Apidologie 38 (2007) 584–594 Available online at: c INRA, EDP Sciences, 2007 www.apidologie.org DOI: 10.1051/apido:2007054 Original article Effect of forager-deposited odors on the intra-patch accuracy of recruitment of the stingless bees Melipona panamica and Partamona peckolti (Apidae, Meliponini)* Felipe Andrés León Contrera, James C. Nieh University of California San Diego, Division of Biological Sciences, Section of Ecology, Behavior, and Evolution, Mail Code 0116, 9500 Gilman Drive, La Jolla, CA 92093-0116, USA Received 24 July 2007 – Revised and Accepted 2 October 2007 Abstract – We show that the stingless bees Melipona panamica and Partamona peckolti haveahighpre- cision of intra-patch recruitment that is influenced by forager-deposited odor marks in the field. We trained foragers to a 2.5 M sucrose feeder in the center of an array of five identical feeders (20 cm between feeders) at different distances and directions from the nest and measured the distribution of recruit visitations. In the free-foraging phase foragers could odor mark a filter paper around the experimental feeder, and in the odor-removal phase we substituted it each five minutes by a clean one. Significantly more recruits in both species chose the experimental feeder over the controls in the distance and directional arrays (P 0.034) and odor removal significantly decreased precision of recruitment in both species (P 0.034). Scent marks in both species thus play a significant role in orienting recruits to already known profitable food sources. odor marking / recruitment / stingless bees / intra-patch precision / Meliponini 1. INTRODUCTION trail deposition (which indicates the direction from the food source to the nest of some stin- Stingless bees (Tribe Meliponini, Michener, gless bees, Kerr, 1960; Lindauer and Kerr, 2000) can recruit nestmates to food sources 1960; Nieh et al., 2003, 2004; Schmidt et al., (Jarau et al., 2000, 2003; Slaa et al., 2003; 2003; Jarau et al., 2004), point source odor Biesmeijer and Slaa, 2004; Aguilar et al., marking of the food source (Aguilar and Som- 2005), and meliponine communication of food meijer, 2001; Nieh et al., 2003; Schmidt et al., source location by foragers has been a sub- 2003, 2005; Hrncir et al., 2004; Jarau et al., ject of interest since the pioneering work 2005), pilot flights (Aguilar et al., 2005) to of Lindauer and Kerr (1958, 1960). To re- simple alerting (which evidently stimulate for- cruit, they use several different mechanisms agers to search for food, without providing any of food communication, ranging from sound location information, Lindauer and Kerr, 1958; pulse production (which may encode the dis- Kerr et al., 1963). tance and quality of food source: Esch et al., However, relatively little is known about the 1965; Esch, 1967; Aguilar and Briceño, 2002; precision of location communication in stin- Nieh et al., 2003; Hrncir et al., 2006), scent gless bees and the influence of the different Corresponding author: F.A.L. Contrera, mechanisms of communication on this preci- [email protected] sion. To date, no studies have examined the Present Address: Embrapa Amazônia Oriental – precision of newcomer recruitment with such Laboratório de Botânica. Tv. Dr. Enéas Pinheiro s/n. feeder arrays, although this is the standard C.P 48, CEP 66095-100, Belém/PA, Brazil. protocol in Apis mellifera studies (von Frisch, * Manuscript editor: Stan Schneider 1967; Towne and Gould, 1988). To test the Article published by EDP Sciences and available at http://www.apidologie.org or http://dx.doi.org/10.1051/apido:2007054 Odor marks in two stingless bee species 585 precision of recruitment communication, it is gless bee species, Melipona panamica Cock- necessary to offer bees more than two feed- erell, 1919 and Partamona peckolti (Friese, ers to choose from because the distribution of 1901). We tested the precision of recruitment foragers arriving at the feeders is thereby lim- in small-scale feeders arrays (20 cm separa- ited to two points and cannot be accurately as- tion among feeders) in order to determine how sessed (von Frisch, 1967; Towne and Gould, these species choose among clumped food 1988). For example, sampling of an error dis- sources. Melipona panamica foragers can re- tribution can yield a biased estimate if limited cruit nestmates to a good food source at a to two points. Sampling of multiple points is specific three-dimensional location (Nieh and desirable, although there are practical consid- Roubik, 1995), deposit attractive odor marks erations that limit the number of feeders in a on the food source (but not a scent trail), and given array, and thus most investigators have produce recruitment sounds that may provide settled on using arrays with a minimum of five location information inside the nest (Nieh, feeders (von Frisch, 1967). 1998; Nieh and Roubik, 1998). Colonies of M. In the meliponine bee Scaptotrigona mex- panamica usually contains 500 to 800 adult icana, Sánchez et al. (2004) used five-feeder individuals (Roubik, 1992; Nieh and Roubik, arrays to measure recruitment precision in re- 1995), nests in hollow trees, and occur from activated foragers. Reactivated foragers have northern Colombia to southern Costa Rica previously experienced the feeder at a similar (Roubik, unpubl. data). location and may therefore be searching for the In contrast, bees in the genus Partamona feeder on their own (based upon a previously appear to use a different mechanism of com- acquired search image) or act on new infor- munication. The only study concerning food mation provided by recruiters (Biesmeijer and source communication in Partamona was per- De Vries, 2001). Sánchez et al. (2004) found formed by Kerr (1969) in the species P. h e l- high precision in the feeder choice of reacti- leri (referred as Trigona (Partamona) cupira vated foragers (approximately 98% of foragers helleri). He reported that workers from P. chose the correct feeder in a five feeder di- helleri were able to communicate the direc- rectional array and approximately 89% made tion of the food source (79% of the recruits the correct choice in the five feeder distance reached the experimental feeder located 37 m array) when feeders were placed 50 m from from the nest in a paired-feeder experiment, the nest and space 5 m from each other. This N = 67 recruits). Noting a strong odor en- high precision of reactivated forager orienta- veloping recruiters, he speculated that foragers tion is likely related to forager odor marking of might release attractive aerial odors in flight to the feeder (Sánchez et al., 2004). Other studies assist recruit orientation, although this hypoth- (Aguilar and Sommeijer, 2001; Schmidt et al., esis still needs to be tested. Partamona peck- 2003, 2005; Hrncir et al., 2004) using paired olti usually nests upon cavities and crevices feeders showed that newcomers of some sting- located in epiphytes and other substrates, and less bees (M. favosa, M. seminigra, Nannotrig- several entrances may be found clumped. This ona testaceicornis and S. aff. depilis) can odor species inhabits the Pacific coast, from the mark the food source. In these studies, most northwestern forests of Peru and Ecuador into workers ( 63%) preferred to visit a feeder Colombia, Panama and Venezuela, and it was that was already visited by workers of the same found in great altitudes (2000 m in the Andes, species, when two feeders were offered to the Camargo and Pedro, 2003). bees in a dual choice experiment. However, no studies have yet examined the effect of odor mark removal on recruits in a feeder-array ex- 2. METHODS periment. We therefore used fan-shaped feeder arrays This study was performed on Barro Colorado to study the spatial precision of recruitment Island at the Smithsonian Tropical Research Insti- in a small patch of feeders and the influence tute (STRI), Republic de Panama, from Septem- of odor marks on this precision in two stin- ber to November 2005. We used one colony of P. 586 F.A.L. Contrera, J.C. Nieh peckolti and two colonies of M. panamica.TheP. not in contact with the sugar solution, we placed a peckolti colony (P1) was located in its natural nest Whatman filter paper (5.5 m inner diameter, 12 cm (in a log crevice) on a balcony outside the bee lab- outer diameter) that extended a further 3 cm around oratory (9◦9.923’N, 79◦50.193’W). The M. panam- from the feeder base and thus allowed bees to walk ica colonies (M1 and M2) were placed inside ob- on it and potentially deposit odor marks. In prelimi- servation colonies (description in Nieh and Roubik, nary trials, we placed an additional 2.5 cm circle of 1995) one inside the bee laboratory, and the other filter paper on top of the feeder to determine if bees on the balcony. Colony M1 corresponds to colony would land and deposit odor marks, as is observed D (approximately 2000 workers) and colony M2 in several stingless bee species (Nieh, 2004). How- to a new colony G in the sequence (approximately ever, we never observed bees landing on top of the 800–1000 workers) published in Nieh and Sanchez feeder (8 hours observation time) despite substan- (2005). Colony M1 was connected to the outside tial recruitment, and thus we used only filter paper with a 1 cm inner diameter, 10 cm long vinyl tube. under the feeder base during our experiments. Observations were made between 0800 and 1600 In the initial odor collection phase, we allowed h and we closed the entrances of M. panamica foragers to visit the feeder for 20 minutes, elicit- colonies not under study during our experiment to ing a large amount of recruitment. After the 20 min, avoid feeder visitation by non-subject colonies.