The world woodlouse flies (Diptera, Rhinophoridae)

Abstract The world Rhinophoridae are catalogued, recognising 33 genera and 177 species. Nomenclatural information is provided for all genus-group and species-group names, including lists of synonyms and name-bearing type data. Species distributions are recorded by country. A key to the world genera is presented. Four new genera are erected to accommodate five new species, which do not fit within any of the current generic concepts in Rhinophoridae, according to the results of a morphology-based phylogenetic analysis: Marshallicona Cerretti & Pape with type species Marshallicona quitu Cerretti & Pape, gen. et sp. nov. (Ecuador); Maurhinophora Cerretti & Pape with type species Maurhinophora indoceanica Cerretti & Pape, gen. et sp. nov. (Mauritius); Neotarsina Cerretti & Pape with type species Neotarsina caraibica Cerretti & Pape, gen. et sp. nov. (Trinidad and Tobago) and Neotarsina andina Cerretti & Pape, sp. nov. (Peru); Kinabalumyia Cerretti & Pape with type species Kinabalumyia pinax Cerretti & Pape, gen. et sp. nov. (Malaysia, Sabah). The genus Aporeomyia Pape & Shima (type species Aporeomyia antennalis Pape & Shima), originally assigned to Tachinidae, is here reassigned to Rhinophoridae based on a reassessment of the homologies of the male terminalia. The following five species-group names, which were previously treated as junior synonyms or nomina dubia, are recognised as valid species names: Acompomintho caucasica (Villeneuve, 1908), stat. rev. [from nomen dubium to valid species]; Acompomintho sinensis (Villeneuve, 1936), stat. rev. [from nomen dubium to valid species]; Stevenia bertei (Rondani, 1865), stat. rev. [from nomen dubium to valid species]; Stevenia sardoa Villeneuve, 1920, stat. rev. [from junior synonym of Rhinophora deceptoria Loew, 1847 to valid species]; Stevenia subalbida (Villeneuve, 1911), stat. rev. [from junior synonym of Rhinophora deceptoria Loew, 1847 to valid species]. Reversal of precedence is invoked for the following case of subjective synonymy to promote stability in nomenclature: Rhinophora lepida (Meigen, 1824), nomen protectum, and Musca parcus Harris, 1780: 144, nomen oblitum. New generic and specific synonymies are proposed for the following two names: Mimodexia Rohdendorf, 1935, junior synonym of Tromodesia Rondani, 1856, syn. nov. and Ptilocheta tacchetti Rondani, 1865, junior synonym of Stevenia obscuripennis (Loew, 1847), syn. nov. The following new combinations are proposed: Acompomintho sinensis (Villeneuve, 1936), comb. nov. [transferred from Tricogena Robineau-Desvoidy, 1830]; Tromodesia guzari (Rohdendorf, 1935), comb. nov. [transferred from Mimodexia Rohdendorf, 1935]; Tromodesia intermedia (Rohdendorf, 1935), comb. nov. [transferred from Mimodexia Rohdendorf, 1935]; Tromodesia lindneriana (Rohdendorf, 1961), comb. nov. [transferred from Mimodexia Rohdendorf, 1935]; Tromodesia magnifica (Rohdendorf, 1935), comb. nov. [transferred from Mimodexia Rohdendorf, 1935]; Tromodesia obscurior (Rohdendorf, 1935), comb. nov. [transferred from Mimodexia Rohdendorf, 1935]; Tromodesia pallidissima (Rohdendorf, 1935), comb. nov. [transferred from Mimodexia Rohdendorf, 1935]; Tromodesia setiventris (Rohdendorf, 1935), comb. nov. [transferred from Mimodexia Rohdendorf, 1935] and Tromodesia shachrudi (Rohdendorf, 1935), comb. nov. [transferred from Mimodexia Rohdendorf, 1935].


Introduction
Rhinophoridae are a small oestroid family with 33 genera and 177 species, recognised as of the present catalogue. The family was earlier considered by several authors to be entirely of Old World distribution and including very few native species outside the Palaearctic Region. This notion was at least partly due to the then superficially sampled tropical and southern hemisphere subtropical faunas, which are continuously revealing new taxa (Crosskey 1977, Colless 1994, Pape 1998a, Pape and Arnaud 2001, Cerretti and Pape 2012, Cerretti et al. 2014a. Adult Rhinophoridae present no unique autapomorphies ( Fig. 1), thus they are difficult to key out in conventional family-level keys (e.g., Crosskey 1977, Marshall 2012. However, their larvae are highly specialised woodlouse parasitoids and, to the extent they are known, provide unambiguous evidence for family affiliation and monophyly (Bedding 1973, Pape and Arnaud 2001, Cerretti et al. 2014a). The woodlouse-parasitising habit is unique not only within Diptera but within all Insecta. The distribution of Rhinophoridae largely matches that of woodlice, and the peak of rhinophorid diversity and abundance appears to be the Turano-Mediterranean area of the western Palaearctic (see below).
The family Rhinophoridae is a member of the Oestroidea, but its phylogenetic position within this clade is still unresolved (Pape 1992, Rognes 1997, Pape and Arnaud 2001, Kutty et al. 2010, Winkler et al. 2015, Cerretti et al. 2017, Kutty et al. 2019. The most recent analyses based on morphology indicate a sister-group relationship to Rhiniidae (Pape and Arnaud 2001) or to the monotypic New Zealand family Mystacinobiidae, with Rhiniidae subordinate to these (Cerretti et al. 2017), which find most arguments from first instar larval morphology, which is very superficially known for the latter group (Ferrar 1987, Rognes 2002. Interestingly, the most recent analyses based on molecular data retrieve Rhiniidae as sister to Bengaliinae (Cerretti et al. 2017, Kutty et al. 2019, and the extensive transcriptome data of Kutty et al. (2019) point to a phylogenetic grouping of Rhinophoridae with the macrolarviparous Helicoboscinae and Ameniinae, which is unexpected from a morphological as well as a biological point of view. The phylogenetic position of Rhinophoridae within Oestroidea remains ambiguous. Peris and González-Mora (2007) compiled the first world catalogue for the family Rhinophoridae, but with insufficient distributional data and without data on type localities and types. Taking into account the substantial number of genera and species recently described from outside the Palaearctic, which have not been included in any regional catalogue, the significant recent exclusions (i.e., Alvamaja chlorometallica Rognes and the five species of the Phyto carinata species group now assigned to the genus Morinia Robineau-Desvoidy, see Cerretti et al. 2019) as well as a new inclusion (Aporeomyia Shima and Pape, as proposed herein) of taxa, we consider it both helpful and timely to provide a fully updated world catalogue of the entire family. We also take this opportunity to describe four new genera to accommodate five new species, produce a key to all genera, and perform a genus-level phylogenetic analysis of the Rhinophoridae in order to support the newly proposed genera in a cladistic framework.

Systematics
Taxa, specimens All genera were studied, based on an extensive representation of the included species, in order to construct a key to the genera of the world and perform a genus-level phylogenetic analysis. Dissections of male terminalia were performed according to the procedure described by Cerretti and Shima (2011). Briefly, this procedure involves the removal of the abdomen, partial clearing in 10% KOH, dissection and rinsing of terminalia, reattachment of the abdomen to the specimen, and final storage of the terminalia in a glycerine-filled microvial pinned with the source specimen. Figures 2 and 3 summarise the morphological terminology of the adults (except terminalia) and the measurements applied in the present work. Morphological terminology of terminalia (both male and female) and preimaginal instars follow Cumming and Wood (2017) and Pape and Arnaud (2001), respectively.
Data from each type specimen are given verbatim, with information for each line separated by a slash ( /) ; in cases with more than one label, these are separated with a double slash ( //). Additional information of relevance, but not appearing on the label(s), is given in brackets. Finally, the acronym of the repository is cited in parentheses.

Photography, SEM
Photographic images of habitus, head, wing, legs, abdomen and male terminalia were produced using a MZ 12.5 stereoscopic microscope (Leica, Germany) equipped with a DS-L1 Nikon digital camera (Nikon, Tokyo). Photographic images of terminalia were produced using a DM LS microscope (Leica, Germany) equipped with the camera described above. Focus stacking with the image stacking software CombineZM (Hadley, UK) was used to merge 15-45 photos of each specimen/structure, taken at different focal planes, into high-resolution images. Additional images were produced with a Hitachi TM1000 environmental scanning electron microscope (ESEM) from uncoated pinned specimens, using Adobe Photoshop for colouration of specific structural details.

Phylogenetic analysis
We adapted the data matrix of Cerretti et al. (2014a) to assess the most probable phylogenetic affinities and associated support of the newly proposed genera. The matrix consists of 99 morphological characters: 82 from adult stage (1-52: body of both sexes, except terminalia; 53-81: male terminalia; 82: female terminalia), 17 from first instar larva (Table 1); rhinophorid diversity is represented by a sample of 57 species (comprising at least one species for each genus), plus Musca spp. (Muscidae) and a selection of oestroid taxa as outgroups (Table 2; the nexus format of the data matrix is available from MorphoBank Project 3576). In the case of Kinabalumyia pinax gen. et sp. nov., described in the present paper based on two males from Sabah (Malaysia), female characters were scored from a female from Palawan (Philippines), which is here considered congeneric but is left unidentified (Kinabalumyia sp. 1, see below); the two sets of characters were concatenated and treated as a single operational taxonomic unit (OTU) in all our analyses, which is shown in the tree under the name of Kinabalumyia pinax for practical reasons.
The data matrix was produced in Mesquite version 3.03 (Maddison and Maddison 2015). Inapplicable or unknown states were coded as '-' and '?', respectively (Table 3). Cladistic analyses were conducted with TNT version 1.5 (Goloboff and Catalano 2016). Heuristic searches were run under equal weights and under implied weighting (k-values: 3-10, 15, 20), with the 'traditional search' option under the following settings: General RAM of 1 GB, memory set to hold 1,000,000 trees, setting 1000 replicates with tree bisection-reconnection (TBR) branch swapping and saving 1000 trees per replicate. Multistate characters were treated as unordered. Character state changes (apomorphies) were optimised in WinClada version 1.00.08 (Nixon 2002) on the fittest tree obtained with a k-value of 4, using the unambiguous transformation algorithm. We chose the term "global apomorphies" for the uncontradicted and unreversed apomorphic character states, whereas "local apomorphies" was used for the homoplasious character states due to convergence or reversal.

Format
The present catalogue lists all nominal genera and species of Rhinophoridae, providing details about name-bearing types and with known distributions updated from both recent literature and our own identifications of museum specimens.
Valid taxa are arranged hierarchically and alphabetically, according to genus and species (subfamilial and tribal classification is considered premature given the difficulties in interpreting adult homologies and defining monophyletic groupings). Synonyms, including unjustified emendations and incorrect original and subsequent spellings, are listed chronologically for all names.

ADULT (body parts, except terminalia) 1
Each genus-group name is listed with the following formatting and information: genus name (in square brackets if unavailable, italics if available, bold + italics if valid), author, year, page, type species with author and date, form of type fixation with author and date. Each type species is given in its original binomen (Recommendation 67B of the "International Code of Zoological Nomenclature", henceforth "the Code", ICZN 1999), followed by its valid name, if different, in square brackets. Incorrect original spellings are given teste their First Reviser. Incorrect subsequent spellings encountered during this study are cited from their earliest occurrence.
Species are listed alphabetically by valid name followed by synonyms, nomina nuda, unjustified emendations and incorrect spellings listed chronologically. The genus Bezzimyia Townsend is likely polyphyletic and the species have been grouped into two species groups (namely Group A and Group B) and listed alphabetically within each group. The valid specific epithet is given in bold and italics followed by author and year. Each available name is given in italics in its original combination and spelling followed by author, year (with letter if applicable, to match with References), and page. Given next is the type locality in modern spelling, followed by information about the namebearing type, consisting of status (holotype, lectotype, neotype or syntypes), sex, and acronym of type repository. Additional information may be given under "Remarks". Distribution is given hierarchically and alphabetically according to biogeographical region and by country, but with larger countries separated into states/provinces and offshore islands listed separately from the mainland. Archipelagos may be listed by island when data are available. European distribution follows Fauna Europaea (https:// fauna-eu.org/, see Pape et al. 2015).
Type localities are cited from largest to smallest geographic area or place. Country and state/province names are given only in their modern equivalents. Coordinates given in an original publication are cited as an integral part of the type locality, in their original format.
For data on the number and sex of name-bearing types other than an unambiguous fixation of a holo-, lecto-or neotype, we follow the format proposed by O'Hara and , slightly modified as explained below.
Type(s), male: One or more males. This citation is used for a species described from the male sex without indication whether the type series comprised a single male (i.e., a holotype) or more than one male (i.e., syntypes). Type(s), female: One or more females. Type(s), unspecified sex: One or more specimens with no indication of sex. Syntypes, [number] male[s] and [number] female[s] (e.g., "3 males and 2 females"): Species described from both sexes, with the exact number of males and females specified and without a designated holotype. Syntypes, males and females: Species described from both sexes, with more than one specimen of each sex but without specified numbers and without a designated holotype. Syntypes, male(s) and female(s): Species described from both sexes, with no indication of the number of specimens of either sex, neither the exact number nor whether only one or more than one. Syntypes, males: Species described from more than one male, without indication of the specific number of males and without a designated holotype. Syntypes, females: Species described from more than one female, without indication of the specific number of females and without a designated holotype. Syntypes, unspecified number and sex: Species described from more than one specimen but without indication of sex or number of specimens and without a designated holotype.

Avoidance of assumption of holotype and lectotypifications
Recommendation 73F of the Code (ICZN 1999), "Avoidance of assumption of holotype", recommends, "where appropriate" and when it is possible that the nominal species-group taxon was based on more than one specimen, to designate a lectotype rather than assume a holotype; and Article 74.6 of the Code deems that an assumed monotypy where the original description neither implies nor requires that there were syntypes, is deemed to be a lectotype designation if it is considered subsequently that the original description was based on more than one specimen. We follow O'Hara et al. (2009) andO'Hara and in using the term "lectotype designation" for an explicit lectotypification where the author used either the term "lectotype" or an exact translation or equivalent expression (e.g., "the type"), and the term "lectotype fixation" for an implicit lectotypification by inference of holotype as well as for cases where the original work reveals that the taxon had been based on more than one specimen none of which were designated as holotype and an author subsequently used the term "holotype" in a way that explicitly indicated that he or she was selecting from the type series that particular specimen to serve as the name-bearing type.  (Linnaeus, 1758) Italy [MZUR] Rognes (1991) Rognes (1997) Calliphoridae/Helicoboscinae Eurychaeta muscaria (Meigen, 1826) Italy, Latium and Veneto [MZUR] Rognes (1986) Rognes 1991) Calliphoridae/Luciliinae Lucilia sericata (Meigen, 1826) Italy, Sardinia [MZUR] Rognes (1991)

Key to world genera of Rhinophoridae
The following is a standard dichotomous key with couplets containing the main characters chosen for the key path, with alternative states in the respective entries of the couplet, listed by their supposed diagnostic strength. Occasionally, supplementary information is given in square brackets where considered helpful to secure an identification. The key includes the genus Alvamaja as well as the Afrotropical Morinia "carinata species group" (Pape 1997, Rognes 2010, all of which were recently transferred from Rhinophoridae to the re-established oestroid family Polleniidae . As these polleniids are nearly indistinguishable from rhinophorids, and given the lack of comprehensive keys to world oestroid families and genera, their inclusion is deemed helpful. They are given within square brackets. The key also includes the genus Aporeomyia, which was originally tentatively assigned to the Tachinidae based on the morphology of the phallus, but which is here moved to Rhinophoridae based on a reassessment of the homology of relevant phallic sclerotisations (see below). Shiny black or brown flies, usually with bright yellow antenna or, more rarely, with variously patterned legs; with very short setae on head, scutum and abdomen, of approx. the same size as the smallest clothing setae ( Fig. 4C-E). Vein CuA+CuP not reaching wing margin. Male: postpedicel characteristically rounded-subtriangular (axe-head-like and shorter than maximum distal width) ( Fig. 6C-F), the thickened posterior margin abutting the enlarged, sunken face or anterior margin forming two or more lobes; arista very short (i.e., distinctly shorter than length of postpedicel) to nearly absent, and arising apically or sub-apically on postpedicel ( Fig. 6C-F). Female: tergite 7 and sternite 7 forming a strongly sclerotised, dorsoventrally flattened and evenly curved, usually long oviscapt ( Fig. 13K-N First postsutural supra-alar seta present and well developed, as long as or longer than notopleural setae (Fig. 13B, F)  Bend of vein M 1 distinctly rounded and very close to wing margin (Fig. 9G). Antenna distinctly longer than compound eye height, and facial ridge longer than frons (Fig. 6K)  Anterior katepisternal seta more than 1/2 as long as posterior seta. Second aristomere slightly elongated, approx. 1.5 times as long as wide (Fig. 6G, H). Male: fore tarsus not elongated or compressed (Fig. 4F); wing membrane hyaline or evenly smoked (Fig. 9D) Schiner 43 Setae present on lunule (Fig. 13C), notopleuron and katepimeron simultaneously. Median process of ventral sclerotisation of distiphallus interrupted proximally and not connected to ventral plate (Fig. 12G) 12A, C). Male terminalia as in Fig. 12D

New taxa
Four new genera are here erected to accommodate five new species, which do not fit within any of the current generic concepts within Rhinophoridae according to our phylogenetic analysis (see Fig. 20, and "Phylogeny and suprageneric classification" section below). Erecting these new genera is considered superior to alternatives involving generic lumping, as it will facilitate our communication and visualisation of the everincreasing morphological diversity of the world rhinophorids.  Fig. 20). Global apomorphies: none; local apomorphies: 13:1, 14:1, 21:2; 30:2, 42:1, 44:1. Diagnosis. Head: head higher than long in lateral view. Facial ridge 1.1 times as long as frons. Ocellar setae virtually absent. Frons approx. 0.9 times as wide as compound eye in dorsal view. Median (= inner) vertical setae strong and crossed. Five medially crossed frontal setae, slightly reclinate, descending to approx. half level of pedicel. Fronto-orbital plate with some short setulae. Two proclinate orbital setae. One upper lateroclinate orbital seta. Parafacial bare, at its narrowest point at most 1.5 times as wide as maximum diameter of arista. Vibrissal angle receding. Vibrissa well developed, arising slightly below level of lower facial margin. Lower facial margin sunken and not visible in lateral view. Facial ridge slightly and evenly convex with a row of setae on lower 2/3, decreasing in size dorsally. Face deeply concave, antennae hidden in lateral view. Antenna long and narrow, much longer than height of gena. Postpedicel narrowly elongated approx. 5 times as long as pedicel.     Arista bare (or apparently so). Arista thickened in proximal 2/5-1/2; second aristomere at most as long as wide. Lunule bare. Gena, in profile, approx. 1/5 as high as compound eye. Palpus absent.

Maurhinophora
Thorax: prosternum bare. Postpronotum with three setae arranged in triangle. Three postsutural supra-alar setae (first postsutural supra-alar seta well developed, i.e., longer than posterior notopleural seta and approx. the same size as anterior notopleural seta). Scutellum with one pair of well-developed basal setae and one pair of strong, horizontal and crossed apical setae; basal setae placed at level of apical setae. Anatergite with a tuft of short setulae below lower calypter. Subscutellum moderately swollen, not fully sclerotised. Posterior lappet of metathoracic spiracle larger than anterior lappet (as in Baniassa). Lower calypter distinctly tongue-shaped (ground-plan trait of Rhinophoridae) (Fig. 2E). Costal sector cs 2 setose ventrally. Costal spine not differentiated from general costal setae. Costal sector cs 5 clearly shorter than costal sector cs 2 . Vein R 1 entirely setulose dorsally. Vein R 4+5 with setulae dorsally extending from base to approx. level of bend of vein M 1 . Bend of vein M 1 well developed, rounded and well removed from wing margin. Crossvein dm-m forming a right angle with proximal section of M 4 . Vein CuA+CuP not reaching wing margin. Preapical anterodorsal seta of fore tibia longer than preapical dorsal seta. Fore tarsus not compressed. Tibiae of mid and hind leg normally developed. Mid tibia with one, short subdistal anterodorsal seta. Hind tibia with 3 dorsal preapical setae.
Etymology. The generic name is a composite word formed from the first part of the name of the island Mauritius, to which the known species is restricted, and the name Rhinophora, which is the type-genus for the family-group name Rhinophoridae. The name should be treated as a feminine noun. Description. Female. Body length: 5.5 mm. Colouration: head, thorax (including tegula, basicosta, wing veins and legs) yellow; abdomen mostly yellow with blackish brown transversal bands on posterior 1/3 of syntergite 1+2, and posterior 1/2 of tergites 3-5. Head: frontal vitta as wide as fronto-orbital plate. Two strong proclinate orbital setae (one additional short proclinate orbital seta is present posterior to the main ones). Parafacial narrow with only one setula below lower frontal seta. Lateral (= outer) vertical seta well developed though not strongly differentiated from strongest uppermost postocular setae. Prementum approx. 2 times as long as wide; labella not elongated, normally developed. Thorax: one posthumeral seta (medial); 1 + 3 supra-alar setae (posterior postsutural supra-alar weak); 0 + 2 intra-alar setae; 2(3) + 3 dorsocentral setae; 0 + 1 acrostichal setae. Two strong, diverging katepisternal setae. Legs: fore tibia with 1 posterior seta. Abdomen: mid-dorsal depression on syntergite 1+2 confined to anterior half.

Maurhinophora indoceanica
Distribution. Afrotropical -Mauritius. Etymology. The species epithet is derived from the name of the Indian Ocean and should be treated as a Latin adjective.
Remarks. We consider the description of a new species based on a single female as warranted due to the quite remarkable habitus and the occurrence on a small oceanic island, which means that the likelihood of complications due to the lack of male material can be considered low.        Fig. 20). Global apomorphies: none; local apomorphies: 2:1, 9:1, 43:1, 51:1, 52:1, 56:1, 66:0, 68:1.

Marshallicona
Diagnosis. Head: head higher than long in lateral view. Facial ridge 0.4 times as long as frons. Ocellar setae virtually absent. Frons approx. 0.8 times as wide as compound eye in dorsal view. Median vertical setae strong, subparallel. Six to eight medially crossed frontal setae, slightly reclinate, descending to level of upper margin of scape. Fronto-orbital plate with few short setulae confined to upper half. Four or five proclinate orbital setae (posterior two slightly lateroclinate). One weak upper lateroclinate orbital seta (usually not distinguishable from uppermost frontal setae). Parafacial bare, at its narrowest point approx. as wide as width of postpedicel. Vibrissal angle receding. Vibrissa well developed, arising at level of lower facial margin. Lower facial margin not sunken though not visible in lateral view. Facial ridge concave with decumbent setulae on lower 1/3-2/5. Face slightly concave, antennae not hidden from view in profile. Antenna approx. as long as height of gena. Postpedicel sub-ovoid, approx. 1.5 times as long as pedicel. Arista bottlebrush-like, trichia longer that maximum diameter of arista. Arista thickened in proximal 1/5 or less; second aristomere at most as long as wide. Lunule hidden by inner anterior margins of fronto-orbital plate. Gena, in profile, approx. 1/2 as high as compound eye. Palpus stout, clavate, with a few thin setulae on apical 1/4. Thorax: prosternum bare. Postpronotum with two setae. One postsutural supra-alar seta (i.e., first and third post sutural supra-alar setae absent). Scutellum with one pair of well-developed basal setae and one pair of strong, horizontal and crossed apical setae; basal setae placed dorsally with respect to apical setae. Anatergite bare. Subscutellum moderately swollen, not fully sclerotised. Metathoracic spiracular lappets virtually absent. Lower calypter distinctly tongue-shaped (ground-plan trait of Rhinophoridae) (Fig. 2E). Costal sector cs 2 setose ventrally. Costal spine not differentiated from general costal setae. Costal sector cs 5 longer than costal sector cs 2 (Fig. 15E). Vein R 1 dorsally setose on distal 1/4. Base of R 4+5 entirely bare. Bend of vein M 1 shallow, well removed from wing margin. Crossvein dm-m forming an acute angle with proximal section of M 4 . Vein CuA+CuP not reaching wing margin. Preapical anterodorsal seta of fore tibia approx. as long as preapical dorsal seta. Fore tarsus not compressed. Tibiae of mid and hind leg normally developed. Mid tibia with one anterodorsal seta. Hind tibia with 3 dorsal preapical setae.
Abdomen: slightly elongated, virtually without microtomentum and without distinct marginal and discal setae.
Male terminalia (Fig. 15F-H): posterior margin of sternite 5 with a deep median notch; lateral lobe rounded posteriorly. Tergite 6 bare, medially divided into two hemitergites; tergite 6 divided from syntergosternite 7+8 by a wide membrane. Connection between sternite 6 and syntergosternite 7+8 on right side membranous. Cerci well developed, not fused medially. Basal 1/3 of cerci convex and covered with short setae; distal 2/3 straight and narrowly digitiform (branches symmetrically diverging and well separated). Surstylus well developed, lobe-like in lateral view; lateral side of surstylus broadly convex at approx. mid length. Surstylus not fused to epandrium. Bacilliform sclerite firmly fused to laterobasal margin of surstylus. Hypandrial arms not fused medially. Connection between phallic guide and pregonite sclerotised. Postgonite without anterior seta. Epiphallus well developed and attached dorsomedially to basiphallus. Extension of dorsal sclerite of distiphallus entirely fused medially into a single sclerite and proximally fused to dorsal sclerite of distiphallus. Median process of ventral sclerotisation of distiphallus present, divided medially into two hemisclerites, which are both proximally fused to ventral plate of distiphallus. Acrophallus simple and scale-like spinules present lateroventrally.  Cerretti & Pape, sp. nov., by present designation. Etymology. The generic name is a composite word formed from the name of our colleague and friend Steve Marshall, who collected the type series and took the photo of a living specimen (Fig. 1C), and from the Latin noun 'icona', meaning image, in honour of Steve's remarkable skills in natural history photography. The name should be treated as a feminine noun. Description. Male. Body length: 4.5-5.5 mm. Colouration: head mostly black in ground colour, antenna and palpus brown; occiput and posterior 2/3 of fronto-orbital plate shiny without microtomentum, anterior 1/3 of fronto-orbital plate and parafacial covered with silver reflecting microtomentum; thorax (including tegula, basicosta, wing veins and legs) black in ground colour, microtomentum virtually absent; abdomen mainly yellow except brownish posterior margin of tergites 1+2, 3 and 4 and on posterior 1/2-2/3 of tergite 5. Wing membrane infuscate around veins. Head: frontal vitta slightly narrower than fronto-orbital plate (measured at midlength). Parafacial entirely bare below lower frontal seta. Lateral vertical seta not differentiated from strongest uppermost postocular setae. Prementum stout, not longer than wide; labella broad. Thorax: one posthumeral seta (medial); 1 + 1 supra-alar setae; 0 + 2 intra-alar setae; 2(3) + 3 dorsocentral setae; 0-3 + 1 acrostichal setae. One or 2 katepisternal setae. Legs: fore tibia without posterior seta. Abdomen: mid-dorsal depression on syntergite 1+2 confined to anterior half.

Neotarsina
Diagnosis. Head: head higher than long in lateral view. Facial ridge 0.6 times as long as frons. Ocellar setae virtually absent. Frons 0.3-0.5 (male), 0.8-0.9 (female) times as wide as compound eye in dorsal view. Median vertical setae converging or crossed, though very short, at most as long as antenna. Ten to 20 short, medioclinate frontal setae, descending to level of upper margin of scape. Fronto-orbital plate bare or with scattered setulae interspersed between frontal setae. Proclinate orbital setae absent. Upper reclinate orbital seta absent. Parafacial bare, at its narrowest point 0.8-1.2 times as wide as width of postpedicel. Vibrissal angle receding. Vibrissa weak, i.e., barely distinguishable from setae of subvibrissal ridge, arising at level of lower facial margin. Lower facial margin not sunken though not visible in lateral view. Facial ridge concave with decumbent setulae on lower 1/5-2/5. Face slightly concave, antennae not hidden in lateral view. Antenna shorter than height of gena. Postpedicel sub-ovoid, approx. 1.0-1.7 times as long as pedicel. Arista bare. Arista thickened on proximal 1/10-1/5 of its length; second aristomere at most as long as wide. Lunule bare. Gena, in profile, 2/5-1/2 as high as compound eye. Palpus very short 1-2 times as long as wide, bare.
Thorax: prosternum bare. Postpronotum with 2-3 setae. One postsutural supra-alar seta (i.e., first and third post sutural supra-alar setae absent). Scutellum with one pair of basal setae and one pair of, crossed, horizontal apical setae; basal setae placed dorsally with respect to apical setae. Anatergite bare. Subscutellum moderately swollen or flat, not fully sclerotised. Metathoracic spiracular lappets small, subequal in size and directed outwards. Lower calypter distinctly tongue-shaped (ground-plan trait of Rhinophoridae) (Fig. 2E). Costal sector cs 2 usually setose ventrally. Costal spine not differentiated from general costal setae. Costal sector cs 5 approx. as long as costal sector cs 2 . Vein R 1 dorsally bare. Base of R 4+5 entirely bare. Bend of vein M 1 indistinct; i.e., M 1 evenly curved forward without forming a distinct bend. Crossvein dm-m forming a right angle with proximal section of M 4 . Vein CuA+CuP not reaching wing margin. Preapical anterodorsal seta of fore tibia longer than preapical dorsal seta. Fore tarsus strongly laterally compressed in both sexes. Tibiae of mid and hind legs laterally compressed and distinctly keeled dorsally. Mid tibia without anterodorsal setae. Hind tibia with three dorsal preapical setae.
Abdomen: lightly elongated, varying from slightly microtomentose to virtually without microtomentum. Marginal and discal setae not differentiated from general setulae.
Male terminalia: posterior margin of sternite 5 with a deep median notch; lateral lobe rounded posteriorly. Tergite 6 plate-like, with median marginal setae; tergite 6 divided from syntergosternite 7+8 by a membrane. Connection between sternite 6 and syntergosternite 7+8 fused on right side. Cerci well developed, basally broad, narrowing toward apex and apically pointed, well separated medially. Surstylus well developed, subtriangular in lateral view; lateral side of surstylus not or only slightly convex. Surstylus fused to epandrium. Bacilliform sclerite articulated (i.e., not fused) to laterobasal margin of surstylus. Hypandrial arms not fused medially. Connection between phallic guide and pregonite membranous. Postgonite without anterior seta. Epiphallus well developed and attached dorsomedially or dorsomedially to basiphallus. Extension of dorsal sclerite of distiphallus divided medially into two hemisclerites which are proximally not fused to dorsal sclerite of distiphallus. Median process of ventral sclerotisation of distiphallus present, not interrupted, i.e., running from the ventral plate to tip of phallus, and not divided medially. Acrophallus simple and scale-like spinules not differentiated.  Cerretti & Pape, sp. nov., by present designation. Etymology. The generic name is a composite word formed from the Latin word 'neo' meaning new, and 'tarsina' [from Latin 'tarsus' and Greek: 'tarsos', the flat part of a human foot] as in the last part of the genus-group name Macrotarsina, in reference to the modified, laterally-compressed fore tarsus characterising the two new species described below. The name should be treated as a feminine noun. Notes on type specimens. The holotype lacks postpronotal and scutellar setae but is otherwise in good condition. The paratype is in fair general condition but lacks most of the setae on the thorax.

Neotarsina andina
Description. Male. Body length: ca. 4 mm. Colouration: head mostly black except antenna and palpus which are pale yellow, parafacial, gena and lower occiput covered with dense silver reflecting microtomentum; fronto-orbital plate polished shiny black, except narrowing microtomentose along medial margin; thorax entirely black or dark brown in ground colour; thoracic pleura covered with silver reflecting microtomentum; femora mostly yellow except dark brown proximally, tibiae and tarsi dark brown; calypters whitish; tegula and basicosta pale yellow; wing membrane hyaline; halter yellow; abdomen black in ground colour, tergites 3 and 4 laterally covered with silver reflecting microtomentum. Head: frontal vitta measured at mid-length approx. 1.7 times as wide as fronto-orbital plate at same level. Postpedicel 1.4-1.7 times as long as pedicel. First and second aristomeres approx. as long as wide. Frontal setae medioclinate, slightly proclinate near antennal insertion. Median vertical setae short, approx. 0.2 times as long as compound eye height. Parafacial approx. 1.2 times as wide as postpedicel. Facial ridge bare or with few short setae above vibrissa. Gena, in profile, 2/5 as high as compound eye. Frons 0.5 times as wide as eye in dorsal view. Prementum stubby, approx. 2 times as long as wide; labella not elongated, normally developed. Palpus short, clavate. Thorax: two postpronotal setae; two posthumeral setae; 0 + 1 supra-alar setae (i.e., first and third postsutural supra-alar setae absent); 0 + 2 intra-alar setae; 2 + 3 dorsocentral setae (first presutural dorsocentral barely distinguishable from general setulae); one presutural acrostichal seta; postsutural acrostichal setae apparently not differentiated. Fore tibia approx. 1.5 times as long as first tarsomere. Costal sector cs 2 nearly bare ventrally. Abdomen: Tergite 5 very short, approx. 0.5-0.6 times as long as tergite 4. Male terminalia: Surstylus narrowly triangular in lateral view. Cerci relatively narrow with a clear, though shallow, bottleneck restriction at approx. mid length; apical 1/5 of cerci strongly narrowing and pointed. Female. Body length: 5 mm. Female differs from male as follows: Head: fronto-orbital plate entirely shiny and parafacial with less dense microtomentum. Frontal vitta measured at mid-length approx. 2.0 times as wide as fronto-orbital plate at same level. Frons 0.8 times as wide as eye in dorsal view.

Distribution. Neotropical -Trinidad and Tobago (Trinidad).
Etymology. The species epithet is derived from the Spanish word for Caribbean and should be treated as a Latin adjective.
Diagnosis. Head: head higher than long in lateral view. Facial ridge approx. 1.3 times as long as frons. Ocellar setae present though small. Frons 0.9-1.1 times as wide as compound eye in dorsal view. Median vertical setae strong and slightly converging. Two or three slightly convergent and reclinate frontal setae, descending to approx. half level of pedicel. Fronto-orbital plate nearly bare with one proclinate orbital seta (no sexual dimorphism). One upper lateroclinate orbital seta. Parafacial bare, at its narrowest point approx. twice as wide as maximum diameter of arista. Vibrissal angle strongly receding. Vibrissa well developed, arising slightly below the level of lower facial margin. Lower facial margin not sunken and slightly visible in lateral view. Facial ridge strongly concave with two or three short setulae above vibrissa. Face concave but antennae not hidden and clearly visible in lateral view. Antenna long and wide (in lateral view), much longer than height of gena. Postpedicel 4.5-6.5 times as long as pedicel. Postpedicel axe-shaped in male, more or less stick-like in female. Arista bare (or apparently so). First and second aristomere slightly thickened and strongly elongated. Lunule bare. Gena, in profile, approx. 3/5 (male), 1/2 (female) as high as compound eye. Palpus pretty short, dark brown.
Thorax: prosternum bare. Postpronotum with two setae. Two postsutural supra-alar setae (first postsutural supra-alar seta absent). Scutellum with one pair of strong lateral setae and one pair of short, crossed and horizontal apical setae. Anatergite with a tuft of short setulae below lower calypter. Subscutellum moderately swollen, mostly, though not entirely, sclerotised. Metathoracic spiracular lappets nearly undeveloped. Lower calypter distinctly tongue-shaped (ground-plan trait of Rhinophoridae) (Fig. 2E). Costal sector cs2 setose ventrally. Costal spine not differentiated from general costal setae. Costal sector cs 5 clearly longer than costal sector cs 2 . Vein R 1 entirely bare. Vein R 4+5 bare dorsally. Bend of vein M 1 absent. Crossvein dm-m forming a right angle with proximal section of M 4 . Vein CuA+CuP reaching wing margin. Preapical anterodorsal seta of fore tibia slightly longer than preapical dorsal seta. Fore tarsus not compressed. Tibiae of mid and hind leg normally developed. Hind tibia with three dorsal preapical setae.
Distribution. Oriental -Indonesia (Bali), Malaysia (Sabah), Philippines (Palawan). Type species. Kinabalumyia pinax Cerretti & Pape, sp. nov., by present designation. Etymology. The generic name is a composite word formed from the name of the type locality of the type species, Mount Kinabalu, and from the Greek word μύγα (miga), meaning fly. The name should be treated as a feminine noun.  1964-250 (NHMUK). Paratype ♂: same data and repository as holotype.
Distribution. Oriental -Malaysia (Sabah). Etymology. The species epithet, which should be treated as a noun in apposition, is derived from the Greek noun pinax, meaning painting, in reference to the remarkable colour pattern of the thorax and abdomen. Remarks. This female specimen, in fair condition, strongly resembles the two males from Sabah described as K. pinax above and may be conspecific with them. However, the sparse material available, the lack of females from the type locality, and the wide geographic separation provide sufficient taxonomic uncertainty to suggest caution. We therefore prefer to treat this specimen as unidentified awaiting further material.

Kinabalumyia sp. 2 (undescribed)
Material examined. 1 ♂: Indonesia, Bali, Bratan L. env., 1250m, Febr, 2014, O. Kosterin (ZMUM); photo only ( Fig. 19), image also available from www.diptera.info (photo_id=9408). Remarks. The photo shows a male specimen assessed as belonging to the genus here newly described as Kinabalumyia. Shape of the antenna and colouration of the body suggest that it is not conspecific with K. pinax and therefore represents an undescribed species.

Diversity and distribution
The Palaearctic Region hosts by far the most diverse rhinophorid fauna, with 92 of the 177 described species, and these are heavily concentrated to the western part. The Afrotropical Region contains 33 species, followed by the Australasian and Neotropical regions with 24 and 21 species, respectively. Only seven species (plus the two undescribed) have been recorded from the Oriental Region, and the by far most speciespoor region is the Nearctic with only four native rhinophorids, all belonging to the genus Bezzimyia. This genus reaches its peak of diversity in the Neotropical Region. Indeed, the distribution of the Nearctic species of Bezzimyia is limited to the southernmost parts of the USA (Arizona, Texas, Georgia, Florida), corresponding with the range of the few native Nearctic woodlice (see Jass andKlausmeier 2000, Schmalfuss 2003), which are the most probable hosts of Bezzimyia and relatives. The near-absence of native woodlice in most of the temperate parts of North America has been given as in need of explanation (Jass and Klausmeier 2000), although it is rather the remarkable diversity of woodlice in the western Palaearctic that asks for scrutiny ("Paradoxically, their present centre of distribution and evolution is the region of the Mediterranean Sea", Hatch 1947: 175; see also Sfenthourakis and Hornung 2018). Faunistic connections are scarce among biogeographic regions, as virtually no rhinophorid species is so far known to be distributed across two or more regions, with the exclusion of a few anthropogenic introductions outside their native distributional range. In particular, three rhinophorids have been introduced from Europe to the Americas (Downes 1965, Mulieri et al. 2010, Wood et al. 2018, namely: Melanophora roralis (Linnaeus), which is now widespread in the New World; Stevenia deceptoria (Loew), recently reported from Ohio (USA) and northern Argentina; and Phyto discrepans Pandellé, known from Canada. While the two former species are now well established, little is known about the status of the latter. Moreover, Melanophora roralis, which is a native West-Palaearctic element, has been introduced to Japan and Cape Verde Islands, where it is now established (Crosskey 1977, Kato andTachi 2016).

Phylogeny and suprageneric classification
Monophyly of Rhinophoridae and differences with previous analyses As stressed by Cerretti and Pape (2012) and Cerretti et al. (2014a), the rhinophorid phylogeny based on currently available morphological data is poorly resolved, while a comprehensive phylogenetic study based on molecular data is still under completion (Gisondi et al., unpublished). Nevertheless, refined morphological evidence, coupled with new discoveries, are producing interesting results. Our taxon sample includes at least one species of all recognised rhinophorid genera, thereby comprising 14 more ingroup species than that of Cerretti et al. (2014a) (Table 1, asterisked names).
Our analyses reconstructed Rhinophoridae (clade A) as monophyletic based on eight unambiguous character state changes [four traceable in the adult: arista with short trichia (1:0, local apomorphy), first postsutural supra-alar seta short (25:1, local apomorphy), subscutellum moderately swollen (26:1, global apomorphy), lower calypter tongue-shaped (36:1, local apomorphy); four in the first instar larva: antenna long and tapering (85:1, global apomorphy), posterior part of anal division modified as a terminal sucker (86:1, global apomorphy), mandibles toothed or serrated (88:1, global apomorphy), and parastomal bar of cephaloskeleton long and slender (91:1, local apomorphy)], confirming the split of the family into two main subclades: the Phyto group (clade B), the first instar larvae of which move by somersaulting, and the Stevenia group (clade E), whose first instars move in a leech-like crawling fashion (see Bedding 1973, Pape 1986, Pape and Arnaud 2001, Cerretti et al. 2014a (Fig. 20). Analyses by Cerretti et al. (2014a) found the New Caledonian endemic genus Rhinodonia as sister to the remaining rhinophorids, and the Australasian genera Axinia and Bixinia as sister taxa nested within the Stevenia group. Our analyses still recovered Bixinia within the Stevenia group but notably diverged from Cerretti et al. (2014a) in retrieving Rhinodonia and Axinia as sister clades and part of an Oriental-Australasian radiation of the Phyto group. The differences between analyses stand likely by the double effect given by the new taxa included in the present version of the matrix and by the application of the implied weighting function implemented in TNT [k/(es+k), where k = concavity constant and es=extra step, i.e., as k increases, the function approximate the linear, equally weighted ("unweighted") function (Goloboff 1993)]. Under implied weighting, the increase in fit (with differing strength depending upon the k-value) caused by the reduction in number of steps for a character is higher for characters with low homoplasy because each step represents a larger fraction of the total homoplasy for that character. In Fig. 20 one of these apomorphies, shared by the Australasian taxa except Bixinia (clade C), is the lack of a bend of vein M 1 (45:1). The position of Axinia as a member of the Phyto group has recently been corroborated by molecular data . However, the affinities of Bixinia are more problematic, as both its separation from Axinia and its position within a Neotropical clade (I) appear unlikely, as 'intuitively' assessed by its relative similarity to other Australasian taxa (see below).

Melinda gentilis
Pollenia paupera  Figure 20. Phylogeny of Rhinophoridae based on morphological evidence. Single tree obtained under IW (total fit: 62.33), enforcing a k-value of 4, with unambiguous character state changes mapped above branches [black squares: global apomorphies (i.e., uncontradicted and unreversed apomorphic character states; white squares: local apomorphies (i.e., homoplasious character state due to convergence or reversal)] and new taxa highlighted in bold.

Phyto group and Stevenia group
The Phyto group (Fig. 20, clade B) was supported by four unambiguous character state changes [anal division of the first instar larva with longitudinal cuticular ridges in posteroventral position (95:1, global apomorphy) and a tongue-like projection in posterodorsal position (96:1, global apomorphy), cephaloskeleton of the first instar with longitudinal incision on parastomal bar (92:1, local apomorphy), and adult male with well-developed proclinate orbital setae (9:1, local apomorphy)]. The clade is composed exclusively of Old World elements. Remarkably, most of the characters whose derived condition supported the monophyly of clade B (Phyto group) are from the first instar larva, which is known for only three (i.e., Paykullia, Melanophora, and Phyto) of the 12 genera composing the group. The Stevenia group (clade E) is nearly worldwide in distribution. Its monophyly was supported by four unambiguous character state changes [first instar larva with prolegs (93:1, global apomorphy), mandible provided with three or more teeth ( (80:1, global apomorphy), and parafrontal plate with proclinate orbital setae (9:1, local apomorphy)] and includes only the endemic Afrotropical genus Ventrops (Cerretti & Pape, 2012). Finally, clade I is weakly supported by three local apomorphies [female without proclinate orbital setae (10:0), male surstylus fused to epandrium (66:1), and postgonite without anterior seta (69:1)], and is mostly composed of Neotropical taxa, with the notable exception of the Australian genus Bixinia, whose phylogenetic position seems unlikely and needs further study. Ventrops (i.e., clade H) was reconstructed as sister to clade I based on two local apomorphies [fronto-orbital plate with more than two proclinate orbital setae (11:1), and extensions of dorsal sclerite of distiphallus entirely fused mid-dorsally from each other into a single sclerotisation (71:1)]. The preimaginal stages and the natural history of Ventrops are unknown and more data are needed to clarify its phylogenetic position within Rhinophoridae.
The recent molecular-based phylogenetic reconstruction by Cerretti et al. (2019) recovered monophyly for both the Phyto group and the Stevenia group.
The Aporeomyia conundrum The genus Aporeomyia was originally assigned to the family Tachinidae by Pape and Shima (1993), despite sharing with Rhinophoridae two local apomorphies, i.e., a tonguelike lower calypter (36:1) and anterior and posterior lappets of metathoracic spiracle subequal in size and directed outward (30:1). The original placement of this unusual taxon was based on the presence of a moderately convex subscutellum and a distiphallus characterised by an undivided extension of the dorsal sclerite, as in tachinids (Pape and Shima 1993). Aporeomyia is here reassigned to Rhinophoridae based on a reassessment of the homologies of the thorax and male terminalia. The subscutellum in Aporeomyia is moderately convex, but not bulging and fully sclerotised as in tachinids, which indicates a rhinophorid affinity. Notably, several rhinophorids also have a moderately convex subscutellum similar to that of Aporeomyia, in particular among species within various genera of the Phyto group, such as Phyto, Comoromyia, Malayia, Rhinopeza, and Kinabalumyia gen. nov. This character state is not restricted to this group: the Australasian genus Bixinia has a bulging subscutellum similar to that of tachinids, but its inclusion in Rhinophoridae is beyond any reasonable doubt (Kutty et al. 2019). As discussed by Pape and Shima (1993: 78, fig. 5), the extension of the dorsal sclerite of the distiphallus in Aporeomyia antennalis Pape & Shima is undivided, as typical for nearly all tachinids and a few other non-tachinid oestroids (see also Rognes 1997, Pape andArnaud 2001). However, recent morphology-based cladistic analyses including a growing number of rhinophorid taxa indicate that an undivided extension of the dorsal sclerite of the distiphallus characterises several rhinophorid species belonging to different genera [e.g., Rhinophora lepida (Meigen), some Bezzimyia, Shannoniella, Trypetidomima, some Melanophora], thus confirming this trait as homoplasious within oestroids, and among rhinophorids in particular (Pape and Arnaud 2001, Cerretti and Pape 2012, Cerretti et al. 2014a. Moreover, the apparently undescribed species of Aporeomyia from Sabah included in our analysis is characterised by an extension of the dorsal sclerite of distiphallus that is longitudinally divided, even if shallowly, into two branches, suggesting that the two conditions can occur even in closely related species. We retrieved Aporeomyia nested within the Phyto group (clade B), as sister to the Oriental genus Kinabalumyia gen. nov. based on three unambiguous character state changes [first aristomere at least four times as long as wide (6:2, global apomorphy), wing vein CuA+CuP reaching wing margin (50:1, local apomorphy), and male with connection between tergite 6 and syntergosternite 7+8 fused (58:1, global apomorphy)]. The monophyly of Aporeomyia is based almost entirely on the strongly autapomorphic postpedicel with a tripartition (Fig. 6B) similar to that of the Neotropical Bezzimyia lobata Pape & Arnaud, but with sufficient differences to be considered a unique autapomorphy of Aporeomyia. In conclusion, we found strong evidence supporting Aporeomyia as a member of the Rhinophoridae as well as generic monophyly.
The position of Kinabalumyia gen. nov. within the Phyto group The Oriental genus Kinabalumyia gen. nov. was supported by four unambiguous character state changes [distiphallus with a helmet-shaped, partly sclerotised envelope (81:1, global apomorphy), palpus reduced (21:1, local apomorphy), and male with: sternite 5 without posterior notch (53:1, local apomorphy), and connection between sternite 6 and syntergosternite 7+8 fused on right side (59:1, local apomorphy)] and it was reconstructed as sister to the Oriental genus Aporeomyia (see above). This clade was in turn retrieved as sister to the Australasian clade D (Rhinodonia + Axinia) based on two local apomorphies [base of wing vein R 4+5 bare (43:1), and male cerci medially fused into a syncercus (61:1)]. We think that the helmet-shaped envelope of distiphallus is a strong autapomorphy supporting the erection of a new genus for the examined specimens. We have examined four specimens of Kinabalumyia gen. nov. from three localities scattered from Bali in Indonesia to Palawan in the Philippines, suggesting that this genus is widespread at least throughout the non-continental part of the Oriental Region, with further species probably awaiting discovery. Clade C (Fig. 20) was supported by two unambiguous character state changes [bend of vein M 1 indistinct (45:1, global apomorphy) and metathoracic spiracular lappets absent (30:0, local apomorphy)] and it reconstructed the Australasian clade (clade D) arising from a grade composed of Australasian (Rhinopeza) and Oriental (Aporeomyia, Kinabalumyia gen. nov.) taxa. Despite being weakly supported by one local apomorphy [arista thickened at least on basal 3/4 (5:1)], clade C is in turn sister to the Oriental genus Malayia. If this reconstruction were accurate, the hypothesis of Cerretti et al. (2014a) about the origin of the Australasian rhinophorid fauna would need to be reconsidered. In fact, our phylogeny favours an Old World ancestor for most of the Australasian taxa, the only notable exception being Bixinia of Neotropical descent.
The Stevenia group and the position of Maurhinophora gen. nov., Marshallicona gen. nov. and Neotarsina gen. nov.
The monotypic genus Maurhinophora gen. nov. was erected to accommodate the new species M. indoceanica sp. nov. from Mauritius, based on a single female. The genus is supported in this work by six local apomorphies [facial plate and lower facial margin deeply sunken (13:1 and 14:1), palpus absent (21:2), lappets of metathoracic spiracle unequal in size, i.e., posterior one distinctly larger (30:2), wing vein R 1 with a row of setulae along whole length (42:1), and base of wing vein R 4+5 with a row of setae reaching at least crossvein r-m (44:1)] and was reconstructed as sister to the monotypic endemic Afrotropical genus Queximyia. Maurhinophora gen. nov. and Queximyia were part of the large clade F. The latter group was based on a global apomorphy of the first instar larva [first instar larva with slightly flattened body shape (90:1)], while an included subclade (clade G), also comprising Maurhinophora gen. nov. + Queximyia, was instead supported by one global apomorphy of the male terminalia [acrophallus distinctly tripartite, i.e., with three openings (79:1)]. This phylogenentic reconstruction suggests that the first instar and adult male of Maurhinophora gen. nov. (both unknown) might share these traits as well.
Two of the four newly described genera are Neotropical in distribution. During the last decades, the native New World rhinophorid fauna has increased remarkably from one genus and two species to the present six genera and 25 species (Pape and Arnaud 2001, Cerretti et al. 2014a, Nihei and Andrade 2014, Nihei et al. 2016, and more are most likely awaiting discovery. Although based on a large array of local apomorphies, our analysis corroborated the non-monophyly of the genus Bezzimyia, as previously suggested by Cerretti and Pape (2012) and Cerretti et al. (2014a). The monotypic genus Marshallicona gen. nov. was supported by seven local apomorphies [arista microtrichia bottlebrush-like in male (2:1), male with proclinate orbital setae (9:1), base of wing vein R 4+5 bare (43:1), abdominal tergites 3 and 4 without median marginal setae (51:1 and 52:1), abdominal tergite 6 bare in male (56:1), connection between surstylus and epandrium membranous (66:0), and phallic guide fused to pregonites (68:1)] and was reconstructed as sister to the clade Trypetidomima + Shannoniella.
Remarks. Treated as unplaced to genus by Herting (1961: 34) but here resurrected as a valid species in the genus Acompomintho. Herting (1961) did not examine the holotype of F. caucasica and he claimed that the original description was inconclusive as to the generic placement of caucasica. In particular, Herting mentioned that despite similarities in colouration and habitus with Tricogena rubricosa (Meigen) [Frauenfeldia Egger is a junior synonym of Tricogena, see below], the original description of F. caucasica lacked any mention of robust setae on the lower half of the parafacial (characterising Tricogena and Stevenia, but not all described Acompomintho), or any information about other diagnostic characters. We have not been able to locate and examine the holotype of caucasica, but we consider the original description to match Acompomintho better than Tricogena. Distribution. Palaearctic -Russia (North Caucasus).
angustifrons Villeneuve, 1912 Stevenia angustifrons  Remarks. Ptilocheta bertei was listed as a nomen dubium by Herting (1993), but it is here resurrected as a valid species based on our examination of the holotype male, which was recently rediscovered in the Rondani collection at MZF. The specimen is pinned and in good condition. Following the key to Palaearctic Stevenia provided by Cerretti and Pape (2007), Stevenia bertei runs to couplet 18 and is separable from S. etrusca by male with three proclinate orbital setae (anterior one distinctly longer), one upper reclinate orbital seta, abdomen yellow laterally on tergites 1+2 and 3, and abdominal tergites 1+2, 3 and 4 without median discal setae. Distribution. Palaearctic -Croatia [new record], Italy (mainland).