Andrena (Lepidandrena) Hedicke 1933

Abstract

SubgenusLepidandrenaHedicke, 1933 Genetics. Many sequences were available covering all species of the southern Balkans. Most species were well-resolved with high bootstrap support of> 95 for species such as Andrena rufizona Imhoff, 1834, Andrena dorsalis Brullé, 1833, Andrena florivaga Eversmann, 1852, Andrena curvungula Thomson, 1870, and Andrena paucisquama Noskiewicz, 1924. However, other species require discussion (Fig. 16). Andrena pandellei Pérez, 1895 formed a monophyletic clade with bootstrap support of 88, but showed moderate average intraspecific genetic variation of 4.80 % (range 0.15–8.05 %). Three sequences from Bulgaria clustered together with bootstrap support of 100, with three sequences from Portugal, Germany, and Slovenia grouped together with bootstrap support of 90. Andrena pandellei was originally described from Algeria (lectotype: Theniet el Had) from a cedar forest in the Tell Atlas mountain range. The species is very rare in North Africa, and Warncke (1967 b: 215) described Andrena pandellei ssp. europaea Warncke, 1967 for the much more abundant, widespread, and commonly encountered European population. No authors have recognised population structuring within Europe, and hence a broad concept of A. pandellei is maintained for the time being. Warncke adopted a broad species concept for Andrena gamskrucki Warncke, 1965 (locus typicus: Greece, Larisa), later describing Andrena gamskrucki ssp. eburnea Warncke, 1975 (locus typicus: Turkey, Ayvalik) and Andrena gamskrucki ssp. impasta Warncke, 1975 (locus typicus: Turkey, Birecik). Subsequent authors have treated these as distinct taxa (Dubitzky et al. 2010), and DNA barcodes support this distinction, at least for the two sampled taxa. Sequences of A. gamskrucki from mainland Greece were separated from sequences of A. impasta from Lebanon by an average genetic distance of 7.52 % (range 7.29–7.75 %). Unfortunately no barcodes are available for A. eburnea, but based on the morphological differences (see identification key), the conclusions of Dubitzky et al. (2010) can be followed. Andrena gamskrucki is treated here in a narrow sense as a species endemic to the southern Balkans. See species entry in the Updated Faunal List for discussion of A. eburnea on the East Aegean Islands. One additional morphologically-unrelated Lepidandrena sequence from southern Greece, NATBE 001-25, was placed as sister to A. gamskrucki and A. impasta, and was strongly isolated from all other Lepidandrena sequences by an average genetic distance of 12.18 % (range 10.97–13.23 %). It shows a genital and sternal morphology with aspects of A. pandellei and A. paucisquama, and is newly described below as Andrena haladai sp. nov. The greatest problems come in the mocsaryi - group of species. These seem to be taxa strongly associated with Ornithogalum (Asparagaceae) (e. g. Westrich 2010). It contains Andrena mocsaryi Schmiedeknecht, 1884 (types lost, but described from Hungary), Andrena baetica Wood, 2020 (locus typicus: Portugal, Portalegre), Andrena caprimulga Warncke, 1975 (locus typicus: Turkey, Horasan), and probably Andrena elisaria Gusenleitner, 1998 (locus typicus: Turkey, Hakkâri: Uludere) and Andrena statusa Gusenleitner, 1998 (locus typicus: Turkey, Hakkâri: Uludere). This group can be recognised due to the combination of the outer surface of the female galea which is strongly roughened and dull and the male sternum 8 which is flat and not divided into horizontal and vertical parts (as is the case in the florivaga - group sensu Schmid-Egger and Scheuchl 1997). Five clades were observed, all with bootstrap support> 93 (Fig. 16). One contained A. baetica (Spain and Portugal), one contained A. caprimulga from Turkey plus A. statusa from Lebanon, one contained A. mocsaryi from Austria and Greece (Peloponnese) (mocsaryi # 1), one contained A. mocsaryi from Italy (mocsaryi # 2; see also Mendicino et al. 2026), and one contained A. mocsaryi from the Peloponnese only (mocsaryi # 3). The three clades of “ A. mocsaryi ” were all strongly separated. mocsaryi # 1 was separated from mocsaryi # 2 by an average genetic distance of 5.86 % (range 4.88–6.33 %) and from mocsaryi # 3 by an average genetic distance of 9.87 % (range 8.61–10.70 %). mocsaryi # 2 was separated from mocsaryi # 3 by an average genetic distance of 8.83 % (range 8.51–9.17 %). The situation is particularly confusing because the two mocsaryi lineages that can be found in sympatry in the Peloponnese (nos. # 1 and # 3) show the greatest observed separation of 10.70 % (sequences NATBE 030-25 and BGENL 1535-24). This genetic distance is substantial, and may indicate the presence of a cryptic species. Unfortunately, all of the barcoded Greek specimens of “ A. mocsaryi ” are female (all selected males failed), and in general it has only been possible to examine two male specimens from Greece compared to 36 female specimens. It seems that most collectors travel to Greece slightly too late to encounter males. If a cryptic species truly is present, then it is mocsaryi # 1 that represents the true A. mocsaryi since this clade contains a specimen from Austria, close to the terra typica of Hungary. More male material is required to understand what is happening in these different mitochondrial lineages of A. mocsaryi. More broadly, the three mocsaryi lineages do not form a monophyletic clade. Andrena baetica was nested within, but was strongly separated from mocsaryi # 1 (5.66 %, range 4.88–6.43 %), mocsaryi # 2 (5.75 %, range 4.95–6.55 %), and mocsaryi # 3 (9.71 %, range 9.58–9.88 %). Morphologically, in principle there is no strong barrier to including A. baetica within a broad concept of A. mocsaryi as a geographically isolated population (differences are mostly related to female scutal hair length and structure and male clypeus colour), but the topology of the tree does not allow this given the strong genetic separation of A. caprimulga which is also nested within the broad A. mocsaryi sensu lato clade, in addition to the differences in the structure of the male S 8. The inclusion of sequences of A. caprimulga is important to exclude the possibility that one of the “ lineages ” of A. mocsaryi in the Peloponnese is actually A. caprimulga. However, A. caprimulga sequences were strongly separated from the lineages present in the Peloponnese, namely mocsaryi # 1 (7.47 %, range 6.43–8.16 %) and mocsaryi # 3 (9.74 %, range 8.87–10.35 %). Morphologically, A. caprimulga is differentiated from A. mocsaryi by a small but distinct difference; it was not well-expressed by Warncke in his writing, but he did illustrate the key difference which is that the 8 th sternum is produced into three distinct points in A. caprimulga, and is rounded with a tiny apical point in A. mocsaryi (illustrated Warncke 1975: 95; see also Schmid-Egger and Scheuchl 1997: 156 for the character state of A. mocsaryi). None of the examined males from the southern Balkans have shown this distinctive A. caprimulga structure of the 8 th sternum. This difference has been slightly obscured because in the OÖLM collection, although Blank and Kraus (1994) indicated two female (holotype and paratype) and one male (paratype) specimens in the collection of Warncke during its accession, only the two female specimens are currently present. Moreover, the female paratype is stylopised and displays intersex features (e. g. a partially yellow-marked clypeus; Fig. 17). The sequence of A. statusa from Lebanon is separated from the Turkish sequences by an average of 1.01 % (range 0.91–1.07 %), but the topology indicates that it is nested between them. Examination of a paratype male from Israel (directly; photographs also available on zobodat. at) indicate that the 8 th sternum shows the typical A. caprimulga shape, with three distinct points. This is consistent between Levantine specimens, newly examined and sequenced specimens from the Taurus mountains (RMNH), and Warncke’s original type series from Horasan (as illustrated by Warncke 1975). Consequently, A. statusa is formally synonymised under A. caprimulga below. Finally, Pisanty et al. (2018) already expressed doubts about the nominal differences described between A. elisaria and A. statusa based on co-occurrence of the taxa in Israel, conclusing that they may simply be dark and red-marked specimens of A. statusa (= A. caprimulga). Some male Lepidandrena specimens from south-eastern Turkey (OÖLM / TJWC) with extensive red maculations show a S 8 that is laterally rounded – these may represent A. elisaria, but the description combined with the holotype designated in the female sex is currently limiting. No taxonomic action for A. elisaria is taken at the present time, awaiting the generation of additional genetic sequences from Turkey.

Authors 1

1. Thomas James Wood corresponding Aachen Naturalis Biodiversity Center

   Naturalis Biodiversity Center

   Affiliation as printed

   Naturalis Biodiversity Center, Darwinweg 2, 2333 CR, Leiden, Netherlands