ENTOMOLOGIST'S MONTHLY MAGAZINE 27 TAXONOMX DISTRIBI/TION AND NESTING BIOLOGY OF WSPA AFFINIS (L.) AND WSPA MOCSARYANA DU BUYSSON (HYM., VESPTNAE) BY MICHAEL E. ARCHER Vespa ffinis (Linnaeus, L764) and 14 mocsaryana du Buysson, L905 are hornets of the mainland and islands of south east Asia. In a cladistic study of Vespa Archer (1994) found that these two species formed a holophyletic group (Mayr & Ashlock, 1991). This paper reports on the taxonomic, distributional and nesting biology of V. ffinis and V. mocsaryana, reviews the colour subspecies of 14 affinis, and proposes informal names for the colour subspecies of 14 ffinis. 'TAXONOMY AND DISTRIBUTION V. affinis and 14 mocsaryana are united by the female derived character state of an elongated first gastral tergum which is half as long or longer than wide. This character state also has been evolved in parallel by the 14 tropica group (Archer, t991), from which it can be separated by the semicircular shape of the lateral apical margin of the clypeus (fig. 1A). In the V. tropica group the lateral apical margin of the clypeus has a triangular shape. PNC Fig. 1. Vespa ffinislmocsaryarut. A, apical margin of female's clypeus; B, ventral view of the -sixth (6) and seventh (7) gastral sterna of the male; C, lateral view of pronotum (PNC = pronotal carina, PTC = pretegular carina). The males of 14 ffinis and 14 mocsaryana can be separated from the other species of Vespa by the following combination of characters: L, apical margin of the seventh gastral sternum with one emargination (fig. 1B); 2, apical margin of the sixth sternum with a deep emargination, semicircular in shape (fig. 1B); 3, pretegular carina complete (fig. 1C); 4, pronotal carina much interrupted by the pronotal pit (fig. 1C). V. ffinis and 14 mocsaryana may be separated from each other as follows: 27th March, 1997 Vol. 133 (1997) 28 1. ENTOMOLOGIST'S MONTHLY MAGAZINE Mesoscutal punctures small and clearly separted from each other, the distances between punctures greater than the puncture diameter. Shaft of the aedeagus not bulbous (fig. V. mocsaryana du Buysson Mesoscutal punctures large and often touching each other, the distances between 28) - i1*::'::'::i'::: Fig. 2. i:1":::: 1':TT: ::::: :: :i: ::i":::: ::1o"" (o' i!)m,* rf j Dorsal view of the male genitalia (SA = shaft of aedeagus): A, Vespa affinis; B, Vespa mocsaryana. Taxonomy and distribution of 14 ffinis The taxonomy of V. ffinis has been largely determined by Bequaert (1936) and van der Vecht (1957). Bequaert (1936) showed that du Buysson (1904, 1905) confused two species under V. cincta Fabricius, 1775, viz.: V. tropica (Linnaeus, 1758) and 14 ffinis. Bequaert (1936) included within V. ffinis the following colour varieties: nominate ffinis, indosinensrs Perez, I9"l.O, alduini Gu6rin, 1831, and picea du Buysson, 1905. Bequaert also described the new colour varieties hainanensls and continentalis. Van der Vecht (1957) added the new subspecies: nigriventris, rufonigrans, moluccana, archboldi and alticincta. V. affinis, as understood through its eleven subspecies, would seem to be one species and not several species as was found for V. tropica, a similarly widespread homet (Archer, 1991). The distribution of V. affinis (fig. 3,A.) is as follows: Ceylon, India (Kerala, Tamil Nardu, Karnataka, Maharashtra, Bihar, West Bengal), Sikkim, Assam, China (Hubei, Shanghai, Fujian, Guangdong, Hunan, Hainan Island), Hong Kong, Taiwan, Japan (Ryukyu Islands), Burma, Thailand, Laos, Vietnam, Sumatra, west Java, Philippines (Palawan), Papua New Guinea including New Britain and New Ireland, Indonesia (Irian Jaya, Moluccas, Sulawesi, Kalimanton), and Malaysia (Peninsular, Sabah, Sarawak). ENTOMOLOGIST'S MONTHLY MAGAZINE 29 0 ( /. '\\ ( ^(/ \IJffi, ar Fig. 3. - \-\_lor.a, e, Distribution maps of; A, Vespa affinis;8, Vespa mocsaryarul The colour forms of V. ffinis can be separated by variation in the colour of the vertex, clypeus, and the first, second and third gastral terga (Table 1,). Specimens showing intermediate colours between several of the TABLE 1. AFFINIS. Colour - THE COLOUR VARIATIONS OF THE SUBSPECIES OF Vertex Clypeus Form affinis indosinensis Red to dark red Red to dark red Red to dark red Red & rufonigrans black Dark red continentalis lninanensis Red Red Black Orange & red Dark red to black Red & black Black Red & black nigriventris First Gastral Tergum Black Orange & red Orange & red Orange & red Orange Second Gastral Tergum VESPA Third Gastral Tergum Orange Black Orange Black Orange Orange Orange Orange Black & orange Black Black Black & black alduini moluccana archboldi picea Dark red & black Black Red Black Black Red Red Black Black Dark red & black Orange & red Dark red & black alticincta Red Orange Black & black Black Red & dark red Orange BIack & black Orange Black Dark red & black Black Red & dark red BIack 30 ENTOMOLOGIST,S MONTHLY MAGAZINE colour subspecies of tr4 ffinis have been found. Intermediates betweenV. ffinis ffinis and 14 ffinis continentalis in the Bombay region, between V. affinis ffinis and V. ffinis indosinensrs in northern Burma, and between V. ffinis rufonigrans and V. ffinis archboldi in western New Guinea have been seen. V. ffinis moluccana would seem to be an intermediate colour form between V. ffinis rufonigrans and V. ffinis alduini. The colour subspecies V ffinis archboldi falls within the variable nominate colour subspecies V. ffinis ffinis. Because of these colour forms it is suggested the subspecies nomenclature be replaced by "informal names" (Table 2) following the arguments of Wilson & Brown (1es3). TABLE 2. THE INFORMALAND FORMAL (SUBSPECIES) COLOUR FORM AFFINIS. NAMES FOR YESPA Informal Name Subspecies India & China affinis continentalis hainanensis indosinensis Southern India Hainan Island Indo-Malaya & Sumatra rufonigrans nigriventris alduini Borneo to western New Guinea Palawan Buru Island Ceram & Aru Islancis Northern New Guinea Southern New Guinea picea New Britain alticincta moluccann archboldi Taxonomy and distribution of 14 mocsaryana The distribution of. V. mocsaryana (fig.38) is as follows: Sikkim, Assam, China (Sichuan, Anhui, Fujian), Thailand, Laos, north Vietnam, Peninsular Malaysia, Burma (Tenasserim), Sumatra. The Chinese records from Sichuan and Anhui are from an unpublished map of van der Vecht (pers. comm., 1981). No colour subspecies of 14 mocsaryana have been proposed, but van der Vecht (1959) noted an increase in the presence of yellow on the gaster on specimens from Tenasserim and Peninsular Malaysia. On these specimens, tergum one has a complete apical yellow band and tergUm two an interrupted yellow band with both bands widened laterally. Sterna two to four ea-ch has a yellow band, but on sterna three and four the yellow band is more-or-less reduced in the middle. Sternum five has a trace of a yellow band. I have found a similar yellow development on a specimen irom north Vietnam, with a yellow band on terga one to three and sterna two to four, although this yellow did not exist in the middle on tergum two and the sterna. The yellow apical bands became more extensive on a specimen from Laoi: on the gastral terga the apical yellow bands graduatly become more extensive from terga one to six so that the sixth iergum is entirely yellow. Similarly development of the-yellow occurs on the sterna so that the sixth sternum is nearly entirely yellow. ENTOMOLOGIST'S MONTHLY NESTING BIOLOGY OF MAGAZINE 3I V. MOCSARYANA Very little is known of the nesting biology of V. mocsaryana other than that it is restricted to mountain forests (Matsuura, 1990). NESTING BIOLOGY OF V. AFFINIS The nesting biology of V affinis has been studied in the lowland {opics: Silgapore (1'20N) (Chan, 1972, Martin, 1993, pers. comm.), Bangkok, Thailand (13"45N) (seeley & Seeley, 1981, Gnmukayakul, I287),--Sumatra (0"555) (Matsuura, L983, 1990), New Guinea 15"ZlS1 (spradbern 1986) and the Philippines (9"25N) (starr & Jacobson,'1990) subtropics: Calcutta (22'35N) (Chopra, 1,925), Taiwan (24'N) 9_._d (Matsuura, L973, Yamane, 1977, Kuo, 1984) and the Ryukyu Iilandg southern Jgpan (26"30N) (Martin, r99za, L99zb, r99ic, 1992, pers. comm., 1993,1993, pers. comm.). This paper will investigate the nesting biology differences between tropical and sub-tropical populations of V. affinis. Nesting Sites Tropical and subtropical nesting sites are usually aerial: on trees and shrubs, but sometimes on the outsides of buildings or in attics. In southem Japan from r34 nest sites found, r23 (gl.s%) were attached to a cylindrical structure such as a twig or plant stem. The remaining eleven nests were attached to the mid-ribs of leaves. Nests attached to stems less than 2.5 mm in diameter usually failed, either because the increasing weight of the nest bent the stem until the nest reached the ground, or th6 nest became detached in bad weather during the spring. Most nests were within 1 m of the ground although some were at a height of lz.4 m. Nests less than 12 cm above the ground failed to rear sexuals as nest expansion stopped when the nest reached the soil surface. The workers did nbt seem able to excavate a cavity in the soil. High level nests were usually in dense bushes or well sheltered from the monsoon rains and strong winds. Exposed nests were easily destroyed by such severe weather. In Sumatra,2r4 nests were all aerial with 197 (9z.lvo) in open sites (6g on shrubs , r22 on trees, 7 on buildings) and L7 in the enclosed space'of attics. The height above the ground ranged from 0.5 m to 30 m. In Taiwan, Kuo recorded nests on trees, shrubs, bamboo, under eaves, and in roof spaces of houses. In New Guinea, nests were recorded most frequently on trees but also on the roofs of houses, under eaves of buildings, attached to vine-covered walls, and inside sheds and outbuildings. In the Philippines, 20 nests were recorded on trees and shrubs and from 1 m to 16 m above the ground but usually 2-3 m above the ground. Nests were attached to a single branch, except in dense vegetation when several branches passed through the nests. Nests attached to slender, flexible vegetation swayed on windy days. In Singapore, chan recorded 316 nests, mostly found on trees and 32 ENToMoLoGIST's MoNTHLY MAGAZINE shrubs (69.9Vo) but also on buildings, including 1o_ofs and vegetated fences. Nests were found from ground level up to 26 m but most were below 17 m. Colonial life history characteristics - Subtropics In southern Japan during the autumn newly-emerged quee-ns remain in the colony for a6out LTI4 days during which.time their body weight increases ty about 4OVo due mainly to fat deposition. The males remain in the colony for 8-LL days with a 38% increase of body weight. During Novembei and December the sexuals leave the colonies, and after mating the males die and the queens enter over-wintering sites, e.g- lmong 9ty' dead wood piles. Queeni remain in the over-wintering sites for four to five months living off their fat stores. They emerge during mid-April and by the end otafrit nest initiation by a single queen starts. The queen colony lasts for abolt 34 days with the adult workers appearing from the beginning of June. Large iell building starts from the end of_August, with the -and queens emerging at the end of September. Adult firsi adult males sexual production t"a"h"s a peak during the first half of November. The colonyLnds during Decembel. The length of colonial life is from April to December, eight months or about 240 days. In Taiwan and Calcutta the length of colonial life is also about eight months. - Tropics oi tropical colonies is longer by about_ two- to Colonial life history characteristics The duration three months, but colonies are still annual. In Singapore, Martin found colonies in all months of the year, but peak activity during September.- In New Guinea, queens initiited colonies from AugUst until November, and colonies *"t" found producing workers from October until December, although some queen iolonies *ere found up to December. Production of the sexuals was delayed until the rainy season started in December. Colonies reached p"uk development during the following_Jule and July. In Sumatra, Matsuura (in Ross-& Matthews, I99l) found, during the rainy season (mid-Novembei), that of 35 colonies, eleven we-re guegl colonies, fS tnoird only adult-worker production, and nine had adult sexuals emerging. Queen usurpation in the subtropics In southern Japan evidence of intra-specific usurpation between queens was found at L6 out of 55 colonies. Usurpation evidence was the presence of dead queens under the colonies, and in one caseactual fighting between the queens was observed. Fourteen of the usurped colonies failed to rear ,r*rrilt, but two developed normally. Usurpation occurred during June when the worker population was smail, indeed at the time of, or just after, the emerg"n". of tht first workers. From a sample of seven, the resident qrrc.t wJn the fight on four occasions and the invader on three occasions. ENTOMOLOGIST'S MONTHLY MAGAZINE 33 Larger queens were not more likely to win the fight and the presence of workers did not always help the resident queen in winning the fight. In Taiwan, Kuo observed queen usurpation fights. In one fight in early June when a few workers were present both the resident and invader queens died from their injuries so the colony was orphaned. Later another invader queen was rejected by the workers. The colony expanded normally but as worker-laid eggs are unfertilized, so maturing as males, the colony ended in July. Polygynous colonies in the tropics In the tropics, a colony usually has more than one egg-laying queen (polygyny). Polygynous colonies may be founded by several queens (pleometrosis and primary polygyny). If one queen initiates a colony (haplometrosis) but is later joined by other egg-laying queens, the polygyny is secondary. In Sumatra, L30 colonies were collected from 1981 until 1984. Of 52 queen colonies 4l (78.8Vo) were polygynous with a mean of 3.9 queens (range 1-10) per colony. In eight of these colonies from 198L, all the queens were inseminated but only one queen (occasionally two) had fully developed ovaries and was laying eggs. In42 colonies with adult workers, 38 (90.5Vo) were polygynous with a mean of.4.4 queens (range I-12) per colony. Of 36 colonies with adult sexuals, 33 (9I.7Vo) were polygynous, with a mean of 4.6 queens (range L-I4) per colony. In the colonies from 1981" with adult sexuals, all the queens were inseminated and with fully developed ovaries and able to lay eggs. No marked dominance behaviour was observed between the queens. It was not known if a colony was initiated by u single queen and then quickly joined by other queens or whether pleometrosis was shown. In two queen colonies, a larger worker with worn wings was present which could not have been reared in the current colony. In these cases the worker had initiated the colony with one or several queens. In New Guinea, of seven queen colonies, five were polygynous with a mean of 3.1 queens (range L*6) per colony. Of eight colonies with adult workers, seven were polygynous with a mean of 4.5 queens (range 1-15) per colony. Of four colonies with adult sexuals, three were polygynous with a mean of 3.0 queens (range 1-5) per colony. Thus 15 (78.9Vo) of the L9 colonies were polygynous. Except for the colony with 15 queens, all the queens were inseminated, with fully developed ovaries, and capable of laying eggs. In the colony with 15 queens, l3 were inseminated, of which eleven had fully developed ovaries. It was considered that these colonies had been initiated by several queens (pleometrosis). In Singapore, Martin found two colonies from a sample of five that were polygynous. Queen colonies - Subtropics In southern Japan, Martin observed 27 queen colonies, eight for more 34 ENTOMOLOGIST,S MONTHLY MAGAZINE than one week and one for 27 days, until the emergence of the first worker. The queen nest consisted of a single comb attached to the vegetation by a thick petiole coated in a shiny secretion. The secretion is produced by the sixth and seventh gastral sternal glands and acts as a barrier to ants. The comb was surrounded by an envelope which was bell-shaped with an elongated entrance or vestibule. The vestibule varied in length from L-5 cm (mean 3.1 cm, n = 15).The vestibule delayed the entrance of small ants onto the comb and probably prevented large ants from entering the nest. The vestibule is removed when the first workers emerge, when the workers readily deter ants from getting onto the combs. In southern Japan, the egg stage lasted for 6.8 (range 6.5:7.5) days, the larval stage eighf to more than22 days but mostly for L2-t6 days, and the sealed brood stage for I7.3 (range 16-18) days). The developmental period is therefore usually 3HO days. During the first ten days, the cell buitding rate was I.2-1,.5 cells per day but later this rate dropped to 0.75-0.85 cells per day. During the early stages of the queen colony, the queen lays L-2 eggs per day. On the emergence of the first worker the queen nest typically consisted of 28 (four egg, 13 larval, eight sealed brood, and three empty) cells. On emergence the workers remain in the nest for two days before starting to forage. In Taiwan, Kuo found queen colonies varied from ten to 30 cells with a development period of 4U41. days (egg 6 days, larva 15 days, sealed brood L9-2O days). Queen colonies - Tropics In New Guinea, Spradbery found that the cell building rate in one monogynous colony (0.3 cell per day) was lower than one polygynous colony (0.7 cell per day). Generally there was a significant positive linear relationship between the number of foundress queens and the number of cells in a colony. In Sumatra, Matsuura recorded polygynous colonies with 2249 cells. In Bangkok, Tonmukayakul recorded a worker developmental period of 38 days (egg and larval stages L9 days, sealed brood stage 19 days) in a colony in which the first workers had recently emerged. Nest development and size - Subtropics southern Japan, Martin divided his sample of nests into two subsamples. The first subsample (51 colonies) consisted of large nests In reaching mean mature size of about 4000 cells in the latter part of October. The second subsample (55 colonies) consisted of small nests reaching a mean mature size of about 2500 cells in the latter part of November. Both large and small colonies were successful in rearing sexuals. Nest growth started with a period of slow growth reaching about 500 cells in large colonies by early August, and in small colonies by early September. Both large and small colonies then showed rapid growth to maximum size, by mid-October for large colonies and by the latter part of November for ENTOMOLOGIST'S MONTHLY MAGAZINE 35 small colonies. Cell building rates were estimated to start at about 0.45 cells per worker per day in large colonies, and about 0.23 cells per worker per day in small colonies. Cell building rates per worker per day gradually decreased during the development of the colony. Normally, a nest consisted of from six to ten combs although sometimes more combs were present. The small cells occupied the upper combs and the large cells the lower combs. Each small cell could be used to rear more than one individual. Matsuura & Yamane (1990) recorded that mature nests in Japan usually had four to five combs and 800-1500 cells. In Taiwan, Matsuura recorded one large nest of 6178 cells. Nest development and size - Tropics Mature nest size has been recorded as follows: in Sumatra 12,058 cells (9 combs), 9877 cells (8 combs); in New Guinea 9732 cells (5 combs), 12,580 cells (8 combs), and 45,065 cells (L2 combs); Martin in Singapore L0,927 cells; in the Philippines 12,000 cells (9 combs); and Seeley & Seeley in Thailand 9600 cells (6 combs). These eight nests give a mean "1.5,230 cells (or 10,968 cells if the very large nest of 45,065 cells is excluded) on 8 combs. Immature brood - Subtropics In southern Japan, the faster development of the large colonies was reflected in the characteristics of the immature brood. The egg laying rate per day peaked at about 55 eggs during late September in large colonies and at about 30 eggs during early October in small colonies. The number of eggs peaked at about 430 during late September, and larvae peaked at about 930 during early October in large colonies. In small colonies, the number of eggs peaked about 240 during early October and the number of larvae at about 460 during mid-October. In large colonies worker sealed brood peaked at about 450 during midOctober and September, male sealed brood at about 430 during early -small colonies q.reen sealed brood at about 210 during late October. In worker sealed brood peaked at about 130 at the end of September, male sealed brood at about 22O durine early November, and queen sealed brood at about 120 during mid-November. The lowest larval/worker ratio was between two to four larvae per worker for both small and large colonies. The ovarian index (which is a measure of egg laying rate) was higher in queens from larger colonies than in queens from smaller colonies. Towards the end of the colony development workers were often seen removing withered or neglected larvae from the nest. In southern Japan, the worker developmental period was 35-39 days (egg stage 7 days, larval stage 12-16 days, sealed brood stage 16 days). Immature brood - Tropics In New Guinea, the number of immature stages for each colony rearing 36 ENToMoLoGIST's MoNTHLY MAGAZINE sexuals was higher than in the sub-tropics: eggs from 77VI709, larvae from 205f3103, and sealed brood from I57V3096. The larval/worker ratio was 0.8-1.1 larvae per worker. Martin, in Singapore, showed that in two polygynous colonies the number of eggs present in a colony was less than the number predicted from an examination of the ovarian index of the queens. However the number of eggs present in a colony was in excess of the predicted laying power of a single queen so several queens must have contributed to the present. In one monogynous colony the number and predicted "ggr of eggs were very similar. number Workers - Subtropics In southern Japan, the number of workers gradually increased to about 80 during early August. Afterwards there was an accelerating increase to about 740 workers by early October in large colonies, and about 300 workers just before mid-October in small colonies. The acceleration was followedby a rapid decline in worker numbers, which was more abrupt in large colonies. It was estimated that the total number of workers reared wai about t20O in large colonies, and 540 in small colonies. The worker length of life was usuaily Iar-IS days but varied from 7-28 days. In Taiwan, Kuo recorded a more rapid increase compared with southern Japan: to 7f80 workers by early July, 20f300 workers by early August, with a peak of 60O-L000 workers by early September. Worker length o! life varied from I0-l25 days. Also from Taiwan, Matsuura recorded 1095 workers from a large colony with an estimated 2I0O workers having been reared. Workers - Tropics From New Guinea much larger worker populations in a colony were found, from about 2700 to over 5000. Queens and Males Littte information is known about the development of queens and males and then only from the subtropics. However adult queens and workers can be cleaily separated by a size difference in the tropics and subtropics. In iouthern Japan, in large colonies adult males peaked at about 300 after mid-Octobei and adult (ueens at about 250 during mid-November. In small colonies adult males peaked at about 200 and adult queens at about 85 during mid-November. InTaiivan, Yamane found higher peaks with 500 or more males and 400 or more queens. In southern Japan, Martin estimated the large colonies produced 1440 males and 540 queens and small colonies produced 540 males and L20 queens. ENTOMOLOGIST'S MONTHLY MAGAZINE 37 DISCUSSION Subtropical colonies have a shorter life cycle (say 240 days) and reach a smaller mature nest size (usually up to 4000 cells) compared with the longer life cycle (say 320 days) and larger mature nest size (in excess of L0,000 cells) of tropical colonies. The development of subtropical colonies is synchronised by the need for the queens to overwinter. Tropical colonies, without this need, are not synchronised, although the occurrence of a wet season can lead to the synchronisation of rearing sexual brood. Subtropical colonies show haplometrosis and monogyny, iuith the queens fighting between themselves for nest ownership (usurpation). Tropical colonies are usually polygynous (either showing haplometrosis with secondary polygyny or pleometrosis with primary polygyny), and usurpation has not been recorded. The larger size of tropical colonies must ultimately be linked to the availability of food for a longer time each year due to more favourable weather conditions, so that the life cycle can be extended. The larger size of tropical colonies is not a consequence of a reduced length of worker developmental period, which is very similar in tropical and subtropical colonies. These similar developmental lengths of workers are probably due to the good colonial temperature regulation. With such favourable weather conditions why do tropical colonies remain annual colonies, rather than become perennial, as is shown by some of the paravespulid wasp species which have expanded their ranges into warmer climates (Ross & Matthews, 1991)? If Vespa evolved from a temperate, overwintering, species thenVespa could be carrying a phylogenetic constraint of annual development. The queens of the tropical species, V. philippinensis de Saussure, L854 still build up fat bodies after emergence (Starr, L987), which in temperate species are used as an overwintering food store. These fat stores could be a relic from an evolutionary past, or the food stores could have acquired an unknown new function. The retention of an annual cycle also could be used to inhibit the build up of pathogenic or parasitic attack. The larger size of tropical colonies could be a consequence of polygyny with several egg layers. Previously it was recorded that polygynous colonies show a quicker development of the queen colony. However, later during the worker rearing phase, cell building rate limits egg laying rate, and, compared with subtropical colonies, sexual production is delayed. When sexual production begins the larva worker ratio is more favourable in tropical colonies and the cell building rate is probably not limiting, so the sexual production can be greatly increased. Polygyny could also have evolved as a defence of the queen and early worker colonies against parasites and predators such as birds, ants and other hornet species since polygynous colonies are left unattended for shorter periods of time: (I7Vo) compared with monogynous colonies (50Vo) (Martin, 1993, pers. comm.). Such colonies seem to have good defences against ants, and Martin in Singapore reported that Dr N.S. Nling 38 ENTOMOLOGIST,S MONTHLY MAGAZINE observed that the queens of a queen colony successfully repelled an attack by a single worker of V tropica. Polygyny could also have evolved as a defence against queen usurpation and a more rapid rebuilding of the nest after its destruction or damage. Certainly usurpation has not been recorded in polygynous colonies. ACKNOWLEDGEMENTS I am grateful to Dr S. Martin and his wife Yuko for translating the Chinese paper by Kuo. REFERENCES Archer, M.E., 1991. Taxonomy and bionomics of the Vespa tropica group (Hym., Vespinae), Entomologist's mon. Mog., 127:225-232;1.994, A phylogenetic study of the species of the genus Vespa (Hymenoptera: Vespinae), Ent. scand.,24:469478. Bequaert, J.,'1.936, The common Oriental hornets, Vespa tropica and Vespa affinis, and their color forms, Treubin,lS:32945L. Buysson, R. du., 1904, Monographie des gu€pes ou Vespa, Annls Soc. ent. Fr.,72; 26V288;1905, Monographie des gu6pes ou Vespa, Annls Soc. ent. Fr.,73:485-556, 565434. Chan, K.L., 1972, The hornets of Singapore: their identification, biology and control , Singapore med. J.,13: 178-187. Chopra' 8., I9?5, Note on a nest of the common Indian homet Vespa cincta Fab., J. Bombay nat. Hist. Soc., 30: 85&860. Kuo, M.C., 1984, Studies of Vespidae in Taiwan. 1. A study of the ecology of Vespa formosann Sonan. J. Natn. Chianyi Inst. Agric., l0:. 73-92. Martin, 5.J., 1992a, Colony failure in the Social Wasp Vespa afrnis (Hymenoptera, Vespidae), Jap. J. Ent., 60: 67148L; 1992b, Development of the embryo nest of Vespa affinis in southern Japan, Insectes Sociaux,3g:45-57;'1,992c, Colony defense against ants in Vespa,lnsectes Soci.atac, 39 99-lI2;1993, Weight changes in adult hornets, Vespa affinis (Hymenoptera: Vespidae), Insectes Socinux, 4O 363468. Matsuura, M., 1973, Nesting habits of several species of the genus Vespa in Formosa, Kontyit, 4l: 286-293; 1983, Preliminary report on the polygynous colonies of Vespa affinis indosinensis (Hymenoptera, Vespidae) in Sumatra, Kontyfi,51: 80-82; 1990, Biology of three Vespa species in central Sumatra (Hymenoptera, Vespidae), In Sakagami, S.E., Ohgushi, R. & Roubik, D.W. (eds), Natural history of social waips and bees in equntorinl Sumatra, Hokkaido Univ. Press, Sapporo. Matsuura, M. & Yamane, Sl<., 1990, Biology of the vespine wasps, Springer-Verlag, Berlin. May6, E. & Ashloclr, P.D., 1991, Principles of Systematic Zoology, McGraw-Hill, New York. Ross, K.G. & Matthews, R.W., 1991, The social biology of wasps, Comstock Publishing Associates, Ithaca. Seeley, T.D. & Seeley, R.H., 1981, A nest of a social wasp Vespa affinis in Thailand (Hymenoptera: Vespidae), Psyche, 87: 299-304' Spradbery, J.P., 1986, Polygyny in the Vespinae with special reference to the hornet Vespa affinis picea (Hymenoptera, Vespidae) in New Guinea, Monitore zool. ital.,20:. 101-1.18. Stam' C.K., '1,987, A colony of the hornet Vespa philippinensis (Hymenoptera: Vespidae), Insectes Sociaux,34: l-9. Starr, C.K. & Jacobson, R.S., 1990, Nest structure in Philippine Hornets (Hymenoptera, Vespidae,Vespa spp.),Jap. J. Ent.,58:125-143. Tonmukayakul,A., 1987, Life cycle of a hornet: Vespa afinis, Siriraj Hosp. Gaz.,38 399402 Vecht, J. van der, 1957,The Vespinae of the Indo-Malayan and Papuan areas (Hymenoptera, Yespidae), Zool. Verh., Leiden,34 1-183; 1959, Notes on Oriental Vespinae, including some species from China and Japan (Hymenoptera: Vespidae), Zool. Meded. Leiden,36 205-232. Wilson, E.O. & Brown, W.L., 1953, The subspecies concept and its taxonomic application, Sysr. Zool., 2: 99-171,. Yamane, So., 1977, On the collecting technique of vespine nests, based '1.974 (Hymenoptera, chiefly on practices through a survey in Taiwan from '1972 to Vespidae), S eibutzu Kyoza i, Kikona i, 12: 42-59. The University College of Ripon & York St John, [.ord Mayor's Walk, York, YO3 7EX. October 26th, L994.