Austroplebeia cassiae (Cockerell)

Dollin, Anne E., Dollin, Leslie J. & Rasmussen, Claus, 2015, Australian and New Guinean Stingless Bees of the Genus Austroplebeia Moure (Hymenoptera: Apidae) — a revision, Zootaxa 4047 (1), pp. 1-73 : 54-60

Austroplebeia cassiae (Cockerell)

( Figures 1–17 , 19–20 , 22–25 , 27–30 , 32–33 )

Trigona cassiae Cockerell 1910: 247 .

Trigona (Plebeia) australis— Michener 1961: 5 (proposed synonym of Trigona cassiae ); Austroplebeia cassiae — Moure 1961: 195, 196, 197; Michener 1990: 96, 101, 105, 133, 134. Trigona (Plebeia) cassiae — Michener 1965: 230;

Trigona symei Rayment 1932: 106 ;— Rayment 1935: 737 (more detailed redescription as new species). New Synonymy. Austroplebeia symei — Moure 1961: 197; Michener 1990: 134.

Trigona (Plebeia) symei — Michener 1965: 230.

Diagnosis. In workers and males, frons and mesepisternum with plumose hair ( Figs 22 e, 23e, 25 b, d, f), head width ≥ 1.70 mm ( Figs 10 , 14 ). In workers, HBW 0.35–0.44 mm ( Fig. 11 ). Most workers have the following characteristics: thorax markings Grades 0–2 ( Figs 5 , 7 ), pronotal collar and legs without maculations, plumose hair on clypeus dense, integument barely visible among hairs ( Fig. 22 e), bristles on basitarsus III and scutellum brown. In males, gonostylus gently curved, thorax markings Grades 6 and 11–13, mesepisternum plumose hairs narrow and feather-like, with short closely set barbs ( Fig. 25 f), HBW 0.22–0.28 mm ( Fig. 13 ), HTL 1.19–1.29 mm ( Fig. 14 ). In queens, short dense hair covering T4–T6 ( Fig. 27 a). Distribution QLD ( Fig. 28 d).

Description. As in general Austroplebeia description, except for the following characters.

Worker. Body 3.4–4.5 mm. Forewing (including tegula) 4.1–4.7 mm mm. All other measurements are listed in Table 4 .

Structure. Head width 1.70–2.01 mm, 1.2–1.3x head length. Interocellar distance 1.4–1.7x ocellorbital distance. Clypeus length 0.5x width. Malar area 0.2–0.3x basal mandibular width. Mandible preapical teeth more widely spaced than in A. cincta and slightly more deeply incised than in A. essingtoni or A. australis ( Fig. 17 g). Scape length 0.7–0.8x alveolus–lateral ocellus distance.

Scutellum length 0.4–0.5x width. Wing venation similar to that shown in Fig. 18 . First transverse cubital vein vestige present in 11% of wings examined and slight trace in 69% (n=98). Vein M of forewing terminates at or before position of anterior end of first recurrent vein in 99% of wings examined (n=98). Cell second Cu of forewing completely indicated by pigmented vein traces in 79% of wings examined (n=97). Hamuli 5 to 6 per wing. HBW broad ( Figs 11 , 19 h), 0.35–0.44 mm.

Sting rudiments. Gonostyli shoulder width moderately narrow, 0.11–0.13 mm. Apex of sting stylet broadly rounded or angular, occasionally with a narrowly rounded or acute protrusion. Sting lancet length moderately long, 0.28–0.34 mm ( Figs 15 , 16 g).

Colour. Face markings usually Grade 3, ranging 1–4 ( Figs 1 , 3 ). Thorax markings usually Grades 0–2, ranging 0–5 ( Figs 5 , 7 ), Grade 2 maculations: 41% brown, 59% cream.

Other areas with cream markings: pronotal lobe (sometimes yellow-brown); sometimes labrum dull cream, tegula and tibia I with a cream spot. Scape, pronotal collar, metasoma and other leg segments without maculations.

Labrum yellow-brown, ranging dark brown to cream. Mandible: orange-brown medially, ranging red-brown to cream ( Fig. 9 ); dark brown to orange-brown apically. Scape dark brown, sometimes red-brown. Tegula red-brown, ranging black to brown, sometimes with a small cream spot. Basal and medial terga black to red-brown. Apical terga dark brown, ranging black to brown ( Fig. 9 ). Sterna ranging black to brown. Tibia I dark brown or red-brown. Distitarsus usually red-brown to orange-brown.

Pilosity. As in A. essingtoni , with the following differences:

Face with numerous appressed short whitish plumose hairs (often worn thin on elevated areas) admixed with sparse erect simple hairs (ca. 50 Μm). Plumose hair on paraocular area and supraantennal area longer, wellbranched and dense, hiding most of the surface. Density of plumose hair on clypeus only slightly finer than that on paraocular area, in unworn specimens hiding most of surface and often obscuring upper margin of clypeus ( Fig. 22 e). Vertex bristles slightly brown or brown, 75–200 Μm.

Scutum short hairs ca. 50 Μm, longer hairs 100 Μm, lateral and anterior fringes, 150–225 Μm, hair colour varying from pale to slightly brown. Scutellum longer bristles usually brown and 100–300 Μm. Mesepisternum longer simple hairs up to 300 Μm. T6 simple hairs up to 175 Μm.

In mature, fully coloured specimens: simple hairs of anterior and posterior corbicular fringes of tibia III usually slightly or distinctly brown, hairs on apical half of anterior fringe distinctly brown, penicillum and rastellum bristles brown, basitarsus III inner surface with coarse brown bristles, ranging from mid-brown to deep opaque brown.

Geographic variation. The extent of thorax colour markings in A. cassiae was 0–4% ( Fig. 10 ).

Average colony head width in workers was generally 1.72–1.90 mm. However, two colonies (Q31, Q32) at Conjuboy, QLD, had unusually high average head widths of 1.94–1.98 mm, the largest observed in this study ( Fig. 10 ).

In some colonies at the southern end of the A. cassiae range, worker bees displayed mixed characteristics in thorax colouration, basitarsus bristle colour, clypeus hair density and HTW/HBW scores. For instance, those in colony D46 had dark thorax colouration typical of A. cassiae (0%), but had relatively fine clypeus hair and golden to slightly brown bristles on the inner basitarsus III that were similar to A. australis .

Male. Body 4.0– 4.7 mm. Forewing (including tegula) 4.1–4.5 mm. All other measurements are listed in Table 5 .

Structure. Head width 1.75–2.00 mm. Interocellar distance 2.2–2.8x ocellorbital distance. Clypeus length 0.5– 0.6x width. Mandible preapical tooth weaker than in A. cincta , similar to A. australis ( Fig. 17 l). Scape length 0.7– 0.8x alveolus–lateral ocellus distance. Flagellum relatively longer than in A. cincta , flagellum length 4.1–4.5x scape length.

Hamuli per wing 5–6. HBW moderate, 0.22–0.28 mm. HTW moderate, 0.43–0.48 mm ( Fig. 13 ). HTL long, 1.19–1.29 mm ( Fig. 14 ).

Terminalia and genitalia. Graduli of S3 and S4 gently curved, but more acutely curved and almost touching antecosta in some males from Cape York, QLD. S5 gradulus transverse and usually very close to or touching antecosta, apical margin smooth or slightly curved. S6 with median body width variable but relatively broad and parallel sided or with slight medial constriction, apically pointed ( Figs 20 d–e, j, n). S7 quadrangular, apical margin transverse (similar to Fig. 20 q). Gonostylus gently curved. Base of penis valve relatively broad and short (similar to Figs 21 c, e).

Colour. Face markings usually Grade 17, sometimes 18 ( Figs 2 , 4 ). Thorax markings usually Grade 11, sometimes 6 or 12–13 ( Figs 5 , 8 ).

Other areas with cream or yellow markings: ventral scape, labrum and pronotal lobe, all femora and tibiae; usually mandible, a spot on tegula, apical terga, sterna and basitarsi; sometimes trochanters and spots on upper edge of pronotal collar.

Scape dark brown to brown dorsally, cream ventrally. Flagellum red-brown to yellow-brown. Tegula redbrown, yellow brown or transparent, sometimes with a cream spot. Metanotum ranging dark brown to cream. Basal and medial terga black to brown. Apical terga cream, ranging red-brown to yellow ( Fig. 9 ). Sterna ranging black to cream. Unmarked areas of tibiae dark brown to yellow-brown.

Pilosity. As in A. essingtoni , with the following differences:

Clypeus with an even layer of short fine plumose or simple hair ( Fig. 23 e). Vertex bristles, 100–275 Μm.

Scutum with numerous ca. 50 Μm and sparse longer simple hairs (ca. 125 Μm), up to 250 Μm laterally. Scutellum simple hairs up to 275 Μm. Mesepisternum plumose hairs narrow and feather-like, with barbs shorter and more closely spaced than in A. australis ( Figs 25 d, f).

In mature, fully coloured specimens, basitarsus III inner surface bristles pale, golden or brown.

Surface sculpture. As in A. essingtoni .

Queen. Body 4.5–6.6 mm. Forewing (including tegula) 4.1–4.7 mm.

Structure. HBW 0.23–0.29 mm.

Colour. Face markings up to Grade 3 ( Fig. 24 e). Thorax markings up to Grade 3.

Pilosity. Clypeus with moderate to dense simple hair, some plumose hair near apical margin. Scutellum bristles slightly brown to brown, and up to 425 Μm. T1–T3 with sparse fine short hairs, longer coarse bristles on lateral and dorsal T2–T5 generally absent. T4–T6 with short dense hair ( Fig. 27 a), T6 with an apical fringe of long (up to 325 Μm) simple hairs. Basitarsus III inner surface bristles golden to brown.

Material examined. Holotype of Trigona cassiae : ( BMNH 17b.1137, worker), labelled as follows: “ Type (red border), “B.M. TYPE / HYM. / 14B.1137, “ Trigona / cassiae / Ckll TYPE, “Turner Coll. / 1912–111, “Mackay / 12.99 / Cassia , 900.

Provenance of types: Rowland Edwards Turner (1863–1945) and/or his brother, Gilbert Turner (ca. 1866– 1903), collected the T. cassiae holotype (“Turner, 900) in December 1899 in Mackay, QLD, from flowers of Cassia (Leguminosae) . Cockerell (1930) documented the intensive collecting work of the Turner brothers. The Turner collection of Australian bees was deposited in the BMNH where Cockerell examined and later borrowed it. The holotype is still in BMNH.

According to the type description ( Cockerell 1910), a male was also collected from the same location by Turner in November 1891 (“Turner, 318). However, this specimen could not be found in the BMNH collection nor their database (D. Notton, pers. comm., 2015).

Holotype of Trigona symei . ( ANIC, worker), labelled as follows: “ TYPE, “ Trigona / symesi Raym / NT. QUEENS, “ AUST. NAT. / INS. COLL.

Provenance of type: According to the detailed species redescription in Rayment (1935), the T. symei holotype was collected in 'North Queensland'. The exact locality and the collector are unknown. Rayment dedicated this new species to Geoffrey Syme of Melbourne, in appreciation for his interest in preserving the fauna and fauna of Victoria. This may have been Sir Geoffrey Syme (1873–1942), the Melbourne newspaper proprietor. The holotype is in ANIC.

Other worker and male specimens examined. Queensland: Nest samples, collected by AED & LJD except where otherwise stated: Cape York near Bamaga, Aug 1983 (Nest F3, 20 workers; Nest F6, 20 workers, 6 males; Nest F32, 20 workers, 6 males); Cape York on Jardine River, Aug 1983 (Nest F43, 20 workers, 5 males; Nest F44, 20 workers, 1 male); Coen, 25 Aug 1983 (Nest F47, 20 workers, 1 male); Alec Springs, 15k N of Strath-Haven Station, Nest G13 (5 workers); Laura, 21 Feb 1982 (Nest Q22, 20 workers); 5k S Rossville: Jun 2010 (Nest D12, 9 workers, 1 male, L. Roberts); Nov 2012 (Nest D18, 20 workers; Nest D19, 20 workers; Nest D21, 20 workers; Nest D22, 20 workers; Nest D24, 14 workers); Koolatah Station, 23–24 Sep 1998 (Nest M18, 20 workers; Nest M19, 20 workers, 2 males; Nest M20, 5 workers; Nest M21, 19 workers); 15k NW of Dunbar Station, 24 Aug 1990 (Nest G7, 20 workers; Nest G9, 15 workers); Conjuboy, Feb 1982 (Nest Q31, 9 workers; Nest Q32, 20 workers); Rockhampton, 7–8 Sep 1996 (Nest C14, 20 workers; Nest C17, 20 workers; Nest C28, 29 workers; Nest C29, 11 workers); Duaringa, Sep 1987 (Nest N23, 20 workers, 15 males); 17–22 Oct 1987 (all by E. Adams: Nest N24, 8 workers; Nest N25, 9 workers; Nest N26, 11 workers), 15 Nov 1987 (Nest N31, 10 workers, E. Adams), 10–12 Sep 1996 (Nest C31, 20 workers; Nest C36, 20 workers; Nest C37, 20 workers; Nest C38, 5 workers; Nest C42, 20 workers; Nest C43, 20 workers), 12 Oct 1996 (Nest C58, 20 workers); Bororen, 8 Sep 1990 (Nest G34, 5 workers); Gin Gin, 18–21 Oct 1996 (Nest C79, 12 workers, 3 males; Nest C91, 10 workers); from Rainbow Beach but kept at Ipswich, Sep 2013 (all by C. Heather: Nest D36, 18 workers; Nest D45, 21 workers; Nest D46, 16 workers; Nest D47, 15 workers).— Samples collected from flowers: 61k N Coen. 2 Nov 1988, K Walker, NMV, 1 worker; McIlwraith Range, 30k E Coen, 3 Nov 1988, K Walker, NMV, 2 workers, 25k E Coen, 12 Nov 1988, K Walker, NMV, 1 worker; 13k SE Coen, 1 Nov 1988, K Walker, NMV, 2 workers; 45k S Coen, 30 Oct 1988, K Walker, NMV, 1 worker; Silver Plains, 55k SE Coen, 1 Nov 1988, K Walker, NMV, 3 workers, 42k SE Coen, 1 Nov 1988, K Walker, NMV, 1 worker; 27k SE Coen, 13 Nov 1988, K Walker, NMV, 5 workers; 28k S Cooktown, 15 Nov 1988, K Walker, NMV, 3 workers; Mt Carbine, 27 Oct 1988, K Walker, NMV, 1 worker; 30k N Normanton, 21 Aug 1990, AED & LJD, GS-10: 8 workers; 20k E Georgetown, 17 Aug 1990, AED & LJD, GS-5: 4 workers; 66k S Townsville, 17 Apr 2005, M Batley, AMS, 1 worker; 10k S Bowen, 14 Feb 1982, AED & LJD, QS-5: 8 workers; Halliday Bay, 50k NW Mackay (label states '50k NE Mackay'), 19 Sep 1983, NW Rodd, AMS, 2 workers; 36k N Clairview, 20 Apr 2005, M Batley, AMS, 1 worker; 14k SW Rainbow Beach, 23 Jan 2006, M Batley, AMS, 1 worker.

Queen specimens examined. Nest samples collected by AED & LJD, 18 queens: Queensland: Cape York near Bamaga, Aug 1983 (Nest F6, 2 virgin; Nest F32, 1 gravid, 8 virgin); Cape York on Jardine River, Aug 1983 (Nest F43, 1 gravid; Nest F44, 1 virgin); Conjuboy, Feb 1982 (Nest Q32, 1 gravid); Duaringa, Sep 1987 (Nest N23, 1 gravid, 2 virgin), 10–12 Sep 1996 (Nest C37, 1 virgin).

Remarks. The Austroplebeia populations with the darkest colouration (Thorax Grades predominantly 0–2, Fig. 5 ) were called the 'symei' group in our preliminary studies ( Halcroft et al. 2015), because the species name currently in common use for these colonies in QLD is A. symei . However, the present study has concluded that the names A. cassiae and A. magna sp. nov. should be used for these populations in QLD and NT respectively.

The A. cassiae and A. symei holotypes. The A. cassiae holotype was collected by Turner in 1899 at Mackay, QLD. This locality is near the centre of the QLD 'symei' group distribution ( Fig. 28 d). With a large HBW of 0.41 mm ( Fig. 11 ; Table 4 ), the A. cassiae holotype is consistent with the QLD 'symei' group. Most of the clypeus pilosity is contaminated but it appears dense.

The A. cassiae holotype specimen is now discoloured. The scutum and head, normally black, are reddish brown, rendering scoring of colour maculations unreliable. However, in his type description, Cockerell (1910) provided a quite detailed account of its original colouration. He stated that the specimen was 'very like [ Tetragonula ] carbonaria '—a very dark species without any cream maculations. He noted that the axillae were black and the scutellum had a broad dusky yellow band: equivalent to Thorax Grade 2 in this study. He added that the hair on the scutellum was 'fuscous' (or very dark brown). These characteristics all suggest that the A. cassiae holotype belongs to the QLD 'symei' group.

The holotype of A. symei is also consistent with the QLD 'symei' group in both colouration (Face Grade probably 4, Thorax Grade 2, Figs 1 , 3 , 5 , 7 ) and structure ( Fig. 11 ; Table 4 ). Furthermore it has dense hair on the clypeus, and dark bristles on the vertex, scutellum and inner basitarsus III, all characters consistent with the QLD 'symei' group. Its eyes are collapsed, preventing a meaningful head width measurement. It was collected in north QLD within the distribution of the QLD 'symei' group ( Fig. 28 d).

We consider that the holotypes of both A. cassiae and A. symei both belong to the QLD 'symei' group. As A. cassiae was described 22 years before A. symei , we conclude that the name of the QLD 'symei' group bees, by priority, should be A. cassiae .

The male Turner specimen described by Cockerell. In his type description, Cockerell (1910) also included a male specimen that was collected by Turner in November 1891. It was from Mackay, QLD, the same locality as the holotype, but was collected eight years earlier. We were unable to locate this specimen.

Cockerell described this male as “scutellum not pale banded”. All A. cassiae males examined in the present study had prominent cream or yellow maculations on the dorsal scutellum (Grades 6–13, Figs 5 , 8 , 25 d). So it is unlikely that this Turner male actually belonged to the species A. cassiae . Instead it may have belonged to the Tetragonula , Australia's other genus of stingless bees, in which the males lack cream maculations.

Earlier synonomy of A. cassiae and A. australis names. Another complication with the Austroplebeia nomenclature is that A. cassiae has sometimes been regarded as a junior synonym of A. australis . Michener (1961) discussed colour distinctions proposed by Cockerell (1930) then informally synonymised A. cassiae with A. australis . This synonymy was mentioned again in Wille & Michener (1973), but the two species were listed separately in Michener (1965, 1990) and Cardale (1993).

Cockerell (1930) had cited the description in Friese (1898) of the red-brown clypeus and yellow mandibles of A. australis . He suggested that these characters could be used to distinguish A. australis from A. cassiae . However, Michener (1961) found that the colouration of the clypeus and mandibles was variable. Hence he suggested that A. cassiae was a synonym for A. australis .

This study agrees with Michener (1961) that the colour of the clypeus and mandibles varies substantially within both species ( Fig. 9 ). It is not possible to separate A. cassiae and A. australis using these characters. However, this study has also shown that the A. cassiae holotype belongs to QLD populations that do differ markedly from A. australis in structure ( Figs 11 , 15 ), pilosity ( Figs 22 d–e, 25 e–f) and thorax colouration ( Figs 7 , 10 ). Hence A. cassiae and A. australis should be regarded as distinct species.

Workers with mixed characteristics in southern QLD. Austroplebeia cassiae was observed to retain distinct characteristics in a large, long-established sympatric population with A. australis near Duaringa in central QLD ( Fig. 12 ). However, as reported above, workers in some A. cassiae colonies in southern QLD displayed mixed characteristics in colouration, pilosity and structure. Some of these colonies had been transported into an area predominantly populated by A. australis and maintained there for some time prior to being sampled. Brito et al. (2014) reported some artificial hybridisation between closely related Australian Tetragonula species, which may have been facilitated by anthropogenic colony movements. Such mechanisms may also be contributing to the mixed characteristics observed in some A. cassiae colonies in southern QLD.

......continued on the next page *Variables: As given in Table 2.

Other than the types, the following worker specimens were measured. A. cassiae , N=16, from nests C37 (4x), F6 (4x), M19 (4x), Q32 (4x). A. magna , N=12 unless otherwise stated, from nests N1 (4x), N8 (4x), D65 (2x), D66 (2x).

Figures

FIGURE 1. Face colour marking grades used to classify Austroplebeia workers. In many specimens, these cuticle markings may be obscured by hair on the clypeus and paraocular area. Visualisation of these markings can be improved by tilting the bee's head back and viewing the face obliquely.

FIGURE 2. Face colour marking grades used to classify Austroplebeia males.

FIGURE 3. Variation in worker face colour grades observed within all Austroplebeia taxa except A. cincta. Data show 300 workers from 114 colonies, in addition to 50 workers from 10 flower samples of A. australis – Western Colour Morph. Colour grades are as illustrated in Fig. 1.

FIGURE 4. Variation in male face colour grades observed within all Austroplebeia taxa except A. cincta and A. australis – Western Colour Morph. Data show 186 males from 38 colonies. Colour grades are as illustrated in Fig. 2.

FIGURE 5. Thorax colour marking grades used to classify workers and males of Austroplebeia.

FIGURE 6. Number of different thorax colour marking grades found amongst workers within individual colonies of Austroplebeia. Data are based on 122 colonies and scores of 10 – 49 workers per colony.

FIGURE 7. Variation in worker thorax colour grades observed within all Austroplebeia taxa except A. cincta. Data show 2503 workers from 158 colonies, in addition to 59 workers from 15 flower samples of A. australis – Western Colour Morph. Colour grades are as illustrated in Fig. 5.

FIGURE 8. Variation in male thorax colour grades observed within all Austroplebeia taxa except A. cincta and A. australis – Western Colour Morph. Data show 196 males from 38 colonies. Colour grades are as illustrated in Fig. 5.

FIGURE 9. Variation in colour of the clypeus, mandible (middle part) and apical terga in workers and males of all Austroplebeia taxa. (No males are known in A. cincta NG and A. australis – Western Colour Morph.) Data are based on 120 workers and 62 males.

FIGURE 10. Ratio of head width and extent of thorax colour marking in workers of all Austroplebeia taxa except A. cincta. Data from 121 colonies are shown, in addition to five flower samples from A. australis – Western Colour Morph and one A. australis flower sample with mixed Western / Central characteristics. Data points are averages from 10 – 49 workers per colony, or 5 – 12 workers per flower sample. Data from eight individual type specimens are also shown. The T. symei and T. ornata types, in which the eyes are collapsed preventing accurate head width measurement, are not shown. The T. essingtoni holotype has a missing head, so data from the paratype are shown. 1. Colony F 31 from the northern tip of Cape York Peninsula, QLD, which has characteristics similar to A. australis – Eastern Colour Morph despite its geographic isolation. Names given in brackets in the legend are groups in our Halcroft et al. (2015) preliminary analysis.

FIGURE 11. Ratio of tibia III width and basitarsus III width in workers of all Austroplebeia taxa. Data from 173 colonies are shown, in addition to 26 flower samples: A. australis – Western Colour Morph (14 x), A. australis with mixed Western / Central characteristics (4 x) and NG A. cincta (8 x). Data points are averages from 2 – 4 workers per colony, or 1 – 12 workers per flower sample. Data from 12 individual type specimens are also shown. 1. Colony F 31 from the northern tip of Cape York Peninsula, QLD, which has characteristics similar to A. australis – Eastern Colour Morph despite its geographic isolation. Names given in brackets in the legend are groups in our Halcroft et al. (2015) preliminary analysis.

FIGURE 12. Ratio of tibia III width and basitarsus III width in workers in a large sympatic population of A. australis – Eastern Colour Morph and A. cassiae near Duaringa, QLD. Data points are averages from two workers per colony. Labelled data points are nests for which the species identification was confirmed in the molecular analysis of Halcroft et al. (2015). Names given in brackets in the legend are groups in our Halcroft et al. (2015) preliminary analysis.

FIGURE 13. Ratio of tibia III width and basitarsus III width in males of all Austroplebeia species. Data points are averages from 2 – 3 males per colony. Data from the male paratypes of A. magna sp. nov. are also shown. Names given in brackets in the legend are groups in our Halcroft et al. (2015) preliminary analysis.

FIGURE 14. Ratio of head width and tibia III length in males of all Austroplebeia species. Data points are averages from 2 – 3 males per colony. Data from the male paratypes of A. magna sp. nov. are also shown. Names given in brackets in the legend are groups in our Halcroft et al. (2015) preliminary analysis.

FIGURE 15. Worker sting rudiments metric data: ratio of maximum length of sting lancet (A) and shoulder width of gonostyli (B) in all Austroplebeia taxa. Data points are averages for 1 – 3 workers per nest or flower sample. 1. The geographically isolated colony, F 31, from the northern tip of Cape York Peninsula, QLD, which has colouring and HTW / HBW similar to A. australis – Eastern Colour Morph. 2. Colonies C 58 and D 55 from the ‘ intermediate’ group of Halcroft et al. (2015) which are consistent with A. australis – Eastern Colour Morph in this analysis. 3. CS- 8 workers, with Grade 4 Thorax Colour Markings, collected from flowers in the MacDonnell Ranges, NT. They are consistent with A. australis – Central Colour Morph in this analysis, despite their relatively dark colouration. 4. GS- 10 workers collected from flowers on the far western edge of the A. cassiae range. 5. NS- 7 workers collected from flowers on the far eastern edge of the A. magna sp. nov. range. Names given in brackets in the legend are groups in our Halcroft et al. (2015) preliminary analysis.

FIGURE 16. Austroplebeia worker sting rudiments: a, A. cincta NG; b. A. cincta QLD; c, A. essingtoni; d, A. australis – Eastern Colour Morph; e, A. australis – Central Colour Morph; f, A. australis – Western Colour Morph; g, A. cassiae; h, A. magna sp. nov. All drawn to same scale. Abbreviations: Gon—gonostylus; Sty—sting stylet (bulb); Lan—sting lancet (first valvula or first ramus).

FIGURE 17. Austroplebeia antennae, mandibles and labial palpi. Figs a – c. Antenna (basal part) of A. essingtoni: a, worker; b, male; c, queen. Figs d – h. Left mandible of worker: d, A. cincta QLD; e, A. essingtoni; f, A. australis; g, A. cassiae; h, A. magna sp. nov. Figs i – m. Left mandible of male: i, A. cincta QLD; j, A. essingtoni; k, A. australis; l, A. cassiae; m, A. magna sp. nov. Fig. n. Left mandible of A. australis queen. Figs a – n drawn to same scale. Figs o – q. Labial palpus of worker (mesal margins uppermost, small hairs omitted): o, A. cincta; p, A. cassiae; q, A. magna sp. nov. Figs o – q drawn to same scale. Abbreviations: F 1 — first flagellomere; F 2 — second flagellomere; F 3 — third flagellomere; F 2 L—second flagellomere length (showing how this measurement was made).

FIGURE 18. Austroplebeia wings. Dotted lines indicate pigmented vein traces. Figs a – c. Forewings, showing variation observed in the vestige of the first transverse cubital vein: a, A. magna sp. nov. forewing with a short vein vestige; b, Detail of an A. australis – Eastern Colour Morph forewing with a thickening of vein M at this position; c, Detail of an A. australis – Eastern Colour Morph forewing with a long vein vestige. Fig. d. Hindwing of A. magna sp. nov. Fig. e. Diagram of a forewing showing how the following measurements were made: (1) forewing length without tegula; (2) forewing width; (3) wing diagonal (Sakagami 1978); (4) 1 st abscissa of M length; (5) 1 st abscissa of Cu length. Figs a – d drawn to same scale: scale bar = 0.25 mm. Abbreviations: M—vein M; 1 st R—first recurrent vein; 2 nd Cu—second cubital cell; Ham—hamuli.

FIGURE 19. Austroplebeia legs. Figs a – b. Worker tibia III and basitarsus III of A. australis – Central Colour Morph: a, inner surface; b, outer surface. Fig. c. Cross-sectional view of A. australis worker tibia III, inner surface uppermost. Fig. d. Worker leg II of A. australis – Eastern Colour Morph, showing mesotibial spur near junction of tibia and basitarsus. Figs e – f. Male tibia III and basitarsus III of A. australis – Central Colour Morph: e, inner surface; f, outer surface. Fig. g. Queen tibia III and basitarsus III of A. australis – Central Colour Morph, outer surface. Fig. h. Variation in size of worker basitarsus III (shaded segment), all drawn to same scale: A. cincta QLD is shown on the left, then remaining species are shown in order of increasing basitarsus III width. Abbreviations: Kei—keirotrichia; Pos—posterior corbicular fringe; Ant—anterior corbicular fringe; Rasrastellum; Cor—corbicula; Pen—pencillum.

FIGURE 20. Austroplebeia male sterna and terga, shown in ventral view except for Fig. r. Figs a – f. S 4: a, A. cincta; b, A. essingtoni; c, A. australis; d – e, A. cassiae showing variation in position of gradulus; f, A. magna sp. nov. Figs g – j. S 5: g, A. cincta; h, A. essingtoni, arrow indicates protruding apical lobe; i, A. australis; j, A. cassiae. Figs k – o. S 6: k, A. cincta; l, A. essingtoni; m, A. australis; n, A. cassiae; o, A. magna sp. nov. Figs p – q. S 7: p, A. cincta; q, A. australis. Fig. r. T 7 of A. cassiae, dorsal view. Figs a – f, scale bar = 0.5 mm; figs g – r, scale bar = 0.33 mm. Abbreviations: Gra—gradulus; Med—median body.

FIGURE 21. Austroplebeia male genitalia. Figs a – c. Genitalia, dorsal view on left, ventral view on right: a, A. cincta; b, A. essingtoni; c, A. magna sp. nov. Figs d – e. Apical abdomen and protruding parts of genitalia, viewed laterally, as visible in many pinned male specimens: d, A. essingtoni, showing distinct curve in apical gonostylus; e. A. magna sp. nov., with gently curved apical gonostylus. Figs a – c drawn to same scale. Abbreviations: Pen—penis valve; Gon—gonostylus.

FIGURE 22. Austroplebeia worker faces: a, A. cincta NG; b, A. cincta QLD; c, A. essingtoni, showing a relatively heavily marked specimen within the species range; d, A. australis – Eastern Colour Morph; e, A. cassiae; f, A. magna sp. nov. Scale bars = 0.25 mm.

FIGURE 23. Austroplebeia male faces: a, A. cincta QLD; b, A. essingtoni; c, A. australis – Eastern Colour Morph; d, A. australis – Central Colour Morph; e, A. cassiae; f, A. magna sp. nov. Scale bars = 0.25 mm.

FIGURE 24. Austroplebeia queen faces: a, A. cincta QLD; b, A. essingtoni; c, A. australis – Eastern Colour Morph; d, A. australis – Central Colour Morph; e, A. cassiae; f, A. magna sp. nov. Scale bars = 0.25 mm.

FIGURE 25. Austroplebeia lateral thoraces: Figs a – b. Worker mesosomal side: a, A. cincta QLD; b, A. cassiae. Figs c – d. Male mesosomal side: c, A. cincta QLD; d, A. cassiae. Figs e – g. Detail of mesepisternum pilosity on male mesosomal side (photographs by K. Walker): e, A. australis; f, A. cassiae; g, A. magna sp. nov. Figs e – g shown at same magnification; all scale bars = 0.25 mm. Abbreviation: Met—metepisternum.

FIGURE 27. Pilosity on apical part of Austroplebeia queen abdomen, viewed laterally: a, A. cassiae; b, A. magna sp. nov. Scale bars = 0.25 mm.

FIGURE 28. Observed distribution of all Austroplebeia species in Australia and the island of New Guinea: a, A. cincta; b, A. essingtoni; c, A. australis; d, A. magna sp. nov. and A. cassiae. In all maps, square symbols represent nest samples and circles represent samples collected from flowers or traps. The type locality for each named species, including those now synonymised, is indicated with arrows. Where the type locality was a general area such as ' Central Australia' or ' Cape York', these areas are indicated with shading.

FIGURE 29. Austroplebeia nest entrances: a, A long A. cincta tunnel (120 mm) near Rossville, QLD; b, The tunnel of an A. essingtoni colony that has nested in a cliff face crevice 65 k S of Maningrida, NT (photograph by Aung Si); c, A translucent fresh tunnel of an A. australis colony near Tara, QLD (photograph by M. Halcroft); d, An A. australis colony without an entrance tunnel, near Tara, QLD (photograph by M. Halcroft); e, A thicker and older tunnel of an A. cassiae colony near Rossville, QLD; f, The tunnel of an A. magna colony 50 k S of Maningrida, NT (photograph by Aung Si); g, The longest tunnel recorded in this study: the 430 mm tunnel of an A. cincta colony that was being harassed by green ants near Rossville, QLD; h, Another long tunnel (ca. 120 – 130 mm) of an A. magna sp. nov. colony 50 k S of Maningrida, NT. Green ants can also be seen on this tunnel (photograph by Aung Si); i, The 160 mm tunnel of an A. cassiae colony near Rossville, QLD. The green ants seen on this tunnel have caught two forager bees.

FIGURE 30. Nest tunnels photographed at night (ca. 11 pm) near Rossville, QLD: a, An A. cassiae tunnel closed by a lacy mesh of fine resinous droplets (this tunnel closure style is also common in A. australis); b, An A. cincta tunnel which was left completely open; c, An A. cincta tunnel partially closed by an almost solid wall of cream waxy flakes.

FIGURE 32. Internal nest characteristics and brood combs of species other than A. cincta: a, Nest of an A. essingtoni colony at Broome, WA (photograph by T. Heard); b, Brood comb of an A. australis colony in a natural log near Ti Tree, NT; c, Brood comb of a boxed colony of A. cassiae near Rossville, QLD; d, Brood comb of an A. magna sp. nov. colony in a narrow tree cavity, 50 k S of Maningrida, NT (photograph by Aung Si). Abbreviations: Lar—larval brood cells; Coc—cocoons; Pol—pollen pots; Hon—honey pots; Bat—batumen layer.

FIGURE 33. Honey and pollen storage pots in Austroplebeia colonies: a, Translucent honey pots in an A. cincta colony near Rossville, QLD; b, Honey pots with thick, strongly pitted walls in an A. australis colony near Tara, QLD (photograph by M. Halcroft); c, Honey pots in an A. magna sp. nov. colony 50 k S of Maningrida, NT (photograph by Aung Si); d, Pollen pots in an A. cincta colony near Rossville, QLD; e, Pollen pots with thickened pitted walls in an A. australis colony near Tara, QLD (photograph by M. Halcroft); f, Pollen pots with strongly pitted walls in an A. australis colony near Kilcoy, QLD; g, Pollen pots made from red-brown cerumen in an A. cassiae colony at Conjuboy, QLD.

Tables

TABLE 4. Structural dimensions from workers of A. cassiae and A. magna sp. nov. Measurements in mm. A " – " indicates that measurements were not available.

VAR* 1	A. cassiae A. cassiae , Holotype 4.56	A. symei , Holotype 4.10	Other workers Mean SD 3.84 0.21	A. magna sp. nov. Holotype Paratype 4.01 4.01	Other workers Mean (N) SD 4.13 0.23
2 3 4	1.90 1.60 1.24	– 1.46 1.14	1.85 0.08 1.52 0.06 1.22 0.06	1.78 1.81 1.49 1.51 1.18 1.21	1.81 0.02 1.50 0.03 1.20 0.02
5 6 7	1.20 1.30 1.04	1.10 1.22 0.96	1.17 0.05 1.27 0.05 1.01 0.05	1.15 1.16 1.24 1.25 0.99 0.99	1.14 0.04 1.24 0.02 0.99 0.03
8 9 10	0.17 0.42 0.26	0.15 0.44 0.22	0.16 0.01 0.42 0.02 0.27 0.01	0.18 0.18 0.42 0.43 0.23 0.23	0.17 0.01 0.42 0.02 0.25 0.03
11 12 13	0.23 0.81 0.19	0.20 0.76 0.19	0.20 0.01 0.77 0.04 0.17 0.02	0.23 0.24 0.76 0.77 0.16 0.18	0.22 0.01 0.75 0.02 0.17 0.01
14 15 16	0.38 0.75 1.06	0.40 0.76 0.94	0.39 0.02 0.78 0.04 0.97 0.03	0.40 0.39 0.78 – 0.95 0.96	0.41 0.01 0.78 0.02 0.95 0.03
17 18 19	0.10 0.32 –	0.10 0.26 –	0.07 0.02 0.31 0.03 1.26 0.06	0.05 0.05 0.30 0.32 1.20 –	0.05 0.01 0.30 0.02 1.22 (11) 0.03
20 21 22	0.64 0.11 0.16	0.61 0.11 0.12	0.62 0.03 0.12 0.01 0.14 0.02	0.61 0.61 0.11 0.11 0.14 0.14	0.61 0.02 0.12 0.01 0.15 0.01
23 24 25	0.13 0.15 –	0.09 0.14 –	0.11 0.01 0.17 0.01 0.11 0.01	0.11 0.10 0.15 0.15 0.10 0.11	0.12 0.01 0.16 0.01 0.11 0.01
26 27	– 0.13	– 0.10	0.15 0.01 0.12 0.01	0.14 0.14 0.11 0.11	0.14 0.01 0.12 0.01