Taxonomic notes on the afrotropical genera Hapalogenius Hagedorn, Hylesinopsis Eggers, and Rhopalopselion Hagedorn (Coleoptera, Curculionidae, Scolytinae)

Abstract Taxonomic confusion among the afrotropical scolytine genera Hapalogenius Hagedorn, Hylesinopsis Eggers and Rhopalopselion Hagedorn, and their synonyms is discussed with especial reference to the catalogues of Wood and Bright (1992), and Alonso-Zarazaga and Lyal (2009). A key is given to separate the three genera recognised, and the species considered to be included in each genus are listed. Hylesinopsis is resurrected from synonymy with Hapalogenius, and shown not to be closely related to it. Chilodendron Schedl is considered to be a synonym of Hylesinopsis and not of Xylechinus Chapuis. The following new synonymy is proposed at specific level: Hapalogenius africanus (Eggers) (= Hapalogenius lesnei Eggers, = Metahylesinus brincki Schedl); Hapalogenius fuscipennis (Chapuis) (= Hapalogenius bimaculatus Eggers); Hapalogenius oblongus (Eggers) (= Metahylesinus striatus Schedl); Hylesinopsis fasciata (Hagedorn) (= Kissophagus punctatus Eggers); Phrixosoma niger Eggers (= Hapalogenius niger Schedl). The following species are returned to Hylesinopsis from Hapalogenius to which they were transferred by Alonso-Zarazaga and Lyal (2009): Hylesinopsis alluaudi (Lepesme), Hylesinopsis angolensis (Schedl), Hylesinopsis arabiae (Schedl), Hylesinopsis atra (Nunberg), Hylesinopsis confusa (Eggers), Hylesinopsis decellei (Nunberg), Hylesinopsis dubia Eggers, Hylesinopsis emarginata (Nunberg), Hylesinopsis fasciata (Hagedorn), Hylesinopsis ficus (Schedl), Hylesinopsis granulata (Lepesme), Hylesinopsis hirsuta (Schedl), Hylesinopsis joveri (Schedl), Hylesinopsis pauliani (Lepesme), Hylesinopsis punctata (Eggers), Hylesinopsis saudiarabiae (Schedl). The following new combination is given: Hylesinopsis leprosula (Browne) from Cryphalus Erichson. New distributional records are given for some species.


Introduction
Th ere has been considerable confusion in the literature about the relationships and limits of the scolytine genera Hapalogenius Hagedorn, Hylesinopsis Eggers, Rhopalopselion Hagedorn, and some other scolytine nominal genera from the Afrotropical region.In this paper, I attempt to resolve some of this confusion, and give some resultant taxonomic changes.Th e conclusions are based on the study of type material and other specimens from the following institutions: Deutsches Entomologisches Institut, Müncheberg (DEI), Hungarian Natural History Museum, Budapest (NHMB), Musée Royale de l'Afrique Centrale, Tervuren (MRAC), Museum and Institute of Zoology, Polish Academy of Sciences, Warsaw (MIZW), Museum fur Naturkünde der Humboldt Universität, Berlin (MNB), National Collection of Insects, Pretoria (NCIP), Natural History Museum, London (NHML), Naturhistorisches Museum, Wien (NMW), Transvaal Museum, Pretoria (TMP), Zoological Museum of Lund University (ZMLU), supplemented by specimens sent for identifi cation by B. Jordal (University of Bergen, Norway), and in my own collection (RAB).
Recently, Alonso-Zarazaga and Lyal (2009) recognised that Wood and Bright (1992) had placed the type species of Hapalogenius in synonymy with Hylesinopsis fuscipennis (Chapuis), and consequently reinstated Hapalogenius as the valid name for the genus Hylesinopsis.Th is resulted in a large number (38) of recombinations of species transferred from Hapalogenius to Hylesinopsis.Alonso-Zarazaga and Lyal (2009) retain Rhopalopselion as a valid genus, with the same twenty-four species included by Wood and Bright (1992).Wood (1986) has suggested that Hylesinopsis sensu Wood and Rhopalopselion sensu Wood are rapidly evolving genera, and that they could either be amalgamated into a single, large genus, or split up into a number of small genera.I believe that the most satisfactory solution at present is to distinguish three genera, based on morphological and biological criteria: Rhopalopselion, Hapalogenius and Hylesinopsis.Th e fi rst two of these genera are closely related, the third, Hylesinopsis, is quite distinct from them.Figures 1-3 illustrate a representative species of each genus.Th e three genera may be distinguished using the following key, which also serves to diagnose the genera: 1.
Antennal  resembling H. bimaculatus, and specimens in which the setae are almost entirely dark.
In the absence of any other distinguishing characters, I conclude that H. bimaculatus is a synonym of H. fuscipennis.Th e shape of the median row of setae on the elytral interstriae varies from almost circular to somewhat elongate and truncate in diff erent individuals.Th e species is known only from South Africa and Mozambique.Wood and Bright (1992) give Zimbabwe, but this appears to be in error for Mozambique.Th e only hosts recorded are Millettia grandis ('Umzimbiti' of Hagedorn 1912) (Leguminosae), and an unidentifi ed tree 'sandalo'.Hagedorn (1912) briefl y described and illustrated the gallery system under bark.

Hapalogenius oblongus (Eggers)
Pseudophloeotribus I have examined the holotype of Phrixosoma nigra (NHML), and the two syntypes of Hapalogenius niger and a further specimen standing under this name in the Schedl collection (NMW).Both species were described from Uganda, from the same host species (Harungana madagascariensis (Clusiaceae)), and are clearly synonymous.Schedl (1952) appears not to have noticed that the eyes of his species are bipartite, and that the antennal club is asymmetrical with a partly septate fi rst segment -characteristics of Phrixosoma and not of Hapalogenius.Schedl (1963a) briefl y describes the biology of the species, and illustrates the gallery system (as Bothryperus niger).
Hylesinopsis Eggers, stat.res.Th ese genera share the type of eye, antenna and protibia given in the key above.Trypographus and Chilodendron were synonymised with Hylesinopsis by Wood (1983).Wood and Bright (1992: 92) give Chilodendron as a synonym of Hylesinopsis, but its type species, C. planicolle, is also listed on page 118 under the genus Xylechinus Chapuis, with a note that the transfer of the species was to be published by Wood in a paper in press in 1992.However, this paper was apparently never published.Alonso-Zarazaga and Lyal (2009), apparently basing their decision on the notes in Wood and Bright (1992), synonymise Chilodendron with Xylechinus.I have examined a syntype of C. planicolle (NMW), and fi nd that the synonymy given by Alonso-Zarazaga and Lyal ( 2009) appears to be excluded by the 6-segmented funicle (always 5-segmented in Xylechinus), entire eye (always emarginate in Xylechinus), fore tibia without socketed teeth, and plumose metepisternal setae (scalelike in Xylechinus), even though the pronotum lacks asperities (as in some Xylechinus) (Wood 1982(Wood , 1986)).Until further detailed studies are made of the species here included in Hylesinopsis, I prefer to leave Chilodendron, and its single included species, as a synonym of that genus.Th e lack of close relationship of this genus to Hapalogenius, based on morphology, seems to be corroborated by some analyses based on molecular data.In the phylogenetic tree of Farrell et al. (2001: Fig. 6), the single species of Hylesinopsis studied (H.dubia) is widely separated from the two species of Hapalogenius included (H.oblonga, H. seriata) (both labelled as Hylesinopsis sp. in Farrell et al. 2001).In the phylogenetic tree of McKenna et al. (2009: Fig.2), the same species of Hylesinopsis (H.dubia) (Mc Kenna, pers. comm. 2009) is widely separated from the two genera (Alniphagus, Hylesinus) currently included in the tribe Hylesinini.In both cases, Hylesinopsis seems to be more closely related to genera included in the subfamily Scolytinae sensu Wood by Wood (1986) and Wood and Bright (1992).One phylogenetic tree (Jordal et al. 2008, Fig. 4) suggests a closer relationship between Hylesinopsis dubia, Hapalogenius seriata, and Hylesinus varius (F.), but in other analyses the relationship between these species is unresolved (Jordal et al. 2008).Th e tribal classifi cation of the Scolytinae sensu Alonso-Zarazaga and Lyal needs revision (e.g.Jordal et al. 2008, Alonso-Zarazaga andLyal 2009), and no attempt to place Hylesinopsis in an existing tribe is made here.
Th e species are normally associated with trees of the family Moraceae (Ficus, Morus, Bosqueia, Treculia).Th ere are only three records from other families, one each from Anacardiaceae, Meliaceae and Rosaceae.Th is narrow host range contrasts with the wide host range of Hapalogenius and Rhopalopselion.
On the basis of the limited distributional data available, nearly 50% (8 ex 17) of the species appear to be confi ned to montane habitats above 1500m.Th is includes the following species: alluaudi, confusa, emarginata, fasciata, granulata, pauliani, punctata, saudiarabiae.Eight species appear to be more lowland species: angolensis, arabiae, atra, decellei, dubia, fi cus, joveri, leprosula.H. planicolle was described from Mt. d'Ambre in Madagascar, but no altitude is given.I have compared two specimens of K. punctatus (NMW), which had been compared with the damaged holotype by Eggers and Schedl respectively, with a series of specimens of H. fasciata in my own collection from Tanzania and Nigeria.Th e latter had earlier been compared to a syntype of that species, and other specimens from East Africa in NHML.Eggers (1932) distinguished the two species by the more elongate shape, stronger shine, more distinct puncturation, and the presence of granules on the basal part of the elytra.Comparisons suggest that K. punctatus lies at one end of the range of variation found in H. fasciata.Th e small diff erences noted by Eggers (1932) are insuffi cient to separate K. punctatus as a separate species, and the latter is, therefore, placed in synonymy.Wood and Bright (1992) cite a holotype for H. fasciata.However, Hagedorn described the species from "compluria specimina", indicating that he had a series of syntypes before him.
I have examined the holotype (MRAC), and twenty-one paratypes (MRAC, NHML).It is not clear why Browne (1980) assigned this species to the genus Cryphalus Erichson.Such an assignment within the tribe Cryphalini is ruled out by the six-segmented funicle, the elongate eyes, the lack of a visible scutellum, the raised and crenulate basal margin of the elytra, and other characters.Th e species is here removed from Cryphalus and transferred to Hylesinopsis.

Rhopalopselion Hagedorn
Fig. 3 When the species described in Hapalogenius are omitted, the remaining eleven species included in Rhopalopselion in Wood and Bright (1992) form a cluster of closely related species distinguished by the quadrate pronotum with strong asperities at the antero-lateral corners, and the large quadrate scutellum.Th e apical visible sternite has a triangular, median projection.Th e beetles are strongly built, black in colour, and 2.5-4.5 mm long.All those with known habits are xylophagous (Schedl 1960, Browne 1963).Like Hapalogenius, the genus belongs in the tribe Hylesinini sensu Wood (1986a) In addition to the type species, Rhopalopselion bituberculatum Hagedorn*, I consider the following species to belong in the genus: R. atrum Eggers, R. confusum Eggers, R. conjungens Schedl*, R. dentatum Nunberg*, R. grande Schedl, R. immune Eggers*, R. intermedium Schedl, R. nitidum Schedl, R. orientale Schedl*, R. thompsoni Schedl*.(* -type(s) examined).Th e remaining species listed under the genus by Wood & Bright (1992) belong in the genus Hapalogenius (see above) in which almost all were originally described.It may be noted here that the holotype of R. bituberculatum is in DEI and not MNB as stated by Wood and Bright (1992).

New records of Hapalogenius and Hylesinopsis
Th e following new records extend the known geographical distribution of the species.