Andrena (Notandrena) oblonga Warncke, 1967 stat. nov.

Andrena langadensis ssp. oblonga Warncke, 1967 b: 236, ♀ ♂ [Libya, OÖLM, examined] (Figs 40, 41)

Material examined.

EGYPT • 1 ♀; Dekhela [Ad Dakhilah]; 20 Feb. 1917; Storey leg.; OÖLM (paratype of Andrena langadensis ssp. oblonga); LIBYA • 2 ♀; Cirenaica, Ras El Hilal [Ra’s Al Hilal]; 26 Apr. 1967; G. Dellacasa & P. Maifredi leg.; MSNG / TJWC • 1 ♀; Cyrenaica, Bengasi; 14 Mar. 1922; G. C. Krüger leg.; OÖLM (holotype of Andrena langadensis ssp. oblonga) • 1 ♀; Cyrenaica, R. U. Agrario; 1 Jun. 1925; G. C. Krüger leg.; OÖLM (paratype of Andrena langadensis ssp. oblonga) • 1 ♀; Cyrenaica, R. U. Agrario; 17 Mar. 1929; G. C. Krüger leg.; OÖLM (paratype of Andrena langadensis ssp. oblonga) • 1 ♀; Cyrenaica, R. U. Agrario, Bu Teija; G. C. Krüger leg.; MSNG • 1 ♂, 1 ♀; Cyrenaika, Mars el Brega; 8–11 Feb. 1942; Kirchberg leg.; OÖLM (paratype of Andrena langadensis ssp. oblonga) • 1 ♀; Libia occ., u. Ramla; 1–31 Mar. 1938; G. C. Krüger leg.; MSNG.

Notes.

This group of species was treated by Warncke (1965 a, 1965 b, 1967 b) as a pan-Mediterranean taxon under the name A. langadensis. However, Warncke (1967 b) made several incorrect treatments and assessments of Notandrena species, leading to confusion and requiring taxonomic action to stabilise names and concepts (Gusenleitner and Schwarz 2002; Wood and Monfared 2022; Wood 2023 b). One stabilised species is Andrena (Notandrena) griseobalteata Dours, 1872. Although the modern concept of this species is clear, there is one remaining issue which is that Warncke (1967 b: 288) treated A. meridionalis as a junior synonym (as A. erythrocnemis auctorum), and this remains the current placement of the taxon (Gusenleitner and Schwarz 2002: 323); this treatment requires dissection.

Andrena meridionalis is a replacement name for Andrena gravida Dours which is a junior primary homonym of A. gravida Imhoff, 1832 (and also A. gravida Eversmann, 1852). Dours described the species from “ Iles de l’archipel grec. Algérie ”. Whilst A. griseobalteata is common in Greece, it is not known from North Africa (Wood 2023 b); it is not clear if the original type series contained multiple taxa, or has been misinterpreted. A neotype was designated from southern France, and A. griseobalteata was considered to be a North Mediterranean species distributed from Spain to Turkey (Wood 2023 b). The identity of A. griseobalteata and A. meridionalis has always been unclear and confused because the collection of Dours was destroyed in the bombing of Amiens in the First World War (e. g. see Wood 2023 b); this is what necessitated the neotype designation for A. griseobalteata. However, what is less known is that some of Dours’ type material survived as specimens were traded with other hymenopterists such as Giovanni Gribodo (1846 – † 1924), and a number of syntypes for taxa described by Dours have been located in the MSNG collection (see Wood and Bossert 2025).

Examination of the Gribodo collection produced a female specimen labelled as “ Andrena gravida Typus Dours ” and nominally coming from Algeria (Fig. 33). Due to the short and broad head, pronotum with a humeral angle, and densely punctate terga, it can be placed into the Subgenus Notandrena, as is expected. However, the punctation of the scutum is dense, but not as dense as in A. griseobalteata (punctures in this species almost confluent), and instead conform to those seen in A. langadensis sensu lato. It is beneficial to reproduce the female description here:

“ ♀. Noire; antennes noires; mandibules ferrugineuses au bout. Tête arrondie, ponctuée; les poils cendré roux. Corselet assez fortement ponctué, les poils en dessus roux, plus pâles sur les côtés et en dessous. Abdomen elliptique très-élargi au niveau des 2 e et 3 e segments, nu, assez fortement ponctué. Bord inférueur des 2 e, 3 e, 4 e segments cilié de poils couchés, courtes, d’un blanc de neige. Les poils du 5 e segment et de l’anus sont roux doré. En dessous, le bord des segments est cilié de longs poils cendrés. Pattes noires. Tibias intermédiaires et postérieurs ferrugineux hérissées de poils blancs frisés, brosse cendré roux. Ailes assez limpides; point calleux testacé, noire au centre; côte, nervures brunes ”.

The description matches the specimen in key aspects, specifically the quite strongly punctured scutum, the hairs above red-brown (the probable meaning of roux in this case), those on the mesepisternum paler, metasoma with the discs of T 2–3 quite densely punctate, apical margins of T 2–4 with short hairbands of snow-white hair, and legs dark with the mid and hind tibiae ferruginous. Given these consistencies, and the known presence of Dours material in the MSNG collection (Wood and Bossert 2025), the specimen is hereby designated as the lectotype, and considered not to be conspecific with A. griseobalteata due to this sparser scutal punctation. As A. meridionalis is the oldest name in the langadensis - group of species, the challenge becomes 2) Where precisely was A. meridionalis described from? 2) Should all taxa in this group be lumped under the name A. meridionalis? and 3) Are multiple good species involved?

Examination of the relevant type material (MSNG; OÖLM) shows that the MSNG lectotype of A. meridionalis is morphologically most similar to A. langadensis due to the relatively long tergal hairbands that are poorly defined and laterally exceed the length of the marginal areas before continuing onto the base of the following tergum (Figs 33 F, 34 D), the brownish pubescence of the head and (to a lesser extent) mesosoma (Figs 33 B, 33 C, 34 B, 34 C), and moderately dense tergal punctation and brown terminal fringe (Figs 33 F, 34 D). The problem is that A. langadensis s. str. is only known from the southern Balkan Peninsula, and the taxon found in the West Mediterranean (ssp. albipila) has never been confidently recorded from Algeria, only Spain and Morocco (Gusenleitner and Schwarz 2002; Fig. 42).

There is therefore an incongruence between the reported collecting locality and the morphology presented by the lectotype of A. meridionalis. Given that Dours reported the species from “ Iles de l’archipel grec. Algérie ”, numerous Andrena species he described from “ Iles de l’archipel grec ” were based on material from Corfu (see Wood and Bossert 2025, section on Subgenus Ulandrena Warncke, 1968 below), and A. langadensis s. str. is known from Corfu based on material from multiple collectors (see Suppl. material 2), it is entirely plausible that this specimen is actually from Corfu (or at least Greece) and is simply mislabelled as coming from Algeria. Given this combined evidence, the position is taken that A. meridionalis was described from Greece as fixed by lectotype designation, and that it is senior to A. langadensis syn. nov.

The issue now becomes, what is the status of the taxa described as subspecies of A. langadensis? Examination of material from Lesbos has produced a female specimen which is referable to A. langadensis ssp. clanga. It shows the characters given by Warncke (1965 b), namely the bright orange antennae (Fig. 38 B; ventral surface), narrow tergal hairbands, and more strongly punctate terga (Fig. 38 D). When considering A. meridionalis and A. langadensis ssp. clanga, the density of the tergal punctation consistently differs, suggesting specific status for each taxon. When including A. langadensis ssp. oblonga (melanic, lacking tergal hairbands) and A. langadensis ssp. albipila (tergal hairbands consistently narrower and pubescence brighter, terga less densely punctate than A. langadensis ssp. clanga), as well as considering their non-overlapping geographic distributions (Fig. 42), the conclusion is that four distinct allopatric species are involved. Consequently, A. albipila stat. nov., A. clanga stat. nov., and A. oblonga stat. nov. are all elevated to species-status. An identification key for the females is presented below; males are currently insufficiently characterised, but can be identified through geographic provenance and co-occurrence with females.

Finally, it is necessary to discuss the species Andrena (Notandrena) stellaris Warncke, 1965. This species was described from Aksaray in Anatolian Turkey, and the published distribution is Central Anatolia (Aksaray, Niğde, Ürgüp, Yeşilhisar, Sultanhani / Kayseri; Warncke 1974 a; Gusenleitner and Schwarz 2002). This species was reported from Lesbos by Grace (2010) without any further information or precise specimen details, and then in Rasmont et al. (2017) with personal communication from Erwin Scheuchl. For Anatolian Andrena species, unless they are very widespread (and hence are present in Anatolia as only a small part of their range), they do not typically descend to the Mediterranean coast. The presence of A. stellaris, a species with a small distribution range on the high Central Anatolian steppe, on Mediterranean Lesbos is considered implausible given the complete lack of published records or examined specimens from Mediterranean Turkey. Given the presence of A. clanga on Lesbos, it is strongly suspected that these listings of A. stellaris are in fact misidentifications of A. clanga. The two species are quite similar, but in the female sex A. clanga can be recognised due to the denser scutal punctures which are separated by 0.5–1 puncture diameters (in A. stellaris with scutal punctures separated by 1–3 puncture diameters) and the denser tergal punctures, on the disc of T 2 with punctures separated by <0.5–1 puncture diameters (in A. stellaris with the punctures of the disc of T 2 separated by 1–2 puncture diameters). Since no actual evidence of A. stellaris on Lesbos has ever been published, it is here excluded from the Andrena fauna of the Balkans pending publication of precise information including diagnosis against A. clanga; Quod gratis asseritur, gratis negatur.

Distributions (Fig. 42).

Andrena meridionalis: Albania and Greece (including Crete); Andrena albipila: Spain and Morocco; Andrena clanga: Greece (East Aegean Islands), Turkey, Cyprus, Syria, Israel, and Jordan; Andrena oblonga: Libya and Egypt.