Taxonomy of Osmia (Pyrosmia)
Although Zanden (1991a) and Warncke (1992b) revised the Osmia species of the subgenus Pyrosmia, the taxonomy of the group is still in a rather poor state. The following new findings were incorporated into the Palaearctic osmiine bee website:
i) The synonymization of Osmia elbaba Warncke, 1992 with O. cyanoxantha Pérez, 1879 by Zanden (1996b) is erroneous. O. elbaba represents a species of its own differing from O. cyanoxantha in the female by the shorter ocellooccipital distance and the less densely shagreened terga 5-6 and in the male by the strongly differing pilosity and shape of the sterna and the uniformly cylindrical last antennal segment. The three specimens from northeastern Algeria, which Ferton (1914) reared from a snail shell and assumed to belong to O. leucopyga Ducke, 1899, are in fact members of O. elbaba as shown by the characters listed by Ferton, i.e. the purple scutellum (metallic green in O. leucopyga), the whitish pilosity of tergum 1 (yellowish in North African O. leucopyga) and the weakly shagreened terga 5-6 (strongly shagreened in O. leucopyga). Thus, to the present knowledge O. elbaba occurs in Morocco, Algeria and Tunisia and it nests in snail shells.
ii) The synonymization of Osmia leucopyga Ducke, 1899 with O. lobata Friese, 1899 by Warncke (1992b) is erroneous. O. lobata represents a species of its own, which is well characterized by the shagreened basal area of the propodeum (polished in O. leucopyga), and the apically emarginated median process of male tergum 7, which is unique among Osmia species of the subgenus Pyrosmia. The female from Israel described by Zanden (1991b) and keyed out in his identification key does most probably not belong to O. lobata since the author mentions that the basal area of the propodeum is polished and that the body length is only 7 mm, which appears to be too small for O. lobata, whose male holotype measures 9 mm in length. Thus, to the present knowledge O. lobata is known so far only from Algeria and no reliable records exist for the other Maghreb countries, southwestern Europe or the Levant. The nesting biology as well as the female of O. lobata are still unknown.
iii) Osmia leucopyga Ducke, 1899 was described based on a single female from Algeria. Zanden (1991b) incorporated the male of O. leucopyga in his identification key. However, the characters mentioned by him do not refer to O. leucopyga but are typical for O. elbaba Warncke, 1992, as revealed by series of both sexes of O. elbaba from the Maghreb. In 2021, T. Wood collected females of O. leucopyga at high elevations in the Sierra Nevada in southern Spain together with males, which closely correspond to the males of O. cyanoxantha Pérez, 1879. Upon closer investigation, these males were found to slightly differ from O. cyanoxantha from northern Italy and the Balkans in few characters and are thus most probably the unknown males of O. leucopyga. A pronounced morphological similarity between the males of O. cyanoxantha and O. leucopyga would not be surprising since the females of the two species are also very similar. Reliable records of O. leucopyga Ducke, 1899 exist for the Iberian Peninsula (Portugal, Spain), Morocco and Algeria, the species possibly also occurs in Tunisia. Based on two single observations, the species nests in preexisting cavities, e.g. in abandoned cells of Megachile (Chalicodoma) or in burrows in the soil (Zanden, 1991b).
iv) Osmia submicans Morawitz, 1870 is a widespread species occurring also on the Canary Islands. Here, three subspecies are recognized, i.e. O. s. canaria Mavromoustakis, 1957 (Gran Canaria, Tenerife, El Hierro), O. s. columbina Zanden, 1996 (La Palma) and O. s. lanzarotae Warncke, 1992 (Fuerteventura, Lanzarote). These three subspecies distinctly differ morphologically among each other on the one hand and with O. submicans from continental Europe on the other hand and thus probably warrant species status. However, as there are also morphological differences between members of the same subspecies inhabiting different islands (e.g. O. s. canaria on Gran Canaria and on Tenerife), these three taxa are provisionally kept as subspecies pending a new in-depth taxonomic study.