BULGARIA • 1♂; Blagoevgrad, Gorno Spanchevo, 1.5 km E; 420 m a.s.l.; 22 May 2024; T.J. Wood leg.; TJWC • 1♂; Stara Kresna; 20 Jun. 1987; Karas leg.; OÖLM (ex. collection B. Tkalců); GREECE • 1♂, 1♀; Western Macedonia, Kottas, 1.6 km E of Vatochori; 850 m a.s.l.; 13 Jun. 2024; T.J. Wood leg.; RMNH • 3♂, 1♀; Western Macedonia, Kottas, 1.6 km E of Vatochori; 850 m a.s.l.; 13 Jun. 2024; C. Praz leg.; PRUN; IRAN • 1♀; Zentralprovinz, 10 km nördlich Karadj [Karaj]; 1100 m a.s.l.; 23 May 1977; Holzschuh & Resel leg.; OÖLM; TURKEY • 1♂; 10 km W of Urgup [Ürgüp]; 15 Jun. 1998; Ma. Halada leg.; OÖLM • 2♂; 50 km S Kars, Pasli; 1 Jul. 1997; Ma. Halada leg.; OÖLM • 2♂; Bolu, S Karayokuş Gec Hocadag; 40.3503°N, 31.3003°E; 760 m a.s.l.; 19 Jun. 1986; E. Hüttinger leg.; OÖLM • 1♂, 1♀; Elazig [Elâzığ]: Hazar-See; 1 Jun. 1978; M. Schwarz leg.; OÖLM • 1♂; Konya, 10 km S Karaman; 1100 m a.s.l.; 19 Jun. 1985; M. Schwarz leg.; OÖLM • 1♂; Konya: Karaman; 11 Jun. 1978; M. Schwarz leg.; OÖLM • 2♀; Madenşehir [Madenşehri]/Konya; 6 Jun. 1971; K. Warncke leg.; OÖLM • 2♀; Madenşehir [Madenşehri]/Konya; 22–23 Jun. 1984; K. Warncke leg.; OÖLM • 1♀; Malatya, 3 km SE of Kubbe Gecidi, st. 2641; 1700 m a.s.l.; 3 Jul. 2000; H. v. Oorschot, H. v. d. Brink, P. Oosterbroek leg.; RMNH; ZMA.INS.5144676 • 2♂; Maraş [Kahramanmaraş], Afsin [Afşin]; 8 Jun. 1985; OÖLM • 1♂; Meram, Konya; 14 Jun. 1968; K. Kudas leg.; OÖLM • 1♂; Nevsehir [Nevşehir], Urgup [Ürgüp]; 16 Jun. 1977; K. Warncke leg.; OÖLM • 1♂; Nevşehir, 10 km NE of Nevşehir; 1200 m a.s.l.; 5–7 Jul. 1982; H. v. Oorschot & H. v. d. Brink leg.; RMNH; ZMA.INS.5144674 • 1♂; Nordhand des östl. Hama dag, südl. Igdir (Kars); 1300–1600 m a.s.l.; 27 Jun. 1971; Reinig leg.; OÖLM; • 1♂; Prov. Ankara, Baglum [Bağlum Güzelyurt]; 1400 m a.s.l.; 13 Jul. 1961; J. Leinfest leg.; RMNH; ZMA.INS.5144675.

There is currently no modern identification key for Anthophora that can be used in south-eastern Europe. The work of Friese (1897) is badly out of date due to the many taxonomic changes which have been made and must still be made in the Anthophorini. Anthophora onosmarum is included in this work (as Podalirius s. str.), but does not key out well as the mandibles can be partially yellow-marked, whereas Friese considered them to be entirely dark. Within the European fauna, A. onosmarum is best diagnosed due to the combination of its elongate head, in direct frontal view with the lower margin of the clypeus ventrally projecting below the lower margin of the compound eyes (Figs 3, 4), with the inner margins of the compound eyes slightly diverging ventrally (most strongly pronounced in the male sex, Fig. 8), head therefore relatively elongate, only 1.15–1.20 times wider than long (in comparison other Anthophora (Paramegilla) species with the head much wider than long, typically between 1.50–1.85 times wider than long), malar space slightly expanded but only subequal to length of antennal segment 4, clypeus bulging in profile view (extending as far in front of the compound eye as the diameter of the compound eye itself; Figs 1, 2), length of tongue very long, almost extending beyond tip of metasoma (Fig. 2), in female sex with metasoma covered with either adpressed pale yellow-orange pubescence (Fig. 5; intensity of colour depending on the age of the specimen) or adpressed black pubescence (Fig. 6), but if with black pubescence then without lateral white hairbands, male with extensive yellow markings on face, these covering the labrum, clypeus, lower paraocular areas to a level reaching the antennal insertions, ventral surface of antennal scape, and sometimes with a small dot on the mandibles (Fig. 8), mid legs without hair patches of fringes, hind basitarsi essentially unmodified (without the presence of teeth) only slightly thickened apically, and genital capsule with gonocoxae produced into long apical projections which come to a curved apical point, with long, thin, and hyaline finger-like gonostyli (Fig. 10). A revised key to European Anthophora will hopefully be produced in the near future (TJW, in prep.).

Figures 1–6. 

Anthophora onosmarum Morawitz, 1876 females, light (Turkey, Madenşehri; 23 Jun. 1984; K. Warncke leg.; OÖLM) and dark (Turkey, Madenşehri; 22 Jun. 1984; K. Warncke leg.; OÖLM) forms). 1. Light form, profile; 2. Dark form, profile; 3. Light form, face, frontal view; 4. Dark form, face, frontal view; 5. Light form, metasoma, dorsal view; 6. Dark form, metasoma, dorsal view.

Figures 7–10. 

Anthophora onosmarum Morawitz, 1876 male (Bulgaria, Stara Kresna; 20 Jun. 1987; Karas leg.; OÖLM). 7. Profile; 8. Face, frontal view; 9. Metasoma, dorsal view; 10. Genital capsule, dorsal view.

A total of nine female specimens were examined, these showing two distinct colour morphs – four specimens showed black pubescence (Greece, central Turkey; Figs 1, 3, 5), and five showed yellow pubescence (central and eastern Turkey, Iran; Figs 2, 4, 6). There was a weak geographic gradient with darker individuals in the west and lighter individuals in the east, but on two occasions (Madenşehri, 6^th June 1971 and 22–23^rd June 1984) Klaus Warncke collected one black and one yellow female at the same place. The specimens illustrated in Figs 1–6 are from Madenşehri from the 22–23^rd June 1984. Very little is known about colour variation in Anthophorine bees, and we make no decisive comments here, other than to say that for at least A. onosmarum this seems to represent simple variation, as it is not accompanied with structural differences. Morawitz (1876) described the female as being of the yellow form: “Das Weibchen ist schwarz, das Gesicht, die Schläfen und die Brust weiss, das Hinterhaupt, die obere Fläche des Thorax und das erste Abdominalsegment sehr dicht fuchsroth behaart” [The female is black, the face, the gena and the mesosoma has white hairs, the dorsal part of the thorax and the first metasomal segment dense, bright orange-red hairs]. In some Andrena species (Andrenidae), specimens from the Balkans are darker, for example Andrena (Hoplandrena) clusia ssp. prilepensis Warncke, 1973 (described from North Macedonia) which is almost completely melanic relative to the nominate subspecies which was described from Azerbaijan. This can also be seen in Eucera (Cubitalia) breviceps (Friese, 1911) (Apidae) (Aubert et al. 2024a). In any case, male A. onosmarum appear to be much more consistent in terms of their colouration, with no melanic individuals observed so far. This would appear to fit the overall pattern of reduced colour variation in male anthophorines relative to females (e.g. Brooks 1983).

We obtained a full-length (658 base pairs) DNA barcode from a male specimen (BOLD accession number HYMAA898-24; www.boldsystems.org); when submitted to the identification tool in the Bold Systems, the closest matches (87.4–89.5% similarity) were to Anthophora (Paramegilla) balneorum Lepeletier, 1841, A. (Paramegilla) nigrovittata Dours, 1872, and A. (Dasymegilla) quadrimaculata Panzer, 1798. The tree-based identification suggested that A. onosmarum was sister to a clade containing A. balneorum and A. nigrovittata. These results confirm that the obtained barcode is distinct from all other species represented on BOLD, and tentatively suggest phylogenetic affinities with A. balneorum and A. nigrovittata.

A single male of A. onosmarum was captured in south-western Bulgaria (Blagoevgrad, Gorno Spanchevo) visiting Onosma sp. (given the challenging nature of Onosma identification, it was not possible to confidently determine to species level; Teppner 1991) growing on piles of discarded building materials mixed with soil on 22^nd May 2024. The surrounding habitat was composed of a small area of dry exposed slopes adjacent to a mountain road surrounded by deciduous woodland; Onosma was observed only on the refuse piles. Despite remaining at the site for several hours, no further individuals of A. onosmarum were seen.

On 13 June 2024 several individuals (approximately 10 males and five females) of A. onosmarum were observed in northern Greece (1.6 km E of Vatochori) on dry slopes with abundant Onosma sp. (Figs 11, 12). The individual Onosma plants formed spaced clumps, and male A. onosmarum were patrolling between clumps (Fig. 13), stopping only occasionally for nectar (Figs 14–16). Males were extremely active and difficult to photograph. Some males stopped to rest by finding thin upstanding stems of dead or dry vegetation and gripping them with their jaws. Males were not observed landing on the ground; presumably this would require more energy to return to the air, or it would increase the time taken to become airborne and therefore reduce their ability to quickly react to a passing female.

Figures 11–12. 

11. Anthophora onosmarum Morawitz, 1876 habitat in northern Greece on 13^th June 2016; 12. Onosma sp. (Boraginaceae).

Figures 13–16. 

Anthophora onosmarum Morawitz, 1876 male behaviour visiting Onosma sp. (Boraginaceae) flowers.

Females showed a similar behaviour, rapidly moving between Onosma clumps, usually with their tongue extended (Fig. 17). Upon encountering a flower, the females hang upside down from the corolla, using their long tongue to access the nectaries (Fig. 18). Females were also observed tightly gripping the corolla (whilst also keeping their tongue inside, Fig. 19) and were heard to emit a short high pitched buzz. This buzz released pollen which could be observed falling onto the underside of the metasoma as the bee released its grip on the corolla (Fig. 20). This pollen is presumably then groomed into the hind tibial scopa during flight.

Figures 17–20. 

Anthophora onosmarum Morawitz, 1876 female behaviour visiting Onosma sp. (Boraginaceae) flowers.

Based on the long tongue (necessary to reach the deep nectaries of the Onosma flowers), the flower-buzzing behaviour, and the generally strong behavioural association with this plant genus (also mentioned in Morawitz 1876), we consider A. onosmarum likely to be narrowly oligolectic on this plant genus. It is not impossible that other genera of Boraginaceae are used; evidence from other bee groups is mixed. Osmia (Osmia) apicata Smith, 1853 (Megachilidae) is broadly oligolectic on Boraginaceae with a preference for Onosma (Haider et al. 2013), but this species does not buzz the flowers and scrapes out the pollen with its forelegs (Gogala and Surina 2011). Hoplitis (Hoplitis) onosmaevae Aubert, 2024 (Megachilidae) and E. breviceps however do buzz flowers of Onosma (Aubert et al. 2024a, 2024b), and seem to be narrow oligoleges of Onosma. In the new world, Boraginaceae-associated Perdita species (Andrenidae) also show variation within their use of this botanical family, from broad to very narrow specialisation (Portman et al. 2016). Ultimately, pollen analysis on collected specimens or more extensive behavioural observations would been needed to decisively conclude on this question.

Based on the examined specimens, collected males outnumber collected females by slightly more than 2:1 (21 examined males, nine examined females), and the flight period ranges from 22^nd May to 13^th July, with a median date of 16^th June. This would fit with our observation on 13^th June 2024 in northern Greece, where we observed a mixture of patrolling males showing signs of wear, combined with pollen collecting females, implying that nest provisioning was well underway.

Whilst A. onosmarum is here newly reported for Bulgaria, Greece, and Europe as a whole, inspection of specimens from the Borek Tkalců collection (OÖLM) revealed that an undetermined male was actually collected in Bulgaria as early as 1987. Searches in the IBER collection in Sofia in 2023 and 2024 (TJW) did not uncover any further specimens, but searches were not exhaustive.

Greece*, Bulgaria*, Turkey*, Georgia, Iran (Rasmont 2014; Ascher and Pickering 2024) (Fig. 21).

Figure 21. 

Distribution map for Anthophora onosmarum Morawitz, 1876 (green points) based on examined specimens and verified literature records.

The locus typicus is “Azchur” which is probably the village of Atskuri in southern Georgia (41.73°N, 43.16°E) and which is alternatively spelt “Azkur” or “Ahiska”. This village is located at an altitude of 900 metres above sea level. Iran is listed based on a GBIF record from the Donald Baker collection from the Snow Entomological Collection at the University of Kansas. Its details are “Iran: centr. Alborz, Kandavan Pass, nr. Pol-e-Zanguleh, 2200 m. 8 VII 1967. STA 8. Baker Exp.”. This record comes from just 40 kilometres to the north-east of the specimen we report here from north of Karaj. Rasmont (2014) also lists southern Russia (North Caucasus) which is plausible (although the species is not listed from Russia by either Levchenko et al. 2017 or Proshchalykin et al. 2023), and Libya, this being not at all plausible clearly the result of an encoding or other type of data error. We therefore take a conservative approach in our distributional listing based on confirmed specimens. Given the collecting locality in north-western Greece, A. onosmarum is almost certainly present also in Albania and North Macedonia. Where altitude data are available, A. onosmarum is found between 420–2200 metres above sea level, with 7/10 records with altitude information reported above 1000 metres above sea level.