Euandrena Hedicke, 1933

Gautam, R. K., Uniyal, V. P. & Wood, Thomas J., 2024, A critical revision of the Andrena Fabricius, 1775 of India, with the description of two new species (Hymenoptera: Andrenidae) from Uttarakhand, European Journal of Taxonomy 948, pp. 1-59 : 5-10

Problems with the subgenus Euandrena Hedicke, 1933

The subgenus Euandrena Hedicke, 1933 is taxonomically complex, as many cryptic taxa are present, and morphological recognition is highly challenging, even in comparatively well-studied regions such as Europe (e.g., Praz et al. 2019). The situation around the Himalayas (from the western to eastern extent) is currently unsatisfactory due to the large number of available names of uncertain status. Currently, seven red-marked species of Euandrena are known, specifically 1) Andrena familiaris Smith, 1878 (described from East Turkestan in what is now western China), 2) A. communis Smith, 1879 (described from Mussoorie in northern India), 3) A. mephistophelica Cameron, 1897 (described from Mussoorie in northern India), 4) A. murreensis Cockerell, 1923 (described from Murree in northern Pakistan), 5) A. euphorbiacea Scheuchl, 2005 (described from Yunnan in southern China), 6) A. humlaensis Scheuchl, 2005 (described from Nepal), and 7) A. kathmanduensis Tadauchi & Matsumura, 2007 (described from Nepal). Modern descriptions ( Grünwaldt et al. 2005; Tadauchi & Matsumura 2007) unfortunately do not provide diagnoses against these older Himalayan taxa, and it is not clear if these authors were aware of these available names, their likely identities, or indeed how any of these names relate to each other.

It is necessary to deal with the four oldest available names, and definitively locate the location of type material as the listings in Gusenleitner & Schwarz (2002) are unfortunately not always correct due to historical uncertainty, and the huge size of the genus which has not always allowed for each reported type locality to be specifically investigated. The first name, A. familiaris , was described from East Turkestan [neighbourhood of Yárkand = Yarkant County, Xinjiang, western China] ( Smith 1878: 2). The species was described in the male sex only, and is illustrated ( Smith 1878: fig. 3). Although Gusenleitner & Schwarz (2002) indicate that the type of A. familiaris is present at the NHMUK, no such specimen matching this collection information is present and Donald Baker (unpublished thesis: 272–273) concludes that type material is likely to be in the ZSI. Examination of the ZSI website (https://zsicollections.in/search?genus= Andrena ) shows that the male type of A. familiaris is present there (reference specimen code ZSI0000008548).

However, in the NHMUK type catalogue, there is an entry for A. familiaris (B.M. type 17a 1306). Inspection of this specimen ( Fig. 1 ) shows that it was collected from Masuri [= Mussoorie] in northern India, and it is a female. A handwritten label (not in the handwriting of Smith) indicates “ Andrena familiaris (Type) Sm.”. An additional handwritten note by Donald Baker correctly indicates that this is a false type, since it is a female, and it comes from Masuri rather than Yárkand. However, the specimen matches perfectly the collection information for A. communis which was described from Masuri, collected at an elevation of 7000 feet ( Smith 1879: 51). Andrena communis was described only in the female sex, and as the NHMUK specimen matches Smith’s description, the specimen is here considered to be the holotype of A. communis . No other specimens labelled as “ A. communis ” are known in the NHMUK catalogue (type collection or main collection). Since Smith (1879) is titled “Descriptions of new Hymenoptera species in the collection of the British Museum”, and all other species of Andrena described in this publication are present in the NHMUK, it is considered unlikely that the type might actually be in another institution.

Gusenleitner & Schwarz (2002) list the NHMUK as the type depository for A. communis , but these authors did not inspect this material. They doubtful associate A. communis with the subgenus Oreomelissa Hirashima & Tadauchi, 1975 , probably on the basis of Bingham (1897) who confusingly uses a character of a punctate propodeum in his identification key, later leading to A. communis . Based on this holotype specimen, A. communis is clearly a member of the subgenus Euandrena , lacks distinct punctures on the propodeum, and is relatively small at 9 mm in length.

In Bingham’s (1897) identification key, A. communis is placed close to A. mephistophelica , and is separated by colouration, with A. mephistophelica having the tergal margins and bases lightened, rather than the red markings also covering the disc of T2, as in A. communis ( Fig. 1D ). Andrena mephistophelica was also described from Mussouri [= Mussoorie], and Cameron (1897) indicates its body length as 11–12 mm. Gusenleitner & Schwarz (2002) again indicate that type material is at the NHMUK, but it is actually in the Rothney Collection (OUMNH). However, the exact location of Andrena type material from the Rothney collection is unclear, and it is not currently available for study (J. Hogan, in litt.). Based on the description and the redescription by Bingham (1897: 443), A. mephistophelica is very close to A. murreensis which was described from Murree in northern Pakistan with a female body length of 9–11 mm ( Cockerell 1923). Examination of the NHMUK collection produced a specimen of A. murreensis labelled by Cockerell with his distinctive handwriting and a yellow “co-type” label, from Murree at an elevation of 7500 feet. The specimen was collected by Dutt and has an additional label of “156”. This corresponds to Cockerell (1923: 265), who listed eight numbered specimens from the Fletcher collection. The specimen is hereby designated as a lectotype ( Fig. 2 ; see below for full specimen details). An additional specimen is present in the USNM collection with the same collecting information (https://collections.nmnh.si.edu/search/ento/) with the number “50”. It is part of the syntypic series, but is not designated as the lectotype.

Genetic analysis of five samples of red-marked Euandrena from northern Pakistan provisionally identified as A. communis (one female) and A. murreensis (three females, one male) plus two sequences of A. murreensis available from BOLD (PCYU, two female specimens identified by TJW) produced three lineages with strong posterior support ( Fig. 3 ). Three “ A. murreensis ” sequences form a clade with bootstrap support of 98 separated from A. communis + the remaining three “ A. murreensis ” sequences, these latter three sequences forming a clade with bootstrap support of 99. Morphologically, examination of the two “ A. murreensis ” clades plus the lectotype of A. murreensis shows subtle but consistent differences. Specifically, 1) true A. murreensis has the clypeus distinctly punctate, but punctures are sparse and scattered, separated by 1–3 puncture diameters, particularly medially where large impunctate areas are present ( Fig. 4C ), whereas the comparison clade shows a more evenly punctate clypeus, with punctures separated by 1–2 puncture diameters over the entire disc ( Fig. 4D ); 2) true A. murreensis has the scutum predominantly shiny, shagreened laterally and anteriorly but shining over 60% of the scutum ( Fig. 4E ), whereas the comparison clade has the scutum strongly and consistently shagreened, with around 20% of the disc weakly shining medially ( Fig. 4F ); and 3) true A. murreensis has the propodeal triangle more strongly depressed and the dorsolateral parts of the propodeum more strongly rugose, thus reciprocally more strongly contrasting, whereas the comparison clade has both the propodeal triangle less strongly depressed and the dorsolateral parts of the propodeum less strongly rugose, so they only weakly contrast each other.

It remains unclear if the other lineage can be called A. mephistophelica , or whether A. mephistophelica is the senior synonym of A. murreensis , and the remaining lineage refers to one of the more recently described species of red-marked Euandrena ( Grünwaldt et al. 2005; Tadauchi & Matsumura 2007), or perhaps is even undescribed. Based on the description, the next best candidate for the non- murreensis clade would be A. humlaensis given its comparatively densely punctate clypeus ( Grünwaldt et al. 2005: 363), whereas A. euphorbiacea and A. kathmanduensis appear to be more similar to A. murreensis based on the sparse clypeal punctures. For the present time, this second lineage is referred to as “aff. A. murreensis ”. The two lineages can be found in direct sympatry, for example at Shogran (both historical and modern specimens, see below).

The true A. murreensis clade was separated from the aff. A. murreensis clade by 7.20% (range 7.14– 7.31%), and A. communis was separated from the true A. murreensis clade by 8.23% (range 8.10– 8.37%) and the aff. A. murreensis clade by 6.85%. It is important to note that these distances in and of themselves (the so called “barcoding gap”) should not be simply accepted as proof of species-level differences, as high intraspecific divergence can occur within widely distributed species (e.g., Hickerson et al. 2006; Collins & Cruickshank 2013; Zhang et al. 2017). However, the species of Euandrena here have only moderate ranges (restricted to the Himalayas), show consistent morphological differences, and present a tree topology that indicates that genetic distance is not the sole difference, with the sequence of A. communis sitting between the two more morphologically similar larger A. murreensis and aff. A. murreensis clades.

On the basis of these consistent morphological differences, combined with these divergent DNA barcodes and tree topology, the conclusion is therefore drawn that at least three red-marked species of Euandrena can be found in the Western Himalayas. Two of these species can be confidently referred to as A. communis (the smaller individuals with all of T2 red-marked; Fig. 1D , contrast Figs 4G–H ) and A. murreensis (larger individuals with the base and apexes of the terga red-marked, but not the tergal discs, Fig. 4G , the clypeus with deep scattered punctures, Fig. 4B , and the scutum predominantly shiny, Fig. 4E ) on the basis of currently available type material.

Full details on examined specimens are given below under each species entry, but in conclusion, 1) A. familiaris is currently only confidently known from the type series in the ZSI (probably a single male specimen) from the locus typicus in western China, 2) A. communis can be placed in the Euandrena and is known from northern India and Pakistan, 3) A. mephistophelica remains an unclear taxon, but is likely referable to one of the two large-bodied Euandrena genetic clades, and 4) A. murreensis is known from across northern Pakistan and India into Nepal (see below for examined material). The more recently described taxa from the Eastern Himalayas require type investigation, combined with additional genetic analysis from Nepal and southern China.

Figures

Fig. 1. Andrena (Euandrena) communis Smith, 1879, holotype, ♀ (NHMUK). A. Label information. B. Habitus, lateral view. C. Face, frontal view. D. Body, dorsal view.

Fig. 2. Andrena(Euandrena) murreensis Cockerell, 1923, lectotype, ♀ (NHMUK). A. Label information. B. Habitus, lateral view. C. Face, frontal view. D. Scutum, dorsal view. E. Hind tibia, lateral view. F. Terga, dorsal view.

Fig. 3. Phylogenetic tree (maximum likelihood) of Andrena Fabricius, 1775 subgenus Euandrena Hedicke, 1933 based on the mitochondrial COI gene. Numbers above branches represent bootstrap support (values of <75 are omitted).

Fig. 4. A, C, G, E. Andrena (Euandrena) murreensis Cockerell, 1923, ♀ (TJWC). A. Habitus, lateral view. C. Face, frontal view. E. Scutum, dorsal view. G. Terga, dorsal view. – B, D, F, H. Andrena (Euandrena) aff. murreensis, ♀ (TJWC). B. Habitus, lateral view. D. Face, frontal view. F. Scutum, dorsal view. H. Terga, dorsal view.