Ethology 111, 1116—1125 (2005) 2005 Blackwell Verlag, Berlin
Scraping Sounds Produced by a Social Wasp (Asteloeca ujhelyii, Hymenoptera: Vespidae) Fa´bio S. Nascimento*, Michael Hrncir , Adam Tolfiskià & Ronaldo Zucchi* *Departamento de Biologia, FFCLRP, Universidade de Sa˜o Paulo, Ribeira˜o Preto, SP Brazil; Institute of Zoology, Biocenter, University of Vienna, Vienna, Austria; àDepartment of Pomology and Apiculture, Agricultural University, Krakow, Poland Abstract A distinct behavior performed by workers of the social wasp Asteloeca ujhelyii is described: guard wasps, sitting outside of the nest around the entrance hole, scratch the nest envelope with their forelegs, thereby producing an audible Ôscraping soundÕ. Field observations revealed a relationship between the occurrence of the scraping behavior and the occurrences of wasps entering and leaving the nest. The scraping was experimentally reproduced in the laboratory by moving a dissected wasp’s leg parallel to the plane of the nest envelope. The resulting substrate-borne vibrations propagate very well over a distance of a few centimeters. The amplitude of these vibrations depends on the velocity of the scraping movements. The possible role of the scraping behavior in the waspsÕ communication is discussed. Correspondence: Fa´bio S. Nascimento, Departamento de Biologia, FFCLRP, Universidade de Sa˜o Paulo, Av. Bandeirantes, 3900, 14040-901, Ribeira˜o Preto, SP Brazil. E-mail:
[email protected] Introduction In neotropical swarm-founding wasps, most species live in large and complex colonies, and sophisticated mechanisms of communication have evolved to coordinate the activities among nestmates (Jeanne 1991). Although chemically mediated communication during both alarm recruitment and swarm trail migration have been well studied in epiponine wasps (Naumann 1975; Jeanne 1981; West-Eberhard 1982; O’Donnell et al. 1997), the occurrence and communicative function of sounds or substrate-borne vibrations remains poorly known in almost all taxa. So far, the best-known example of such a Ômechanical communicationÕ was studied in Synoeca surinama, which produces an audible warning signal by vibrating their wings on the nest envelope in response to a disturbance of the nest, (Evans & Eberhard 1970). Moreover, recent studies
Scraping Sounds in a Social Wasp
1117
showed that Parachartergus colobopterus workers produce an alarm signal by tapping their abdomina against the nest carton (Strassmann et al. 1990; Jeanne & Keeping 1995). In several species of Polistes, vibrations produced by body oscillations and drumming against the nest surface are known to be used for communication (Hermann & Dirks 1975; Gamboa et al. 1978; Gamboa & Dew 1981; Pratte & Jeanne 1984; reviewed in Starr 1991; Savoyard et al. 1998). Oscillatory movements of the body are generally performed by dominant females when inspecting brood cells, and occur during interactions between adults and the brood (Gamboa & Dew 1981; Pratte & Jeanne 1984). In Vespa orientalis, audible sounds generated through abdominal vibrations against the comb signal the start of morning activities. Tapping the comb with the tarsi of the mid- and hindlegs occurs in V. tropica during feeding the larvae (Ishay et al. 1974; Ishay 1977). The aim of this paper is to describe the scraping behavior performed by the guard workers of Asteloeca ujhelyii, and to investigate whether activity of foragers affects this scraping behavior. Moreover, by moving a dissected wasp’s leg on the nest envelope, we imitated the behavior and studied the transmission of substrate-borne vibrations on the nest envelope generated by the scraping action.
Materials and Methods Species
Asteloeca ujhelyii, first described as Polybia ujhelyii by Ducke (1909), is a medium-sized wasp (body length, 11–12 mm), found in Amazonian Bolivia, French Guiana, and Brazil (Carpenter et al. 2004). Dome-like astelocyttarous nests are small to medium-sized (230–450 cells) and covered with a fragile-glossy envelope (Wenzel 1998). Colonies of A. ujhelyii can contain from a dozen to a few hundred adult individuals. The nest entrance is like an upward collar, which allows the passage of a single wasp at a time. Additional information on taxonomy and biology of Asteloeca species has recently been reported in Carpenter et al. (2004) and Nascimento et al. (2004). Study Colony, Scraping Recordings, and Worker Specialization
Field studies were conducted between Jul. and Aug. 2001 at Fazenda Catuaba (1004¢36¢¢S 6737¢40¢¢W), 20 km east of Rio Branco, Acre State, Brazil. A mature colony (labeled as colony A) was the main subject in this study. Fifty-eight adult females and the nest were collected after the experiments, but 116 females had been individually marked and observed throughout the observations. Additional observations of scraping behavior without video recordings were made with another young nest (colony B: 52 individuals) located at a different site (Xapuri – Acre State, 1049¢S 6822¢W). Voucher specimens are deposited at the collections of Departamento de
1118
F. S. Nascimento, M. Hrncir, A. Tolfiski & R. Zucchi
Biologia, Universidade de Sa˜o Paulo, Ribeira˜o Preto, Brazil and American Museum of Natural History, NY, USA. The scraping behavior and resulting sound were recorded with a digital Hi8 Sony camcorder (TRV110) equipped with a microphone. Recording sessions were taken either during morning or afternoon periods resulting in a total 60 h of field observations. The video recordings were analyzed by detecting the events of foraging wasps entering and leaving the nest, and recording the time spent by other wasps with scraping within 10 s intervals preceding and following each of these events. Additionally, the amount of scraping was determined during the remaining 10 s intervals. If scraping occurred, the number of wasps involved was recorded. To test whether the occurrence of the scraping behavior was a response to the transit of wasps, a Spearman’s correlation test was used to verify the association between the number of scraping wasps and the number of exiting and entering workers per sequence of behavior. To avoid ambiguous analysis, we only considered entrance/exit-scraping behavior as a sequence when they occurred within 3 s of observation. All behaviors performed by marked wasps on the nest envelope were scanned and recorded. Worker specialization on scraping behavior was verified by comparing the distribution of the number of repetitions of the behavior by individuals against the expectations of a Poisson’s distribution (Visscher et al. 1999). We compared the frequency of scraping performed by workers facing the nest entrance (named scrapers) and all other wasps sitting on the nest envelope. In addition, during 2 d we observed the performance of the following outside nest tasks: ÔscrapingÕ, ÔfanningÕ, Ôreceiving preyÕ, and Ôreceiving liquidÕ. Only workers engaged in each behavior on at least four occasions during a given day were included in the analysis. Individual task specialization was calculated using the Shannon–Wiener index (Lehner 1979; O’Donnell & Jeanne 1990): X H ðxÞ ¼ pðxÞ log2 pðxÞ where p(x) is the proportion of a worker’s effort devoted to a given task (x). Artificial Vibration Signals and Envelope Properties
To imitate the scraping action in the laboratory, we fixed a dissected wasp’s foreleg on the tip of a vibrator (V106, Ling Dynamic Systems, Royston, Hertfordshire, UK). Only the tarsal claws touched and scraped on the substrate near to the nest entrance. The movement of the leg was parallel to the envelope. Different amplitudes of the leg’s movement resulted in different scraping velocities (15.1, 8.0, and 3.8 mm/s). The vibrations (perpendicular to the surface of the envelope, henceforth: vertical vibrations) elicited by the scraping were measured at five different points on the envelope, using a portable laser vibrometer (Polytec PDV 100, Walbronn, Germany) connected to a notebook. The five points were: close to the source of the scraping (0 cm, reference point) and at distances of 1, 2, 3 and 4 cm from the scraping point.
Scraping Sounds in a Social Wasp
1119
Signal Analysis
To analyze the scraping sounds recorded in the field, and the substrate-borne vibrations generated in the laboratory experiments, we used the sound-analyzing computer programs spectra pro 3.3 and sound forge 4.0, respectively.
Results Scraping Behavior
During daytime, there was a group of 8–37 workers (x SD ¼ 17.73 ± 12.47 wasps, n ¼ 3 nests, 24 h/colony) sitting on the outer surface of the A. ujhelyii nests. Most of the workers remained almost completely inactive but one to four (x SD ¼ 2.2 ± 1.14 wasps, n ¼ 3 nests, 24 h/colony) of them frequently performed scraping behavior. This distinct behavior consisted of workers repeatedly moving their forelegs back and forth while the tarsal claws had contact with the envelope surface. The performing scraping behavior was positioned close to the nest entrance and facing it. Occasionally, they inserted their antennae into the entrance hole. The number of scraping workers during periods of intense activity (measured by the number of wasps leaving or entering the nest per hour) tended to be higher than during periods of foraging inactivity (rs ¼ 0.71, p < 0.0001; y ¼ 3.51 + 1.71x; n ¼ 27 h). The scraping time was significantly shorter during the 10 s intervals without activity (2.9 ± 3.82 s; n ¼ 406 intervals; Fig. 1a) than during intervals preceding and following an event of wasps entering (7.7 ± 3.03 s; n ¼ 106 intervals; Fig. 1b; Mann–Whitney test: Z ¼ 10.62, p < 0.001) or leaving the nest (6.9 ± 3.53 s; n ¼ 114 intervals; Fig. 1c; Mann–Whitney test: Z ¼ 9.33, p < 0.001). However, there was no difference between the intervals with wasp entering and the intervals with wasp leaving the nest (Mann–Whitney test: Z ¼ 1.66, p ¼ 0.096, n1 ¼ 106, n2 ¼ 114). During the intervals preceding events of a wasp entering the nest, the time of scraping was shorter than during intervals following the event (x SD: 7.0 ± 3.40 and 8.4 ± 2.44 s, respectively; Mann–Whitney test, Z ¼ )2.14, p < 0.05, n1 ¼ n2 ¼ 53). Moreover, during intervals preceding the forager’s departure from the nest wasps scraped significantly less than after the event (x SD: 6.0 ± 3.81 and 7.7 ± 3.04 s, respectively; Mann–Whitney test, Z ¼ )2.93, p < 0.01, n1 ¼ n2 ¼ 57). Scrapers were highly specialized workers. They devoted 95% ± 4.3 (n ¼ 26) of the time observed to this task. The Shannon–Wiener specialization index was 0.7 ± 0.27 (n ¼ 26), supporting specialization among scrapers. Other wasps on the nest occasionally scraped the envelope, but the distributions were far different from a random Poisson’s distribution (Fig. 2a. Colony A: D ¼ 0.73, p < 0.01, n ¼ 76; Colony B: D ¼ 0.63, p < 0.01, n ¼ 22). The group of specialized workers scraped significantly more than the other wasps sitting on the nest envelope (Colony A: t-test ¼ 7.74, df ¼ 74, p < 0.001; Colony B: t-test ¼ 7.97,
1120
F. S. Nascimento, M. Hrncir, A. Tolfiski & R. Zucchi
Fig. 1: Distributions of scraping times during: (a) 10 s intervals without wasps entering or leaving the nest, (b) intervals preceding and following an event of wasp entering the nest and (c) intervals preceding and following an event of wasp leaving the nest. There was no scraping in more than 50% of intervals without wasp entering or leaving the nest (a) in contrast to less than 10% of intervals with wasp entering or leaving the nest (b and c)
df ¼ 49, p < 0.001). There were overlapping of tasks functions performed by workers on the nest, but scraping workers were specialized in their function at least 3 d within the period of observation. In general, workers engaged in receiving food and water before starting as scrapers, and in foraging activities later (F.S. Nascimento, unpub. data).
1121
Scraping Sounds in a Social Wasp 60 Colony A Colony B
Number of workers
50
40
30
20
10
0 0
10
20
30
40
50 60 70 80 Scraping per individual
90
100
110
120
Fig. 2: Number of wasps that scraped the nest envelope during 3 d of observation in the colony A (black bars) and colony B (white bars)
Fig. 3: Example of the scraping sound produced by three workers of A. ujhelyii. The sonogram shows the broad frequency spectrum of the scraping noise. Vertical lines indicate the entire period of the scraping, DF indicates the dominant frequency of the spectrum that can be attributed to the scraping. Note the noisy background that impeded a more detailed analysis of the scraping sound
Scraping Noise
The scraping behavior of the wasps resulted in an audible, broad-banded frequency sound (Fig. 3) which could be perceived by human ear even from a distance of 2 m from the nest. A dominant frequency of about 1430 Hz (1428 ± 69 Hz, n ¼ 10; Fig. 3) contained within the broad-banded noise could be attributed to the scraping.
1122
F. S. Nascimento, M. Hrncir, A. Tolfiski & R. Zucchi
Artificial Scraping and Transmission Properties of the Nest Envelope
In the laboratory, we imitated the scraping by moving a wasp’s leg (mounted onto a vibrator) parallel to the nest envelope with only the tarsal claws touching the substrate. Scraping the envelope this way near the nest entrance resulted in vertical vibrations of the substrate (Fig. 4a) which could be measured at any point of the envelope. Independent of the velocity of the leg’s movement, the substrateborne vibrations elicited by the scraping movement showed a main frequency of 700 Hz and its harmonics (1400 and 2100 Hz) (Fig. 4b). However, the velocity amplitude of the vertical substrate vibrations depended on the velocity of the leg’s movement. At the scraping point (0 cm, reference point), the velocity of the
Fig. 4: Artificial scraping. A wasp’s leg attached to a vibrator was used to imitate the scraping behavior observed in the field. (a) The movement of the leg parallel to the nest envelope (top) resulted in vibrations perpendicular to the substrate (bottom). (b) The main frequency (MF) and its harmonics (H1, H2) of the substrate vibrations elicited by the scraping were similar at all studied scraping velocities (15.1 mm/s, open triangles; 8.0 mm/s, open squares; 3.8 mm/s, filled circles). The amplitude (dB) is given in relation to the maximum measured (0 dB). (c) The nest envelope amplified frequencies around 500 Hz. The relative transmission (dB/cm) was calculated as the frequency amplitude measured at a given distance in relation to the amplitude of this frequency at the point of scraping (reference, 0 dB). Symbols are same as in (b).
1123
Scraping Sounds in a Social Wasp
Table 1: Substrate-borne vibrations elicited by scraping. Artificial scraping movements of three different movement velocities (15.06, 7.98, and 3.76 mm/s peak–peak) induced vibrations of the nest envelope. The velocity amplitude (peak–peak) of these substrateborne vibrations was measured at different distances from the point of scraping (0 cm) near the nest entrance Scraping velocity (mm/s)
Response (mm/s) 0 cm
1 cm
2 cm
3 cm
4 cm
15.06 7.98 3.76
6.73 3.52 1.73
5.10 2.39 1.14
4.20 1.97 1.03
2.48 1.13 0.55
1.62 0.79 0.37
vertical vibrations was about half of the scraping velocity, and the vibration amplitude decreased with increasing distance to the scraping point (Table 1). In order to study which frequencies of the substrate vibrations elicited by the scraping are best transmitted across the nest envelope, we analyzed the envelope’s transmission properties. Indifferently of the velocity of the scraping movement, vibration frequencies of around 500 Hz were amplified across the envelope (Fig. 4c). The transmission of frequencies above 700 Hz differed between the different scraping velocities. However, the transmission of frequencies below 700 Hz was similar at all studied velocities of the leg’s scraping movements (Fig. 4c). Discussion In the present study we described for the first time a behavior performed by guards of A. ujhelyii wasps. The guards occasionally produced audible sounds by scraping the nest envelope with their forelegs. We clearly demonstrated that this behavior occurred mainly when other wasps were entering or leaving the nest. The scraping started before a wasp arrived or left the nest and continued for some time afterwards. The scraping before a wasp entered the nest was triggered by clearly visible returning foragers. However, it is more difficult to explain how the scrapers, sitting outside the nest, perceived foragers approaching the nest entrance from inside. Although we do not know what happens inside the nest we assume that scrapers receive a signal just before foragers go out of the nest. We demonstrated that substrate-borne vibrations are elicited when a dissected wasp’s leg is moved repeatedly along the surface of the nest envelope. Unfortunately, we had no opportunity to record the velocity of the scraping movements by the wasps in the field. However, our experiments demonstrate that both the main frequency of the substrate vibrations generated by scraping and the amplification of frequencies below 700 Hz by the nest envelope were independent from the velocity of the scraping movement (Fig. 4). Moreover, artificial movements from the leg with 8.0 mm/s produced a similar frequency found in the field, thus it could indicate that in nature wasps move their legs
1124
F. S. Nascimento, M. Hrncir, A. Tolfiski & R. Zucchi
very close to this velocity during the scraping. In hymenopterans the ability to detect substrate-borne vibrations has been predominantly attributed to the subgenual organ, a chordotonal organ in the proximal part of the tibia of each leg (Autrum & Schneider 1948). This vibration sensitive organ reacts to vibrations in the axial direction of the tibia. Hence the sensory cells are predominantly stimulated by vertical vibrations (perpendicular to the substrate on which the animal stands) (Rohrseitz & Kilpinen 1997). The subgenual organ of honey bees (Apis mellifera) is most sensitive to substrate-borne vibrations at frequencies of 500 Hz (Kilpinen & Storm 1997). Interestingly, the nest envelope of A. ujhelyii amplifies vibration frequencies around 500 Hz (Fig. 4c). Unfortunately, no detailed study on the vibration sensitivity in wasps is available. However, our findings that substrate vibrations are elicited by scraping movements, and that the nest envelope transmits and even amplifies frequencies that could be relevant for the animal’s perception, give reason to suggest that the scraping could well be used as signal in these wasps. Previous studies in Vespidae showed that the vibrations transmitted via the nest envelope can provide information about food availability (Pratte & Jeanne 1984 and references therein) and about the presence of enemies. In fact, some evidence in Synoeca spp. and P. colobopterus indicate that a generalized warning signal can be transmitted across the nest envelope in response to human disturbances, army ant attacks or parasitoids (Chadab 1979; West-Eberhard 1982; Strassmann et al. 1990; Jeanne & Keeping 1995). We assume that scraping behavior in A. ujhelyii could be associated with colony defense as well. It is performed by guards and occurs mainly when wasps pass the nest entrance. We hypothesize that the small nest entrance is guarded both inside and outside the nest envelope. The vibration produced during scraping behavior can inform the guards on the other side of the envelope whether that the wasp passing the nest entrance is a nestmate or an enemy. This would make defense against enemies more effective. However, more data about the behavior of wasps during colony defense are needed to verify this hypothesis. Acknowledgements Part of this study was presented in the XXVIII International Ethological Conference. The authors thank to Jim Carpenter and two anonymous reviewers for helpful comments on the manuscript, and Jim Hunt for stimulating discussion on the scraping behavior. Financial support to F.S.N. was provided by Fapesp (Procs. 02/12540-9 and 03/10663-5).
Literature Cited Autrum, H. & Schneider, W. 1948: Vergleichende Untersuchungen u¨ber den Erschu¨tterungssinn der Insekten. Z. vergl. Physiol. 31, 77—88. Carpenter, J. M., Nascimento, F. S., Mateus, S., Noll, F. B. & Kojima, J. 2004: A revision of the genus Asteloeca (Hymenoptera: Vespidae; Polistinae). Am. Mus. Nov. 3427, 1—12. Chadab, R. 1979: Early warning cues for social wasps attacked by army ants. Psyche 86, 115—123. Ducke, A. 1909: Deux vespides nouveax du Muse´um National Hungrois. Ann. Hist.-Nat. Mus. Natl. 7, 626—627.
Scraping Sounds in a Social Wasp
1125
Evans, H. E. & Eberhard, M. J. 1970: The Wasps. University of Michigan Press, Ann Arbor. Gamboa, G. J. & Dew, H. E. 1981: Intracolonial communication by body oscillations in the paper wasp, Polistes metricus (Hymenoptera: Vespidae). J. Kansas Entomol. Soc. 54, 153—155. Gamboa, G. J., Heacock, B. D. & Wiltjer, S. L. 1978: Division of labor and subordinate longevity in foundress associations of the paper wasp, Polistes metricus (Hymenoptera: Vespidae). J. Kansas Entomol. Soc. 51, 343—352. Hermann, H. R. & Dirks, T. F. 1975: Biology of Polistes annularis (Hymenoptera: Vespidae). I. Spring behavior. Psyche 82, 97—108. Ishay, J. 1977: Acoustical communication in wasp colonies (Vespinae). Proc. 15th Int. Cong. Entomol. 2, 406—435. Ishay, J., Motro, A., Gitter, S. & Brown, M. B. 1974: Rhythms in acoustical communication by the oriental hornet, Vespa orientalis. Anim. Behav. 22, 741—744. Jeanne, R. L. 1981: Chemical communication during swarm emigration in the social wasp Polybia sericea (Olivier). Anim. Behav. 29, 102—113. Jeanne, R. L. 1991: The swarm-founding Polistinae. In: The Social Biology of Wasps (Ross, K. G. & Matthews R. W., eds). Comstock, Ithaca, pp. 191—231. Jeanne, R. L. & Keeping, M. G. 1995: Venom Spraying in Parachartergus colobopterus: a novel defensive behavior in a social wasp (Hymenoptera: Vespidae). J. Insect Behav. 8, 433—442. Kilpinen, O. & Storm, J. 1997: Biophysics of the subgenual organ of the honeybee, Apis mellifera. J. Comp. Physiol. A 181, 309—318. Lehner, P. N. 1979: Handbook of Ethological Methods. Garland STPM Press, New York, NY. Nascimento, F. S., Tannure-Nascimento, I. C. & Zucchi, R. 2004: Behavioral mediators of the cyclical oligogyny in the Amazonian swarm-founding wasp Asteloeca ujhelyii (Hymenoptera, Vespidae, Epiponini). Insect. Soc. 51, 17—23. Naumann, M. G. 1975: Swarm behavior: evidence for communication in social wasps. Science 189, 642—644. O’Donnell, S. & Jeanne, R. L. 1990: Forager specialization and the control of nest repair in Polybia occidentalis Oliver (Hymenoptera: Vespidae). Behav. Ecol. Sociobiol. 27, 359—364. O’Donnell, S., Hunt, J. H. & Jeanne, R. L. 1997: Gaster-flagging during colony defense in neotropical swarm-founding wasps (Hymenoptera: Vespidae, Epiponini). J. Kansas Entomol. Soc. 70, 175—180. Pratte, M. & Jeanne, R. L. 1984: Antennal drumming behavior in Polistes wasps (Hymenoptera: Vespidae). Z. Tierpsychol. 66, 177—188. Rohrseitz, K. & Kilpinen, O. 1997: Vibration transmission characteristics of the legs of freely standing honeybees. Zoology 100, 80—84. Savoyard, J. L., Gamboa, G. J., Cummings, D. L. D. & Foster, R. L. 1998: The communicative meaning of body oscillations in the social wasp, Polistes fuscatus (Hymenoptera, Vespidae). Insect. Soc. 45, 215—230. Starr, C. K. 1991: The nest as the locus of social life. In: The Social Biology of Wasps (Ross, K. G. & Matthews, R. W., eds). Comstock, Ithaca, NY, pp. 520—539. Strassmann, J. E., Hughes, C. R. & Queller, D. C. 1990: Colony defense in the social wasp, Parachartergus colobopterus. Biotropica 22, 324—327. Visscher, P. K., Shepardson, J., McCart, L. & Camazine, S. 1999: Vibration signal modulates the behavior of house-hunting honey bees (Apis mellifera). Ethology 105, 759—769. Wenzel, J. W. 1998: A generic key to the nests of hornets, yellowjackets, and paper wasps worldwide (Vespidae: Vespinae, Polistinae). Am. Mus. Nov. 3224, 1—39. West-Eberhard, M. J. 1982: Communication in Social Wasps: Predicted and Observed Patterns, with a Note on the Significance of Behavioral and Ontogenetic Flexibility for Theories of Worker ÔAltruismÕ. In: La communication chez les socie´te´s d’insectes (De Haro, A. & Espadaler, X., eds). Universidad Auto´noma de Barcelona, Bellaterra, pp. 13—36. Received: December 10, 2004 Initial acceptance: March 22, 2005 Final acceptance: April 8, 2005 (S. Forbes)
