Relation between size and foraging range in stingless bees (, Meliponinae) Mgl van Nieuwstadt, Ce Ruano Iraheta

To cite this version:

Mgl van Nieuwstadt, Ce Ruano Iraheta. Relation between size and foraging range in stingless bees (Apidae, Meliponinae). Apidologie, Springer Verlag, 1996, 27 (4), pp.219-228. ￿hal-00891340￿

HAL Id: hal-00891340 https://hal.archives-ouvertes.fr/hal-00891340 Submitted on 1 Jan 1996

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

Relation between size and foraging range in stingless bees (Apidae, Meliponinae)

MGL van Nieuwstadt CE Ruano Iraheta

1 Proyecto Regional de Apicultura y Meliponicultura, Universidad Nacional de Costa Rica, Apdo Postal 475-3000, Heredia, Costa Rica; 2 Facultad de Ciencias Agronomicas, Universidad de El Salvador, San Salvador, El Salvador

(Received 20 March 1996; accepted 15 July 1996)

Summary— We estimated the maximum foraging distances of four species of neotropical stingless bees using two experimental methods. In the first experiment marked bees, which were released at different distances from the nest, were recaptured. The second method consisted of training bees on an artificial nectar source which was subsequently moved away from the nest till visitation ceased. In both exper- iments we found a positive linear relationship between the head width of the different species and their foraging range. However, the maximum foraging distances estimated by the capture-recapture exper- iment were about 300 m greater than those estimated by training the bees at an artificial source. It is concluded that the distance delimiting the area in which 95% of the foraging activity occurs lies in between the distances indicated by these two methods.

Apidae / Meliponinae / / flight distance / body size

INTRODUCTION ited. As a result food competition is restricted to a certain number of neighbouring colonies and The stingless bees (, Apidae, (Hubbell Johnson, 1977). Meliponinae) are a pantropical group of Because of the sessile character of a eusocial which live in perennial colony the pollination capacity of stingless colonies. In their nests they build up bees is limited too (Free, 1993). is reserves of pollen and honey, which help not dispersed further than the foraging area them to survive dearth periods. Like all ani- of the colony. Notwithstanding the high den- mals which forage from a central place sities of stingless bees in the tropics this (Schoener, 1979), the foraging area is lim- may limit their importance as ,

* Present address: Department of Comparative Physiology, Utrecht University, PO Box 80 086, 3508 TB Utrecht, the Netherlands especially as far as small species are con- The investigations were carried out in April cerned. and May 1995, at the end of the dry season, and consisted of two experiments. The first experi- Although the maximum foraging range ment was a modified version of the experiments and the distribution of the foragers of a nest done by Gary (1971 ) and Roubik and Aluja of stingless bees are of fundamental impor- (1983). We determined the fraction of bees that tance for our understanding of a bee com- returned after being released at various distances munity and the importance of stingless bees from the nest. Bees feeding on an artificial nectar source in front of the nest were captured and as pollinators, detailed data on this subject given a distance-specific colour mark. The cap- are scarce. Some scattered information is tured bees were fed a 50% honey water solution. available from studies that have been per- Subsequently 30 bees for each distance were formed in the neotropics, usually by train- released at different distances from the nest. The ing bees on an artificial nectar source (Kerr, ones that actually flew away were counted. We 1959; Lindauer and Kerr, 1960). More recaptured the returning workers of Ta by replac- ing the original colony by an empty nest box, detailed information has been gathered on which was provided with the entrance tube of the fasciata and Melipona (Cephalotrig- original colony and which contained some honey. ona) capitata by releasing marked foragers The released workers of the other three species at different distances from the nest (Roubik were recaptured on the nectar source in front of and Aluja, 1983). But no comparison has the nest for 2.5-3 h after the release. The exper- been made between the results obtained iment was repeated twice with Pc and four times with Nt and Tc. by the releasing of marked bees and by Ta, distances were also estimated training bees on an artificial source, nor has Foraging by training foragers on an artificial nectar source. any attempt been made to uncover the rela- The source consisted of a styrofoam platform tion between the size of the various stingless with about 40 small holes which floated on a bee species and their respective foraging clove-scented 35% sugar solution. To facilitate ranges. In this study, we investigate the for- orientation, the source was marked with coloured aging ranges of several species of sting- paper. As soon as the source was frequented by less bees and estimate the relation between at least 25 bees, it was moved away from the nest in steps of 5-10 m till visitation ceased. The the size of a species and its foraging range. number of visiting bees was counted every 10 m. Foragers of other bee species were removed from the source. The experiment was repeated MATERIALS AND METHODS three times with Nt, Ta, and Tc and four times with Pc.

The research was done in Pozo Azul, a village Before fitting linear regression lines to the results of the we in the neotropical dry forest zone in Guanacaste, capture-recapture experiment Costa Rica (10°10’24"N; 85°00’24"W; altitude excluded the distances from which no bees 170 m). The village is situated in the tropical dry returned, since these might have distorted the of the between head width forest zone and is surrounded by a hilly land- analysis relationship scape with a savannah woodland vegetation. and foraging distance (Roubik and Aluja, 1983). In the source Along the rivers are small strips of secondary experiment, we estimated the area gallery forest. We studied four species of stingless in which 75, 95 and 100% of the bees foraged bees: Trigona (Trigona) corvina Cockerell, Par- cumulatively. We compared the results of the two tamona aff cupira Smith, Trigona (Tetragonisca) experiments by using two-tailed t-tests with angustula Illiger, and Nannotrigona testaceicornis repeated measurements in order to assess the of of the distribution of bees in perilampoides Cresson. Species names are degree similarity the two abbreviated as Tc, Pc, Ta and Nt, respectively. experiments. Nomenclature follows Roubik (1992). The We estimated the relation between the size colonies studied had been present at the research of the bee species and their foraging range. As an site for at least 3 months before the experiments estimate of body size we used the head width, were conducted. because this is a well-defined characteristic which can be measured unequivocally and, within the species varied from 623 to 853 m. In three tribe of stingless bees, relates well to the body cases the regression line indicates that the size. percentage of returning bees would be between 60 and 70% if the bees were released at 0 m from the nest. In the case of RESULTS Ta, where the returning bees were captured by changing the maternal nest box, the per- The percentages of returning bees released centage of recaptured bees was lower. at different distances from the nests are The distribution patterns of foragers shown in figure 1. In all cases we obtained trained at the artificial nectar source differ significant results from linear regression for the different species. The number of for- lines that were fitted through the data (table agers of Tc and Nt decreases gradually I). Maximum foraging distances of the four with increasing distance (fig 2). The num- ber of foragers of Pc and Ta decreases species, the distances predicted by the cap- more rapidly at the shorter distances, while ture-recapture experiment were significantly at the longer distances a low number of larger than the distances reached with the bees persistently visits the source. Espe- artificial nectar source (Nt: df = 5, t-value = cially some foragers of Ta were very per- 19.5, P < 0.0001; Ta : df= 5, t-value = 13.4, sistent in visiting the source. In all cases, P < 0.0001; Pc: df = 4, t-value = 8.68, P = the number of bees increased over the first 0.001; Tc: df= 5, t-value = 4.08, P= 0.0095). 10-50 m of the experiment, because at the We analysed the relationship between of the we started beginning experiment, the head width and the range within which moving the nectar source before the maxi- 95% of the bees returned cumulatively mum number of visitors had been reached. (including the results of Roubik and Aluja We compared the distances at which (1983), table II, fig 3), based on the pre- each species had attained 95% cumulative dicted values of the linear regressions. foraging activity as predicted by the two Although the distances reached by species experiments, using a two-group t-test. For all in our study are not strictly related to the head width of the species, the overall rela- nectar source could not be compared with tionship is highly significant (P= 0.0026). other studies. The distances delimiting the We also compared the distance that area in which 95% of the cumulative forag- delimited the area in which cumulative for- ing occurred and head width are significantly related in this aging at the artificial source was 95% and experiment (P = 0.0013). the head width of the species (fig 3). The The results of the two experimental meth- results of the experiment with the artificial ods are remarkably similar. The slopes of both regression lines are almost identical. in which 95% foraging activity occurs in the On the other hand, the maximum foraging natural situation lies in between the two cal- distances found in the capture-recapture culated regression lines, because both experiment are about 300 m longer than in methods have their shortcomings. the with the artificial source. experiment Releasing bees at a certain distance from the nest probably results in an overestima- tion of the actual foraging distance because DISCUSSION the bees have to fly only one way. On the other hand, the percentage of returnees Our study, combined with the data pre- from long distances might be underesti- sented by Roubik and Aluja (1983), show mated if the bees are not familiar with the that there is a positive linear relation neighbourhood in which they are released, between the maximum flight range of sting- and the costs of reorientation are high. less bees and their head width. However, The chance that a bee will accidentally the linear relations obtained with the two reenter a familiar area depends on the experimental methods differ by some 300 m. search methods and orientation mecha- Probably, the distance delimiting the area nisms it uses. Many ranging from isopods and ants to birds, and probably also bees, are known to explore their environ- ment by ’random walks’ (Waterman, 1989; Wehner et al, 1996). The relative impor- tance of the various mechanisms used for orientation during the exploration may vary widely, depending on the availability and utility of specific cues in the environment. The use of the polarisation of the sun- light (review by Wehner and Rossel, 1985) is of no use if the bee does not already know the direction from the place where it is released to her nest. If bees depend on visual cues in the landscape for their orien- tation, the percentage of returnees will decrease rapidly with increasing distance from the familiar area (Roubik and Aluja, 1983). A worker that finds itself outside its original foraging area may orientate herself with a familiar landmark inside this area. However, landmark orientation is limited by bee vision and memory (Gould, 1991; Dyer et al, 1993), and is not likely to be very suc- cessful over long distances in a dense and monotonous forest vegetation. Under such circumstances the use of odours (Lindauer and Kerr, 1960; Wenner et al, 1991) may be an important orientation mechanism. The outcome of the capture-recapture technique is not only influenced by biologi- cal and physiological limitations of the bees, spatial distribution of the returnees in dif- but also by characteristics of the experi- ferent ways. As explained above, orienta- mental design. The percentage of bees tion may be more difficult in a forest because recaught both by collecting metal marks with the orientation mechanisms the bees rely a magnet (Roubik and Aluja, 1983) and by upon may be more restricted in a dense capturing returning bees on a nectar source vegetation than in a more open landscape. in front of the nest is between 60 and 70%. On the other hand, overheating during flight In contrast, the use of an empty ’trap’-nest caused by solar radiation is much less of a was less efficient, probably because not all problem in a forest than in an open vege- returning bees entered it. The first two meth- tation. If such differences had the potential ods are therefore preferable. to influence the results of the experiments, The other experiment, in which bees their influences have probably levelled off each other. were trained at an artificial source, was likely to yield an underestimation of the natural The relationship between head width and flight range, because some bees were lost foraging range cannot be extrapolated to during the experiment because they could Apidae other than stingless bees. The for- not relocate the source after it had been aging distances for several species of the moved. The concentration of the provided genus Apis, inferred from dances for natural sugar solution was chosen within the range food sources, differ considerably from the of naturally collected nectar (Roubik et al, distances expected from our experiments 1995), but a higher sugar concentration (Visscher and Seeley, 1982; Dyer and See- might have stimulated the foragers to con- ley, 1991). No relation between body size tinue source visitation over a longer dis- and foraging distance was found in the com- tance. Moreover, the artificial source is visu- parison of four Apis species (Dyer and See- ally much less attractive than a tree in full ley, 1991). The lack of a relationship in these bloom. Therefore, the motivation to continue two comparisons may be an effect of mor- foraging will cease at a shorter distance. phological or physiological features, such In spite of the highly significant relation- as a relatively longer wing length, a larger ship, the variation between head width and wing area/body weight ratio, and a higher the maximum distance of return in the various wing beat frequency. These features, in species studied is considerable. Especially interaction with the body size, define the Ta had a larger flight range than expected, maximum flight range of a species (Roubik, whereas Tc had a relatively short flight range. 1989, p 83). Probably over a long distance, overheating Although the foraging range of a nest of causes a fall out among the returning for- stingless bees is limited, it competes for food agers (Heinrich, 1980). The air temperature sources with a number of neighbouring during the experiments reached about 35 °C. colonies. The number of intraspecifically com- Possibly Ta, with its light body colour, was peting colonies is difficult to be estimated for able to withstand the high insolation and air small non-agressive foragers, because they temperature for a longer time than the other are randomly dispersed or even clumped three species, which all have dark coloured (Roubik, 1991). However, the number of bodies (Willmer and Corbet, 1981). intraspecifically competing colonies can be Although the results are comparable, it estimated for species which are distributed must be stressed that the difference in veg- uniformly. Such species most often use etation between the experiment of Roubik pheromone trails in their communication for and Aluja (1983) and our capture-recap- food sources and aggressively defend ture experiment may have influenced the sources (Hubbell and Johnson, 1977). According to our calculations (fig 3), 95% scholarship which made this study possible. We of the foraging activity of the uniformly dis- thank MJ Sommeijer, JC Biesmeijer, D Koedam, persed species observed by Hubbell and LA Sánchez Chaves and two anonymous refe- rees for their comments on earlier versions of the Johnson (which vary by a factor 1.4 in head article, and SM McNabb for her linguistic advice. width) occurs in an area of 2.9 km2 (SD = 1.4; four species), the 95% foraging range for each species varying from 740 to Résumé — Relation entre la taille et la 1 240 m. By using the colony densities mea- distance de butinage chez les abeilles sured by Hubbell and Johnson (1977), we sans dard Les dis- calculated that in that area there will be (Apidae, Meliponinae). tances de chez les abeilles sans several tens of competing colonies (ave- butinage dard revêtent une importance pour rage 54; SD = 31; four species) of the same capitale notre de la entre species. The size of the home range of compréhension compétition colonies, l’influence de la déforestation sur la these species was positively related to the estimated colony weight (Hubbell and John- disponibilité des ressources et l’efficacité dans les communautés son, 1977), and in a slightly lesser extent pollinisatrice végé- to bee weight (van Nieuwstadt, personal tales naturelles et les cultures. Il existe pour- tant d’estimations de la distance maxi- observation). Since the species with a larger peu male de des différentes flight range occur in lower densities, the cal- butinage espèces culated number of competing colonies within et de la distribution de l’activité de butinage dans l’aire ainsi délimitée. Nous avons the 95% foraging range does not show a correlation with head width (Spearman rank estimé les distances de butinage de quatre correlation coefficient; Z=-0.693; P= 0.49; espèces néotropicales d’abeilles sans dard n = 4). (Trigonini) à l’aide de deux méthodes expé- rimentales. Dans la première expérience Because forager activity is highest in the nous avons utilisé la méthode de vicinity of the nest, the intensity of competi- capture- recapture (Gary, 1971) : les abeilles, qui ont tion between two colonies depends on the été lâchées à diverses distances du nid, sont distance they are removed from each other du nid. Le seconde (Roubik 1991). More than 75% of the fora- recapturées près méthode consiste à conditionner les abeilles ging activity normally occurs within 40% of à une source artificielle de the maximum foraging distance (Roubik, nectar qui est ensuite progressivement du nid 1991; van Nieuwstadt, unpublished data), éloignée jusqu’à ce ait de visites. Les while further away from the colony the fora- qu’il n’y plus distances délimitent l’aire d’où revien- ger density is low. It can therefore be conclu- qui ded that for uniformly dispersed species nent 95 % des abeilles dans l’expérience par varient de 623 à only the ten (SD = 7) colonies that are on capture-recapture 853 m tableau Dans average present in that range should be (fig 1, I). l’expérience les distances déli- considered as serious intraspecific compe- par conditionnement, qui mitent titors for the available resources. l’aire dans laquelle 95 % de l’activité cumulée de butinage a lieu s’étend de 90 à 490 m (fig 2). Dans les deux séries d’expé- ACKNOWLEDGMENTS riences nous avons trouvé une relation posi- tive linéaire entre la largeur de la tête des différentes espèces et la distance de buti- We thank the people of Pozo Azul for providing nage (fig 3). Mais les distances maximales the colonies we used in this study. CERI thanks de trouvées la méthode de the Proyecto Regional de Apicultura y Meliponi- butinage par cap- cultura (Utrecht University, the Netherlands and ture-recapture dépassent d’environ 300 m Universidad Nacional, Costa Rica) for awarding a celles obtenues par la méthode de condi- tionnement. On en conclut que les distances Dressurversuchen, in denen zu 95% eine maximales atteintes par des colonies qui allmählich zunehmende Sammelaktivität butinent naturellement se situent entre les auftrat, reichten die Entfernungen von 50m chiffres donnés par les deux méthodes. bis 490m (Abb 2). In beiden Experimenten Parce que la plus grande partie de l’activité fanden wir eine positive lineare Korrelation de butinage a lieu dans un rayon égal à 40% zwischen der Kopfbreite der verschiedenen environ de la distance maximale de buti- Arten und der Reichweite ihrer Trachtflüge nage, on peut conclure qu’une colonie d’une (Abb 3). Jedoch war bei der Wiederfang- espèce uniformément dispersée et de taille methode die maximale Flugweite um 300m moyenne (largeur de la tête comprise entre größer als die, die durch eine Dressur auf 2,4 et 3,3 mm) entre en compétition avec künstliche Futterstellen erreicht wurde. Es environ dix colonies de la même espèce. wurde daraus geschlossen, daß das Maxi- mum der Distanz, die von natürlich sam- Apidae / Meliponinae / abeille sans dard / melnden Bienen erreicht wird, zwischen den distance de vol / aire de butinage / taille Entfernungen liegt, die durch die beiden corporelle / Methoden gefunden wurden. Da die größte Sammelaktivität innerhalb von etwa 40% Relation zwischen Zusammenfassung — der maximalen Flugweite liegt, kann man Größe und Sammelflugweite bei sta- schließen, daß ein Volk einer gleichmäßig chellosen Bienen (Apidae, Meliponinae). verteilten, mittelgroßen Art (Kopfbreite von Grundlegende Kenntnisse über den Bereich, 2,4 - 3,3 mm) mit im Schnitt 10 Völkern der den stachellosen Bienen mit Trachtflügen gleichen Art in Konkurrenz steht. abdecken, sind für unser Verständnis über die Konkurrenz zwischen den Völkern, über Apidae / Meliponinae / stachellose Biene / die Folgen des Abholzens der Wälder für Flugdistanz / Körpergröße die Nutzbarkeit von Trachtquellen, über die Bestäubungsleistung in natürlichen Lebens- gemeinschaften und über Ernteerträge wich- REFERENCES tig. Es gibt jedoch nur wenige Schätzungen über die maximale Distanz der Trachtflüge FC, NA, Richard A bee von verschiedenen Arten und über die Ver- Dyer Berry (1993) Honey spatial memory: use of route-based memories after dis- teilung der Flugaktivität innerhalb dieses placement. Anim Behav 45, 1028-1030 Bereichs. Wir schätzten die Flugdistanz für Dyer FC, Seeley TD (1991) Dance dialects and foraging 4 Arten neotropischer stachelloser Bienen range in three Asian species. Behav Ecol (Trigonini) mit zwei verschiedenen Metho- Sociobiol 28, 227-233 den. Im ersten Experiment benutzten wir Free JB (1993) Pollination of Crops. 2nd edition, die Wiederfangmethode nach Gary (1971): Academic Press, London NE retrieval of ferrous labels in a Die in verschiedenen Entfernungen vom Gary (1971) Magnetic capture-recapture system for honey bees and other Nest Bienen wurden wur- freigelassenen insects. J Econ Entomol 64, 961-965 den in der Nähe des Nestes wiedergefan- Gould JL (1991) The ecology of honeybee learning. In: gen. Bei der zweiten Methode dressierten The Behaviour and Physiology of Bees (LJ Good- wir Bienen auf künstliche Futterstellen, die man, RC Fisher, eds), CAB International, 306-322 anschließend schrittweise vom Nest fort- Heinrich B (1980) Mechanisms of body temperature reg- bewegt wurden bis die Bienenbesuche auf- ulation in honeybees, Apis mellifera. II. Regulation of hörten. Die die einen Bereich thoracic temperature at high air temperatures. J Exp Entfernungen, Biol 85, 73-87 in dem 95% der Bienen den umgrenzten Hubbell SP, Johnson LK (1977) Competition and nest zurück ins Volk fanden variiert zwi- Weg spacing in a tropical stingless bee community. Ecol- schen 623 m und 853 m (Abb 1, Tab I). In ogy 58, 949-963 Kerr W E (1959) Bionomy of Meliponids -VI- Aspects of Schoener TW (1979) Generality of the size-distance food gathering and processing in some stingless relation in models of optimal feeding. Amer Nat 114, bees. In: Symposium on Food Gathering Behaviour 902-914 of Ed at Cornell Univ, Ithaca, NY, Hymenoptera. Visscher PK, TD of USA, 24-31 Seeley (1982) Foraging strategy honeybee colonies in a temperate deciduous forest. Lindauer M, Kerr WE (1960) Communication between Ecology 62, 1790-1801 the workers of stingless bees. Bee World 41, 29-41, 65-71 Waterman TH (1989) Navigation. Scientific Amer- ican Library, HPHLP, New York Roubik DW (1989) Ecology and Natural History of Trop- ical Bees. Cambridge University Press, Cambridge Wehner R, Rossel S (1985) The bee’s celestial com- pass - a case study in behavioural neuobiology. In: Roubik DW (1991) Aspects of Africanized honey bee Fortschritte der Zoologie 31, Experimental ecology in tropical America. In: The African Honey- Behavioural Ecology and Sociobiology (B Hölldobler, bee (M Spivak , DJC Fletcher, MD Breed, eds), West- M Lindauer, G 11- view Press, Boulder, CO, 259-281 eds), Fischer Verlag, Germany, 53 Roubik DW (1992) Stingless bees: a guide to Panama- Wehner Michel Antonsen P Visual nian and Mesoamerican species and their nests. In: R, B, (1996) naviga- Insects of Panama and Mesoamerica (D Quintero, A tion in insects: coupling of egocentric and geocentric Aiello, eds), Oxford University Press, 495-524 information. J Exp Biol 199, 129-140 Roubik DW, Aluja M (1983) Flight ranges of Melipona Wenner AM, Meade DE, Friesen LJ (1991) Recruitment, and Trigona in tropical forest. J Kansas Entom Soc search behavior, and flight ranges of honey bees. 56, 217-222 Am Zool 31, 768-782 Roubik DW, Yanega D, Aluja SM, Buchmann SL, Inouye Willmer PG, Corbet SA (1981) Temporal and micro- DW (1995) On optimal nectar foraging by some trop- climatic partitioning of the floral resources of Justicia ical bees (Hymenoptera: Apidae). Apidologie 26, aurea amongst a concourse of pollen vectors and 197-211 nectar robbers. Oecologia 51, 67-78