Are Appearances Deceiving? Morpho-Genetic Complexity of the Eumerus tricolor Group (Diptera: Syrphidae) in Europe, with a Focus on the Iberian Peninsula

Simple Summary Hoverflies are a diverse group (6200+ species) of two-winged flies that provide important ecosystem services (e.g., plant pollination and regulation of populations of pest insects). The hoverfly genus Eumerus has caught the attention of scientists and entomology amateurs due to its high number of species worldwide (250+) and its wide range of body sizes and shapes. Nevertheless, some species show a similar overall appearance making both their separation from each other and the study of their classification/biology difficult. Thus, the aim of the present work is to update the knowledge of the Eumerus hoverflies in the Iberian Peninsula by assessing the diversity of the Eumerus tricolor group under an integrative approach (i.e., combining different techniques and data sources). We found and described two new species from Spain and provided the most comprehensive identification key for all known European species of this Eumerus group. Abstract Eumerus Meigen, 1822 is one of the largest Syrphidae genera in the Palaearctic Region, with the highest levels of taxonomic diversity found in the Eumerus tricolor species group. Despite its high diversity, the interspecific levels of morphological variability can be low. Additionally, some species may show certain levels of intraspecific variability. Hence, species delimitation may become challenging. In this work, we assessed the diversity of the E. tricolor group in the Iberian Peninsula through an integrative analysis of nomenclature, morphology and the 5′ (COI-5′) and 3′ (COI-3′) end regions of the Cytochrome c oxidase subunit I gene. Two new species, Eumerus ancylostylus Aguado-Aranda & Ricarte sp. n. and Eumerus petrarum Aguado-Aranda, Nedeljković & Ricarte sp. n., were described, and their intra- and interspecific variations discussed. In addition, the first barcodes of Iberian members of the E. tricolor group were obtained, and the distribution ranges of all species were mapped within the study area. The systematic position of the new species is discussed based on the resulting COI-based trees. The male genitalia of Eumerus hispanicus van der Goot, 1966 and Eumerus bayardi Séguy, 1961 were studied and illustrated. A lectotype was designated for Eumerus lateralis (Zetterstedt, 1819). An updated dichotomous key for all known European species of the E. tricolor group is provided. The egg of E. petrarum sp. n. is also described.


Morphological Study
All species were, at the first stage, identified with the available dichotomous keys for Eumerus [15,[29][30][31]. For the undescribed species not suiting the species included in the keys, their taxonomic statuses were investigated. In the new species descriptions, first, a diagnosis was provided and second, a detailed morphological description of males (based on the holotype) and females, if available. Then, species distribution, biology and special remarks of the new species are indicated. To complete the new species descriptions, the following measures were taken on several adult specimens ('n') with the software Leica Application Suite X (LAS X) ® v3.0.4.16529: Body length ('l'); basoflagellomere ratio (length:width; see figure 4a of Aguado-Aranda et al. [23]); and the maximum width of the vertical triangle ('vtw') measured at a mid-point on the ocellar triangle. Male genitalia were dissected and prepared following Ricarte et al. [32] and then stored in glycerine in plastic microvials. All photographs of the specimens and male genitalia were taken with a Leica DFC 450 camera attached to a Leica M205 C binocular microscope. Male genitalia of Eumerus bayardi Séguy, 1961, E. hispanicus, E. sabulonum and Eumerus aff. sabulonum were hand drawn from printed photographs.
For eggs of Eumerus aff. sabulonum, a detailed description of their external surface was undertaken from images of scanning electron microscopy (SEM). The eggs were fixed and preserved in ethanol 70%. They were dehydrated using an ethanol/water series of 2 baths of 70% for 5 min each, followed by one bath of 80%, 90% and 96% ethanol for 5 min each and 2 baths in absolute ethanol of 5 min each. The dehydrated eggs were mounted on aluminum stubs with a double-sided adhesive carbon tape and sputter-coated in a Quorum Morphological terminology used in the adult descriptions follows Thompson [33] except for the term "hair/s," which is used in replacement of "pilis/pile," as well as the terms "fossette" and "notopleural sulcus" which follow Doczkal & Pape [34]. The terminology for male genitalia follows Doczkal [35], and for egg follows Chandler [36]. For those illustrations cited from the literature (not shown in the present paper), 'figure' (lower case) is used, while ' Figure' (upper case) is used for those which are original from this work (shown in the present paper).

Molecular Study
DNA was extracted both from 1 mesoleg and 1 metaleg of 2 males of Eumerus aff. grandis, 3 specimens (2 males and 1 female) of E. grallator, 2 specimens (1 male and 1 female) of E. hispanicus, 8 specimens (5 males and 3 females) of Eumerus aff. sabulonum, 5 specimens (4 males and 1 female) of E. sabulonum, 1 male of E. tarsalis and 2 specimens (1 male and 1 female) of E. tricolor. DNA of one male of Eumerus aff. grandis (code CEUA00017798) and two males of E. grandis were extracted from a wing. The NZY Tissue gDNA Isolation kit, following the manufacture's protocol for animal tissues, was used for two males of Eumerus aff. grandis and all specimens of E. grallator, E. hispanicus, E. sabulonum, Eumerus aff. sabulonum, E. tarsalis and E. tricolor. Otherwise, the DNA of 1 male of Eumerus aff. grandis (code CEUA00017798), 2 males of E. bayardi and 2 males of E. grandis was extracted using the QIAGEN DNeasy Blood and Tissue kit with some modifications in the manufacturer's protocol.
PCR amplifications of the 5 (COI-5 ) and 3 (COI-3 ) end regions of the Cytochrome c oxidase subunit I gene were performed using universal and custom primers ( Table 1). The COI-5 fragments "A," "B" and "C" were amplified for "older" specimens (collection date previous to the year 2000) in which DNA is heavily fragmented and "recent" specimens of which the complete COI-5 was not achieved to amplify. All amplifications were carried out in a total volume of 25 µL containing 1 × of Buffer reaction, 0.4 mM of dNTPs, 0.2 µM of each primer, 0.65-2 mM of MgCl 2 and 1-2 units of DNA polymerase. Thermocycler conditions followed those used by Grković et al. [37] and Grković et al. [25] for COI-5 , and Chroni et al. [12], but with annealing at 48 • C, for COI-3 . Otherwise, the PCR profile for the barcodes of E. bayardi, all fragments of COI-5 and COI-3 (specimen vouchers CEUA_S71-72, S83-84, S86-87, S93 and S231-232) consisted of an initial denaturalization of 3 min at 94 • C; 45 cycles of 45 s at 94 • C, annealing of 45 s at 45 • C and 1 min at 72 • C; with a final elongation step of 5 min at 72 • C.
All PCR products were visualized with an electrophoresis process in a 1-1.5% agarose gel, except those of E. bayardi, which were visualized and purified in an E-Gel CloneWell 0.8% agarose gel (Life Technologies Corporation, Austin, TX, USA). Purifications of COI-5 fragments and COI-3 (vouchers above indicated) products followed the USB ® ExoSAP-IT ® method (Affymetrix, Inc., Santa Clara, CA, USA). Sequencing reactions of the fragments, COI-5 of E. bayardi and COI-3 products (vouchers above indicated) were carried out with the BigDye TM Terminator v3.1 Cycle Sequencing kit (Thermo Fisher Scientific, Inc., Waltham, MA, USA). The rest of the PCR products were purified and sequenced at Macrogen (Macrogen, Inc., Seoul, Republic of Korea).
The sequences were edited by eye with the program Sequencher v5.4.6 (Gene Codes Corporation 2017, Ann Arbor, MI). Then, COI-5 and COI-3 sequences of Eumerus alpinus Rondani, 1857, Eumerus minotaurus Claussen & Lucas, 1988, Eumerus sogdianus Stackelberg, 1952 and at least one male and female specimen of each European species of the E. tricolor group available at the public repositories GenBank and BOLD [38] were downloaded (see Supplementary Material S2). Alignments were performed manually and checked with the program AliView v1.25 [39]. The final COI-5 and concatenated (COI-5 +COI-3 ) matrices had lengths of 569 and 1289 bp, respectively. Then, Neighbor-Joining (NJ) and Maximum Likelihood (ML) analyses for both datasets were carried out in MEGA7 [40] Insects 2023, 14, 541 5 of 33 with 1000 bootstrap replications using the Maximum Likelihood Composite and General Time Reversible (GTR) models, respectively. A GTR+G model (with gamma distribution) was used for the COI-5 matrix, and a GTR+G+I model (with invariant sites) was used for the COI (COI-5 +COI-3 ) matrix. The resulting trees were rooted based on a Xanthogramma citrofasciatum (De Geer, 1776) sequence. Table 1. List of primers used in the molecular study of the E. tricolor species group.

Gene Region
Primer Sequence Reference

Integrative Approach
As a result of the literature review, 66 species of the E. tricolor group were found to occur in the Palaearctic Region (Table 2). Regarding the Ibero-Balearic area, 128 new specimens were collected of the following species: 2 males of Eumerus aff. grandis, 1 male of E. azabense, 1 male of E. bayardi, 26 specimens of E. grallator (25 males and 1 female), 4 specimens of E. hispanicus (3 males and 1 female), 31 specimens of E. sabulonum (27 males and 4 females), 41 specimens of Eumerus aff. sabulonum (36 males and 5 females), 5 males of E. tarsalis and 17 specimens of E. tricolor (13 males and 4 females). In the molecular analyses, 21 species of the E. tricolor group were considered in the COI-5 dataset and 17 species in the COI (COI-5 +COI-3 ) dataset (see Supplementary Material S2). Apart from those of the undescribed species (Eumerus aff. sabulonum, Eumerus aff. grandis), we generated the first known COI sequences for the following species: E. bayardi, E. grallator, E. hispanicus, E. sabulonum and E. tarsalis. In addition, we also obtained new COI sequences for E. grandis and COI-5 for E. tricolor. All resulting trees had virtually the same topologies (Figures 1 and 2). The COI-based trees showed the highest bootstrap values (>70) for the two new species that clustered in single clades (Figure 2), supporting our morphological hypotheses. Therefore, based on the combination of the nomenclatural, morphological and molecular evidence, the new species were characterized.   Stackelberg, 1952 Ukraine EN     Diagnosis. This species can be distinguished by the following combination of features (only males, unknown female): large body size (>9.5 mm); eyes hairy and touching along a line ( Figure 3C,D); wide vertical triangle (0.58-0.61 mm) ( Figure 4A); basoflagellomere trapezoidal in shape; mesonotum with moderately long, golden yellow hairs ( Figure 3A), which are of two lengths intermixed on the posterior part, with the shortest hairs surpassing half the length of the longest ones; posterior surstylar lobe arched anteriorly ( Figure 6A). This species is most similar to E. grandis (see Taxonomic notes).

New Species
Etymology. The specific epithet 'ancylostylus' originates from the combination of the Latinised Greek words 'ankylos' (curved) and 'stylos' (stylus, of the male genitalia) and refers to the curved/round anterior part of the posterior surstylar lobe in male.       Description. MALE (holotype). Measurements (mm): l = 9.56; vtw = 0.58. Head. Eyes contiguous along a line. Eye hairs moderately short and disperse ( Figure 3C,D). Face and frontal triangle black; face grey pollinose. Frontal triangle and face covered with dense, golden hairs. Ocellar triangle isosceles. Vertical triangle black; anterior half (including ocellar triangle) with long, erect and black hairs. Apex and posterior half of the vertical triangle covered with whitish-yellow hairs ( Figure 3C,D). Occiput black, with whitish yellow hairs which become shorter (than those on the vertical triangle) and whiter toward the laterals; occiput slightly grey pollinose along eye margin and covered in short, black hairs. Scape and pedicel black; pedicel with black hairs ventrally, longer than those on the dorsal side of the scape. Basoflagellomere trapezoidal in shape (length:width ratio = 1:1), slightly striated radially and black; basoflagellomere with a round fossette at the distal margin ( Figure 3C). Pedicel and basoflagellomere sparsely white pollinose (pollinosity more obvious under artificial white lighting). Arista black and bare. Thorax. Mesonotum and pleura black. Scutum covered with golden yellow hairs, slightly shorter than those on the vertical triangle; hairs on the posterior half of the scutum almost uniformly long, with the shortest hairs surpassing half of the length of the longest ones; scutum with two medial, white pollinose vittae slightly exceeding the transverse suture; width of the vittae less than 1/8 of the total width of scutum. Notopleural sulcus absent. Disc of scutellum with golden yellow hairs of the same length as that on the scutum; posterior margin of the scutellum with small teeth-like protuberances, each bearing a long black hair apically. Posterior anepisternum, anterior anepimeron and katepisternum on its postero-lateral and ventral areas bearing densely-arranged long golden yellow hairs. Katatergum with a discrete bunch of white hairs. Pleuron grey pollinose except the anterior anepimeron posterolaterally, the upper half of the posterior anepimeron and the katepimeron anterolaterally. Metasternum covered with long, yellowish-white hairs. Femora black; apices of pro-and mesofemora light brown. Tibiae black; pro-and mesotibiae dark brown basally. Tarsi black dorsally and yellow ventrally. Profemur with long, directed backward, golden yellow hairs on its posterior side. Mesofemur with long, directed backward, golden yellow hairs on its posterior side, which are curved at their apex. Pro-and mesofemur with few long, directed backward, black hairs intermixed with the yellow ones on their posterior sides. Metacoxa densely covered with long and white hairs anteriorly and bare posteriorly. Metafemur with dense and yellow hairs, with black hairs postero-apically; hairs on ventral side of the metafemur at least twice the length of the subapical spinae. Metafemur with an anteroventral row of nine spinae and a posteroventral row of six spinae apically. Metatibia with a basoventral ridge covered with short, reclined setae. Wing membrane extensively microtrichose; posterior margin of the wing with dense, short and brown hairs; margin of ventral calypter with rather long and yellow hairs; hairs on the margin of dorsal calypter shorter than those on the ventral calypter; halter brownish yellow. Abdomen. Tergum I black. Terga II-IV black, lateral sides reddish-orange; terga II-IV with a pair of slightly curved, white pollinose maculae. Terga II-IV covered with short, semi-reclined, black hairs.
Tergum II with white hairs on lateral margins (hairs at the anterior corners rather long). Tergum III covered with short and black hairs on lateral margins (hairs on the basal part of the maculae white). Tergum IV with short and black hairs on the anterior half of the lateral margins, with white hairs on the posterior half of lateral margins. Sterna I-IV dark brown; anterior margin of sternum I black. Sternum IV flat (without projections); posterior margin of sternum IV slightly V-shaped. Sterna II and III covered with white hairs. Sternum IV covered with short, reclined, white hairs. Genitalia. Epandrium with a simple posterior surstylar lobe, straight posteriorly and rounded anteriorly, and slightly inclined in the direction opposite to cerci. Cercus with long hairs. Hypandrium with two large, round bulges and three little, almost rounded, hyaline expansions at the base ( Figure 5B).     Biology. Adults fly from June to early August at altitudes of over 1300 m asl. The predominant vegetation in the Cantabrian locality of E. ancylostylus sp. n. was mowing meadows and hazelnut woods [46]. The specimens from the Pyrenees were found in a valley, on a side mountain track, amongst vegetation consisting of herbs, scrubs and mixed woodland, and relatively close to a river [44].
Taxonomic notes. The new species resembles morphologically E. grandis, which was described from an undetermined number of specimens from 'Europe' (unknown precise locality). The type material of E. grandis is lost [11]. According to Peck [47], E. grandis has two synonyms: Eumerus annulatus, originally described as Syrphus annulatus Panzer, 1798, and Eumerus varius Meigen, 1822. Syrphus annulatus sensu Panzer was revealed to be a primary homonym of S. annulatus Fabricius, 1798, recently synonymized with Merodon natans (Fabricius, 1794) [48]. Grković et al. [11] confirmed the status of E. annulatus (Panzer, 1798) as a synonym of E. grandis. In addition, E. varius was excluded from the list of synonyms of E. grandis since the examined male of E. varius from the supposed type series was actually E. tricolor. This male was designated as lectotype of E. varius in order to stabilize this species concept [11].
Eumerus ancylostylus sp. n keyed out as E. grandis in the keys for the south-eastern European species of the E. tricolor group [11] and in those of the Eumerus from Switzerland and surrounding parts of central Europe [30]. Moreover, the new species keyed out as E. annulatus, later described by Meigen as E. grandis [11], in the Stackelberg's key for the Palaearctic species of Eumerus [31]. Nevertheless, the new species differs from E. grandis in the basoflagellomere ratio, which is wider in E. ancylostylus sp. n. (1 : 1-1.1, n = 3) than in E. grandis (1 : 0.8-0.9, n = 3), the width of the vertical triangle, which is wider in E. ancylostylus sp. n. ( Figure 4A,B), the length of the hairs on posterior half of the scutum, which are all almost of equal length in E. ancylostylus sp. n. but clearly different in length in E. grandis (short hairs do not exceed half of the length of the longest hairs) and the shape of the posterior surstylar lobe, rounded anteriorly in E. ancylostylus sp. n. but more square- Biology. Adults fly from June to early August at altitudes of over 1300 m asl. The predominant vegetation in the Cantabrian locality of E. ancylostylus sp. n. was mowing meadows and hazelnut woods [46]. The specimens from the Pyrenees were found in a valley, on a side mountain track, amongst vegetation consisting of herbs, scrubs and mixed woodland, and relatively close to a river [44].
Taxonomic notes. The new species resembles morphologically E. grandis, which was described from an undetermined number of specimens from 'Europe' (unknown precise locality). The type material of E. grandis is lost [11]. According to Peck [47], E. grandis has two synonyms: Eumerus annulatus, originally described as Syrphus annulatus Panzer, 1798, and Eumerus varius Meigen, 1822. Syrphus annulatus sensu Panzer was revealed to be a primary homonym of S. annulatus Fabricius, 1798, recently synonymized with Merodon natans (Fabricius, 1794) [48]. Grković et al. [11] confirmed the status of E. annulatus (Panzer, 1798) as a synonym of E. grandis. In addition, E. varius was excluded from the list of synonyms of E. grandis since the examined male of E. varius from the supposed type series was actually E. tricolor. This male was designated as lectotype of E. varius in order to stabilize this species concept [11].
Eumerus ancylostylus sp. n keyed out as E. grandis in the keys for the south-eastern European species of the E. tricolor group [11] and in those of the Eumerus from Switzerland and surrounding parts of central Europe [30]. Moreover, the new species keyed out as E. annulatus, later described by Meigen as E. grandis [11], in the Stackelberg's key for the Palaearctic species of Eumerus [31]. Nevertheless, the new species differs from E. grandis in the basoflagellomere ratio, which is wider in E. ancylostylus sp. n. (1:1-1.1, n = 3) than in E. grandis (1:0.8-0.9, n = 3), the width of the vertical triangle, which is wider in E. ancylostylus sp. n. (Figure 4A,B), the length of the hairs on posterior half of the scutum, which are all almost of equal length in E. ancylostylus sp. n. but clearly different in length in E. grandis (short hairs do not exceed half of the length of the longest hairs) and the shape of the posterior surstylar lobe, rounded anteriorly in E. ancylostylus sp. n. but more square-shaped in E. grandis ( Figure 6A,B). In addition, E. ancylostylus sp. n. differs clearly from E. grandis in COI sequences.
We also examined the type material of Eumerus lateralis (Zetterstedt, 1819) (Figure 8), described under the genus Pipiza Fallén, 1810, because Meigen [49] stated in his description of E. grandis: "Whether this species is Pipiza lateralis Falléni (v. Kon. Vet. Ak. Handl. 1819. St. I. Nro. 38) I cannot state for lack of an accurate description of it". After examination of the type material of E. lateralis (1 male and 1 female), we confirmed that all were conspecific and the male differed from E. ancylostylus sp. n. in the basoflagellomere ratio, which is wider in E. ancylostylus sp. n. (1:1-1.1, n = 3) than in E. lateralis (1:0.8), the width of the vertical triangle, slightly wider (0.1 mm) in the new species ( Figure 4A,C), the length of the hairs on posterior half of the scutum, which are all almost of equal length in E. ancylostylus sp. n. but different in length in E. lateralis (short hairs are at most as long as half the length of the longest hairs) and the shape of the posterior surstylar lobe, which is square-shaped in E. lateralis, as in E. grandis ( Figure 6B,C). The male from the type series of E. lateralis was designated as lectotype to stabilize the morphological concept of this species. Fresh material of E. lateralis from the type locality or close was not available to us. Given the similarity among these species and in the absence of molecular data on E. lateralis, we cannot confirm if E. lateralis is the same species as E. grandis or not. Nevertheless, based on the morphological similarities, E. lateralis and E. grandis may actually be the same species. Thus, fresh material of E. lateralis should be studied in the future to clarify the taxonomic status of this species. shaped in E. grandis ( Figure 6A,B). In addition, E. ancylostylus sp. n. differs clearly from E. grandis in COI sequences.
We also examined the type material of Eumerus lateralis (Zetterstedt, 1819) ( Figure  8), described under the genus Pipiza Fallén, 1810, because Meigen [49] stated in his description of E. grandis: "Whether this species is Pipiza lateralis Falléni (v. Kon. Vet. Ak. Handl. 1819. St. I. Nro. 38) I cannot state for lack of an accurate description of it". After examination of the type material of E. lateralis (1 male and 1 female), we confirmed that all were conspecific and the male differed from E. ancylostylus sp. n. in the basoflagellomere ratio, which is wider in E. ancylostylus sp. n. (1 : 1-1.1, n = 3) than in E. lateralis (1 : 0.8), the width of the vertical triangle, slightly wider (0.1 mm) in the new species ( Figure  4A,C), the length of the hairs on posterior half of the scutum, which are all almost of equal length in E. ancylostylus sp. n. but different in length in E. lateralis (short hairs are at most as long as half the length of the longest hairs) and the shape of the posterior surstylar lobe, which is square-shaped in E. lateralis, as in E. grandis ( Figure 6B,C). The male from the type series of E. lateralis was designated as lectotype to stabilize the morphological concept of this species. Fresh material of E. lateralis from the type locality or close was not available to us. Given the similarity among these species and in the absence of molecular data on E. lateralis, we cannot confirm if E. lateralis is the same species as E. grandis or not. Nevertheless, based on the morphological similarities, E. lateralis and E. grandis may actually be the same species. Thus, fresh material of E. lateralis should be studied in the future to clarify the taxonomic status of this species.   Diagnosis. This species can be distinguished by the following combination of characters: bare eyes; in males, eyes not touching along a line ( Figure 10A); wide vertical triangle (0.44-0.61 mm) ( Figure 10E), posterior surstylar lobe slightly constricted ( Figure 11B), interior accessory lobe of the genitalia triangular-shaped ( Figure 11B); in females, vertical triangle slightly elevated dorsally in lateral view ( Figure 10D), occiput usually with a little orange macula on its dorsal part near the vertex. This species is most similar to E. sabulonum (see Taxonomic notes). dark orange. Terga II-IV covered with short, reclined and white hairs (black on middle area of terga II-III); anterior corners of tergum II with long and white hairs. Tergum II with a pair of rounded, white pollinose maculae; maculae on terga III-IV slightly curved. Sterna I-II orange; anterior margin of sternum I black. Sternum III orange brownish. Sternum IV black and flat (without projections); the posterior margin of sternum IV slightly convex. Sterna II-IV with short, semi-reclined white hairs. Genitalia. Epandrium with a simple posterior surstylar lobe, spatula-shaped and slightly inclined anteriorly ( Figure 11B). Cercus with long hairs. Interior accessory lobe triangular-shaped ( Figure 11B). Hypandrium simple, curved almost 90° in the middle ( Figure 11A). FEMALE. Same as males, besides sexual dimorphism, except for the following characters: pollinosity on face mostly present along eye margin; basoflagellomere usually 1.2 wider than long; hairs on scutum slightly shorter.
EGG. Measurements (mm): length = 0.6-0.91 (n = 10). White when recently laid but yellowish-white when older. The chorionic sculpture consists of irregularly oval-shaped units with a sinuous margin and punctured surface raised by columns. Units do not touch each other (Figure 12). Distribution. Sierra Nevada, in Granada province (Figure 7). Biology. Adults fly from late May to July and can be found in a wide altitudinal range (1400-3000 m) but are mainly associated with high mountain areas (>1900 m). The predominant vegetation in localities at low altitudes was woodlands (Quercus sp.) but with grasslands ('Borreguiles') at high altitudes.
Taxonomic notes. Eumerus petrarum sp. n. was run through the key of Eumerus from Switzerland and surrounding parts of central Europe [30] and the Stackelberg's key to the Palaearctic Eumerus [31], keying out as E. sabulonum. The new species differs from E. sabulonum in the vertical triangle, which is wider in E. petrarum sp. n. (0.44-0.61 mm, n = 10) than in E. sabulonum (0.33-0.42 mm, n = 10); the shape of the posterior surstylar lobe, laterally constricted in E. petrarum sp. n. but square-shaped in E. sabulonum ( Figure 13B); and the interior accessory lobe of the male genitalia, triangular-shaped in E. petrarum sp. n. but square in E. sabulonum ( Figure 13B); in females, occiput with a little orange macula on its dorsal part close to the vertex, which is highly reduced (sometimes absent) in E. petrarum sp. n. but usually more noticeable in E. sabulonum. Furthermore, based on the examined material, E. petrarum sp. n. shows intraspecific variation in the dorsal coloration of protarsomere II (from yellow to black); the coloration of mesotarsomere III (from yellow to black); the number of spinae on the anterior row of the metafemur (six to nine); in males, the morphology of the surstylus (from more to less constricted laterally; Figure 13C); in females, body length is variable (8.04-9.48 mm; n = 5). Regarding the chorionic sculpture, the eggs of E. petrarum sp. n. are similar to those of E. sabulonum but differ on their surface, which is punctured in E. petrarum sp. n. (Figure 12) but smooth in E. sabulonum (see figure  1 of Munk [51]). In addition, the overall appearance of the eggs of E. petrarum sp. n. also Etymology. The specific epithet derives from the Latin 'petra/ae' (rock) and refers to the fact that many specimens were collected in rocky habitats, usually resting or thermoregulating on rocks. The specific epithet 'petrarum' should be treated as a genitive invariable name.
Description. MALE (holotype). Measurements (mm): l = 8.3; vtw = 0.6. Head. Eyes approximated on a point, narrowly separated each other by a distance equalling the diameter of the frontal ocellus. Eye bare. Face and frontal triangle black, white pollinose and covered with white hairs. Vertical triangle black and with long, erect, white hairs. Ocellar triangle isosceles, with black and white hairs intermixed. Occiput black and purple iridescent (under artificial lighting), covered with hairs shorter than those on the vertical triangle; occiput grey pollinose along the eye margin. Scape and pedicel black; pedicel with white hairs which are longer on the ventral side. Basoflagellomere square (ratio length:width = 1:1), concave ventrally, slightly striated radially and black; basoflagellomere with a round, dark brown fossette at the distal margin. Pedicel and blasoflagellomere sparsely white pollinose (pollinosity more obvious under artificial white lighting). Arista black, bare. Thorax. Mesonotum and pleura black. Scutum covered with short, semi-reclined white hairs; scutum with two medial white pollinose vittae slightly exceeding the transverse suture; width of the vittae less than 1/8 of the total width of the scutum. Notopleural sulcus absent. Disc of scutellum with white hairs, with hairs of the same length as those on the scutum; posterior margin of the scutellum with small teeth-like protuberances, each bearing a short white hair apically. Posterior anepisternum, anterior anepimeron and katepisternum on its postero-lateral area with dense, long white hairs. Katatergum with a discrete bunch of white hairs. Pleuron grey pollinose except the posterior anepisternum antero-basally, the katepisternum anterolaterally and the posterior anepimeron. Femora black, apices light brown. Tibiae black, basal third light brown. Proand mesobasotarsomeres yellow. Second pro-and mesotarsomeres yellow, black at the base. Third pro-and mesotarsomeres black, ventrally yellow and black at the base. Fourth and fifth pro-and mesotarsomeres black. Metafemur with white hairs, ventrally with hairs of different lengths, including some which are as long as the subapical spinae. Metafemur with an anterior row of nine spinae and a posterior row of six spinae apicoventrally. Metatibia with a basoventral ridge covered with short, reclined and black setae. Metatarsi black; metabasotarsomere and second and third tarsomeres ventrally yellow. Wing membrane extensively microtrichose; posterior margin of the wing with dense, short and brown hairs; margin of ventral calypter with long and yellow hairs; margin of dorsal calypter with hairs shorter than that of ventral calypter; halter dark yellow. Abdomen. Terga I and IV black; posterior margin of tergum IV brownish orange. Tergum II orange. Tergum III dark orange. Terga II-IV covered with short, reclined and white hairs (black on middle area of terga II-III); anterior corners of tergum II with long and white hairs. Tergum II with a pair of rounded, white pollinose maculae; maculae on terga III-IV slightly curved. Sterna I-II orange; anterior margin of sternum I black. Sternum III orange brownish. Sternum IV black and flat (without projections); the posterior margin of sternum IV slightly convex. Sterna II-IV with short, semi-reclined white hairs. Genitalia. Epandrium with a simple posterior surstylar lobe, spatula-shaped and slightly inclined anteriorly ( Figure 11B). Cercus with long hairs. Interior accessory lobe triangular-shaped ( Figure 11B). Hypandrium simple, curved almost 90 • in the middle ( Figure 11A).
FEMALE. Same as males, besides sexual dimorphism, except for the following characters: pollinosity on face mostly present along eye margin; basoflagellomere usually 1.2 wider than long; hairs on scutum slightly shorter.
EGG. Measurements (mm): length = 0.6-0.91 (n = 10). White when recently laid but yellowish-white when older. The chorionic sculpture consists of irregularly oval-shaped units with a sinuous margin and punctured surface raised by columns. Units do not touch each other (Figure 12).
Distribution. Sierra Nevada, in Granada province (Figure 7). Biology. Adults fly from late May to July and can be found in a wide altitudinal range (1400-3000 m) but are mainly associated with high mountain areas (>1900 m). The predominant vegetation in localities at low altitudes was woodlands (Quercus sp.) but with grasslands ('Borreguiles') at high altitudes.
Taxonomic notes. Eumerus petrarum sp. n. was run through the key of Eumerus from Switzerland and surrounding parts of central Europe [30] and the Stackelberg's key to the Palaearctic Eumerus [31], keying out as E. sabulonum. The new species differs from E. sabulonum in the vertical triangle, which is wider in E. petrarum sp. n. (0.44-0.61 mm, n = 10) than in E. sabulonum (0.33-0.42 mm, n = 10); the shape of the posterior surstylar lobe, laterally constricted in E. petrarum sp. n. but square-shaped in E. sabulonum ( Figure 13B); and the interior accessory lobe of the male genitalia, triangular-shaped in E. petrarum sp. n. but square in E. sabulonum ( Figure 13B); in females, occiput with a little orange macula on its dorsal part close to the vertex, which is highly reduced (sometimes absent) in E. petrarum sp. n. but usually more noticeable in E. sabulonum. Furthermore, based on the examined material, E. petrarum sp. n. shows intraspecific variation in the dorsal coloration of protarsomere II (from yellow to black); the coloration of mesotarsomere III (from yellow to black); the number of spinae on the anterior row of the metafemur (six to nine); in males, the morphology of the surstylus (from more to less constricted laterally; Figure 13C); in females, body length is variable (8.04-9.48 mm; n = 5). Regarding the chorionic sculpture, the eggs of E. petrarum sp. n. are similar to those of E. sabulonum but differ on their surface, which is punctured in E. petrarum sp. n. (Figure 12) but smooth in E. sabulonum (see figure 1 of Munk [51]). In addition, the overall appearance of the eggs of E. petrarum sp. n. also resembles those of Eumerus etnensis van der Goot, 1964. However, they differ in the shape of the units, which are irregular in E. petrarum sp. n., but smooth and rounded with irregular branches in E. etnensis (see figure 2 of Pérez-Bañón & Marcos-García [52], as Eumerus purpurariae Baéz, 1982). According to Peck [47], three species are listed as synonyms of E. sabulonum: Eumerus litoralis Curtis, 1839, Eumerus rubriventris Macquart, 1829 and Eumerus selene Meigen, 1822. The type series of E. litoralis consists of five specimens (three males and two females) collected "on the sand hills near Christchurch" in Great Britain. All syntypes are preserved in the MMV collection. We examined pictures of the habitus and the genitalia of one male (Figure 14), which confirmed the synonymy with E. sabulonum and its differentiation with the new species. Eumerus rubriventris was described based on a single male from the north of France, without a specific locality. According to Horn & Kahle [53], the main section of the Macquart collection was destroyed. The remaining part was divided and preserved between the 'Museum d'Historie Naturelle' (MHNL) in Lille, the 'Museum National d'Historie Naturelle' (MNHNP) in Paris and the Museum of Natural History of the Oxford University (OUMNH) (Xavier Lair, in litt.). There are not records of this species neither in the MNHL (Olivier Boilly, in litt.) nor in the MNHPN and the OUMNH. Therefore, we assume that type of material is lost. Moreover, and considering the original (male) description of E. rubriventris, the vertical triangle is dark green colored, and the mesonotum exhibits an olive-green coloration, differing from the new species in which both are black. The description of E. selene is based on an undetermined number of males from an unknown locality in Europe. The only preserved specimen of this species we have been able to locate is a semi-destroyed female collected in Bavaria, labeled as "E. silene, Meig./Allemagne/1155" and deposited in the MNHNP (https://science.mnhn.fr/institution/mnhn/collection/ed/item/ed4105?listIndex=25&listCount=32, accessed on 15 March 2023). Thus, the type series of this species is apparently lost. Regarding the male description of E. selene, the white maculae on tergum II are smaller than those on tergum III, differing from the new species in which the maculae of tergum II are the widest of the three pairs. According to Peck [47], three species are listed as synonyms of E. sabulonum: Eumerus litoralis Curtis, 1839, Eumerus rubriventris Macquart, 1829 and Eumerus selene Meigen, 1822. The type series of E. litoralis consists of five specimens (three males and two females) collected "on the sand hills near Christchurch" in Great Britain. All syntypes are preserved in the MMV collection. We examined pictures of the habitus and the genitalia of one male (Figure 14), which confirmed the synonymy with E. sabulonum and its differentiation with the new species. Eumerus rubriventris was described based on a single male from the north of France, without a specific locality. According to Horn & Kahle [53], the main section of the Macquart collection was destroyed. The remaining part was divided and preserved between the 'Museum d'Historie Naturelle' (MHNL) in Lille, the 'Museum National d'Historie Naturelle' (MNHNP) in Paris and the Museum of Natural History of the Oxford University (OUMNH) (Xavier Lair, in litt.). There are not records of this species neither in the MNHL (Olivier Boilly, in litt.) nor in the MNHPN and the OUMNH. Therefore, we assume that type of material is lost. Moreover, and considering the original (male) description of E. rubriventris, the vertical triangle is dark green colored, and the mesonotum exhibits an olive-green coloration, differing from the new species in which both are black. The description of E. selene is based on an undetermined number of males from an unknown locality in Europe. The only preserved specimen of this species we have been able to locate is a semi-destroyed female collected in Bavaria, labeled as "E. silene, Meig./Allemagne/1155" and deposited in the MNHNP (https://science.mnhn.fr/ institution/mnhn/collection/ed/item/ed4105?listIndex=25&listCount=32, accessed on 15 March 2023). Thus, the type series of this species is apparently lost. Regarding the male description of E. selene, the white maculae on tergum II are smaller than those on tergum III, differing from the new species in which the maculae of tergum II are the widest of the three pairs. able to locate is a semi-destroyed female collected in Bavaria, labeled as "E. silene, Meig./Allemagne/1155" and deposited in the MNHNP (https://science.mnhn.fr/institution/mnhn/collection/ed/item/ed4105?listIndex=25&listCount=32, accessed on 15 March 2023). Thus, the type series of this species is apparently lost. Regarding the male description of E. selene, the white maculae on tergum II are smaller than those on tergum III, differing from the new species in which the maculae of tergum II are the widest of the three pairs.     Séguy, 1961 ( Figures 17 and 18)    Distribution. New. SPAIN • Almería, Sierra de Gádor. Published records. Revised. PORTUGAL • Beira Litoral, Serra de Santo António [54].
Taxonomic notes. This species was described based on a single male---from the Jabron Valley (France)---of which we examined pictures provided by Dr. Martin Hauser. Speight et al. [55] designated E. bayardi as the valid name for Eumerus micans (Fabricius, 1798), described originally under the genus Syrphus Fabricius, 1775. According to article 57.2 of the International Code of Zoological Nomenclature (ICZN) [56], E. micans is a primary homonym of Syrphus micans (Fabricius, 1794), which was synonymized with Volucella inanis (Linnaeus, 1758) afterward [47]. The name Eumerus micans became then invalid, and the species required a new name. Speight et al. [55] indicated that the type material of E. micans is lost and that, based on the original description, no differences were found between E. micans and E. bayardi. Therefore, they synonymized and proposed E. bayardi as a replacement name for E. micans, following article 60.2 of the ICZN [56].
As a result of our integrative analysis, the status of E. bayardi as a member of the E. tricolor species group is confirmed. Based on the examined specimens, E. bayardi is similar to E. azabense but differs in the facial hairs, which are black in E. bayardi and white in E. azabense; hairs on the scutum, which are entirely white in E. bayardi but black on the central area of the scutum in E. azabense ( Figure 15A); hairs on the basoventral side of the metafemur, white in E. bayardi ( Figure 17D) and black in E. azabense; hairs on the abdomen, white in E. bayardi but mainly black in E. azabense; and the general shape of the surstylus in male genitalia (for E. azabense genitalia, see figure 6b of Ricarte et al. [14]).
Taxonomic notes. The type series of this species consists of 10 specimens collected in Spain [15]. According to the species diagnosis, one of the characters to distinguish it from other similar species is the relative length of the metatibia, which is clearly shorter than the metafemur. After the examination of further material, we have found that the length of these leg segments is variable amongst individuals (metatibia = 2.37-3.1 mm, metafemur = 2.4-3.1 mm, n = 10). The intraspecific variation in the length of these leg segments indicates that this should be avoided as a diagnostic character for species identification.
Furthermore, and based on the original description of the long-legged North African Eumerus afrarius Séguy, 1961, the males of E. grallator appear to be very similar morphologically to those of E. afrarius. Based on the original description of E. afrarius, the only differences found between the two species were the eye distance, which is equal to the diameter of an ocellus in E. grallator but twice the diameter of an ocellus in E. afrarius, and the marginal coloration of the scutellum, which is bluish-black in E. grallator but supposedly red in E. afrarius. Grković et al. [15] did not mention E. afrarius in their revision of the E. binominatus subgroup. Eumerus afrarius was described from a single male collected in Massena (Algeria), apparently preserved in the MNHNP (http://www.diptera.org/Nomenclator/Details/47859, accessed on 15 May 2023). No records of this species were found in the online database of the MNHNP, nor was information about the holotype available upon request. A potential synonymy between these two species should be further investigated in the future due to their high morphological similarity and the geographical proximity between the Iberian Peninsula and the North of Africa. Meigen, 1822 ( Figures 4A and 5B) Distribution. Published records. Not revised. SPAIN • Girona, Queralbs [11].

Eumerus grandis
Taxonomic notes. The only known Iberian record of this species is a single female collected in the Pyrenees of Girona, near the type locality of E. ancylostylus sp. n. Since the female of E. ancylostylus sp. n. is unknown and in the absence of molecular data for the above-mentioned female, the record of E. grandis from the Pyrenees of Girona is likely to belong to E. ancylostylus sp. n. Thus, further fieldwork and molecular analyses are necessary to confirm the presence of E. grandis in the Ibero-Balearic area.  [11]. Taxonomic notes. The only known Iberian record of this species is a single female collected in the Pyrenees of Girona, near the type locality of E. ancylostylus sp. n. Since the female of E. ancylostylus sp. n. is unknown and in the absence of molecular data for the above-mentioned female, the record of E. grandis from the Pyrenees of Girona is likely to belong to E. ancylostylus sp. n. Thus, further fieldwork and molecular analyses are necessary to confirm the presence of E. grandis in the Ibero-Balearic area. Genitalia. Epandrium with a simple posterior surstylar lobe, square-shaped and slightly rounded at the top ( Figure 20C); epandrium with a rectangular-shaped, black pilose interior accessory lobe of the surstylus ( Figure 20C,D). Cercus with long pilosity. Hypandrium with two round bulges basally ( Figure 20B); hypandrium with three pointed, hyaline expansions basoventrally ( Figure 20A,B); hypandrium with a small, triangular expansion basodorsally ( Figure 20B). Distribution. Published records. Not revised. SPAIN • Alicante, Agost, Albatera; Almería, Sorbas, Tabernas [24].
Taxonomic notes. This species has been recently described from material collected in south-eastern Spain. The type locality is "SPAIN: Agost//Barranc de Pina" [24]. Taxonomic notes. Gil-Collado [63] reported this species from the localities of Santa Fe (Granada) and San Marsal (Girona). The record from the first locality was previously published by Andréu [62] based on a single male but collected close to a marsh located in Santa Fe del Montseny (Girona), not in Granada. In addition, based on the known records, this species is present in rainy, high mountain habitats, which is not the case for the Santa Fe site in Granada, southern Spain. Therefore, the citation of E. ovatus from Granada should be a misinterpretation by Gil Collado [63] of the locality data provided by Andréu [ [63].
Taxonomic notes. The type series of E. sabulonum, described under the genus Pipiza, consists of eight specimens, and it is preserved in the entomological collection of the 'Naturhistoriska riksmuseet' in Stockholm (catalog numbers NHRS-GULI000070672-79). All these specimens were collected in Sweden. We could not access the type series upon request, but we included in the molecular analyses a barcode sequence of one male specimen (voucher NORSY404-12) collected in Norway, close to the type locality. We considered this specimen as the E. sabulonum sensu stricto to compare with those collected in the Iberian area.

Key to the European Species of the E. tricolor Group
The key presented below is a result of combining and adapting the key to males of the E. binominatus subgroup [15] and the key to the south-eastern European species of the Taxonomic notes. The type series of E. sabulonum, described under the genus Pipiza, consists of eight specimens, and it is preserved in the entomological collection of the 'Naturhistoriska riksmuseet' in Stockholm (catalog numbers NHRS-GULI000070672-79). All these specimens were collected in Sweden. We could not access the type series upon request, but we included in the molecular analyses a barcode sequence of one male specimen (voucher NORSY404-12) collected in Norway, close to the type locality. We considered this specimen as the E. sabulonum sensu stricto to compare with those collected in the Iberian area.

Key to the European Species of the E. tricolor Group
The key presented below is a result of combining and adapting the key to males of the E. binominatus subgroup [15] and the key to the south-eastern European species of the E. tricolor group [11]. We decided to exclude E. lateralis from the key since its taxonomic status remains uncertain until further molecular evidence is available.  Eumerus ancylostylus sp. n. was recorded from two different localities in the Euro-Siberian region of Spain, on north [26]. Although the two localities are far away from each other, both are montane areas in which the observed vegetation was derived from long periods of human use [44,46]. Similarly, E. petrarum sp. n. was collected in a mountainous area (Sierra Nevada) where this species is present from the montane zone to the highest alpine zones. The finding of these undescribed species is an indicator for the Iberian Peninsula to have acted as a climate refuge, in which the mountains were speciation centers for animals and plants during glacial-interglacial events [71,72]. Regarding the other Iberian members of the E. tricolor group, E. sabulonum was recorded for the first time from the Pyrenees of Navarra, which is the northernmost record in the Iberian territory. This species showed the most cosmopolitan distribution pattern within the Iberian Peninsula since it can be found from meridional areas, almost at sea level, to septentrional mountainous zones. This fact is consistent with the remainder of its distribution range as it is present from North Africa to the European areas of Russia [73]. Eumerus grallator was reported from Madrid and Sierra de Gádor (Almería), which is one of the most southern records of this species in the Iberian Peninsula. According to the literature and the examined material, this species is mainly present in montane areas of the Mediterranean Region. On the contrary, based on the Iberian records, E. ovatus appears to be associated with humid and rainy habitats, but it can also be found in dry and open environments where herbaceous and shrubby vegetation are predominant [73]. Nevertheless, the low number of records of E. ovatus, half of them originating from old publications [11,62,63], may indicate that its Iberian populations are declining like in other European areas [74]. Similarly, the first documented record of E. bayardi from Spain is provided on the basis of a single specimen from Almería. The taxonomic status of this species has been uncertain for a long time, as well as its presence in the Ibero-Balearic region [11,18]. Based on the two known records, E. bayardi seems to be present in low-temperature Mediterranean areas. After the present work, the morphological concept of this species is clarified, but more fieldwork is necessary in order to fill the gap of knowledge about its distribution range and biology. Moreover, the first record of E. azabense from Portugal highlights that incomplete knowledge of species distributions in the Ibero-Balearic area represents a limitation for the conservation of those species that are threatened with extinction [75]. This is particularly alarming for newly discovered species with restricted distributions (e.g., E. petrarum sp. n.). Nowadays, 16 European species of the E. tricolor group are under one of the IUCN threatened categories, of which five are present in the Iberian Peninsula. Thus, further studies in unexplored areas of the Iberian Peninsula will assist in getting a better picture of the distribution of the species of the E. tricolor group in this area and, thus, a more accurate view of their conservation status.

Conclusions
(1) A total of 12 species of the E. tricolor group are present in the Ibero-Balearic region. A lectotype is designated for E. lateralis; (2) Two new species, E. ancylostylus sp. n. and E. petrarum sp. n., are described, illustrated and discussed; (3) From a morphological point of view, levels of intraspecific variability were high for both E. sabulonum and E. petrarum sp. n., while the interspecific variability was low between E. grandis and E. ancylostylus sp. n.; This is the first work devoted to improving the knowledge about the diversity, systematics and distribution of the E. tricolor species group in the Iberian area.
Supplementary Materials: The following supporting information can be downloaded at: https: //www.mdpi.com/article/10.3390/insects14060541/s1, Table S1: List of examined material; Table S2: List of species/specimens analyzed in the molecular study. Data Availability Statement: All sequences generated in this work are available in the publicly accessible repository of GenBank (https://www.ncbi.nlm.nih.gov/genbank/).