Apidologie 35 (2004) 159–182 © INRA/DIB-AGIB/ EDP Sciences, 2004 159 DOI: 10.1051/apido:2004007 Review article

Recruitment communication in stingless bees (Hymenoptera, Apidae, Meliponini)

James C. NIEH*

University of California San Diego, Division of Biological Sciences, Section of Ecology, Behavior, and Evolution, 9500 Gilman Drive, Mail Code 0116, La Jolla, CA 92093-0116, USA

(Received 25 August 2003; revised 28 October 2003; accepted 2 December 2003)

Abstract – The stingless bees (Hymenoptera, Apidae, Meliponini) have evolved sophisticated communication systems that allow foragers to recruit nestmates to good resources. Over the past 50 years, a growing body of research has shown that foragers can communicate three-dimensional resource location, uncovered several potential communication mechanisms, and demonstrated new information transfer mechanisms. Some of these mechanisms are unique to stingless bees and some may provide insight into how the ability to encode location information, a form of functionally referential communication, has evolved in the highly social bees. The goal of this review is to examine meliponine recruitment communication, focusing on evidence for contact-based, visual, olfactory, and acoustic communication and what these mechanisms can tell us about the evolution of recruitment communication in stingless bees.

Meliponini / information transfer / referential communication / recruitment / multimodality

1. INTRODUCTION recruitment communication of the Meliponini and the Apini (Kerr, 1969). Insect societies have evolved information Stingless bees are a monophyletic group transfer systems of remarkable complexity principally found in tropical and subtropical (Wilson, 1971). Highly social bees (honeybees areas of America, Africa, Australia, and parts and stingless bees) use sophisticated methods of Asia (Roubik, 1989). Unlike honeybees, to exploit resources such as pollen, nectar, which consist of approximately 11 species in water, resin, and nest sites (Kerr, 1969). Social one genus, stingless bees consist of several bees can recruit, increase the number of nest- mates searching for a particular resource, to hundred species distributed through more than specific or non-specific locations. Mecha- 36 genera (Michener, 2000). Honeybees and nisms that allow highly social bees to commu- stingless bees are the only highly social bees, nicate resource location have been the subject but stingless bees lead more diverse lifestyles, of intensive study, particularly in honeybees, including obligate necrophagy (Roubik, 1982a; which can communicate the polar coordinates Camargo and Roubik, 1991), and can recruit (distance and direction) of a resource through for resources such as dead animals, pollen, nec- a waggle dance (von Frisch, 1967; Dyer, tar, mud, resin, water, and nests (Roubik, 1989). 2002). Researchers have sought to understand Stingless bees also use a greater diversity of how such complex recruitment communica- recruitment communication systems (ranging tion systems have evolved, and some studies from odor trails to the potential referential have therefore focused on stingless bees, encoding of food location) than honeybees, in exploring possible homologies between the which all studied species use the waggle dance

* Corresponding author: [email protected] 160 J.C. Nieh

(Kerr, 1969; Roubik, 1989; Dyer, 2002). simple, as von Frisch (1967) noted, because Research on stingless bees (Lindauer and Kerr, although they are conceptually straightfor- 1958; Kerr and Esch, 1965; Esch, 1967; Nieh ward, there are several key details that should et al., 2003c) has uncovered intriguing similar- not be overlooked. ities and important differences between recruit- ment communication in the Meliponini and the 2.1. Intranidal observations Apini (honeybees). However, stingless bees have not been as intensively studied as honey- Lindauer and Kerr (1958) and Esch (1967) bees, and thus our understanding of their biol- pioneered the description of motions and ogy is at an early stage relative to what is known sounds produced inside the nest by foragers of the Apini. returning from a food source. In addition, they Within the last decade, interest in stingless performed intriguing preliminary experiments bee recruitment communication has renewed. testing the function of recruitment sounds. The The goal of this review is to provide an over- general method is to train bees to a feeder at a view of the field while focusing on these new specific location and record the behavior of findings, and examining their implications for these foragers when they return inside the nest understanding the evolution of recruitment (Esch, 1967). It is also possible to mark indi- communication in the Apidae. It has been vidually the entire colony and thus examine argued that olfaction is the sole source of ori- the role of intranidal contacts and the eventual entation information for honeybee recruits success of nestmates in reaching the adver- (Wenner, 2002), and this controversy may also tised food source (Seeley, 1995; Hrncir et al., play out in the stingless bees (Dyer, 2002). 2000). Thus one objective of this review is to exam- ine the evidence for olfactory and non-olfac- 2.2. Feeder arrays tory communication. I will begin by discuss- ing experimental methods and review the Karl von Frisch (1967) developed the gen- literature on recruitment to a specific spatial eral methods for investigating bee location location. I will then examine several potential communication. These methods involve train- communication mechanisms, ranking them in ing bees to an artificial feeder filled with their classically hypothesized order of evolu- sucrose solution (the experimental feeder) at a tionary complexity: contact and visually- particular location. Once trained, these bees based communication, followed by olfactory are known as experienced foragers (Biesmeijer communication, and finishing with a discus- and de Vries, 2001). The investigator then sion of acoustic communication (von Frisch, individually marks a fixed number of trained 1967; Kerr, 1969; Wille, 1983). Although the foragers that are regularly censused to insure phylogeny of the Meliponini remains largely that a constant number of bees visit the feeder. unresolved, the classic ordering gives us a All newcomers (unmarked bees) are then starting point to pursue evolutionary ques- counted and captured or immediately marked. tions. These newcomers are presumptive recruits In the literature, multiple synonyms often that must be shown (1) to come from the col- exist for each species name and there can be ony under study (generally by verifying their disagreement on the appropriate synonym return to the colony) and (2) to have had a high (Roubik, 1989; Michener, 1990; Inoue et al., probability of arriving based upon information 1999; Michener, 2000). I have therefore provided in the nest by the experienced forag- followed the assignments of Michener (2000) ers. Both points can be demonstrated if new- and refer the reader to the original source comers from the colony under study arrive literature for more detailed information to only when trained foragers feed and do not determine the appropriate nomenclature. arrive when these trained foragers are cap- tured. If these criteria are fulfilled, then new- comer recruitment has occurred. 2. METHODOLOGY In order to test whether bees can recruit to a First, it is useful to review the methods gen- specific location, at least two identical feeders, erally used. These methods can be deceptively one control and one experimental, are placed Recruitment communication in stingless bees 161 in an array in which only one spatial dimen- array distances are often the result of working sion is altered at a time. Foragers are trained to in environments where, even in times of the experimental feeder, but not to the control dearth, there are sufficient natural food sources feeder. For example, to test directional com- such that foragers cannot be trained to greater munication, the experimental and control distances and still recruit. It is also possible feeders are placed at different directions equi- that foragers do not usually recruit for natural distant from the nest and at the same height food sources beyond a certain distance. This above the ground. If foragers can communi- remains to be determined. cate the tested dimension, then significantly With stingless bees, small colony sizes also more recruits should arrive at the experimental limit the sample size per experiment and the feeder than at the control feeder. In order to number of experiments that can be conducted. control for potential site bias, it is important to A typical M. quadrifasciata colony contains conduct trials with the training feeder at differ- 300–400 total bees, of which only a fraction are ent locations. foragers (Lindauer and Kerr, 1960). Location These methods are elegant, yet not neces- experiments with such small colonies rapidly sarily simple to execute. Highly social bees use up newcomers, turning them into experi- will only recruit under the right conditions of enced foragers. This occurs because a new- colony need and generally only to an artificial comer is defined as a forager that has never pre- feeder if it provides a resource that is more dif- viously visited any feeder (von Frisch, 1967). ficult to obtain in the natural environment (von Such a rigorous definition is necessary because Frisch, 1967; Roubik, 1989; Seeley, 1995). bees are highly flower constant (Seeley, 1995; Thus researchers usually work during periods Slaa and Biesmeijer, in press) and can search of natural food dearth in order to train bees to on their own for food sources, particularly if artificial feeders (Lindauer and Kerr, 1960; they possess a search image acquired through Esch, 1967; Nieh and Roubik, 1995; Jarau prior experience and search as reactivated for- et al., 2000). agers (von Frisch, 1967; Biesmeijer and de Vries, 2001). In location experiments, counting foragers that have previously experienced the 2.3. Experimental challenges feeder at a different place or time as newcomers can lead to misleading results, such as finding In general, the feeder array experiments no or weak communication of food location testing meliponine location communication when experienced foragers search on their own have used feeders placed less than 200 m from for displaced food sources (Nieh et al., 2003b). the nest (Tab. I), whereas the Meliponini prob- If significantly more reactivated foragers con- ably forage at much greater distances for natu- tinue to arrive at the experimental feeder, even ral food sources. Roubik and Aluja (1983), after the feeder positions are changed, it is pos- reported flight ranges of approximately 1.7– sible to conclude that location information is 2.1 km from capture and release studies of being transferred to reactivated foragers. How- stingless bees (Cephalotrigona capitata and ever, if reactivated foragers arrive in equal Melipona panamica) living in a tropical forest. numbers at experimental and control feeders, Van Nieuwstadt and Ruano (1996) discovered one cannot conclude that there is no communi- a positive correlation between bee size (head cation of location because such information width) and foraging distance: 600–900 m for may be present and used by newcomers, but not Trigona corvina, Partamona aff cupira Smithi, by reactivated foragers. The effect of melipo- Trigona (Tetragonisca) angustula, and Nan- nine location communication on reactivated notrigona testaceicornis, as estimated from foragers has yet to be explicitly studied. capture and release studies. Thus it is desirable Typically, honeybee researchers distinguish to test location communication at greater dis- between newcomers and reactivated foragers tances. However, such relatively short feeder- by capturing and killing all newcomers once they become experienced foragers (von Frisch, i P. cupira does not occur in Costa Rica, and thus the 1967; Seeley, 1995). However, this solution species may be either P. bilineata or P. orizabaensis would rapidly destroy a small colony of sting- (Roubik, personal communication). less bees, and thus the usual practice is to mark 162 J.C. Nieh † † † † 1998 shes indicate no Jarau et al. 2000 Jarau et al. 2000 Jarau et al. 2000 Jarau et al. 2000 periments with each periments with Lindauer & Kerr 1960Lindauer & Kerr Lindauer & Kerr 1960Lindauer & Kerr Lindauer & Kerr 1960Lindauer & Kerr Lindauer & Kerr 1960Lindauer & Kerr which bees were trained is given eeder was used. Da used. was eeder tion, and and height ex tion, yes* 12 m high 0 15 Nieh et al. 2003b r all distance, direc = 0.0001). The maximum distance to maximum The = 0.0001). P 30 m 33 94 ? — — — raining Training ≤ = 0.001, ** = 0.001, P s in stingless bees. Results are pooled fo tailed Chi-square test (* test Chi-square tailed 30 m 33 107 yes** 140 m 60 203 40 m 63 118 yes** ≤ ≤ ovide a sense of scale. Multiple control values indicate that more than one control f control one than more that indicate values control Multiple scale. of a sense ovide ? — — — yes** 36 m 14 53 ? — — — 1969 Kerr nono 150 m 11, 211 100 m 125 55 yes** 150 m 27 31, 40 no 144 160 m ? 20 — 27 — ? — Nieh et al. 2000 — — —1958, Kerr & Lindauer no 200 m 79 17 no 150 m 21 25 ? — — —1958, Kerr & Lindauer no 50–70 m 133no 120 50 m 78 45 no 50–140 m 73 65 DISTANCE DIRECTION HEIGHT yes* 300 m 0, 2 19 yes** 300 m 3, 3 78 ? — — —1958, Kerr & Lindauer yes** yes** yes** 175 m 12 103 yes** 175 m 14 180 yes** 40 m high 26 252 Nieh & Roubik 1995, yes** 50 myes** 0–2.5% 97.5–100% 146 myes** yes** 100 m 14yes** 50 m 150 m 0 0–1.1% 98.9–100% 61 6 ? 22 yes** 117 146 m yes** — yes** 100 m 26 150 m — 0 — 2 92 Schmidt et al. 2003 34 42 ? no yes** — 20 m high 12 m high 0 44 — 41 801958, Kerr & Lindauer — Nieh et al. 2003b Nieh et al. 2003a yes** 100 m 3 45 yes** 100 m 3 49 no 0 m high 7 9 Nieh et al. 2003b †† -values are obtained from a 2- are obtained from -values P M. scutellaris T. carbonaria T. Plebeia droryana hel- Partamona leri S. depilis M. quadrifasciata spinipes T. M. panamica T. hyalinata T. mandacaia M. S. postica M. bicolor M. Summary of location communication experiment specificity Locationcommunication SpeciesNon-specific Some specific? feeder #control #exp specific? feeder #control Training #exp specific? feeder #control#exp Paper(s) T Table I. respective species. species. respective 3-D specificity † See text. †† Sample sizes given. † See text. not data available. for each type of experiment in order to pr in order of experiment each type for Recruitment communication in stingless bees 163 and release the newcomers and to not count blebees (Dornhaus and Chittka, 2001). Such marked bees as newcomers (Nieh and Roubik, searching is not location-specific, but is not 1995). necessarily random, because foragers may use a specific search pattern (Wehner, 1992) and Thus a limited time window coupled with small available sample sizes creates challenges because the searches likely involve some level for studying recruitment communication in of olfactory orientation (Dyer, 2002; Wenner, some stingless bee species. In addition, some 2002). species are difficult to work with, being so In honeybees, returning foragers perform a aggressive that they resist resettlement and round dance for food sources close to the nest, generally die in observation nests (T. spinipes, and this round dance leads recruited nestmates Almeida and Laroca, 1988); are difficult to to search in all directions (von Frisch, 1967). train, preferring to attack feeders placed near Only the waggle dance, performed for food the nest rather than feed at them (personal sources at greater distances (typically >100 m observations); are highly sensitive to distur- for Apis mellifera carnica), leads foragers to bances (Sakagami, 1966; Roubik, 1989; Kolmes search at a particular distance and direction and Sommeijer, 1992); or are susceptible (von Frisch, 1967; Gould, 1976; Dyer, 2002). to stress caused by artificial manipulations In bumblebees, it was known that foragers of (Nogueira-Neto, 1997, 1999). some species produce excitatory movements upon returning from a rich nectar source (Wagner, 1907; Brian, 1954; Jacobs-Jessen, 1959), and recent studies show that this 3. RECRUITMENT behavior results in an increase in foraging TO NON-SPECIFIC LOCATIONS activity at non-specific locations (Chittka and Dornhaus, 1999; Dornhaus and Chittka, 1999, The classic hypothesis, in keeping with the 2001, 2004; Dornhaus and Cameron, 2003). theory of ritualization (Tinbergen, 1952; Zahavi, In comparison to the number of meliponine 1980; Krebs and Dawkins, 1984; Bradbury and species cited as communicating resource loca- Vehrencamp, 1988), posits that bee recruit- tion, there are surprisingly few species for ment communication evolved from excitatory which even preliminary feeder array data are behaviors that do not communicate resource published (Tab. I). I therefore focus on species location into communicatory behaviors that for which published data are available. indicate resource location (Kerr, 1960; Esch Recruitment to non-specific locations has been et al., 1965; von Frisch, 1967; Kerr, 1969; reported in Nannotrigona testaceicornis, Michener, 1974). Meliponula (Axestotrigona) ferruginea tesco- Although the communication of resource rum, Frieseomelitta silvestrii, F. flavicornis, location can be beneficial, allowing social and F. freiremaiai (Kerr, 1969) and shown insects to rapidly exploit food sources through recruitment experiments with Plebeia (Hölldobler and Wilson, 1990; Seeley, 1995), droryana (Lindauer and Kerr, 1958). Lindauer these benefits depend upon several factors such and Kerr (1958) performed preliminary exper- as colony need, the spatial and temporal distri- iments with M. scutellaris and their data also bution of resources, and whether the resource suggest no communication of distance or direc- is amenable to exploitation through mass tion in this species (Tab. I). Such non-specific recruitment (Roubik, 1982b; Seeley, 1995; Breed location recruitment may be an ancestral state et al., 2002; Dornhaus, 2002, 2004; Sherman in the evolution of recruitment location com- and Visscher, 2002; Jarau et al., 2003; Slaa, munication (Lindauer and Kerr, 1958; Kerr, 2003; Biesmeijer and Slaa, 2004). Thus loca- 1969). Overall, data on species demonstrating tion communication may not always be advan- a complete inability to communicate distance tageous (von Frisch, 1967; Dyer, 2002; and direction are quite limited (Tab. I). It is Dornhaus and Chittka, 2004), and searching for unclear if this paucity of data reflects a rare advertised resources at non-specific locations strategy or if researchers have selectively appears to exist in all social bee groups (Kerr, focused on species with some ability to com- 1969), including the primitively eusocial bum- municate distance or direction. 164 J.C. Nieh

4. SPECIFIC LOCATION height. In paired feeder experiments, M. COMMUNICATION panamica shows an excellent ability to com- municate height, recruiting nestmates to the The communication of resource location correct height, even for height differences as may represent a more derived state in the evo- small as 10 m (Nieh and Roubik, 1995). lution of bee recruitment communication sys- Melipona bicolor only recruited nestmates to tems (Kerr, 1960; Esch et al., 1965; von the correct height when the feeder was at can- Frisch, 1967; Kerr, 1969; Michener, 1974). opy height (12 m high); and M. mandacaia Table I shows that the communication of could not recruit to the correct height, regard- resource location is widespread among the less of whether the feeder was 12 m high or on stingless bees, but can occur to different the ground (Nieh et al., 2003b). Food resources degrees of specificity, even to the extent of can be dispersed at different canopy levels, and indicating height (which honeybees do not thus the ability to specify height could be communicate, Dyer, 2002). It is unclear advantageous, particularly in competitive envi- whether increasing levels of location specifi- ronments with tall canopies (Roubik, 1993; city or accuracy reflect an evolutionary trend, Nagamitsu and Inoue, 1997; Roubik et al., because location specificity may also be tai- 1999; Eltz et al., 2001). lored to the habitat and niche occupied by It is unclear whether there is variation in the different species over evolutionary time ability to communicate distance and direction (Nagamitsu and Inoue, 1997; Eltz et al., 2001, among the Melipona. Jarau et al. (2000) 2002). reported that M. quadrifasciata communicated For example, T. carbonaria is thought to be direction and distance only when the food a relatively basal species (on the basis of behav- source was respectively ≤30 m and ≤40 m ioral traits), and foragers can guide nestmates from the nest (Tab. I). Melipona scutellaris to the correct direction, but not the correct dis- communicated direction up to 140 m from the tance (Nieh et al., 2000). In contrast, Scaptot- nest (the maximum distance tested), but com- rigona postica has a good ability to communi- municated distance only up to 30 m from nest. cate three-dimensional location and can recruit In these experiments, both species appeared to nestmates to a feeder placed 20 m above the search initially at random locations before ground (Lindauer and Kerr, 1958). Within showing some ability to communicate specific the Melipona, there is evidently variation in the location (Jarau et al., 2000). These results are ability of different species to communicate surprising given that precise distance commu- three-dimensional food location (Tab. I). How- nication should be more important for more ever the phylogenetic relationship between distant food sources than for nearby food these three groups remains unclear (Wille, sources (Weidenmuller and Seeley, 1999; 1979, 1983; Michener, 1990; Camargo and Dyer, 2002). Because experienced foragers Pedro, 1992; Fernandes-Salomao et al., 2002; may have been counted as newcomers (sample Costa et al., 2003). sizes exceeded the total number of foragers in Detailed studies of habitat-specific varia- the colonies used, Jarau et al., 2000), the inter- tions in location communication have only pretation of these results is unclear (Nieh et al., been conducted with the Melipona. Melipona 2003b). panamica lives in tropical forests with high canopies (up to 40 m high, Croat, 1978), whereas M. bicolor lives in an Atlantic Rain- 5. MECHANISMS OF INFORMATION forest habitat with intermediate canopy heights TRANSFER (approximately 15 m, Wilms et al., 1997), and M. mandacaia (not to be confused with M. 5.1. Contact and visual communication mandaçaia, Nogueira-Neto, personal commu- nication) lives in an arid Caatinga habitat where 5.1.1. Contact food sources are seldom more than a 5–6 meters above the ground (Martins and Aguilar, 1992; Trophallactic contact (food exchange) may Rizzini, 1997). These species possess a corre- be the most primitive form of meliponine com- spondingly graded ability to communicate munication (Hart and Ratnieks, 2002) and Recruitment communication in stingless bees 165 could provide information about food quality of the leader is hypothesized to be the primary and odor, as it does in honeybees (De Marco orientation mechanism (Esch, 1967; Kerr, and Farina, 2003). For example, Hrncir et al. 1969). Esch et al. (1965) reported observing (2000) showed that the number of trophallac- M. quadrifasciata foragers perform a zigzag tic (food exchange) contacts between M. quad- piloting flight that pointed in the direction of rifasciata and M. scutellaris recruiters and the feeder, but which the nestmates did not fol- potential recruits increases before the recruit low beyond 30–50 m (Esch, 1967). Kerr successfully reaches the feeder. More such (1969, 1997) proposed that the whitish reflec- experiments should be conducted to determine tive abdominal hairs of some species may the amount and type of information that forag- facilitate orientation to experienced foragers ers need to receive before they arrive at the as they leave the nest. Melipona panamica for- advertised resource. agers appear to communicate direction outside Studies of meliponine nest recruitment the nest because separating experienced forag- behavior report jostling (bees appearing to ers from potential recruits as they exit the nest purposefully run and bump into nestmates), resulted in equal numbers of newcomers zigzag running (changing direction several arriving at the control and experimental feed- times while running), and spinning (clockwise ers, located in opposite direction from the and counterclockwise body rotations) by nest (Nieh and Roubik, 1998). However zig- recruiting foragers (Lindauer and Kerr, 1958; zag piloting flights have not been observed Kerr, 1960; de Bruijn and Sommeijer, 1997; in M. panamica (Nieh and Roubik, 1998), Nieh, 1998a; Hrncir et al., 2000). In honey- M. scutellaris (Hrncir et al., 2000), or of bees, physical contact, particularly antennal M. quadrifasciata (Hrncir et al., 2000). contact, may be important in the transmission of waggle dance information (Rohrseitz and 5.2. Olfactory communication Tautz, 1999). In stingless bees, body contacts may help to transfer relevant odors and com- Several odor sources can influence forager municate excitement about the relative quality orientation in social bees: food odor, olfactory of the food source. Hrncir et al. (2000) showed signals, cues deposited by foragers, and locale that the number of bees visiting a feeder is odors—odors of the environment surrounding highly correlated with the mean number of jos- the food source (von Frisch, 1967; Goulson tles performed by recruiting M. quadrifasciata et al., 2000; Goulson et al., 2001; Dyer, 2002; and M. scutellaris foragers inside the nest. In Nieh et al., 2003b). To date, no studies have addition, antennal contacts may help transfer specifically isolated and examined the effect acoustic information to recruits (Nieh, 1998a; of locale odors on stingless bee foraging. Nieh and Roubik, 1998). Melipona panamica Whenever evidence for attractive odor nestmates hold their antennae closely around marking of food sources has been sought, it and sometimes touching the vibrating wings has been found in the highly social bees (Kerr and body of a recruiting forager (Nieh, 1998a). et al., 1963; von Frisch, 1967; Kerr, 1972; Nieh, 1998b; Aguilar and Sommeijer, 2001; 5.1.2. Visual Stout and Goulson, 2001; Jarau et al., 2002; Nieh et al., 2003d; Schmidt et al., 2003; Hrncir Visual communication plays a role in et al., 2004). However, stingless bees are the recruitment orientation outside the nest. Slaa only bees known to produce long odor trails et al. (2003) have shown that foragers of Trig- beginning near the nest and extending to the ona amalthea, Tetragonisca angustula, Parta- ii food source (reaching a reported 900 m in mona cupira , and Oxytrigona mellicolor dis- T. amalthea, Kerr, 1960). Although odor trails play local enhancement and orient towards the usually lead to nectar or pollen resources, obli- visual presence of nestmates on a food source. gate necrophages such as T. necrophaga Foragers of several species may also guide reportedly use odor trails to indicate their food experienced foragers for at least part of the sources (Roubik, 1982a). Bumblebees also distance to the food source, and visual tracking produce odor trails, but these are limited to the close vicinity of the nest (Cameron and ii Ibid. Whitfield, 1996; Cameron et al., 1999). Within 166 J.C. Nieh the stingless bees, there is a gradation of strat- with leaves was stretched across the pond. egies, ranging from complete odor trails that Foragers began to odor-mark the leaves and begin in the vicinity of the nest and extend to recruited. Kerr et al. (1963) performed a simi- the food source, short odor trails that only lar set of experiments on a grass field with extend a short distance away from the food S. bipunctata and T. spinipes, using wires source in the direction of the nest (Nieh et al., attached to the tops of bamboo poles. 2003a), and odor-marking of the food source alone (Nieh, 1998b; Hrncir et al., 2004). Odor 5.2.2. Short odor trails trail marking is reported in many meliponine species (Lindauer and Kerr, 1958, 1960; Kerr Kerr and Rocha (1988) hypothesized that and Cruz, 1961; Kerr et al., 1963; Kerr and M. rufiventris and M. compressipes could Esch, 1965; Kerr, 1960, 1969, 1972, 1973, deposit a short odor trail consisting of one to 1994; Blum et al., 1970; Kerr et al., 1981; Kerr three putative “urine” marks (anal droplets) and Rocha, 1988), but I will focus on those extending up to 8 m from the food source species for which there are data showing that (Kerr, 1994). However, no experiments were the putative odor marks attract nestmates. conducted to test the attractiveness of these putative marks. Recently, Nieh et al. (2003a) 5.2.1. Complete odor trails reported that T. hyalinata foragers produced short odor trails, that extended up to 27 m from Lindauer and Kerr (1958) discovered that the food source towards a nest 146 m away. stingless bees could produce an odor trail con- Displacing a rope odor-marked by T. hyalinata sisting of odor droplets deposited periodically foragers resulted in a significant increase in on vegetation between a food source and the the number of newcomers visiting a control nest. The distance between odor marks ranges feeder, as did displacing the odor-marked from 1 m to 8 m in different species (Kerr, feeder alone. 1969). In all described cases, such odor mark- ing occurs after the forager has finished feed- 5.2.3. Odor trail polarization ing and as she returns back to the nest (Lindauer and Kerr, 1958; Kerr, 1969, 1972, Lindauer and Kerr (1958) first hypothe- 1973; Kerr et al., 1981; Nieh et al., 2003a). sized that stingless bee odor trails might Foragers prefer to land on prominent vegeta- increase in concentration at the food source to tion, and thus odor trails can be studied on indicate the exact food location. Kerr et al. ropes or wires hung with leaves and elevated (1963) introduced the term “polarity” to above the ground (Lindauer and Kerr, 1958; describe this effect, and experiments show that Kerr et al., 1963). The literature on odor trails Scaptotrigona postica (Kerr et al., 1963) and has focused on complete odor trails that begin T. hyalinata (Nieh et al., 2003a) newcomers 7.5 m (Lindauer and Kerr, 1958) to 35 m (Kerr will ignore a feeder placed within the odor et al., 1981) from the nest and extend to the trail, between the nest and the training feeder (Tab. I). Thus the newcomers have some food source. However, little is known about means of determining the correct endpoint. the detailed spatial structure of these trails Recently, Schmidt et al. (2003), showed that (Lindauer and Kerr, 1958, 1960). Scaptotrigona depilis is able to find the odor- Lindauer and Kerr (1958) provide the only marked feeder, even when it is displaced away detailed descriptions of complete odor trails. from the putative odor trail or into the putative They followed and recorded the spatial pattern odor trail. The ability of newcomers to reach of odor marks left by S. postica, recording a the correct endpoint may be due to differences fairly even spacing of odor marks with in the concentration of odor marks deposited increasing distance from the feeder (see on the target (Lindauer and Kerr, 1960). Figs. 10–12 in Lindauer and Kerr, 1958). They also performed a series of ingenious experi- 5.2.4. Target-only odor marking ments showing that these odor marks were necessary for recruitment. Foragers were Target-only odor marking, defined as odor- trained to the opposite side of a pond, but did marking of the food source alone, is wide- not recruit over the open water until a rope spread among the social bees (Kerr et al., Recruitment communication in stingless bees 167

1963; Stout and Goulson, 2001). However, under a dense forest canopy. To date, no because some meliponine species also use studies have examined this hypothesis. odor trails, one must show that meliponine for- agers mark the target, but do not deposit an 5.2.6. Odor mark sources odor trail, to demonstrate target-only odor marking. To do this, one may observe that for- Meliponine odor marks have several agers do not land to deposit a putative odor glandular sources (Cruz-Landim et al., 1998) trail or show that foragers do not need or use and may also consist of cues such as anal an odor trail to recruit across a water gap, droplets. Using feeder choice experiments, because odors cannot be effectively deposited Jarau et al. (2002) and Hrncir et al. (2004) or retain their spatial structure on water have shown that tarsal gland extracts from M. (Lindauer and Kerr, 1958; Nieh and Roubik, seminigra are attractive to foragers. Nieh et al. 1995). The latter form of evidence is more (2003d) reported that M. mandacaia foragers conclusive. produce an attractive ventro-abdominal odor Although target-only odor marking may be in addition to other attractive odor sources, widespread among stingless bees, good evi- including an odor cue, anal droplets. dence has only been obtained for M. panam- Anal droplets are clear droplets of fluid ica. In this species, foragers odor-mark the excreted from the anus of foragers, usually food source (Nieh, 1998b), but do not deposit after foragers have finished feeding. They odor trails because they could recruit to the have been described in several species: M. rufiv- correct direction across a large water gap that entris and M. compressipes (Kerr and Rocha, prevented odor trail deposition (Nieh and 1988), M. favosa (Aguilar and Sommeijer, Roubik, 1995). On the basis of visual observa- 1996, 2001), M. panamica (Nieh, 1998b), and tions, Frieseomelitta silvestrii, F. flavicornis, M. mandacaia (Nieh et al., 2003d) and are F. freiremaiai (Kerr, 1969), M. panamica generally produced in greater number at more (Nieh and Roubik, 1995), M. scutellaris, M. dilute food sources for which foragers do not quadrifasciata (Hrncir et al., 2000), M. man- recruit (M. mandacaia) or which lead foragers dacaia, and M. bicolor (Nieh et al., 2003b) to greatly reduce their rate of recruitment foragers have not been observed to land and (M. panamica). However, anal droplets can deposit odor trails. It is unclear whether M. attract nestmates in paired feeder assays (Aguilar and Sommeijer, 2001; Nieh et al., rufiventris and M. compressipes deposit short 2003d). These droplets may therefore be odor trails, because the attractiveness of the excreta (Nieh, 1998b) and serve as cues, not putative marks was not tested (Kerr and signals (Nieh et al., 2003d). Further work may Rocha, 1988; Kerr, 1994). reveal that meliponine food-marking odors Target-only odor marking may account for generally consist of multiple odor sources, as the ability of M. bicolor foragers to specifi- has been shown in honeybees (Free et al., cally recruit nestmates to the top of a 12 m 1982; Free and Williams, 1983). high tower, but not to its base, because of dif- Lindauer and Kerr (1958) reported that ferences in the active space of odor marks mandibular glands extracts could attract S. pos- caused by increased wind or other microcli- tica foragers to a food source and were used in mate differences (Bradbury and Vehrencamp, depositing odor trails as well as marking the 1988) between the base and the top of the feeder. Subsequent studies found a similar tower (Roubik, 1989; Nieh et al., 2003b). attractive role for mandibular gland secretions in T. spinipes (Kerr et al., 1981) and T. hyalinata 5.2.5. Aerial odor trails (Nieh et al., 2003a) odor trails. However, a recent study by Jarau et al. (2004) found attrac- Kerr (1969, 1994) hypothesized that tion to labial gland excretions deposited at food foragers of some species such as Partamona sources, not to mandibular gland excretions. helleri deposit aerial odor trails, creating an Thus the role of mandibular gland secretions in “odor tunnel” as they fly to the food source. odor marking should be more rigorously eval- Kerr (1994) proposed that this mechanism uated. The composition of mandibular gland could function during windless conditions secretions has been examined in species such 168 J.C. Nieh as T. silvestriana (Johnson et al., 1985), Scap- strate (Michelsen et al., 1986a; Sandeman totrigona tubiba and S. postica, and is com- et al., 1996; Visscher et al., 1999; Nieh and plex, consisting of over 25 compounds that Tautz, 2000). exert different behavioral effects (Kerr, 1969) There is evidence from preliminary play- and increase in diversity with worker age in back experiments that meliponine recruitment Meliplebeia denoiti (de Korte et al., 1998) and sounds play a role in activating nestmates that S. postica (Cruz-Landim and Ferreira, 1968). have previously experienced the feeder (reac- A major component is 2-heptanol (Smith and tivated foragers) to go out and forage (Lindauer Roubik, 1983), which, in a synthetic prepara- and Kerr, 1958; Esch, 1967). In addition, cor- tion, attracted T. spinipes foragers to a feeder relations have been found between recruitment (Kerr et al., 1981). Citral is also present in the sounds and the quality and location of the food mandibular gland secretions of many species source (Esch et al., 1965; Nieh and Roubik, (Kerr, 1969), and a synthetic preparation of 1998; Aguilar and Briceño, 2002; Hrncir et al., citral can attract T. subterranea foragers to 2002; Nieh et al., 2003c). feed (Blum et al., 1970). In the robber stingless bee, Lestrimelitta, high concentrations of citral released during attacks may assist in the orien- 5.3.1. Sound production tation of Lestrimelitta recruits towards the vic- To date, successful foragers of 18 melipo- timized nest (Wittman et al., 1990; Sakagami nine species have been observed to produce et al., 1993). pulsed sounds upon returning to the nest: Frie- seomelitta silvestrii, F. flavicornis, F. freire- 5.2.7. Odor-mark deposition maiai, Leurotrigona muelleri, M. bicolor, M. costaricensis, M. mandacaia, M. seminigra Stingless bees can deposit odors in several merrillae, M. panamica, M. scutellaris, M. ways. Mandibular gland secretions may be seminigra, Nannotrigona testaceicornis, Ple- deposited by rubbing the substrate with the beia droryana, Cephalotrigona capitata, S. mandibles (Kerr, 1969), and tarsal gland postica, Tetragonisca angustula, Trigona secretions by walking on the substrate (Jarau (Hetetrotrigona) carbonaria, and Meliponula et al., 2002). When depositing odor marks on (Axestotrigona) ferruginea tescorum (Lindauer surrounding foliage, foragers generally land and Kerr, 1958; Esch, 1967; Kerr, 1969, 1994; briefly (average of 1.4 ± 0.7 s for T. hyalinata, Nieh and Roubik, 1998; Hrncir et al., 2000; Nieh et al., 2003a) and rub their mandibles Aguilar and Briceño, 2002; Hrncir et al., 2002; (Lindauer and Kerr, 1958; Kerr, 1972; Kerr Nieh et al., 2003c). These sounds have air- and and Rocha, 1988) and tongue (T. hyalinata, substrate-borne components (Esch, 1967) and Nieh et al., 2003a) against the substrate. may be produced as foragers vibrate their Foragers can also groom when depositing odor wings and thoracic muscles, much as honey- marks, but it is unclear if grooming deposits bees produce sound during the waggle dance odors (Kerr, 1994; Nieh et al., 2003d). (Michelsen, 2003). No detailed physiological investigations of sound production mecha- 5.3. Acoustic recruitment behavior nisms have yet been conducted with stingless bees. Sound production is an intriguing aspect of In Meliponula (Axestotrigona) ferruginea meliponine recruitment behavior. Although tescorum and Nannotrigona testaceicornis, studies document the existence of recruitment Kerr (1969, p. 141) reported a form of acoustic sounds in several species, these sounds have chorusing, “in these species, when a worker, been quantitatively studied in relatively few. having discovered a source of food, enters the Bees do not possess tympana, and thus do not hive, it produces a characteristic sound. hear the pressure component of sound Immediately thereafter the bees closest to it (Michelsen et al., 1986b). However, the begin to produce the same sounds, and then Meliponini may share the ability of honeybees others join in, and within less than 1 min all the to detect the particle-velocity component of hive will be buzzing, and all available bees sound (Kirchner, 1994; Michelsen, 2003) and leave the hive looking for food”. This behavior sound vibrations transmitted through the sub- has not been investigated in greater detail. Recruitment communication in stingless bees 169

Kerr et al. (1963) also reported that M. 2003c). In M. panamica, Nieh and Roubik, quadrifasciata foragers produce sounds on (1998) found correlations between the height food sources. In M. panamica, foragers pro- above ground and the distance to a rich (2.5 M) duce sounds upon departing from a feeder, and food source and the duration of sound pulses the rate of sound production is positively cor- during two phases of the recruitment perform- related with sucrose concentration, suggesting ance: respectively, the food-unloading phase that feeder sounds may be related to forager and the dance phase. During the food-unload- perceptions of food quality (Nieh, 1998b). ing phase, when returning foragers transfer However, the significance of feeder sounds their gathered nectar to nestmates, sound remains a mystery. pulses were significantly longer if the food source was on the ground than when it was 5.3.2. Effect of food location 40 m above the ground in the forest canopy. During the dance phase, when recruiting for- Recently, Hrncir et al. (2003) have shown agers run rapidly and can spin clockwise and that M. seminigra foragers, like honeybees counterclockwise inside the nest, the sound (Srinivasan et al., 2000; Esch et al., 2001), can pulses increased in duration with increasing use optic flow to measure the distance to a distance to the food source. food source. Esch et al. (1965) discovered that Hrncir et al. (2003) state that the term some stingless bees (M. seminigra merrillae “dance” is inappropriate as applied to these and M. quadrifasciata) may encode the dis- spinning motions because they do not commu- tance to the food source in the duration of nicate distance or direction. However, the term recruitment sound pulses. Foragers of both “dance” is also applied to forms of honeybee species increased the duration of sound pulses behavior related to foraging that do not com- as distance to a food source increased. This municate distance or direction, such as the discovery created great excitement because tremble dance (Seeley, 1992; Nieh, 1993; honeybees were the only animals previously Seeley et al., 1996) and round dance (von known to referentially encode food distance Frisch, 1967; Waddington, 1982; Waddington (von Frisch, 1967). The discovery of a melipo- and Kirchner, 1992; Kirchner et al., 1998). nine distance-encoding communication sys- Moreover, correlations are found between dis- tem would provide a new animal model of tance and sounds produced during this spinning functionally referential communication and phase in M. panamica (Nieh and Roubik, perhaps yield insights into the evolution of the 1998). Thus the communication of distance waggle dance in the closely related honeybees may occur during the dance phase of M. (Esch et al., 1965; Esch, 1967). Some 80 years panamica, as it does in the honeybee waggle after von Frisch’s (1923) discovery of the dance (von Frisch, 1967). waggle dance, the ability of honeybee foragers Nieh and Roubik (1998) also performed to communicate distance and direction with experiments testing the ability of newcomers to the waggle dance has largely been accepted, find the food source at the correct direction, dis- following numerous tests (Gould, 1976; tance, and height based upon information trans- Michelsen et al., 1989; Esch et al., 2001; mitted inside the nest (see above). These exper- Sherman and Visscher, 2002; Dornhaus and iments prevented newcomers from following Chittka, 2004). However, the ability of any recruiters, even for a short distance, to the food Melipona species to communicate distance source and, in the distance and height experi- through sounds has not been conclusively ments, removed odor-marks deposited by demonstrated. foragers on the food sources. When such poten- The evidence for the acoustic communica- tial sources of information were excluded, tion of distance is primarily based upon corre- M. panamica newcomers did not arrive at the lations between sound pulse duration and food correct direction (equal numbers arrived in distance in different species: M. seminigra opposite directions), but significantly more did merrillae, M. quadrifasciata (Esch, 1967), M. arrive at the correct distance and height. These panamica (Nieh and Roubik, 1998), M. cos- experiments did not disturb locale odors or taricensis (Aguilar and Briceño, 2002), M. putative odor trails. Aguilar and Sommeijer mandacaia, and M. bicolor (Nieh et al., (2001) argued that these separation experiments 170 J.C. Nieh did not exclude the possibility of newcomers Waddington and Kirchner, 1992). There is using odor trails or locale odors, because such evidence that nestmates are sensitive to varia- odors cannot be removed. However, if new- tions in recruiter motivation in honeybees (von comers can use odor trails or locale odors alone, Frisch, 1967) and stingless bees (Nieh, 1998a, they should have arrived in the correct direction b). In M. panamica, recruitment performances during the separation treatment. They did not. for food of different qualities attract different One could object that the separation treat- numbers of potential recruits (Nieh, 1998a). ment was disruptive and prevented newcomers More potential recruits cluster with their from using odor trail and locale odor informa- antennae held closely around the vibrating tion. However, significantly more newcomers wings of sound-producing recruiters returning arrived at the correct distance and height in from a rich 2.5 M food source than from a separate experiments during the separation poorer 1.0 M food source (Nieh, 1998a) and treatment, even when forager-deposited odors recruitment rates are significantly higher for on the feeder were removed (Nieh and Roubik, richer food sources (Nieh, 1998b). 1998). Thus the existence of forager-deposited There are also questions concerning the odor trails does not consistently account ability of some Melipona species to encode for the results of the direction, distance, and distance in recruitment sounds. Hrncir et al. height separation experiments with M. panam- (2000) performed a detailed analysis of recruit- ica (Nieh, 1998a, b; Nieh and Roubik, 1998). ment sounds and found no evidence for dis- Furthermore, there is no evidence in M. tance encoding in M. quadrifasciata or in panamica for forager orientation over long M. scutellaris. It is quite possible that different distances to forager-deposited odors (Nieh and species possess different communication Roubik, 1998). Forager-deposited odor marks mechanisms. However, they were unable to on the feeder are only able to influence forager replicate the results of Esch et al. (1967) with orientation within 12 m of the feeder (Nieh, M. quadrifasciata. The smaller sample sizes 1998b). Thus distance and height information used by Hrncir et al. (2000) may partly account are evidently transmitted inside the nest, for differences between their results and those whereas direction information is evidently of Esch et al. (1967). It is also unclear if M. quadrifasciata foragers successfully recruited transmitted outside the nest. while Hrncir et al. (2000) recorded sounds, and It remains unclear if the information content thus it is uncertain if recruitment communica- of M. panamica recruitment sounds is suffi- tion was associated with the measured sounds cient to account for the specificity of location (Nieh et al., 2003c). communication. There are high variances in the sound pulse durations, particularly in the food unloading pulses whose durations may be cor- 5.3.3. Effect of forager motivation related with resource height (Nieh and Roubik, Recently, Hrncir et al. (2002) have shown 1998). Correlations between distance and other that stingless bee recruitment sounds can vary temporal parameters such as pulse duration with forager motivation. Melipona seminigra also exist in M. costaricensis, although the foragers reduced the amount of recruitment variances of the putative distance-encoding sound produced as food quality decreased by sounds are again quite high at each distance decreasing the pulse duration and increasing (Aguilar and Briceño, 2002). the interpulse interval (the time between two Such high variances may be due to the anal- sound pulses). Limited sucrose solution flow ysis of weak and strong performances by for- also led to shorter pulses and longer inter- agers with varying degrees of motivation. In pulses than unlimited flow, and pulse duration honeybees, variances in the strength of recruit- increased with increasing foraging duration ment performances have long been noted (von (Hrncir et al., 2002). Nieh et al. (2003c) Frisch, 1967) and detailed investigations have reported that food quality (sucrose concentra- shown that forager motivation affects parame- tion) has a similar effect on M. mandacaia and ters such as the return phase of the waggle M. bicolor recruitment sounds, decreasing the dance (Seeley et al., 2000) and the reversal pulse duration and increasing the interpulse rate in the round dance (Waddington, 1982; duration as sucrose concentration decreases. Recruitment communication in stingless bees 171

This raises the question of whether pulse inside the nest (Tautz and Rohrseitz, 1998) or duration is a reliable measure of distance, to odors or the putative piloting flights of for- given that it varies with both distance and food agers outside the nest. Thus greater study of quality. However, Nieh et al. (2003c) reported the different communication channels will that M. mandacaia and M. bicolor foragers did likely reveal more about the processes that not successfully recruit newcomers to sucrose guide colony foraging. concentrations sufficiently low to affect the recruitment sounds. Thus potential recruits are evidently able to discriminate between recruit- 6.1. Communication channels ment performances for low and high quality food sources, and may use the interpulse inter- 6.1.1. Contact and vision val to discriminate between sound pulses that reliably encode distance and those that do not Inside the nest, the function of recruiter (Nieh et al., 2003c). motions and contacts between recruiters and potential recruits remains little understood. However, the experiments of Hrncir et al. 6. CONCLUSION (2000) show that jostling behavior and contacts between the recruiter abdomen and the anten- nae of potential recruits are associated with Thus stingless bee foragers may use multi- successful recruitment. Thus further examina- ple sensory modalities (touch, vision, olfac- tion of intranidal contact behavior may be quite tion, and audition) to transfer information con- revealing. cerning the existence of a resource and, in some cases, to communicate its specific 3-dimen- The existence of visual following (piloting) sional location (Tab. I). These mechanisms of in the communication of resource location information transfer allow the colony to collec- remains unclear. Such piloting is thought to be tively and appropriately allocate foraging the ancestral state in the evolution of recruit- resources (Roubik, 1989; Biesmeijer et al., ment communication, and may be used by 1998, 1999a; Biesmeijer and Ermers, 1999; some extant species (Esch et al., 1965; Esch, Biesmeijer and Slaa, 2004). Stingless bees, like 1967; Kerr, 1969). The difficulty of following honeybees, act as a superorganism with decen- at high flight speeds or of maintaining visual tralized control of foraging based upon the deci- contact while bypassing obstructions or flying sions of individual foragers (Seeley, 1983, through dense vegetation may have contrib- 1985, 1989a, b; Biesmeijer and de Vries, 2001; uted to evolution of other strategies (Nieh, Biesmeijer and Slaa, 2004). Thus the transfer 1999), but the advantages and disadvantages of recruitment information should be reliable of following have not been experimentally and, inside the often noisy and crowded nest evaluated. (Nieh, 1999), multimodal communication may provide information redundancy (Hölldobler 6.1.2. Olfaction and Wilson, 1978, 1990) enabling foragers to reliably alert nestmates to the existence of a Little is known about the behavior and sen- good resource. sory physiology of how stingless bees detect In those species with specific location com- odors in the field (Stort and Moraes, 1997, munication, multimodal information transfer 1998). In some meliponine species, experi- may also enhance the ability of nestmates to enced foragers may guide recruits who then find a resource when noise degrades informa- use the odor trail as a backup source of infor- tion sources (Nieh, 1999). Different sensory mation (Lindauer and Kerr, 1958; Kerr et al., channels are affected by noise to different 1981). Trigona hyalinata recruits arrive in degrees and also have different characteristics large groups that may be guided by experi- of temporal persistence, coherence, and infor- enced foragers, with final orientation assisted mation content (Bradbury and Vehrencamp, by a short odor trail (Nieh et al., 2003a). A 1988). These can be exploited in attracting similar grouping effect has been observed potential recruits to recruitment performances in stingless bee (T. corvina: Roubik, personal 172 J.C. Nieh communication) and honeybee foraging tion. However, in some species, recruitment (Tautz and Sandeman, 2003). Odor trails are sounds have high temporal variances and particularly interesting because, unlike worker appear to encode both food quality (motiva- ants, bees fly to the resource and must detect tion) and food location (referential informa- the odor droplets from some distance away. tion). It is therefore unclear whether nestmates Thus far, studies have shown that some odor can reliably obtain location information from marks are attractive and some are repellant such sounds. Some of this variance may be due (Villa and Weiss, 1990; Goulson et al., 2001), to the inclusion of less motivated perform- that they are deposited in such a way (i.e. ances (that are ignored by potential recruits) mouthpart rubbing to deposit secretions of into the acoustic analysis (Nieh and Roubik, cephalic glands and walking to deposit tarsal 1998; Hrncir et al., 2000, 2002, 2003; Nieh gland secretions) that is suggestive of their et al., 2003c). However, high variances may source, and that extracts of cephalic glands can also be an inherent property of recruitment attract foragers. It would be desirable to chem- sounds produced by highly motivated forag- ically assay these odor marks to demonstrate ers. Given that the evolution of functionally that they actually contain compounds pro- referential communication is hypothesized to duced by cephalic glands and to study olfac- have occurred through the ritualization of tory sensilla (Stort and Moraes, 1997, 1998) in motivationally related behaviors, such signals greater detail. The active space and effective can possess both motivational and referential duration of the odor plumes created by odor components and be classified within a motiva- marks (both in odor trails and also in target- tional to referential continuum (Marler et al., only odor-marks) should also be described. In 1992). Current debate on the information con- this respect, a modification of lab techniques tent of Melipona recruitment sounds (Hrncir (Vázquez et al., 2003) to produce field electro- et al., 2002, 2003; Nieh et al., 2003c) can be antennograms may be informative. explored with this continuum concept, partic- ularly with respect to how the variance in tem- poral structure is affected by motivation and 6.1.3. Audition whether this variance prevents receivers from obtaining reliable information. For example, if One of the most important questions to be Melipona foragers only communicate motiva- answered is whether some stingless bee spe- tion, yet encode location information that is cies use functionally referential communica- not used by receivers, researchers will have tion, the ability to abstractly encode environ- uncovered a presumptive intermediate stage in mental information in signals understood by the evolution of functionally referential com- receivers (Marler et al., 1992; Blumstein, munication. Detailed studies of motivational 1999). This debate focuses on the ability of behaviors may thus reveal potential early certain Melipona species to transfer location stages in the evolution of functionally referen- information inside the nest (Esch et al., 1965; tial communication. Esch, 1967; Nieh and Roubik, 1998; Hrncir It also is unclear whether and how stingless et al., 2000, 2002; Nieh et al., 2003c). In bees hear air-borne sounds and vibrations, M. panamica, intranidal information is evi- although the preliminary experiments of dently necessary for recruits to find the indi- Lindauer and Kerr (1958) and Esch (1967) cated food source at the correct distance and suggest that recruitment sounds may play a height (Nieh, 1998b; Nieh and Roubik, 1998). role, and the experimental methods used to Locale odors were not sufficient to guide study honeybee audition provide guidance recruits to the correct direction (Nieh, 1998b; (Kirchner, 1994; Michelsen, 2003). The devel- Nieh and Roubik, 1998). Thus locale odor opment of a sound playback system based alone does not account for the ability of upon the work of Esch (1967) would enable us M. panamica recruits to orient to the correct to systematically investigate acoustic commu- three-dimensional resource location. nication. Such experiments have proven diffi- In some species, the temporal structure of cult in honeybees (Michelsen et al., 1989; pulsed recruitment sounds is correlated with Michelsen, 2003), but they may be somewhat distance and (in M. panamica) height informa- simpler with stingless bees, given that the Recruitment communication in stingless bees 173 motion of the recruiter does not appear to con- molecular data are providing new information tain any location information (Nieh, 1998a; and alternative topologies (Fernandes-Salomao Hrncir et al., 2000), and because sound alone et al., 2002; Costa et al., 2003). It is not within is evidently sufficient to excite experienced the scope of this review to consider all of these foragers to go to a feeder (Lindauer and Kerr, topologies in detail, but I hope to give a sense 1958; Esch et al., 1965; Esch, 1967). of the issues. On the basis of morphological and behavioral characters, Wille (1979, 1983) 6.2. Phylogeny and evolution proposed that the Melipona, in which some species encode food distance in pulsed sounds The goal of understanding the evolution of (Tab. I), were fairly derived within the recruitment communication in the Apidae Meliponini, although the relationship of this depends upon a good phylogeny of this group. genus with other groups such as Scaptotrigona Unfortunately, there is disagreement concern- and Trigona (that mainly appear to use odor ing the evolutionary relationships between trails) is unresolved. Michener (1990) pro- groups in the Apidae, and eusociality and posed a topology with Melipona as a basal recruitment communication may have evolved group and Scaptotrigona and Trigona as more independently in honeybees and stingless bees derived. Subsequently, Camargo and Pedro (Winston and Michener, 1977; Chavarria and (1992) placed the Melipona as a more derived Carpenter, 1994; Cameron and Mardulyn, group but basal to the Nannotrigona (which 2001; Cameron, 2003). However, even if reportedly does not communicate food loca- advanced eusociality has evolved twice, prim- tion, Kerr, 1969) and Scaptotrigona. itive sociality (that was lost in the communal Recently, Costa et al. (2003) have pub- Euglossini) may still have been the plesiomor- lished topologies based upon mitochondrial phic state for all four groups in the Apidae 16s rDNA sequences (550 bp fragment). In the (Bombini, Euglossini, Meliponini, and Apini, Scaptotrigona Michener, 2000). Recent work on the Bombini maximum likelihood analysis, Melipona suggests that the Apidae may have shared a is more derived than which is more primitively social ancestor that excited nest- derived than Nannotrigona (Costa et al., mates after discovering food (Dornhaus and 2003). However, a parsimony analysis of the Chittka, 2004). Upon returning from a rich food same molecular data provides yet another source, bumblebees can motivate nestmates to topology for these groups (Costa et al., 2003). forage through food-alert runs, bouts of run- The addition of more genes to such analyses ning through the nest and interacting with other may increase our resolution, and groups are bees (Dornhaus and Chittka, 1999, 2001; currently working on molecular phylogenies Dornhaus and Cameron, 2003). This food-alert of the Meliponini (Roubik, personal commu- behavior is similar to the excitatory zigzag and nication; Cameron, personal communication). jostling behaviors performed by foragers of Given the topological uncertainties, we are several meliponine species, including species not currently able to make detailed statements that do not communicate food location about the mappings of communication charac- (Lindauer and Kerr, 1958; Nieh, 1998a; Hrncir ters. Nonetheless, the data continue to support et al., 2000). Thus the communication of food the hypotheses of Lindauer and Kerr (1958), quality through a graded series of excitatory von Frisch (1967), Kerr (1969), Esch (1967), behaviors (movements and sounds) may be ple- and Wille (1983) that the ancestor to the sting- siomorphic to the four tribes in the Apidae but less bees (and possibly of the Apidae) made was lost in the solitary Euglossini. Nieh et al. excitatory movements at the nest after return- (2003c) have termed this the “excitable ances- ing from a good food source, and possessed tor” hypothesis and predict that decreasing some form of odor marking (Kerr and Esch, food quality will increase the duration between 1965; Kerr, 1969). Meliponine species that food-alert runs in bumblebees and stingless deposit mandibular gland secretions as food- bees, even in stingless bee species that do not marking odors also use the same secretions as communicate food location. alarm pheromones (Kerr, 1969; Smith and Within the stingless bees, a well-resolved Roubik, 1983; Johnson et al., 1985). Mandib- phylogeny remains an elusive goal, although ular gland pheromones are involved in alarm 174 J.C. Nieh and defense in a wide range of bee groups, 1997) or the spatial distribution and clumping including honeybees (Collins et al., 1989; of food (Johnson, 1981) in different distances Brockmann et al., 1998; Winston and Slessor, and directions could play a role in the ability of 1998) and the solitary Colletidae and Halicti- different species to recruit nestmates to a set of dae (Duffield et al., 1984), and thus mandibu- dimensions, either through shaping communi- lar gland odor-marking to indicate food cation mechanisms or the search behavior of resources may be apomorphic, evolving from recruits (see work on honeybees, Roubik et al., a plesiomorphic alarm or defensive pherom- 1999; Dornhaus and Chittka, in press). The one widely found in bees. However, the recent role of habitat has yet to be considered in work of Jarau et al. (2004) suggests that rigor- detail, but the diversity of meliponine species ous verification of the attractive properties of and habitats suggests that this may be a prom- mandibular gland pheromone are necessary ising line of inquiry. before this scenario can be further explored. Competition for resources may also have Within the Meliponini, excitatory sound imposed an important selective pressure on production that does not encode food location stingless bee recruitment (Johnson, 1974; may be a plesiomorphic trait (Lindauer and Johnson and Hubbell, 1974; Johnson and Kerr, 1958; Esch, 1967; von Frisch, 1967; Kerr, Hubbell, 1987; Nagamitsu and Inoue, 1997; 1969; Wille, 1983). Whether such sound pro- Slaa et al., 1997; Biesmeijer et al., 1999b). duction is plesiomorphic to Meliponini and Given that foragers orient towards and are Apini is unclear. It may have evolved sepa- highly attuned to odor marks, there is an rately in both groups (Dyer, 2002). To further intriguing possibility that some species may explore these evolutionary issues, it would be use olfactory eavesdropping to exploit good good to understand more about the precise resources visited by other colonies. Lindauer mechanisms of communication and the detailed and Kerr (1958) trained a colony of S. postica phylogeny of genera such as the Melipona, in and a colony of S. xanthotricha to feeders which species form a clear taxonomic group placed such that the odor trails of both colo- but also display a range of recruitment commu- nies crossed. In this experiment, two S. postica nication strategies (Pisani et al., 1977; Correa- foragers arrived at the S. xanthotricha feeder Rego, 1990; Nieh and Roubik, 1995; Nieh and 28 S. xanthotricha foragers arrived at the et al., 2003b, in press). S. postica feeder. It is unclear how many of each species would have arrived at the feeder 6.3. Selective pressures of the opposite species in the absence of odor The selective pressures that have shaped marks, but this experiment suggests that inter- recruitment communication and given rise to specific odor-mark recognition may exist. functionally referential communication in hon- Nieh (1999) proposed that competition and eybees and, potentially, in stingless bees are olfactory eavesdropping of odor trails may poorly understood. Thus questions of why such have contributed to the evolution of concealed communication systems have evolved and how intranidal communication, and thus to the evo- functionally referential communication sys- lution of functionally referential communica- tems are advantageous remain fertile areas for tion as referential information replaced odor new hypotheses. Potential selective pressures trail information. To date, no studies show that include resource spatial distribution (habitat), less conspicuous odor trails decrease competi- and competition (Kerr, 1951; Johnson and tion at the indicated resource. However, the Hubbell, 1974; Hubbell and Johnson, 1977, short odor trail of T. hyalinata may represent 1978; Johnson, 1980, 1981; Roubik and an intermediate communication strategy Buchmann, 1984; Nagamitsu and Inoue, 1997; between full trail marking and target-only Wilms and Wiechers, 1997; Biesmeijer et al., odor marking (Nieh et al., 2003a). Whether 1999b; Steffan-Dewenter and Tscharntke, this strategy is less conspicuous than full-trail 2000; Eltz et al., 2001, 2002; Liow et al., 2001). marking, and therefore advantageous, remains The ability to communicate different to be tested. dimensions of food location may be influ- The role of competition in the evolution of enced by habitat. The height of food sources honeybee referential communication has not above the ground (Nagamitsu and Inoue, been considered, but given the recent finding Recruitment communication in stingless bees 175 that waggle dancing is more important in trop- un nourrisseur. Mais on rencontre plusieurs défis ical than in temperate habitats, it would be lorsqu’on travaille avec des abeilles sans aiguillon. interesting to explore the effects of competi- Certaines espèces d’abeilles sans aiguillon recrutent tion for the tropical resources that were found des membres du nid pour un lieu donné, alors que d’autres ne le font pas et le nombre de dimensions to be more spatially clumped than the temper- qui peuvent être communiquées varie (peut-être en ate resources (Dornhaus and Chittka, 2004). liaison avec l’habitat) (Tab. I). Les mécanismes de In general, the behavioral ecology of communication du recrutement incluent le contact, meliponine foraging is now being studied in la vision, l’olfaction et peut-être l’audition greater detail (Biesmeijer and Slaa, in review), (Lindauer et Kerr, 1958 ; Esch, 1967 ; Kerr, 1969). Les butineuses peuvent contacter leurs consoeurs à and the relative benefits and roles of different l’intérieur du nid pour les alerter de l’existence de recruitment strategies are beginning to be elu- bonnes sources de nourriture (Hrncir et al., 2000) ou cidated (Jarau et al., 2003). There are promis- les orienter vers la présence visuelle d’autres buti- ing signs that the field of stingless bee research neuses sur une source de nourriture (Slaa et al., is growing. With more detailed studies on 2003). Les butineuses de certaines espèces peuvent the sensory physiology and mechanisms of déposer des odeurs qui marquent les bonnes sources recruitment information transfer, we may de nourriture et constituer un chemin odorant com- reach firmer ground and begin to understand plet entre le nid et la source de nourriture (Lindauer et Kerr, 1958) ou un chemin odorant court qui part the evolution of the amazing recruitment com- de la source en direction du nid (Nieh et al., 2003a) munication systems used by social bees. ou ne marquer que la source de nourriture comme cible (Nieh et Roubik, 1995 ; Nieh, 1998b). Enfin les butineuses de nombreuses espèces produisent à ACKNOWLEDGEMENTS l’intérieur du nid des sons de recrutement intenses et pulsés. Chez certaines espèces, les sons sont cor- I would like to thank the insightful comments rélés à la distance et à la hauteur de la source de Felipe A.L. Contrera, David Roubik, Santiago nourriture (Aesch et al., 1965 ; Nieh et Roubik, Ramírez, and two anonymous reviewers. Their sug- 1998 ; Aguilar et Briceño, 2002 ; Nieh et al., gestions have substantially improved the quality of 2003c). On n’est pas certain que ces sons communi- this review. I would also like to thank Tom Seeley, quent le lieu de la source. Des questions restent en David Roubik, Cole Gilbert, Ron Hoy, and Kern suspens concernant le contenu de l’information et la Reeve for their encouragement, support, and enthu- fiabilité de ces sons. Pourtant des expériences anté- siasm in leading me into the field of stingless bee rieures en play-back suggèrent que les sons de biology. recrutement peuvent faire retourner à une source de nourriture des butineuses expérimentées (Lindauer et Kerr, 1958 ; Esch, 1967). Résumé – Communication du recrutement chez Les butineuses des abeilles sans aiguillon peuvent les abeilles sans aiguillon (Hymenoptera, Api- ainsi utiliser des modalités sensorielles multiples dae, Meliponini). Les mécanismes qui permettent aux abeilles sociales supérieures de communiquer le pour indiquer de façon appropriée les sources de lieu d’une bonne source de nourriture ont fait l’objet nourriture. Puisque la phylogénie des abeilles sans d’intenses études et peuvent donner un aperçu de la aiguillon n’est pas encore claire (Michener, 1990 ; façon dont la capacité à encoder l’information d’un Camargo et Pedro, 1992 ; Costa et al., 2003), il est lieu, forme de communication fonctionnellement difficile de faire des déclarations précises sur l’évo- référentielle (Marker et al., 1992 ; Blumstein, lution de la communication du recrutement. Néan- 1999), a évolué chez les abeilles sociales supérieu- moins, il se peut que l’ancêtre des abeilles sans res (Esch, 1967). Ainsi le but de cette revue est aiguillon (et peut-être des Apidae) ait fait des mou- d’examiner la communication du recrutement chez vements d’excitation une fois rentrée au nid après les abeilles sans aiguillon, en se concentrant sur les avoir trouvé une bonne source de nourriture, qu’elle divers types de communication. ait possédé une forme de marquage odorant (Kerr et Pour étudier la communication du recrutement, les Esch, 1965 ; Kerr, 1969), par des éléments qui ont chercheurs dressent des abeilles à butiner sur un pu évoluer vers les phéromones d’alarme (Kerr, nourrisseur placé dans un lieu donné et à enregistrer 1969). Les pressions de sélection sur la communica- le comportement de ces butineuses à l’intérieur du tion du recrutement peuvent inclure la distribution nid (Esch, 1967). Afin de tester si les abeilles peu- spatiale des sources de nourriture et la compétition, vent recruter pour un lieu donné, on place au moins issue de l’utilisation potentielle de marques odoran- deux nourrisseurs identiques, l’un témoin et l’autre tes étrangères, qui a pu contribuer à l’évolution de expérimental, de façon à ce qu’une seule dimension la communication cachée interne au nid. Ainsi, au spatiale soit modifiée à la fois (von Frisch, 1967). Il cours de l’évolution, l’information sur les chemins est important de ne compter comme recrues que les odorants aurait pu être remplacée par une commu- abeilles qui n’ont jamais expérimenté auparavant nication fonctionnellement reférentielle. 176 J.C. Nieh

Meliponini / transfert d’information / 1998). Es ist noch nicht klar, ob diese Töne den Ort recrutement / communication référentielle / des Futters angeben. Auch bleiben Fragen in Bezug multimodalité auf den Inhalt der Information und der Zuverlässig- keit dieser Töne bestehen. Frühere Versuche mit Rückspielungen dieser Töne weisen darauf hin, dass Zusammenfassung – Rekrutierungskommunika- diese Rekrutierungstöne erfahrene Sammelbienen tion bei Stachellosen Bienen (Hymenoptera, Api- veranlassen, zur Futterstelle zurückzukehren (Esch, dae, Meliponini). Die Mechanismen der 1967; Lindauer und Kerr, 1958). Kommunikation über den Ort einer guten Nahrungs- Demnach können Stachellose Bienen mehrere sen- quelle bei hoch sozialen Bienen sind intensiv unter- sorische Modalitäten nutzen, um Nahrungsquellen sucht worden. Diese Kenntnis kann uns einen ausreichend anzuzeigen. Da die Phylogenie der Sta- Einblick geben, wie sich eine funktionale, ortswei- chellose Bienen noch ungeklärt ist (Camargo und sende Kommunikation und die Fähigkeit zur Ent- Pedro, 1992; Costa et al., 2003; Michener, 1990), ist schlüsselung der Information über den Ort es schwierig detaillierte Feststellungen über die (Blumstein, 1999; Marler et al., 1992) in hoch sozia- Entwicklung der Rekrutierungskommunikation zu len Bienen entwickelt (Esch, 1967) hat. Dement- machen. Ursprüngliche scheinen die Vorfahren der sprechend wird hier ein Überblick über die Stachellose Bienen (und wahrscheinlich der Apidae) Untersuchungen der Rekrutierungskommunikation anregende Bewegungen im Nest gemacht zu haben, bei Stachellosen Bienen mit einer Ausrichtung auf wenn sie von einer guten Futterstelle zurückkehrten, die verschiedenen Mechanismen der Kommunika- und sie hatten eine Art der Duftmarkierung (Kerr, tion gegeben. 1969; Kerr und Esch, 1965) mit Elementen, die sich Zur Untersuchung von Rekrutierungskommunika- aus den Alarmpheromonen entwickelt haben könn- tion dressieren Forscher Bienen auf eine Futterstelle ten (Kerr, 1969). Selektionsdruck auf die Rekrutie- und notieren das Verhalten der Sammlerinnen im rungskommunikation könnte durch die räumliche Nest (Esch, 1967). Um zu testen, ob Bienen andere Verteilung der Nahrungsquellen und durch die aus Nestgenossinnen zu einer spezifischen Stelle rekru- der Möglichkeit der Fremdnutzung der Duftspuren tieren können, werden mindestens zwei identische entstehende Konkurrenz geformt worden sein, diese Futternäpfe, ein Kontroll- und ein Versuchsnapf, so könnte zur Evolution von verborgener Kommunika- plaziert, dass jedesmal nur eine der räumlichen tion innerhalb des Nestes beigetragen haben. So Dimensionen verändert wird (von Frisch, 1967). Es könnte während der Evolution die Information über ist wichtig, nur Bienen als Neulinge zu werten, die Duftwege von einer auf funktionell hinweisenden niemals zuvor Erfahrungen an einem Futternapf Kommunikation ersetzt worden sein (Nieh, 1999). gemacht haben. Bei Stachellose Bienen stellen sich bei dieser Arbeit Meliponini / Informationstransfer / referentielle mehrere Herausforderungen. Einige Arten der Sta- Kommunikation / Rekrutierung / Multimodalität chellose Bienen können Nestmitglieder zu einem bestimmten Ort rekrutieren, andere dagegen nicht. Zusätzlich variiert (wahrscheinlich abhängig vom REFERENCES Lebensraum) die Zahl der Dimensionen, die vermit- telt werden können (Tab. I). Mechanismen der Rekrutierungskommunikation beinhalten Kontakt, Aguilar I., Briceño D. (2002) Sounds in M. Sehen, Riechen und vielleicht auch Hören (Esch, costaricensis (Apidae: Meliponini): effect of sugar concentration and nectar source distance, 1967; Kerr, 1969; Lindauer und Kerr, 1958). Sam- Apidologie 33, 375–388. melbienen können mit ihren Nestgenossinnen im Nest Kontakt aufnehmen, um sie auf die Existenz Aguilar I., Sommeijer M. (2001) The deposition of einer guten Nahrungsquelle aufmerksam zu machen anal excretions by Melipona favosa foragers (Hrncir et al., 2000) oder sie können sehen, dass sich (Apidae: Meliponinae): behavioural observations andere Sammelbienen an der Nahrungsquelle concerning the location of food sources, aufhalten (Slaa et al., 2003). Sammlerinnen einiger Apidologie 32, 37–48. Arten können gute Futterstellen auf unterschiedliche Aguilar I., Sommeijer M.J. (1996) Communication in Weise mit Duftstoffen markieren: es kann eine voll- stingless bees: are the anal substances deposited ständige Duftspur von der Futterstelle zum Nest by Melipona favosa scent marks? Proc. Sect. Exp. (Lindauer und Kerr, 1958), eine kurze Duftspur nur Appl. Entomol. Netherlands Entomol. Soc. 7, 57– ein kleines Stück in Richtung zum Nest gelegt wer- 63. den (Nieh et al., 2003a), oder es erfolgt nur eine Ziel- Almeida M.C., Laroca S. (1988) Trigona spinipes markierung an der Nahrungsquelle (Nieh, 1998b; (Apidae, Meliponinae): taxonomy, bionomy, and Nieh und Roubik, 1995). Schließlich können Sam- trophic relationships in restricted areas, Acta melbienen vieler Arten laute, pulsierende Rekrutie- Biol. Parana. 17, 67–108. rungstöne im Nest erzeugen. In einigen Arten sind Biesmeijer J.C., de Vries H. (2001) Exploration and diese Töne mit der Entfernung und der Höhe der Fut- exploitation of food sources by social insect terstelle korreliert (Aguilar und Briceño, 2002; Esch colonies: a revision of the scout-recruit concept, et al., 1965; Nieh et al., 2003c; Nieh und Roubik, Behav. Ecol. Sociobiol. 49, 89–99. Recruitment communication in stingless bees 177

Biesmeijer J.C., Ermers M.C.W. (1999) Social Camargo J.M.F., Pedro S.R.M. (1992) Sistemática de foraging in stingless bees: how colonies of Meliponinae (Hymenoptera, Apidae): sobra Melipona fasciata choose among nectar sources, polaridade e significado de algns caracteres Behav. Ecol. Sociobiol. 46, 129–140. morfológicos, in: Cruz-Landim C., Chaud-Netto J. (Eds.), Anais do Encontro Brasileiro de Biesmeijer J.C., Slaa E.J. (2004) Information flow Biologia de Abelhas e outros Insectos Sociais, and organization of stingless bee foraging, UNESP, São Paulo, pp. 45–49. Apidologie 35, 143–157. Camargo J.M.F., Roubik D.W. (1991) Systematics Biesmeijer J.C., van Nieuwstadt M.G.L., Lukács S., and bionomics of the apoid obligate necrophages: Sommeijer M.J. (1998) The role of internal and the Trigona hypogea group (Hymenoptera: external information in foraging decisions of Apidae; Meliponinae), Biol. J. Linn. Soc. 44, 13– Melipona workers (Hymenoptera: Meliponinae), 40. Behav. Ecol. Sociobiol. 42, 107–116. Cameron S.A., Whitfield J.B. (1996) Use of walking Biesmeijer J.C., Born M., Lukacs S., Sommeijer M.J. trails, Nature 379, 125. (1999a) The response of the stingless bee Melipona beecheii to experimental pollen stress, Cameron S.A., Whitfield J.B., Cohen M., Thorp N. worker loss and different levels of information (1999) Novel use of walking trails by the input, J. Apic. Res. 38, 33–41. Amazonian bumble bee, Bombus transversalis (Hymenoptera: Apidae), in: Byers G.W., Hagen Biesmeijer J.C., Richter J.A.P., Smeets M.A.J.P., R.H., Brooks R.W. (Eds.), Entomological Sommeijer M.J. (1999b) Niche differentiation in contributions in memory of Byron A. Alexander, nectar-collecting stingless bees: the influence of Vol. 24, University of Kansas Natural History morphology, floral choice and interference Museum Special Publications, pp. 187–193. competition, Ecol. Entomol. 24, 380–388. Chavarria G., Carpenter J.M. (1994) Total evidence Blum M.S., Crewe R.M., Kerr W.E., Keith L.H., and the evolution of highly social bees, Cladistics Garrison A.W., Walker M.M. (1970) Citral in 10, 229–258. stingless bees: isolation and functions in trail- laying and robbing, J. Insect Physiol. 16, 1637– Chittka L., Dornhaus A. (1999) Comparisons in 1648. physiology and evolution, and why bees can do the things they do, Cien. Dia 2, 1–17. Blumstein D.M. (1999) The evolution of functionally referential alarm communication: multiple Collins A.M., Rinderer T.E., Daly H.V., Harbo J.R., adaptations; mutiple constraints, Evol. Commun. Pesante D. (1989) Alarm pheromone production 3, 135–147. by two honeybee (Apis mellifera) types, J. Chem. Ecol. 15, 1747–1756. Bradbury J., Vehrencamp S.L. (1988) Principles of animal communication, Sinauer Associates, Inc., Correa-Rego M.M. (1990) Revisão do genero Sunderland, Massachusetts. Melipona Illiger, 1806: genitalia e esternos pre- genitais de machos (Meliponinae, Apidae, Breed M.D., Stocker E.M., Baumgartner L.K., Vargas Hymenoptera), in: Departamento de Filosofia, S.A. (2002) Time-place learning and the ecology Ciências e Letras, Universidade de São Paulo, of recruitment in a stingless bee, Trigona Ribeirão Preto. amalthea (Hymenoptera, Apidae), Apidologie 33, 251–258. Costa M.A., Del Lama M.A., Melo G.A.R., Sheppard W.S. (2003) Molecular phylogeny of the stingless Brian A.D. (1954) The foraging of bumble bees, Bee bees (Apidae, Apinae, Meliponini) inferred from World 35, 61–67; 81–91. mitochondrial 16S rDNA sequences, Apidologie Brockmann A., Brückner D., Crewe R.M. (1998) The 34, 73–84. EAG response spectra of workers and drones to Croat T.B. (1978) Flora of Barro Colorado Island, queen honeybee mandibular gland components: Stanford University Press, United States of the evolution of a social signal, America. Naturwissenschaften 85, 283–285. Cruz-Landim C.d., Ferreira A. (1968) Mandibular Cameron S.A. (2003) Data from the elongation gland development and communication in field factor-1-a gene coroborates the phylogenetic bees of Trigona (Scaptotrigona) postica pattern from other genes revealing common (Hymenoptera: Apidae), J. Kans. Entomol. Soc. ancestry of bumble bees and stingless bees 41, 474–481. (Hymenoptera: Apinae), in: Vandame R., Rabanales M.C.M.R., Díaz M.C.M.A.G., Conesa Cruz-Landim C.d., Moraes Regina L.M.S.d., Salles I.B.T.J.G., Guillén M.C.D.S. (Eds.), III. Heliana C., Reginato Rejane D. (1998) Note on Seminario Mesoamericano sobre Abejas sin glands present in Meliponinae (Hymenoptera, Aguijón, El Colegio de la Frontera Sur, Apidae) bees legs, Rev. Bras. Zool. 15, 159–165. Tapachula, Chiapas, México, pp. 132–135. de Bruijn L.L.M., Sommeijer M.J. (1997) Colony Cameron S.A., Mardulyn P. (2001) Multiple foraging in different species of stingless bees molecular data sets suggest independent origins (Apidae, Meliponinae) and the regulation of of highly eusocial behavior in bees individual nectar foraging, Insectes Soc. 44, 35– (Hymenoptera: Apinae), Syst. Biol. 50, 194–214. 47. 178 J.C. Nieh

de Korte M., Weissenbacher B.K.H., Crew R.M. eicosen-1-ol to foraging honeybees, J. Apic. Res. (1998) Chemical signals in a stingless bee 21, 151–156. Trigona (Meliplebdeia) denoiti Vachal (Hymenoptera: Apidae: Meliponinae), J. Gould J.L. (1976) The dance-language controversy, Entomol. Soc. South Afr. 51, 9–16. Q. Rev. Biol. 51, 211–244. de Marco R.J., Farina W.M. (2003) Trophallaxis in Goulson D., Stout J.C., Langley J., Hughes W.O.H. forager honeybees (Apis mellifera): resource (2000) Identity and function of scent marks uncertainty enhances begging contacts? J. Comp. deposited by foraging bumblebees, J. Chem. Physiol. A 189, 125–134. Ecol. 26, 2897–2911. Dornhaus A. (2002) When do dances make a Goulson D., Chapman J.W., Hughes W.O.H. (2001) difference? Significance of honey bee Discrimination of unrewarding flowers by bees; recruitment depends on foraging distance, direct detection of rewards and use of repellent Entomol. Gen. 26, 93–100. scent marks, J. Insect Behav. 14, 669–678. Dornhaus A., Cameron S. (2003) A scientific note on Hart A.G., Ratnieks F.L.W. (2002) Task-partitioned food alert in Bombus transversalis, Apidologie nectar transfer in stingless bees: work 34, 87–88. organisation in a phylogenetic context, Ecol. Entomol. 27, 163–168. Dornhaus A., Chittka L. (1999) Evolutionary origins of bee dances, Nature 401, 38. Hölldobler B., Wilson E. (1990) The ants, The Belknap Press of Harvard University Press, Dornhaus A., Chittka L. (2001) Food alert in Cambridge, Massachusetts. bumblebees (Bombus terrestris): possible mechanisms and evolutionary implications, Hölldobler B., Wilson E.O. (1978) The multiple Behav. Ecol. Sociobiol. 50, 570–576. recruitment systems of the African weaver ant Oecophylla longinoda (Latreille) (Hymenoptera: Dornhaus A., Chittka L. (2004) Why do honey bees Formicidae), Behav. Ecol. Sociobiol. 3, 19–60. dance? Behav. Ecol. Sociobiol. 55, 395–401. Hrncir M., Jarau S., Zucchi R., Barth F.G. (2000) Duffield R.M., Wheeler J.W., Eickwort G.C. (1984) Recruitment behavior in stingless bees, Melipona Sociochemicals of bees, in: Bell W.J., Cardé R.T. scutellaris and M. quadrifasciata. II. Possible (Eds.), Chemical ecology of insects, Sinauer mechanisms of communication, Apidologie 31, Associates, Inc., Sunderland, Massachusetts, 93–113. pp. 387–428. Hrncir M., Jarau S., Zucchi R., Barth F.G. (2002) Dyer F.C. (2002) The biology of the dance language, Dependence of sound signals in Melipona Annu. Rev. Entomol. 47, 917–949. seminigra with factors other than food location, Eltz T., Brühl C.A., van der Kaars S., Chey V.K., in: Billen J. (Ed.), XIV Int. Congr. IUSSI: The Linsenmair K.E. (2001) Pollen foraging and golden jubilee proceedings, Science Council of resource partitioning of stingless bees in relation Japan, Entomological Society of Japan, Hokkaido to flowering dynamics in a Southeast Asian University Coop, Hokkaido University, Sapporo, tropical rainforest, Insectes Soc. 48, 273–279. Japan, p. 124. Eltz T., Brühl C.A., van der Kaars S., Linsenmair K.E. Hrncir M., Jarau S., Zucchi R., Barth F.G. (2003) A (2002) Determinants of stingless bee nest density stingless bee (Melipona seminigra) uses optic in lowland dipterocarp forests of Sabah, flow to estimate flight distances, J. Comp. Malaysia, Oecologia 131, 27–34. Physiol. A 189, 761–768. Esch H. (1967) Die Bedeutung der Lauterzeugung für Hrncir M., Jarau S., Zucchi R., Barth F.G. (2004) On die Verständigung der stachellosen Bienen, Z. the origin and properties of scent marks deposited Vergl. Physiol. 56, 408–411. at the food source by a stingless bee, Melipona seminigra Friese 1903, Apidologie 35, 3–13. Esch H., Esch I., Kerr W.E. (1965) Sound: an element common to communication of stingless bees and Hubbell S.P., Johnson L.K. (1977) Competition and to dances of the honey bee, Science 149, 320– nest spacing in a tropical stingless bee 321. community, Ecology 58, 949–963. Esch H.E., Zhang S., Srinivasan M.V., Tautz J. (2001) Hubbell S.P., Johnson L.K. (1978) Comparative Honeybee dances communicate distances foraging behavior of six stingless bee species measured by optic flow, Nature 411, 581–583. exploiting a standardized resource, Ecology 59, 1123–1136. Fernandes-Salomao T.M., Muro-Abad J.I., De Oliveira Campos L.A., Fernandes de Araujo E. Inoue T., Roubik D.W., Suka T. (1999) Nestmate (2002) Mitochondrial and nuclear DNA recognition in the stingless bee Melipona characterization in the Melipona species panamica (Apidae, Meliponini), Insectes Soc. 46, (Hymenoptera, Meliponini) by RFLP analysis, 208–218. Hereditas 137, 229–233. Jacobs-Jessen U.F. (1959) Zur Orientierung der Free J.B., Williams I.H. (1983) Scent-marking of Hummeln und einiger anderer sozialer und flowers by honeybees, J. Apic. Res. 22, 86–90. solitärer Apiden, Z. Vergl. Physiol. 41, 597–641. Free J.B., Williams I., Pickett J.A., Ferguson A.W., Jarau S., Hrncir M., Zucchi R., Barth F.G. (2000) Martin A.P. (1982) Attractiveness of (Z)-11- Recruitment behavior in stingless bee, Melipona Recruitment communication in stingless bees 179

scutellaris and M. quadrifasciata. I. Foraging at Kerr W.E. (1997) Sex determination in honey bees food sources differing in direction and distance, (Apinae and Meliponinae) and its consequences, Apidologie 31, 81–91. Braz. J. Genet. 20, 601–611. Jarau S., Hrncir M., Zucchi R., Barth F.G. (2002) Foot Kerr W.E., Cruz C.d. (1961) Funções diferentes print pheromones used to mark food sources by tomadas pela glândula mandibular na evolução stingless bees, in: Billen J. (Ed.), XIV Int. Congr. das abelhas e em Trigona (Oxytrigona) tataira IUSSI: the golden jubilee proceedings, Hokkaido em especial, Rev. Bras. Biol. 21, 1–16. University Coop, Hokkaido University, Sapporo, Japan, p. 16. Kerr W.E., Esch H. (1965) Comunicação entre as abelhas sociais brasileiras e sua contribuição para Jarau S., Hrncir M., Schmidt V.M., Zucchi R., Barth o entendimento da sua evolução, Cienc. Cult. 17, F.G. (2003) Effectiveness of recruitment 529–538. behavior in stingless bees (Apidae, Meliponini), Insectes Soc. 50, 365–374. Kerr W.E., Rocha R. (1988) Comunicação em Melipona rufiventris e Melipona compressipes, Jarau S., Hrncir M., Zucchi R., Barth F.G. (2004) A Cienc. Cult. 40, 1200–1203. stingless bee uses labial gland secretions for scent trail communication (Trigona recursa Smith Kerr W.E., Blum M., Fales H.M. (1981) 1863), J. Comp. Physiol. A, in press. Communication of food sources between workers of Trigona (Trigona) spinipes, Rev. Bras. Biol. Johnson L.K. (1974) The role of agonistic behavior in 41, 619–623. the foraging strategies of Trigona bees, in University of California, Berkeley, Berkeley, Kerr W.E., Ferreira A., Simões de Mattos N. (1963) California, USA. Communication among stingless bees — additional data (Hymenoptera: Apidae), J. N.Y. Johnson L.K. (1980) Foraging strategies and the Entomol. Soc. 71, 80–90. structure of stingless bee communities in Costa Rica, in: Jaisson P. (Ed.), Proc 1st Int. Symp. Kirchner W.H. (1994) Hearing in honeybees: the organized by the IUSSI and the Sociedad mechanical response of the bee's antenna to near Mexicana de Entomologia, Vol 2. Centre field sound, J. Comp. Physiol. A 175, 261–265. Régional de Publications de la 5e Circonscription Kirchner W.H., Lindauer M., Michelsen A. (1998) du C.N.R.S., Cocoyoc, Morelos, Mexico, pp. 31– Honeybee dance communication: acoustical 58. indication of direction in round dances, Johnson L.K. (1981) Effect of flower clumping on Naturwissenschaften 75. defense of artificial flowers by aggressive Kolmes S.A., Sommeijer M.J. (1992) Ergonomics in stingless bees, Biotropica 13, 151–157. stingless bees: changes in intranidal behavior Johnson L.K., Hubbell S.P. (1974) Aggression and after partial removal of storage pots and honey in competition among stingless bees: field studies, Melipona favosa (Hym. Apidae, Meliponinae), Ecology 55, 120–127. Insectes Soc. 39, 215–232. Johnson L.K., Hubbell S.P. (1987) Defense of food Krebs J.R., Dawkins R. (1984) Animal Signals Mind- supply by eusocial colonies, Am. Zool. 27, 347– Reading and Manipulation, in: Krebs J.R., Davies 358. N.B. (Eds.), Behavioural Ecology: An Evolutionary Approach, 2nd ed., Sinauer Johnson L.K., Haynes L.W., Carlson M.A., Fortnum Associates, Inc., Sunderland, Mass., pp. 380–402. H.A., Gorgas D.L. (1985) Alarm substances of the stingless bee, Trigona silvestriana, J. Chem. Lindauer M., Kerr W.E. (1958) Die gegenseitige Ecol. 11, 409–416. Verständigung bei den stachellosen Bienen, Z. Vergl. Physiol. 41, 405–434. Kerr W.E. (1951) Bases para o estudo da genética de poplações do Hymenoptera em geral e dos Lindauer M., Kerr W.E. (1960) Communication Apinae sociais em particular, An. Esc. Sup. between the workers of stingless bees, Bee World Agric. Luiz de Queiróz 8, 219–354. 41, 29–41; 65–71. Kerr W.E. (1960) Evolution of communication in Liow L.H., Sodha N.S., Elmqvist T. (2001) Bee bees and its role in speciation, Evolution 14, 326– diversity along a disturbance gradient in tropical 327. lowland forests of south-east Asia, J. Appl. Ecol. 38, 180–192. Kerr W.E. (1969) Some aspects of the evolution of social bees, Evol. Biol. 3, 119–175. Marler P., Evans C.S., Hauser M.D. (1992) Animal signals: motivation, referential, or both? in: Kerr W.E. (1972) Orientação pelo sol em Trigona Papoucek M. (Ed.), Nonverbal vocal spinipes, Cienc. Cult. Suppl. 341–342. communication: comparative and developmental Kerr W.E. (1973) Sun compass orientation in the approaches, Cambridge University Press, stingless bees, Trigona (Trigona) spinipes Cambridge, pp. 66–86. (Fabricius, 1793) (Apidae), An. Acad. Brasil. Martins C.F., Aguilar J.B.V. (1992) Visits at a feeding Ciênc. 45, 301–308. stations during the dry season of Africanized Kerr W.E. (1994) Communication among Melipona honey bees and native social insects in the workers (Hymenoptera: Apidae), J. Insect Behav. Brasilian Caatinga (Hymenoptera: Apidae), 7, 123–128. Entomol. Gen. 17, 9–15. 180 J.C. Nieh

Michelsen A. (2003) Signals and flexibility in the Nieh J.C., Contrera F.A.L., Ramírez S., Imperatriz- dance communication of honeybees, J. Comp. Fonseca V.L. (2003b) Variation in the ability to Physiol. A 189, 165–174. communicate 3-D resource location by stingless Michelsen A., Andersen B.B., Kirchner W.H., bees from different habitats, Anim. Behav. 66, Lindauer M. (1989) Honeybees can be recruited 1129–1139. by a mechanical model of a dancing bee, Nieh J.C., Contrera F.A.L., Rangel J., Imperatriz- Naturwissenschaften 76, 277–280. Fonseca V.L. (2003c) Effect of food location and Michelsen A., Kirchner W.H., Andersen B.B., quality on recruitment sounds and success in two Lindauer M. (1986a) The tooting and quacking stingless bees, Melipona mandacaia and vibration signals of honeybee queens: a Melipona bicolor, Behav. Ecol. Sociobiol. 55, quantitative analysis, J. Comp. Physiol. A 158, 87–94. 605–611. Nieh J.C., Ramírez S., Nogueira-Neto P. (2003d) Michelsen A., Kirchner W.H., Lindauer M. (1986b) Multi-source odor-marking of food by a stingless Sound and vibrational signals in the dance bee, Melipona mandacaia, Behav. Ecol. language of the honeybee, Apis mellifera, Behav. Sociobiol. 54, 578–586. Ecol. Sociobiol. 18, 207–212. Nogueira-Neto P. (1997) Vida e criação de abelhas Michener C.D. (1974) The social behavior of the bees, indígenas sem ferrão, Editoria Nogueirapis, São Harvard University Press, Cambridge, Mass. Paulo. Michener C.D. (1990) Classification of the Apidae Nogueira-Neto P. (1999) Stressful situations and their (Hymenoptera), Univ. Kans. Sci. Bull. 54, 75– consequences on sex determination and on the 164. behavior of stingless bees (Meliponinae, Apidae, Hymenoptera), Rev. Etol. 1, 65–68. Michener C.D. (2000) The bees of the world. Johns Hopkins University Press, Baltimore and Pisani J.F., Kerr W.E., Cunha R.A. (1977) Estrutura London. multidimensional dos Meliponídeos, Acta Amaz. 7, 233–245. Nagamitsu T., Inoue T. (1997) Aggressive foraging of social bees as a mechanism of floral resource Rizzini C.T. (1997) Tratado de fitogeografia do partitioning in an Asian tropical rainforest, Brasil, 2nd ed., Âmbito Cultural, Rio de Janeiro. Oecologia 110, 432–439. Rohrseitz K., Tautz J. (1999) Honey bee dance Nieh J.C. (1993) The stop signal of honey bees: communication: waggle run direction coded in Reconsidering its message, Behav. Ecol. antennal contacts? J. Comp. Physiol. A 184, 463– Sociobiol. 33, 51–56. 470. Nieh J.C. (1998a) The food recruitment dance of the Roubik D.W. (1982a) Obligate necrophagy in a social stingless bee, Melipona panamica, Behav. Ecol. bee, Science 217, 1059–1060. Sociobiol. 43, 133–145. Roubik D.W. (1982b) Seasonality in colony food sto- Nieh J.C. (1998b) The role of a scent beacon in the rage, brood production, and adult survivorship: communication of food location in the stingless studies of Melipona in tropical forest (Hymenop- bee, Melipona panamica, Behav. Ecol. Sociobiol. tera: Apidae), J. Kans. Entomol. Soc. 55, 789– 43, 47–58. 800. Nieh J.C. (1999) Stingless-bee communication, Am. Roubik D.W. (1989) Ecology and natural history of Sci. 87, 428–435. tropical bees, Cambridge University Press, New York. Nieh J.C., Roubik D.W. (1995) A stingless bee (Melipona panamica) indicates food location Roubik D.W. (1993) Tropical pollinators in the without using a scent trail, Behav. Ecol. canopy and understory: field data and theory for Sociobiol. 37, 63–70. stratum "preferences", J. Insect Behav. 6, 659– 673. Nieh J.C., Roubik D.W. (1998) Potential mechanisms for the communication of height and distance by Roubik D.W., Aluja M. (1983) Flight ranges of a stingless bee, Melipona panamica, Behav. Ecol. Melipona and Trigona in tropical forest, J. Kans. Sociobiol. 43, 387–399. Entomol. Soc. 56, 217–222. Nieh J.C., Tautz J. (2000) Behavior-locked signal Roubik D.W., Buchmann S.L. (1984) Nectar analysis reveals weak 200–300 Hz comb selection by Melipona and Apis mellifera vibrations during the honeybee waggle dance, J. (Hymenoptera: Apidae) and the ecology of nectar Exp. Biol. 203, 1573–1579. intake by bee colonies in a tropical forest, Oecologia 61, 1–10. Nieh J.C., Tautz J., Spaethe J., Bartareau T. (2000) The communication of food location by a Roubik D.W., Inoue T., Hamid A., Harrison R. (1999) primitive stingless bee, Trigona carbonaria, Height communication by Bornean honey bees Zoology 102, 239–246. (Apiformes: Apidae; Apini), J. Kans. Entomol. Nieh J.C., Contrera F.A.L., Nogueira-Neto P. (2003a) Soc. 72, 256–261. Pulsed mass-recruitment by a stingless bee, Sakagami S.F. (1966) Techniques for the observation Trigona hyalinata, Proc. Roy. Soc. London B of stingless bees by a using a special hive, Papéis 270, 2191–2196. Avulsos Zool., São Paulo 19, 151–162. Recruitment communication in stingless bees 181

Sakagami S.F., Roubik D.W., Zucchi R. (1993) and heterospecifics in stingless bees, Ecol. Ethology of the robber stingless bee, Lestrimelitta Entomol. 28, 369–379. limao (Hymenoptera: Apidae), Sociobiology 21, Smith B.H., Roubik D.W. (1983) Mandibular glands 237–277. of stingless bees (Hymenoptera: Apidae), J. Sandeman D.C., Tautz J., Lindauer M. (1996) Chem. Ecol. 9, 1465–1472. Transmission of vibration across honeycombs Srinivasan M.V., Zhang S., Altwein M., Tautz J. and its detection by bee leg receptors, J. Exp. (2000) Honeybee navigation: nature and Biol. 199, 2585–2594. calibration of the "odometer", Science 287, Schmidt V.M., Zucchi R., Barth F.G. (2003) A 8510853. stingless bee marks the feeding site in addition to Steffan-Dewenter I., Tscharntke T. (2000) Resource the scent path (Scaptotrigona aff. deplis), overlap and possible competition between Apidologie 34, 237–248. honeybees and wild bees in central Europe, Seeley T.D. (1983) Division of labor between scouts Oecologia 112, 288–296. and recruits in honeybee foraging, Behav. Ecol. Stort A.C., Moraes A.M.M.B. (1997) Sensory Sociobiol. 12, 253–259. structures of the antennae of Nannotrigona Seeley T.D. (1985) The information-center strategy of testaceicornis (Apidae: Meliponinae), J. N.Y. honeybee foraging, in: Hölldobler B., Lindauer Entomol. Soc. 105, 230–235. M. (Eds.), Experimental Behavioral Ecology, G. Stort A.C., Moraes A.M.M.B. (1998) A study of the Fischer Verlag, New York, pp. 75–90. sensory structures of the antennae of Seeley T.D. (1989a) The honey bee colony as a Scaptotrigona postica workers (Hymenoptera- superorganism, Am. Sci. 77, 546–553. Apidae), Rev. Bras. Biol. 58, 163–167. Seeley T.D. (1989b) Social foraging in honey bees: Stout J.C., Goulson D. (2001) The use of conspecific how nectar foragers assess their colony’s and interspecific scent marks by foraging nutritional status, Behav. Ecol. Sociobiol. 24, bumblebees and honeybees, Anim. Behav. 62, 181–199. 183–189. Seeley T.D. (1992) The tremble dance of the honey Tautz J., Rohrseitz K. (1998) What attracts honeybees bee: message and meanings, Behav. Ecol. to a waggle dancer? J. Comp. Physiol. A 183, Sociobiol. 31, 375–383. 661–667. Seeley T.D. (1995) The wisdom of the hive: the social Tautz J., Sandeman D.C. (2003) Recruitment of honeybees to non-scented food sources, J. Comp. physiology of honey bee colonies, Harvard Physiol. A 189, 293–300. University Press, Cambridge, Mass. Tinbergen N. (1952) Derived activities: their Seeley T.D., Kuehnholz S., Weidenmueller A. (1996) causation, biological significance, origin, and The honey bee's tremble dance stimulates emancipation during evolution, Q. Rev. Biol. 27, additional bees to function as nectar receivers, 1–32. Behav. Ecol. Sociobiol. 39, 419–427. van Nieuwstadt M.G.L., Ruano I.C.E. (1996) Seeley T.D., Mikheyev A.S., Pagano G.J. (2000) Relation between size and foraging range in Dancing bees tune both duration and rate of stingless bees (Apidae, Meliponinae), Apidologie waggle-run production in relation to nectar- 27, 219–228. source profitability, J. Comp. Physiol. A 186, 813–819. Vázquez R.G., Rincón M.R., Rojas J., Malo E., Cruz L. (2003) Respuesta electrofisiologica y Sherman G., Visscher P.K. (2002) Honeybee colonies conductual de la abeja scaptotrigona mexicana a achieve fitness through dancing, Nature 419, los volatiles florales de Coffea canephora var. 920–922. robusta, in: Vandame R., Rabanales M.C.M.R., Slaa E.J., Wassenberg J., Biesmeijer J.C. (2003) The Díaz M.C.M.A.G., Conesa I.B.T.J.G., Guillén use of field-based social information in eusocial MCDS (Eds.), III. Seminario Mesoamericano foragers: local enhancement among nestmates sobre Abejas sin Aguijón, El Colegio de la and heterospecifics in stingless bees, Ecol. Frontera Sur, Tapachula, Chiapas, México, Entomol. 28, 369–379. pp. 79–82. Slaa E.J., Biesmeijer J.C. (in press) Flower constancy: Villa J.D., Weiss M.R. (1990) Observations on the A review of definitions, measurements and data use of visual and olfactory cues by Trigona spp. analysis, in: Dafni A., Kevan P.G. (Eds.), Manual foragers, Apidologie 21, 541–546. of field methods in pollination ecology, Visscher P.K., Shepardson J., McCart L., Camazine Enviroquest, Cambridge, Ontario, Canada. S. (1999) Vibration signal modulates the behavior Slaa E.J., Nieuwstadt M.G.L.v., Pisa L.W., of house-hunting honey bees (Apis mellifera), Sommeijer M.J. (1997) Foraging strategies of Ethology 105, 759–769. stingless bees (Apidae, Meliponinae): the relation von Frisch K. (1967) The dance language and between precision of recruitment, competition orientation of bees, 2nd printing, 1993 Belknap and communication, Acta Hortic. 437, 193–197. Press, Cambridge, Mass. Slaa E.J., Wassenberg J., Biesmeijer J.C. (2003) The Waddington K.D. (1982) Honey bee foraging use of field-based social information in eusocial profitability and round dance correlates, J. Comp. foragers: local enhancement among nestmates Physiol. A 148, 297–301. 182 J.C. Nieh

Waddington K.D., Kirchner W.H. (1992) Acoustical the introduced Africanized honey bee in the and behavioral correlates of profitability of food Brazilian Atlantic rain forest, Apidologie 28, sources in honey bee round dances, Ethology 92, 339–355. 1–6. Wilms W., Wendel L., Zillikens A., Blochtein B., Wagner W. (1907) Psychobiologische Untersuchun- Engels W. (1997) Bees and other insects recorded gen an Hummeln, I und II, Zoologica 19, 1–239. on flowering trees in a subtropical Araucaria Wehner R. (1992) Arthropods, in: Papi F. (Ed.), forest in southern Brazil, Stud. Neotrop. Fauna Animal homing, Chapman and Hall, New York, Environ. 32, 220–226. pp. 45–144. Wilson E.O. (1971) The insect societies, Belknap Weidenmuller A., Seeley T.D. (1999) Imprecision in Press of Harvard University Press, Cambridge, waggle dances of the honeybee (Apis mellifera) Mass. for nearby food sources: error or adaptation? Winston M.L., Michener C.D. (1977) Dual origin of Behav. Ecol. Sociobiol. 46, 190–199. highly social behavior among bees, Proc. Natl. Wenner A.M. (2002) The elusive honey bee dance Acad. Sci. (USA) 74, 1135–1137. "language" hypothesis, J. Insect Behav. 15, 859– Winston M.L., Slessor K.N. (1998) Honey bee primer 878. pheromones and colony organization: gasps in Wille A. (1979) Phylogeny and relationships among our knowledge, Apidologie 29, 81–95. the genera and subgenera of the stingless bees Wittman D., Radtke R., Zeil J., Lübke G., Franke W. (Meliponinae) of the world, Rev. Biol. Trop. 27, (1990) Robber bees (Lestrimelitta limao) and 241–277. their host chemical and visual cues in nest defense Wille A. (1983) Biology of the stingless bees, Annu. by Trigona (Tetragonisca) angustula (Apidae: Rev. Entomol. 28, 41–64. Meliponinae), J. Chem. Ecol. 16, 631–641. Wilms W., Wiechers B. (1997) Floral resource Zahavi A. (1980) Ritualization and the Evolution of partitioning between native Melipona bees and Movement Signals, Behaviour 72, 77–81.

To access this journal online: www.edpsciences.org