Taxonomic revision of the assassin-fly genus Microphontes Londt , 1994 ( Insecta , Diptera , Asilidae )

The genus Microphontes Londt, 1994 (Diptera: Asilidae: Brachyrhopalinae) is revised. Currently, three species are known from Namibia and western South Africa, i.e. Microphontes megoura Londt, 1994 from north-western South Africa, Microphontes safra Londt, 1994 from Namibia and Microphontes whittingtoni Londt, 1994 from western South Africa. Four new species, Microphontes ericfisheri sp. n. from the Little Karoo of South Africa, Microphontes gaiophanes sp. n. from the Namib desert of Namibia and Microphontes jasonlondti sp. n. and Microphontes kryphios sp. n. from western South Africa, are described. Distribution, occurrence in biodiversity hotspots sensu Conservation International and seasonal incidence are discussed. Descriptions/redescriptions, photographs and identification keys are provided and made openly accessible in data repositories to support future studies of the included taxa. An unusual flight pattern of male Microphontes gaiophanes sp. n. is discussed. A unique morphological feature on tergite 8 of Microphontes females, termed postero-paramedian T8 pores, is described, illustrated and discussed.


Introduction
Microphontes Londt, 1994 is endemic to Namibia and western South Africa (Londt 1994) and flies inhabit sandy habitats along rivers or on sand dunes (Figs 1-4).Microphontes was initially based on three species known from 18 specimens, i.e.Microphontes megoura Londt, 1994 from north-western South Africa, Microphontes safra Londt, 1994 from Namibia and Microphontes whittingtoni Londt, 1994 from western South Africa (Fig. 5).It had not been studied in any detail since then and was only included in updated identification keys to the Afrotropical Stenopogoninae by Londt (2013) and Asilidae by Londt and Dikow (2017) and in reference to the description of the genus Antiscylaticus Londt, 2010 from western Africa (Londt 2010).
Recently, specimens of a small assassin-fly species belonging to Microphontes were collected by the junior author on sand dunes on the eastern edge of the Namib Desert in Namibia (Figs 1-2) and that initiated a closer look at this genus.Furthermore, specimens had accumulated in natural history collections that extended the range and are here recorded for the first time (Fig. 5).
We present a taxonomic revision of Microphontes based on 44 specimens from 14 collecting events in desert to semi-desert localities in western South Africa and Namibia and describe four new species.

Material and methods
Morphological features were examined using a Leica MZ8 with LED goose-neck lights and a Zeiss SteREO Discovery.V12 stereo microscope with an LED ring-light.Wing length is measured from the tegula to the distal tip of the wing.The female and male terminalia were first excised and macerated in 10% potassium hydroxide (KOH) at 55 °C followed by neutralisation in acetic acid (glacial, CH 3 COOH) and rinsing in distilled water (H 2 O).They were temporarily stored in 75% ethanol (C 2 H 5 OH) for examination and illustration and sealed in polyethylene vials containing 100% glycerine (C 3 H 8 O) and attached to the specimen's pin.

Species descriptions and re-descriptions
Species descriptions are based on composites of all specimens and not exclusively on the holotype and are compiled from a character matrix of 218 features and 359 character states assembled with Lucid Builder (version 3.5) and exported as natural-language descriptions.These species descriptions have been deposited in the Zenodo data repository and can be accessed in XML-format following the SDD (Structure of Descriptive Data) standard.The structure of wings and terminalia is only described once for the genus (with a few exceptional characteristics) and additional species-specific features should be interpreted from the provided photographs.The description of the pubescence pattern, particularly of the abdomen, is based on the examination of the specimens with an LED ring-light while the photographs were taken with a twin-flash (more similar to the lighting with two goose-neck lights).Care needs to be taken to study the pattern in both dorsal and lateral view.All taxon names have been registered in ZooBank (Pyle and Michel 2008).

Specimen occurrence data
Specimen occurrence data were captured in a custom FileMaker Pro database and the following data are exported and presented for each specimen (where available): country, state/province, county, locality, geographic co-ordinates (formatted in both degrees minutes seconds and decimal latitude/longitude), elevation (in metres), date of collection (format: yyyy-mm-dd), habitat information, perching behaviour, sampling protocol (if other than hand netting), collector, catalogue number (a unique specimen identifier and any other identifying number), depository (institution code), number of specimens and sex.Each specimen is listed with a unique specimen identifier (either an institutional catalogue number or an AAM-XXXXXX number used by the junior author) that will allow the re-investigation as well as provide a unique Life Science Identifier (LSID).The occurrence of all species is illustrated in distribution maps plotted with SimpleMappr with all of those localities for which co-ordinates are available.Type localities are plotted with a square symbol while all other specimens are plotted with a circular symbol.The distribution map includes Biodiversity Hotspots sensu Conservation International (Mittermeier et al. 1998;Myers et al. 2000;Mittermeier et al. 2005).The specimen occurrence data are deposited as a Darwin Core Archive (DwC-A) in the Global Biodiversity Information Facility (GBIF) using the Integrated Publishing Toolkit (IPT) at the NMNH.

Photographs
Whole habitus photographs of pinned specimens were taken using a GIGAmacro Magnify 2 (http://www.gigamacro.com/gigapixel-macro-imaging-system/)system, a Canon EOS D5 full-frame DSLR, a Canon MP-E 65 mm f2.8 macro-lens and illuminated by a twin-flash.Individual RAW format images were stacked using HeliconFocus Pro (version 6.7.1) utilising Method C (pyramid) and exported in Adobe DNG-format.In some instances, greasy specimens were placed in acetone ((CH 3 ) 2 CO) for 24-48 hours for degreasing to enhance the study of pubescence and image quality (see, e.g. .
Detailed photographs of morphological features were taken using a Zeiss Ste-REO Discovery.V8 stereo microscope, a Zeiss PlanApo S 1.5× lens, illuminated with visiLED ring-lights and a transillumination base and an attached Olympus OM-D E-M1 Micro Four Thirds camera.Individual RAW-format images were stacked using Affinity Photo with default settings.
All photographs have been deposited in Morphbank Biological Imaging.These images will be automatically harvested by the Encyclopedia of Life (EOL) and are available under the respective species page.

Key
The dichotomous, interactive key has been built with Lucid Phoenix and can be accessed on Lucidcentral and the junior author's research web-site.

Institutions providing specimens
Institutions providing specimens are listed below, together with the abbreviations used in the text when citing depositories (institutionCode) and the people who kindly assisted: INHS Illinois Natural History Survey, Urbana-Champaign, IL, USA (Thomas McElrath); NMSA KwaZulu-Natal Museum, Pietermaritzburg, South Africa (Tricia Pillay); NMNW National Museum of Namibia, Windhoek, Namibia; SAMC Iziko South African Museum, Cape Town, South Africa (Aisha Mayekiso); USNM National Museum of Natural History, Smithsonian Institution, Washington, DC, USA; ZMUC Natural History Museum of Denmark, Copenhagen, Denmark (Thomas Pape).

Microphontes ericfisheri sp. n.
http://zoobank.org/483F65F1-9050-4227-B601-793BEA4A377CFigs 6-8, 56-57, 71 Etymology.The species is named after Eric M. Fisher who is one of the most knowledgeable Nearctic and Neotropical Asilidae taxonomists, present and past, to recognise his contributions to the study of assassin flies.This species was unveiled at the 9 th International Congress of Dipterology (25-30 November 2018) in Windhoek, Namibia during the Asilidae symposium organised to honour a contemporary colleague, Jason Londt and entitled, "Taxonomy and phylogeny of Asilidae -honouring 40 years of Afrotropical research by Jason Londt" on 27 November 2018.
Diagnosis.The species is distinguished from congeners by the distinctly wider than long and transversely rectangular abdominal tergites, the entirely setose postpronotal lobes and dorsal and posterior anepisternum (no macrosetae), the brownpubescent appearing abdomen (in dorsal view), features of the male terminalia such as the long postero-median projection on the hypandrium and shape of the gonostyli and distribution in the Little Karoo in southern South Africa.
Antenna: light brown; scape 1.5× as long as pedicel, short and long yellowish setose and macrosetose ventrally; pedicel short yellowish setose dorsally and long yellowish macrosetose ventrally; postpedicel and stylus broken off.
Distribution, biodiversity hotspots, phenology and biology.Known only from the type locality in the Little Karoo of South Africa (Fig. 71).A rarely collected species known only from a single specimen and collecting event in 2015 (Table 1).The species is endemic to the Succulent Karoo biodiversity hotspot.Adult flies are active in mid summer in a winter rainfall region (Table 2).Nothing is known of the biology.
Diagnosis.The species is distinguished from congeners by the more or less square abdominal tergites, the short macrosetose dorsal anepisternum, the extensively macrosetose ante-and postpronotum, the overall brown colouration and the grey pubescent female abdominal tergite 8.
Antenna: light brown, lightly grey pubescent; scape 1.5× as long as pedicel, short and long yellowish setose and macrosetose ventrally; pedicel short yellowish setose ventrally and dorsally; postpedicel cylindrical (same diameter throughout), 1.5× as long as scape and pedicel combined, asetose; stylus comprised of 1 element, 0.28× as long as postpedicel, asetose; apical seta-like sensory element situated apically on stylus.Table 2. Seasonal incidence of Microphontes species through number of specimens collected and unique collecting events in each month (data given as # specimens/# collecting events when more than one specimen has been collected).Months abbreviated starting with July.
Distribution, biodiversity hotspots, phenology and biology.Known only from the type locality (Fig. 72).A rarely collected species known only from a single collecting event in 2017 (Table 1).Adult flies are active in summer in an arid region on the eastern edge of the Namib Sand Sea (Table 2).Not known to occur in any biodiversity hotspot.
Biological data were gathered during observations in the field at the type locality.Male flies were more active and flew in a very specific pattern of an up-and-down flight, similar to a roller coaster, around the grass boulders or vegetation (see habitat in Figs 1-2).Furthermore, the male flies would land on the sand in open spaces between the grass boulders and perch or rest with their pro-, mes-and metathoracic legs held sideways and up (see Fig. 9).The females appeared less active and were observed to perch or rest on dry, low vegetation close to the grass boulders (Fig. 10, only a single female was collected perching on sand).

Microphontes jasonlondti sp. n.
http://zoobank.org/9CFF13BC-0398-40A7-AE83-AC0035309201Figs 26-31, 58-59, 71 Microphontes whittingtoni Londt, 1994 (in part) Etymology.The species is named after Jason G.H. Londt who is without doubt the most knowledgeable Afrotropical Asilidae taxonomist, present and past, to recognise his contributions to the study of assassin flies and who also collected the type series.This species was unveiled at the 9 th International Congress of Dipterology (25-30 November 2018) in Windhoek, Namibia during the Asilidae symposium organised in his honour entitled, "Taxonomy and phylogeny of Asilidae -honouring 40 years of Afrotropical research by Jason Londt" on 27 November 2018.
Diagnosis.The species is distinguished from congeners by the only slightly transversely rectangular abdominal tergites, the general brown colouration and a long gonocoxite extending beyond midpoint of the epandrium (in lateral view) in the male terminalia.Description.Head: wider than high, brown; vertex and compound eyes at same level; facial swelling indistinct, only lower facial margin slightly developed, silver pubescent; mystax white macrosetose, restricted to lower facial margin, short, reaching tip of proboscis; ommatidia of same size; postgena posterior margin simple, smooth; frons (at level of antennal insertion) slightly diverging laterally, grey pubescent, light brown setose; ocellar tubercle greyish-brown pubescent, light brown setose; vertex brown pubescent, yellowish to light brown macrosetose; median occipital sclerite (m ocp scl) with several yellowish macrosetose; postocular (pocl) setae slightly angled anteriorly distally, yellowish macrosetae; occiput predominantly light brown pubescent, yellowish setose; compound eye posterior margin (in lateral view) straight or slightly curved throughout.
Material examined.SOUTH AFRICA: Northern Cape: 1♀ Renoster River, 24 km N Sutherland, 32°12'18"S, 020°41 '41"E, 1290 m, 1998-11-11, hard  Distribution, biodiversity hotspots, phenology and biology.Known only from three localities in the Northern Cape of South Africa (Fig. 71).A rarely collected species known only from three collecting events in 1986 and 1998 (Table 1).The species is primarily distributed in the Succulent Karoo biodiversity hotspot, but also occurs in the Nama Karoo outside of any hotspot (Fig. 71).Adult flies are active in early summer in a winter rainfall region (Table 2).Biological data available pertain to the habitat preferences.Specimen occurrence data indicate that the species occurs in habitats near rivers.
Remarks.The male specimen from near Williston (NMSA-DIP-4764) was studied by Londt (1994) and assigned paratype status of the then newly described species M. whittingtoni.The male terminalia were dissected, but not illustrated by Jason Londt and, when we studied the specimen and terminalia attached in a micro-vial, it became clear that the male terminalia of the holotype of M. whittingtoni (NMSA-DIP-4777) do not correspond to the terminalia of this specimen.In particular, the holotype of M. whittingtoni has a short postero-median projection on the hypandrium (Fig. 66) whereas this specimen has a very long hypandrial projection (Fig. 58, see also Remarks under M. kryphios sp.n.).
Diagnosis.The species is distinguished from congeners by the small size with a wing length of 3-4 mm, distinctly wider than long and transversely rectangular abdominal tergites, partly macrosetose postpronotal lobes, setose dorsal and posterior anepisternum and features of the male terminalia such as the long postero-median projection on the hypandrium and shape of the gonostyli.
Antenna: orange, lightly grey pubescent; scape approximately as long as pedicel, short and long yellowish setose and macrosetose ventrally; pedicel short yellowish setose ventrally and dorsally; postpedicel cylindrical (same diameter throughout), approximately 2× as long as scape and pedicel combined, asetose; stylus comprised of 1 element, 0.15× as long as postpedicel, asetose; apical seta-like sensory element situated apically in cavity on stylus.
Distribution, biodiversity hotspots, phenology and biology.Known from three nearby localities in the Northern Cape of South Africa (Fig. 71).A rarely collected species known only from three collecting events (Table 1).Adult flies are active in early summer in a winter rainfall region (Table 2).The species occurs within the Succulent Karoo biodiversity hotspot, but also outside of it in the Nama Karoo biome.Biological data available pertain to the mating position and habitat preferences.A pair has been collected in copula (NMSA-DIP-4766) and remarkably the flies remained in the copulation position after capture in a tail-to-tail position.Specimen occurrence data indicate that the species prefers sandy habitats along rivers.
Remarks.The holotype and paratype of M. kryphios sp.n. were studied by Londt (1994) and assigned paratype status of the then newly described species M. whittingtoni.The male holotype of M. kryphios sp.n. was dissected and the male terminalia illustrated by Jason Londt (see Figs 68-70) to represent M. whittingtoni.When we studied the specimen and terminalia attached in a micro-vial, it became clear that the male terminalia of the holotype of M. whittingtoni (NMSA-DIP-4777) do not correspond to the illustrations published in figures 51-53 by Londt (1994), which were based on the 23 km N Middelpos specimen that we assign to be the holotype of M. kryphios sp.n.In particular, the holotype of M. whittingtoni has a short postero-median projection on the hypandrium (Fig. 66) whereas the holotype of our new species M. kryphios sp.n. has a very long hypandrial projection (Fig. 68, see also Remarks under M. jasonlondti sp.n.).Londt, 1994 http://zoobank.org/5CEE3768-26EB-4BCA-8629-E26423B0F7D9Figs 38-43, 60-61, 71 Microphontes megoura Londt, 1994: 87.Diagnosis.The species is distinguished from congeners by the small size with a wing length of 3.2-4.1 mm, distinctly wider than long and transversely rectangular abdominal tergites, partly macrosetose postpronotal lobes, macrosetose posterior anepisternum and distribution in Namaqualand in western-most South Africa.
Distribution, biodiversity hotspots, phenology and biology.Known only from the type locality in the Northern Cape of South Africa (Fig. 71).A rarely collected species known only from a single collecting event in 1936 (Table 1).The species is en-demic to the Cape Floristic Region biodiversity hotspot.Adult flies are active in early summer in a winter rainfall region (Table 2).Nothing is known of the biology.Londt, 1994 http://zoobank.org/0F1497C1-868C-48B7-8816-D6415CC1ED97Figs 3-4, 44-49, 64-65, 72 Microphontes safra Londt, 1994: 87.Diagnosis.The species is distinguished from congeners by the more or less square abdominal tergites, the short macrosetose dorsal anepisternum, the setose ante-and postpronotum, the overall orange colouration and white pubescence and the apubescent female abdominal tergite 8.
Material examined.NAMIBIA: Hardap: 2♀ 1♂ Aandster Farm, 25°21'34"S, 016°06 '04"E, 1000 m, 1974-02-16, vegetated  Distribution, biodiversity hotspots, phenology and biology.Known from two localities in Namibia (Fig. 72).A rarely collected species known only from three collecting events (Table 1).Adult flies are active in summer in a summer rainfall region on the eastern edge of the Namib Sand Sea (Table 2).Not known to occur in any biodiversity hotspot.Biological data available pertain to the habitat and perching behaviour.All specimens have been collected on partly vegetated sand dunes (Figs 3-4) and data for one specimen indicate that the species perches directly on the sand.
Remarks.A photograph of this species was erroneously included by Londt and Dikow (2017) (their figures 91-92, specimen USNMENT00832231) under the genus Afroholopogon Londt, 1994, which was a misidentification by T. Dikow at the time.The original type series of M. safra collected on the Aandster farm (now part of the NamibRand Nature Reserve) originates from two collecting events on a joint expedition by Drs Mike Irwin (3 specimens) and Leif Lyneborg (1 specimen).These specimens were collected on separate dates and, at separate sites on the Aandster farm as co-ordinates in the quarter-degree system (Larsen et al. 2009), were provided on the label as 2516Ac (M.Irwin) and 2515Bd (L.Lyneborg).Since it is impossible to decipher the exact location of the dunes from these co-ordinates, we plot a single locality (25°21'34"S, 016°06'04"E) for all specimens.
Remarks.The boundaries of this species have been redefined based on the study of all primary and secondary type specimens originally studied by Londt (1994) and additional specimens collected in the south-western Northern Cape Province.We postulate that M. whittingtoni represents a group of morphologically similar species and we have split the type series into three species.Londt (1994) cleared the male terminalia of two paratype specimens and illustrated one of them in his figures [51][52][53].The holotype of M. whittingtoni does not have the long postero-median projection on the hypandrium (see Fig. 66) that was illustrated by Londt (1994) based on the paratype from 23 km N Middelpos (Fig. 68), which now represents the holotype of M. kryphios sp.n. (see also Remarks under M. jasonlondti sp.n.). of Southern Africa (Table 1,Figs 5,(71)(72).Species of Microphontes are very similar in external morphological features and are difficult to distinguish.Several specimens studied were greasy making identification difficult as the pubescence pattern could not be properly studied.A single specimen that appears to be teneral has been studied that could represent an undescribed species.The distribution of this specimen is plotted in Fig. 71 and the occurrence data are as follows: 1♂ South Africa: Western Cape: Willie Nel Farm, 11 km W Clanwilliam, 32°10'47"S, 018°53'29"E, 1999-11-30, Malaise trap, Parker, F. (INHS-Diptera-3932, INHS).

Male flight pattern in M. gaiophanes sp. n.
The peculiar flight pattern of M. gaiophanes sp.n. males with a roller coaster-like up-anddown movement in open spaces on partly vegetated sand dunes (Figs 1-2) is not used to capture prey although it should be noted that no prey records exist.Our hypothesis is that this pattern is an advertising flight to attract females, which are perching or resting close to the larger vegetation on the sand dune.Lavigne and Holland (1969) and Dennis and Lavigne (1975) describe a very similar undulating flight behaviour, termed searching flights, for Efferia staminea (Williston, 1885) and Proctacanthus micans Schiner, 1867 (both Asilinae) and Scleropogon neglectus (Bromley, 1931) (Stenopogoninae), for example.It has more recently been described for Diogmites crudelis Bromley, 1936 (Dasypogoninae) by Dennis (2015).These published observations include three different subfamily taxa of rather large assassin flies, but this behaviour has not been observed in smaller robber flies possibly due to the difficulty in following smaller flies in the field (pers.comm.S. Dennis).

Morphological similarities
Species are, in general, morphologically very similar, but can be placed in two groups.M. gaiophanes sp.n. and M. safra are distinct in a number of features from all other species and are also currently representing the northernmost species.Both species share the square abdominal tergites making the abdomen longer than the wings (Figs 11,15,48), the macrosetose dorsal anepisternum, the laterally macrosetose frons, the ♂ terminalia rotated by 90° (Figs 62,64), the entire development of ♂ T9 and S9 (Figs 17-19) and the ♂ terminalia essentially being closed in ventral view (epandrium, gonocoxite and hypandrium being tightly associated, Figs 63, 65).
: 3.0-5.5 mm, hyaline, evenly microtrichose; C circumambient (developed around entire wing), anterior wing margin straight; R 2+3 distally relatively straight, r 1 open; R 4 terminating anterior to wing apex, relatively straight, stump vein (R 3 ) absent (except in left wing of one specimen); r 4 open, R 4 and R 5 more or less parallel; R 5 terminating posterior to wing apex; r 5 open; M 1 terminating posterior to wing apex; cell d closed by base of M 2 and m-m, M 2 and m-m not aligned, r-m situated either in proximal or distal half or in centre; m 3 open; cua open; alula reduced in size to small lobe; microtrichia on posterior wing margin arranged in a single plane.

Figures 9 -
Figures 9-10.Photographs of M. gaiophanes sp.n. in nature on sand dune as in Figs 1-2: 9 ♂ resting on sand (note that pro-, mes-and metathoracic legs are held sideways and up so that tarsi do not touch sand, Morphbank #861774) 10 ♀ resting on dry vegetation just above ground (#861777).Photographs by T. Dikow.

Table 1 .
Collecting event summary for Microphontes species.