Trachusa / Subgenus Paraanthidium Friese, 1898

Anthidium (Paraanthidium) Friese, 1898: 101 . Type species: Apis interrupta Fabricius, 1781, by designation of Cockerell (1909: 269).

Protanthidium Cockerell & Cockerell, 1901: 49. Type species: Megachile steloides Bingham, 1896 (= Anthidium longicorne Friese, 1902).—Combination with Paraanthidium first suggested by Mavromoustakis (1937); Pasteels (1972) retained Trachusa (Protanthidium) as subgenus; synonymized with Trachusa (Paraanthidium) by Michener & Griswold (1994).

Protoanthidium Cameron, 1902: 125. Type species: Protoanthidium rufobalteatum Cameron, 1902, by designation of Sandhouse (1943).

Trachusa (Philotrachusa) Pasteels, 1969: 22 . Type species: Anthidium aquiphilum Strand, 1912 (= Trachusa aquifilum, Pasteels, 1969, misspelling).—Synonymized by Pasteels (1984) and confirmed by Michener & Griswold (1994).

Trachusa (Massanthidium) Pasteels, 1969: 24 . Type species: Trachusa flavorufula Pasteels, 1969 . syn. nov.

The generic keys provided by Pasteels (1969, 1984) are inconsistent in several aspects and are therefore not used here. Michener & Griswold (1994) in their classification of Old World Anthidiini confirmed the status of Paraanthidium Friese as a subgenus of Trachusa Panzer. They distinguished it from the subgenera Massanthidium Pasteels, Metatrachusa Pasteels and Congotrachusa Pasteels by the distinctly arcuate outward subantennal suture and the bifid, Y-shaped gonoforceps of the male, and from the subgenus Orthanthidium Mavromoustakis by the shape of the scutellum and the eyes and the position of the ocelli. Michener (2007) reinforced this classification (Table 1), but noted that Massanthidium resembles Paraanthidium, and that the relationships of Massanthidium remain in doubt until males are found.

It should also be noted that Michener & Griswold (1994) uses the bifid gonostylus as one of the key characters for Paraanthidium, while the shape of the male genitalia is actually not known for most of the species of this subgenus. The shape of the genostyli is for example known only for one of the six Asian species of the subgenus.

The subantennal suture in Trachusa namibiensis sp. n. is nearly straight in the female and slightly arcuate outwards in the male (Figs 1, 4). The gonoforceps of the male are deeply bifid and Y-shaped (Fig. 6d). Males of T. namibiensis sp. n. therefore run in the keys of Michener & Griswold (1994: 323) and Michener (2007: 533–534) to the subgenus Paraanthidium Friese. Taking the distinctly carinate omaulus with a carina extending onto the ventral surface of thorax close to middle coxa into account, females run, by contrast, to the subgenus Massanthidium Pasteels. It must be pointed out that it cannot be clearly determined what should be called in respect to the subantennal suture “nearly straight” and what is “distinctly arcuate outwards”; this character varies between the species so far attributed to the subgenus Paraanthidium (e.g., various photographs in Kasparek 2017a). It further needs to be considered that couplet 3 of Michener’s (2007) key is based, additionally to the shape of the subantennal suture, on the shape the gonoforceps, but males are unknown for the currently known three species of Massanthidium . In the key, the only available character feature to distinguish Paraanthidium from Massanthidium is therefore the form of the subantennal suture, but this feature is variable and does not allow to unambiguously attribute the new species to one of the two subgenera.

The length of the omaular carina has also been used to distinguish Massanthidium from Paraanthidium . It extends in Massanthidum ventrally onto the thoracic venter, extremely close to the middle coxa, while it never extends ventrally onto the venter in Paraanthidium (Michener 2007) . The omaulus is sharply angular in T. aquiphila and carinate in Trachusa namibiensis sp. n. In both species, it reaches the thoracic venter. The difference is not particularly pronounced and obvious enough to use this character for subgeneric classification.

Massanthidium further differs from Paraanthidium in the minute arolia (Michener 2007). While it is small in T. namibiensis sp. n., the size of the arolia is variable within Paraanthidium, and this character seems not to be an unambiguous feature for separating these two subgenera. Trachusa namibiensis sp. n. thus combines characters of both Paraanthidium and Massanthidium (Table 1), and the character features used to distinguish these two subgenera are overlapping. It is therefore not justified to maintain this subgeneric classification. Massanthidium Pasteels should be regarded as a junior synonym of Paraanthidium Friese, with T. eburneomaculata Pasteels, T. flavorufula Pasteels and T. massauahensis Pasteels assigned to this subgenus. All sub-Saharan species of Trachusa are herewith assigned either to Trachusa ( Paraanthidium Friese) (five species) or to T. ( Congotrachusa) (one species). A recent analysis of molecular data conducted by Litman et al. (2016) supports Congotrachusa as a distinct subgenus, not closely related to Paraanthidium .

The possibility to erect a new subgenus with T. namibiensis sp. n. as the only species and with characters shared both by Paraanthidium and Massanthidium was discarded because of the gradual transition of most character features.

Trachusa namibiensis sp. n. shares with T. aquiphila four-segmented maxillary palpi. All other species of Trachusa (Paraanthidium), including the species which had been assigned to Trachusa (Massanthidium) have three segments (Kasparek 2017a).

The characters of the subgenus Paraanthidium sensu Michener (2007) are herewith extended in the following way: female mandible with 4–7 teeth (before: 4–5 teeth); pronotal lobe carinate to lamellate; omaulus sharply angular to carinate or even (in the upper part) lamellate, with carina extending in some species onto the thoracic venter; arolia present but in some species minute; gonoforceps of male bifid. As further species of Trachusa attributed provisionally to the subgenus Paraanthidium await description from the Indomalayan and the West Palaearctic realms (Kasparek, unpubl.), a full review of the diagnostic characters of Paraanthidium is not given until this new information is fully available.

The distribution of the subgenus Trachusa (Paraanthidium) extends from Western Europe over the Himalaya to Malaysia and east China (species distribution maps in Kasparek 2017a). Trachusa (P.) aquiphila was so far the only species known from sub-Saharan Africa, and more than 6000 km distant from the nearest occurrence of the subgenus in North Africa. With the assignment of the East African species (which were hitherto regarded as being members of the subgenus Massanthidium) to the subgenus Paraanthidium, the geographic gap is closed. From an evolutionary perspective, it would have been difficult to explain how it comes that the southern African species are closer to the West Palaearctic species than to the East African species.