Journal of Insect Behavior, Vol. 12, No. 4. 1999

Evidence of Entrance Sharing and Subterranean Connections in Andrena agilissima (Hymenoptera; Andrenidae) M. Giovanetti,1 F. Andrietti,1,3 A. Martinoli,2 and F. Rigato1 Accepted October 30, 1998: revised February 1, 1999

Andrena agilissima is a bee that nests on sloped or vertical earth walls. Data from 3 years of fieldwork presented here show that this species is communal and uses the same nesting site for many years. On an earth wall on Isola d'Elba (Tuscany, Italy), covered with many bee nest entrances, we studied an aggregation of Andrena agilissima to determine if bees share entrances, if the sharing provides evidence for subterranean connections, and if bees involved in sharing show a division of labor. We marked several females coming out of the same entrance, as well as recapturing the same females using different entrances. No division of labor was observed in our observations. KEY WORDS: Andrena agilissima: ethology; Hymenoptera; nest sharing.

INTRODUCTION Michener (1974) defined aggregations as grouped nests in a restricted area, where each female makes her own nest. Aggregations are quite common among bees and wasps. Among those species that live in aggregation, different behaviors characterize the female interactions. In some cases, each female provisions her own nest, the only contacts among conspecifics being agonistic (solitary life). In other cases, the same nest is used by more than one individual, with each female provisioning her own cells and tolerating the other conspecifics living Dipartimento di Biologia, Via Celoria, 26. 20133 Milano, Italy. Dipartimento di Biologia Strutturale e Funzionale, Via Ravasi, 2 21100 Varese, Italy. 3To whom correspondence should be addressed. Fax: ++39-2-26604446. e-mail: andriett@imiucca. csi.unimi.it. 1

2

423 0892-7553/99/0700-0423$16.00/0 © 1999 Plenum Publishing Corporation

424

Giovanetti, Andrietti, Martinoli, and Rigato

with her (communal colonies). One can find females being closely related, living in the same nest, and cooperating in its construction and in the provision of the cells. In those cases, no hierarchy among females living together is defined and each female has developed ovaries (quasisocial colonies). Other species show a deeper definition of the interactions with a division of labor among nestmates (semisocial colonies). The division of labor is one of the characteristics which leads to complex social structures. These definitions (Michener, 1974) outline a change from solitary to social life in sequential steps. Andrena is a very large genus, including about 2000 species in the Old World (Iwata, 1976). For most species few data are available on nest-building behavior. They are mostly solitary bees; each female digs its own nest into horizontal or sloping ground and provisions its brood (offspring) cells with pollen and nectar. Aggregations of nests have been reported in few andrenid species (Ayasse et al., 1990; Visscher and Danforth, 1993; Neff and Simpson, 1997). However, their nesting habit may be solitary even in case of large aggregations. Occasionally two or more females may cohabit the same nest as in Andrena erythronii (Michener and Rettenmeyer, 1956). Only a few species are known to be communal, with several to many reproductive females sharing the same nest (Paxton and Tengo, 1996; Danforth, 1991; Danforth et al., 1996). In these species, each female builds and provisions her own cells (e.g., every female has developed ovaries) but shares the entrance and the main tunnel of the nest with other females. We discovered an aggregation of Andrena agilissima Scopoli on an earth wall, where many nest entrances were very close one to the other. As far as we know, there are only few other reports of A. agilissima, nesting on sandy walls (Schmiederknecht, 1882-1884; quoted by Rode, 1962) or on outside house walls (Rode, 1962), where many females were forming an aggregation. Those papers do not give useful information on the bees' interactions and therefore cannot help in proving communality. Westrich (1989), in Germany, suggested that this species is communal, but no studies have confirmed this since then. We asked the following questions: (1) Do females living in the same aggregation share the entrances which lead to subterranean burrows? (2) If so, are there subterranean connections among burrows? and (3) Is there any identifiable division of labor among bees? STUDY AREA The study area was located on Isola d'Elba (Tuscany, Italy), where the aggregation of A. agilissima was found on an earth wall 2.5 m high and 6 m wide, facing southwest. On this wall many entrances to subterranean burrows were interspersed with natural crevices, not always easily distinguishable from the bee entrances. Some of the entrances were probably from previous years,

Entrance Sharing and Subterranean Connections in Andrena agilissima

425

since they were not used during our observations. The distribution of entrances on the wall was not homogenous, the central part of the wall having the highest density. The wall was completely devoid of vegetation. On Isola d'Elba the precipitation (which influences the consistency of the earth wall starting erosion process) in May-June is usually less than 25 mm; the overall annual precipitation is usually about 700 mm. MATERIALS AND METHODS This study was carried out during the month of May in the years 1994-1996. Observations were made every day for a week each year from 0800 to 1700 (solar hours), the time when bees were active at the site. A preliminary study was carried out in May 1993. To recognize each bee individually, all individuals leaving from the nest entrances under observation were captured by means of a net and marked with nontoxic dyes applied to the thorax and the abdomen. Each bee bore a single mark or a unique combination of colors. In this way all the insects could be individually recognized when recaptured and their activity recorded. A number identified each nest entrance. Four days of intensive marking activity was enough to count all the bees coming out of a single entrance, as stated on the base of longer observation periods. To verify the presence of subterranean connections, we mapped the nest entrances every year and injected some smoke (produced by wet paper) into some of them, checking from which openings it came out. In 1996, we excavated a small part of the wall to describe the internal structure of the nest and the distribution of cells and to confirm the presence of connections among tunnels. RESULTS Not all the entrances on the wall were checked, but each year we demonstrated that more than one bee (except one single case) used the same entrance (Fig. 1). To calculate the average of individuals coming out from each entrance, we used the data obtained from the first 4 days of observations in each year. Thus we could compare the data of different years, considering that every year different entrances were checked for a different number of days. The first 4 days were comparable for all the years and were a period long enough to mark every bee coming out of a single entrance, as reported under Materials and Methods. In 1994 we checked 7 nests in the first 4 days (range of marked bees: 2-40 per nest), in 1995 we checked 25 nests (range: 1-42 per nest), and in 1996 we checked 3 nests (range: 8-33 per nest). In Table I we summarize the total number of checked entrances, the total number of females marked and recaptured, and the average number of individuals using the same entrance.

Giovanetti, Andrietti, Martinoli, and Rigato

426

Fig. 1. Number of entrances (vertical axis) used by different number of females (horizontal axis). For example, in 1995 four entrances were each used by 15 bees. Data obtained in the first 4 days of observation in 1994-1996.

Bees were tolerant one to another; we observed no aggressive behavior when a bee entered a tunnel while another bee was waiting to leave. A possible internal use of the nest by the bees is consistent with their high level of intranidal mating (Paxton et al., 1999). No data are available about the provisioning of cells and possible intraspecific cleptoparasitism. Recaptured individuals also showed the use of more than one entrance during their activity. Sometimes a bee entered an opening and came out of another a few minutes later, demonstrating the presence of subterranean connections; however, these may be restricted to the dense patch of the wall. However, in the case of Table I. Data Collected in Different Years During 7-Day Observation Periods (Rows 1-3)

Total checked entrances Total marked bees Total recaptured bees Average No. of individuals per entrance a (SD) a

1994

1995

1996

22 199 135 18

204 750 434 14.4

3 71 71 23.3

(15.2)

(10.6)

(13.4)

Averages refer to bees which were marked in the first 4 days of observation.

Entrance Sharing and Subterranean Connections in Andrena agilissima

427

Fig. 2. Number of bees (vertical axis) using different numbers of different entrances (horizontal axis). For example, in 1994 three bees used three entrances each. Data obtained in at least 4 days of observation in 1994 and 1995.

the nests observed in 1996, in a region of the wall where the nest density was lower then in the regions observed in 1994 and 1995, the bees marked coming out of one entrance used the same one during the entire period of observation. Figure 2 shows the number of entrances used by each marked bee, provided that she was observed for at least 4 days, for the years 1994 and 1995. Data presented in Figs. 1 and 2 do not deviate significantly from a normal distribution, on the basis of the one-sample Kolmogorov-Smirnov test. The aggregation of Andrena agilissima consisted of many hundreds of entrances in an area of about 15 m2. In the high-density area observed in 1995 (4 m2) the number of active entrances was 154 (giving an average density of 38.5 entrances/m2), but we do not know whether, during the season, some old entrances (sometimes recognizable) could have been reactivated. Our data are consistent with those found by Rode (1962), who gives a figure of 72 entrances/m 2 for the total number (either active or not) present in an area of 6.5 m2. Based on the mapping data, the position of the entrances on the earth wall (either used or not) was constant during the 3 years of this study. The experiment with the injection of smoke inside nest entrances showed that, when smoke enters just one entrance, it comes out of others. This method is not completely reliable, because the rock material that forms the wall has many natural crevices. If two tunnels are connected by a natural fracture, the smoke can spread into both and give misleading results.

428

Giovanetti, Andrietti, Martinoli, and Rigato

In October 1996 we excavated some nests, starting from one entrance. The excavation demonstrated that internally the tunnels are connected and that the cells are distributed in small groups. Some cells were filled with earth, as were some tunnels. We never observed any burrowing activities, i.e., no new tunnels were seen to be dug out from the outside. However, we often saw some sand coming out of nest entrances. The presence of earth-filled cells and tunnels proves that the tunnels are reorganized and reused over the years and explains the few records of new entrances. Figure 3 shows the distribution of cells and tunnels. DISCUSSION The genus Andrena is widespread and information on its biology is lacking for most species. Among the Andrena that have been studied, some are solitary (Wafa et al., 1972; Litt, 1988), whereas others present some kind of simple social organization (Osgood, 1989; Rozen, 1973). This group of bees could therefore be useful in understanding the development of social interactions in species whose nests are grouped. For Andrena agilissima our report is the first where interactions among females are well documented, as far as we know. Sharing entrances and burrows could be a useful way to save time in excavating, especially if the rock is hard (Brockmann, 1979). It could also be a way to maintain the entrance under better control: the presence of a bee at the entrance during the day, ready to enter or leave, would increase the defense of that entrance against parasites and predators (Lin and Michener, 1972; Rosenheim, 1990). This can develop a division of labor, as defined in some bees (Abrams and Eickworth, 1981), where a guard is always present while other individuals are foraging. In the case of A. agilissima, bees were not always present at the nest entrance and a well-defined division of labor vis-a-vis guards was never observed. The presence of many bees moving around the entrances did not appear efficient in deterring parasites such as Bombylius (Andrietti et al., 1997) and other flies that were recorded (Disney et al., 1999). Brockmann's assumption (1979), that the tolerance of many females in a nest may be advantageous for saving time in excavating, combined with a tendency for bees to return to the sites where they were born as suggested by Michener (1974), seems the most suitable explanation for Andrena agilissima. Further evidence arises from the results of excavation, in which we observed the earth to be full of burrows and cells, suggesting a long time use of some nest sites. In fact the general arrangement of the external entrances was very constant from 1993 to 1996. Changes in some areas were due mainly to rainfall, rather than to bee activity. The connections among tunnels are so deep and so numerous that we can infer that numerous interactions must take place among females. Interactions

Entrance Sharing and Subterranean Connections in Andrena agilissima

429

Fig. 3. Dissection of 30 cm2 of the nesting surface: (a) cells filled with earth; (b) cells with an unidentified cocoon, probably a parasite; (c) empty cells. 1-6, external openings (see text).

were also observed at the nest entrance, with two or three females waiting to leave and exchanging positions always without aggression. To prevent aggression, there is likely an identification system among nestmates, perhaps involving chemicals. The means by which tunnels are constructed and come into communal use requires further investigation. Andrena agilissima females cohabit, using the same entrances and main tunnels. Each bee makes foraging trips, so it is possible that each provisions only her own cells. The kind of interactions among females could be similar to those

430

Giovanetti, Andrietti, Martinoli, and Rigato

among other Andrenidae species (Danforth, 1991). They could have developed a high tolerance among individuals that facilitates nonaggressive interactions in casual encounters inside the burrows and at the nest entrance, without necessitating a dominance hierarchy and differences in labor. This could be an interesting point for theories on the development of sociality helping to establish other ways to define a social group (as proposed by Michener, 1985), instead of reasoning simply in terms of close relatedness (Hamilton, 1964) as the theories which deal with eusocial insects. As to relatedness, some genetic work on the females using the same entrances would provide information on possible kin relationships. More work must be done on theoretical questions about the evolution of sociality to provide answers on the role played in social behaviors by kin selection and/or the aggregation of nests. ACKNOWLEDGMENTS We thank Drs. Robert Paxton, Rita Cervo, and Diane DeSteven for an earlier comment on this paper, Dr. Pagliano for the identification of the specimens, the students of "Campagna Naturalistica 1994" and "Campagna Naturalistica 1995" for their help in collecting data during the field seasons, and Dr. Jim Dalling for further comments on the paper.

REFERENCES Abrams, J., and Eickwort, G. C. (1981). Nest switching and guarding by the communal sweat bee Agaposlemon virescens (Hymenoptera: Halictidae). Insectes Soc. 28(2): 105-116. Andrietti, F., Martinoli, A., and Rigato, F. (1997). Quantitative data concerning the oviposition of Bombylius fimbriatus Meigen (Dip.: Bombyliidae), a parasite of Andrena agilissima (Scopoli) (Hym.: Andrenidae). Entomol. Rec. 109: 59-63. Ayasse, M., Leys, R., Pamilo, P., and Tengb, J. (1990). Kinship in communal nesting Andrena (Hymenoptera: Andrenidae) bees is indicated by composition of Dufour's gland secretions. Biochem. Syst. Ecol. 18: 453-460. Brockmann, H. J. (1979). Nest-site selection in the great golden digger wasp, Sphex ichneumoneus L. (Sphecidae). Ecol. Entomol 4: 211-224. Danforth, B. N. (1991). Female foraging and intranest behavior of a communal bee, Perdita portalis (Hymenoptera: Andrenidae). Ann. Entomol. Soc. Am. 84(5): 537-548. Danforth, B. N., Neff, J. L., and Barretto-Ko, P. (1996). Nestmates relatedness in a communal bee, Perdita texana (Hymenoptera: Andrenidae), based on DNA fingerprinting. Evolution 50(1): 276-284. Disney, R. H. L., Scanni, B., Scamoni, E., and Andrietti, F. (1999). A new species of scuttle fly (Diptera: Phoridae) whose larvae are kleptoparasites of a bee (Hymenoptera: Andrenidae). Giornale italiano di Entomologia (in press). Hamilton, W. D. (1964). The genetical evolution of social behavior. J. Theor. Biol. 7: 1-16. Iwata, K. (1976). Evolution of Instinct. Comparative Ethology of Hymenoptera, Amerind, New Delhi. Lin, N., and Michener, C. D. (1972). Evolution of sociality in insects. Q. Rev. Biol. 47(2): 131-159. Litt, R. (1988). Observations sur Andrena fulva Schrk. (Apidae, Hymenoptera). Rev. Vervietoise Hist. Nat. Spring: 22-30.

Entrance Sharing and Subterranean Connections in Andrena agilissima

431

Michener, C. D. (1974). The Social Behavior of the Bees, Belknap Press of Harvard University Press, Cambridge, MA. Michener, C. D. (1985). From solitary to eusocial: Need there be a series of intervening species? Fortschritte Zool. 31: 293-305. Michener, C. D., and Rettenmeyer, C. W. (1956). The ethology of Andrena erythronii with comparative data on other species. Univ. Kans. Sci. Bull 37: 645-684. Neff, J. L., and Simpson, B. B. (1997). Nesting and foraging behavior of Andrena (Callandrena) rudbeckiae Robertson (Hymenoptera: Apoidea: Andrenidae) in Texas. J. Kans. Entomol. Soc. 70(2): 100-113. Osgood, E. A. (1989). Biology of Andrena crataegi Robertson (Hymenoptera: Andrenidae), a communal nesting bee. J. N.Y. Entomol. Soc. 97(1): 56-64. Paxton, R. J., and Tengo, J. (1996). Intranidal mating, emergence, and sex ratio in a communal bee Andrena jacobi Perkins 1921 (Hymenoptera: Andrenidae). J. Insect Behav. 9(3): 421-440. Paxton, R. J., Giovanetti, M., Andrietti, F., Scamoni, E., and Scanni, B. (1999). Mating in a primitively social bee, Andrena agilissima (Hymenoptera: Andrenidae). Ecol. Ethol. Evol. (in press). Rode, H. (1962). Solitarbienen als Schadlinge am Mauerwerk. Anz. Schadlingskunde 35: 72-73. Rosenheim, J. A. (1990). Density-dependent parasitism and the evolution of aggregated nesting in the solitary Hymenoptera. Ann. Entomol. Soc. Am. 83(3): 277-286. Rozen, J. G. (1973). Biology notes on the bee Andrena accepta Vierek (Hymenoptera: Andrenidae). J. N.Y. Entomol. Soc. 81: 54-61. Schmiederknecht, O. (1882-1884). Apidae Europeae, Gumperda, Berlin. Visscher, P. K., and Danforth, B. N. (1993). Biology of Callipsis pugionis (Hymenoptera: Andrenidae): Nesting, foraging and investment sex ratio. Am. Entomol. Soc. Am. 86(6): 822-833. Wafa, A. K., Rashad, S., and Moustafa, M. A. (1972). On the nesting habit of Andrena ovatula (K.). Egypt. Dtsch. Ent. Z. 19(IV-V): 303-306. Westrich, P. (1989). Die Wildbienen Baden-Wuertembergs. S.T.: Die Gattungen und Arten, Eugen Ulmer, Stuttgart, pp. 469-470.