Phylogeny and Taxonomy of the Risa Genus-group (Diptera: Ephydridae), with Description of a New Genus from Australia

. The systematic and somewhat controversial history of Risa Becker is presented, and its relationship with Diasemocera Bezzi (tribe Psilopini, Ephydridae) is documented by morphological evidence and an association with host plants in the family Amaranthaceae. The tribe Risini Papp (as Risidae) is synonymized with Psilopini Cresson. Notorisa gen. nov., from Australia, is described (type species: Notorisa mcalpinei sp. nov.; Australia. Victoria: Big Desert National Park, near Lake Hindmarsh; 36°03.7'S 141°54.8'E). Achaetorisa Papp is retained as a subgenus within Risa and includes five species, including two new combinations: Risa brevicornis (Papp) comb. nov., Risa salsolae (Mathis & Zatwarnicki) comb. nov., and two new species: R. brevirostris sp. nov. (Israel. Ẕomet Zohar; 31°08.5'N 35°21.6'E) and R. nettae sp. nov. (Israel. Ẕomet Zohar; 31°08.5'N 35°21.6'E). A fourth new species is described in the subgenus Risa: R. (Risa) kotrbae sp. nov. (Israel. Ẕomet Zohar; 31°08.5'N 35°21.6'E).


Introduction
The genera and species included in the "the Risa group" have been the source and subject of controversy and debate since their first description. Specimens of these taxa are real, but their hypothetical and phylogenetic placement within the higher classification of Acalyptratae (Diptera: Schizophora) has been historically unstable, depending on which character suites were employed and analyzed and interpretations of variable characters. This situation is perhaps to be expected when dealing with specimens that are tiny (body length 1-3 mm), dark colored and shiny, rarely collected, and demonstrate considerable reduction in setation, wing venation, and structures of the male and female terminalia. This reduction is sometimes by convergence in the phenotypic expression of some characters, especially those of mouthparts, which are moderately to remarkably long and with a geniculate proboscis. Specimens are generally rare in collections, and some species are known only from their type localities.
Although the proposed classifications of the Risa group lack complete resolution, progress has been made, and in recent decades, the focus has narrowed to differing placements of the group within the superfamily Ephydroidea (Wiegmann et al., 2011;Winkler et al., 2022). Further resolution of this phylogenetic and classificatory puzzle is the overall objective of this paper in addition to describing new taxa. First, however, some historical background. Becker (1907) described Risa and its only included species, R. longirostris Becker (a male and female from Biskra, Algeria). Becker (1907) first placed Risa in the family Ephydridae, subfamily Agromyzinae. Thereafter, however, there has been little agreement on the placement of the genus. Melander (1913) was the first to associate Risa with Milichiidae and therein to the Madizina group. Hennig (1937) redescribed Becker's types and suggested that Risa be placed in the family Milichiidae (subfamily Carninae). Despite having misgivings, Hennig (1965: 184) continued placement in Milichiidae ("Die Gattungen Horaismoptera und Risa, die ich früher (1937) zu den Carnidae ("Carninae") stellte, haben stark nach hinten verlängerte Labellen und dürften daher zu den Milichiidae gehören." English: "The genera Horaismoptera and Risa, which I formerly (1937) included in the Carnidae ("Carninae"), have strongly elongated labellum directed backwards, and should therefore belong to the Milichiidae.") and again in 1973 (p. 61) Milichiidae (subfamily Madizinae). Papp (1977) was the first to attribute familial status for Risa (Risidae). Later, Papp (1984) cataloged the family, noting that Achaetorisa brevicornis Papp was parasitic on its host caterpillar. J. F. McAlpine (1989) questioned the necessity for a separate family in his discussion on the phylogeny of the Risidae and included skeptical comments on the parasitic relationships noted previously. J. F. McAlpine (1989) suggested that Risa is no more than a genus of Milichiidae. More recently, Kotrba & Mathis (2009) described the internal female reproductive tract of Risa, discussed the relationship of Risa with Ephydridae, and first suggested its membership in the subfamily Discomyzinae Acloque. Although the phylogenetic relationships and familial status of Risa are now better resolved, there is much need for further research at both the descriptive and phylogenetic levels. These taxa generally remain poorly understood and studied.
At the species level, additional taxa and locality data were added. Steyskal (1968) recorded a specimen from Egypt, and Papp (1977) recharacterized the type species, designating a lectotype, and described R. mongolica. Papp (1980) then described R. longicornuta and a second genus and its included type species, Achaetorisa brevicornis Papp. The latter species was described from 20 specimens from Morocco, some allegedly reared from a caterpillar found on Halogeton sativus (L.) Moq. (Amaranthaceae). Ozerov (1984Ozerov ( , 1992 described three new species in Risa: R. asiatica, R. flavipalpis, and R. nartshukae but subsequently synonymized the latter with R. longirostris. More recently, Mathis & Zatwarnicki (2017) described Achaetorisa salsolae from the United Arab Emirates and Krivosheina and Ozerov (2019) added Risa flavicoxa from central Asia. Thus, until now the so-called family has included seven valid species in two genera. No fossils are known.
The Risa group of genera occurs in arid regions of the Old World and has greatest diversity along the great desert belt, extending from northwest Africa through the Near East to Mongolia.
Our purposes herein are to describe additional new species and to describe a new genus and species from Australia, Notorisa mcalpinei. The latter genus and species have further contributed to our knowledge of the group and provided additional evidence that this group should be placed in the shore-fly family Ephydridae, subfamily Discomyzinae, tribe Psilopini Cresson. Risa and Notorisa are somewhat anomalous genera, demonstrating several structurally derived conditions, especially in external characters of the head that have posed difficulties in placing the family within the Ephydroidea, including, as noted previously, recognition as a separate family or tribe (Papp, 1977). Risa, for example, was included prominently in analyses and discussions of the Ephydroidea (Chandler, 1987;Grimaldi, 1990;Wiegmann et al., 2011;Winkler et al., 2022). An improved phylogenetic framework available for the Ephydroidea and for the family Ephydridae (Chandler, 1987;Zatwarnicki, 1992;Mathis & Zatwarnicki, 1995) provides the context for this paper, including placement of these genera within the shore-fly tribe Psilopini Cresson (see "Discussion" associated with the tribal diagnosis).
The impetus for this paper came from discovery of new species of Risa Becker from Israel and the Sinai (Egypt) and of a new genus and species from Australia. Also contributing significantly was the availability of larger series, including numerous specimens of the new species, especially from the Middle East and Australia. The availability of large series was tremendously increased with discovery of a biological association between Risa and plants of the family Amaranthaceae (including the former goosefoot family Chenopodiaceae).
In summary, although Risa has hitherto been enigmatic and rare in collections, it now appears to be more common and widespread than previous locality data had indicated. An important outcome of this study is the establishment of a strong association between Risa and various species of the plant family Amaranthaceae. This has already supplied us with information about where potentially to look for these flies. Wherever an association with a plant species was established, many specimens were easily and relatively quickly collected. It may be that all the species are associated with Amaranthaceae, although this remains unsubstantiated for a majority of species. About half of the species have been known from a single locality, indicating that additional species probably exist and await discovery and description.

Methods and materials
The descriptive terminology, with the exceptions noted in Mathis (1986) and Mathis & Zatwarnicki (1990a), follows Cumming & Wood (2017). Because specimens are small, usually less than 3.5 mm in length, study and illustration of the male terminalia required use of a compound microscope. We have followed the terminology of Zatwarnicki (1996) for most structures of the male terminalia, which differs from the terminology that other workers in Ephydridae have used (see references in Mathis, 1986, and Mathis & Zatwarnicki, 1990a, 1990b such as the use of gonostylus instead of surstylus. The term surstylus is based on the erroneous assumption that the "clasping lobe" in Cyclorrhapha does not have a muscular connection with the epandrium. There is no evidence of its secondary origin in Eremoneura and its musculature is easily homologized throughout all Brachycera. We follow the terminology for mouthparts that was proposed by Clausen & Cook (1971) and Zatwarnicki & Ryczko (2014). We use the term basal flagellomere for the large antennomere beyond the pedicel. We prefer this term over "first flagellomere" as there may be more than one flagellomere involved, and basal does not imply a number or numbers. We likewise do not use "postpedicel" (Stuckenberg, 1999) for this antennomere because at least the multisegmented arista is beyond the pedicel in addition to the large antennomere, and postpedicel is thus ambiguous and lacking precision. Terminology for structures of the male terminalia is provided directly on Figs 16-19. Species descriptions are composite and are not based solely on the respective holotypes.
External morphological characters were drawn using an ocular grid attached to a stereoscopic dissecting microscope. Internal genitalia features were drawn using a camera lucida with a Nikon Labophot2 and were rendered using the graphic software Micrografx Designer 7.0. Photographs of habitus were taken with a Canon EOS 7D (18 Mp) camera with a zoom objective 70-200/f4 and a Nikon 10×/0.3 Plan Fluor microscopic objective. The images were then combined, using licensed stacking software, Zerene Stacker.
Five head and two venational ratios that are used in the descriptions are defined below (all ratios are based on three specimens (the largest, smallest, and one other): Costal vein ratio: the straight-line distance between the apices of R 2+3 and R 4+5 /distance between the apices of R 1 and R 2+3 ; M vein ratio: the straight-line distance along vein M between crossvein dm-cu and r-m divided by the distance apicad of crossvein dm-cu; Antennal ratio: length of antenna/longest diameter of eye; Basal flagellomere ratio: length/width (sometimes referred to as "height"); Gena-to-eye ratio is the genal height measured at the maximum eye height divided by the eye height; Labellar ratio: length of labellum divided by length of ventral head margin; Medial-to-lateral vertical setal ratio: length of medial vertical seta divided by length of lateral vertical seta.
This study was based on specimens from the following museums (with acronyms used in the text and names of curators and assistant persons given in parentheses):  Enderlein, 1934: 191. Type genus: Heringium Enderlein, 1934(= Clanoneurum Becker, 1903. Clanoneurinae Enderlein, 1936: 168 Diagnosis. The tribe Psilopini is distinguished from Discomyzini and other tribes of Ephydridae by the following combination of characters: Small to moderately small shore flies, body length 1.50-2.50 mm; usually mostly black, shiny, microtomentum usually sparse, especially on frons and scutum. Head: Ocellar seta aligned behind anterior ocellus, sometimes only slightly so; reclinate fronto-orbital seta inserted behind, sometimes weakly developed or lacking, lateroclinate (sometimes well developed, divergent, and slightly proclinate), usually much less than ½ length of ocellar seta; arista with 7-14 dorsal rays; conformation of face variable, usually mostly smooth, sometimes distinctly carinate, shallowly pitted or rugose; gena, including midportion, setulose, its posterior margin rounded, mouthparts sometimes moderately to distinctly geniculate, length of labellum equal to haustellum.
Thorax: Mesonotum generally sparsely microtomentose to bare, subshiny or shiny, although variable; supra-alar seta usually lacking, if present, well developed, subequal to postalar seta; arrangement of acrostichal setulae variable; prescutellar acrostichal setae usually present, widely set apart and aligned anterior of posteriormost dorsocentral setae; scutellum usually sparsely setulose; anterior and posterior notopleural setae inserted at about the same level. Wing hyaline to milky white; costal length variable; crossvein dm-cu sometimes lacking.
Discussion. The position of the Risa group of genera within the Ephydridae is supported by the following characters: Head: Postocellar setae absent. Thorax: Anal vein atrophied; hence anal cell absent. Abdomen: Spiracles 2-5 included in tergites; abdominal spiracle 7 absent in both sexes; tergites 6 and 7 of male reduced, much shorter than preceding tergite; sternites 6 and 7 of male lost; sternite 8 of male completely lost (Fig. 1).
Structures of the male terminalia of the Risa group include a few characters that are also found apparently in the shore-fly tribe Atissini (Ephydridae: Hydrelliinae). Externally, moreover, there is some resemblance also with specimens of Gymnomyzini (Ephydridae: Gymnomyzinae), and the primitive type of cibarium and lack of presutural dorsocentral setae suggest placement in Discomyzinae or Gymnomyzinae. However, the presence of a subepandrial plate, setose pregonites, which are attached to subepandrial plate, and lack of an ejaculatory apodeme are more typical for Discomyzinae, and like other genera of Psilopini, the pseudopostocellar setae are weakly developed with their length being considerably less than half that of the ocellar setae and the stem of the radial vein lacks setulae.
The tribe Psilopini is closely related to Discomyzini, and these two tribes constitute the subfamily Discomyzinae ( Fig.  1). Worldwide, the tribe Psilopini comprises 14 genera and 159 species (Mathis & Zatwarnicki, 1995 and electronic updates). As noted in the above synonymy, we place the tribe Risini as a junior synonym of Psilopini and refer to the taxa formerly placed in Risini as the Risa group of genera. Like other genera of Psilopini, the pseudopostocellar setae are weakly developed with their length being considerably less than half that of the ocellar setae and the stem of the radial vein lacks setulae.
In this study, we include both external and internal morphological characters, and based on this assessment, we suggest that the Risa group of genera is a monophyletic lineage that is also supported by several hypopygial characters: hypandrium fused with epandrium; subepandrial plate fused with epandrium; fusion of phallapodeme with aedeagus, forming aedeagus-phallapodeme complex; fusion of postgonites to each other; and fusion of postgonites to aedeagus-phallapodeme complex.
Within the tribe Psilopini, the Risa group is closely related to the genus Diasemocera Bezzi. This conclusion is based on the following characters: Seven pseudotracheae ( Fig. 10) and cibarium that are similarly shaped and with an extended central process (Figs 11,12). Other genera within Psilopini, such as Clanoneurum Becker, Cnestrum Becker, and Trimerina Macquart, have seven pseudotracheae, but these genera have a simple cibarium. Only Diasemocera has an elongated cibarial process (Zatwarnicki, 2018).
The other genera referred to above (Clanoneurum, Cnestrum, Diasemocera, and Trimerina) also have an association with the plant family Amaranthaceae. For example, D. leucostoma (Meigen) has been reared from Beta vulgaris L., and most species of Diasemocera breed in saline habitats where plants of the genera Chenopodium L., Seidlitzia L., and Suaeda Forsskål proliferate. We acknowledge, however, that specific host-plant information is not presently available for many species.  2); mouthparts geniculate, with long haustellum and rather short labellum (Figs 5-7); wing venation normal (Fig. 8).
Antenna moderately elongate (Figs 5-7); antennal ratio about 0.4; basal flagellomere ratio 2-2.5; pedicel lacking a well-developed, spine-like seta anterodorsally; pedicel elongate with lobe on medial side of dorsal seam more prominent than on lateral side; distal articular surface less deeply concave, lacking any suggestion of a cup-like cavity; conus relatively narrow but strongly projected, arising close to medial dorsal lobe; foramen faces laterally; basal flagellomere has no basal stem, basal foramen on a slight scabrous prominence on lateral margin of basal hollow, hollow capacious tilted medially, and its floor with a narrow caecum-like extension; pore of sacculus in a ventral position; beyond base of arista; arista about as long as antenna, rather thick at base, bearing short dorsal hairs. Face vertically carinate (Figs 5, 7), carina produced toward ventral margin as a rounded point; 1 large facial seta near ventral margin; ventral margin of face concave. Clypeus exposed in ventral facial emargination. Gena low, gena-to-eye ratio about 0.8; 1 large genal seta. Eye vertically oval, higher than width, appearing bare of interfacetal setulae. Proboscis geniculate but with length of labellum slightly more than half haustellum; 7 pseudotracheae ( 6-7 ventral sensillae (Figs 11, 12); palpus black; lacinia digitiform, bearing numerous setulae ventrally (Fig. 10).
Etymology. Notorisa is a combination of "notios," which is Greek for southern, and Risa, which is also of Greek derivation and means nose (Becker, 1907:404), in reference to the pronounced facial carina.

Distribution.
Notorisa is known to occur thus far only on the island continent of Australia, where it is reasonably widespread (see below).
Discussion. This distinctive genus was first noticed by D. K. McAlpine (2002, 2011, = his "Genus E"), who showed us specimens while on a visit to the Australian Museum. David had collected several specimens and had accumulated others, which he set aside as an unusual and probably undescribed genus of shore flies. He had labelled the series with "? Ephydridae." By good fortune and coincidence, we had begun a study of Risa a few months before our field work in Australia and we immediately recognized a close resemblance with Risa. We further observed that in many features Notorisa is intermediate between Risa and other genera of Psilopini, especially those that have an association with Amaranthaceae. Our subsequent, more detailed studies have confirmed that this similarity has a phylogenetic basis, reflecting our proposal of a sister-group relationship.
As noted on the cladogram (Fig. 1) and generic diagnosis above, Notorisa is characterized by having a digitiform lacinia that bears numerous setulae ventrally (Fig. 10). The finger-like lacinia is a synapomorphy for Notorisa. Other genera in Psilopini, including species of Risa, have a narrowed and tapered lacinia.
Abdomen: Structure: Sternite 3 rectangular with rounded corners; sternite 4 mostly rectangular, anterior corners flat, making anterior margin trapezoidal; Coloration: Strongly and generally bronzish black to black. Male sternites 3-5 longer than wide, sternites 3-4 mostly rectangular, parallel sided; sternite 5 with width of posterior margin twice that of anterior margin, posterior margin shallowly concave, closely associated with anterior margin of hypandrium. Male terminalia  as in generic description with these details: Epandrium well developed, in lateral view (Fig. 17) higher than wide, widest at midheight, anterior margin angulate, posterior margin shallowly curved, ventral margin narrowly rounded, in posterior view ( Fig.  16) U-shaped; cercus (Fig. 17) in lateral view long and narrow, at least 4 times higher than wide, shallowly curved, in posterior view (Fig. 16) dorsal margin slightly recurved, pointed medially, ventral margin truncate, medial margin with dorsal ⅔ shallowly concave; gonostylus (these paired structures could be terminal portion of epandrium that is now separated by an unsclerotized gap) free at ventral margin of epandrium, in lateral view (Fig. 17) slender, shallowly curved, bearing 5-7 setulae, in posterior view (Fig. 18) longer than wide, apical ⅔ tapered to narrowed apex; pregonite a well sclerotized structure, in lateral view (Fig. 19) with basal and apical position both evenly tapered to acute point, both slightly curved, apical half recurved, bearing a few setulae in recurved margin, apical ⅓ digitiform, bearing tiny setulae along shallow concavity, in ventral view (Fig. 18) with medioapical margin truncate, thereafter toward base somewhat rectangular, basal portion narrowed to acute, basal apex; subepandrial plate in lateral view (Fig. 19) narrow, basal ⅔ nearly straight, apical third angled posteriorly, in ventral view (Fig. 18) irregularly quadrate, lateral sides narrow, angulate, posterior margin with lateral angles produced laterally as short, tapered, acutely pointed; anterior margin widely and evenly truncate; secondary process attached to latero-anterior margin of subepandrial plate robustly digitiform, longer than wide, bearing 9-11 short setulae; aedeagus and phallapodeme fused, aedeagal portion in lateral view (Fig. 19) irregularly triangular, tapered to point, phallapodemal portion gently arched posteriorly, broadly rounded apically, with a short, digitiform projection at midheight, in ventral view (Fig.  18) aedeagus broadly oval, apical margin broadly rounded, posterior narrowed, phallapodemal portion dagger-like; postgonites ovate, bearing 4-6 setulae; hypandrium in lateral view (Fig. 19) with length over twice width, anterior margin rounded, posterior margin irregularly linear with angulate depressions. Female: tergites 6-7 bearing several setulae along posterior margin, but not fringe-like; tergite 8 less sclerotized dorsolaterally; sternite 8 more or less pentagonal (length/width ca 0.7) with obtusely pointed apex directed anteriorly (Fig. 13), more strongly sclerotized along anterior margin, with anteromedial protrusion that bears 12-13 short, stout, black, sometimes knob-like setulae; posterior margin very shallowly concave; posterior and laterad of protrusion bearing numerous setulae with 1 larger, submarginal pair; subanal plate present, shallowly V-shaped, bearing some setulae; ventral receptacle with a helmet-shaped operculum and curved stalk, not bearing ventral appendages (Fig. 14); cerci short, rounded, bearing several setulae and 1 large pair of setulae inserted anteriorly (adjacent to a structure that appears like a double setal base without a seta); medial margin of cerci sharply defined.  Etymology. The species epithet, mcalpinei, is a genitive patronym to honor David K. McAlpine, collector of most of the type series. We take great pleasure in recognizing David's many contributions to the study of Diptera, especially families within Schizophora.

Distribution and natural history. Risa is strictly an
Old World genus, now comprising ten described species. The geographic range for the genus extends from Spain and Morocco eastward across North Africa and Israel to Mongolia in the far east. The genus seems to be restricted to arid regions where it is associated with host plants of the family Amaranthaceae.
Discussion. The female terminalia of Risa are peculiar and are a primary source of evidence, supporting the inclusion of this genus in the Ephydridae. This combination of characters is not known to occur in any other family of Schizophora. In freshly pinned material it is possible to see a row of white hairs that requires high magnification for clarification of its structure (SEM?). The structure and sclerotization of the female terminalia are such that even in shriveled specimens the apex of the abdomen is exposed and more or less in a natural position. In shriveled males, the lateral margins of the abdomen fold such that the abdomen appears triangular and the terminalia are hidden behind the folds.
The monophyly of Risa is corroborated by the following characters that are mostly synapomorphies (many of these characters were identified previously by Griffith, 1972;D. K. McAlpine, 2002D. K. McAlpine, , 2011: Head: Reclinate fronto-orbital seta reduced or lacking; pore of the sacculus in a ventral position on the basal flagellomere; face distinctly carinate, sometimes carina somewhat pointed; mouthparts moderately to distinctly geniculate; elongate prementum of the proboscis; labellum equal to haustellum. Thorax: Wing often milky white; costa short, extended to vein R 4+5 ; subcosta fused with R 1 distally; crossvein dm-cu absent. The proboscis in species of Risa is elongated, which is apparently an adaptation to nectar feeding as adults. Similar adaptations of the proboscis have developed in the Milichiidae, hence the placement of Risa by some authors in this family (see illustrations and descriptions in the seminal work of Brake, 2000).  ( Subgenus Achaetorisa Papp Achaetorisa Papp, 1980: 420, 421  Diagnosis. The subgenus Achaetorisa is distinguished from the nominate subgenus Risa by the following combination of characters: Head: Face with a medial carina, in lateral view carina rounded. Thorax: Prescutellar acrostichal setae absent, only setulae present; scutellar disc lacking setulae. Costal section III (between R 2+3 and R 4+5 ) bearing approximately 10-35 setulae along anterior margin.

Discussion.
Achaetorisa is given status here as a wellcorroborated, monophyletic subgenus. As a terminal clade, however, its recognition as a subgenus leaves the nominate subgenus as a paraphyletic group that includes, among other lineages, Achaetorisa. Papp (1980) described Achaetorisa and distinguished it from Risa using several characters. The discovery of additional species of Risa, and thus an enhanced opportunity to study critically characters for this group, revealed that Achaetorisa brevicornis is congeneric with other species of Risa and that Achaetorisa should therefore be consigned to subgeneric status within Risa. Papp (1980), however, did not emphasize the distinctive and paired ventral appendages of the female ventral receptacle in R. brevicornis. These appendages appear to be unique to the Achaetorisa group.

Figs 46-56
Diagnosis. This species is distinguished from congeners by the following combination of characters: Small shore flies, body length 1.14-1.50 mm.

Figs 57-64
Diagnosis. This species is distinguished from congeners by the following combination of characters: Small shore flies, body length 1.40-1.56 mm.