Taxonomic revision of Mordellistenahirtipes species complex with new distribution records (Insecta, Coleoptera, Mordellidae)

Abstract A taxonomic revision of species related to Mordellistenahirtipes Schilsky, 1895 is presented. Five species among the M.hirtipes complex are recognised: M.hirtipes Schilsky, 1895, M.pseudohirtipes Ermisch, 1965, M.purpurascens Costa, 1854, M.balearica Compte, 1985, and M.irritans Franciscolo, 1991. Descriptions, differential diagnoses, an identification key, and new distributional records are provided. Principal Component Analysis is performed for visualisation of differentiation between taxa. The following taxonomic acts are proposed: Mordellistenapodlussanyi Czető, 1990 and M.aegea Franciscolo, 1949 are proposed as junior subjective synonyms of M.hirtipes Schilsky, 1895; M.fageli Ermisch, 1969 and M.pseudohirtipeskrotosensis Czető, 1990 are proposed as junior subjective synonyms of M.pseudohirtipespseudohirtipes Ermisch, 1965; M.geronensis Ermisch, 1977 and M.istrica Ermisch, 1977 are proposed as junior subjective synonyms of M.purpurascens Costa, 1854.


Introduction
Mordellistena Costa, 1854 is the largest genus within the family Mordellidae Latreille, 1802, comprising approximately 800 described species (Horák 2011). Adults are commonly found on flowers where they feed on pollen and nectar. Larvae are found in the stems of herbaceous plants or in decaying wood.
The Western Palaearctic species are conventionally assigned to species groups proposed by Ermisch (1956Ermisch ( , 1969b, based on combinations of morphological characters. There are no studies focused on the phylogeny of Palaearctic species of genus Mordellistena, and thus the phylogenetic relationships among its species remain unresolved. The present study is focused on nine morphologically related taxa, which represent a complex within M. confinis species group (Ermisch 1956, 1969b, Batten 1977. The unique character shared by all species of the complex is the expanded second segment of maxillary palpi in males, bearing very long setae on the ventral surface (Figs 4A,6C,7C). Additional characters are a strongly convex body and unique shape of parameres (Figs 5D-G, 6E-H, 7G-J, 8, 9).
Species belonging to this complex were described by Costa (1854), Schilsky (1895), Ermisch (1965Ermisch ( , 1969a, Compte (1985), and Franciscolo (1949Franciscolo ( , 1991. Most of descriptions appeared to be insufficient for proper identification and after examination of type material, it became clear that some of the taxa are conspecific. In the present paper, we provide redescriptions of M. hirtipes Schilsky, 1895, M. pseudohirtipes Ermisch, 1965, and M. purpurascens Costa, 1854. Important diagnostic characters are visualised in drawings and photographs. We also performed Principal Component Analysis (PCA) based on morphometric measurements. This method is widely used in taxonomical research of invertebrates to help separate putative species in difficult species-complexes, to visualise differentiation between species and to evaluate the importance of peculiar morphometric characters (Kucharczyk et al. 2012, Przybycień andWacławik 2015).
Examination of material from several localities in Western Palaearctic revealed new distribution records and new biological information for M. hirtipes Schilsky, 1895, M. pseudohirtipes Ermisch, 1965, and M. purpurascens Costa, 1854.

Materials and methods
Dried specimens were relaxed in water with a few drops of acetic acid to allow for the dissection. Specimens were observed using Leica MZ16 stereomicroscope with magnification up to 120×, illuminated with diffuse light (neon bulb, 6400 K). Dissected body parts for drawings were temporarily mounted on slides in glycerine. Drawings were prepared using Leica drawing tube attached to Leica DM 1000 microscope, scanned and traced in Adobe Illustrator CS6. Dissected body parts were after examination mounted on the same card as respective specimen using dimethyl hydantoin formaldehyde (DMHF) or put to genitalia microvials filled with glycerine and pinned with the respective specimen. Measurements were taken using ocular micrometre. Intervals of measured

Data resources
The data underpinning the analyses reported in this paper are deposited at GBIF, the Global Biodiversity Information Facility, https://doi.org/10.15468/pkhkul

Mordellistena
Habitus illustrated in Fig. 1. Body slender, widest at the end of anterior third of elytra. Integument black. Head and pronotum covered with yellowish pubescence; pubescence on elytra yellowish in proximal half, gradually darkened towards apices, sometimes with reddish or violet metallic sheen; venter covered with yellowish pubescence, darkened along posterior margins of ventrites 3-5.
Sexual dimorphism. Females are usually more robust; with shorter antennae. Maxillary palpomere II is not expanded in females and without long setae on ventral side. Terminal maxillary palpomere is shorter in females, with angles more rounded. Protibiae are not expanded in females, without long setae in basal portion.
Biology. Adults were found on the flowers of Daucus sp. (Apiaceae) and Helichrysum sp. (Asteraceae) on dry grasslands and in urban environment.
Remarks. Franciscolo (1949) described M. aegea based on three specimens from Kos island (Greece). Batten (1977) mentioned that this species does not belong to the micans group because the antennomere IV and V are equal in length. Examination of holotype revealed that this specimen belongs to M. hirtipes. We consider this taxon as a junior synonym of M. hirtipes. Czető (1990) described M. podlussanyi based on a single male specimen from Crete. He mentioned in the original description that the maxillary palpomere II is not dilated. Examination of the holotype actually revealed, that the palpomere is expanded, and any other differences which could separate this taxon from M. hirtipes were found. This interpretation is also supported by the results of PCA analysis (Fig. 10A). We propose M. podlussanyi as a junior synonym of M. hirtipes.

Mordellistena
Habitus given in Fig. 2. Body strongly elongate, slender, widest just behind humeri. Integument black, mouthparts sometimes paler. Pubescence on head pale yellowish, on pronotum yellowish to dark grey, on elytra yellowish in anterior 1/2, darkened towards apices, or completely dark grey, sometimes with reddish or purplish metallic sheen, on pygidium dark grey, on venter yellowish, darkened along posterior margins of abdominal ventrites.
Sexual dimorphism. Females are more robust than males, their protibiae are not expanded in basal 1/3 and without fringe of long setae. Maxillary palpomere II is not expanded in females and without long setae on ventral side. Terminal maxillary palpomere is slenderer in females and its inner angle is situated more distally than in males. Antennae are somewhat shorter in females.
Biology. Adults were collected on the flowers of Apiaceae plants on dry grasslands. Plaza (1983) mentioned that M. pseudohirtipes was collected in Spain on following plant species: Thapsia villosa, Daucus carota (both Apiaceae) and Ruta montana (Rutaceae).

Remarks.
Mordellistena fageli was placed in the pentas-group in the original description, based on the dark pubescence and three ridges on the metatarsomere II. In fact, as Horák (1983) already mentioned, this species belongs to M. hirtipes complex, based on the strongly convex body and expanded maxillary palpomere II. Examination of type material did not reveal any characters which could separate this taxon from M. pseudohirtipes. In the plot from PCA analysis (Fig. 10A), M. fageli is placed just next to the cluster of M. pseudohirtipes, in the same plane along the PC 1 axis. We consider these taxa as conspecific and propose M. fageli as a junior synonym of M. pseudohirtipes. Czető (1990) described M. pseudohirtipes krotosensis based on two male specimens from Crete island. Characters such as length of the body, length of pygidium and colouration of pubescence, that he used for differentiation of the subspecies, are subjects of individual variability. Results of PCA (Fig. 10A) show, that holotype of M. pseudohirtipes krotosensis is placed within the cluster of the nominotypical subspecies. After examination of holotype we consider this subspecies as a junior synonym of M. pseudohirtipes pseudohirtipes.
In Ermisch's collection, there is a series of specimens named Mordellistena lopezi. Such species has not been described, and in fact, all the specimens belong to M. pseudohirtipes, except the one labelled as "Type", which belongs to M. purpurascens.  Habitus illustrated in Fig. 3. Body strongly elongate, slender, widest behind anterior 1/4 of elytra. Integument black, anterior margin of clypeus and mandibles somewhat paler. Pubescence on head and thorax yellowish; on elytra yellowish in anterior half, gradually darkened apically; on venter yellowish, darkened along posterior margins of ventrites 3 and 4 and completely dark grey on terminal ventrite and pygidium.
Sexual dimorphism. Females are more robust, with protibiae not expanded and without fringe of long setae in basal part. Maxillary palpomere II not expanded in females and without long setae on ventral side. Terminal maxillary palpomere is wider in males, with its inner angle situated approximately in the middle (Fig. 7C), in females it is generally slenderer, with its inner angle situated in terminal 1/3 (Fig. 7D). Antennae are shorter in females; antennomeres V-X ~1.60× as long as wide in males, ~1.30× in females.
Distribution. Croatia, France, Greece, Italy, Montenegro, Morocco, Spain. Mordellistena purpurascens is reported here for the first time from Greece and Montenegro. Odnosum (2003Odnosum ( , 2005Odnosum ( , 2010 reported M. purpurascens from Kazakhstan, Turkmenistan, Tajikistan, and Ukraine. However, based on the figures of parameres provided in all the three mentioned studies, it is obvious that he referred to a different species (see Discussion).
Biology. Adults were collected by the first author in Montenegro, in urban environment of Bar on the flowers of Daucus sp. (Apiaceae).
Remarks. Mordellistena purpurascens was described by Costa (1854) and referred to be found in several localities in former "Regno di Napoli" (southern parts of present Italy). Series of M. purpurascens in Costa's collection in MZFN contains only two specimens. One of them with the original label "Mordellistena purpurascens n. Napoli" is designed here as a lectotype. The other specimen labelled "S. Severina" without identification label belongs to a different species from the gemellata-group (sensu Ermisch 1956). Genitalia of the lectotype were examined for the first time for the purposes of the present study (Fig. 7G).  briefly described two new species M. geronensis and M. istrica as a part of an identification key. He differentiated these species from each other based on the shape of the apical part of the median lobe (expanded in M. istrica, not expanded in M. geronensis). Shape of the apical part of median lobe depends on the observation method (dry/wet, card mounted/slide mounted). After examining the series of slide mounted median lobes of both taxa, we did not find any differences in the shape. Examination of the male genitalia from type specimens of M. purpurascens (Fig. 7G), M. geronensis (Fig. 7H) and M. istrica (Fig. 7I) (Compte 1985), holotype should be deposited in MNCN. However, despite of the effort of the curator, the specimen was not found.
Diagnosis. Mordellistena balearica was described based on a single male specimen from Mallorca. According to the original description, this species closely resembles M. pseudohirtipes and can be distinguished from this species by longer antennae (antennomeres V-X two times longer than wide) and different shape of parameres ( Fig. 8) (Compte 1985). All characters mentioned in the original description suggest that this taxon is conspecific with M. pseudohirtipes; unfortunately, the authors did not have the opportunity to study the type.
Distribution. Known only from type locality. Remarks. Compte (1985) mentioned following information: "This specimen, together with other specimens collected by P. López in Majorca, which current location I don't know, was studied by the specialist Mr Ermisch, who considered it as a new species for science, called M. balearica, a name that seems to have remained in litteris". There are several specimens collected by P López in Mallorca in Ermisch's collection (SNSD) labelled by Ermisch as M. balearica which in fact all belong to M. thuringiaca Ermisch, 1963. Franciscolo, 1991 Fig. 9 Mordellistena irritans Franciscolo, 1991: 168-173 (Franciscolo 1991). Not examined.

Mordellistena (s. str.) irritans
Diagnosis. Mordellistena irritans can be assigned to M. hirtipes complex based on the expanded maxillary palpomere II in males and the shape of parameres. This species can be distinguished from all other species in the complex by the characteristic shape of the left paramere with dorsal branch parallel-sided and rounded at apex (Fig. 9).
Distribution. Known only from the type locality.

Discussion
The family Mordellidae is taxonomically very challenging and thus rather poorly known. Most of the original descriptions are insufficient for proper identification and differentiation of the species, especially those published before the 1950s (before K Ermisch provided a more precise method of description). There are still some species which were described as several different taxa, sometimes even by the same author (e.g., M. pseudohirtipes Ermisch, 1965= M. fageli Ermisch, 1969. Characters used for the differentiation of these taxa were usually misinterpreted (e.g., the shape of the median lobe in M. geronensis M. istrica Ermisch, 1977) or they are subjects of the intraspecific variability (e.g., the dark coloration of the pubescence in M. fageli Ermisch, 1969). In other cases, the insufficient descriptions in combination with overlooking of the type specimens led to a misinterpretation of the taxa. It can be seen for example in some species described by Achille Costa (1854). Revision of the type specimens in his collection deposited in MZFN revealed that several species described by him were incorrectly interpreted by the subsequent authors as completely different species, one of them is M. purpurascens Costa, 1854 treated in the present paper. This species was considered by the subsequent authors as a synonym of either M. pumila (Gyllenhal, 1810) (Gemminger and Harold 1870) or M. micans (Germar, 1817) (Emery 1876;Baudi di Selve 1877;Heyden et al. 1883Heyden et al. , 1906Schilsky 1898;Csiki 1915). Later it was treated again as a valid species by , Ermisch in Kaszab (1979) and Batten (1977) but none of these authors had studied the types and it is obvious, based on their figures of the genitalia that the specimens considered by them as M. purpurascens belong to a different species. Their misinterpretations were later followed by Odnosum (2003Odnosum ( , 2005Odnosum ( , 2010 who published several new distribution records for M. purpurascens which were then included in the catalogue by Horák (2008). Only the examination of the lectotype of M. purpurascens Costa, 1854 done by the first author revealed that it is conspecific with the types of M. geronensis M. istrica Ermisch, 1977. As it can be seen from this example, examination and redescriptions of the type specimens are essential for the future studies, especially in such taxonomically difficult and species-rich family as Mordellidae.
We live in the era of the global biodiversity crisis caused by the anthropogenic interventions in the natural ecosystems. But how does these changes affect the diversity and distribution patterns of Mordellidae beetles is not known. Despite of the great effort of the authors such as Ermisch (e.g., 1956Ermisch (e.g., , 1965Ermisch (e.g., , 1969a, Horák (e.g., 1990Horák (e.g., , 2008 and Odnosum (e.g., 2003Odnosum (e.g., , 2005Odnosum (e.g., , 2010 who have summarised and published a vast number of distributional records, the information about distribution and ecology of many Palearctic species is still very poor and several species are reported only from a single locality stated in the original description. It is very important to gather and provide new distributional and ecological records, however, it is also essential to pay effort to correct identification of the specimens to guarantee the accuracy of the published biological data.