Morphological, Genetic and Biological Evidences to Understand Meromacrus Rondani Diversity: New Species and Early Stages (Diptera: Syrphidae)

Simple Summary Hoverflies or flower flies, technically known as syrphids, are insects providing important ecosystem services. They are used as indicators of biodiversity and habitat quality, as well as models for evolution studies. The better syrphids are known the better can be used with different aims. Meromacrus is a genus of showy syrphids that pollinate plants and decompose organic materials in the Americas. However, their classification and biology are still being investigated. In this study, morphology and DNA were used in combination to define species concepts. Two species new to science were discovered, one from Mexico and the other from Peru. The immatures (e.g., larvae or pupae) and breeding sites of these species were also described, with the larva of the Peruvian species happening to be the first ever found in a rotting cactus. To assist those working with immatures, we provide here an identification key to Meromacrus species. Our work represents the starting point for a modern revision of the Meromacrus classification and provides data that, in future, can be used to interpret evolutionary relationships within this genus. Abstract Meromacrus is a genus of conspicuous syrphids with saprophagous larvae, ranging from the southern United States to Argentina and Chile. However, this genus is in need of a taxonomic revision. Adults reared from larvae collected in Mexico and Peru, and other material available at different institutional collections were examined. Meromacrus cactorum sp. nov., from Peru, Meromacrus yucatense sp. nov., from Mexico, their puparia and breeding sites were described. A key to Meromacrus puparia is provided. The holotypes of Meromacrus canusium, Meromacrus gloriosus, Meromacrus laconicus and Meromacrus melmoth were also examined. The name Meromacrus draco is proposed as a junior synonym of M. gloriosus. Larvae of M. cactorum sp. nov. were found in decaying columnar cacti in Peru, while those of M. yucatense sp. nov. in a rot-hole of a Ceiba pentandra stump. Obtained results on both taxonomy and biology of these species serve as a first step towards a revision of the entire genus.

Although molecular evidence has proven useful to resolve taxonomic problems in the eristalines e.g. [6,7], current species concepts in Meromacrus are based only on classic morphology of adults and their phylogenetic relationships are unknown. The study of the immature stages of syrphids not only provide an important set of characters on which to build more robust phylogenies [8] but is also critical to understand the species requirements in different ecosystems. The typical long-tailed larvae of Meromacrus syrphids are saprophagous in rot holes of various tree species, water pockets in banana leaf axils [4,9], bromeliads [10] and Heliconia L. flower bracts [11], in decaying banana stems [4,12] and coffee pulp [9]. Larvae of different Meromacrus species can be found coexisting in the same tree hole [4]. The larvae/puparia of only six species are described [9]. Adult flies are frequent flower visitors in natural environments [11], but also seem to play an important role in the pollination of some cultivated plants such as Meromacrus melansoni Blatch in Blatch et al. [4] in mango trees (Anacardiaceae) [4,13].
The aim of the present study is to stablish the bases for a future revision of the taxonomy and phylogeny of the genus Meromacrus. The specific objectives are: (a) to describe new species and address some pending taxonomic problems, (b) to describe new puparia and breeding sites, (c) to update the existing early stage identification key, (d) to explore the molecular support of some species, based on COI (cytochrome c oxidase I mitochondrial gene).

Fieldwork and Rearing of Early Stages
In Yucatan (Mexico), Meromacrus larvae were collected from a rot hole in a Ceiba pentandra (L.) Gaertn. tree (Malvaceae) by Javier Quinto. Larvae were reared by placing them in a plastic cage of 30 × 20 × 8 cm containing part of the original rot-hole substrate (water and woody debris). This cage was placed inside another of 45 × 37 × 24 cm covered by a mesh and filled with a thin layer of dry small woody material from the forest soil of the collection site, to facilitate pupation of larvae. This dry layer was checked daily for puparia. Puparia were marked with a label and after 4 days, when the pupal spiracles protruded entirely, they were individualized in separate small plastic pots. Emerged adults were allowed to die in their pots. The date of pupation and the date of adult emergence were recorded for each specimen. Larvae and puparia were reared under environmental conditions. Larvae were collected on 15 March 2014 and they all pupated between 18 March and 8 April, i.e., a range of 22 days of pupation from the date of the larva finding. The pupal stage lasted 6-31 days and all adults emerged between 25 March and 26 April.
In Trujillo (Peru), larvae were collected in the decaying columnar cacti Espostoa melanostele (Vaupel) Borg (Cactaceae) by M. Ángeles Marcos-García on 17 January 2005, and then reared in plastic cages of 24 × 20 × 34 cm containing part of the original substrate where the larvae were found. Cages were covered with a mesh to allow the entrance of oxygen. They were reared under environmental conditions in shade. Puparia were individualized in Petri dishes until adult emergence. Adults reared from larvae With material sequenced at CNC, raw sequence reads were scored using Sequencher 5.4.6 (2018) and aligned using Mesquite [24]. In some cases, BOLD (Barcode of Life Data System) alignments were also manually checked and corrected using Mesquite. The sequence data obtained are stored online on the BOLD database (www.boldsystems.org). Data are publicly accessible in the Meromacrus revision dataset, available at http://www.boldsystems.org (dx.doi.org/10.5883/DS-SYRMEROM). Sequences obtained for this project are also available on GenBank (https://www.ncbi.nlm.nih.gov/genbank/), with accession numbers listed in Table 2. Neighbour-joining (using the BOLD algorithms) was used to explore species concepts for ingroup taxa. Uncorrected pairwise genetic distances (p-distance) (see Supplementary  Table S1) were calculated with Mega7 [25]. Maximum likelihood, utilizing RAxML v8 [26], was used to create a preliminary phylogenetic hypothesis. The model calculated and used in this analysis was GTR + G + I. Bootstraps were calculated using 1000 replicates. The most likely tree is presented in Supplementary Figure S1. Quichuana calathea Shannon, 1925 and Tigridemyia curvigaster (Macquart, 1842) were used as outgroups for the likelihood analysis.   Etymology. The specific epithet 'cactorum' refers to the cacti, which are the breeding sites of thi species.
Puparium. Shape and size ( Figure 3). Subcylindrical, tapered posteriorly, with a typical eristalin long tail. Light brown. Tegument slightly punctured with spicules. 6 pairs of prolegs on sma • Diagnosis. This new species meets all characters and remarks stated for the genus Meromacrus in Blatch et al. [4], except for its wholly pilose anepimeron and virtually hyaline wing. This species can be separated from other congeneric species by the following combination of characters: general body colouration black; eyes approximated along a very short length, and separated by a distance equal to the diameter of a largest facet (only males); axe-shaped orange antenna, with pedicel longer-sometimes slightly-than basoflagellomere; face with a medial dark brown to black vitta; occiput with yellow tomentose pile on the dorsal 1/3; scutum with an inconspicuous line of yellow tomentose pile along each transverse suture and notopleuron, continued along the posterior margin of posterior anepisternum; posterior margin of scutum with two maculae of sparse yellow tomentose pile; postalar callus with sparse tomentose pile posteriorly; swollen metafemur, as broad as the width of tergum 4; metatibia curved and broad; elongate abdomen, with orange maculae at least in tergum 4; terga 2-4 with a narrow yellow fasciae on the posterior margin; male genitalia as in Figure 2A. Head ( Figure 1A,C). Eye with larger facets near eye contiguity; vertical triangle with dark brown to black pile, except for the short white pile on its anterior corner and the long white pile posterior to ocellar triangle; ocelli ellipsoidal, light brown; ocellar triangle slightly elevated in lateral view, and anterior corner of the vertical triangle not elevated; eyes approximated along a very short length, 4-5 facets long, and separated by a distance equal to the diameter of a largest facet; dark brown to black frontal triangle, with white pile; brown lunule; axe-shaped orange antenna, with black basoflagellomere along its dorso-apical margin; scape and pedicel with white pile; light orange arista; trapezoidal basoflagellomere, shorter than pedicel ( Figure 1D); face with a medial dark brown to black vitta, elsewhere orange and sparsely pollinose, with silver white pile ( Figure 1C); ventral tubercle of face slightly marked but visible; black gena, with two orange maculae-one larger than other-on each eye margin; occiput with yellow tomentose pile on the dorsal 1/3, elsewhere light yellow pilose anteriorly and white pilose posteriorly; occiput sparsely pollinose, black except for the narrowly orange eye margin on the dorsal 1/3. Thorax. Black scutum, black pilose except for the white pile on the anterior margin, and an inconspicuous line of yellow tomentose pile along each transverse suture and notopleuron ( Figure 1B), continued along the posterior margin of posterior anepisternum; postalar callus with long white pile posteriorly intermixed with two or three tomentose pile; posterior margin of scutum with two maculae of sparse yellow tomentose pile ( Figure 1B); scutum with two inconspicuous medial grey-pollinose vittae extending along the anterior 3/4 of scutum length; scutellum brown, blackish laterally, with both short black and long white pile intermixed; extensively black pleuron; posterior anepisternum, katepisternum, anepimeron and metasternum with white to light yellow pile. Wing. Hyaline, extensively microtrichose, with narrow bare areas in cells R and BM basally; stigmal crossvein conspicuous; spurious vein as thick as close veins; orange pilose basicosta and black pilose tegula; calypter white centrally and light brown along the margin, with white pile; light orange halter. Legs. Anterior part of all coxae white pilose; basal part of all femora with a well-defined macula of black setulae antero-ventrally; orange pro-and mesofemora, black dorsally; metafemur orange anteriorly, but black dorsally and posteriorly; white pilose pro-and mesofemora, with some black pile in mesofemur ventrally; white pilose metafemur, with thick black pile on its ventro-posterior margin basally, and its ventro-anterior margin apically (apical part with some longer black pile); swollen metafemur, as broad as the width of tergum 4 ( Figure 1A); tibiae extensively orange, except for the extensively black metatibia (orange apically) ( Figure 1A); all tibiae white pilose, except for a few very short setulae in the mesotibia basally and some black setae at the mesotibia apex; metatibia curved and broad, with a triangular projection posteriorly, at the apex; tarsi orange, except for the black dorsal part of tarsomeres 3-5, all tarsi white to light yellow pilose; claws black apically. Abdomen ( Figure 1E). Elongate; terga black except for the orange lateral maculae in the anterior part of tergum 2, and the lateral margins of terga 3 and 4; dorsum of abdomen metallic, Insects 2020, 11, 791 9 of 30 with greyish blue reflections; terga 2-4 with two inconspicuous maculae of white pollinosity on the anterior margin and a narrow bare yellow fascia on the posterior margin; all terga black pilose, except for the white to light yellow pile on antero-lateral areas of each tergum and lateral margins; pleural membranes orange; sterna extensively orange, with long orange pile. Genitalia. Posterior surstylar lobe broad and roundish, black pilose ( Figure 2A). FEMALE. Range of female sizes (n = 13.5-13.75 mm). Similar to male except for the following characters: frons with a fascia of sparse white pollinosity; frons orange and white pilose on the ventral 3/4; basoflagellomere nearly as long as pedicel; grey pollinose vittae of scutum even less conspicuous than in male; posterior part of postalar callus with yellow tomentose pile connecting with a tomentose fascia on posterior margin of scutum; cells R and BM with bare areas basally; basal part of metafemur without black pile; metatibia without a triangular projection posteriorly, at the apex; terga 2-4 with two maculae of tomentose yellow pile on the anterior margin, united in tergum 2; at least tergum 4 with some orange parts ( Figure 1F).

•
Taxonomic notes. M. cactorum sp. nov. does not key out using Hull (1942) due to its hyaline loop of vein R4 + 5 and the two medial grey-pollinose vittae on scutum. However, this species appears to belong to the group of 'very dark, black or almost black flies' referred to in the couplet 1 of the key in Hull [20]. Within this group, M. cactorum sp. nov. can be readily separated from Meromacrus melmoth Hull, 1937 and Meromacrus pluto Hull, 1942 by the shape of basoflagellomere, which is about as long as wide, blunt apically in M. cactorum sp. nov. (Figure 1D), while elongate, slightly concave dorsally and curved at its pointed apex in M. melmoth and M. pluto [Hull [20]: Figure 13].  [27]] has the metafemora less thickened than M. cactorum sp. nov. and the basoflagellomere broadly rounded, not trapezoidal as in the new species ( Figure 1D). M. cactorum sp. nov. is also similar to Meromacrus brunneus Hull, 1942 due to the general shape of antenna and very thickened metafemur, but M. cactorum sp. nov. has the wing extensively hyaline, and tomentose pile on transverse suture, notopleuron ( Figure 1B      Subcylindrical and slightly curved tubes, dark brown and less shiny than the anterior spiracles, ≈1.9 mm long, more than 6 × longer than broad at the base. Straight, slightly curved at the tip. Surface reticulated, with 14-16 bands of spiracles arranged almost at the base of the tube, absent on the ventral surface. Each band with 8-12 tubercles, each one bearing 5-8 oval spiracular openings. PRP ( Figure 5E,F). Subcylindrical to oval in cross section, ≈167 µm broad near the apical end of the structure. Surface clear and smooth, without any apparent transverse ridge (maybe hidden by the tegument). Spiracular plate domed, with two twisted central scars, two pairs of curved openings and four pairs of feathery interspiracular setae, highly divided and covering the distal perimeter of the PRP.     [4] and it can be separated from other congeneric species by the following combination of characters: antenna orange; basoflagellomere oval, over 1.3 times longer than width (holotype) ( Figure 7C); face with a medial black vitta; scutum with a tear-shaped macula of golden-yellow tomentose pile on the anterior margin, next to each postpronotum, a line of golden-yellow tomentose pile along each transverse suture and notopleuron, continued along the posterior margin of posterior anepisternum and dorsal margin of katepisternum; postalar callus with a tuft of golden-yellow tomentose pile connecting with a semicircular fascia of tomentose pile along the entire posterior margin of scutum; legs extensively orange, with a black carina on the basal 1/3 of metatibiae ventrally; tergum 2 with two lateral slender triangular whitish-yellow markings; tergum 1 with two triangular maculae of golden yellow tomentose pile; terga 3 and 4 with two oval maculae of tomentose pile on the anterior margin of each terga; male genitalia as in Figure 2C. Etymology. The specific epithet 'yucatense' refers to the state of Yucatan (Mexico), where th type locality of this species is found. Puparium. Shape and size. Subcylindrical, tapered posteriorly, with a typical eristaline long tai  Head ( Figure 7A,B). Eye with larger facets near eye contiguity; ocellar triangle slightly elevated in lateral view, with dark brown to black pile progressively longer towards the occiput; ocelli ellipsoidal, light brown; anterior corner of the vertical triangle not elevated in lateral view, slightly white pollinose and with short silver-white pile; eye contiguity 16-17 facets long; dark brown frontal triangle, with black pile, white pollinose and with silver white pile laterally; light brown lunule; orange antenna, slightly darkened in the dorsal part of basoflagellomere ( Figure 7C); scape and pedicel with black setulae of different lengths; light orange arista; oval basoflagellomere, about 1.3 times longer than wide ( Figure 7C); face with a medial black vitta ( Figure 7B), elsewhere white pollinose, with silver white pile; ventral tubercle of face inconspicuous, nearly absent; gena light orange with darker areas; occiput with light-orange tomentose pile, except the area just behind the vertical triangle. Thorax. Scutum black, with brown postpronotum; scutum with a tear-shaped macula of golden-yellow tomentose pile on the anterior margin, next to each postpronotum; scutum with a line of golden-yellow tomentose pile along each transverse suture and notopleuron (inner end of line widened), continued along the posterior margin of posterior anepisternum and dorsal margin of katepisternum ( Figure 7D); postalar callus with a tuft of golden-yellow tomentose pile connecting with a semicircular fascia of pile of the same kind along the entire posterior margin of scutum; scutum with a medial grey-pollinose vitta extending along the anterior 3/4 of scutum length, a fainter grey-pollinose vitta from each tear-shaped tomentose macula to the transverse suture, and an equally faint pollinose macula next to each postalar callus; scutellum brown, darker on the anterior margin, with short black pile all over, except for a line of light brown pile on its posterior margin; posterior anepisternum with golden regular yellow pile, next to the tomentose line; katepisternum with regular yellow pile, longer ventrally; anepimeron with fine yellow pile, and black pile postero-dorsally; metasternum black pilose. Wing. Wholly microtrichose, brown pigmented on the anterior margin, except cell C; brown pigmentation darker apically than basally, and not extending beyond the apical end of cell R2 + 3; stigmal crossvein conspicuous; spurious vein as thick and sclerotised as close veins; orange pilose basicosta and black pilose tegula; calypter white centrally and black along the margin, with light brown pile; white halter. Legs. Extensively orange ( Figure 7A), with a black carina on the basal 1/3 of metatibiae ventrally; anterior part of all coxae with both black and orange pile intermixed; basal part of all femora with a well-defined macula of black setulae antero-ventrally, more anterior than ventral in metafemora; all femora with black pile ventrally, and a bare line apico-ventrally; metafemur with setulae apico-ventrally; dorsal part of all femora with black pile, specially abundant in metafemora apically; tibiae extensively orange pilose, with scattered short black pile; all tarsomeres with at least one or two black pile dorsally, usually extensively black pilose; all tarsi orange pilose ventrally; claws black apically. Abdomen. Terga black except for two lateral slender triangular whitish-yellow markings on tergum 2 ( Figure 7E); all terga with short black pile, except the following parts: tergum 1 with two triangular maculae of golden yellow tomentose pile; terga 3 and 4 with two oval maculae of tomentose pile on the anterior margin of each terga, each macula nearly reaching the midpoint of tergum; regular yellow pile present on the anterior corner of tergum 2 and along the lateral margins of terga 2-4; pleural membranes and sterna black; all sterna with long yellow pile, except for the black pile of sternum 4. Genitalia. Posterior surstylar lobe elongated, straight apically, slightly expanded before the round apex; basal part of surstylus with a triangular expansion that curves inwards; surstylus black pilose all over, with a patch of thicker setae on the inner part; superior lobes of hypandrium anteriorly curved, pointed at apex ( Figure 2C). FEMALE. Unknown.  Figure 6). In addition, these two species differ in the shape of the surstylus, as shown in Figures 2C and 10A of Blatch et al. [4].

•
Etymology. The specific epithet 'yucatense' refers to the state of Yucatan (Mexico), where the type locality of this species is found. • Puparium. Shape and size. Subcylindrical, tapered posteriorly, with a typical eristaline long tail. Brown in colour. Tegument slightly punctured with spicules. 6 pairs of prolegs on small cones, with numerous crochets. 10.9 mm long (10.71-11.2), 5.49 mm high (5.44-5.53) and 6.47 wide (6.24-6.64) (n = 3). Head skeleton (Figure 8). In general, of the filter-feeding type [19], heavily sclerotised only in the area between the dorsal bridge and the tentorial arm. Dorsal cornu shorter than ventral cornu. In profile view, dorsal bridge area in obtuse angle. Mandible without hooks. Anterior spiracles ( Figure 9A). Straight structures, light brown and shiny, striated surface along the tube, 3× longer than broad at the base, slightly curved at the end. Numerous respiratory openings on a plate at the ventral tip of the tube. Surface of the plate reticulated and smother than the rest of the entire structure, ridges concentrically arranged around the spiracular openings.   • Biology and habitat. Larvae were found in a traditional henequen (Agave fourcroydes Lem., Asparagaceae) hacienda. By the 1850s, the henequen industry collapsed and natural vegetation colonized large areas cultivated with henequen. The studied hacienda is now embedded in a heterogeneous landscape matrix, including remnants of tropical secondary dry forest with large old trees, and agriculture and livestock areas in which crop rotations and different types of management take place.

•
Larvae were collected in a single Ceiba pentandra stump with a large water-filled tree hole containing abundant wood decay ( Figure 10). This stump was the result of a recent pruning at ground level of an old tree (the margins were burned to prevent regrowth), exposing the hole that the trunk had inside. • Larvae were collected in a single Ceiba pentandra stump with a large water-filled tree hole containing abundant wood decay ( Figure 10). This stump was the result of a recent pruning at ground level of an old tree (the margins were burned to prevent regrowth), exposing the hole that the trunk had inside. Larvae of three Meromacrus species, M. gloriosus, M. laconicus and M. yucatense sp. nov., were found coexisting in the same hole. All the larvae of M. yucatense sp. nov.

A New Synonymy in the Genus Meromacrus
Meromacrus gloriosus Hull, 1941 Meromacrus draco Hull, 1942 syn. nov. Figure 11 Insects 2020, 11, x 19 of 30 a T-shaped black macula on the anterior margin of tergum 2 (as M. draco), but yellow pile on terga 3 and 4 were nearly as abundant as in the holotype of M. gloriosus. The other female from Arizona was similar to the male holotype of M. gloriosus in having the tegula wholly yellow pilose and the metafemur extensively yellow pilose dorsally, anteriorly and posteriorly. However, the metafemur was black centrally and the tergum 2 had a T-shaped black macula on the anterior margin, as in M. draco. The holotype of M. gloriosus and both examined females had conspicuous yellow tomentose maculae on the anterior margin of terga 3 and 4, consistent with a specimen of M. draco from Hidalgo, Mexico. All this variation in the stated characteristics (otherwise, all examined specimens of M. draco and M. gloriosus were similar) seem to support the existence of a single variable taxon, as shown by the COI study of two gloriosus-like specimens from USA (106257: female; 106256), two draco-like specimens (UA1ME: male; UA2ME: female) from Yucatan, Mexico and a female from Costa Rica (UA5ME) plus six specimens from Costa Rica identified as M. draco in BOLD systems; all 11 specimens analysed had very similar COI sequences (see Section 3.4.4). On the basis of this evidence, both morphological and molecular, we propose M. draco as junior synonym of M. gloriosus.   [4], who also examined the male holotype of M. draco at the American Museum of Natural History. In our specimens, the female frons is brown to black on the posterior half to two thirds; tegula with black pile anteriorly (at least one or two); basicosta orange pilose; metafemur black centrally along a variable length (usually narrowly orange basally and on the apical fourth) and black pilose except for the yellow pile dorsally on baso-anterior half; tergum 2 with two lateral orange maculae of variable extension, with a T-shaped black macula on the anterior margin or a H-shaped black maculae extending from the anterior to the posterior margin; terga 3-5 from black to reddish black; tergum 3 with two maculae of tomentose pile on the anterior margin; tergum 4 with two smaller tomentose maculae on the anterior margin, usually inconspicuous, sometimes virtually absent; terga 3 and 4 extensively short black pilose; sterna black to brownish black. The examined males shared the same genitalia (see Figure 9A-C in Blatch et al. [4]).
The male holotype of Meromacrus gloriosus (Figure 11), from the USA (New Mexico), is in general lighter than the Costa Rican and Mexican specimens, and differs from them in the following characters: tegula wholly yellow pilose; metafemur wholly orange and yellow pilose, just with black setulose pile ventrally; tergum 2 almost wholly orange ( Figure 11A); terga 3 and 4 with more abundant short yellow pile, extending towards the central parts of terga; sternum 1 yellow posteriorly; sternum 2 yellow, with a central black macula. However, the holotype has the same genitalia morphology as the specimens examined from Costa Rica and Mexico. We also examined two females from the USA (Texas and Arizona, respectively). According to the key in Hull [20] these two females would not key out further than couplet 26. The female from Texas had the metafemur wholly orange but the short black pile were more abundant on its apico-posterior third than in the holotype. In addition, this female had black pile on the tegula and a T-shaped black macula on the anterior margin of tergum 2 (as M. draco), but yellow pile on terga 3 and 4 were nearly as abundant as in the holotype of M. gloriosus. The other female from Arizona was similar to the male holotype of M. gloriosus in having the tegula wholly yellow pilose and the metafemur extensively yellow pilose dorsally, anteriorly and posteriorly. However, the metafemur was black centrally and the tergum 2 had a T-shaped black macula on the anterior margin, as in M. draco. The holotype of M. gloriosus and both examined females had conspicuous yellow tomentose maculae on the anterior margin of terga 3 and 4, consistent with a specimen of M. draco from Hidalgo, Mexico.
All this variation in the stated characteristics (otherwise, all examined specimens of M. draco and M. gloriosus were similar) seem to support the existence of a single variable taxon, as shown by the COI study of two gloriosus-like specimens from USA (106257: female; 106256), two draco-like specimens (UA1ME: male; UA2ME: female) from Yucatan, Mexico and a female from Costa Rica (UA5ME) plus six specimens from Costa Rica identified as M. draco in BOLD systems; all 11 specimens analysed had very similar COI sequences (see Section 3.4.4). On the basis of this evidence, both morphological and molecular, we propose M. draco as junior synonym of M. gloriosus.  . Specimen in poor condition, apparently disturbed by a liquid, headless, without right wing, left pro-and metatibiae, left pro-and metatarsus, and right legs except for the mesofemur; meso-and metafemora partly eaten by Anthrenus.

•
Taxonomic notes. This species was described from a female of 'Cape of Good Hope' (Africa) under the genus Milesia (Walker, 1849). In his revision of the genus Meromacrus, Hull [20] redescribed the holotype and addressed the supposed type locality error on the basis of the exclusively Neotropical distribution of the known species of this genus. Blatch et al. [4] also examined the holotype of M. canusium and redescribed the species, stating that the tegula is orange pilose, metafemur extensively black, terga 2-4 with black vittae, terga 3-4 with a small yellow tomentose fasciate macula on the anterior margin, tergum 4 black pilose apico-medially and all sterna brown coloured. Neither Walker [29] nor Hull [4] addressed the tomentose macula on the anterior margin of terga 3-4. We examined the holotype and, even though the specimen is in poor condition, a close examination of it reveals no maculae of tomentose pile on these terga ( Figure 12A,B); in addition, the tegula is black pilose, metafemur, terga 2-4, tergum 4 pile and all sterna wholly orange. We did not find specimens with the same combination of characters as the holotype of M. canusium and possibly neither did Blatch et al. [4], who apparently considered the differences with the holotype as to be intraspecific variability. This species is most similar to M. draco sensu Blatch et al. [4], from which these authors distinguished it by the mainly orange abdomen (mainly dark brown to black in M. draco) and in the female, the wholly orange frons (brown in dorsal 2/3 in M. draco). Apart from the holotype, Blatch et al. [4] only found a male and two females of the putative M. canusium, while 19 males and 21 females fit their M. draco concept.
The holotype of M. canusium is also similar to that of M. gloriosus. However, the holotype of M. canusium has the tegula wholly black pilose, terga 2-5 ( Figure 12A,B) and all sterna wholly orange and tergum 4 wholly yellow pilose, while in M. gloriosus the tegula is yellow pilose (at most with sparse black pile anteriorly), terga and sterna are partly black or reddish black and tergum 4 with extensive areas covered in black pile. It might be that the holotype of M. canusium is an extreme variant of M. gloriosus but given the uncertain origin of the M. canusium holotype and its apparently unique combination of characters, we maintain this species as valid until morphological and molecular analyses of new holotype-like specimens are undertaken.     (Figure 13), including its genitalia in comparison with our specimens. In the holotype, the legs are darkened ( Figure 13A,B) but in our specimens the legs are red, somewhat black apically in the femora. The shape of the cercus and surstylus in the holotype differs slightly from that found in the Mexican and Costa Rican specimens. In addition, the cercus and surstylus in the Mexican and Costa Rican material also displays certain variability in shape: the cercus can be round to trapezoidal, and the surstylus can be narrower or wider apically or even wedge-shaped apically. We consider this as intraspecific variability, since a specimen with holotype-like cercus (UA7ME) and two other specimens with different cercus shape (UA8ME and UA10ME) were shown to be conspecific in the COI gene tree (see Section 3.4.4).
ects 2020, 11, x 23 of   Taxonomic notes. Distinctive species due to the shape of basoflagellomere ( Figure 14B), which is wider than long, and the male genitalia ( Figure 2B). The CEUA specimen did not yield a genetic sequence, but another specimen from CNC did, and shows that this species clearly differs in COI from the other analysed species (Figure 15).

Gene Trees and Pairwise Comparisons
Uncorrected pairwise genetic distances are presented in a table as Supplementary Material. Intraspecific variation ranged from 0.00 to 1.67% and averaged 0.25%. Interspecific variation ranged from 0.61% to 14.81% and averaged 8.76%. A neighbour joining tree is shown in Figure 15 and was used along with pairwise distances to explore taxonomic issues. Hypothesized relationships between Meromacrus species are illustrated on a maximum-likelihood gene tree (Supplementary Figure S1). Both trees show the two new species as independent clades ( Figure 15  Uncorrected pairwise genetic distances are presented in a table as Supplementary Material. Intraspecific variation ranged from 0.00 to 1.67% and averaged 0.25%. Interspecific variation ranged from 0.61% to 14.81% and averaged 8.76%. A neighbour joining tree is shown in Figure 15 and was

Discussion
After this study, which represents the first step towards a revision of the genus Meromacrus (which is being carried out by the authors of the present work), the number of valid species in this genus is 44. Two morphologically distinct species were described, M. cactorum sp. nov. and M. yucantense sp. nov., and M. draco was proposed as junior synonym of M. gloriosus on the basis of morphological and COI evidence. The status of two species genetically analysed in this paper (specimens CNC464847 and INBIOCRI001204119) is still pending of confirmation ( Figure 15).
Adults and puparia of M. cactorum sp. nov. and M. yucatense sp. nov. differ considerably in morphology. Conspicuous differences can be found in the antenna shape ( Figures 1D and 7C), length of eye contiguity ( Figures 1C and 7B), size and density of yellow tomentose pile in thorax and abdomen ( Figures 1B,E,F and 7D,E), metafemur size ( Figures 1A and 7A), shape of male genitalia (Figure 2A,C), etc. Hull [20] grouped the Meromacrus species he studied according to the presence or absence of conspicuous markings of yellow to brownish tomentum. The two new species each agree

Discussion
After this study, which represents the first step towards a revision of the genus Meromacrus (which is being carried out by the authors of the present work), the number of valid species in this genus is 44. Two morphologically distinct species were described, M. cactorum sp. nov. and M. yucantense sp. nov., and M. draco was proposed as junior synonym of M. gloriosus on the basis of morphological and COI evidence. The status of two species genetically analysed in this paper (specimens CNC464847 and INBIOCRI001204119) is still pending of confirmation ( Figure 15).
Adults and puparia of M. cactorum sp. nov. and M. yucatense sp. nov. differ considerably in morphology. Conspicuous differences can be found in the antenna shape ( Figures 1D and 7C), length of eye contiguity ( Figures 1C and 7B), size and density of yellow tomentose pile in thorax and abdomen ( Figure 1B,E,F and Figure 7D,E), metafemur size ( Figures 1A and 7A), shape of male genitalia (Figure 2A,C), etc. Hull [20] grouped the Meromacrus species he studied according to the presence or absence of conspicuous markings of yellow to brownish tomentum. The two new species each agree with the characters of these two morphological groups (see new species descriptions). Hull [20] recognises that the dark or black-bodied species have brown pigmentation inside the loop of the vein R4+5, but our dark species, M. cactorum sp. nov., has the wing hyaline. We interpret that the hyaline loop is the natural state of this character in M. cactorum sp. nov., even though all studied specimens were reared. The two suggested morphological groups of Hull [20] appear to have no COI support. For example, M. cactorum sp. nov., with very slight markings of tomentum, clusters together in the ML tree with M. laconicus and M. yucatense sp. nov. (Supplementary Figure S1), both with obvious markings of tomentum. Further molecular markers and species should be analysed to test the phylogenetic significance of Hull's putative groups.
Meromacrus loewi and M. zonatus are the closest taxa genetically (0.61-0.98% different) but cluster separately on the tree ( Figure 15) and are morphologically distinctive. Meromacrus zonatus has golden-yellow tomentose pile on the head (occiput and frons), while M. loewi has not. In addition, M. loewi has less golden-yellow tomentose pile on the scutum, making the white-pollinose stripes of the scutum more visible than those of M. zonatus. Meromacrus acutus and M. gloriosus are also genetically close (0.47 to 2.29% different) but cluster separately ( Figure 15) and are morphologically distinctive [9,20]. Meromacrus draco and M. gloriosus are interdigitated on the NJ tree ( Figure 15) and have 0.00-1.57% pairwise differences; this supports the morphological justification presented above that these taxa are synonymous. Similarly, the single specimen of M. panamensis is nested within M. laconicus ( Figure 15) and differs from them by 0.00-0.46%; this supports the morphological decision to synonymize these species made by Blatch et al. [4]. All other species have significant barcode gaps of more than 3% (see Supplementary material). The two new species, M. cactorum sp. nov. and M. yucatense sp. nov., are closely related to M. laconicus based on COI evidence (Supplementary Figure S1). Although Meromacrus is not monophyletic based on this analysis (M. cingulatus falls between the outgroup taxa, Supplementary Figure S1), a more comprehensive analysis using multiple markers is needed to confirm or refuse this preliminary result. Nonetheless, and suggesting the possible non-monophyly, M. cingulatus morphology differs from that of all other species represented in the COI-based trees, for example in having several yellow fasciae in the terga 3-4.
The puparia of M. yucatense sp. nov. and M. cactorum sp. nov. key out to Meromacrus in the key to the genera of Neotropical long-tailed syrphid larvae of Pérez-Bañón et al. [9], and they also have the shared characters stated for the Meromacrus species examined by these authors. Meromacrus yucatense sp. nov. anterior spiracles ( Figure 9A) are clearly similar to those of the other known Meromacrus puparia, with all the respiratory openings arranged on a flat plate in the spiracle ventral surface. However, M. cactorum sp. nov. puparia have the respiratory openings of the anterior spiracles differently arranged, with paired openings on slightly protruding areas along the ventral curved spiracle surface ( Figure 5A). In addition, the number of respiratory openings is clearly lower in M. cactorum sp. nov. than in other species, since M. cactorum sp. nov. has up to 10 openings while the known puparia of other Meromacrus species have at least double number of openings [9]. A higher number of respiratory openings might be an adaptation to live in aquatic media where the concentration of-diluted-oxygen is lower than in sites more exposed to the aerial media, such as the decaying cactus where M. cactorum sp. nov. was found. In the same way, the different characters found on the head skeletons of the two new species might be regarded as an indicator of their feeding media. Meromacrus yucatense sp. nov. does not have mandibular hooks while M. cactorum sp. nov. may use its mandibular hooks and its more sclerotised head skeleton to grasp firmer materials to obtain food rather than only filtering the fluid media as M. yucatense sp. nov.
Pupal spiracles of the two new species also look quite similar to those described previously of other species. Meromacrus cactorum sp. nov. and M. yucatense sp. nov. tubercle bands do not reach the ventral surface ( Figures 5D and 9D Figure 9D), which makes M. yucatense sp. nov. easily distinguishable. In addition, M. yucatense sp. nov. has small and scarce setae covering the surface of the bands of the pupal spiracles ( Figure 9B), similarly to M. draco but in higher number than this species. This may be another evidence of the close relationship between the genera Meromacrus and Habromyia, as already indicated by Pérez-Bañón et al. [9]. Habromyia coerulithorax Williston, 1888 has a higher number of pupal spiracular setae than any described species of Meromacrus. However, Meromacrus yucatense sp. nov., in addition to M. gloriosus, seems to be another morphological intermediate between the presence and the absence of pupal spiracular setae.
The images presented in this paper show some significant differences in the PRP morphology of the two new species ( Figure 5E,F and Figure 9E,F). The PRP of M. cactorum sp. nov. and M. yucatense sp. nov. are very different, especially in their shape and interspiracular setae. M. cactorum sp. nov. has subcylindrical to oval shaped PRP in cross section near its apical end, while M. yucatense sp. nov. has a strongly flattened oval-shaped perimeter along the entire PRP tube. Apart from this, the interspiracular setae along the perimeter of the spiracular plate show different forms. Those of M. cactorum sp. nov. are fan-looking and multibranched, but M. yucatense sp. nov. interspiracular setae are pectinate and have two morphotypes, one with two branches and the other uniramous. Further research on PRP SEM images might provide additional characters to separate larvae/puparia of other Meromacrus species such as those described in Pérez-Bañón et al. [9], who described only the anterior and/or pupal spiracles.
The number of Meromacrus species for which their early stages are known increases now to eight, i.e., 18% of described species in this genus. Our still-poor knowledge of Meromacrus larval biology involves a reasonably wide range of plants and breeding sites (see Introduction and Results) that suffice to anticipate an evolutionary history hypothesis strongly supported in the adaptation of larvae to novel breeding sites/plants, as in Copestylum  Figure S1), suggesting that larval ecology is relatively consistent throughout the genus. Larvae of M. yucatense sp. nov. were found in the water-filled hole of a stump. Most known larvae of Meromacrus are also found in different kinds of water holes or pockets above ground level. However, larvae of M. cactorum sp. nov. were found in decaying parts of cacti. These are the first larvae of Meromacrus ever found in cacti, but not in decaying plant materials, since other larvae of this genus have been reported from banana stems and coffee pulp [9,12]; nonetheless, the species thought to be associated with banana stems (M. gloriosus, as M. draco) was actually collected from aground cavity filled with mud and incidentally containing a banana plant stem inside which its larva was found [4] and might not be then genuinely associated to banana stems. The putative high adaptability of Meromacrus larvae to breeding sites in different plants, together with this paper findings (two new species found as larvae in two sporadic sampling events) suggest the high number of Meromacrus species awaiting discovery in the New World. The findings as larvae of M. cactorum sp. nov. and M. yucatense sp. nov. also reinforce the idea of early stage sampling as an important method to inventory biodiversity and find out species requirements in the Neotropical ecosystems.

Conclusions
Morphological, genetic and biological evidence was combined from specimens deposited in different collections to shed light on the systematics of the New World genus Meromacrus. According to the objectives of the present study, conclusions are as follows: (a) Meromacrus taxonomy was partly revised, with two species new to science (M. cactorum sp. nov. and M. yucatense sp. nov.) and M. draco being synonymised under M. gloriosus. The male genitalia of M. ruficrus was figured for the first time to facilitate its unequivocal identification based on genitalia characters.
(b) The two new species were reared from saprophagous larvae collected in rot-holes (M. yucatense sp. nov.) and decaying cacti (M. cactorum sp. nov.), representing the first Meromacrus larva ever found in cacti. Larvae of M. cactorum sp. nov. appear to have specific morphological adaptations to their breeding site and substrate, while those of M. yucatense sp. nov. have a morphology most similar to that of other Meromacrus species. (c) The existing identification key to Meromacrus puparia was further completed with the addition of the two new species' puparia. With these additions, the utility of this key increases and diversity surveys based on early stages become even more feasible than prior to this study. (d) A NJ tree-with 16 named and unnamed taxa putatively assigned to the genus Meromacruscompiling all COI data available to authors of the present paper was produced to show how the new species clearly diverge from other named species and to support the proposed synonymy.
In summary, this study becomes the first step towards a taxonomic, biological and phylogenetic revision of the genus Meromacrus, in such a way these flies can be used in future as bioindicators and models of adaptive radiations.