A phylogenetic and taxonomic assessment of Afrotropical Micracidini (Coleoptera, Scolytinae) reveals a strong diversifying role for Madagascar

Afrotropical bark beetle genera in the tribe Micracidini are revised and an identification key provided. The new classification is based on phylogenetic analyses of five molecular markers (COI, EF-1α, 28S, PABP1, CAD) in combination with morphological characters. Five new genera are erected and one genus synonymized, resulting in a total of 11 valid genera: Lanurgus Eggers, 1920, Traglostus Schedl, 1938, Pseudomicracis Eggers, 1920 (=Saurotocis Wood, 1984 syn. nov.), Phloeocurus Wood, 1984, Afromicracis Schedl, 1959, Dendrochilus Schedl, 1959, Neomicracis Jordal gen. nov., Leiomicracis Jordal gen. nov., Diplotrichus Jordal gen. nov., Pseudolanurgus Jordal gen. nov., Microlanurgus Jordal gen. nov. The following new species are described to be included in the new genera: Leiomicracis aurea Jordal sp. nov., Neomicracis squamigera Jordal sp. nov., both from Tanzania, and Microlanurgus bicolor Jordal sp. nov. and Microlanurgus ater Jordal sp. nov., from Madagascar. The following new synonyms and new combinations are proposed: Afromicracis dubius (Schedl, 1950) (=Afromicracis angolensis Schedl, 1962 syn. nov.), Afromicacis elongatulus (Schedl, 1977) comb. nov.,Afromicracis jasminiae (Schedl 1957) comb. nov. (=Dendrochilus mikaniae Schedl 1957 syn. nov.), Afromicracis robustus (Schedl 1957) comb. nov. (=Dendrochilus arundinarius Schedl 1957 syn. nov., =Hypothenemus bambusae Browne, 1970 syn. nov., =Dendrochilus filum Schedl, 1977 syn. nov.) (all from Dendrochilus), Afromicracis setifer (Schedl 1957) comb. nov. (Mimiocurus), Lanurgus longipilis (Schedl, 1958) comb. nov., Lanurgus pubescens (Schedl, 1961) comb. nov. (both from Traglostus), Diplotrichus catenatus (Schedl, 1953) comb. nov.,Diplotrichus elongatus (Schedl, 1950) comb. nov.,Diplotrichus euphorbia (Schedl, 1961) comb. nov.,Diplotrichus gracilis (Schedl, 1958) comb. nov.,Diplotrichus minor (Schedl, 1950) comb. nov (=Lanurgus frontalis Schedl, 1953 syn. Nov.), Diplotrichus obesus (Schedl, 1953) comb. nov., Diplotrichus pygmaeus (Schedl, 1965) comb. nov., Diplotrichus rugosipes (Schedl, 1961) comb. nov., Diplotrichus subdepressus (Schedl, 1965) comb. nov., Diplotrichus widdringtoniae (Schedl, 1962) comb. nov. (all from Lanurgus), Diplotrichus ignotus (Schedl, 1965) comb. nov. (Pseudomicracis), Pseudolanurgus harunganae (Schedl, 1961) comb. nov. (=Lanurgus cribrellus Schedl, 1965 syn. nov.), Pseudolanurgus bugekeae (Schedl, 1957) comb. nov. (both from Pseudomicracis), Pseudolanurgus minutissimus (Schedl, 1961) comb. nov. (Lanurgus), Pseudomicracis dispar (Schedl, 1961) comb. nov., Pseudomicracis tomicoides (Schedl, 1961) comb. nov. (both from Saurotocis). The following taxa were transferred to genera in other tribes: Acanthotomicus intermedius (Schedl, 1977) comb. nov., Xylocleptes villiersi (Lepesme, 1942) comb. nov. (both from Dendrochilus); Eidophelus agnathus (Schedl, 1942) comb. nov., and Eidophelus ciliatipennis (Schedl, 1979) comb. nov. (all from Miocryphalus). The following five species were included in Karlseniusgen. nov. (Trypophloeini): Karlsenius klainedoxae (Schedl, 1957) comb. nov., Karlsenius nitidum (Schedl, 1965) comb. nov., Karlsenius nigrinum (Schedl, 1957) comb. nov., and Karlsenius attenuatus (Eggers, 1935) comb. nov. (from Miocryphalus), and Karlsenius ghanaensis (Schedl, 1977) comb. nov. (from Eidophelus). A time-tree and biogeographical analysis suggested that Madagascar was colonized only once in Micracidini, from East Africa soon after the origin of the tribe in late Cretaceous. Multiple re-colonisations from Madagascar to the mainland have contributed to further diversification of a tribe which is otherwise highly restricted in geographical distribution.


Introduction
Most bark beetles that live in dry, old twigs and branches are of little economic importance. Hiding away in their largely cryptic or forgotten habitats, these beetles are often overlooked in field collections. The diversity of tiny dryadapted bark beetles is nevertheless high and numerous species have been described from these kinds of habitats, often based just on a single or a few collecting events. In Africa and Madagascar such habitats are typically occupied by species in the genera Hypothenemus Westwood and Afrocosmoderes Johnson and Jordal (Trypophloeini), Glostatus Schedl (Xyloctonini), and several genera in the tribe Micracidini. The taxonomy of tiny and obscure beetles is often in flux, or largely ignored, as one may find it difficult to diagnose species properly. This is mainly the case for Micracidini where recent phylogenetic work has pointed towards many errors in the classification, particularly in the Afrotropical fauna (Jordal and Kaidel 2017).
Members of Micracidini are found in the Neotropics and southern parts of the Nearctic, and in the Afrotropical region, including Madagascar. More than 160 species have been described from the first two regions, whereas Africa and Madagascar are much less explored with fewer species known. The Afrotropics is nevertheless expected to have a similar diversity of species if given more taxonomic attention. Such inventories are accumulating, but the circumscription of meaningful genera has been a challenge. Micracidini included previously 13 genera globally (Wood 1986), with recent additions of two new Neotropical genera (Bright 2010(Bright , 2019 and inclusion of two genera from other tribes (Jordal and Kaidel 2017). Five additional new genera are here defined by molecular and morphological characters, and one genus is synonymized, with a new total of 21 genera (Table 1).
The biology of many species in Micracidini is unusual and deviates from the more typical bark beetles in several aspects. They are often bigamous where two females pair up with a colonizing male. This is a rare mating system in nature; in bark beetles, it seems more common to establish harem polygamy if more than one female should be accepted (Kirkendall 1983;Kirkendall et al. 2015). Females in this tribe are often gracefully ornamented, with long golden setae in the forehead and on the antennal scapus which is often enlarged or extended into a spine (Fig. 1). They use these structures in tactile communication with the male to gain access to a tunnel entrance in the bark. After mating, the two females extend the engravings of two independent tunnels where they deposit eggs in individual egg niches. Secondary sexual characters in the head region are species specific and provide the most useful characters to distinguish closely related species. Males in some genera can be more elaborated on the elytral declivity, with tubercles and mucronate elytral apices. In Cactopinus, and occasional other micracidine species, the male rather than the female has a concave frons, with horn-like projections (Atkinson 2010;Jordal and Kaidel 2017).
Taxonomists have focussed mainly on sexually evolving characters as a mean to define species. As a potential consequence, the classification at higher taxonomic levels became poorly developed. There are now obvious mixtures of species in each genus that belong to different genera in Micracidini (Jordal and Kaidel 2017) or even from other tribes (Johnson et al. 2020). Taxonomic inconsistencies were likely fostered by a general lack of diagnostic characters at the genus level. These problems are perhaps connected to a very ancient origin of Micracidini, which is considered the oldest monophyletic tribe in Scolytinae (Jordal and Cognato 2012;Pistone et al. 2018;Jordal and Kaidel 2017).
Molecular data from five genes are included in this study to provide a robust phylogeny which enable assessment of morphological consistency in groups of similar evolutionary age. New diagnostic features are supported by phylogenetic analyses and incorporated in a new identification key to genera which will improve accurate identification. Molecular data also allow for testing biogeographical scenarios-more specifically the role of Madagascar in the diversification of Afrotropical micracidines. Are species in Madagascar monophyletic with respect to the mainland, with few, or even a single, very ancient origin? If not, are Malagasy taxa paraphyletic, and of more recent origin, with adventive African lineages on the island? Alternatively, multiple recolonizations of the mainland could have happened, which is more in accordance with historical geophysical conditions (Ali and Huber 2010).

Samples
Type material of the type species for each genus was examined in museum collections in Vienna (NHMW), Tervuren (RMCA), and Paris (MNHN). In the preparation of revisions of the Afrotropical genera, the main types for each species have been studied (holotype, holotype with allotype and/or paratypes, or Egger's 'type', except Miocryphalus ciliatipennis Schedl, 1979). Fresh material useful for DNA analyses (Table 2) was collected in several African countries between 1998 and 2015, after obtaining the necessary collecting and export permits as required at the time of field work.

Molecular data
Sequence data were obtained from five genes: Cytochrome Oxidase I (COI, 690 bp), Elongation Factor 1a (EF-1a, 822 bp), Carbamoyl-phosphate synthetase 2, aspartate transcarbamylase, and dihydroorotase (CAD, 458 bp), Poly-A binding protein 1 (PABP1, 435 bp), and the large ribosomal subunit (28S, 720 aligned nucleotide positions). DNA extraction, PCR and sequencing followed Mugu et al. (2018). The ribosomal alignment was made in Muscle (Edgar 2004) given default settings and a portion of ambiguous alignment sites were pruned in GBlock (Castresana 2000) using the most liberal settings possible (allow smaller blocks, allow gaps, allow less strict flanking regions, allow contiguous non-conserved positions). GenBank accession numbers are listed in Table 2.

Phylogenetic and biogeographical analyses
Nucleotide sequences from all five genes were concatenated and analysed in combination with the 36 morphological characters in MrBayes v 3.2.6 (Ronquist and Huelsenbeck 2003). Best models for each gene partition were selected in MrModelTest (Nylander 2004). Models for morphological characters were given a gamma-distribution based on empirical character variation. Stationarity in likelihoods was visualized in Tracer (Rambaut et al. 2014) and was reached before 3 million generations set as burn-in.
Approximate age for clades were estimated in Beast 1.10.4, with the xml file prepared in Beauti (Drummond and Rambaut 2007). Rates were calibrated with time estimates from a Biogeographical scenarios were tested using model selection in RASP (Yu et al. 2020) using the time tree estimated in the Beast analysis. Areas in Africa were defined using biogeographical regions proposed by Linder et al. (2012): Southern African, Zambesian (eastern Africa), and Congolian (western Africa) regions, in addition to Madagascar and the Neotropical/Nearctic regions. The best model was selected among DEC, DIVAlike and BayArea using BioGeoBEARS, with or without the extra parameter for jump dispersal (+j). Because of the uncertain foundation for testing +j models (Ree and Sanmartín 2018), it was presumed that jump dispersal was more realistically involved in overseas splits, while vicariance motivated models can be more relevant in the inference of widespread taxa on the African mainland. Analyses with or without jump dispersal integrated in the model were furthermore compared to a Bayesian Binary MCMC analysis in RASP (Yu et al. 2015), using default parameters.

Phylogeny
Bayesian inference of 3226 nucleotides and 36 morphological characters resulted in a well-resolved tree topology with high posterior probability for most nodes (Fig. 45). The inclusion of morphological characters did not change the tree topology, but increased node support for three nodes. Micracidini were monophyletic, with the Neotropical genera forming the sister group to all but one of the Afrotropical species. The excepted species formed the sister group to all Micracidini and is here described as a new genus, Leiomicracis.
The remaining Afrotropical taxa were distributed on two main clades. One was composed mainly by species in Afromicracis and its sister group, Dendrochilus. These two genera combined were placed as sister to a single species which is here described in the new genus Neomicracis. The other main clade included the genera Lanurgus and Pseudomicracis and three additional lineages described as new genera. One such lineage consisted of two tiny species with spatulate setae on odd-numbered interstriae only. This new genus, Microlanurgus, is not closely related to any other taxa (Fig. 45). Some species which are currently placed in Lanurgus have bifid setae on the metaventrite (and abdominal ventrites I-II) and all these species formed a lineage described as a new genus, Diplotrichus. This  included here, only Dendrochilus has elongated antennal clubs, otherwise seen in a few species of Pseudomicracis (previously Saurotocis). 8. Club sutures, anterior face: 0 two transverse, 1 two procurved, 2 two bisinuate, 3 one transverse and one recurved, 4 two partial, 5 first partial and second transverse, 6 terminal only, 7 all pubescent, 8 entirely corneous. CI= 0.62, RI=0.76. Most species have two procurved or transverse sutures marked by dense setae. A single suture near apex is typical for Afromicracis, two bisinuate sutures define Pseudolanurgus, a continuous corneous area over segments 1 and 2 with suture 1 partial defines Neomicracis, two slightly recurved sutures define Microlanurgus, and a suture-free and shiny club defines Dendrochilus. Pronotum 9. Pronotal asperities on anterior half: 0 present, 1 absent on anterior fifth, 2 few and coarse. CI=1.0, RI=1.0. All except two genera have asperities rather evenly distributed on the anterior half of the pronotum. In Microlanurgus the asperities are few and coarse-grained and not fully reaching the front margin, while in Leiomicracis the anterior fifth is smooth without any asperities. 10. Pronotal setae: 0 hair-like, 1 bristle-like, 2 scale-like. CI=0.13, RI=0.58. Varies between bristle-like and hairlike within genera. Particularly broad setae are observed in Neomicracis and the Diplotrichus catenatus species group. Elytra 11. Interstrial setae: 0 absent, 1 in regular rows, 2 confused, 3 on odd-numbered interstriae only. CI=0.33, RI=0.69. These setae are densely placed and confused at least on the posterior-lateral areas of the elytra, but more often on all interstriae, in Lanurgus and a few other species.
Microlanurgus is unique in the tribe bearing spatulate setae only on odd-numbered interstriae. 12. Interstrial setae, shape: 0 hair-like, 1 bristle-like, 2 spatulate. CI=0.29, RI= 0.58. Most genera have mainly spatulate setae, but occasionally hair-like or bristle-like shapes occur in some Afromicracis and in several of the Neotropical genera. 13. Strial setae: 0 absent, 1 tiny in rows, 2 broad and confused. CI=0.5, RI= 0.86. These setae are broader, almost scale-like or plumose, in Lanurgus and a few Pseudomicracis males; apparently missing in most Hylocurus. 14. Elytral apex: 0 rounded, 1 mucronate, 2 slightly extended flange. CI=0.33, RI=0.73. Typically mucronate in Micracis, Hylocurus, Micracisella, and in In the latter it takes the shape of a bottleneck. Legs 17. Lateral edge of protibia: 0 straight, 1 twisted. CI_0.20, RI=0.60. Most scolytines and some micracidines have their protibia flattened with denticles on its lateral edge. In most micracidines the inner mucro is large and the lateral edge is twisted and rotated posteriorly towards the mucro, but is not twisted in Hylocurus, Micracis, Micracisella, Cactopinus, Thysanoes and Phloeocurus. 18. Protibial denticles: 0 along the edge only, 1 mainly on posterior side. CI= 0.25, RI=0.40. Placed along the edge in outgroups, Cactopinus, one Afromicracis, and at least partly so in Micracisella. 19. Protibial transverse denticles: 0 none, 1 one, 2 two, 3 three, 4 four+. CI= 0.31, RI=0.70. The number of denticles that deflects from the lateral edge towards the inner mucro varies within the tribe, but nearly all species have at least one. The common state in Afrotropical species is two denticles, with typically three in Lanurgus and Dendrochilus, and a single denticle in Leiomicracis, and one posterior and one lateral in Microlanurgus. is straightened out in many other tribes, but generally zig-zag shaped in all Micracidini. Hind wings 25. Precostal setae: 0 zero, 1 one, 2 two, 3 three. CI=0.50, RI=0.87. The normal condition in Micracidini is three closely set setae near the junction of subcosta and costa. Afromicracis, Dendrochilus, Neomicracis, and possibly Microlanurgus, have only a single seta. 26. Setae on radial cell (pterostigma): 0 none, 1 one, 2 two, 3 three. CI=0.43, RI =0.71. All Micracidini have one, two, or three setae on the apical portion of the radial cell, most commonly two setae, but frequently three setae in Pseudomicracis. Male aedeagus 27. Apophyses length: 0 1-2 times as long as aedeagal body, 1 extremely elongated, 2 shorter than aedeagal body. CI=0.33, RI=0.80. Shorter than aedeagal body in Thysanoes, no more than twice as long in Neomicracis, Lanurgus, Microlanurgus, Micracis, Hylocurus and Micracisella; extremely elongated in all others. 28. Tegmen: 0 ring, 1 Y-shaped, 2 tube-shaped, 3 absent. CI=1.0, RI=1.0. The normal condition is a closed tegmen forming a ring, with or without a manubrium (median struts). This feature is extremely difficult to assess in many taxa, and is apparently absent in Diplotrichus, Pseudolanurgus and Pseudomicracis; it is prolonged as a tube in Cactopinus, Stenoclyptus and Phloeocleptus, and variably reduced to a small Y-shaped piece in Dendrochilus and Afromicracis. 29. Manubrium (median struts): 0 absent, 1 no longer than tegmen thickness, 2 several times longer than tegmen thickness. CI=0.4, RI=0.8. For those with a normal tegmen, the manubrium is long and thin in the Neotropical genera, and in Neomicracis.
genus was more closely related to Pseudomicracis and a lineage erected as a new genus, Pseudolanurgus, which has a less mucronate elytral apex compared to Pseudomicracis. These three genera together were defined as a group by having a greatly enlarged and prolonged flagellum in the male aedeagus with the loss of tegmen and spiculum gastrale. Saurotocis, here represented by the type species S. dispar (Schedl, 1961), was furthermore nested within Pseudomicracis with maximum support. A separate parsimony analysis of 36 morphological characters resulted in 68,947 trees. Nearly all genera as supported by molecular data were monophyletic, with two genera forming polytomies not contradicting monophyly (Afromicracis and Diplotrichus). Most characters performed well in terms of high retention indices, with 28 characters obtaining a retention index above 0.6 irrespective of optimization on the Bayes combined analysis topology or one of the parsimony tree topologies. Characters with the lowest performance had RI between 0.4 and 0.6 and these less optimal characters included features of the male frons (character 2), shape of pronotal and interstrial setae (10, 12), shape of the lateral edge of the protibiae (17) and the position of its denticles (18), the shape of the suture and teeth on the apical plate of the proventriculus (32, 33) and its closing teeth (34).

Biogeography
Model selection based on the AIC criterion in RASP using the Beast tree suggested DIVAlike (with or without the additional parameter for jump dispersal) as the best model for biogeographical analysis. The BBM reconstruction (Fig. 46) was similar to the DIVAlike+j analysis. Both of these differed from the DIVAlike analysis primarily by the simultaneously dispersal and vicariance events in accordance with jump dispersal theory, while DIVAlike always estimated dispersal one node before the inferred vicariance event, suggesting expansion of the ancestral area with subsequent vicariant splits in geographical isolation.
BBM estimated an origin of the tribe in Eastern Africa, which occurred about 78.5 Ma (Table 4), and several subsequent range expansions took place from this area. Colonization of the Neotropics occurred in a single event not long after the origin of the tribe, around 69.3 Ma. Madagascar was colonized around 59.6 Ma, from eastern Africa, and thereafter colonized Southern Africa from Madagascar twice (Lanurgus 58 Ma, Diplotrichus subclade 29.5 Ma) and Eastern Africa once (Pseudolanurgus 31.5 Ma). The Afromicracis clade likely originated in Eastern Africa and expanded its ranges towards the west 47.4 Ma and more recently moved several times between various African regions. This clade never expanded its range outside the African mainland.
Diagnosis. Eyes entire, short (longer in Micracis and Micracisella), female scapus usually with a small to large tuft of golden setae, funiculus 5-or 6-segmented, club moderately to strongly flattened, usually with two sutures, occasionally without; labial palps long, basal segment obliquely triangular and thereby separating diverging palps, segment 3 narrow, twice as long as segment 2. Pronotum weakly to strongly dome-shaped, asperate on anterior half. Elytra usually with rows of scale-like setae (hair-like in Cactopinus, some Hylocurus and some Afromicracis). Protibiae either parallelsided or twisted, with one or more denticles apically on posterior side. Hindwings with either 1 or 3 setae on precosta, 2-9 on costa, and usually 2 (1-3) on distant margin of the radial cell (pterostigma). Scutoscutellar suture parallel to the scutellar grove for most of its length. Pleural suture zigzag shaped. Postnotum separated from metanotum by a membrane. 30. Flagellum: 0 absent, 1 sac-like, 2 long, thin and coiled, 3 extremely enlarged. CI = 0.60, RI = 0.95. The flagellum is unusually long and enlarged in Pseudomicracis, Pseudolanurgus and Diplotrichus, much thinner in Neomicracis, Dendrochilus and Afromicracis, in the latter very long and coiled. In Micracis, Hylocurus and Micracisella the inner sac is inflated, which may or may not be a homologous structure. 31. Spiculum gastrale: 0 simple rod, 1 rod with apical fork, 2 absent. CI = 0.67, RI = 0.97. Missing in Pseudomicracis, Pseudolanurgus and Diplotrichus. The speculum contains an apical fork in all Neotropical genera except Thysanoes, and simple rod elsewhere. Proventriculus 32. Proventriculus with median suture: 0 tight, 1 wide open. CI = 0.20, RI = 0.50. All Micracidini has a longitudinally divided apical plate, but the median suture is occasionally rather narrow as in some Lanurgus and Cactopinus. 33. Apical plate with: 0 sharp teeth, 1 rows of tubercles, 2 smooth. CI = 0.50, RI = 0.50. In most species the teeth on the apical plate are very sharp and placed in lines. Occasionally they are less sharp and more granulate as in Cactopinus, and nearly so in Leiomicracis and Microlanurgus. 34. Closing teeth: 0 finely plumose, 1 serrated. Most micracidines have plumose or weakly dentated closing teeth. In Dendrochilus the closing teeth are not spreading but compact in a serrated pattern. 35. Crop spines: 0 absent, 1 tiny and dispersed, 2 bundle of long spines, 3 few spines on a tubercle per plate, 4 one or two spines per plate. CI = 0.36, RI = 0.74. Micracidini species typically has bundles of enlarged spines in the crop, pointing posteriorly towards the lumen of the proventriculus. These spines are small and dispersed in Cactopinus, Stenoclyptus, Micracisella, Neomicracis and Leiomicracis. In Dendrochilus and Afromicracis these spines are shorter and placed on a tubercle, in Dendrochilus with only one or two such spines. 36. Termination of lateral teeth: 0 gradual, 1 U-shaped and strongly enforced. CI = 1.0, RI = 1.0. Lateral teeth on the apical plate are usually faint but form an enforced and sharp U-shaped ridge in Lanurgus.
Anterior plate of the proventriculus longitudinally divided and broadly separated (tight suture in some Lanurgus and most Cactopinus), transversely set by sharp teeth, or ridges (Cactopinus), closing teeth plumose or serrated, sometimes compacted. Male genitalia usually strongly elongated by very long apophyses, tegmen y-or ring-shaped, or absent; if Etymology. From Greek Leios, meaning smooth, referring to the lack of asperities on the anterior part of the pronotum, and the Greek genus name Micracis. Gender feminine as explained for Micracis by Alonso-Zarazaga and Lyal (2009).
Diagnosis. Typical micracidine by the twisted lateral edge of the protibiae with only a single socketed denticle present on the posterior face; inner mucro large, curved posteriorly. Female scapus longer than funiculus, with a triangular tuft of setae on its dorsal side. Antennal funicle 5-segmented; club setose, without sutures. Pronotum asperate in middle, smooth on posterior half, lateral thirds and anterior sixth. Interstriae with a single row of scale-like setae. Setae on ventrites simple, unifid. Hindwing precosta with a single seta, radial cell with two setae. Proventriculus with anterior plate much longer than posterior plate, longitudinally, broadly divided, with bundles of strong, long crop teeth; apical teeth enforced, forming U-shaped structure.

Distribution. Tanzania.
Remarks. Leiomicracis is phylogenetically isolated as the sister group to all other Micracidini. The basal position in Micracidini was also confirmed by adding DNA sequences to the previously published 182 Scolytinae taxa dataset of Pistone et al. (2018). The new genus is distinguished from all other genera in the tribe by the smooth anterior sixth of the pronotum and by the scattered long setae forming a triangular tuft on the scapus. Etymology. Latin adjective aurea meaning golden or splendid, referring to the golden-brown colour of the most mature specimens, and the elegant golden setae on the female scapus.
Diagnosis. Female scapus with long golden setae on its dorsal side increasing in length towards its distal end, tuft appearing triangular. Pronotum with asperities in middle, reaching only to anterior sixth. Protibiae with a single denticle on its posterio-lateral side. Proventriculus with closing teeth compacted and coarsely serrated.
Description, female. Length 1.1-1.3 mm long, 2.4-2.5 × as long as wide. Colour light golden or bronze brown (possibly young individuals). Frons flat to weakly impressed, surface reticulate, obscurely punctured; scattered fine setae, slightly thicker and denser on epistoma. Antennal scapus longer than funiculus, dilated apically, with a tuft of increasingly long golden setae on its dorsal side; funiculus 5-segmented, pediculus about as long as segment 2 and 3 combined; club flat, 1.6 × as long as wide, pubescent, with a faint trace of one procurved suture. Pronotum distinctly humpbacked, on median half from summit to anterior sixth with sharp asperities, on posterior half, laterally, and anterior sixth smooth, with Myr strial setae. Ventral vestiture of fine, rather long setae. Legs. Protibiae evenly rounded on lateral and inner sides, ending in an apical central mucro which is rather strong, curved posteriorly; one additional tiny socketed denticle on posterior side close to tarsal insertion. Metatibiae narrow, sides parallel, with distinct, fine, long inner mucro, and three transversely set apical denticles. Proventriculus. Apical plate 2.5 × as long as posterior plate, with regular transverse rows of sharp triangular or obovate teeth. Apical teeth connected in an U-shaped formation with 2-4 longer, sharp teeth. Closing teeth short, rather smooth.
Biology and distribution. Only known from the lowland type locality near the Udzungwa mountains in Tanzania. Six young female specimens were taken from a small twig of an unknown host plant.

Etymology. The Greek prefix Neo-, new, reflects the designation of a new genus in Micracidini. Gender feminine (Alonso-Zarazaga and Lyal 2009).
Diagnosis. Typical micracidine with twisted protibia which has apical denticles on its posterior side, and a long tuft of setae on a dorsally strongly elongated female scapus. Antennal funicle 6-segmented; club with first suture absent on median 3/4 and the second suture complete, transverse, surface of segment 1 and 2 fused, corneous. Elytral apex slightly inflated. Setae on pronotum in both sexes and on ventrites III-V in the males broad and scale-like. Apical margin of metatibia transverse. Precosta in hind wings bearing a single seta, radial cell with two setae. Proventriculus with fine, dispersed crop spines; apical plate shorter than posterior plate. Male genitalia with tegmen Y-shaped, manubrium long; spiculum gastrale a simple rod.

Distribution. Tanzania.
Remarks. This genus is superficially similar to the new genera Pseudolanurgus and Diplotrichus described below, having an inflated elytral apex, but differ from those genera by the elongated horn on the scapus. Neomicracis is unique in the tribe by having corneous first two segments of the antennal club (first suture interrupted), and by the broad scale-like setae on ventrites III-V in the males. It differs further from the sister group, which consists of Dendrochilus and Afromicracis, by the 6-segmented funiculus.
Etymology. Latin adjective composed by squama, meaning scale, and -ger, meaning bearing, referring to the broad scale-like setae associated particularly with asperities on the pronotum.
Diagnosis, female. Pronotum with very broad scale-like setae. Scapus short, bearing dorsally an elongated spine with long setae pointing inwards; funiculus 6-segmented; frons smooth and slightly impressed on central third; scutellum with four short setae in a transverse row. Male similar to female except scapus elongated, frons flat, reticulated, and ventrites III-V with very broad and long scale-like setae.
Description, female. Length 1.1-1.2 mm long, 2.4-2.6 × as long as wide. Colour brown. Frons flat, weakly impressed and impunctate on central one-third, coarsely reticulated elsewhere; scattered fine setae present around impressed area, slightly thicker and much longer, dense setae on epistoma. Antennal scapus shorter than funiculus, strongly elongated on its dorsal side into a long triangular horn, with a tuft of long forwardly pointing setae; funiculus 6-segmented, pediculus about as long as segment 2-4 combined; club flat, as long as wide, with suture 2 transverse, suture 1 absent in median three quarter, segments 1 and 2 fused, corneous, shiny. Pronotum short, much wider than long, summit slightly elevated, anterior slope short, steep; anterior half with sharp, almost quadrated asperities, each associated on its posterior side with a broad scale-like setae; on posterior half generally smooth, reticulated, in median posterior area slightly granulated; vestiture consisting of fine, semi-erect setae. Basal and lateral margin rounded. Scutellum U-shaped, wider than long, with 3-4 short setae. Elytra finely rugose, finely granulated, on declivity granules sharper; punctures faint, confused. Vestiture consisting of regular interstrial rows of slightly curved, pointed or slightly spatulate setae, and fine, short semi-erect strial setae. Ventral vestiture of fine, rather long setae. Legs. Protibiae with lateral edges evenly rounded towards a large, posterio-laterally curved terminal mucro, appearing slightly twisted; one lateral socketed denticle and one additional denticle on posterior side close to tarsal insertion. Metatibiae narrow, sides parallel, with a distinct, fine, long inner mucro, and two transversely set apical denticles.
Proventriculus. Apical plate slightly shorter than posterior plate, with regular transverse rows of sharp triangular. Apical teeth not apparent. Closing teeth smooth, long; femoral teeth sharp, connected to base of closing teeth.
Male. Similar to female except length 1.0-1.1 mm, 2.2-2.5 × as long as wide; frons rugose, lightly impressed on lower half; upper margin of impressed area marked by two small tubercles; antennal scapus short, rounded; setae on abdominal ventrites III-V very broad and long, one transverse row on each ventrite, pointing posteriorly. Genitalia elongated, apophyses as long as penis body, tegmen open dorsally, Yshaped, with long manubrium; spiculum gastrale a simple rod, as long as penis.
Biology and distribution. Only known from the type locality in the Udzungwa mountains in Tanzania. It was taken from a 1 cm thick twig of an unknown host plant where it was excavating longitudinal but irregular egg tunnels.

The Afromicracis clade
This group consist of two genera defined by a long, narrow scapus with scant setae in both sexes, and a 5-segmented funiculus. The precosta on the hind wing has only a single seta, the inner mucro of the metatibiae is long and thin, and in males the tegmen is minute and Y-shaped, and a very long, thin and often coiled flagellum runs through the aedeagus.
Proventriculus with closing teeth compact and serrated.

Distribution. Congo, Tanzania.
Remarks. The type species differs from other species previously included in this genus by the elongate, shiny and suture-free antennal club, and by the peculiar bottleneckshaped apical part of the last ventrite. Two very similar, but undescribed, species grouped as the sister to Afromicracis by maximum support. Both genera share the presence of a long scapus in both sexes, 5-segmented funicle, and a reduced Yshaped tegmen. Most species assigned to Dendrochilus by Schedl belong to Afromicracis (see below).
Afromicracis Schedl, 1959. =Miocryphalus Schedl, 1939, synonym by Alonso-Zarazaga and Lyal (2009, name unavailable Type species. Afromicracis kenyaensis Schedl, 1959. (Figs. 56, 59, 62) Diagnosis. Antennal scapus in both sexes straight, longer than funiculus, slightly dilated but not strongly inflated; in females with a fine tuft of scant long setae; funiculus 5segmented; club on anterior face usually with one procurved suture close to apex. Elytra with scale-like or hair-like setae, always in single rows. Protibiae twisted, with 2 or 3 denticles on latero-posterior side near apex. Ventral setae always simple, never bifid or plumose. Proventriculus with cluster of few crop spines attached to a tubercle base. Male genitalia with a long flagellum, sometimes shorter and broad, sometimes very long and coiled; tegmen reduced to absent, apophyses more than twice as long as aedeagal body.

Distribution. Tropical Africa.
Remarks. Afromicracis was removed from synonymy under Miocryphalus by Alonzo-Zarazaga and Lyal (2009), with the latter taxon deemed nomen nudum. Many species were erroneously assigned to Afromicracis (see Wood and Bright, 1992) and are now placed in many different genera in different tribes. Although species are small in size, they do have sufficient characters for reliable classification as noted in the diagnosis.
Etymology. Adding the Greek prefix Micro-, meaning very small, to the genus name Lanurgus (masculine), referring to the size of species which are shorter than 0.8 mm.
Diagnosis. This genus consists of very small and stout species (<0.8 mm), with the typical twisted shape of micracidine protibiae which have only one lateral denticle and a single additional denticle on its posterior face. The scapus is much shorter than the 6-segmented funiculus, in females with a little tuft of setae on its dorsal side. The antennal club has two sutures where suture 1 is transverse and suture 2 slightly recurved, appearing ring-like. Unique features include a pronotum which is strongly domeshaped, with few but large asperities which are not fully reaching the anterior margin; basal and lateral margins rounded. Elytra have scattered, erect, spatulate setae on odd-numbered interstriae only. Scutellum triangular, as wide as long. Male genitalia have tegmen shaped as a complete ring without manubrium, and the apophyses are not longer than the aedeagal body. Proventriculus with apical plate longitudinally divided, much shorter than posterior plate, with few sharp triangular teeth. Lateral, apical and femoral teeth absent; closing teeth lightly serrated, masticatory brush appears spiny.

Distribution. Madagascar (southern parts).
Remarks. This is the only micracidine genus with erect elytral setae only on odd-numbered interstriae. The proventriculus and protibiae are typical micracidine and molecular data firmly placed the two included species in the tribe ). There are no obvious close relatives based on morphological comparison or by phylogenetic analyses, but one may note that the male genitalia is quite similar to species in Lanurgus. (1). Holotype and 4 paratypes in CAS, 2 paratypes each in ZMBN and NHMW.
Etymology. The Latin adjective bicolor, meaning twocoloured, refers to the light yellow and black colours of the elytra.
Diagnosis. Pronotum with about 20 coarse asperities, 2 or 3 of these near the anterior margin. Elytra bicolored, spatulate setae present on odd-numbered interstriae only, each separated by 2-3 times the length of a seta.
Description, female. Length 0.6-0.7 mm long, 2.0-2.2 × as long as wide. Colour yellow and black; primarily yellow on pronotum and posterior part of elytra, venter, and anterior part of elytra black. Frons flat, shiny and impunctate on central half from epistoma to upper level of eyes, reticulate elsewhere; scattered, short, fine setae present around glabrous central area. Antennal scapus much shorter than funiculus, rounded, with a scant tuft of long setae on its dorsal side; funiculus 6-segmented, pediculus about as long as segment 2 and 3 combined; club flat, slightly longer than wide, suture 1 transverse, suture 2 recurved, appears ring-like in dorso-lateral view. Eyes separated by 2.7-2.9 × their width. Pronotum strongly dome-shaped, on anterior two-thirds with large, rounded asperities, only 2-3 of these reaching near anterior margin. Vestiture consisting of mixed spatulate and fine setae, longest setae near pronotal summit. Elytra smooth, shiny, stria not impressed, punctures shallow, large compared to body size, interstrial punctures obscure. Vestiture consisting of rows of erect, spatulate setae on odd-numbered interstriae, and scattered, fine and very short strial setae. Ventral vestiture of fine, rather long setae. Legs. Protibiae with lateral edges subparallel, inner mucro strong, curved posteriorly; two tiny socketed denticle transversely set apically on posterior side close to tarsal insertion. Metatibiae broader apically, with a distinct, fine, long inner mucro, and one apical and two laterally placed thin denticles.
Male near identical to female, except fewer and shorter seta on scapus.
Biology and distribution. Many specimens were sifted from leaf mould and rotten wood. It is only known from two localities in in southern Madagascar characterized by dry vegetation types.
Etymology. The Latin adjective ater means black, referring to the colour of the species.
Diagnosis. Frons impressed below upper level of eyes. Pronotum with about 15-20 coarse asperities, 4-5 of these near the anterior margin. Elytra and pronotum black, legs yellow; spatulate setae present on odd-numbered interstriae only, each separated by 2-3 times the length of a seta.
Description, female. Length 0.7-0.8 mm long, 2.0-2.1 × as long as wide. Colour black, legs yellow. Frons rounded from vertex to just below upper level of eyes, then impressed from a faint transverse rim to epistoma; impressed area glabrous, smooth and impunctate, striate and rugose elsewhere, with fine, short scattered setae. Antennal scapus much shorter than funiculus, rounded, with a scant tuft of long setae on its dorsal side; funiculus 6-segmented, pediculus about as long as segment 2 and 3 combined; club flat, slightly longer than wide, suture 1 transverse, suture 2 recurved, appears ringlike in dorso-lateral view. Eyes separated by 2.4-2.5 × their width. Pronotum strongly dome-shaped, on anterior twothirds with large, rounded asperities, only 4 or 5 of these reaching near anterior margin; posterior third lightly rugose. Vestiture consisting of mixed spatulate and fine setae, longest setae near pronotal summit. Elytra smooth, shiny, stria not impressed, punctures shallow, separated by 2-3 × their diameter, interstrial punctures absent. Vestiture consisting of rows of erect, lightly curved, spatulate setae on odd-numbered interstriae, and scattered, fine and very short strial setae. Ventral vestiture of fine, rather long setae. Legs. Protibiae with lateral edges subparallel, inner mucro strong, curved posteriorly; two tiny socketed denticle transversely set apically on posterior side close to tarsal insertion. Metatibiae broader apically, with a distinct, fine, long inner mucro, and one apical and two laterally placed thin denticles.
Male near identical to female, except fewer and shorter seta on scapus.
Biology and distribution. Only known from the type locality in dry thorny shrub vegetation in south-western Madagascar where it was sifted from leaf mould and rotten wood. Habitat occupation is similar to the close relative M. bicolor, both were collected in large numbers on two occasions, and localities were less than 300 km apart.
Remarks. Pronotum and elytra are very much as in Lanurgus, but female Traglostus differs particularly by the two peculiar long spines on the mandibles directed dorsally. Until DNA data become available it seems prudent to treat this taxon as a valid genus. Four previous members of the genus are now in Lanurgus, two of which are synonymized under other species in that genus (Beaver 2011).
Remarks. Traglostus pubescens share all typical features with Lanurgus and is transferred to that genus. Traglostus longipilus is more atypical by the transverse sutures on the antennal club, very long vestiture, and two hornlike projections on the epistoma. However, it is more similar to other Lanurgus species than the type species of T. exornatus.
The 'flagellum' clade Three genera are defined by an enormously expanded and prolonged flagellum in the male aedeagus, where the tegmen and spiculum gastrale are missing. The last ventrite is slightly to strongly compressed or truncated. Most species are found in Madagascar, the remaining in south to south-eastern parts of Africa.  Schedl, 1958. (Figs. 75, 78, 81) Etymology. From Greek Diplo-, meaning double, and thrix, meaning hair, referring to the split setae on ventral sclerites.
Diagnosis. Antennal scapus inflated, in females rarely with a dorsal spine, funiculus 6-segmented, antennal club with two procurved sutures. All interstriae with a single row of setae (except catenatus group), strial setae very fine, in rows. Protibiae with 3-4 lateral and apical denticles: metatibiae rounded near apex. Ventrites 1 and 2 and part of the metaventrite with bifid setae. Male genitalia with long apophyses, exceedingly long and enlarged flagellum, tegmen and spiculum gastrale missing.

(Lanurgus)
Diplotrichus widdringtoniae (Schedl, 1962) comb. nov. (Lanurgus) =Glostatus perplexus Schedl, 1982(syn. Beaver 2011 Distribution. South Africa, Madagascar. Remarks. The bifid condition of ventral setae is unique among Afrotropical micracidines and is the most reliable character matching the molecular data. Some species related to D. ignotus has an extended flange on the elytral apex, but they all have bifid setae and are therefore not Pseudolanurgus. The South African clade of species consist of the type species and one more known species, and several very similar undescribed species, suggesting a recent recolonization of the mainland (see Fig. 46).
Remarks. This genus can be distinguished from the closely related Pseudomicracis Eggers, 1920 by the less procurved and slightly bisinuate sutures on the antennal club and by the slightly extended, but not mucronate, apex of the elytra. It is further distinguished from Diplotrichus by the simple hair-like setae on ventrites, and from Lanurgus by the very different male genitalia. The shared distribution between Madagascar and Tanzania is known for several scolytine genera where more recent colonisations of the mainland have taken place (e.g. Johnson et al. 2020;Jordal 2013).
Diagnosis. Antennal scapus short, inflated, in females with a long tuft of golden setae; funiculus 6-segmented, club usually with two strongly procurved sutures, suture along first segment reaching basal and not lateral margin, corneous part usually longer than broad. Apex of elytra strongly mucronate; ventrite 5 truncated or impressed. Setae on ventral sclerites never split.
Remarks. This genus is distinguished from all Afrotropical genera, except Phloeocurus, by the strongly mucronate elytral apex, and from the latter genus and the mucronate American genera by the round and twisted lateral margin of the protibiae which has only few apical teeth. Some undescribed species of Diplotrichus have an extended (but not mucronate) elytral apex, and a narrow and elongated corneous first segment on the antennal club, but these species have split setae on the ventral sclerites as in all other Diplotrichus.
The type specimens of the type species P. elsae were lost during the world war attack on Hamburg and no other specimens are known from collections. The description states that the apex of the elytra is prolonged and mucronate. The antennal club of the type species has possibly only weakly procurved sutures and may reach the lateral instead of basal margin. The type of P. camerunus is also lost. The description of P. camerunus is in Latin and the diagnostic characters are ambiguous, except body length is 3.5 mm. it will likely remain as a ghost taxon. With the possibility, albeit less likely, that African Pseudomicracis might be Pseudolanurgus, we seek a solution requiring the fewest possible taxonomic changes and thereby nomenclatural stability.
Both molecular and morphological data supported the synonymy of Saurotocis (Fig. 45). The two Saurotocis species transferred here, and Pseudomicracis pennatus, are generally similar in that males have rather deviating shapes of the elytra with excessive tubercles and confused, broad interstrial and strial setae (Fig. 85). The females of these three species (Fig. 84) are more similar to other species of Pseudomicracis in this respect.
Phloeocurus Wood, 1984 Type species. Hylocurus africanus Schedl, 1957. Monotypic. (Figs. 92-95) Diagnosis. Antennal scapus short, inflated, with a dense tuft of setae; funiculus 6-segmented, club with two weakly procurved sutures at middle and near apex. Frons convex in both sexes. Body elongated, elytral apex mucronate, surface of elytra asperate with short spines or sharp granules on declivity, interstrial setae scale-like, in rows, intermixed with confused smaller setae. Protibiae broad, squared and flattened, with transverse row of denticles at apex; inner mucro much enlarged, curved posteriorly. Remarks. This monotypic genus is unique among the Afrotropical members of Micracidini by having a broadly flattened, squared protibiae with apical, transversely set denticles, and a mucronate elytral apex similar to Micracis. Molecular data was not available but based on the strict Neotropical-Afrotropical split between genera, irrespective of morphological similarities, it is hypothesised that the similarity to Neotropical micracidines is not reflecting a close evolutionary relationship.

Key to the Afrotropical genera of Micracidini
Includes species which have protibiae with one or more apical denticles placed on the posterior face of the protibiae, rough asperities on most of the anterior half of the pronotum, entire eyes, and a small to large tuft of setae on the female scapus. known to have a rounded club without sutures (Johnson et al. 2016). It seems problematic to place all Karlsenius species in Hypothenemus, particularly because K. nigrinum is dimorphic in the frons, which indicate normal outbreeding. As Afrocosmoderes and Hypothenemus are sister genera (Johnson et al. 2020;Pistone et al. 2018), it seems equally possible that Karlsenius is the closest relative to one or the other. Karlsenius nitidum has no antennae available on the holotype but otherwise exhibits the general body shape of Hypothenemus and Karlsenius. Because this species is near identical to K. klainedoxae, it is tentatively placed in the same genus. Even though this paper relates to species previously placed in Micracidini, it is worth noting that Eidophelus ghanaensis (Schedl, 1977) (see Johnson et al. 2020) is fairly similar to the type species of Karlsenius and therefore transferred to this genus.
Note added in press: a 28S partial sequence place Karlsenius klainedoxae in Trypophloeini, but clearly separate from both Hypothenemus and Afrocosmoderes.

Discussion
Previous phylogenetic studies have indicated a very early origin of Micracidini which is likely the oldest tribe within the subfamily, dating back to the Cretaceous (Jordal and Cognato 2012;Jordal et al. 2008;Pistone et al. 2018). The time estimates provided here may be slightly exaggerated, as some studies have indicated a younger age for the subfamily, around 90-70 Ma (McKenna et al. 2009;Shin et al. 2017). However, irrespective of the root age, Micracidini will not be as young as tertiary in age in any of the proposed scenarios because the tribe is almost as old as the subfamily. Earlier mistakes in the classification can therefore be explained in part by slow morphological evolution, confounded by an ancient origin and subsequent radiations within the tribe. One may note that genera such as Afromicracis, Lanurgus and Diplotrichus could need further splitting into multiple genera, but no consistent morphological differences have yet been found to justify erection of further genera.
Many new morphological characters were discovered and were helpful in diagnosing genera and groups thereof. It is quite unusual that the shape of setae is diagnostic for a genus, as documented by the split setae on ventrites in the new genus Diplotrichus. This feature is nevertheless the most consistent and is the only reliable external character distinguishing Diplotrichus from Lanurgus and Pseudolanurgus. One may also note that parts of the antennae are very informative. Not necessarily the presence or absence of an elongated horn on the scapus, but rather the length of the scapus compared to the funiculus. Furthermore, the number of funicular segments accurately diagnose clades in Micracidini; we know this feature may vary within some genera (Jordal 1998;Jordal et al. 2004), or even within a species (Johnson et al. 2020). Similarly, the sutures in the antennal club also reflect evolution quite accurately, while past taxonomic work on this tribe never took this information into account.
Despite the recognition of many new diagnostic characters, Afrotropical micracidines are still a challenge to identify correctly. In view of their considerable old age, the limited differentiation of external morphological characters is puzzling, although by no means exceptional among beetles, or other animals. The limited variation in external characters has therefore motivated investigation of internal morphological characters, which has previously proven a rich source of taxonomically informative characters in Scolytinae (Johnson et al. 2020;Jordal 2009;Jordal and Kaidel 2017). Male genitalia differ substantially between externally similar species in the genera Lanurgus and Diplotrichus and reflect more accurately than external characters the ancient split between these two lineages. While aedeagal apophyses are of normal length in Lanurgus and Microlanurgus, and the tegmen and spiculum gastrale are present, the apophyses are extremely elongated and the tegmen and spiculum gastrale absent or not recognizable in the 'flagellum' group where Diplotrichus belong. Another internal structure, the proventriculus, is easily characterized for the tribe, but Lanurgus differ from most other genera by the narrow suture and enforced lateral to apical teeth in the anterior plate. The two genera Afromicracis and Dendrochilus have multiple crop spines attached on a tubercle base as another example of easily recognized shared characters. They also have many species with a coiled, thin flagellum in the male genitalia, and together with Neomicracis have only a single seta on the hindwing precosta.
All these examples demonstrate the great utility of internal anatomical structures in assessing relationships and placing new taxa in the correct genus, or group of genera. In the molecular era, the application of morphological character is therefore by no means in vain. Rather the contrary, taxonomically informative characters are immediately apparent once a solid phylogeny is established based on broadly sampled molecular and morphological data. This is of great value when molecular data are not possible to obtain.

The Malagasy-African faunal interchange
Madagascar and the Zambesian region of the African mainland stand out as the historically most important drivers of Afrotropical diversity in Micracidini. The Zambesian region is the most likely ancestral area for the tribe, given the endemic distribution of several basal taxa. Subsequent radiations, for instance of the largely Zambesian dominated Afromicracis clade into the Congolian and Southern African regions, further support this view. These are the most widely distributed micracidine genera, but nevertheless restricted to the African continent.
Two particularly important range expansions out of the Zambesian region resulted in markedly increased diversification. The first significant event involved the colonization of