 1992) suggest Meliponini for which imbibement rates were these bees prefer 51 over 34% sugar solu- recorded had different ranges of optimal tions. However, their nectar flowers (see profitability (table IV). Smaller species had Roubik, 1989), being tubular and unlikely lower uptake rates, similar to trends among to change in sugar concentration through solitary bees and Bombus (Harder, 1983). evaporation through the day (table III) very The smallest Meliponini seemed less likely seldom contain nectar of >50% sugar. This to profit from viscous nectar. Whereas T added constraint implies that Euglossini are muzoensis fed most profitably at solutions of less likely to forage optimal nectars, but only 30-45% sugar, M beecheii did so at con- because these are seldom available. centrations near 65% sugar. T fulviventris In addition to bee tongue length, other benefitted from concentrations of equally aspects of feeding structure likely affect nec- 30-60%. Concentrations of 45-60% were tar profitability. The flabellum of T fulviven- for the other optimal species. tris is enlarged and very different from other Comparison between species’ perfor- meliponines (Michener and Roubik, 1993), mance in feeding experiments and their mode which may allow them to load the proboscis and mean nectar sugar harvest in nature using 30-60% sucrose with equal profit (see suggests modes are good predictors of opti- Harder, 1986, 1988), and also accept more mal caloric value. Examining the majority dilute solutions (tables I, IV). Tetragona have a relatively wide proboscis, which may allow such sugar concentration changes in inso- them to specialize in more viscous sugar lated flowers visited by Melipona. solutions (T dorsalis, table I). T necrophaga, Suboptimal nectar is frequently taken by with its liquified flesh foraging habit socially foraging species. Bees that perfo- (Camargo and Roubik, 1991), is therefore rate the flower to obtain nectar-virtually all accustomed to viscous substances imbibing of Trigona ss but also some Parta- and use also appears to very viscous nectar. mona-often used low-sugar nectars. This It uses non-floral nectar apparently only resource is even less rewarding, due to the sources (D Roubik, unpublished data, pollen extra investment made in removing floral from the 6 analyzed honey). Among tissue and defending the resource. These now 4 can be called Melipona studied, high bees also visit flowers normally, but sugar sugar specialists. The 2 others, M panamica concentration in their nectars, along with and M fuliginosa, profit from nectars of vary- those of non-robbing M panamica, dis- and to have the most ing quality appear played mode averages much lower than recruitment rapid forager capability. the means (table I). This result agrees with In the above examples, some stingless previous data for T ferricauda Cockerell bees harvest nectar ignored by other for- (18.4% mean, in the wet season) and Par- agers. Small bees like Nannotrigona would tamona peckolti Friese (given as T gain by using resources unattractive to others, musarum), with a wet season mean of because the proportion of a full foraging 19.3% sugar (Roubik et al, 1986), and also load from single flower could compensate for M panamica (Roubik and Buchmann, both time and effort. Despite higher caloric 1984). T corvina is an aggressive group values, such nectar is likely to be relatively forager that apparently did not discriminate scarce or inaccessible. Conversely, when resource quality in small patches (fig 5). All potential caloric harvest rate is very rapid, appear to profit more than do other species cooperating social bees may be more willing from dilute nectar. If relatively high rates of to accept less profitable nectar, performing harvest are realized, then caloric reward is as ’rate maximizers’ (Stephens and Krebs, comparable to that obtained from sweeter 1986). Long-term rate maximizers appar- nectars harvested at slower rates. Bees are ently forage inefficiently for some periods, behaving as rate maximizers. but net caloric intake is acceptable to them. The amount of nectar within flowers is Roubik and Buchmann (1984) give exam- measured by foragers, along with the sugar ples of nectar concentration being similar concentration; their product can indicate all day for Melipona collection 20 or 60% profitability (Real, 1981; Southwick, 1982). sugar, which demonstrates 2 extremes in Risk aversion has been detected for Bom- foraging. Therefore, some flowers do offer bus (Real, 1981), defined as avoidance of nectar near the optimal concentration for flower species that vary greatly in reward bees through the day. Others offer nectar at a given time. However, such flowers may of low sugar concentration, or gradually still be visited if their nectar sugar concen- increase the sugar concentration (Corbet tration increases (Southwick, 1982; Roubik et al, 1979). Increases of 35-60% are known and Buchmann, 1984). The studies of Real in tropical flowers, despite high humidity and (1981) and recent studies of risk-sensitive low insolution (eg, Hybanthus prunifolius foraging (Banschbach and Waddington, [Schult] Schulze, Violaceae, see Roubik and 1994) use sugar concentrations below opti- Buchmann, 1984). Data for Verbesina mality or modes that we observed; thus their gigantea Jacq, Coll (Compositae, D Yanega results should be tested using more prof- and D Roubik, unpublished results) show itable nectar. Non-sugar nectar components have been la réaction des abeilles à 10 solutions thought to affect optimal foraging (Baker, d’acides aminés (tableau V, fig 5). La teneur 1978). In our study, amino acids did not en sucres des nectars optimaux se situait have a deterrent effect for robbing bees like entre 35 et 65%, ce qui correspond au buti- Trigona and Partamona. Because the con- nage modal (fréquence la plus élevée de centrations we prepared were up to 10 times nectar naturellement butiné). Les abeilles that observed among neotropical flowers ont accepté aussi des nectars non optimaux (Baker et al, 1983), the amino acids seem et optimisaient le taux de prélèvement. Chez unlikely to serve either as attractants or 2 espèces, le mode était nettement situé repellents. en-dessous de la moyenne, ce qui prouve Bee pollination management strategies qu’elles sélectionnent à l’occasion des nec- tars dilués. Pour la and selective breeding programs for nectar pollinisation, l’hétérogé- néité et la douceur du nectar sont sugar content in crops (Davis, 1995) might optimale vraisemblablement des mécanismes clés focus on the mode average sugar content in nectar taken by the target pollinator under pour fidéliser les butineuses. Les Euglos- sini en une teneur en favorable conditions. All Euglossini, and préféraient moyenne sucres de 38%, les de 44% et some Meliponini likely to perform as thieves Meliponini les Centridini de 48%. Toutes les abeilles or robbers will rarely visit flowers providing ont récolté des nectars à 30-45% de nectar sugar concentrations over 60%. sucres, certaines aussi des nectars à 10-15% ou Desirable pollinators would seek greater rewards. 65-70% de sucres. D’autres n’ont pas butiné de nectar ayant plus de 60% de sucres. Les solutions d’acides aminés aux concentra- ACKNOWLEDGMENTS tions comprises entre 35 et 80 mM n’ont pas affecté l’attractivité, mais Melipona a évité l’acide glutamique, la glycine, la sérine, We acknowledge support from the Scholarly Stud- l’alanine et la proline. Ceci correspond à ies Program (Smithsonian Institution) to DWR une réaction à une solution sucrée à and help in field studies from E Moreno, W Devia, à une solution sucrée à R Villanueva, A Búrquez, J Ackerman, L Esco- 20-40% par rapport bar, and collaboration from INDERENA for work 50%. in Colombia. The comments of an anonymous reviewer were helpful. nectar / composition chimique / butinage optimal / pollinisation / zone tropicale / Apidae Résumé — Du butinage optimal de nec- tar par certaines abeilles tropicales (Hymenoptera: Apidae). Dans des habi- Zusammenfassung — Optimales Nek- tats riches du point de vue floristique, on a tarsammeln von einigen tropischen Bie- étudié la concentration en sucres du nec- nen (Hymenoptera, Apidae). Für 13 Arten tar butiné par 13 espèces d’Euglossini, 16 der Euglossini, 16 Arten der Meliponini und espèces de Meliponini et 8 de Centridini 8 Arten der Centridini (Apidae) wurde die (Apidae). Pour 6 espèces de Meliponini on Zuckerkonzentration von in floristisch rei- a calculé la rentabilité des solutions sucrées chen Habitaten gesammeltem Nektar ermit- (énergie consommée) en conditions expé- telt. Für 6 Arten der Meliponini wurde die rimentales et on l’a comparée avec la teneur Profitabilität von Zuckerlösungen anhand en sucres des nectars récoltés en condi- der Aufnahmeraten ermittelt und mit dem tions naturelles (figs 1-4 ; tableaux I-IV). Zuckergehalt von natürlich gesammeltem Des tests de butinage ont permis d’étudier Nektar verglichen (Abb 1—4, Tabellen I-IV). Darüberhinaus wurde die Reaktion von Meli- Banschbach VS, Waddington KD (1994) Risk-sensitive ponini-Arten auf 10 Aminosäure-Lösungen foraging in honey bees: no consensus among indi- viduals and no effect of colony honey stores. Anim in Sammel-Testanordnungen untersucht Behav 47, 933-941 (Abb 5, Tabelle V). Der Zuckergehalt von Camargo JM de Franco (1994) Biogeografia de optimalem Nektar lag je nach Art zwischen Meliponini (Hymenoptera, Apidae, Apinae): a fauna 35-65%, dies entsprach jeweils dem Modus amazônica, In: Primiero encontro sobre abelhas de Ribeirão Preto Zucchi et Editora des natürlich Nektars. Die (R al, eds), Legis gesammelten Summa Ltda, Ribeirão Preto, Brazil Bienen nahmen auch allerdings unteropti- Camargo JM de Franco, Roubik DW (1991) Systemat- male Nektare an, in manchen Fällen opti- ics and bionomics of the apoid obligate necrophages: mierten sie die Aufnahmerate. Bei zwei der the Trigona hypogea group. Biol J Linnean Soc 44, Arten lag der Modus deutlich unter dem Mit- 13-39 telwert, was auf die Auswahl Corbet SA, Willmer PG, Beament JW, Unwin DM, Pr&jadnr;s gelegentliche Jones OE determinants of von Nektaren hin- (1979) Post-secretory niedrig-konzentrierten sugar concentration in nectar. Plant Cell Environ 2, deutet. Hautpmechanismen zur Erreichung 293-308 von Blütenstetigkeit zur Bestäubung sind Davis AR (1995) Floral biology and research techniques. wahrscheinlich die Verschiedenartigkeit und In: Pollination of Cultivated Plants in the Tropics (DW die optimale Süße des Nektars. Die Eu- Roubik, ed), FAO, Rome, Italy glossini bevorzugten im Mittel einen Zucker- Eickwort GC, Ginsberg HS (1980) Foraging and mat- ing behavior in Apoidea. Annu Rev Entomol 25, 426- gehalt von 38%, die Meliponini 44% und die 446 Centridini die Modi im Mittel 48%; lagen Endress PK (1994) Diversity and Evolutionary Biology of 3-4% höher. Nektare mit 30-45% Zucker- Tropical Flowers. Cambridge University Press, Cam- gehalt wurden von allen Bienen gesammelt. bridge, UK Einige der Bienen nutzten sogar Nektare Harder LD (1983) Functional differences of the pro- mit 10-15% oder 65-70% Zucker, während boscides of short- and long-tongued bees (Hymenoptera, Apoidea). Can J Zool 61, 1580-1586 andere Arten keinen Nektar mit über 60% Harder LD (1986) Effects of nectar concentration and Zucker sammelten. Aminosäurelösungen flower depth on flower handling efficiency of bum- von 35-80 mM hatten keinen generellen ble bees. Œcologia (Berlin) 69, 309-315 Einfluß auf die Attraktivität, allerdings ver- Harder LD (1988) Choice of individual flowers by bum- mied Melipona Glutaminsäure, Glycin, Serin, ble bee interaction of morphology, time and energy. Alanin und Prolin; ihre Reaktion entsprach Behaviour 104, 60-77 Hurtbert SH, White MD (1993) Experiments with fresh- dabei etwa der von 20-40% Zuckerlösung water invertebrate zooplanktivores: quality of statis- im Kontrast zu 50% Zuckerlösung. tical analyses. Bull Marine Sci 53, 128-153 Inouye DW, Favre ND, Lanum JA et al (1980) The effects Nektar / chemische Zusammensetzung / of nonsugar nectar constituents on nectar energy optimales Sammeln / Bestäubung / Tro- content. Ecology 61, 992-996 pen / Apidae Inouye DW, Waller GD (1984) Responses of honey bees (Apis melifera) to amino acid solutions mimicking floral nectars. Ecology 65, 618-625 Kato M, Roubik DW, Inoue T (1992) Foraging behavior REFERENCES and concentration preference of male euglossine bees (Apidae: Hymenoptera). Tropics 1, 259-264 Baker HK (1978) Chemical aspects of the pollinisation Kearns CA, Inouye DW (1993) Techniques for Pollina- biology of woody plants in the tropics. In: Tropical tion Biologists. University Press of Colorado, Niwot, Trees as Living Systems (PB Tomlinson, MH Zim- USA Univ Press, mermann, eds), Cambridge Cambridge, Michener CD, Roubik DW (1993) Observations on the UK labium of meliponine bees (Hymenoptera: Apidae). Baker HK, Bawa KS, Frankie GW, Opler PA (1983) In: Evolution of Societies: Comparative Soci- Reproductive biology of plants in tropical forests. In: ology of Bees, Wasps and Ants (T Inoue, S Tropical Rain Forest Ecosystems (FB Golley, ed), Yamane, eds), Hakuhin-sha Publishing Co, Tokyo, Elsevier Scientific, Amsterdam, The Netherlands Japan Real LA (1981) Uncertainty and pollinator-plant inter- Roubik DW, Buchmann SL (1984) Nectar selection by actions: the foraging behavior of bees and wasps Melipona and Apis mellifera (Hymenoptera: Apidae) on artificial flowers. Ecology 62, 20-26 and the ecology of nectar intake by bee colonies in Roig-Alsina A, Michener CD (1993) Studies of the phylo- a tropical forest. Œcologia (Berlin) 61, 1-10 geny and classification of long-tongued bees Roubik DW, Moreno JE, Vergara C, Wittmann D (1986) (Hymenoptera: Apoidae). Univ Kansas Science Bull Sporadic food competition with the African honey 55, 124-160 bee: projected impact on neotropical social bees. Roubik DW (1989) Ecology and Natural History of Tropical J Trop Ecol 2, 97-111 Bees. Cambridge University Press, New York, USA Schwarz HK (1932) The genus Melipona. Bull Am Mus Roubik DW (1992) Stingless bees: a guide to Panama- Nat Hist 63, 231-456 nian and Mesoamerican species and their nests Sokal RR, Rohlf FJ (1981) Biometry. Second edition. (Hymenoptera: Apidae; Meliponinae). In: Insects of WH Freeman, San Francisco, CA, USA Panama and Mesoamerica: Selected Studies (D Quintero, A Aiello, eds), Oxford Science Publica- Southwick EE (1982) ’Lucky hit’ nectar rewards and tions, Oxford, UK energetics of plant and pollinators. Comp Physiol Ecol 7, 51-55 Roubik DW (1994) Wild bees of Brunei. In: Tropical Rainforest Research: Current Issues (DS Edwards, Southwick EE, Loper GM, Sadwick SE (1981) Nectar Earl of Carnbrook, ed), Universiti Brunei Darussalam, production, composition, energetics and pollinator Bandar Seri attractiveness in spring flowers of western New York. Am J Bot 68, 994-1002 Roubik DW, Ackerman JD (1987) Long-term ecology of euglossine orchid bees in Panama. Œcologia (Berlin) Stephens DW, Krebs JR (1986) Foraging Theory. Prince- 73, 321-333 ton Univ Press, Princeton, NJ, USA