Afrotropical Cynipoidea (Hymenoptera)

Abstract The Afrotropical Cynipoidea are represented by 306 described species and 54 genera in four families: Cynipidae, Figitidae, Liopteridae and Ibaliidae, the latter represented by a single introduced species. Seven of these genera are only represented by undescribed species in the region. Seven new genus-level synonymies, one genus resurrected from synonymy, 54 new combinations, one combination reinstated, and one new replacement name are presented. We provide identification keys to the families, subfamilies and genera of cynipoid wasps occurring in the Afrotropical region (Africa south of the Sahara, including Madagascar and southern Arabian Peninsula). Online interactive Lucid Phoenix and Lucid matrix keys are available at: http://www.waspweb.org/Cynipoidea/Keys/index.htm. An overview of the biology and checklists of species for each genus are provided. This paper constitutes the first contributory chapter to the book on Afrotropical Hymenoptera.


Introduction
e Afrotropical Cynipoidea are taxonomically and biologically poorly known, a situation typical for wasp taxa from this region. e lack of knowledge in cynipoid systematics is exempli ed by the recent revisions that have increased the number of described species (Pycnostigminae: Bu ngton and van Noort 2007; Anacharoides: Bu ngton and van Noort 2009; Stentorceps: Nielsen and Bu ngton 2011) as well as the description of a new genus (Bu ngton 2012). As a consequence of the underdocumentation of the region's diversity, the process of unraveling the biology of the Afrotropical cynipoid wasps is also in its infancy. Some headway has been made with a recent biological study of the cynipid Rhoophilus loewi Mayr, 1881(van Noort et al. 2006) and the discovery of two true indigenous gall formers, Phanacis neserorum Melika and Prinsloo (2007), and Qwaqwaia scolopiae Liljeblad, Nieves-Aldrey & Melika (2011), the latter meriting the description of a new genus and establishment of a new tribe (Liljeblad et al. 2011). However, the majority of the species and generic level diversity within the Afrotropical cynipoids can be found within the eucoiline Figitidae, a monophyletic group that are primary koinobiont endoparasitoids of muscomorphan Diptera (Ronquist 1999, Fontal-Cazalla et al. 2002, Bu ngton et al. 2007. Cynipoid knowledge was rst systematised in a scienti c form in Europe. To the systematic entomologists in Linnaeus's time, most cynipoids were too small to warrant recognition, and it was mostly the large-sized ibaliids and the cynipids (accessible via their obvious galls) that were studied in the early days. e rst more systematical treatment of gitids was that of Westwood in the 1830s, followed by Hartig, Dahlbom, Giraud, omson and Förster, and eventually the three massively productive authors of the previous turn of century: Ashmead, Cameron and Kie er. Of these, Peter Cameron and Jean-Jacques Kie er were given single cynipoid specimens from European expeditions to di erent parts of Africa, and thereby became the rst to treat Afrotropical taxa of Cynipoidea (Cameron 1904, Kie er 1904, 1910a-d, 1911a-b, 1912, 1913. Single Afrotropical taxa were added by Hedicke (1912), Bridwell (1919) and Kinsey (1919). Later, Louis Weld in the United States made a major e ort to summarise and recheck all cynipoid knowledge and revise all classi cations based on type studies. His thoroughness made his compendium of cynipoidology (Weld 1952, with a major appendix: Weld 1962) an important groundwork, which was also applicable for the Afrotropical fauna. Further, single taxa were added by Weld (1944), Risbec (1956), Masner (1960), Belizin (1973) and Dessart (1976).
Two systematists authored major faunistic e orts, describing a large number of Afrotropical Cynipoidea in a comprehensive format. e rst was carried out by P L G Benoit ( , 1956a, investigating the fauna of Congo and Rwanda. e second was John Quinlan, whose work attempted to document the entire Afrotropical region (Quinlan 1979(Quinlan , 1988). Quinlan's papers followed Nordlander's (1982 and papers therein), providing the rst treatment of Afrotropical cynipoids that employed phylogenetic considerations (Quinlan , 1988. More recently there have been only smaller taxonomic group revisions: Ronquist (1995) on Liopteridae (worldwide); Allemand et al. (2002) on Afrotropical Leptopilina; Ros-Farré and Pujade-Villar (2006) on world Prosaspicera; Liu et al. (2007) on Paramblynotus; Bu ngton and van Noort (2007)  Higher-level cynipoid phylogenetics traditionally follows Ronquist (1999), whose data suggested the rare, Australian-endemic Austrocynipidae was sister-group to Ibaliidae + Liopteridae + (Figitidae + Cynipidae), or (Ibaliidae + Liopteridae) + (Figitidae + Cynipidae). Liopterid phylogenetics were investigated by Ronquist (1995) based solely on morphology; the rarity of many groups of Liopteridae precludes the use of molecular data, so this study remains the most thorough and up to date treatment. Cynipidae phylogenetics has been intensely studied by Liljeblad et al. (2008), Rokas et al. (2002), Ronquist and Liljeblad (2001), as well as the doctoral dissertation of Nylander (2004); Nieves-Aldrey et al. (2009) recently described a new tribe of cynipids, while Liljeblad et al. (2011) provided an updated key to cynipid tribes. A preliminary analysis of the phylogeny of Figitidae demonstrated that the family was monophyletic (Ronquist 1999). Subsequent analyses, employing more thorough taxon and character sampling, found the family monophyletic (Bu ngton et al. 2007) and paraphyletic , the latter study having recovered Cynipidae as sister-group to the gitid subfamilies rasorinae + (Pycnostigminae + Mikeiinae). Further phylogenetic interpretation of cynipoids is beyond the scope of this study, however, it should be pointed out that cynipids are very rare in the Afrotropical region, and the gall-associated gitid subfamilies rasorinae, Parnipinae, Mikeiinae and Plectocynipinae have not been recorded to date.
A few species of Cynipoidea are of agricultural importance. e siricid woodwasp, Sirex noctilio Fabricius is detrimental to commercial pine tree plantations in South Africa (Tribe andCillie 2004, van Noort andPicker 2011), and the ibaliid Ibalia leucospoides (Hochenwarth) has been introduced to both Australia and South Africa for biological control (Hurley et al. 2007, van Noort andPicker 2011). e gall wasp Dryocosmus kuriphilus (Yasumatsu) negatively impacts commercial chestnut (Castanea) production, destroying the chestnut itself through galling, and is spreading rapidly throughout Europe (Brussino et al. 2003), Asia (Yasumatsu 1951, reviewed by Melika 2006 and North America (Payne and Anagnostakis 1993). A few eucoiline Figitidae have been evaluated for biological control of pestiferous Diptera, but only Banacuniculus utilis (Beardsley) (formerly Ganaspidium) and Aganaspis daci (Weld) have been actually utilized commercially (Petcherat andJohnson 1988, Wharton et al. 1998). e world cynipoid fauna has not been catalogued since Dalla Torre and Kie er (1910), but in recent years e orts have been made to provide overviews of regional faunas, starting with the European cynipoid catalogue within the Fauna Europaea database project (Ronquist and Forshage 2004). A North American catalog of Eucoilinae  and an Australian catalog of the entire superfamily have been recently published (Paretas-Martínez et al. 2013). e Australian region comprises 37 recorded genera: one each for Austrocynipidae, Ibaliidae and Liopteridae; two for Cynipidae; and 32 for Figitidae (Paretas-Martínez et al. 2013). While ibaliids are species-poor (with 20 species) and largely con ned to the Holarctic, the Liopteridae are relatively rare in all biogeographic regions with the exception of the Oriental region, which supports an unusually high diversity of Paramblynotus species (Ronquist 1995a). e Holarctic region supports a rich gall-wasp fauna (Askew et al. 2013, Nieves-Aldrey 2001, Ronquist and Forshage 2004, Melika 2006), but Eastern (Abe et al. 2007) and Southeast Asia (Tang et al. 2009, Melika et al. 2011a are just beginning to be explored. e Neotropical cynipids are relatively understudied as well, and recent descriptions suggest our knowledge of this region is in its infancy (Medianero et al. 2011, Nieves-Aldrey et al. 2009. Cynipids are very rare in the Afrotropical region, with the recent descriptions of Aylacini (Melika and Prinsloo 2007), Synergini (van Noort et al. 2006) and Qwaqwaini (Liljeblad et al. 2011) providing a glimpse into a hitherto unknown fauna.
Figitidae are likely the most diverse cynipoid group, though the majority of species remain to be described (Nordlander 1984, Ronquist 1999, Bu ngton et al. 2007. Within Figitidae, the Eucoilinae dominate at both the generic and species level of diversity; tropical areas tend to be more species rich, though Holarctic regions can be very rich, including tall-grass prairie and high-elevation deserts in North America (Bu ngton pers. obsv.). Eucoiline species diversity of Australia is relatively depauperate compared to the rest of the world (Paretas-Martinez et al. 2012). Afrotropical species and generic diversity of Cynipoidea is not as high as that of the Neotropics (Bu ngton and Forshage pers. obsv.), but there is one endemic subfamily (Oberthuerellinae, with three genera), a few endemic (Angustacorpa, Pycnostigmus, Stentorceps, and Nanocthulhu) or almost endemic (Anacharoides, Tylosema) genera, as well as species-rich genera (Leptopilina, Rhoptromeris, Afrostilba). Furthermore, much of this species and generic richness is widespread in the Old World Tropics, with numerous species and genera found in Southeast Asia, the Afrotropical region and the Oceanic region (Lin 1988, Bu ngton and Forshage pers. obsv.).
is treatment is part of the initiative to document Afrotropical hymenopteran richness published as a series in ZooKeys (a peer-reviewed, open-access, rapidly produced journal launched to support free exchange of ideas and information in systematic zoology) (http://www.waspweb.org/Afrotropical_Hymenoptera_book/ index.htm). e virtual book, including well-illustrated dichotomous identi ca-tion keys in each chapter will be published as a series of stand alone peer-reviewed scienti c papers with all chapters linked together as a virtual book on the ZooKeys website. e publication will include links to online interactive Lucid dichotomous and matrix based keys hosted on WaspWeb. e goal of this paper is to provide a current synthesis of Afrotropical cynipoid systematics, including an overview of biological associations, and the rst key to cynipoid genera of the Afrotropical region. e development of this resource is aimed to facilitate future research on this ecologically and agriculturally important superfamily of wasps. To this end, we provide keys to all of the genera of Afrotropical Cynipoidea. As part of this overview assessment we present seven new genus-level synonymies, one genus resurrected from synonymy, 54 new combinations, one combination reinstated, and one new replacement name.

Materials and methods
Character states diagnostic of each taxon were discerned from material in extensive recent collections of African Hymenoptera housed at the Iziko South African Museum, Cape Town; California Academy of Arts and Science, San Francisco; National Museum of Natural History (Smithsonian Institution), Washington DC, and Natural History Museum, London. Historically there are important Hymenoptera collections from the region that are housed in a number of European Museums including the Natural History Museums in Paris, Tervuren, and Berlin, to name a few. We have also made use of material housed in the Museum of Zoology of Lund University, Sweden, the Swedish Museum of Natural History, Stockholm; South African National Collection, Pretoria; Biologiezentrum, Linz; and elsewhere. e wealth of recently sampled Hymenoptera residing in African, European and USA museums has been built up over the last 20 years by extensive and rigorous quanti ed and replicated inventory surveys using a wide diversity of collecting methods (Malaise traps, yellow pan traps, sweeping, pitfall traps, Winkler bag extraction of leaf litter, UV light trapping and tree canopy fogging) carried out across large parts of Africa and Madagascar by Simon van Noort (Iziko South Afrian Museum); Brian Fisher and colleagues (Californa Academy of Sciences); Bob Copeland (a liated with National Museums of Kenya and the National Museum of Natural History); Michael Sharkey (University of Kentucky); and John Noyes (Natural History Museum London). e Hymenoptera from these samples provide an unparalleled resource from which the systematics and diversity of Afrotropical Hymenoptera can continue to be elucidated. All the collections where we have been studying Afrotropical Cynipoidea are listed below.
Freshly collected specimens were point-mounted on black or white, acid-free cards for examination (using a Leica MZ9.5, Z16 or M205c stereomicroscope with incandescent and uorescent light sources), photography and long-term preservation. Representative specimens were imaged using the EntoVision multiple-focus imaging system to illustrate diagnostic characters. Methods for generating these photographs follow those in Bu ngton and van . Di used lighting was achieved using techniques summarized in Bu ngton et al. (2005), Kerr et al. (2009) and Bu ngton and Gates (2009). Scanning electron micrographs were generated using a Hitachi TM3000 desktop scanning electron microscope; specimens were coated in 25-30 nm gold-palladium alloy, or imaged uncoated, using 'analysis' voltage, running in 'compo' mode. All new images generated for this project are deposited in Morphbank.
Morphological terminology follows that of Fontal-Cazalla et al. (2002) and Ronquist and Nordlander (1989); cuticular surface terminology follows Harris (1979). Character matrices were generated and edited using Microsoft Excel; matrices were then used as input into Lucid matrix key production (Penev et al. 2009) Online interactive keys were produced using Lucid and Lucid Phoenix meeting the requirements of publishing both static and dynamic interactive keys under an open access model (Penev et al. 2009). All keys were illustrated using high quality annotated images, highlighting diagnostic characters. e images are integrated into the key above each couplet resulting in a user-friendly output. is key format circumvents the requirement of familiarity with morphological terminology associated with a particular taxonomic group, because the characters are visually illustrated making the keys usable by a wide range of end-users including the lay person. ese keys are available at: http://www.waspweb.org/Cynipoidea/Keys/index.htm. End users can choose between three di erent key formats depending on their personal preference. e keys are available in three formats. Although Lucid Phoenix keys are interactive keys they are still dichotomous and a choice needs to be made at each key couplet to continue. Lucid matrix keys, on the other hand, use a di erent approach where relevant states from multiple character features can be selected independently until identi cation is achieved. For more information concerning Lucid keys visit http://www.lucidcentral.org. is publication is available in 4 different formats: (1) high-resolution, full-colour print version, to satisfy the current requirements of the International Code of Zoological Nomenclature (ICZN), as well as for readers who prefer hardcopy, and for the purposes of paper archiving; (2) PDF to provide an electronic version identical to the printed one, to be archived in BHL and PubMedCentral; (3) HTML to provide links to external resources and semantic enhancements to published texts for interactive reading, and (4) XML version based on the TaxPub XML schema to provide archiving document format for PubMedCentral and a machine-readable copy of the contents to facilitate future data mining (Penev et al. 2010b).
e Afrotropical region is a relatively uncontroversial concept ( Fig. 1) conforming to the old Ethiopian region of Sclater and Wallace from the earliest days of zoogeography, with the name changed as of Crosskey and White (1977). As in all the authoritative versions of delimitations, Madagascar and the islands of the western Indian Ocean are included, as is the southern part of the Arabian Peninsula, the South Atlantic islands of Ascension, St. Helena, Tristan da Cunha and Gough, as well as the Cape Verde Islands and the Gulf of Guinea islands (Darlington 1957, Crosskey 1980. Crosskey and White (1977) de ned the northern limit of the region by the 10 inch (254 mm) precipitation isohyet. With climate change and aridi cation this boundary is constantly in a state of ux. Here we use the boundary between the arid and hyperarid climatological zones as de ned by the World Meterological Organization and United Nations Environment Program derived from mean monthly precipitation and potential evapotranspiration surfaces (Desanker and Magadza 2001). is corresponds to the northern edge of the Sahel region bordering with the southern limits of the Sahara desert, a boundary which is usually demarcated by the 150 mm isohyet (White 1983, Fensholt et al. 2013. White (1983) and Linder (2012) further subdivide the sub-Saharan region into cohesive biogeographical subentities based on plant and vertebrate data. Patterns of invertebrate distributions would be expected to correlate with the environmental partitioning of the region.
We have not included major systematic revisionary work in the preparation of this paper, but have implemented taxonomic nomenclatural acts, as required, to bring Figure 1. e Afrotropical region is depicted in green. e northern limits correspond to the boundary between the arid and hyper-arid climatological zones, delimited by the 150 mm precipitation isohyet. the systematic treatment of the Cynipoidea in line with a contemporary assessment of the superfamily. e main rationale behind this initiative is to produce workable, accessible keys to generic level based on current taxonomic knowledge, a contemporary state-of-the-art resource that will be available to facilitate future systematic revisions. With continued ongoing sampling new taxon discoveries are being made all the time and this treatment will be out of date as soon as it is published. However, the online taxon treatment and identi cation keys available on WaspWeb will be expanded and updated as taxonomic progress is made. is is one of the major advantages of electronic output.
All images presented in this paper are available through http://morphbank.net and http://www.waspweb.org using the link to individual collections. Synonyms (for species-level taxa as well as higher taxa) are cited only to the extent they are relevant for the discussion of the knowledge of the Afrotropical fauna. For each genus, we summarise the knowledge of its biology and its worldwide distribution, as well as provide a list of Afrotropical country records with references. In most of the gitids, our knowledge is not yet at the level where the diversity has been systematically treated on the species level. As a result, only a minority of the specimens (and thus of the country records) are assignable to a species name; for these, it is the genus-level treatment which is the basic unit and which contains the most information. Since this is a checklist rather than an actual catalogue, we do not give full references to original descriptions, cite type repositories, full synonymies and nomenclatural histories (except in the cases involving nomenclatural acts made here); we provide standard authorship designation and list the genus if di erent to the original combination in a parenthesis. In the most extensive treatment of Afrotropical taxa before this, Quinlan (1979Quinlan ( , 1988 was sometimes inconsistent in terms of gender agreement in the scienti c names, and a certain number of gender endings are changed accordingly. Further, certain genera have been inconsistently treated in terms of gender throughout scienti c history, and we strive to add consistency here. is project is the result of more than 40 years' worth of cumulative research by all three authors. SvN sampled extensive areas across Africa from 1992 to present for Cynipoidea, supplemented by MB and other initiatives listed above. Management of curation of the extensive recent collections of material in Iziko SAMC was conducted by SvN, of those in USNM by MB. MF spent the last 10 years reviewing generic and species concepts and generating generic keys for Eucoilinae; MB has spent the last 10 years similarly studying the entire Figitidae, but especially the genera of Diglyphosematini. MB and SvN have been publishing on endemic African groups of Cynipoidea for some time, and SvN about the region's Cynipidae. Responsibilities for taxonomic assessment (specimen examination and identi cation of material in said museums as well as numerous international museums, including assessment of generic concepts and delimitation, dichotomous and matrix key formulation and catalogue production) were generally divided as follows: Liopteridae and Figitidae:  (MB, MF and SvN). Imaging of type material and representative taxa in various international institutions was carried out by SvN and MB. New SEM images were done by MB (others downloaded from MorphBank). Lucid key production and image plate production for the keys was done by SvN. Keys to genera of gitids was originally developed by MF; keys to liopterids by SvN and MB. e format and production of this project was conceptualised by SvN in consultation with MB and MF as part of the Afrotopical Hymenoptera Initiative virtual book project (http://www.waspweb.org/ Afrotropical_Hymenoptera_book/index.htm). All authors collaborated on the nal writing and editing of the paper.

List of depositories
We have considered it relevant to list in some detail the collections referred to in this paper, especially in terms of holdings of Afrotropical Cynipoidea. Cynipidae e Cynipidae are represented in the Afrotropical region by three of the world's eight tribes: Aylacini, Synergini and Qwaqwaiini, and four described species. Phanacis contains a number of undescribed species indicating that the genus is likely to be far richer in southern Africa than currently recorded.

MZLU
Biology. Afrotropical cynipids are biologically better known than other African cynipoids and include an endemic, specialist lethal inquiline (van Noort et al. 2007) and gall formers of both herbs and trees (Liljeblad et al. 2011, Melika andPrinsloo 2007).
Distribution. e family is represented in all biogeographical regions with the majority of species occurring in the northern hemisphere (Liljeblad et al. 2008(Liljeblad et al. , 2011. Remarks. Phanacis Förster is closely related to Timaspis Mayr, which is considered to be a junior synonym by some authors (Eady and Quinlan 1963, Melika 2006, Melika and Prinsloo 2007, but not others (Nieves-Aldrey 1994, 2001. Diagnosis. Phanacis is immediately distinguishable from the other cynipid genera present in the Afrotropical region by the open marginal cell, which may be semi-closed (completely closed in both Qwaqwaia and Rhoophilus). e areolet, if de ned, is triangular, but may be inconspicuous or absent in some species. It is large and distinct in both other genera, triangular in Rhoophilus, elongate rhomboidal in Qwaqwaia. Phanacis lacks an occipital carina as in Rhoophilus, but has free metasomal tergites, whereas Rhoophilus has tergites 2 and 3 fused.
Biology. Phytophagous: Phanacis neserorum oviposit in young, soft stems of Chrysanthemoides monolifera (L.) Norl. (Asteraceae), larvae developing in cells in the pith without any visible gall formation. Adults emerge in July and August. Gall formation shows no swelling or external deformation of the stem with surface emergence holes the only indication of infestation (Melika and Prinsloo 2007). e introduced Phanacis hypochoeridis is a gall former in stalks of Hypochaeris radicata L. (Asteraceae) (Melika and Prinsloo 2007).

Rhoophilus Mayr, 1881
Remarks. Rhoophilus is morphologically similar to the Holarctic inquiline genera Synergus, Saphonecrus, and Synophrus, all of which typically attack oak cynipid galls. e transverse ridges of the mesoscutum, and a mesopleuron sculptured with longitudinal ridges in Rhoophilus closely resemble characters in several species of the Synergus/Saphonecrus complex. A sister group relationship between Rhoophilus and these three oak inquiline genera was hypothesized by Ronquist (1994) and Liljeblad and Ronquist (1998). Diagnosis. Rhoophilus has a closed marginal cell, a character shared with Qwaqwaia. e areolet is triangular as in Phanacis, but larger and more distinct, whereas it is elongate rhomboidal in Qwaqwaia. Rhoophilus lacks an occipital carina as in Phanacis (present in Qwaqwaia), but has tergites 2 and 3 fused, whereas these are free in Phanacis.
Distribution. South Africa: Western, Eastern and Northern Cape Provinces. Biology. Inquiline forming secondary cells in galls induced by Scyrotis moths (Cecidosidae) on Searsia (formerly Rhus) (Anacardiaceae) shrubs and trees. e larval cells expand into the hollow interior of the host gall resulting in death of the gall inducing moth larva (van Noort et al. 2007 Remarks. Qwaqwaia has a dorsally wide pronotum, a character shared with members of the Synergini and the Aylacini (especially some species of Phanacis). However, the presence of only two prominent teeth on the right mandible and a parascutal carina that extends anteriorly all the way to the notaulus separate Qwaqwaia scolopiae from species of these two tribes (Liljeblad et al. 2011 Biology. Afrotropical gitids are primarily endoparasitoids of Muscomorpha Diptera, attacking the early instar stages of their hosts and emerging from the host puparium ). e two exceptions to this pattern are the Anacharitinae and Charipinae, of which the former are Chrysopidae parasitoids (Neuroptera), and the latter are hyperparasitoids of braconids and chalcidoids in plant lice (Sternorrhyncha) . Hosts are unknown for the Emargininae and Pycnostigminae, as well as numerous species of Eucoilinae. Ecologically, gitids are mainly associated with three environments: aphid communities on plants, where they attack aphid enemies (Charipinae, Anacharitinae, Aspicerinae); that of plant leafminers, attacking the mining ies (many Eucoilinae); and arguably the most common is various decomposing matter, where saprophagous ies are the hosts (many Eucoilinae, Figitinae).
Distribution. e family is represented in all biogeographical regions, also on subantarctic islands. e largely undescribed tropical faunas are di cult to compare, but eventually the Neotropical fauna might very well turn out to be the richest of all (Bu ngton and Forshage pers. obsv.).  Anacharitinae e Anacharitinae are represented in the Afrotropical region by four genera containing 3 described, and at least one undescribed species. Additional undescribed species are present in world collections.
Distribution. e subfamily is widespread, with many described species in the Holarctic and Neotropics, and fewer elsewhere. In the Afrotropical and Oceanic regions, there are certainly many species undescribed, and they are likely to be present also in the Oriental regions even though not recorded as such yet. Diagnosis. Immediately distinguishable from other Afrotropical anacharitines by the presence of a distinct and complete medial longitudinal furrow containing crosscarinae on the mesoscutum. e ventral part of the mesopleuron is coriaceously sculptured, a unique anacharitine character state.  Remarks. Rare in Afrotropical region. e genus is often di cult to separate from Xyalaspis, and requires revision. Diagnosis. A variable and rather unsatisfactorily circumscribed genus. Some representatives are quite similar to Xyalaspis while some have more of the super cial appearance of small Figitinae. e scutellum may be pointed posteriorly but forms far less of a spine, and is less strongly foveolate so that a circumscutellar carina may follow all the way around the scutellum. Head is less transversal and triangular than in other Anacharitinae.

Key to Afrotropical anacharitine genera
Distribution. Mainly Holarctic, but present locally also in the Neotropical and Afrotropical regions. Afrotropical records: Democratic Republic of Congo, Kenya, Zimbabwe (here).

Species richness.
Only undescribed species in the Afrotropical region, as Kie er's Aegilips capensis (at current state of knowledge) is a Xyalaspis.
Diagnosis. Characteristic anacharitines with elongate, smooth petiole and distinctly transverse and triangular head. e scutellum does not overhang the propodeum; it has more or less reticulate sculpture (never strongly foveolate) and often has a posterior carina that forms a distinct posterodorsal edge. e mesopleura are typically more sculptured than in other Anacharitinae, and the metasoma ends in a more pointed way (the others typically more abruptly).
Distribution. Mainly Holarctic, but also common in Australia and New Zealand, and locally present in Afrotropical region. Afrotropical records: Madagascar (Pujade-Villar 2012), Democratic Republic of Congo, Kenya, South Africa, Uganda (here).

Xyalaspis Hartig, 1843
Remarks. Not common in the Afrotropical region. e genus is often di cult to separate from Aegilips and requires revision.

Diagnosis.
Xyalaspis are often easy to recognise by their very distinct scutellar spines, but several taxa have more blunt or moderate spines that are close to character states found in Aegilips. In these cases, the scutellum of Xyalaspis is characterised by a heavier foveolate sculpture, where no circumscutellar carina is obvious. As currently circumscribed, the genus is somewhat heterogenous in the Afrotropical region, and a distinct species group is characterised by a strongly sculptured mesoscutum with longitudinal carinae as well as strong genal carinae. Biology. Parasitoids of aphidivorous Neuroptera larvae (Fergusson 1986, Miller andLambdin 1985).

Species richness.
Xyalaspis capensis (Kie er, 1912), comb. n. (Aegilips) (South Africa, Zimbabwe) (Type supposedly in ZMBH, but not found there. However, the original description has been deemed su cient for generic placement here) Xyalaspis subsaharica Mata-Casanova & Pujade-Villar, 2014 (Nigeria, Uganda) Several undescribed species from elsewhere in the region. Distribution. e subfamily is widespread. Although the majority of described species are Palearctic this is a biased distribution with both the Nearctic and Neotropical faunas probably being more diverse than the Palearctic. e species numbers in the Afrotropical and Oriental regions are signi cantly lower (but with many undescribed species), while the subfamily has not yet been recorded at all in the Oceanic region. projections. e elongate petiole is somewhat variable within Aspicerinae, though the state in Anacharoides is longer than in most other genera. e only two taxa Anacharoides may be confused with are Callaspidia and Pujadella; both of these latter genera have mesoscutal sculpturing that is remininscent of Anacharoides; however, close examination of the scutellar morphology easily separates these taxa (Bu ngton and van Noort 2009). Callaspidia has not been recorded from the Old World Tropics;

Diagnosis.
is taxon can be di cult to seperate from Prosaspicera. Aspicera hartigi, the only species recorded from the Afrotropical region thus far, has a much shorter scutellar spine than Prosaspicera. Additionally, Aspicera lacks the characteristic inner-orbital carina that Prosaspicera has. Finally, A. hartigi (from Yemen) is distinctly bi-chromatic, with a orange-brown mesoscutum and black mesopleuron; African Prosaspicera, to our knowledge, are all black. Neralsia and Xyalophora are Figitinae and lack the characteristic ligulate T2 of the metasoma that aspicerines have. Furthermore, Neralsia and Xyalophora lack the setiferous pit on the frons, a putative de ning feature of Aspicerinae (Ros-Farré et al. 2000).
Distribution. Mainly Holarctic, but marginally present in the Afrotropical region. Afrotropical records: Yemen (here).
Species richness. Aspicera hartigi Dalla Torre, 1889 (Yemen). is species has also been reported from the UAE (Bu ngton 2010 as Aspicera sp.), Saudi Arabia and Algeria (here)  records (summarized below) as well as phylogenetic analyses suggest this taxon is more closely related to the genera in Aspicerinae, and we maintain that classi cation here. Diagnosis. Distinguished from other Figitidae by the characteristic 'clam-shell' petiolar lamina present on the dorsal half of the petiole. is lamina can often cover the junction between the petiole and the nucha. Super cially, Melanips appears to be a cynipid, but can be distinguished from the Afrotropical cynipids by lacking an areola in the forewing, and by having a dorsally smooth mesopleuron. In addition, Melanips has a setose mesoscutum.
Distribution. Mainly Holarctic but transgressing into the Old World Tropics; in the Afrotropical region found so far only in Kenya and Yemen (here) but expected to show up elsewhere.
Biology. Parasitoid of aphidivorous Brachycera larvae (Evenhuis 1968, Fergusson 1986; label data of several specimens in BMNH (from Kenya and India) records them as reared from Lecopis (Chamaemyiidae), some with host remains.

Prosaspicera Kie er, 1907
Remarks. Revised by Díaz (1979) and by Ros-Farré and Pujade-Villar (2006). Diagnosis. is taxon is most easily confused with Afrotropical Aspicera, but can be distinguished from that taxon by having a much longer scutellar spine, easily as long as the petiole and distinctly overhanging it (much shorter in Aspicera, not overhanging the petiole). Further, Aspicera has not yet been recorded from equatorial Africa, and appears to be restricted to arid portions of Mediterranean Africa and the southern Arabian Peninsula (here). Prosaspicera can, to a lesser extent, be confused with thegitines Neralsia and Xyalophora (all having reasonably well-developed scutellar spines); however, gitines lack the ligulate metasomal T2, and well as the facial impression, and these two characters separate Prosaspicera from gitines with scutellar spines.
Species richness. Prosaspicera antennata (Benoit, 1956c) (Aspicera) (Democratic Republic of Congo, Ethiopia) Prosaspicera optiva Quinlan, 1979 (Democratic Republic of Congo, Ethiopia) Prosaspicera paragicida (Benoit, 1956c)   Diagnosis. ese are the characteristic tiny pale charipines, but in fact they vary a lot in size and colour (also within species!). Colours vary from very dark brown through middle browns and reds to pale yellow, frequently with the head in a paler hue than the rest of the body. e absence of a mesopleural line is the best way to separate them from the otherwise often similar Phaenoglyphis. Dilyta and Apocharips also lack the mesopleural line, but can be separated by their conjoined two apical antennal articles, or by their characteristic metasoma, which is mostly covered by a syntergite and is oval in shape. e metasoma of a representative of Alloxysta typically shows the posterior tergites separate, and is truncated at the end, often with a (cynipid-like) oblique slash. Furthermore, most Alloxysta are larger, paler and more pubescent (on the pronotum, metapleura and coxae) than most Dilyta, and the mesoscutum is smoothly convex (rather than the hump of Dilyta).
Distribution. Worldwide, but most abundant in the Holarctic region. Afrotropical records: Democratic Republic of Congo (Benoit 1956e), Kenya (Evenhuis 1974), Madagascar (Ferrer-Suay et al. 2012), Burundi, Rwanda, South Africa, Uganda, Zimbabwe (Ferrer-Suay and Pujade-Villar 2013) Ethiopia, Ghana, Namibia, Yemen, Zambia (here). Apparently some species are synanthropically widespread, but there is no reason to assume that none of the forms present in Africa are native.
Biology. Hyperparasitoids attacking aphelinid and aphidiine wasps on aphids. (Gutierrez and van den Bosch 1970, Evenhuis passim, Andrews 1978, Fergusson 1986 Remarks. Rare. Diagnosis. Similar to Dilyta, and most easily separated by wing ventation characters: an elongate marginal cell, a subcosta running near mid-width of wing in basal part and having a distinct curve near the basal cross-vein. Other distinguishing characters, easily observed in larger specimens but hardly seen in smaller, is the scutellum with distinct lateral and posterior carinae, and the syntergite where T2 is barely visible as a separate small anterior sclerite. Distribution. Palearctic, Neotropical and Afrotropical. Afrotropical records: Ethiopia (Silvestri 1915 Diagnosis. Mostly tiny and dark brown charipines. Separated from Apocharips by lacking lateral ridges on scutellum, having at most a small posterior ridge, and by usually having T2 discernable as a separate sclerite. May be confused with Alloxysta, but has an oval-rounded metasoma covered in the major part by a syntergite (no posterior tergites visible) and has the two apical articles of the female antenna immovable conjoined. Furthermore, Alloxysta are usually larger, paler, has more pubescence, and have a more smoothly convex mesoscutum; Dilyta have a characteristic anterior hump in lateral view. Distribution
Diagnosis. Similar to Alloxysta, but usually with a habitus more resembling other gitids. Easily recognised among charipines through the possession of a mesopleural carina.
Distribution. Worldwide, but most abundant in Holarctic. Afrotropical records: South Africa (Gaston et al. 2003). Seemingly introduced. Biology. Hyperparasitoids attacking aphelinid and aphidiine wasps on aphids (Kierych 1979;Quinlan and Evenhuis 1980;Fergusson 1986 (Weld 1960, Díaz 1978. Distribution. e subfamily is represented in all biogeographical regions except the Antarctic, but with very few species in the Holarctic and the majority of species occurring in the Afrotropical region (present study).

Remarks.
Emargo and oreauella (and Bothriocynips) were described at di erent times in di erent groups that are now considered di erent gitid subfamilies, obviously with the authors being unaware of the other generic names (Díaz 1978, Girault 1930, Quinlan 1960). e Neotropical Bothriocynips was made a junior synonym of the Neotropical and Afrotropical Emargo by Pujade-Villar et al (2002). However nothing has been suggested to distinguish Emargo from oreauella except distribution (Emargo Neotropical and Afrotropical, oreauella Australian), and there is no morphological evidence to support keeping them separate. Kovalev (1994), within a consideration of broad evolutionary trends in the Cynipoidea, found that some of the characters separating some of the Afrotropical Emargo species in Quinlan's key (Quinlan 1988) were important enough to merit separation on a higher taxonomic level and erected not just new genera but new tribes for two of Quinlan's species from Madagascar (and a new family for the whole group). Ronquist (1995) changed the status of Kovalev's Emarginidae into a subfamily, and made Kovalev's Weldiolini and Quinlaniini synonyms of it, but Kovalev's genera have remained to this date, in spite of being raised in the absence of an actual morphological study as well as of consideration of the global variation. Here we consider that at the current level of knowledge they constitute mere recognisable species or species groups among others within a single genus. us, here we synonymise all genera of Emargininae, and the senior name is oreauella Girault. Most of the resulting new combinations are species hitherto classi ed as Emargo Weld, including the Neotropical type species oreauella eciton (Weld, 1960), comb. n. Bothriocynips was already made a synonym of Emargo but is a new synonym of oreauella, and its Neotropical type species is now oreauella recisa (Díaz, 1978), comb. n. Quinlaniana and Weldiola were both monotypical for taxa from Madagascar and cited in the new combinations T. pexa and T. capito respectively below.
Diagnosis. Members of the subfamily Emargininae, now coinciding with the genus oreauella, uniformly possess an emarginate apical margin of the forewing. e only other cynipoids to have such a character are species of Kleidotoma, but being eucoilines, the latter have a distinct scutellar plate with posterior midpit. Some emarginines have what appears to be a raised scutellar plate on the scutellum, but in these species, there is clearly an entire lack of a glandular pit. Emarginines also typically have: a very abrupt, compact marginal cell in the forewing; distinct notauli; dense 'foamy' setae on the propodeum and base of the metasoma; and large, setiferous pits on the agellomeres of the male antenna. Much species level work remains to be conducted on the Afrotropical emarginines. However, we have noted the following characteristics of currently unnamed species groups: Species group A, which appears to be common in Madagascar, comprises species that possess notauli, and have a thin, complete, lamella along the posterior margin of the scutellum. e lamellae enclose a rather deep 'trough'; further, the lamella is so thin that without signi cant magni cation, it will not be visible. e presence of this resulting 'trough' may have led Quinlan (1988) to erroneously conclude Emargininae are actually Eucoilinae.
Species group B also possesses notauli, as well as having rather spectacular projections o the posterior margin of the scutellum. Two sub-groups can be recognized; one with a simply bifurcate scutellar margin; and a second group possessing a single projection o the posterior margin of the scutellum. Both subgroups contain the largest physical specimens of oreauella, with some reaching nearly 1.5 mm; further, nearly all members of this group have dark, dusky wings. Quinlan's oreauella pexa, considered by Kovalev to constitute the tribe Quinlanianini, belongs in this species group. Species group B appears to be endemic to Madagascar.
Species group C possesses notauli, and is similar in overall appearance to species group A, however, species in group C lack the posterior lamella on the scutellum. Nevertheless, group C species do have a deep, marked depression on the posterior margin of the scutellum. Quinlan's oreauella capito, considered by Kovalev to constitute the tribe Weldiolini, belongs in this species group. Species group C is common in continental Africa (throughout the tropical belt), as well as Madagascar.
Species group D is the most widespread of the four genus groups, found equally common in Madagascar and continental Africa. e distinguishing feature of this group is the general lack of clear notauli. In some species, it is clear that the mesoscutum is perfectly smooth; in others, there are faint traces of notauli, but never as clearly indicated as in species groups A-C. e posterior margin of the scutellum in these species is rounded, lacking any remarkable morphology. oreauella laverna, T. micipsa, T. palloris, and T. vacuna would be included in this group.
Distribution. Pantropical, apparently especially diverse in Madagascar. Presence in the Oriental region has not been published hitherto but is con rmed here. Afrotropical records: Cameroon, Democratic Republic of Congo, Madagascar, South Africa, Zambia, Zimbabwe (Quinlan 1988), Central African Republic, Ghana, Nigeria, Rwanda, Tanzania, Uganda (here).
Biology. Afrotropical eucoilines are koinobiont endoparasitoids of Muscomorpha Diptera larvae . A large group of Eucoilinae are parasitoids of Agromyzidae, mostly leafminers in the canopy, but the majority attack various families of ies typically in decomposing habitats (carrion, dung, fruit, leaf litter, sea wrack etc) but also in living plants, mushrooms and algae (Drosophilidae, Phoridae, Sepsidae, Ephydridae, Muscidae, Calliphoridae etc). However, it must be noted that the biology of the majority of eucoiline species remains unknown (Forshage and Bu ngton pers. obs.).
Distribution. e subfamily is represented in all biogeographical regions (including subantarctic islands). e majority of described species are Palearctic, and indeed there is a striking diversity in some genera throughout the Holarctic, but species diversity in the tropical regions is far richer though mostly yet undescribed, and possibly the Neotropics has the largest diversity of all regions (Bu ngton and Forshage pers. obs.).  Entire wasp laterally compressed, narrow, with a protruding head (a); head more than twice as long as wide, eyes hairy and scarcely protruding from outline of head capsule (a

14
Posteroventral corner of metapleuron glabrous, obliquely angled, corner often raised and forming a more or less triangular surface facing posterolaterally (a). Antennal F2 modi ed in males, more or less asymmetric, more so than F1 (or rarely only as much as F1) (b

Diglyphosematini
Diglyphosematini is a characteristic and rather easily circumscribed tribe, which was overviewed and revised on the generic level by Bu ngton (2011) Hairy ring at base of syntergum absent; metasoma downturned towards ventral position. Most easily confused with Diglyphosema and Gronotoma; distinguished from these genera by the presence of a distinctly impressed mesopleural triangle, and downturned metasoma. In both Diglyphosema and Gronotoma, the mesopleural triangle is present, but the dorsal and ventral margins are gently rounded, and the metasoma is directed more posteriorly. Distinguished from Paradiglyphosema by the possession in the latter of a genal carina and postero-lateral projections on the scutellum.
Five species groups are recognized within the genus, all based on the morphology of the scutellum. is character system provides a wealth of taxonomic information, and from the hundreds of specimens examined for this paper, the characters appear to be stable through space and time. e most commonly encountered species group is the A. nitida group, characterized by having a distinctly concave dorsal surface of the scutellar plate.
is is best seen in lateral view, and can be characterized as reminiscent of a gentle 'wave' or 'ski jump'. From our examination of the type series of all of  species, several species in this group will be synonymized in a more thorough treatment of the genus. e A. dubia species group is characterized by having an extremely short, narrow scutellar plate, revealing much of the dorsal surface of the scutellum. Some specimens in this group approach an Ealata in appearance, but lack other characters of that genus. e A. bucca species group is second in diversity to the A. nitida species group, and super cially looks similar. However, in the A. bucca species group, the scutellar plate, in pro le, is perfectly at (wave or ski-jump shaped in the A. nitida species group), with a large, deep glandular release pit. e last species group to be recognized here is the A. fercula species group. On Afrostilba conversa , comb. n. (Eucoilidea) (Cameroon, Democratic Republic of Congo, Ghana, Kenya, Madagascar, Mauritius, Nigeria, South Africa, Tanzania, Uganda, Zimbabwe) Afrostilba lacerta , comb. n. (Eucoilidea) (Democratic Republic of Congo, Ivory Coast) Afrostilba lana , comb. n. (Eucoilidea) (Democratic Republic of Congo) Afrostilba marcellus   , comb. n. (Eucoilidea) (Democratic Republic of Congo, Madagascar, Nigeria) Afrostilba perangusta , comb. n. (Eucoilidea) (Democratic Republic of Congo, Zambia, Zimbabwe) Afrostilba trulla , comb. n. (Eucoilidea) (Democratic Republic of Congo, South Africa) Afrostilba tyrus , comb. n. (Eucoilidea) (Cameroon)
Diagnosis. Protuberances absent on malar space. Dorsal margin of pronotal plate with a distinct emargination. Notauli reduced, only present anteriorly and sometimes posteriorly. Scutellar plate small, with a mound like protuberance anterior of glandular pit. Dorsal surface of scutellum broadly rounded both laterally and posteriorly, distinct posterior aspect of scutellum absent. Separated from all other Diglyphosematini by the presence of a single, broad protuberance anterior of the glandular pit of the scutellum.
Biology. Controversial, reviewed in Bu ngton (2011). Putatively reared from Tephritidae (unlike the rest of the Diglyphosematini, which attack leaf-mining Agromyzidae), but there are no isolated rearings to con rm this.
Diagnosis. Genal carina present. Lateral pronotal carina present (at least ventrally). Parascutal impression complete, with a distinct interruption anteriorly. Notauli present and well developed. Laterodorsal projections of scutellum present. Dorsal and ventral margins of mesopleural triangle distinct. is genus can be distinguished from all other Diglyphosematini by the presence of laterodorsal projections on the scutellum and a complete genal carina. ese characters are only shared with some Zaeucoilini (found in New World tropics), but the latter group is not found in the Afrotropical or Oriental regions, nor do Zaeucoilini possess notauli (Bu ngton 2009).

Species richness.
Two undescribed species: one in Cameroon, Central African Republic and Uganda, one in Kenya and Somalia.

Eucoilini
Current evidence suggests that Eucoilini is possibly not monophyletic. e tribe may be paraphyletic visavis the Trichoplastini and the two tribes may eventually have to be merged. It includes a majority of the larger-sized representatives of the subfamily.

Afrodontaspis Weld, 1961
Remarks. Rare. Afrodontaspis was rst characterised as super cially similar to the Neotropical Trissodontaspis Ashmead, 1903(Weld 1962) and then  noted it had a nities with Bothrochacis. In fact Afrodontaspis shares most of the systemati-cally important character states with Bothrochacis, and may possibly turn out to be an ingroup there (and accordingly a junior synonym); however, preliminary phylogenetic analyses (Baião and Forshage, unpublished) have not con rmed this, and we maintain it as a separate genus.
Diagnosis. Large, black or reddish wasps with very little pubescence on body and wings; unmistakable in their heavy striation and scutellar spine, yet sharing most of the systematically informative characters with Bothrochacis. Like Bothrochacis with striate vertex, and like some Bothrochacis with vermiculate coxae and notched scutellum, but Bothrochacis always has many setae and no striation on the mesosoma. Linoeucoila are similar but smaller, without a scutellar spine, and with many setae on the mesosoma. e scutellar spine is thinner and more pointed than in Trichoplasta. General appearance may also approach members of Aspicerinae and Figitinae with striate sides and reduced wing pubescence, but these lack the characteristic eucoiline scutellum. Distribution. Endemic to the Afrotopical region: Burundi, Democratic Republic of Congo (Benoit 1956a), Kenya, Uganda ), Burkina Faso, Cameroon, Central African Republic, Gambia, Rwanda, South Africa, Tanzania, Zambia (here).
Biology. Unknown, but like its close relatives, and like other species with glabrous wing membranes and striated body, it can be expected to attack dung-dwelling ies.
Species richness. Afrodontaspis lanata  (Democratic Republic of Congo, Kenya) Afrodontaspis striatissima (Benoit, 1956a)  Remarks. Not uncommon. Bothrochacis has been a small and easily recognisable genus since its original description. Quinlan's Afrotropical Eucoila species have been known for some time to not t into Eucoila, but nothing has been published about them since their original description. Here we note that they share all the important characteristics of Bothrochacis except the very peculiar notched scutellum, and include them here.
Most of the characteristics of the genus in the wider sense are also found in the genera Afrodontaspis and Linoeucoila, but preliminary phylogenetic analyses (Baião and Forshage, unpubl.) have not supported their merging into one genus. Diagnosis. Most of the large and dark (black or distinctly bicolored) eucoilines in the Afrotropical region belong to Bothrochacis. ey can be distinguished from the few others (Afrodontaspis, Linoeucoila, Aganaspis, Trybliographa and perhaps some Trichoplasta) by having plenty of large truncate setae on pronotum and mesoscutum, and stout setae on the subcosta of the wing. Like other genera of Eucoilini, they have subalar pits and glabrous oblique metapleural corners. ey have a striate vertex, often have vermiculate-reticulate sculpture on the coxae and large scutellar foveae, usually strong reduction of wing pubescence, and sometimes a notched scutellum. e core group are the "erythropoda species group", coinciding with Bothrochacis sensu Quinlan, the species possessing a notched scutellar plate. ese are large, typically bicolored, have hairless wings, smooth coxae, and large scutellar foveae.
Appearing like a less modi ed stem group there is the "veleda species group", consisting of species described by Quinlan in Eucoila, typically of smaller size, with an evenly arched scutellar plate and smaller scutellar foveae, often with some pubescence on wings. Specimens belonging to this group are sometimes quite similar to Trybliographa, rare in the region, common elsewhere, but easily separated by the characters listed above.
Biology. One species is recorded as attacking Muscidae in dung (Bridwell 1919), and due to phylogenetic inference as well as inference based on parallell occurrence of morphological particularities (elsewhere in the Eucoilinae as well as in the Figitinae) it seems reasonable to assume that most of them have similar habits.

Remarks. Very common throughout Africa.
Diagnosis. Usually medium-sized and rather stout eucoilines, with a more or less reduced hairy ring. Often super cially similar to Ganaspis. Easily separated from the former in males, since Leptopilina males have antennal F1 hardly modi ed and F2 moderately modi ed, while Ganaspis have F1 distinctly modi ed and F2 not at all; but for females (or males with antennae not visible) it is important to examine the metapleural corner: in Leptopilina it is oblique and hairless, in Ganaspis rectangular and hairy. Many Leptopilina which have a narrow marginal cell may be confused with Rhoptromeris, which have the male F2 more strongly modi ed and only rarely have a reduced hairy ring, and which are easy to separate by having lateral bridges on the pronotum (open in Leptopilina, closed in Rhoptromeris).
Species groups have been suggested by Nordlander (1980) and Allemand et al. (2002), but these have little consequence for the identi cation of the genus as such.
Distribution. Worldwide. Afrotropical records: Cameroon, Democratic Republic of Congo, Madagascar, Nigeria, Seychelles, South Africa, Uganda, Zambia, Zimbabwe   Lee et al. 2009, Novkovic et al. 2011, including in Africa (Allemand et al. 2002, Nordlander 1980  Leptopilina boulardi (Barbotin, Carton & Kelner-Pillault, 1979 Remarks. Rare in the Afrotropical region. Linoeucoila was rst described with a number of species from Taiwan (Lin 1988), and considered close to Trybliographa, mainly di ering by their aciculate, vermiculate or striate body sculpture. e name has only been used later when the genus was included in the tribe Eucoilini (Forshage et al. 2007). Type specimens have not been available for loan, but specimens from Taiwan corresponding to the original description of the genus (mostly undescribed species) have been studied, and it has been considered that this is a taxon present throughout the Oriental region and also present in Africa. However, it has been di cult to decide how to circumscribe the genus or to separate it from Trybliographa except by the body sculpture, and it may well turn out to be one or several lineages within Trybliographa. Despite conforming to the hitherto known diagnostic characters of Linoeucoila, African specimens have some di erences from the Oriental species, and may eventually have to form a genus of their own.

Diagnosis.
Large and dark (bicolored) eucoilines, most similar to Bothrochacis, but lacking the characteristic subcostal setae and pentiful truncate mesosomal setae of that genus, and having striation on sides of mesosoma. Striation of vertex is reticulate or lateral, not radiate as in Bothrochacis. Very similar to Trybliographa except for the sculpture of the integument. With subalar pits, an oblique and glabrous metapleural corner, more or less reduced wing pubescence, large scutellar foveae, and vermiculate sculpture on coxae.
Distribution. Mainly Oriental, but also occuring in the Afrotropical region (hitherto unpublished). Afrotropical records: South Africa, Uganda (here).
Biology. Unknown, but many close relatives, and many gitids with striate sides and reduced wing pubescence are parasitoids of dung-breeding ies.

Species richness.
Undescribed species only.

Trybliographa Förster, 1869
Remarks. Rare in the Afrotropical region. Only one species encountered so far, but more are to be expected. Diagnosis. Characteristically large eucoilines, dark and mostly strongly built, separated from several other genera with which confusion is otherwise possible by the possession of distinct subalar pit, as well as a metapleural corner which is hairless, oblique, and more or less upturned (forming a small, more or less triangular glabrous posterolateral surface called the metapleural triangle). Aganaspis which is often supercially similar always has a distinct tuft of hairs on the metapleural corner. Furthermore, in males, Aganaspis just like other Ganaspini have the antennal F1 modi ed, while Trybliographa have F2 modi ed. In Africa, the major confusion risk is in fact those specimens of the closely related Bothrochacis that have less reduced wing pubescence. Unlike Trybliographa, they typically have stout setae on the subcosta, truncate setae on the pronotum and mesoscutum, large scutellar foveae, and vermiculate sculpture on the coxae. e single Trybliographa species encountered so far in the Afrotropical region though, is very easy to recognise by the fuscate marginal cell.
Distribution. Worldwide, but by far most diverse in the Holarctic. Afrotropical region: Madagascar (here).

Species richness.
Trybliographa australiensis Ashmead, 1900 (Madagascar) Ganaspini is is the largest and most di cult tribe of Eucoilinae. e generic limits between the major genera are not clear and a small selection of autapomorphic forms currently have generic status which could just as well be moved into the major genera. We have considered a major generic-level revision of this group beyond the scope of this work, and probably impossible without a thorough phylogenetic analysis.
Diagnosis. Large, strongly built, black or darkly brown eucoilines. e genus was originally erected (Lin 1987) for Oriental species with a very large scutellar plate reaching the posterior end of the scutellum, and a high pronotal plate with an emarginate (bilobed) dorsal rim protruding well over the pronotal-mesoscutal suture. But the exploration of apparently closely related forms, especially in South America, has made the genus far more di cult to circumscribe in terms of unambiguous diagnostic characters, and indeed, blurred the boundaries towards the heterogenous Ganaspis. Most Aganaspis, however, are large and resemble Trybliographa in general habitus, but are easy to distinguish from the latter based on the distinct hairtuft on the metapleural corner, the small and shallow subalar pit, and the modi ed antennal F1 in males. e posterior metapleural margin is uninterrupted (but sometimes depressed in the middle) and usually somewhat oblique in the ventral part; the metacoxae usually have semi-long hairlines (but sometimes only small tufts); the sctutellar plate is usually very large and the scutellar foveae usually large (but sometimes far more normal sized).
Distribution. Widespread, but primarily East Asian and Neotropical. Afrotropical records: Central African Republic, Democratic Republic of Congo, Kenya, Reunion, South Africa, Tanzania (here).

Didyctium Riley, 1879
Remarks. Among the most common genera of eucoilines in the Afrotropical region (yet not recorded by Quinlan). Not always certainly distinguished from Ganaspis, Hexacola and Endecameris.
Diagnosis. Didyctium are usually recognisable through the combination of a concave scutellar plate and the characteristically half-open marginal cell. Furthermore, unlike in typical representatives of the closely related Ganaspis, the head is transverse (not deep), the mesosoma is short (propodeum not extended), the coxae bear small hair tufts (not elongate hairlines), the scutellar plate is relatively small, the lateral bars of the scutellum are striate, and the posterior margin of the metapleuron is straight (not with a circular or elongate incision). However, the diverse Ganaspis may vary in all these traits. Didyctium females usually have very unusual antennae, where the agellomeres are strongly di erentiated into very short annelli and very long club articles. Very small specimens of Didyctium approach the character states of Endecameris, and the boundary between the two genera is uncertain.
Biology. Attacks Phoridae and other ies in concealed habitats (Beardsley 1989, Riley 1879; no host records are from Africa.

Species richness.
Didyctium naivashae (Kie er, 1913b), comb. n. (Cothonaspis) Type in MNHN studied by MF (Kenya, Rwanda) Numerous African species remain to be described.  presence of a mesopleural line separates Endecameris from Micreriodes and Rhoptromeris. In Endecameris, the scutellar plate is narrow, there is often a reduction in wing pubescence, and there is often a remarkable reduction of the number of antennomeres (the latter not yet observed in African specimens). e antennae of female Endecameris are similar to those of Didyctium, with a strong di erentiation of the agellomeres into proximal very short articles (more or less annelli) and distal elongate club articles with distinct white rhinaria.
Distribution. Mainly in the Oriental and Oceanic regions, but also occuring in the eastern Palearctic and in the Afrotropical region. Afrotropical records: Cameroon, Central African Republic, Gambia, Guinea-Bissau, Madagascar, Mauritius, Sierra Leone, South Africa, Uganda, Yemen, Zimbabwe (here).

Species richness.
Several species that are probably undescribed are present in Africa.

Ganaspis Förster, 1869
Remarks. Common in Africa as elsewhere, yet only recorded in the key to genera in . e genus is very di cult to circumscribe, not always certainly distinguished from Aganaspis, Didyctium and Hexacola. At the same time, many smaller genera of the Ganaspini may actually be ingroups in the genus as currently conceived. us, a thorough analysis of the boundaries of this genus will most likely highlight the need to either synonymise a number of apomorphic small genera back into Ganaspis, or to recognise species groups of Ganaspis as separate genera. However, the di culty to circumscribe the latter and to delineate them from the other major genera of Ganaspini suggests that this cannot be done without a thorough phylogenetic analysis. Of the several genera that Lin (1988) described in his revision of Taiwanese eucoilines, we have recognised Gastraspis (easily recognisable at least in females), but we have not been able to establish whether the plentiful African Ganaspis with very wide scutellar plates could be assigned to his Epochresta, and thus we have not treated the latter as an African genus.
Diagnosis. Ganaspis is a vast and rather morphologically heterogenous genus, and can be regarded as currently comprising all of the typical Ganaspini that are not "different" enough to warrant a genus of their own. Typical Ganaspis are small, somewhat pale, compactly built eucoilines. Other characteristics include: a deep, more or less globular, head; a rather elongate mesosoma; more or less long hairlines on meso-and metacoxae; scutellum with a foveolate dorsal surface; a large, at or convex scutellar plate; smooth lateral bars; a posterior metapleural margin with a circular or elongate excision; a narrow but distinct petiolar rim; broad wings with a rather truncate or faintly excised apex; and a deep fore wing marginal cell with curved sides -but most or all of these characters may vary within the genus. e typical Ganaspis are very often super cially similar to Leptopilina (also common parasitoids of Drosophilidae), but usually rather easily separated from them by having a modifed F1 in male antennae, and a distinct hair tuft on the metapleural corner. In order to facilitate recognition and sorting of Ganaspis in the Afrotropical region, we recognise the following morphological types of Ganaspis as characteristic within the Afrotropical fauna.
e apparently most common appearance of Ganaspis is one rather close to the European type, with long coxal hairlines, a circular or elongate excision of posterior pronotal margin, a wide scutellar plate not reaching the posterior end of scutellum, rather pale colour (middle brown body and yellow or pale brown legs). In some species, including most of the African taxa there are distinct patches of dense white pubescence on the pronotum and axillulae; marginal cells of wings are relatively homogenous with stong dominance for a short deep closed type with curved sides; some forms have a striking tooth on the metapleural edge, distinctly bi-or tricolored antennae, or large scutellar foveae.
A similar morphological type can also be seen in forms with an even larger scutellar plate, that is distinctly convex, but with a posterior depression and often an elongate-oval glandular release pit. ese species are often relatively large in size. ey may coincide with Lin's (1988) genus Epochresta (but types in TARI have not been available for loan).
Occasional specimens, usually of small size, lack the coxal hairlines and have only short hair tufts.
Many undescribed species, especially from Madagascar, are rather large and remarkably slender-elongate in build and pale in color, somewhat approaching the habitus of Chrestosema and related genera. Occasionally these may have very unusual features such as an elongated petiolar rim. Some of these are similar in appearance to the characteristic Neotropical Ganaspis "neotropica-group".
Other morphological types conform to the most common type in most respects, but di er, for example, in scutellar morphology. Some tiny specimens resemble Endecameris or Didyctium, others Hexacola, many Aganaspis.
Biology. Usually attacking Drosophilidae in various habitats, but sometimes also fruit-infesting Tephritidae or other ies (Nordlander and Grijpma 1991, Melk and Govind 1999, Vass and Nappi 2000. None of the host records are from Africa.
Diagnosis. Small Ganaspis-type eucoilines most easily characterised by the elongate female metasoma. ese specimens also show the following characteristics: short coxal hair tuft; a rather large, at scutellar plate and relatively large scutellar foveae; metapleural margin with an elongate excision; wing with a very slightly emarginate apex and an elongate marginal cell. Associated males are very di cult to discern within the range of variation present in Ganaspis.
Distribution. Described from Taiwan, but here reported also from Africa. Afrotropical records: Central African Republic, Kenya (here).

Species richness.
Undescribed species in the region. Remarks. Common throughout the Afrotropical region. Not always easily distinguished from Ganaspis and Didyctium. Daruna was erected by Benoit (1956a) for a specimen from Ruanda (the generic name being an anagram thereof ), which, according to the original description, had dual glandular release pits of the scutellar plate, both the common posterior pit, and the central pit characteristic of most Diglyphosematini. Inspection of the holotype (in RMCA) by MB and MF revealed only the normal posterior glandular pit though, and nothing separates Daruna from a normal Hexacola. Linguistic gender of Hexacola is, according to etymology in the original description, neuter, but many species names have been given in the feminine previously. Diagnosis. Hexacola are usually very characteristic habitus-wise, being rather elongate eucolines with a characteristically globular head; a characteristic wing which is rather narrow with a narrow, closed, mostly triangular marginal cell; a very characteristic scutellum which is convex in its entirety; a very wide, convex, scutellar plate covering most of the surface; mostly striate sculpture on the narrow surrounding dorsal surfaces of the scutellum; and narrow, oblique scutellar foveae. Certain forms, however appear intermediate between the typical Hexacola and Dicyctium or Ganaspis. Occasionally the general bodyshape as well as scutellar striae of Hexacola may cause confusion with Kleidotoma, though the wings and metapleura are very di erent, or with Rhoptromeris, which can always be recognised by the pronotal plate with closed lateral bridges (and modi ed F2 in males).
Diagnosis. Small elongate eucoilines without a hairy ring, with narrow wings with narrow triangular marginal cells. May be confused with certain Leptopilina with strongly reduced hairy ring, but Cothonaspis are far more elongate in shape, have globular heads, and a pointed metapleural corner, whereas Leptopilina are stout, have more transverse heads and an oblique metapleural corner. Easily separated from their closest relatives in the region, Kleidotoma, by the reduced hairy ring, wing apex truncate (not incised), and male F2 modi ed (not F1).
Distribution. Mainly Palearctic, but also present in Nearctic and Afrotropical regions, and with a widespread species present throughout the Old World tropics. Afrotropical records: South Africa , Central African Republic, Democratic Republic of Congo, Madagascar, Uganda, Yemen (here).

Kleidotoma Westwood, 1833
Remarks. Common throughout the Afrotropical region. Diagnosis. Usually elongate eucoilines of varying size, in most cases unmistakable through their incised wing apex alone (shared only with oreauella of Emargininae), but in forms with indistinctly incised wings (or where the wing shape is not readily visible) there are several other diagnostic characters: patchily reduced wing pubescence and the reduced wing venation with a small triangular marginal cell; fore wing veins of uneven width; distinctly pointed metapleural corner; narrow scutellar plate; and longitudinally striate dorsal surface of the scutellum. May be confused with Hexacola, who share the striate scutellum, the globular head, the often strongly modi ed male F1 and the sometimes narrow triangular marginal cell, but Hexacola typically have a very large and convex scutellar plate and narrow oblique scutellar foveae -and always a rectangular metapleural corner and a non-incised wing apex. Kleidotoma is a taxon that is di cult to overview, and globally there are very few recognisable species-groups that are not obviously arti cial. ere is a general spectrum from tiny, often brown, species with little reduction of wing pubescence, and large, often black, species with very strong reduction of wing pubescence. And there are the aberrant brachypterous taxa (often ripicolous species occuring in algae or wrack, sometimes on isolated islands; but at least in the Holarctic also in ground-dwelling species in grasslands). e wingless or brachypterous forms among Kleidotoma are the only cynipoids with this state found in the Afrotropical region so far (though in other regions, Rhoptromeris and Alloxysta occasionally show brachyptery too, and such specimens may be found in the Afrotropical region). However, beyond what is already said here, we are not at the level of knowledge to start discussing species groups in Kleidotoma in a meaningful way.
Biology. Parasitoids of various ies in concealed and mostly decomposing habitats; debris, dung, carrion, fruit, fungi, grass, seawrack, aquatic plants etc. Broad host range, common hosts in other parts of the world include Drosophilidae, Sepsidae, Ephydridae and others (Baker 1979, Beardsley 1990, Belizin 1963, Burghele 1959 Trichoplastini is a tribe that is easily recognisable by the striking morphological synapomorphy of having a pronotal plate with lateral bridges closing the lateral cavities. e tribe may ultimately be synonymised with Eucoilini; the two tribes currently constitute a monophyletic clade. roughout Africa, Rhoptromeris is by far the most commonly encountered Trichoplastini, and in fact, may be the most commonly encountered gitid group.

Angustacorpa Quinlan, 1988
Remarks. Rare. is genus may be an apomorphic ingroup of Trichoplasta, but for the time being we recognize this taxon as a valid genus.
Diagnosis. Very characteristic through their spectacular degree of lateral compression. Otherwise in all respects similar to Trichoplasta (pronotal plate with lateral bridges, narrow scutellar plate, extended posterior lobe of scutellum), and some Trichoplasta do indeed approach the Angustacorpa habitus (indicating the possibility of Angustacorpa being just an apomorphic ingroup), but the true Angustacorpa are always recognisable by their head: compound eyes are placed at mid-height of head, are pubescent, and their outline does not project from the general outline of the head capsule.
Distribution. Endemic to the Afrotropical region: Democratic republic of Congo, Kenya (Quinlan 1988), Cameroon, Central African Republic, Guinea, Republic of Congo, South Africa, Uganda, Yemen (here).
Biology. Host unknown, but expected to attack a dipterous host in a narrow space such as under tree bark.

Rhoptromeris Förster, 1869
Remarks. Ubiquitous; the most common genus of eucoilines in Africa. is taxon keys out three times in the generic key above, which re ects the morphological plasticity and complexity of the genus. ough many Afrotropical species have been described, there are undoubtedly many more undescribed species awaiting description. Further,  reliance on antennal characters to separate species within Rhoptromeris will have to be revisited too, since antennal characters themselves without other characters are frequently unreliable in gitid taxonomy.
Diagnosis. Mostly small and elongate eucoilines, with narrow wings and a narrow triangular marginal cell. Males usually have a strongly modi ed antennal F2. Rhoptromeris are easily recognised by the combination of a pronotal plate with lateral bridges, a scutellum lacking a protruding posterior lobe, and the dorsal surface of the scutellum distinctly striate, especially anteriorly. ese characters separate the genus from the morphologically similar Trichoplasta. However, intermediate forms between these two genera have been examined, creating di culties in separating the two genera; in these cases, focusing on the striate dorsal surface of the scutellum will reliably run these species in the key. If the pronotal plate is not obvious, the diversity of morphotypes of Rhoptromeris in this region may create confusion with several other genera. e most common Afrotropical specimens are small, brown and elongate; have narrow elongate wings with a narrow triangular marginal cell; and have a circular incision near metapleural corner. Some, often larger, have a more elongate incised (or depressed) area along metapleural posterior margin. ese may possibly be mistaken for several other eucoiline genera of a similar general build; Hexacola, Ganaspis, Kleidotoma or Cothonaspis. Rhoptromeris is separated from all of these by the pronotal plate, from all but Cothonaspis by the male antennae, and from most of these genera by differences in the metapleuron and scutellar plate.
Larger taxa are in general stouter, less elongate, have darker colour, broader wings, a deeper marginal cell and a shorter hair fringe on the wings; and approach Leptopilina or some Trybliographa or Trichoplasta in appearance.
A distinct species group is represented by tiny specimens with an indistinct or absent mesopleural line, a very long hair fringe on the wing, and often a shorter marginal cell; these are super cially very similar to Micreriodes and Endecameris, but easily recognised by their pronotal plate.
Another distinct species group have a pointed posterior lobe of the scutellum and are very easily confused with Trichoplasta; separated only by the general build of the scutellum, which has a convex scutellar plate and at least partly smooth or lineate sculpture of the dorsal surface. Of these we have seen a small, stout black taxon from South Africa, a large dark brown taxon from Uganda, and some less characteristic small brown forms from elsewhere.
One or several species groups have protuberances on the face, small often pointed tubercles on the clypeus and/or malar spaces and sometimes modi ed mandibles; they are small or tiny, and more stoutly built than typical Rhoptromeris; their appearance approaches Nanocthulhu or Stentorceps, both of which have more spectacular facial protuberances. Most of these specimens are from South Africa, but we have also seen singletons from East African countries.
Biology. As far as known they attack Chloropidae on grasses and fungi (numerous sources including Bhattacharyya 1957, Quinlan 1978, Nordlander 1978, Nordlander and Grijpma 1991, but no certain records are from within the Afrotropical region.

Species richness.
Rhoptromeris abba  (Democratic Republic of Congo, Kenya) Rhoptromeris afer Quinlan, 1986 (Democratic Republic of Congo, Uganda) lasta, thus with a posterior protrusion on the scutellum, which is another character that separates them from Nanocthulhu. Some species of Rhoptromeris and Hexacola have protrusions from the malar space and clypeus, but they also lack the corniculum of Stentorceps. Distribution. Endemic to the Afrotropical region and mainly occurring in East Africa. Afrotropical records: Kenya , Botswana, Madagascar, Nigeria, Rwanda, Somalia, South Africa, Uganda (Nielsen and Bu ngton 2011), Republic of Congo, Tanzania, Yemen (here).
Diagnosis. Often easily recognisable by the combination of a pronotal plate with lateral bridges, a posteriorly protruding scutellum, and the dorsal surface of the scutellum distinctly foveate. Species with a moderately protruding scutellum can be di cult to distinguish from Rhoptromeris, but focusing on the foveate sculpture of the dorsal surface of the scutellum will run an unknown species through the key e ectively.
Distribution. Holarctic region and Old World Tropics. Appears to have its largest diversity in the Afrotropical region. Afrotropical records: Cameroon, Democratic Republic of Congo, Kenya, Nigeria, Rwanda, South Africa, Tanzania, Uganda, Zimbabwe , Burkina Faso, Cape Verde, Cameroon, Central African Republic, Comoros, Gabon, Gambia, Ghana, Guinea-Bissau, Ivory Coast, Malawi, Republic of Congo, Sierra Leone, Sudan, Yemen (here).
Biology. Specimen label data suggests tropical species appear to attack fruit-infesting Lonchaeidae and Muscidae (label data), but in the Holarctic they are mostly woodassociated, attacking Drosophilidae and Lonchaeidae under tree bark (label data), and these habits are probably present in some Afrotropical taxa too.

Garudella Bu ngton & Forshage, 2014
Remarks. Very rare, in Africa only known from a single specimen in the Republic of Congo. e tribal placement of this taxon is uncertain. Key characters make it run to Trichoplastini, but both habitus and a number of other characters are more similar to Cothonaspis, and in the original description it is tentatively placed in Kleidotomini. It may, along with Cothonaspis and Triplasta Kie er, represent a basal lineage in that tribe, but in the absence of a proper phylogenetic analysis, this remains little more than just a guess.
Diagnosis. e thick basal part of the pronotal plate and the very robust petiole are unique characters in the Eucoilinae. e overall body shape is similar to Cothonaspis, and the wings are similar to Cothonaspis and a number of other genera (Rhoptromeris, Trichoplasta, and many Leptopilina), but Garudella can be separated based on the morphology of the pronotal plate and the uniquely robust propodeal-petiolar complex.
Distribution. Oriental and Afrotropical regions. Afrotropical records: Republic of Congo.

Garudella afrotropica Bu ngton & Forshage, 2014 (Republic of Congo)
Leptolamina group is is a distinct group of eucoiline genera that is currently not assignable to a tribe. Included genera were previously associated with the informal "Chrestosema group" (Nordlander 1982), most representatives of which are now included in the Ganaspini. Buington et al. (2007) recovered Leptolamina and Sirenes as a sister-group clade to what is now Eucoilini and Trichoplastini. With the synonymisation of these two genera, this group currently only contains Leptolamina and Micreriodes; additional sampling in the Old World tropics may yield more genera that would be assignable to this group.

Leptolamina Yoshimoto, 1962
(synonym Sirenes Quinlan, 1988, syn. n.) Remarks. Leptolamina was originally described from the Paci c region (Yoshimoto 1962). Quinlan (1988), in his treatment of the Afrotropical fauna, cites Nordlander (1982) recognising Leptolamina as a member of the Chrestosema group of genera, but does not consider the taxon with regards to the African fauna. e Glauraspidia described by Quinlan in the same work (1988) are in fact Leptolamina (new combinations below). Also in the same work (1988), he described the genus Sirenes to accommodate species of eucoilines that lacked a mesopleural line, but whose surface sculpture was matte, and otherwise conformed to his concept of Glauraspidia. After examining many specimens collected throughout the Afrotropical region, we have determined that Sirenes is at best one end of a morphological spectrum, which also includes Leptolamina. On the Sirenes end, forms are typically larger, slightly more matte, heads slightly more elongate, and lateral depressions of the mesoscutum slightly more developed. Altogether, we have found many intermediate forms, without any distinct morphological features to suggest monophyly of each genus; we hypothesize that the features mentioned above are the result of allometry related to overall body size. Hence, we hereby make Sirenes a junior synonym of Leptolamina (see below).
is decision nds further motivation in the phylogenetic analysis of Bu ngton et al.  Diagnosis. Entire wasp lightly to heavily matte over entire body, but particularly on head and mesoscutum. Mesopleural line entirely absent. Pronotal plate distinctly directed anteriorly, anterior half (just behind head) wider than posterior half. Pronotal fovea indistinct in most cases; when visible, closed laterally (lateral bridge present). Hind coxae often entirely without hairpatch. Setae on wings ranging from normal to a particular form with dark, broad sockets; setae along anterior aspect of marginal cell very stout in larger forms. Face elongate to round, mandibles ranging from blocky, subquadrate to smaller and triangular; malar space with very slight striations running from the ventral margin of the compound eye to the mandibular base. Scutellar plate typically narrow, elongate; occasionally wider, tear-drop shaped. Longitudinal lateral depressions of mesoscutum present in larger specimens, absent in smaller ones. Shares several of its characteristics (matte nish, longidutinal lateral depressions of mesoscutum, very elongate coxae etc) with Chrestosema, but the latter taxon always has a mesopleural line and an elongate hairline along metacoxae. Very small specimens of Leptolamina will approach the appearance of Micreriodes, but the latter is far less "foamy".
Distribution. Mainly Old World tropics, but also in eastern Palaearctic. Afrotropical records: Democratic Republic of Congo, Ivory Coast, Seychelles, Zambia, Zimbabwe , Cameroon, Central African Republic, Comoros, Gabon, Kenya, Madagascar, Mauritius, Republic of Congo, South Africa, Uganda (here).
Biology. Reared from Drosophilidae in Japan (label data).

Micreriodes Yoshimoto, 1962
Remarks. Rare in Afrotropical region. Previously known only from the Paci c, but actually distributed worldwide. Diagnosis. Tiny wasps with a well-developed set of "dwar cation" characters: a globular head, short antennae, sometimes a reduction in antennomere number, a narrow scutellum, very narrow wings with very long hair fringe, wing venation of uneven width, and fore wing marginal cell short and wide open. Similar to Endecameris and to some Rhoptromeris in some or all of these characters, but separated from the latter by the combination of having a pronotum with open lateral foveae, and lacking a mesopleural line.
Distribution. Previously only known from the Paci c region, but are in fact widespread in the Old World tropics and recently a specimen was found in North America. Afrotropical records: Cameroon, Central African Republic, Madagascar, Nigeria, Reunion, Sierra Leone, Uganda (here).
Biology. Hosts unknown, assumed to be Drosophilidae.

Figitinae
e Figitinae are represented in the Afrotropical region by four genera containing 7 described species. A number of undescribed species are present in world collections. Figitinae contains a heterogenous assemblage of genera and is probably not monophyletic (Ronquist 1999, Bu ngton et al. 2007). However, the genera native to the Afrotropical region all belong to a very distinct lineage (big black wasps, with strongly reduced wing pubescence, hairy compound eyes, no hair patches at metasomal base, often with lateral striation of the mesosoma, often with scutellar spines, attacking calyptrate Diptera in dung and carrion), which is safely monophyletic, and in fact, present a very interesting morphological and life history convergence with some genera of Eucoilinae (in the Afrotropical region, namely Bothrochacis). ough in addition to the native genera, there is the genus Lonchidia, of which we have so far encountered only one Afrotropical specimen of a species present in Europe, which may be an accidental introduction or possibly an established population of synanthropic origin. is genus is very di erent from the major faunal component of Figitinae, and represents a separate lineage that renders the subfamily paraphyletic in phylogenetic analyses. It is easily recognisable by its con uent scutellar foveae.
Biology. Host records are lacking for Afrotropical species of Figitinae, however, Bu ngton et al. (2012) cite all con rmed host records for the group, and these all relate to calyptrate ies in dung and carrion.
Distribution. e subfamily is represented in all biogeographical regions (except the Antarctic) with the majority of described species occurring in the Holarctic and Neotropical regions. e paucity of Afrotropical and Oceanic species, as well as the almost total absence of Oriental records, may very well only re ect a poor state of taxonomic knowledge. Remarks. is is a rare genus in the region. e Afrotropical representatives of the genus were recently revised and four of Benoit's species described in 1956 were sunk into synonymy with F. aciculatus (van Noort et al. 2014). On a global scale, it is a poorly circumscribed genus versus several smaller genera, and many of its nominal species are of doubtful identity.

Diagnosis.
Large gitines with reduced pubescence on wings (often completely hairless) and more or less striate mesosomal sides. Easily separated from Xyalophora and Neralsia by the rounded scutellum (no indication of a spine in outline in dorsal view). Sti , stout hairs present across most of body, distally bifurcate.
Distribution. Probably worldwide, but to date no records from the Oriental and Oceanic regions have been published. Afrotropical records: Democratic Republic of Congo (Benoit 1956d), Cameroon, Ethiopia, Kenya, South Africa, Uganda, Yemen (van Noort et al. 2014.

Species richness.
Figites aciculatus (Benoit, 1956d)  Remarks. e only Afrotropical specimen seen so far is from South Africa and may be an accidental introduction. It corresponds to a form present in Europe, which is currently considered as belonging to Lonchidia clavicornis omson, but which di ers from the type specimen in some minor respects. Further studies may possibly show that this is a separate, currently unnamed, species. Diagnosis. Small, rather slender, and more or less strongly pubescent gitines, easily recognised by the con uent scutellar foveae. Pubescence is dense in patches on the sides of the large metasomal tergite, as a collar on the pronotum, on the propodeum, and rather dense also on metapleura and metacoxae. e marginal cell of the forewing is characteristically short, and the antennae in females end with an enlarged apical agellomere.
Distribution. Mostly an Holarctic genus, here reported for the rst time from the Afrotropical region. Afrotropical records: South Africa (here).
Biology. No host records exist. Hosts are expected to be saprophagous Brachycera larvae.

Neralsia Cameron, 1883
Remarks. Rare in the Afrotropical region. e genus is extremely species-rich in the Neotropical region and has recently been revised in a series of papers by Jiménez et al. ( , 2005aJiménez et al. ( , 2005bJiménez et al. ( , 2005cJiménez et al. ( , 2007Jiménez et al. ( , 2008aJiménez et al. ( , 2008b  . Neralsia is also common throughout the Nearctic Region, but species limits have not been thoroughly established (Bu ngton pers. obs.). Diagnosis. Neralsia and Xyalophora are the only known gitines in the Afrotropical region with a scutellar spine. Neralsia can be distinguished from Xyalophora by whether or not the notauli are horizontally striate: smooth in Neralsia, striate in Xyalophora (Jimenez et al. 2008). Also, most Neralsia have a longer, more robust scutellar spine than Xyalophora, but in specimens we have examined, this character varies with overall size of the specimen. is genus also resembles some members of Aspicerinae, most notably Prosaspicera, which also possess a distinct scutellar spine, but can be separated from Prosaspicera by the lack of a facial impression on the head (present in Prosaspicera), and lack of a ligulate metasoma T2.
Distribution. Mainly Neotropical, but with single species in the Nearctic and the Afrotropical regions. Purported records from the Oriental region and the east Palearctic are uncon rmed. Afrotropical records: Central African Republic, South Africa (van Noort et al. 2014).
Biology. Parasitoids of calyptrate Brachycera larvae in decomposing substrates (Díaz et al. 2000, omas and Morgan 1972 ), and an often slightly smaller scutellar spine; this second character, however, is often linked to adult body size and should be used with caution. As in the case of Neralsia, species of Xyalophora can be super cially similar to Prosaspicera (Aspicerinae), but can be separated from that taxon by the lack of a facial impression on the head, as well as the lack of a ligulate metasomal T2. All three African species have the occipital carinae directed towards the ocellar area and separated in the middle by a smooth surface as well as a smooth interocellar area.
Distribution. Probably worldwide, but no records from the Oriental region are published. Afrotropical records: Burkina Faso (Jiménez et al. 2008c);Democratic Republic of Congo, Mali, Namibia, South Africa (van Noort et al. 2014).
Biology. Parasitoids of calyptrate Brachycera larvae in decomposing substrates (Ionescu 1969  Notauli complete from anterior margin of mesoscutum (at junction with pronotum) to posterior margin (at junction with scutellum) (a); parascutal impression present (a); female with more than 10 agellomeres (b Diagnosis. Within the Afrotropical region, Tylosema can be confused with Pycnostigmus. However, Pycnostigmus lack notauli on the mesoscutum, and Tylosema have complete notauli; this is very easy to observe and the character is reliable. Outside the Afrotropical region, Tylosema could be confused with the Palearctic Trjapitziniola, but this latter taxon lacks notauli, and has not been recorded from Africa (but see above).  Distribution. e family is represented in nearly all biogeographical regions with the majority of species occurring in the northern hemisphere (Nordlander et al. 1996). Introduced for biological control into Australia and South America (Nordlander et al. 1996).
Diagnosis. Readily distinguished from all other Afrotropical Cynipoidea by the sheer size of this wasp. Adults easily reach 1.5 cm; the closest, by size, to this group of wasps are some species of Oberthuerella, especially O. lenticularis and O. cyclopia. An additional character that readily seperates Ibalia from other cynipoids (and most other Hymenoptera, for that matter), is the extremely laterally attened mesosoma. In dorsal view, the metasoma of Ibalia is blade-like, housing a long, coiled ovipositor. e large liopterids, mentioned above, all have a distinctly ovate metasoma, never laterally attened. As in the case of the liopterids, ibaliids have a distinctly horizontally strigate mesoscutum, which is hypothesized to be critical in emerging from their wood-boring hosts (Ronquist 1999).
Distribution. Holarctic and transgressing into the northeast Oriental region, introduced elsewhere. Afrotropical records: South Africa (Hurley et al. 2007).
Biology. Parasitoids of woodwasps: Sirex, Urocerus and Xeris (Siricidae) living in conifers. Males mate with females while they are laying eggs. e female inserts her ovipositor down the tunnel bored in pine trees by the host woodwasp larva, to lay an egg either into the egg of the host or into the young host larva. On hatching the ibaliid wasp larva emerges from the body of the host and feeds externally (Hurley et al. 2007).

Species richness.
Ibalia leucospoides (Hochenwarth, 1785) (Ichneumon) ssp. leucospoides Hochenwarth, 1785 (South Africa; extralimital distribution throughout the Holarctic region and introduced elsewhere) Liopteridae e Liopteridae are represented in the Afrotropical region by two of the three world subfamilies: Oberthuerellinae and Mayrellinae, with the former having been recently revised by Bu ngton and van Noort (2012) and the latter by Liu et al. (2007) and van Noort and Bu ngton (2013). A key to Afrotropical liopterid genera was published in Bu ngton and van Noort (2012).
Biology. e biology of the Liopteridae is unknown, though a few published observations suggest hosts could be Coleoptera in rotting wood: two species of Kie eriella Ashmead emerged from logs infested with buprestids (Acmaeodera pulchella (Herbst)) (Weld 1956); a Kieferiella species and a Paramblynotus Cameron species were reared from trees in the family Fabaceae, Prosopis glandulosa Torr. and Dalberghia fusca Pierre, respectively (Ronquist 1995a Distribution. e family is represented in all biogeographical regions except for the Western Palaearctic with the majority of species occurring in tropical or subtropical regions , Ronquist 1995a). e subfamily Liopterinae is restricted to the New World and is centered in the Neotropical region with a few species extending north into the Nearctic region (Ronquist 1995a

Mayrellinae
Globally this subfamily is represented by two genera, Kie eriella and Paramblynotus, with the latter genus occurring in the Afrotropical region (Ronquist 1995a, van Noort and Bu ngton 2012).
Paramblynotus Cameron, 1908(synonyms: Paraegilips Kie er, 1910a, Allocynips Kie er, 1914, Holocynips Kie er, 1916, Diholocynips Rohwer & Fagan, 1917, Mayrella Hedicke, 1922, Paribalia Weld, 1922, Stylobrachys Belizin, 1951, Baviana Barbotin, 1954, Decellea Benoit, 1956b Remarks. e genus was recently revised by Liu et al. (2007) and van Noort and Buington (2013). e latter paper described a further 9 species, including a new species group from Madagascar. Paramblynotus species are rare in collections. Diagnosis. Medium sized to very small cynipoids. Very small species look supercally like cynipids, but careful attention to the relative size of the metasomal terga will help seperate Paramblynotus from cynipids. Some super cially resemble gitids, especially rasorinae (not found in Africa), but can be separated from the latter by having a deeply foveate pronotum and mesoscutum, as well as diagnostically liopterid metasomal terga. Within Afrotropical Liopteridae, Paramblynotus can be distinguished by lacking any scutellar armament, by the lack of any sort of lobe at the base of the tarsal claws, and the presence of an auricula on the side of the scutellum.
Distribution. e genus is represented in all biogeographical regions except for the Western Palaearctic and Australia , Ronquist 1995a. ree species groups are present in, and endemic to the Afrotropical region: the P. trisetosus and P. yangambicolus species groups (two of the seven species groups recognized by Liu et al. (2007)) and the P. seyrigi species group erected by van Noort and Bu ngton (2013).
Biology. e type female of Paramblynotus yangambicola was captured on a Drypetes gossweileri S. Moore (Euphorbiaceae) log in Democratic Republic of Congo (Zaire) (Benoit 1956b). Two females of P. yangambicola from Uganda are labelled "ex Coleoptera"; two other females from Uganda are labeled "ex Lepidoptera" (Ronquist 1995). Inferred association with Lepidoptera and Coleoptera and rearing of Paramblynotus yangambicolous from a rotten log, suggest that Paramblynotus species are parasitoids of beetle larva ). e extensive backward pointing ridges on the pronotum and mesoscutum in a number of species suggest an adaption for exiting from (or burrowing in to nd) concealed hosts in a con ned substrate such as dense leaf litter or rotten logs (van Noort and Bu ngton 2013). Ronquist (1995) proposed that these structures help with host tunnel negotiation. ese e ective backward pointing teeth would facilitate the negotiation of such substrates, preventing slippage and promoting forward movement down the tunnels or through the substrate.

Paramblynotus seyrigi species group
Erected by van Noort and Bu ngton (2013) to accommodate a single species that is likely to be a Madagascan endemic.
Diagnosis. Paramblynotus seyrigi has closest a nities with the two Oriental species groups P. ru collis and P. punctulatus of Liu et al. (2007). e P. seyrigi species group shares the sculptural arrangement of the vertex (large ocelli with three distinct carina extending to or between the toruli) with the two aforementioned Oriental species groups, but the lack of an occipital carina in combination with an absence of a pronotal crest or tooth (uniquely the posterior pronotal margin is represented by a swollen rim), reduced sculpture on the mesoscutum and a unique scutellar foveal character state separate it from these two groups. It is distinct from the two African species groups P. yangambicolus and P. trisetosus in a number of characters including a glabrous mesopleuron without sculpture, F1 equal in length to F2, and the presence of an angled latero-ventral pronotal margin.

Paramblynotus trisetosus species group
is is the most species rich group within the Afrotropical region with 28 described species (Liu et al. 2007, van Noort and. e species group is only known from the African mainland. Diagnosis. Species in this group are typically smaller than those in other species groups, and are the easiest to confuse with Figitidae. ey are characterized by having a at pronotal crest (or, pronotal crest absent); the mesoscutum is foveate-reticulate or with continuous transverse carina with fovea set in rows looking like saw teeth in lateral view; in most species, the speculum is perfectly smooth (gently striate in P. vannoorti); and the median propodeal area is distinctly delimited by lateral propodeal carinae, and posteriorly is not foveoate-reticulate. Careful attention to the metasomal sclerites will prevent confusing trisetosus-group Paramblynotus with Figitidae Liu, Ronquist & Nordlander, 2007 (Namibia) Paramblynotus jacksoni Liu, Ronquist & Nordlander, 2007 (Cameroon) Paramblynotus kekenboschi Liu, Ronquist & Nordlander, 2007 (Democratic Republic of Congo) Paramblynotus maculipennis Liu, Ronquist & Nordlander, 2007 (Democratic Republic of Congo) Oberthuerellinae is subfamily is represented by three genera, Oberthuerella, Tessmannella, and Xenocynips all of which are endemic to the Afrotropical region (Ronquist 1995a, van Noort andBu ngton 2012).

Diagnosis.
Oberthuerella can be readily distinguished from Xenocynips by having distinct metasomal terga (tergites 3-5) with the inter-tergal sutures not fused. e mesopleuron is also distinctly concave, the concavity forming an oblique, shallow femoral Tessmannella Hedicke, 1912 Diagnosis. Female antenna 13-segmented, subclavate; male 14-segmented. Face with reticulate to rugose sculpture and scattered pubescence. Pronotum coarsely rugose with median tooth or spine viewed laterally. Mesonotum with coarse variable sculp-ture, propodeum without pronounced side margins. Segment 1 of metasoma (petiole) three times as long as broad, segments 2-4 short viewed laterally and dorsally, segment 5 the largest. Metafemora with a rounded lobe between medial area and apex, tooth on metafemur angled, hind tibia with a distinct lobe apically, opposite the tibial spines. Scutellum with three foveae. Identi cation. Dichotomous and online interactive keys to species are available in Bu ngton and van Noort (2012) and van Noort (2004van Noort ( -2015. Distribution. Central African Republic, Democratic Republic of Congo, Equatorial Guinea, Gabon, Kenya, Republic of Congo (Bu ngton and van Noort 2012).

Xenocynips Kie er, 1910a
Diagnosis. Metasomal terga 3-5 fused, with intertergal sutures partially visible; lower mesopleuron horizontally striate. Tessmannella is most easily confused with Xenocynips; the fusion of terga in Xenocynips is a very reliable and clearly visible character. Additionally, most species of Xenocynips possess a dorsoventrally striate lateral aspect of the scutellum, posterior to the auricula; this is useful for specimens in which the metasoma is missing.
Identi cation. Dichotomous and online interactive keys to species are available in Bu ngton and van Noort (2012) and van Noort (2004van Noort ( -2015

Discussion
Hyper-diverse insect groups provide a challenge for identi cation, particularly from regions where they are poorly studied and where knowledge of generic and species diversity is wanting. Years of frustrated scrabbling through scattered historical literature to nd poorly illustrated and out of date keys to identify Afrotropical Hymenoptera led to the formulation of the Afrotropical Hymenoptera Initiative (AHI) to address this hiatus (van Noort et al. 2010). is project will provide a sorely needed synthesised resource to enable the identi cation of Afrotropical Hymenoptera to generic level, with a summary of described species and biology and introduction to the relevant literature. Richly illustrated user-friendly web-available keys will provide a tool for coping with the phenomenal biological diversity of the region, the production of which will require major international collaboration between specialists across the included taxa. e Cynipoidea were taken on as the rst phase of the project since the superfamily is reasonably diverse at generic level and currently actively worked on for the region. Hence ouput was attainable within a reasonable time period to set the approach for the remaining superfamilies. Groups such as the Chalcidoidea and Ichneumonoidea will provide a far greater challenge.
With the completion of this phase of the Afrotropical Hymenoptera Initiative project, the taxonomic knowledge base of cynipoids in the Afrotropical region joins a similar knowledge base for the Western Palearctic and Australian regions, with modern keys to genera available for most cynipoid groups (Fergusson 1986, Forshage and Nordlander 2008, Melika 2006, Nieves-Aldrey 2001, Paretas-Martinez et al. 2011. It may be premature to promise the treatment of another major region sometime soon. Since we have observed, and indeed are concluding here, that for Figitidae there is a substantial overlap, with mostly shared genera, between the Afrotropical and Oriental regions, the Oriental region might be the logical next step to address. Nevertheless, major collecting e orts still reveal new genera in Eucoilinae, and for the Cynipidae, the rich fauna of the region is only now being discovered (e.g. Tang et al. 2009, Melika et al. 2011a. When the Oriental fauna is better understood, combining this knowledge with that of the Western Palearctic might possibly make the treatment of the Eastern Palearctic a relatively easy task, but collection e orts from this region are still very scattered and often di cult to access. Additionally, the Oceanic fauna outside of Australia (the Paci c islands) presents a special problem as there are endemic radiations of Eucoilinae that are poorly understood from taxonomic and phylogenetic perspectives. is fauna has only fairly recently been subjected to preliminary assessment (Beardsley 1988(Beardsley , 1989(Beardsley , 1990(Beardsley , 1992a(Beardsley , 1992b and contrary to early assumptions has been shown, at least partly, to be a part of the shared Paleotropical fauna (Forshage and Bu ngton pers. obs.) Additionally, circumscription of the Oceanic fauna is further confounded by what appears to be repeated introductions of eucoilines from the west coast of North America (Bu ngton, pers. obsv.). e cynipid and gitid diversity are both relatively high in North America, and while Figitidae systematics of the region has become more stable in recent years, including an updated catalogue of Eucoilinae ), many taxonomic changes are still underway within Cynipidae, making the generation of an updated key to North American genera still some time o (Melika pers. comm.). However, once North America has been treated, the last major region to require extensive assessment is that of the Neotropics. E orts in the Neotropical region, however, are hampered by a general lack of comprehensive collections, as well as a truly infantile knowledge of diversity with respect to other regions. In a recent trip to Brazil, for instance, MB sorted undetermined Figitidae from a major biodiversity survery in Espirito Santo, recovering an estimated 400 morphospecies and, at the very least, 4 undescribed genera. However, we must note the recent important advances which are being published on cynipoid faunas of Chile  2006) provided a key to genera of the Neotropical region, this key itself is out of date, and likely covers only a portion of the actual generic-level diversity (and no species level diversity). Lastly, a major challenge to understanding the Neotropical cynipoid fauna, as in other tropical regions, rests squarely upon habitat destruction and disturbance, inhibiting our gaining knowledge of many species before they become locally rare and/or extinct.
For the Afrotropical region, patterns of cynipoid diversity have become clearer through this project. e most notable is an over-arching distribution pattern of high taxon a nity between the Afrotropical and Oriental regions. is parallels, and is probably in uenced by, host speci city across the various trophic levels, driven by the underlying high degree of oral similarity between the two regions. Both regions have their share of endemic genera and species, but they also possess a signi cant overlap. Two groups stand out immediately: Paramblynotus (Liopteridae) and Afrostilba (Figitidae). Paramblynotus is particularly diverse in both the Afrotropical (van Noort and Bu ngton 2013) and Oriental regions , with what appears to be very little species overlap (this study); elsewhere in the world, the genus is either rare (Bu ngton and Gates 2013) or not nearly as speciose ). e eucoiline Afrostilba can often be a dominant taxon in bulk sampling e orts, with many stable morphospecies shared between the Afrotropical (many countries sampled; this study) and the Oriental regions ( ailand, Indonesia, Bangladesh, India; Bu ngton pers. obsv.). However, until a revision of Afrostilba has been completed, the scope and accuracy of taxonomic comparisons across these regions is limited. Several other genera of Eucoilinae have been shown to have a distribution extending across the Afrotropical and Oriental regions, recently or in this paper. Originally described from Taiwan, Paradiglyphosema, Linoeucoila and Gastraspis, occur both in the Afrotropical and the Oriental regions. Bothrochacis, Afrostilba, Ealata and Nordlanderia were described from the Afrotropical region, but are now known to occur in both. Even genera such as Endecameris, Micreriodes and Leptolamina, originally described from the Paci c, have been shown to occur both in Asia and in Africa, suggesting that a substantial portion of the fauna is generally Paleotropical. For several of the mentioned genera (Linoeucoila, Gastraspis, Bothrochacis, Endecameris and Micreriodes), the geographical records allowing these observations are rst published here, for others (Afrostilba, Leptolamina) they are dependent on nomenclatural acts made here.
A few gitid lineages show unique diversi cation within the Afrotropical region. e Pycnostigminae are one of the most enigmatic cynipoid lineages sub-endemic to the region with 80% of the species restricted to sub-Saharan Africa (two genera Tylosema and Trjapitziniola are represented by single species in the Mediterranean region). e only known metallic-colored cynipoid, Pycnostigmus mastersonae, falls within this unusual group. Worldwide, Emargininae tend to be a rather rarely encountered and collected taxon (Bu ngton and Forshage pers. obsv.); however, within the Afrotropical region, and particularly Madagascar, the group can often dominate a bulk sample of cynipoids. e species-level diversity within Madagascar is spectacular (as noted above), and this group will make for an exciting species-level revision in the future. e aspicerine Anacharoides, and the anacharitine Acanthaegilopsis, are both genera unique to the Afrotropical region. Within Eucoilinae, a number of notable endemic or particularly species-rich groups have been recorded over the past three decades and are treated in this paper. Specialised morphological adaptations are uniquely exhibited by a number of Afrotropical taxa. Angustacorpa is bizarrely attened, and Stentorceps and Nanocthulhu are both characterized by arguably grotesque, unique protrusions from their frons and clypeal regions; the function of which is unknown. Hyperdiverse groups include Rhoptromeris, Hexacola and Didyctium, all of which are common outside the Afrotropical region, but were previously not known to include the specieslevel diversity observed here. Finally, Leptopilina and Trichoplasta appear to have a larger diversity in the region than elsewhere.
Since assessment of generic diversity, species richness and distribution of Afrotropical cynipoids is in its infancy, with major gaps in sampling e ort and habitat coverage, we can only forward hypothetical conjecture regarding biogeographical patterns based on limited data. Nevertheless, these hypotheses are a starting point and will be tested and revised as we proceed with the documentation of the region's hymenopteran fauna. Van Noort and Bu ngton (2013) hypothesized, based on data in Liu et al. (2007) and , that African Paramblynotus diverged from the remaining Palaearctic members of the genus between the late Oligocene to early Miocene periods (26-23 mya). Expanding this schema to Figitidae is certainly within reason, since Bu ngton et al. (2012) recovered the majority of tribe-to genuslevel diversi cation events within most subfamilies to have occurred at roughly the same time, especially within Eucoilinae. is would result in a few recently diverged, uniquely African groups that potentially evolved and diversi ed as a result of the aridication of the continent and formation of the savanna biome during the Oligoceneearly Miocene c. 33-20 mya (Couvreur et al. 2008, Sepulchre et al. 2006) (e.g. the Oberthuerellinae, Pycnostigminae, the aspicerine Anacharoides, and the eucoiline genera, Stentorceps and Nanocthulhu), as well as a few lineages present in other regions, but with unique species present only in Africa (e.g. Rhoptromeris and Afrostilba). Buington et al. (2012), however, did provide weak evidence of a Gondwanian element to the Afrotropical fauna. In that analysis, the thrasorines (southern South America), the mikeiines (Australian) and pycnostigmines (southern Africa) clustered in a clade. However, the posterior probability was very low (69; gure 1), and the stem-group divergence estimates for the group were centered around 75 mya, somewhat young for a Gondwanian explanation of distribution. Another question worth expanding is that of the origin of cynipoids, other than Paramblynotus, present on Madagascar. Since there is evidence of Madagascar separating from mainland Africa 160-120 mya (Ali et al. 2008), for groups with large numbers of endemic species, such as the Malagasy Emargininae, some form of dispersal event would be expected to have taken place from the great continent. is observation is reinforced by the fact that Bu ngton et al. (2012) recovered the emarginines to have a crown group age of 40 mya, making a hypothesis of vicariance from the African mainland unlikely for this group of cynipoids. However, Madagascar's cynipoid species richness and endemism is likely a combination of numerous historical evolutionary processes, including persistence of paleoendemic lineages and more recent rapid speciation of younger lineages, as holds for many other groups of animals and plants on this island (Vences et al. 2009, Burkei et al. 2013. Studies on endemic Afrotropical cynipoids (liopterids: Bu ngton and van Noort 2012, van Noort and Bu ngton 2013; aspicerines: Bu ngton and van Noort 2008; eucoilines: Nielsen and Bu ngton 2011, Bu ngton 2012; pycnostigmines: Bu ngton and van Noort 2007) recover a similar set of patterns with respect to distribution: A) a general 'hydrophilic belt' of diversity along equatorial Africa, with a 'southern swath' southward along the central eastern seaboard, ultimately ending in the more Mediterranean-esque portion of South Africa (aspicerines, some liopterids, many eucoilines and speci cally Stentorceps); B) a 'hydro-phobic' patchy distribution within South Africa, often focused on the Western Cape (but theoretically this should extend into Namibia and Botswana) with linkages to the arid north-eastern areas of Africa and the middle east (pycnostigmines, some liopterids, some eucoilines). However, it should be pointed out that few new samples are being generated from, or are presently in world collections for the majority of the African countries, and hence any assessment of ner scale biogeographical patterns based on current data is premature.
In terms of species richness hot spots the arid Sahel belt paralleling the southern edge of the Sahara desert and extending south down the eastern side of Africa through Ethiopia, Somalia and Kenya to Tanzania and the south-western aridness of Africa may potentially contain a reasonably diverse assemblage of cynipoids (as does the arid southwest of the United States and north Mexico, and arid Central Asia). With continued collecting in South Africa, this region may perhaps prove to be a hot spot of species richness in Africa. e uniqueness of the Cape Floral Kingdom, containing a diverse range of vegetation types (Mucina and Rutherford 2006), coupled with geographic characteristics typi ed by rugged, highly strati ed mountain ridges (Cape Fold Mountain belt) likely played a role in promoting speciation within the group. Likewise the East African Rift valley extending from Ethiopia in the north to Zimbabwe in the south, encompassing the rugged topography of eastern DRC, Rwanda and Burundi, which started formation with upliftment in the Eocene-Oligocene period (Sepulchre et al. 2006), together with the belt of eastern arc mountains in Tanzania, each with isolated Afromontane forest refugia (Burgess et al. 2007) will likely prove to be another rich area of species diversity and endemism, as holds for plants and vertebrates (Myers et al. 2000, Lovett 2005. ese eastern arc forests have strong connections with the Guineo-Congolian lowland rainforest and may have been isolated for 30 Myr from the start of the breakup of the pan African forest swath as a result of aridi cation in east Africa during the Oligocene-early Miocene (c. 33-20 Myr) (Couvreur et al. 2008). e vast expanse of the relatively homogenous Congo basin lowland rainforest [Congolian Region of Linder et al. (2012) equating to the Guineo-Congolian biome of White (1983)] may prove to harbour fairly widespread species, and hence exhibit low degrees of endemism. e forest has, however, undergone numerous contractions, fragmentations and re-expansion from the mid-Tertiary onwards (c. 33-2 Myr) (Couvreur et al. 2008, Marks 2010 promoting speciation and endemism, and coupled with this high energy tropical ecosystem, cynipoid species richness could arguably be expected to be elevated in the Congo basin. e majority of cynipoid species, world-wide, have unknown biological roles; the exception being Cynipidae where much of the taxonomy of the group is based on rearing records. Afrotropical cynipids are highly depauperate compared to the diverse northern hemisphere richness for the family, but their biology is relatively better known than their parasitoid counterparts in Africa. South Africa harbours a couple of enigmatic cynipids including an endemic, specialist lethal inquiline (van Noort et al. 2007) and gall formers of both herbs and trees (Liljeblad et al. 2011, Melika andPrinsloo 2007). Phanacis is poorly studied and indications, based on undescribed species at hand, are that the genus may be richer in the Cape Floral Kingdom than cur-rently perceived. As mentioned previously, the hosts of Liopteridae remain uncertain, though some association with wood-boring beetles has been discussed extensively in the literature . Figitids fall into three major categories of host preference: those associated with gall forming Hymenoptera (Euceroptrinae, rasorinae, Plectocynipinae, and Mikeiinae), those associated with aphid predators and parasitoids (Charipinae as hyperparasitoids, usually with braconid parasitic wasps as immediate hosts; Anacharitinae on Neuroptera; Aspicerinae on Diptera), and those associated with cyclorrhaphous Diptera (Figitinae and Eucoilinae, which compose the majority of Afrotropical cynipoid diversity). Hosts are completely unknown for Emargininae and Pycnostigminae; the former has been associated with driver ant refuse piles , while the latter, based on phylogeny, has been speculated to attacking gall-forming Hymenoptera (Bu ngton and van Noort 2012).
Eucoilines can be further divided into two major divisions of host use: those which utilize leaf-miners (Agromyzidae) hosts up in the canopy; and those which attack Diptera in habitats like decomposing organic matter (debris, dung, carrion, wood etc), or algae and mushrooms (e.g. Ephydridae, Drosophilidae, Muscidae, Calliphoridae) on the ground. Afrostilba and Nordlanderia are the dominant leaf-miner parasitoids, with Afrostilba common along equatorial Africa as well as down the 'southern swath'; Nordlanderia dominates in much more Mediterranean habitats, and these species are common in the Western Cape and along the southern coast of South Africa. roughout sub-Saharan Africa, the drosophilid-parasitic genera Leptopilina and Ganaspis are extremely common, some of which may be tramp species that are cosmopolitan (Bu ngton and Forshage pers. obsv.). However, throughout sub-Saharan Africa, it is the chloropid, drosophilid, and ephydrid parasitic Rhoptromeris and Hexacola respectively that are frequently the most abundant in bulk samples. Furthermore, there are members of both of these genera, as well as the unusual Nanocthulhu and Stentorceps, that all possess some form of protrusions of the frons and/or clypeal region. e frequency of this condition in the Afrotropical region is unparalleled in other biogeographic regions (Bu ngton and Forshage pers. obs.). While some of these genera are related (Nanocthulhu, Stentorceps, and Rhoptromeris, all belonging to Trichoplastini), Hexacola belongs to Ganaspini, and the facial protrusions are therefore an example of convergence with these other genera. While we do not yet know the hosts for these species, we do know the facial traits are not sexually dimorphic; ergo, we speculate these structures are used in escaping from the host puparium, the surroundings of the puparium (e.g. soil), or both. e authors hope this initial chapter on the Hymenoptera of the Afrotropical region marks a turning point in the larger understanding and appreciation of this incredibly diverse and important order of insects. e nal AHI production will provide an essential resource for identi cation of Afrotropical Hymenoptera by a diverse array of end-users, from specialists, ecologists, and conservationists, to the applied forestry and agricultural sectors, enabling e ective long-term conservation of an economically important and ecologically signi cant component of African and Madagascan ecosystems. Elucidating wasp systematics is a fundamental requirement for the future preservation of ecosystems that play an essential life support function for continued human survival.