Observations on the social wasp Ropalidia montana from peninsular India

J. Biosci., Vol. 17, Number 1, March 1992, pp 1 14. Printed in India Observations on the social wasp Ropalidia montana from peninsular India ROBERT L JEANNE and JAMES H HUNT* Department of Entomology, University of Wisconsin, Madison, Wisconsin 53706, USA *Department of Biology, University of Missouri St. Louis, St. Louis, Missouri 63121, USA MS received 1 July 1991; revised 3 February 1992 Abstract. Four colonies of Ropalidia montana collected in August in the Nilgiri Hills of southern India had adult populations of 32,000-61,000. Queens represented 0 46 1 40% of the populations. All colonies contained males, but in widely varying proportions (1 47 27. 00%). The large adult populations and the fact that the nests were largely filled with brood in all stages of development indicate that the colonies were in a stage of active growth in this season. In December, however, colonies have been reported to have smaller adult populations and little brood. Thus brood production appears to be seasonal in southern India. Predation by the hornet Vespa tropica was observed in 4 of 31 active colonies. Ropalidia montana adults were defenseless in the face of these depredations, which appear to continue at a low level for days or weeks on a given nest. Keywords. Hymenoptera; social wasps; colony size; colony cycle; colony composition; predators; inquilines; Ropalidia montana; Vespa tropica. 1. Introduction The roughly-750 species of social wasps in the vespid subfamily Polistinae fall into two behaviourally defined groups, the independent-founders and the swarmfounders (Jeanne 1980, 1991). Colonies of independent-founding species are initiated by one or several inseminated queens, independently of any workers. Soon after colony founding, one of the founding queens typically becomes the sole egg layer (monogyny). Reproductive dominance is based largely on direct physical attacks by the queens. Colonies of these wasps are small (usually < 100 adults) and shortlived. In the swarm-founding species, a colony is founded by a large number of workers accompanied by a smaller number of queens (polygyny). Reproductive dominance may involve pheromones (West-Eberhard 1977). Colony size in this group ranges from a few dozens to many thousands (Jeanne 1991). Independent founding characterizes 5 genera, with distributions throughout the tropics and well into temperate zones. Swarm founding is the exclusive initiation mode in 23 genera in the subfamily and is overwhelmingly tropical in distribution. Polistine genera belong to either one or the other group, except for the Old World genus Ropalidia, which includes both types of species. If, as is currently believed (West-Eberhard 1978; Carpenter 1991), swarm-founding is the derived state, then swarm-founding behaviour evidently evolved in Ropalidia independently of its origin(s) among the more numerous New World swarm-founding genera of social Vespidae (Jeanne 1980; Carpenter 1991). Ropalidia thus has a potentially Corresponding author. 1

2 Robert L Jeanne and Jame s H Hunt important role in the eventual understanding of the origins of the swarming mode of colony founding and its associated social characteristics (Jeanne 1980), yet it is poorly known biologically. Only a few independent-founding species have been studied in the field (Darchen 1976; Richards 1978; Gadagkar and Joshi 1983, 1984; Ito 1983, 1985; Yamane 1986; Wenzel 1987). Swarm-founding species, which are a minority in the genus, are even less well-known. Ropalidia montana Carl is a swarm-founding species found only in peninsular India, with collection records from the states of Karnataka, Kerala and Tamil Nadu (Das and Gupta 1989). R. montana first entered the scientific literature when Carl (1934) described the species and its unique nest architecture. The only recent paper to deal with the species documents its queen/worker dimorphism (Yamane et al 1983). Recently we had the opportunity to study R. montana in the field. We worked at a site close to where Carl (1934) and Yamane et al (1983) collected their colonies, but we worked at a different time of year. We compare the size and composition of our colonies with those studied by Yamane et al (1983). We also provide the first assessment of the range of colony sizes based on a survey of nests and the first behavioural and life history notes since the brief speculations by Carl (1934). 2. Materials and methods The study was conducted during mid-august, 1990, in the Mudumalai Wildlife Sanctuary, which lies between latitudes 11 32' and 11 42' N and longitudes 76 20' and 76 40' E in the state of Tamil Nadu, near where the boundaries of the three southern states of Kerala, Karnataka and Tamil Nadu meet. The sanctuary is located in an extensive belt of forest on a 1000 m plateau at the base of the Nilgiri Hills. Annual rainfall varies from 200 cm at the western end of the sanctuary to 50 cm in the east. Nests of R. montana were seen only in the wet, western half of the sanctuary, although a few foragers of R. montana were taken at flowers during a brief visit to the drier, eastern region, near the village of Masinagudi. Nests were located with the help of local naturalists. Each was categorized for approximate size. Identification was made of the plant on which it was located, and occurrence of raiding activity by Vespa tropica was checked using binoculars. Four of the colonies were collected. Two of these (nos 4 and 13) were high in trees and had to be taken by first cutting the supporting branch. Because this procedure caused large numbers of the adult population to evacuate, we propped each nest near its original position for approximately an hour after cutting, enabling many of the adults that had fled to return to the nest. Each nest was then carefully lowered into a plastic bag, along with most of its adult population. The remaining two nests (nos 2 and 12) were lower and could be taken by first trimming surrounding branches from the nest, then bagging it the next day with relatively little loss of adults. Because all colonies were collected during daylight hours, undetermined numbers of foraging workers were not collected. Following collection, adults of all four colonies were immobilized by placing cotton soaked in ether or chloroform into the bag with the nest. Most of the anaesthetized adults could then be shaken out of the nest; the remainder were retrieved during dissection of the nest. To estimate cell number, we separated the combs and traced their outlines on

Ropalidia montana 3 paper. The tracings were then numbered, carefully cut out along the trace lines, and weighed. Calibration was made by taking the weight of 100 cm 2 of paper. The cells in both one and four cm 2 samples of comb (5 of each) were counted, yielding an estimate of 14 cells/cm 2, which was used to calculate the estimated total cell number for each nest. Adult population of each colony was estimated volumetrically. The volume of the entire collected population was measured using 50 and 500 ml beakers. The numbers of individual wasps in four 50 ml samples were determined by direct count, and the mean of the four counts was used to convert the total volume into an estimation of the population for the whole colony. Frequencies of males and females in each colony were estimated based on examination under 10 magnification of a sample of 200 to 2,800 individuals. Each sample was selected at random after thorough mixing of the entire (anaesthetized) adult population. We found that males were recognized most easily by the rounded apex and slight median depression to the terminal sternite. Preliminary examination showed that queens had large, distended gasters and that the first gastal tergite was brownish, while in workers this was black. After sorting on these criteria, the subsample of queen-like females was dissected to check ovary condition. Whenever caste status of a female was in doubt, that female was dissected to check ovary condition. 3. Results 3.1 Nest sites, sizes and host plants We located 31 active colonies and 2 abandoned nests of R. montana in the sanctuary. All were on woody vegetation at heights of 2 m or more (table 1). The most commonly used substrate was Xeromphis spinosa (Rubiaceae), a low understorey tree with slender, spiny twigs. Second most commonly used was bamboo (Bambusa arundinacea, Gramineae). In all cases, nests were attached to slender horizontal or pendant twigs. Trees with thicker twigs, such as teak (Tectona grandis, Verbenaceae), despite their abundance in the habitat, were never used. Table 1. Size and host plant data for nests of R. montana observed in Mudumalai Wildlife Sanctuary.

4 Robert L Jeanne and James H Hunt Most nests were in good condition; a few were missing small pieces of envelope from the lower parts of the nest. Most nests had moderate numbers of wasps scattered more or less evenly over the nest envelope (figure 1). Nest no. 12, however, was considerably more densely covered with adults (figure 2). The reason for this was not clear; the colony was not in an alarmed state, although it did contain the largest population of the four we collected. Figure 1. Nest no. 2 in situ. Numerous R. montana workers are scattered over the surface. Near the center of the nest is a single Vespa tropica worker. Five holes made by V. tropica are visible in the nest envelope. 3.2 Size and composition of the collected colonies Colony adult populations in the four collected nests ranged from 32,000 to 61,000 (table 2). The representation of queens was consistently low (0 46 1 40% of male + female population), while that of males was highly variable from colony to colony (1 47 27 00%). In nest no. 4 we were able to compare the per cent males among eclosing adults with that of the adult population at the time of collection. A sample of 220 eclosing adults contained 2 males (0 91%). Fifty pupae were removed from near the center of a large comb, where the brood pattern was irregular (mixed pupae and larvae of

Ropalidia montana 5 Figure 2. Nest no. 12 in situ. Envelope is densely covered with R. montana workers. Table 2. Collection data and statistics for R. montana colonies collected at Mudumalai (sample sizes in parentheses). *Crude estimate; does not include workers away on foraging trips.

6 Robert L Jeanne and James H Hunt various sizes) and sexed; 2 (4%) were males. An uncounted number (more than 50) of pupae from the margins of the same nest comb where the brood pattern was uniform (distinct annuli of same-age larvae and pupae) were also sexed; all were female. Approximately 30 pupae were sexed from regions of uniform-age brood at the margins of large combs from both nest nos 12 and 13; all were female. Together these samples yield a male percentage of < 1 14% among eclosing adults of nest no. 4, compared with 27% in the standing adult population of the same colony. 3.3 Nests and brood The 33 censused colonies spanned a large range of sizes. The smallest nest was spherical and approximately 15 cm in diameter. Larger nests were ovoid, with the broad end uppermost. The largest exceeded 50 cm in height. The envelope consisted of from one to several layers of light, thin, yet tough and flexible, carton. In places, the inner layers of envelope had ragged, free upper edges. Assuming that envelope is constructed from the top down, as in most other paper wasps, this suggests that these inner layers had been partially chewed away. In some nests the light-tan-coloured envelope covering most of the nest was attached to the surface of darker layers remaining at the very top of the nest (figure 3). This darker colour was apparently due to large amounts of oral secretion added to its surface. We found spot attachments of envelope to the combs (but cf. Carl 1934). Figure 3. Top of nest showing darkened carton surrounded by lighter, newer carton attached to its surface. Envelope partially removed. Combs of most nests were parallel to one another in the horizontal plane. In nest no. 12, however, the plane of each comb was built at a slight angle with respect to the one above it (figure 4). Over the entire height of the nest these angles accumulated to a 40 difference between the planes of the uppermost and lowermost

Ropalidia montana 7 Figure 4. Nest no. 12 with envelope removed, showing increasingly angled combs toward the bottom. combs. The lowest combs were horizontal when the nest was in situ. A similar pattern, with about a 40 change in plane of nest comb orientation, was also seen in nest no. 4. The angular difference between the planes of upper and lower combs in nest no. 2 was approximately 5. There was no change in the plane of orientation of combs in nest no. 13. Estimates of the number of cells in the four collected nests ranged from 59,000 to 145,000 (table 2). Notes were taken on the pattern of occurrence of brood in nest nos 4 and 12; both were similar in major respects. The uppermost 5 comb layers were devoid of brood in nest no. 4; in nest no. 12 the first 5 combs had only a few scattered larvae of various sizes and very few pupae. All lower comb layers in both nests were filled with brood. In the large combs at the center of each nest the pattern of brood distribution was of distinct annuli of pupae, with each annulus bordered distally by large larvae and proximally by eggs (figure 5). Cells beyond the outermost annulus lacked the windowed bottoms (figure 6) that indicate that they had been opened by workers for removal of the meconium from the hindgut of the prepupa, then resealed with a clear oral secretion. From an inner annulus of pupae to the annulus of pupae next most distal to it (or to the comb margin), brood stage decreased uniformly from large larvae through small larvae to eggs. There were

8 Robert L Jeanne and James II Hunt Figure 5. Comb from center of a nest, showing concentric annuli of white-capped pupal cells. Successive generations of brood are numbered from the periphery of the comb inward. Figure 6. Back (upper) surface of a comb showing cells with windowed bottoms, indicating that they had been opened by workers for removal of the meconium from the hindgut of the prepupa, then resealed with a clear oral secretion. Non-windowed cells along the edge of the comb contain brood that has not yet pupated.

Ropalidia montana 9 three distinct annuli of pupae in the largest combs, with pupae scattered within the innermost annulus. The lowest 5 combs of both nests contained pupae only in the center of each comb, with larvae decreasing in size toward the margins; cells at the margins contained eggs. The brood pattern in nest nos 2 and 13 was similar to that in nest nos 4 and 12. In nest no. 13 the number of cells having neither brood nor eggs was estimated to have been less than 5% of the total cell number. Honey (or stored nectar) was found in all four nests. The honey was present in two forms. A clear liquid filling nest cells to one-half or two-thirds of their depth was found in fewer than 50 nest cells in the upper comb layers of each colony. An amber-coloured, viscous honey was present as droplets on the walls of nest cells near the cell rim, both in cells lacking brood in the upper comb layers and in cells with eggs or small larvae in the central regions of combs with annuli of pupae. Comb regions where brood was distributed uniformly by size had no honey. Cells with honey droplets represented no more than 1% of the total number of nest cells. Twenty-two larvae in nest no. 13 were sampled for trophallactic saliva by gently probing their mouthparts with a microliter pipet. The volume of saliva obtained ranged from none (n =3) to 1 5µl (n= 1); the mean saliva volume of the 22 larvae was 0 75 ± 0 47µl 3.4 Condition of the queens There was no overlap in ovary condition between queens and workers. The ovaries of queens consistently contained 2 4 opaque eggs in each ovariole, while workers' ovaries were filamentous and lacked any swellings. The spermathecae of randomly sampled queens were all full (n=9). 3.5 Predators and inquilines Four of the 31 colonies we found were under attack by the hornet V. tropica. Nests being raided had one or more holes (2 4 cm diameter) in the envelope through which the hornets passed to reach the brood (figure 1). Usually one or two hornets were visible, either flying around the nest, walking on the envelope, or entering or exiting through one of the holes (figure 1). The hornets showed no interest in the adult R. montana during their visits to the nest. (Dr. John Wenzel, observing in the same area two weeks earlier, saw hornets crushing the heads of adult R. montana on the outside of a nest.) Hornets entered the nest to reach the brood combs, where they chewed down the cell walls to get at the pupae and larvae. We collected and examined two of the nests being raided (nos 1 and 2). A single adult V. tropica was inside nest no. 2 at collection; there were none inside no. 1. In each nest the cells over broad areas of comb surface had been chewed down and were missing brood. Neither of the nests showed any evidence that R. montana had attempted to repair the damage. None of the three colonies we collected that were not under attack (nos 4, 12 and 13) showed any sign in the combs or envelope of Vespa-like damage that had once existed and been repaired, suggesting to us that these colonies had not previously been attacked. R. montana appeared to be completely defenseless against raids by V. tropica. Adult residents of the nest merely stood aside when a hornet landed on or walked

10 Robert L Jeanne and James H Hunt on the surface of the nest (figure 1). When a hornet hovered near the nest, the R. montana adults on the envelope nearby jerked in unison, apparently in response to the vibration and/or air movement. When colony no. 2 was first observed (August 13) several hornets were visiting the nest. A large number (many thousands) of R. montana adults had left the nest and were clustered on foliage within a metre of the nest (figure 7). When next observed (August 16) these wasps were gone. Then on August 19, when the colony was collected, there was a smaller number (perhaps 2,000) of adults again clustered on nearby leaves. A sweep-netted sample of these contained both males and females. The colony was being raided by V. tropica during the entire observation period, but it is not known whether this clustering behaviour was in response to these raids. Figure 7. Adults (males and females) of colony no. 2 clustered on leaves adjacent to the nest. V. tropica raids last at least for several days and possibly for several weeks. When we first saw colony no. 2 on August 13 the raid was well-established (several holes already existed in the envelope), yet raiding was still in progress when we collected the nest on August 19. There were few brood left in the nest when we collected it, so we suspect that the raids persist until virtually no brood remain.

Ropalidia montana 11 Naturalists familiar with the sanctuary knew the location of only one colony of V. tropica. This nest, 30 40 cm in length, was under a concrete bridge approximately 1 km from the nearest known R. montana nest, and several km from the most distant known nest. It is, of course, likely that undiscovered V. tropica nests also existed in the area. We discovered four live adults of the beetle Campsiura javanica (G and P) (Scarabaeidae: Cetoniinae) inside nest no. 13. Two of the beetles were found amongst the debris of envelope and comb material as we dismantled the nest for examination. The two remaining beetles were found between adjacent layers of comb in the upper portion of the nest. We could find no evidence that the beetles had been feeding on the brood. A more detailed account will be published elsewhere. Ten unidentified dipteran puparia (4 6 mm in length) were removed from the cells of nest no. 4. There was no sign of predation by vertebrates on any of the active colonies we observed. 3.6 Swarm emigration After passersby destroyed the nest with rocks at midday on August 16, the adult population of colony no. 1 evacuated and clustered on the leaves and twigs of a nearby branch. The swarm was still present in the morning of August 17. By afternoon, however, it was gone. We searched nearby trees but failed to locate it. Similarly, a swarm left behind when their nest had been collected by Dr John Wenzel on the morning of August 1, 1991 had within 5 h formed a small cluster at the nest site and a much larger cluster approximately 7 meters away. By the next afternoon both clusters were gone (J Wenzel. personal communication). 4. Discussion With the exception of a single colony of Agelaia vicina collected in the state of São Paulo, Brazil, whose adult population was estimated at 1 3 million (D Simões, R Zucchi and N Gobbi, personal communication cited in Jeanne 1991), these colonies are the largest of any polistine wasp collected to date. The next largest on record was a colony of A. areata from Mexico, with a population of 21,800 (Jeanne 1975). The largest recorded colony of Ropalidia is one of R. rornandi cabeti from Queensland, Australia, that contained a population very roughly estimated at 20,000 (Richards 1978). The collected adult populations of two colonies of R. montana from India studied by Yamane et al (1983) comprised 9,816 and 8,497 adults, considerably fewer than in the colonies reported here. Yamane et al's colonies were collected on 6 December 1978 at Top Slip, Annamalai, Tamil Nadu (elev. 700 800 m, 10 20' N, 76 40' E), some 140 km south of our site. Carl states that his two colonies, collected south of Coonoor in the Nilgiri Hills, also in December, had populations of several thousands, which he believed to be "tres réduite" (Carl 1934). In both previous studies, the nests were at least as large as ours; Carl's larger nest was 80 cm long and 58 cm wide, much larger than our largest. One curious difference is that Carl's

12 Robert L Jeanne and James H Hunt photograph clearly shows his large nest attached to a branch approximately 3 cm in diameter, while all our nests were attached to twigs no thicker than a few mm. Although neither previous report estimates how many adults were lost during collection (Yamane et al indicate "undetermined numbers"), it seems evident that in December, colony populations are reduced compared with what we found in August. This, along with the nearly complete absence of brood noted by Carl, suggests that colonies are not in an active growth phase in December. Our colonies, on the other hand, were clearly growing actively adult populations were large and combs were filled with brood. It seems likely that these wasps enter an active production period during the warm, rainy summer months, then shut down during the cool, dry winter, as Carl (1934) speculated. Although not documented, it is worth citing the observation of local naturalists that nests of R. montana cannot be found in the dry months of April and May. This supports the notion that nesting is seasonal and suggests that the nests we saw were initiated since May and that colonies are annual. Seasonal nesting is known to occur in Protopolybia acutiscutis in Panama, which occupies large, conspicuous nests during the dry season, then moves into the forest to pass the wet season in small, inconspicuous nests (Naumann 1970). If seasonal nesting activity is the case, it is curious that males are found in both seasons. We found up to 27% males in August, while Yamane et al found 0 4% in one of their colonies and 47% in the other. Carl found only 4 males in a sample of "about 600" (0 67%) from one of his nests. This indicates that male production is not limited to one season of the year. On the other hand, the vast inter-colony differences in male/female ratio found by us and Yamane et al (1983), along with the reduced (<1 14%) proportion of males among eclosing adults compared with that of the standing adult population (27%) in colony 4, suggest that the rate of male production is not constant. Yamane et al (1983) found a proportional representation of queens similar to that of our colonies ( <1%). Four lines of evidence suggest that nests of R. montana are enlarged during colony growth. First, the range of nest sizes suggests that large size is reached by growth from an initially smaller size, though we recognize that an alternative hypothesis, namely determinate nests built by swarms of greatly different numbers of wasps, cannot be excluded. The second line of evidence is the pattern of brood distribution, which we interpret as demonstration of repeated use (as many as four occupancies) of nest cells in the center of the upper parts of the nests. Outward and downward from the upper center of the nest, cells had been used fewer times, with cells containing larvae or eggs near the comb margins or in the lowermost combs containing their first occupants (these cells lacked the "windows" marking meconium removal from a previous occupant). This pattern of brood distribution is commensurate with nest growth outward and downward from an initially smaller size. It could, however, also be produced in a nest of fixed size if eggs were laid downward and outward over a period of weeks or months. A third, more compelling line of evidence is the progressive change in comb orientation from top to bottom in nests 4 and 12. We believe that as each nest gained weight from growing brood and adult populations, the slender supporting branch bent downward, causing the lower end of the nest to tip toward the trunk of the tree. Because the wasps continued to construct each new comb horizontally, the

Ropalidia montana 13 progressive change in plane of the combs from top to bottom of the nest reflects continuing construction during colony growth. The fourth line of evidence is that the envelope had been enlarged in some nests. This was shown both by the remnants of darker, more weathered layers remaining at the top of some nests, to which lighter-coloured, newer envelope layers had been attached, and by the evidence that inner envelope layers had been partially chewed away. V. tropica apparently preys exclusively on larvae and pupae of other social wasps in the subfamilies Stenogastrinae and Polistinae (Ruiter 1916; Vecht 1957, 1962; Ward 1965; Matsuura and Yamane 1984; Matsuura 1990). Ours is the first record of attacks on large, enclosed nests of a swarm-founding wasp. Although R. montana appears to lack any means of defending its colonies against depredation by V. tropica, it is not clear what effect these attacks have on a given colony. Further study will be required to establish the long-term response of a Ropalidia colony to depredation by Vespa. Are the cells re-lengthened and the nest re-used? Does the adult population emigrate and start over in a new nest elsewhere? Or is the colony inescapably exterminated by the attack? It is unfortunate that we were unable to witness swarm emigration. The rapid departure of swarms from the site after its nest had been destroyed suggests some means of coordinating swarm movement. Presumably the mechanism is unlike the sternal-gland produced trail pheromone of Polybia sericea (Jeanne 1981), since R. montana lacks sternal glands (Jeanne et al 1983). Acknowledgements We thank Dr Raghavendra Gadagkar for arranging our visit to Mudumalai, handling all the logistical details, assistance in the field, and for reviewing this manuscript. The manuscript also benefitted from comments by John Wenzel. Harry Williams, Shyamala, and Sudha cut out and weighed the nest templates. To Harry Williams and the graduate students and staff at the Masinagudi Field Station of the Centre for Ecological Sciences we extend our thanks for their generous hospitality and assistance. To Shivaji, our guide in the field, we owe a special debt of appreciation: without his keen eye for Ropalidia nests our sample would have been much smaller. Research supported by College of Agricultural and Life Sciences, University of Wisconsin, Madison, and by NSF grants BSR 8805971 and INT 9012417. References Carl J 1934 Ropalidia montana n. sp. et son nid. Un type nouveau d'architecture vespienne; Rev Suisse Zool. 41 675-691 Carpenter J M 1991 Phylogenetic relationships and the origin of social behavior in the Vespidae; in The social biology of wasps (ed) K G Ross and R W Matthews (Ithaca: Cornell University Press) pp 7 32 Darchen R 1976 Ropalidia cincta, guepe sociale de la savane de Lamto (Côate d'lvoire) (Hym. Vespidae); Ann. Soc. Entomol. Fr. 12 579--601 Das B P and - Gupta V K 1989 The social wasps of India and the adjacent countries (Hymenoptera: Vespidae); Oriental Ins. Monogr. no. 11 1-292 Gadagkar R and Joshi N V 1983 Quantitative ethology of social wasps: time-activity budgets and caste differentiation in Ropalidia marginata (Lep.) (Hymenoptera: Vespidae); Anum. Behan. 31 26-31

14 Robert L Jeanne and James H Hunt Gadagkar R and Joshi N V 1984 Social organisation in the Indian wasp Ropalidia cyathiformis (Fab.) (Hymenoptera: Vespidae): C. Tierpsychol. 64 15 32 Ito Y 1983 Social behaviour of a subtropical paper wasp, Ropalidia jasciata (F.): field observations during founding stage; J. Ethel. 16 1 14 Ito Y 1985 Colony development and social structure in a subtropical paper wasp, Ropalidia fasciata (F.) (Hymenoptera: Vespidae); Res. Popul. Ecol. (Kyoto) 27 333 349 Jeanne R L 1975 Social biology of Stelopolybia areate (Say) (Hymenoptera: Vespidae); Biotropica 5 183 198 Jeanne R L 1980 Evolution of social behavior in the Vespidae; Annu. Rec. Entomol, 25 371-396 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 (ed) K G Ross and R W Matthews (Ithaca: Cornell University Press) pp 191 231 Jeanne R L, Downing H A and Post D C 1983 Morphology and function of sternal glands in polistine wasps (Hymenoptera; Vespidae); Zoomorphology 103 149 164 Matsuura M 1990 Biology of three Vesta species in central Sumatra (Hymenoptera, Vespidae); in Natural history of social wasps and been in equatorial Sumatra (eds) S F Sakagami, R Ohgushi and D W Roubik (Sapporo: Hokkaido University Press) pp 113 124 Naumann M G 1970 The nesting behavior of Protopolybia pumila in Panama (Hymenoptera: Vespidae), Ph. D dissertation, University of Kansas, Lawrence Richards 0 W 1978 The Australian social wasps (Hymenoptera: Vespidae); Aust. J. Zoo. (Supple.) 61 1 132 Ruiter B A de 1916 De Vesta cincta als Poppenroover; Trap. Nat. 5 190-191 Vecht J van der 1957 The Vespinae of the Indo-Malayan and Papuan areas (Hymenoptera, Vespidae); Zool. Verh. 34 1 83 Vecht J van der 1962 The Indo-Australian species of the genus Ropalidia (Icaria) (Hymenoptera. Vespidae) (Second Part); Zool. Verh. 57 1-72 Wenzel J W 1987 Ropalidia formosa, a nearly solitary paper wasp from Madagascar (Hymenoptera: Vespidae); J. Kansas Entomol. Soc. 60 549 556 Ward P 1965 A raid on a Stenogaster nest by a hornet Vespa tropica, Malay Nat. J. 19 152 West-Eberhard M J 1977 The establishment of reproductive dominance in social wasp colonies; in Proc. 8th Int. Congr. Int. Union Study Soc. Insects, pp 223-227 West-Eberhard M J 1978 Polygyny and the evolution of social behavior in wasps; J. Kansan Entomol. Soc. 51 832-856 Yamane S 1986 The colony cycle of the Sumattan paper wasp Ropalidia (Icariola) rariegata jacobsoni (Buysson), with reference to the possible occurrence of serial polygyny (Hymenoptera: Vespidae); Monit. Zool. Ital. 20 135-161 Yamane S. Kojima J and Yamane Sk 1983 Queen worker size dimorphism in an Oriental polistine wasp, Ropalidia montana Carl (Hymenoptera: Vespidae); Insectes Soc. 30 416-422