Stingless bees - CiteSeerX

Zootaxa 3647 (3): 401–428 www.mapress.com / zootaxa / Copyright © 2013 Magnolia Press

ISSN 1175-5326 (print edition)

Article

ZOOTAXA ISSN 1175-5334 (online edition)

http://dx.doi.org/10.11646/zootaxa.3647.3.1 http://zoobank.org/urn:lsid:zoobank.org:pub:3E2DFCFC-9D75-4245-82F5-9B9FD977160A

Stingless bees (Hymenoptera: Apidae: Meliponini) of the Indian subcontinent: Diversity, taxonomy and current status of knowledge CLAUS RASMUSSEN Department of Bioscience, Aarhus University, Ny Munkegade 114, Bldg. 1540, DK-8000 Aarhus C, Denmark. E-mail: [email protected]

Abstract Eight named species of stingless bees are known from the Indian subcontinent: Lepidotrigona arcifera (Cockerell), Lisotrigona cacciae (Nurse), Lisotrigona mohandasi Jobiraj & Narendran, Tetragonula aff. laeviceps (Smith), Tetragonula bengalensis (Cameron), Tetragonula gressitti (Sakagami), Tetragonula iridipennis (Smith), Tetragonula praeterita (Walker), and Tetragonula ruficornis (Smith). Lectotypes are newly designated for T. bengalensis and T. ruficornis. Keys, comparative notes, and illustrations for species identification are provided. The distribution of stingless bees throughout the Indian subcontinent are summarized and concluding that they are found in most parts of the Indian subcontinent, except at higher elevation or the drier interior regions. Additional collections and studies are urgently needed to clearly define the species limits of the complex “iridipennis” species group. Key words: Tetragonula, Lisotrigona, Lepidotrigona, Trigona, Melipona, India, Pakistan, Bangladesh, Sri Lanka, Nepal, Bhutan, taxonomy, potential distribution

Introduction This is an account of the specimens upon which the eight specific names of stingless bees from the Indian subcontinent are based, as well as references to what has been published about the distribution and biology of the stingless bees of that region. With this information, and after additional collecting of specimens and nest data, it will be possible to prepare a complete revision of the Indian stingless bee fauna even though most of the relevant type specimens are in museums far from India. Bee-keeping involving several species of native honey bees (Apis spp) is a very important enterprise with a long tradition in the Indian subcontinent (including the countries India, Pakistan, Bangladesh, Sri Lanka, Nepal and Bhutan) (Batra 1977, Engel 1999). Less known is the fact that stingless bees have also been kept for centuries in India, Sri Lanka and Nepal (Crane 1999, Kumar et al. 2012). As in other regions where stingless bees occur, colonies have been kept in tree logs, wooden boxes, and clay pots for harvesting small quantities of highly prized medicinal honey, and also for the wax and propolis, produced and gathered by the bees and used for its household and curative properties (Crane 1999). While the medicinal properties of the honey are little explored for the Indian subcontinent when compared to that of the Neotropical species (Rodríguez-Malaver et al. 2009, Choudhari et al. 2012, Surendra et al. 2012), the botanical origin of the honey has been explored (Phadke 1968, Joshi et al. 1998), as have the plants visited by the bees (e.g., Ramanujam et al. 1993, Viraktamath et al. 1999, Devanesan et al. 2002, Danaraddi 2007, Danaraddi et al. 2011). In fact, most research on stingless bees from the Indian subcontinent has focused on their role as crop and native flora pollinators (e.g., Raju et al. 2000, Raju et al. 2009a). Earlier studies of stingless bees have also dealt with worker communication (Lindauer 1956, Lindauer & Kerr 1958, 1960), taxonomy and morphology (George 1934, Sakagami 1978, Jobiraj & Narendran 2004, Danaraddi et al. 2012), as well as their nesting biology (Danaraddi et al. 2009). Recent accounts report five named species of stingless bees from the Indian subcontinent (Sakagami et al. 1990, Jobiraj & Narendran 2004) and presumably several unnamed species that remains to be described (Sakagami et al. 1990, Rasmussen & Cameron 2007, 2010). Additional species from the Indian subcontinent are here removed Accepted by A. Lelej: 3 Apr. 2013; published: 10 May 2013

401

from synonymy raising the total number of named species to eight with the potential to raise further (Lepidotrigona arcifera (Cockerell), Lisotrigona cacciae (Nurse), Lisotrigona mohandasi Jobiraj & Narendran, Tetragonula aff. laeviceps (Smith), Tetragonula bengalensis (Cameron), Tetragonula gressitti (Sakagami), Tetragonula iridipennis (Smith), Tetragonula praeterita (Walker) and Tetragonula ruficornis (Smith)). Stingless bees are widely known in the Indian subcontinent as “dammer bees” or “dammar bees” (dammar is resin from in amongst dipterocarp trees), with additional local names commonly applied, e.g., “putka” in Sikkim and Nepal (Gurung et al. 2003, Singh et al. 2011, Lepcha et al. 2012), “ngap siwor”, “ngap hamang”, and “ngap khyndew” in Khasi language (Pugh 1947), “cherutheneecha” and “arakki” in Kerala (Nair 2003), and “kanameemessa” in Sri Lanka (Karunaratne 2005). The scientific genus-group names for the different stingless bees from the Indian subcontinent have long been Trigona Jurine and Lisotrigona Moure (Michener 2007), but recent data support that Trigona s.l. is not a single evolutionary linage, as it is not a monophyletic clade. Rather, Trigona s.l. encompasses a Neotropical clade and a distantly related Indo-Malayan/Australasian clade (Rasmussen & Cameron 2007, 2010). The solution has been to restrict the usage of Trigona to the Neotropical lineage while treating the names proposed by Moure (1961) for the IndoMalayan/Australasian taxa at the genus level (Rasmussen 2008, Michener 2013), including Lepidotrigona Schwarz and Tetragonula Moure. The latter was proposed with Trigona iridipennis Smith, 1854 from Sri Lanka as the type species. The few genera of stingless bees from the Indian region can easily be separated using published keys (e.g., Sakagami et al. 1990, Michener 2007). However, the species-level identification is not trivial. A notoriously complex genus of stingless bees is Tetragonula, with more than 30 described species and no global identification key (Rasmussen 2008). Specimens of Tetragonula from the Indian subcontinent were treated as a single species in the key by Sakagami (1978). Older keys, such as those of Bingham (1897) and Schwarz (1939), cannot reliably be used for species identification; in particular, the first reference contains several misinterpretations of species and both references include many species from outside the region. No comprehensive summary of stingless bee distribution from the region has been published, with the exception of Sri Lanka (Karunaratne et al. 2005b). Within India, many scattered distribution records exist for states in north and south, east and west, as well as from the distant Andaman Islands (Sakagami 1978, National Horticulture Mission 2012). Stingless bees are also reported from Nepal (Partap 1999, Gurung et al. 2003, Sharma 2004) and Bangladesh (Islam 1997, Hannan 2007), although the exact localities and species identity in those countries needs to be established. In the Neotropical region, stingless bees are known from 34.90°S (Montevideo) in Uruguay (Camargo & Pedro 2007) up to 27.03°N (Álamos, Sonora) in Mexico (Búrques 1997). In Africa they are distributed between 28.54°S (Eshowe) in South Africa up to 18.00°N (Njala) in Sierra Leone (information extracted from gazetteer in Eardley 2004), while in the Indo-Malayan/Australasian region they are known from 36.41°S in Australia (Dollin et al. 1997) up to 24.23 °N in Taiwan (Sakagami & Yamane 1984, 1987, Sung et al. 2006). However, the northernmost global record of extant stingless bees is Dehra Dun, Uttar Pradesh (30.32°N) in India (Sakagami 1978) with several other Indian records above 28°N (Nogueira-Neto 1951, Goel & Kumar 1990, Gurung et al. 2003, Chaudhary & Singh 2007, Gupta et al. 2011). It thus appears that stingless bees are known from most parts of the Indian subcontinent, at least up to 1000 m a.s.l. in India and Nepal (Sinu & Shivanna 2007). In South America and Asia they become rare above 2500 m a.s.l., although exceptionally they have been recorded up to 4000 m a.s.l. in the Andes of Peru and Bolivia (Camargo & Moure 1996, Sung et al. 2006). Given the steady interest in pollination by stingless bees in India and a renewed global interest in cultivation and propagation of stingless bee colonies for conservation, the purpose of the present paper is to summarize information about the described species of stingless bees and provide morphometric data on the type specimens, as well as photographic documentation of their habitus. This will facilitate species identification in the region. In addition, I compile and map all published records of stingless bees from the Indian subcontinent. As such data are fragmentary with many states unrepresented or with few presence points, I further apply species distribution modeling to identify the potential range where these bees may naturally occur. This should inspire more and much needed studies of stingless bees from the region. Hopefully, new collections can be made of these bees for taxonomic purposes, particularly capturing male specimens with associated workers (collected from the same nests) and photographic documentation of the nest entrances. The hidden male genitalia have proven to be valuable for diagnosing species, even in cases where almost no external morphological differences can be recognized (Sakagami 1978). Longer series of workers are necessary to describe and account for the variation in coloration and pilosity within a single species. Lastly, sometimes external nest features such as nest entrances are the most

402 · Zootaxa 3647 (3) © 2013 Magnolia Press

RASMUSSEN

diagnostic field traits observed, distinguishing closely related species (Camargo & Pedro 2003, 2004, Rasmussen 2004, Rasmussen & Camargo 2008). It should also be established as a standard procedure to collect additional workers from the sampled nests into cold-kept ethanol for subsequent DNA studies and molecular confirmation of identification. Only through such careful and rigorous collection and documentation across the subcontinent, can a comprehensive revision of the true diversity of stingless bees from the region be made. Likely the diversity in any parts of the Old World will not reach more than 50 species per site as in highly diverse parts of South America (Roubik pers. comm., Rasmussen & Gonzalez 2009), but there is certainly a potential for encountering several species of stingless bees at any site in India.

MAP 1. All localities from the Indian subcontinent where stingless bees have been studied according to the references listed in appendix 1. Notice occurrences near the border of Pakistan and within Bangladesh and Nepal. Possibly stingless bees can be found in parts of Pakistan and Bhutan as well.

Material and methods Morphological terminology follows that of Michener (2007) and Camargo & Pedro (2009), including the traits measured in the latter. Photographs were prepared using a Canon 7D digital camera with a MP-E 65 mm lens and stacked with the Zerene Stacker v1.04 software package. Measurements in mm were made with an ocular micrometer on a Leica stereomicroscope. Abbreviations used are HW = head width, WL = wing length, WD =

STINGLESS BEES OF THE INDIAN SUBCONTINENT

Zootaxa 3647 (3) © 2013 Magnolia Press ·

403

wing diagonal which is the distance between M-Cu bifurcation and basal tip of marginal cell, HTL = hind tibia length. The primary types of all species described from India and Sri Lanka, except Lisotrigona mohandasi, were examined; the loaning institutions are BMNH = Natural History Museum, London, England; OUMNH = Oxford University Museum of Natural History, Oxford, England; USNM = National Museum of Natural History, Washington DC, USA. To produce distribution map 1, all references listed in Appendix 1 were searched for occurrence records of stingless bees (only records with detailed locality indication or complete coordinates were included). Due to uncertainty of the identification of species of Tetragonula, I have not distinguished between the different species reported and all records have simply been entered as a distribution record at a higher taxonomic rank than genus. A total of 155 unique records were entered as coordinates in DIVA-GIS v7.5 (Hijmans et al. 2001). To estimate the potential full distribution of stingless bees from the Indian subcontinent, I used Maxent v3.3.3k software with default parameters for modeling of species’ geographic distributions (Phillips et al. 2006). Modeling was based on 19 bioclimatic variables from the WorldClim database (Hijmans et al. 2005) at a spatial resolution of 2.5 arc-min (bioclimatic variables included: annual mean temperature, mean diurnal range, isothermality (temperature mean/ range), temperature seasonality (coefficient of variation), max temperature of warmest period, min temperature of coldest period, temperature annual range, mean temperature of wettest and driest quarter, mean temperature of warmest and coldest quarter, annual precipitation, precipitation of wettest and driest period, precipitation seasonality, precipitation of wettest and driest quarter, precipitation of warmest and coldest quarter). Model results were processed and visualized using the DIVA-GIS v7.5 (Hijmans et al. 2001).

Systematics Key to the identification of the known stingless bees from the Indian subcontinent (workers only) 1. -. 2.

-. 3. -. 4. -.

5. -.

Inner surface of hind basitarsus uniformly setose, without a differentiated basal sericeous area (Fig. 2f) . . . . . . . . . . . . . . . . . 2 Inner surface of hind basitarsus with a differentiated basal sericeous area of short, dense, hairs (Figs 1a–b). (Tetragonula Moure 1961) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Head and thorax with dense tessellation and without shiny interspaces; densely plumose (“scale-like”) hairs present on the margin of mesoscutum (Fig. 7c); body and wing each more than 4 mm in length (Fig. 2) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lepidotrigona arcifera (Cockerell 1929) [or other species of the genus Lepidotrigona Schwarz]. Head and thorax with brilliant and polished integument (Fig. 3i); mesoscutum without densely plumose hairs; body and wing each less than 3 mm in length. (Lisotrigona Moure 1961) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3 Head and thorax dark brown to black with metasoma brown (Fig. 3); clypeus with minute hairs; metanotum finely imbricate (see also text for a discussion of the two Indian Lisotrigona species) . . . . . . . . . . . . . . . . . . Lisotrigona cacciae (Nurse 1907) Head and thorax all black with metasoma brown; clypeus with plumose hairs; metanotum reticulate anteriorly. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Lisotrigona mohandasi Jobiraj & Narendran 2004 Larger species with forewing length, including tegulae, between 4.2 and 4.5 mm; head width more than 1.7 mm. Mesoscutum with bands of pubescence weakly defined, not as clearly defined with glabrous interspaces as on Fig. 1d. . . . . . . . . . . . . . . . 5 Smaller species with forewing length, including tegulae, between 3.5 and 4.2 mm; head width 1.5–1.8 mm. Mesoscutum with distinct bands of pubescence separated by broad glabrous interspaces (Fig. 1d) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . “iridipennis” species group (See tables 1 and 2). Antenna ventrally mostly testaceous to ferruginous . . . . . . . . . . . . . . . . . . . . . . . . . . . Tetragonula aff. laeviceps (Smith 1857) Antenna ventrally totally blackish brown to black . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Tetragonula gressitti (Sakagami 1978)

Lepidotrigona arcifera (Cockerell 1929) (Figs 2a–l, map 2) Trigona arcifera Cockerell 1929: 591–592: Holotype (BMNH 17b.1081, worker): examined, “Type” (red border), “B.M. TYPE / HYM. / 17B.1081”, “Trigona / arcifera Ckll / TYPE”. “Trigona / ventralis Sm / ♀ from Sladen”, “Teesta / Bridge / Himalayas / India / 10.1.97”, “T.D.A. Cockerell / B.M.1936-415”. Type locality: INDIA, Sikkim, Teesta bridge (on January 10, 1897) [ca. 27.11°N, 88.47°E].

Provenance: Theodore Dru Alison Cockerell (1866–1948) received a single worker specimen probably directly through the collector, Frederick William Lambert Sladen (1876–1921) (see Sladen 1915). Sladen, who was an apiculturist and bumble bee worker, had collected numerous bees in the Himalayan parts of India (Anon 1921),

404 · Zootaxa 3647 (3) © 2013 Magnolia Press

RASMUSSEN

which were described by Cockerell and other authors (e.g., Blüthgen 1926). Cockerell misread the handwritten specimen label (Fig. 2a) as it is not “Testa” Bridge, but rather the old “Teesta” Bridge across the river Teesta or Tista, flowing for most of the length of the Sikkim state bordering West Bengal. The bridge joins the road between Kalimpong and Darjeeling. The locality is actually near the border of Nepal and thus suggests the hitherto formally unrecorded presence of the genus Lepidotrigona in Nepal.

FIGURE 1. 1A. Scanning electron microscope image demonstrating the inner surface of hind tibia and basitarsus, the latter with the differentiated basal sericeous area of short, dense, hairs characteristic of the genus Tetragonula of the Indian subcontinent (specimen is Heterotrigona itama Cockerell from Malaysia). 1B. Close-up of the hind basitarsus with sericeous area of short, dense, hairs. 1C. Lepidotrigona with diagnostic dense tessellation on head and thorax and densely plumose hairs on the margin of mesoscutum (Lepidotrigona cf. arcifera from India, Nagaland, collection 21). 1D. The “iridipennis” species group is characterized by distinct bands of pubescence separated by broad glabrous interspaces on the mesoscutum (specimen from India, Tamil Nadu, collection 17). Notice that the latter two specimens are laterally glued directly to the entomological pin, thus, avoiding to pin through the mesoscutum and obscure characters from this area.

Notes on the type: The holotype is unfortunately slightly damaged, with the left hind leg (Figs 2f–g), the left antenna (Fig. 2d), and abdomen (Figs 2i–j) glued to a point-mount. The right hind leg and the right antenna are missing. Due to the glue on the abdomen, the extent of the dark band on the pale metasomal tergum 1 cannot be properly determined (Fig. 2i). It appears to form a complete band, with a projection in the center (condition B3 in Fig. 11 of Sakagami 1975); the hairs on vertex, scutellum and mid tibia are very light to pale brown. Comments: This species was tentatively recognized as a valid species within the “ventralis” species group (Rasmussen 2008), although Schwarz (1939) suggested it should be synonymized with L. flavibasis (Cockerell 1929) and Moure (1961) suggested synonymization with L. hoozana (Strand 1913b). Within Lepidotrigona, the “ventralis” species group is characterized by smaller body size (body and forewing length each less than 5 mm),

STINGLESS BEES OF THE INDIAN SUBCONTINENT

Zootaxa 3647 (3) © 2013 Magnolia Press ·

405

FIGURE 2. Lepidotrigona arcifera (Cockerell): holotype of Trigona arcifera. 2A. Labels. 2B. Dorsal view of head and mesosoma (metasoma glued separately to the point mount). 2C. Frontal view of head. 2D. Antenna glued to point mount. 2E. View of propodeum. 2F. Inner (posterior) view of left hind leg, glued to point mount. 2G. Outer (anterior) view of left hind leg. 2H. Left lateral view of head and mesosoma. 2I. Dorsal view of metasoma, notice adherent; hiding part of the basal tergum. 2J. Ventral view of metasoma. 2K. Right ventral view of head and mesosoma. 2L. Right forewing.

406 · Zootaxa 3647 (3) © 2013 Magnolia Press

RASMUSSEN

and the diagnostic densely plumose mesoscutal hairs usually whitish (not yellow) and not extending to the scutellum (Figs 2b, 7c). The scutellum is bordered with rather long, erect hairs. The hind tibia in the “ventralis” group is not as apically expanded (spoon-like) as it is in some of the other species of the genus (i.e., “nitidiventris” species group). No taxonomic revision exists, but species are traditionally distinguished by a combination of body size, integument coloration on the first and apical metasomal terga and coloration of hairs on the vertex, scutellum, and posterior margins of the mid and hind tibiae (Schwarz 1939, Sakagami 1975). Two species of the same species group that occur the nearest and could be expected in the subregion are L. ventralis (Smith) and L. flavibasis. The first species can be recognized by the two lateral and separate dark spots on the otherwise pale metasomal tergum 1 (forms a complete band in L. arcifera) while the second species can be recognized by the semi-circle of black integument, partly enclosing the basal depression on the otherwise pale tergum 1, as in L. arcifera; L. flavibasis can be recognized by the apical metasomal terga brownish to blackish (yellowish on L. arcifera, Fig. 2i) and the fore and middle tibiae with blackish hairs on the external surface (Schwarz 1939). Such characters do show intraspecific variation over a large geographical range, which makes it difficult to separate and delimit the species convincingly. Likely, the separation of the different species has to be done while examining longer series of specimens from more localities, and including associated males, information on nest architecture, and molecular data.

Lisotrigona cacciae (Nurse 1907) (Figs 3a–i, map 2) Melipona cacciæ Nurse 1907: 619: Lectotype (BMNH 17b.1103, worker): examined, “LECTOTYPE” (blue border), “Type” (red border), “B.M. TYPE / HYM. / 17B.1103”, “Melipona / cacciæ / (Nurse)”, “Col. C.G. Nurse / Collection. / 1920-72”, “Hoshung- / -abad / Type”, “LISOTRIGONA / cacciae (Nurse) / det. M. S. Engel, 1999”, “LECTOTYPE / Melipona / cacciae Nurse / design. J. S. Moure, 1961 / (ref: Engel, Oriental Insects, 2000)”. Type locality: INDIA, Madhya Pradesh, Hoshangabad [ca. 23.25°N, 78.2°E].

Provenance: Charles George Nurse (1862–1933) based his description on specimens collected by A.M.F. Caccia, of the Indian Forest Department, and presented to Nurse though the courtesy of Isaac Henry Burkill (1870–1965), reporter on Economic Products to the Government of India since 1902. Comments: Engel (2000) synonymized Trigona scintillans Cockerell, 1920 from Sandakan, Borneo, under L. cacciae (from India). Engel (2000) found no structural differences between the specimens and although coloration varied both within and between specimens from different locations, he pointed out that even the original type series of L. cacciae from a single locality, as described by Nurse (1907), included specimens with variable coloration (“abdomen dark red … [to] almost black in some specimens”). Jobiraj and Narendran (2004) in their account of Lisotrigona from India did not have access to the type specimen of L. cacciae and made a few misinterpretations in their comparisons to this species: in L. cacciae the head is near black in some specimens (not brown to dark brown), the gena is finely punctate (not impunctate) (Fig. 3g), pubescence on clypeus is too short to clearly distinguish whether or not it is plumose (stated as being simple) (Fig. 3d), but the hypoepimeron (immediately above the scrobal groove) has very fine hairs (not without hairs). In addition, the metanotum is imbricate but approaching reticulate sensu Harris (1979). The diagnostic characters for L. mohandasi thus approach L. cacciae, and a direct comparison of the two primary type specimens is necessary to encounter reliable diagnostic characters for their separation.

Lisotrigona mohandasi Jobiraj & Narendran 2004 (map 2) Lisotrigona mohandasi Jobiraj & Narendran 2004: 39, 40–43, figs 1–4: Holotype (Calicut University Zoology Department, worker); paratypes (Jobiraj collection (2)): Not examined. Type locality: INDIA, Kerala, Thrissur, Peechi, Kerala Forest Research Institute (on April 3rd, 2000) [ca. 10.53°N, 76.35°E].

Provenance: T. Jobiraj and Thekke Curuppathe Narendran described this species based on three specimens collected by Dr. K. Mohandas.

STINGLESS BEES OF THE INDIAN SUBCONTINENT

Zootaxa 3647 (3) © 2013 Magnolia Press ·

407

FIGURE 3. Lisotrigona cacciae (Nurse): lectotype of Melipona cacciae. 3A. Habitus, dorsal view. 3B. Frontal view of head. 3C. Habitus, left lateral view. 3D. Lower frontal view of head, including partially hidden mandibles. 3E. Labels. 3F. Habitus, right lateral view. 3G. Lateral view of head including genal area. 3H. Right forewing. 3I. View towards propodeum.

Comments: The holotype of L. mohandasi was not examined. Some of the characters indicated by Jobiraj & Narendran (2004) for separating the present species from L. cacciae are incorrect as discussed under L. cacciae.

408 · Zootaxa 3647 (3) © 2013 Magnolia Press

RASMUSSEN

Tetragonula species of the “iridipennis” species group Diagnosis: Amongst Tetragonula species groups members of the “iridipennis” group are characterized by having a dark mesoscutum with four distinct hair bands separated by broad glabrous interspaces (Fig. 1d), and by their smaller body size, with the forewing length, including tegula, measuring 3.0 to 4.3 mm. In addition, Sakagami (1978) reported that the male gonostylus arises dorsad (not laterad) to the gonocoxite in all species of this group. Comments: Besides the Indian species of Tetragonula (i.e., T. bengalensis (Cameron), T. iridipennis (Smith), T. praeterita (Walker) and T. ruficornis (Smith)), the group also includes the smallest member of the genus, the apparently widespread T. fuscobalteata (Cameron 1908) and T. clypearis (Friese 1909) from Australia and New Guinea (Dollin et al. 1997). A last species in the group is T. pagdeni (Schwarz 1939) described based on the holotype (in USNM) male collected from Nakhon Si Thammarat in southern Thailand (Sakagami 1978), but widespread in all of southeast Asia (Sakagami et al. 1990). Sakagami (1978) reported that T. pagdeni workers were only statistically different from T. iridipennis s.l. from India (the former with paler coloration of the anterior corbicular fringe and more plumose frontal hairs), although males of T. pagdeni did differ in genitalic characteristics from the Indian specimens of T. iridipennis s.l. Preliminary molecular results (Rasmussen & Cameron 2007, 2010) placed four genetically distinct but morphologically “iridipennis”-like specimens from India in a highly supported clade together with specimens identified as T. pagdeni (from Malaysia), T. clypearis (from Australia) and, surprisingly, also T. minor (Sakagami 1978) (from Thailand). Tetragonula minor is a very small member of the genus and is tentatively placed in this species group, as the male is unknown and it cannot be assessed whether the male gonostylus arise dorsad from the gonocoxite. Workers of Tetragonula minor are distinguished from the other members by the darker hairs of the mesoscutum and corbicular fringe. The principal measurements of the species of this species group are given in table 1 (on average HW=1.6–1.8, WL=3.7–4.3, WD=0.9–1.2, HTL=1.4–1.7) and as follows for species not encountered in the Indian subcontinent: T. fuscobalteata: HW=1.35–1.5, WL=3.1–3.4, WD=0.9–1.0, HTL=1.2–1.4; T. clypearis: HW=1.57, WL=3.55, WD=0.97, HTL=1.46; T. minor: HW=1.6, WL=3.6–3.8, WD=1.0, HTL=1.5; T. pagdeni: HW=1.6–1.8, WL=3.9– 4.1, WD=1.0–1.2, HTL=1.4–1.7. The amber fossil Meliponorytes devictus Cockerell 1921 from the Hukawng Valley, Myanmar, has been placed under junior synonymy of T. iridipennis (Zeuner & Manning 1976), but could represent an additional member of this species group. Although Hukawng amber generally is recognized as old (Cruickshank & Ko 2003), this particular piece is copal of much more recent but unknown age (Grimaldi et al. 1995). TABLE 1. Measurements (mm) of the examined primary type specimens of species described from the Indian subcontinent and listed above.

Total length of body

Lepidotrigona arcifera 4.70

Lisotrigona Tetragonula cacciae iridipennis 2.95 3.55

T. T. praeterita ruficornis 3.33 3.45

T. bengalensis 3.55

Width of head

1.89

1.19

1.60

1.52

1.66

1.70

Length of head (clypeal apex-vertex)

1.51

1.01

1.30

1.32

1.32

1.34

Length of compound eye

1.13

0.83

1.10

1.04

1.13

1.13

Width of compound eye

0.47

0.33

0.40

0.40

0.42

0.43

Upper interorbital distance

1.23

0.75

1.00

1.04

1.02

1.02

Maximum interorbital distance

1.34

0.85

1.13

1.12

1.14

1.13

Lower interorbital distance

1.13

0.65

0.82

N/A

0.79

0.87

Diameter of median ocellus

0.08

0.11

0.15

0.14

0.14

0.15

Interocellar distance

0.36

0.27

0.33

0.35

0.38

0.38

Ocellorbital distance

0.32

0.18

0.23

0.24

0.24

0.23

Interalveolar distance

0.19

0.13

0.15

0.15

0.17

0.17

Alveolorbital distance

0.32

0.19

0.30

0.27

0.31

0.26

Alveolocellar distance

0.75

0.54

0.71

N/A

0.67

0.67

......continued on the next page

STINGLESS BEES OF THE INDIAN SUBCONTINENT

Zootaxa 3647 (3) © 2013 Magnolia Press ·

409

TABLE 1. (Continued)

Alveolar diameter

Lepidotrigona arcifera 0.10

Lisotrigona Tetragonula cacciae iridipennis 0.06 0.14

T. T. praeterita ruficornis 0.15 0.12

T. bengalensis 0.12

Length of clypeus

0.36

0.24

0.33

N/A

0.39

0.33

Maximum width of clypeus

0.71

0.42

0.70

N/A

0.71

0.65

Intertentorial distance, width of clypeus 0.56

0.36

0.48

N/A

0.48

0.51

Clypeocellar distance

0.24

0.06

0.96

N/A

0.06

0.14

Length of malar space

0.13

0.02

0.03

N/A

0.05

0.06

Length of scape

0.60

0.37

0.57

0.59

0.57

0.57

Diameter of scape

0.13

0.07

0.10

0.08

0.10

0.10

Diameter of third flagellomere

0.14

0.11

0.12

0.13

0.13

0.13

Length of pedicel + flagellomeres

1.35

N/A

1.24

N/A

1.25

1.30

Length of first flagellomere

0.10

0.06

0.09

0.05

0.07

0.11

Length of second flagellomere

0.10

0.06

0.07

0.12

0.11

0.11

Length of third flagellomere

0.12

0.07

0.10

0.10

0.11

0.11

Length of mandible

0.65

0.45

0.62

N/A

0.62

0.60

Width of mandible

0.21

0.12

0.19

N/A

0.21

0.18

Length of forewing (excluding tegula)

4.25

2.37

3.44

3.44

3.20

3.50

Length of forewing (including tegula)

4.60

2.65

3.80

3.70

3.50

3.80

Width of forewing

1.65

0.95

1.32

N/A

1.28

1.25

Length of pterostigma

0.71

0.36

0.48

0.48

0.51

0.48

Width of pterostigma

0.19

0.12

0.11

0.11

0.13

0.13

Length of marginal cell

1.40

0.89

1.21

1.21

1.23

1.21

Width of marginal cell

0.32

0.18

0.30

N/A

0.30

0.29

Length of first abscissa of M

0.77

0.38

0.58

0.60

0.52

0.55

Length of first abscissa of Cu

0.89

N/A

0.71

0.70

0.70

0.71

Length of wing diagonal

1.33

0.73

1.01

0.95

0.95

0.95

Hamuli

6

6

5

5

5

5

Length of mesoscutum

1.08

0.71

0.87

0.85

0.83

1.00

Width of mesoscutum

1.26

0.93

1.01

1.00

1.05

1.13

Width of scutellum

0.94

0.48

0.57

0.61

0.69

0.79

Length of scutellum

0.28

0.23

0.33

0.24

0.33

0.28

Length of tibia III

1.43

0.86

1.55

1.47

1.43

1.51

Width of tibia III

0.57

0.31

0.54

0.54

0.52

0.54

Length of basitarsus III

0.62

0.35

0.50

0.48

0.48

0.50

Width of basitarsus III

0.26

0.18

0.29

0.27

0.27

0.29

Width of tergum III

1.58

1.05

1.36

1.28

1.32

1.28

Length of hairs on clypeus