Original article

Within-colony size variation of foragers and pollen load capacity in the quadrifasciata anthidioides Lepeletier (, )

M. Ramalhoa V.L. Imperatriz-Fonsecab T.C. Giannini

aDepto. de Botânica do Instituto de Biologia da Universidade Federal da Bahia, Salvador, CEP 40170-240, Bahia, Brazil bDepto. de Ecologia Geral, Instituto de Biociências, Universidade de São Paulo, CEP 05508-900 Sâo Paulo, Brazil

(Received 6 February 1997; accepted 28 January 1998)

Abstract - Within-nest worker size variation in the eusocial stingless bees is a contingent phe- nomenon of unknown adaptive value. We assume that the magnitude of variation represents a com- promise between a minimum population of foragers and foraging efficiency at colony level. In Melipona quadrifasciata, worker size was found to vary according to colony conditions, and the pollen carrying efficiency of an individual was related to its size. On average, the foragers from a weak colony are smaller and are able to carry greater amounts of pollen per unity of body weight (’load capacity’) than the larger foragers from a strong colony. The allometric variation of the corbicula (the pollen carrying structure in the hind tibia) contributes to the observed decrease in pollen load capacity with increased body size. By higher pollen intake per worker, the colonies with smaller bees could increasing the rate of brood production and colony population recovery after population crashes. © Inra/DIB/AGIB/Elsevier, Paris Melipona quadtifasciata / stingless bees / worker size / pollen load / allometric variations

1. INTRODUCTION individual level and variability of tasks performed at colony level [16]. The range Body size variation within the worker of within-colony body size variation is the caste in colonies of eusocial Hymenop- releaser to polymorphism in eusocial tera is related to task specialization at the . Polymorphism is a well-docu-

* Correspondence and reprints E-mail: [email protected] mented phenomenon in colonial inverte- To support this hypothesis, we measured brates and some explanations have been the within-colony body size variation of proposed for its relative rarity among the stingless bee M. q. anthidioides Lepe- social insects [7]. Because worker body letier and compared the efficiency of pol- size variation is not enough to give rise to len carrying between small and large fora- polymorphism in social bees, it has been gers, from weak and strong colonies. viewed as a contingent phenomenon with unknown adaptive value. 2. MATERIALS AND METHODS The stingless bees Meliponinae, a and diverse eusocial widespread highly Foragers of two colonies of M. q. anthi- show subtle within- Apidae group [1], dioides, with Eucalyptus pollen loads were colony worker size variation [14] and divi- captured at the nest entrance, during the same sion of labor coupled with age polyethism period of time. They were individually main- [11]. Kerr et al. [13] related a queen dif- tained in snapcap vials and anesthetized below ferentiation with minimum female 0 °C for 5 min to be weighed. The bees were body with and without their loads size in anthi- weighed pollen Melipona quadrifasciata in an Sartorius balance to an and that analytical accuracy dioides, suggested poor feeding of 0.1 mg. The presence of nectar in the wor- conditions in the colony would produce kers, crop was checked by pressing their abdo- smaller workers and decrease the queen- men against absorbent paper and individuals to-worker ratio. It has been suggested with both nectar and a pollen load were dis- carded. The corbicula area of 60 bees was (Velthuis, pers. comm.) that the body size of the worker that measured under a stereoscopic microscope previous generation Wild M8 using the image analyser ’Mini-mop builds the cells determines the brood cell (Kontron-Bildanalyse)’. size and the newborn worker size distri- We use the terms weak and strong to spe- bution. If a worker size should be so, cify the relative apparent condition of such conservative and of irrespective colony colonies, without any assumption about their condition. The direct manipulation of the growth or reproductive potential. The strong nutritional condition of larvae (e.g. food colony was larger, more populated by larger quantity) probably determines a worker workers, and had a larger brood comb and size rather than the of numerous larger pollen and honey pots than body manipulation the weak brood cell size. colony. We used a t-test to comparare the average Recently, Ramalho et al. [ 18] observed forager weight of the weak and the strong the relationship between a forager body colony of M. q. anthidioides. The pollen load and the tibia area weight and pollen load weight in a study of capacity per body weight flower of bees (Meli- were plotted against body weight. For allome- constancy stingless tric we transformed the cubic used the comparison, (pol- poninae, Apidae). They pollen len load weight) and square (tibia area) dimen- weight/forager body weight ratio (pollen sions to linear dimensions by calculating the load capacity) as a relative measure of the respective root data. We calculated the regres- forager’s pollen carrying efficiency among sion lines and coefficients. species with body sizes ranging from 4 to 100 mg. They found that the smaller the worker the higher its pollen load capacity, 3. RESULTS and suggested that variation in forager size would have an adaptive value at the colony The measured body structures of the level during population crashes: for ins- M. quadrifasciata foragers (table I) were tance, a larger population of smaller wor- significantly correlated with body weight kers could be produced by the same quan- and with each other. Obviously, the cor- tity of food during food stress conditions. relation values depended on the order of values (volume-weight, area and length). 4. DISCUSSION On average, the foragers from the weak colony were significantly lighter (smal- Although stingless bee workers from a ler) than those from the strong one and strong nest have similar sizes, some varia- the smaller carried more foragers pollen tions occur, as et al. unit of load Imperatriz-Fonseca per body weight (higher capa- [10] indicated by the weight variation of than the ones. city) larger workers of Nannotrigona testaceicornis The pollen load capacity (plc) and the just after emergence. Nevertheless, size tibia area per body weight ratio (ta/bw) variations are more conspicuous when were both correlated with body weight examined among colonies of different (table I; figure 1), but not with the other health. morphometric body parameters. There- Camillo-Atique [2] studied weak, fore, the weight of pollen carried per body medium and strong colonies of the sting- weight unit was mainly an outcome of the less bee M. rufiventris rufiventris, esta- allometric development of the pollen car- that some behavioral parameters, rying structure. However, the difference blishing such as the rate of cell construction, of the slope (’b’ parameter) between plc pro- visioning and oviposition, were depen- and ta/bw regression lines (figure 1) sug- dent on conditions. Worker gests that other factors contributed to the colony flight was also influenced higher pollen load capacity of the smal- activity by colony condition Weaker colonies had a ler workers. [8, 9]. shorter period of foraging than stronger Figure 2 shows the weight frequencies colonies in M. bicolor bicolor and in of the 30 workers from strong, medium Schwarziana quadripunctata. Other and weak colonies. aspects, such as brood cell size and food discharge rate during cell provisioning, The proximate factor that causes etc., may be useful to discern the functio- within-nest size variation of workers is nal state of a stingless bee colony [14] the amount of food given to the larval which has been empirically related to the stages. In contrast, variation in worker size workers’ body size and number within may be constrained by several factors such colonies. The variations in worker size as a flight aerodynamics [6] and the tro- among weak, medium and strong colonies phic determination of female caste [24]. of M. quadrifasciata are easily observed in The latter explanation should be applied figure 2. with reservation to all stingless bees because genetic and trophic mechanisms is, the smaller the forager the higher the play a role in caste differentiation [3]. For amount of pollen it can carry per unit of the stingless bees, there is also an inverse body weight. In a study of food gathering relationship between complexity of the by bumblebees, Free [5] observed that recruitment system and within-nest varia- "several of the smaller bees were found tion in worker body size [23]. These fac- to have collected relatively large loads of tors explain why within-colony size varia- pollen for their size". The smaller fora- tion is not so large as to produce a gers also collected larger loads of sugar polymorphic worker caste. However, they syrup in proportion to their body size. do not clarify the functional role of the Obviously, below a maximum value of observed body size variation. plc for a given body size, the individual The approximately 400 species of sting- foragers could adjust their loads, depen- less bees are small to medium sized bees ding on extrinsic factors and risk sensi- (3-13 mm in length) with considerable tive behaviors, which caused a broad dis- variation in colony population size (from persion of plotted points (figure 1). Fewell several hundreds up to more than ten thou- and Winston [4] provided evidence that sands workers in a colony). They also in Apis mellifera the pollen foragers adjust exhibit a high rate of brood production their load size and trip time in relation to and require a high pollen intake. Due to the colony’s pollen storage levels. In colo- their restricted colony home range [ 12, nies of the bumblebee Bombus terricola, 15, 19, 22] and the high local plant diver- deprived of pollen, the foragers carried sity in tropical habitats, they become par- heavier pollen loads instead of increasing ticularly exposed to the turnover of food the number of foraging trips [ 17]. The sting- in their In supply colony surroundings. less bees seem to respond in a similar way such a context the reduction of worker to a reduced foraging period. It is com- of larval food in the size, through quantity mon to observe the workers arriving at the would be a alternative to deal cell, good nest entrances with very large pollen loads with food deprivation if it prevents an after periods of shortened foraging acti- additional decline in the colony popula- vities (cold, rainy days, etc.). At this time, tion. The achieved gain by producing we do not know how the size of the pollen small workers be might outweighed by load is related to the age of the worker and negative side-effects: the small bees could its foraging experience. retrieve food from a profitable food source Observational data on several at a higher cost across a given distance stingless than could the large bees because the for- bee species suggest that reduction in wor- ker size and the decrease in mer may have a higher energy consump- colony popu- lation takes after a few weeks of tion per time of foraging [2 1 ]. Most of all, place to stress conditions the higher energy consumption of the colony exposure (e.g. small workers during foraging activity after transport of colonies from the wild to conditions with a loss in fora- may cause a shortened life span [20]. laboratory ger population, and after brood comb mani- In summary we hypothesize that sting- pulations). Probably there is a decrease in less bee colonies adjust their average fora- the amount and quality of larval food in ger size after food deprivation periods to these weak colonies, as Lacerda et al. [ 14] maximize pollen foraging efficiency. In observed in Geotrigona inusitata. We such a situation it would be better to pro- hypothesize that these simultaneous duce many smaller workers because wor- changes would be adaptive; that is, they ker size is negatively correlated with the contribute to colony survival when pollen pollen load capacity - (plc) (figure 1); that supply is low for two reasons: 1) a higher number of workers discharge smaller d’une colonie forte. Les termes « faible », amounts of larval food in each brood cell « moyenne » et « forte » désignent la [14] so a higher proportion (not necessarily condition relative apparente des colonies a higher absolute number) of individuals en termes de taille des ouvrières, diamètre within the low populated colony may be du nid à couvain, effectif du nid... Chez involved in food discharging; 2) as a M. q. anthidioides la taille des ouvrières result, a smaller proportion of workers varie en fonction des conditions apparentes might be available for pollen foraging. de la colonie et l’efficacité du transport Therefore the weak colonies would be de pollen d’un individu est en relation avec best served if their workers are small sa taille (tableau I). Les butineuses d’une because with the same available food they colonie faible sont en moyenne plus petites would produce more small bees than large et capables de transporter de plus grandes bees, and the smallest ones would carry quantités de pollen par unité de poids cor- greater amounts of pollen per body bio- porel (forte capacité de charge en pollen) mass. que les butineuses plus grandes d’une colo- nie forte. Les colonies de taille moyenne présentent des valeurs intermédiaires pour ACKNOWLEDGMENTS le poids des ouvrières (figure 2). La capa- cité de charge en pollen et le rapport de We acknowledge support from the CNPq la surface du tibia postérieur au poids cor- (521776/96.1). The authors are grateful to porel (ta/bw) sont corrélés négativement Dr H.H.W. Velthuis and Dr A.M.P. Kleinert avec le poids corporel (figure 1). La varia- for their suggestions, as well as to the anony- tion allométrique de la corbicule contri- mous referees. bue à la diminution observée de la capacité de charge en pollen par rapport à la taille corporelle. La variation intra-colonie Résumé - Variation intra-colonie de la observée de la taille corporelle moyenne taille des ouvrières et de la capacité de d’une ouvrière aurait une valeur préadap- charge en pollen chez l’abeille sans dard tative chez les abeilles sans dard. En aug- Melipona quadrifasciata anthidioides mentant le taux de récolte de pollen par Lepeletier (Hymenoptera : Apidae). La unité de travail, les colonies qui ont de variation de la taille des ouvrières à l’inté- petites ouvrières pourraient accroître le rieur de la colonie est un phénomène aléa- taux de production de couvain et accélérer toire chez les abeilles sans dard, dont on ne le rétablissement de l’effectif de la colonie, connaît pas la valeur adaptative. Nous pen- lorsque celle-ci s’est effondrée suite à des sons que l’étendue de la variation repré- manques saisonniers de nourriture dans la sente un compromis entre une population zone de butinage des petites colonies, à la minimale de butineuses et l’efficacité du fondation de nouvelles colonies ou à la butinage au niveau de la colonie. Pour prédation du nid. © Inra/DIB/AGIB/Else- étayer cette hypothèse, nous avons mesuré vier, Paris la variation intra-colonie de la taille cor- porelle, du poids corporel et de la surface Melipona quadrifasciata / abeille sans de la corbicule chez l’abeille sans dard dard / taille corporelle / récolte pollen / M. q. anthidioides et comparé l’efficacité variation allométrique du transport de pollen par des butineuses de petite taille et des butineuses de grande taille issues de deux colonies. Trente buti- Zusammenfassung - Größenvariation neuses ont été capturées respectivement und Pollenbeladungsfähigkeit von à l’entrée du nid d’une colonie faible et Sammlerinnen innerhalb von Völkern der Stachellosen Biene Melipona qua- bachteten Variation der mittleren Körper- drifasciata anthidioides Lepeletier (Api- größe der Arbeiterinnen könnte bei den dae, Hymenoptera). Innerhalb von Völ- Stachellosen Bienen ein prädativer Wert kern der eusozialen Stachellosen Bienen zukommen. Hierdurch könnte die auf die ist die Variation der Arbeiterinnengröße Arbeitseinheit bezogene höhere Pollen- eine durchgehende Erscheinung; ihr adap- eintragsrate in kleinen Völkern die Brut- tiver Wert ist allerdings unbekannt. Wir produktion beschleunigen und damit die nehmen an, daß das Ausmaß dieser Varia- Erholung der Population nach Volkszu- tion einen Kompromiß zwischen mini- sammenbrüchen verbessern, wie diese maler Sammlerinnenpopulation in den durch saisonale Nahrungsknappheit in dem Völkern und der Sammeleffizienz auf dem begrenzten Futtersuchareal kleiner Völ- Volkslevel darstellt. Zur Stützung dieser ker, durch die Gründung neuer Nester, Hypothese haben wir die Körpergröße, und durch Nestberäuberung auftreten. das Körpergewicht und die Körbchen- © Inra/DIB/AGIB/Elsevier, Paris größe innerhalb von Völkern der Stachel- losen Biene Melipona quadrifasciata Melipona quadrifasciata / Stachellose anthidioides Lepeletier vermessen sowie Bienen / Arbeiterinnengröße / Pollen- die Effizienz des Polleneintrags von klei- beladungsfähigkeit / allometrische nen und großen Arbeiterinnen zwischen Variation zwei Völkern verglichen. Am Nesteingang eines schwachen sowie eines starken Volkes wurden jeweils 30 Sammlerinnen abgefangen. Die Bezeichnungen ’schwach’, REFERENCES ’mittelstark’ und ’stark’ beziehen sich hier- bei auf den Zustand [1] Camargo J.M.F., Pedro S., Systematics, phy- augenscheinlichen logeny and biogeography of the Meliponinae der wie die Größe der Arbei- Völker, (Hym., Apidae): a mini-review, Apidologie 23 terinnen, den Durchmesser der Brutwa- (1992) 509-522. ben, die Individuenzahl etc. Bei Melipona [2] Camilo-Atique C., Variabilidade do compor- quadrifasciata variiert die Arbeiterinnen- tamento de Melipona rufiventris rufiventris mit dem Zustand der Völker, hierbei (Hymenoptera, Apoidea), Dissertação de Mes- größe Ribeirão 1974. ist die Polleneintragskapazität eines Ein- trado, Preto, Caste zeltiers von seiner Größe [3] Engels W., Imperatriz-Fonseca V.L., abhängig and Im Mittel sind die Sammle- development, reproductive strategies, (Tabelle I). control of fertility in honeybees and stingless rinnen eines ’schwachen’ Völkchens klei- bees, in: Engels W. (Ed.), Social Insects, An ner und sind in der Lage relativ zu ihrem Evolutionary Approach to Castes and Repro- von Pol- duction, Springer-Verlag, Berlin, 1990, pp. Körpergewicht größere Mengen 167-230. len einzutragen (hohe Pollenbeladungs- als die Arbeiterinnen [4] Fewell J.B., Winston M.L., Colony state and fähigkeit) größeren regulation of pollen foraging in the honeybee, eines "starken" Volkes. Mittelstarke Völ- Apis mellifera, L. Behav. Ecol. Sociobiol., 300 ker zeigen mittlere Gewichte der Arbei- (1992) 387-393. terinnen (Abb 2). Die Pollenbeladungsfä- [5] Free J.B., The collection of food by bumble- higkeit und das Verhältnis der Fläche der bees, Insectes Soc. 2 (1955) 303-311. Tibia des Hinterbeins zum Körpergewicht [6] Hartfelder K., Engels W., Allometric and mul- (ta/bw) waren beide negativ mit dem Kör- tivariate analysis of sex and caste polymor- pergewicht korreliert (Abb 1). Die allo- phism in the Neotropical stingless bee, Scap- totrigona postica, Insectes Soc. 39 (1992) metrische Variation der Corbiculae trägt 251-266. daher zu der beobachteten Abnahme der [7] Harvell C.D., The evolution of polymorphism Pollenbeladungsfähigkeit relativ zum Kör- in colonial invertebrates and social insects, Q. pergewicht bei. Der in den Völkern beo- Rev. Biol. 69 (1994) 155-185. [8] Hilário S.D., Imperatriz-Fonseca V.L., A [16] Oster G.F., Wilson E.O., Caste and ecology in influência do estado da colônia na atividade de the social insects, Princeton Univ. Press., New vôo de bicolor, Anais do 2° Jersey, 1978. Encontro sobre abelhas, Ribeirão Preto, 1996, p. 285. [17] Plowright R.C., Thomson J.D., Lefkovitch L.P., Plowright C.M.S., An experimental study of [9] V.L., P., Imperatriz-Fonseca Darakjian Flight the effect of colony resource level manipula- of Schwarziana activity quadripunctata: tion on foraging for pollen by worker bumble influence of environmental factors, in: Abs- bees (Hymenoptera: Apidae), Can. J. Zool. 71 tracts of IV Behavioral Ecology Congress, Not- (1993) 1393-1396. tingham, 1994, p. 86. [10] Imperatriz-Fonseca V.L., Cruz-Landim C., [18] Ramalho M., Giannini T.C., Malagodi-Braga Pollen harvest Moraes R.L.M.S., Dwarf gynes in Nannotri- K.S., Imperatriz-Fonseca V.L., bee gona testaceicornis (Apidae, Meliponinae)- by stingless foragers (Hymenoptera, Api- behaviour, exocrine gland morphology and dae, Meliponinae), Grana 33 (1994) 239-44. status, 28 reproductive Apidologie (3) (1997) [19] Roubik D.W., Aluja M., Flight ranges of Meli- 113-122. pona and Trigona in tropical forest, J. Kans. [11] Inoue T., Salmah S., Sakagami S.F., Individual Entomol. Soc. 56 (1983) 217-222. variations in worker polyethism of the Sumatran stingless bee, Trigona (Tetragonula) minang- [20] Schmid-Hempel P., Wolf T.J., Foraging effort kabau (Apidae, Meliponinae), Jpn. J. Entomol., and life span in a social . J. Anim. Ecol. 56 64 (3) (1996) 641-668. (1988) 209-18. [12] Kerr W.E., Bionomy of Meliponids, VI. [21] Seeley T.D., The Wisdom of the Hive, Har- Aspects of food gathering and processing in vard University Press, Cambridge, 1995. some stingless bees, Food gathering in Hyme- van Nieuwstadt Iraheta Rela- noptera, Symp. Entomol. Soc. Am., Detroit [22] M.G.L., C.E.R., (1959) p. 24-32. tion between size and foraging range in stin- gless bees (Apidae, Meliponinae), Apidologie, [13] Kerr W.E., Stort A.C., Montenegro M.J., 27 (4) (1996) 219-228. Importância de alguns fatôres ambientais na determinação das castas do gênero Melipona, [23] Waddington K.D., Herbst L.H., Roubik D.W., Anais Acad. Brasil Ciênc 38 (1966) 149-168. Relationship between recruitment systems of [14] Lacerda L.M., Zucchi R., Zucoloto F.S., Colony stingless bees and within-nest worker size condition and bionomic alterations in Geotri- variation, J. Kans. Entomol. Soc. 59 (1986) 95-102. gona inusitata (Apidae, Meliponinae), Acta. Biol. 20 (1991) 109-123. [24] Wheeler D.M., Developmental and physiolo- [15] Michener C.D., The Social Behavior of the gical determinants of caste in social Hyme- Bees, A Comparative Study, Harvard Univer- noptera: evolutionary implications, Am. Nat. sity Press, Cambridge, 1974. 128 (1986) 13-34.