A contribution to the knowledge of the subfamily Panagaeinae Hope, 1838 from Africa. Part 2. Revision of the Craspedophorus leprieuri and C. regalis groups (Coleoptera: Carabidae) Author Häckel, Martin text Zootaxa 2017 4236 2 201 243 journal article 36468 10.11646/zootaxa.4236.2.1 ec0dbb00-2a4a-41c9-9e60-185f972be9bd 1175-5326 321720 BC5E331F-045C-47FF-BA0E-042C69DE3F80 Notes to C. reflexus group Häckel 2016 In the last work devoted to the C. reflexus group ( Häckel 2016: 515) I accepted Basilewsky‘s (1987: 200) synonymization of C. uelensis Burgeon, 1930 with Craspedophorus savagei ( Hope, 1842 ) . Although I could not examine Burgeon’s type the description ( Burgeon 1930b: 159 ) clearly indicated its assignment to the C. reflexus group. In the same work I also synonymized C. raddoni Hope, 1842 and C. savagei Hope, 1842 , on the basis of sexual dimorphism, with C. reflexus reflexus . For C. raddoni I clarified Basilewsky‘s (1987: 200) synonymization, whereas the synonmy of C. savagei was new. After accepting Basilewsky’s conclusion, I provisionally assigned C. uelensis Burgeon to C. reflexus reflexus , although the distribution of Burgeon’s species differs from that of C. reflexus reflexus (and all other synonymized taxa). Now , after examination of the holotype of C. uelensis ( Plate 9, Fig. 70) and of a number of specimens of the C. reflexus group in the Basilewsky collection at MRAC , I must correct that provisional assignment. In the description of C. uelensis Burgeon (1930b: 159) stated: “ It resembles eximius Laf. [= C. reflexus ], and decorsei All. [= C. b. bouvieri ( Rousseau, 1905 ) ], from which is distinguishable by distinctly lesser pronotum and much less distinctly punctate elytra...”, [translated from French ]. In the differential diagnosis Burgeon (1930b: 160) stated: “ C. reflexus is larger and longer… [than C. uelensis ]. In C. bonnyi Bates, 1890 metepisterna are almost squared, pronotal margins sinuate at midlength, hind angles are almost rectangular, and elytra are smooth. In C. b. bouvieri the antennae are dilated and the pronotal base is weakly extended toward peduncle, as in C. regalis ( Gory, 1833 ) and C. decorsei [= C. b. bouvieri ( Rousseau, 1905 ) ” [translated from French ]. From these characters, it seems that C. uelensis differs from C. reflexus not only in the geographic distribution but also in body size. The study of additional specimens at MRAC indicates that species of the C. reflexus complex are exceedingly difficult to distinguish (in this context “ C. reflexus species complex“ is only intended to mean the very similar taxa C. arnosti , C. crampeli , C. impictus , C. reflexus , C. ruvumanus , and C. uelensis ). Some exoskeletal characters (e.g. overall size, size and shape of the pronotum, elytral color pattern) are too variable and can be only applied to certain geographically limited populations; in other species of the complex these characters tend to be constant and have been used in differential diagnoses and keys. For instance in populations of the western savanna near the shore of the Guinea Gulf (roughly in Guinea and Nigeria ) the prevailing type of morphology of both sexes is the “ reflexus - holotype ”, i.e., specimens with a broad pronotum, as broader as the elytra, which show an elytral pattern with four weakly reduced maculae ( Häckel 2016, plate 1, figs. 1–5, 7 and Plate 9: Figs 62, 63 of this paper). Basilewsky arranged his collection at MRAC so that specimens of this complex collected in the region from Guinea to northeastern DR Congo were labeled “ C. reflexus (Fabr.) ”, as far as they had a large pronotum, as wide as the elytra. Other specimens from this territory, often from the same localities and showing a smaller pronotum narrower than the elytra were labeled “ C. savagei (Hope) ”. This arrangement contradicts his own work ( Basilewsky 1987: 200 ), because he synonymized C. raddoni ( Hope, 1842 ) whose holotype male has a large pronotum with C. savagei ( Hope, 1842 ) , whose holotype (female) has a markedly smaller pronotum (Häckel 2016, plate 1, figs. 5, 6). In the same work he also synonymized C. eximius (LaFerté- Sénectere, 1851 ) with C. reflexus , whose types (both sexes) show a broad pronotum (viz Häckel 2016, plate 1, figs. 1, 2). As stated in my first contribution, where C. savagei ( Hope, 1842 ) was synonymized with C. reflexus , in the C. reflexus complex size and width of the pronotum are not reliable characters to identify specimens (Häckel 2016: 524), nor can they distinguish between sexes. After examination of the holotype of C. uelensis (Plate 9, Fig. 70), the hereby discussed holotype of C. crampeli ( Alluaud , 1915, ( Fig. 61 , Plate 9, Fig. 65 left, centre) and of other specimens of this group in the MRAC collection, I propose the following solution: Variable populations of the C. reflexus group inhabiting the northwest coast of Gulf of Guinea from Guinea to Nigeria which apart from similar aedeagus ( Plate 11, Fig. 76) share also some exoskeletal characters, should be left in the nominotypical C. reflexus reflexus ( Fabricius , 1781 see also map in Fig. 81 : 3–11). These populations share fine punctation of striae with punctures not reaching the middle of intervals, which are micropunctate but glossy. The prevailing character is elytral pattern with two maculae, humeral and apical, which are antero-posteriorly reduced and form short transverse fasciae ( Häckel 2016, plate 1, figs 1–7 and here Plate 9, Figs 62, 63). In some specimens the apical fascia may be absent . To the north and east (“Senegambia“, Burkina Faso ) populations maintain the same elytral sculpture but the maculae are much bigger, not antero-posteriorly reduced and roughly circular or quadrate, as in C. reflexus megamacula Häckel, 2016 (Häckel 2016, plate 1, fig. 8 and here Plate 9, Fig. 64, see also map in Fig. 81 : 1–2). Specimens from Nigeria and adjacent western Cameroon (Bamenda) have characters intermediate between the two subspecies. Populations in western (Bafoussam), northern (Yagoua) and in part also central Cameroon (Mbalmayo) include somewhat smaller individuals with relatively smaller pronotum and highly variable elytral coloration, ranging from only slightly reduced macular fasciae (as in most specimens of C. r. reflexus ) through much reduced or absent apical fascia (as in most specimens of another subspecies), to completely black specimens (Fig: 91: 12–14). I do not consider such variability in a population exceptional, it is well documented for the closely related panagaeine Microcosmodes persicus Häckel and Azadbakhsh (2016: 558, figs 6–9) . A shared character of the above noted Cameroonian populations is coarser sculpture with striae punctures reaching the middle of intervals (Plate 9, Fig. 67 centre), in some specimens to the extent of interrupting and obliterating them (Plate 9, Fig. 67 right and left, Fig. 68). Such coarse sculpture justifies the description of C. reflexus rugatus ssp. nov. In central and eastern Cameroon and Central African Republic there are populations of mostly larger specimens, that show elytral sculpture similar to C. r. reflexus , in which the apical fascia is usually either strongly reduced or absent (it is fully developed in rare specimens) (Plate 9, Figs 65 right, 66). Specimens of these populations showing high variability were described as separate species and should be synonymized, as it happens with C. crampeli ( Alluaud, 1915 ) ( Fig. 61 , Plate 9, Fig. 65 left). Assignment of C. crampeli to the C. reflexus group was already discussed in the paragraph devoted to the C. regalis group. Similarly, the most recently described C. arnosti Häckel, 2016 is conspecific with C. crampeli , and must be synonymized. The holotypes of these two taxa differ only in the presence of the apical macula in C. crampeli and its absence in C. arnosti ( Plate 9, Fig. 65 centre and right). The Central African populations of C. reflexus thus must named as C. reflexus crampeli ( Alluaud, 1915 ) (see also map in Fig. 81 : 15–18). The last subspecies of C. reflexus is that inhabiting the vicinity of the East African Rift , i.e. east part of the Orientale Province in DR Congo ( Upper Uele ) , South Sudan (Lolibai) and southwestern Ethiopia ( Gambela , see map in Fig. 81 : 19–21). In the eastern populations the size of the pronotum varies (the smaller type prevails), elytral sculpture is coarser in the west and gradually becomes finer eastward, most specimens are smaller than most C. reflexus crampeli to the west, and all four elytral maculae are present . Not all of these characters are expressed in all specimens, nevertheless they can be distinguished from the homogeneousappearing populations of the closest C. impictus (Boheman, 1848) , that inhabit the southern vicinity of C. reflexus (see map in Fig. 81 : 22–41) and C. ruvumanus Häckel, 2016 ( Fig. 81 : 42–45). Similarly to the preceding subspecies, some specimens of the eastern populations have been described as separate species. C. bozasi Alluaud, 1930 was based on three specimens from southern Ethiopia but I have not found any of them at MNHN . They were most likely gone already in 1987, when Basilewsky (1987: 200) searched the MNHN collection. The fate of these syntypes is unknown to me, and I therefore have not designated a neotype . However , on the basis of the drawing and description ( Alluaud 1930: 6, fig. 3 ) reproduced in my work (2016: 517), I identified a male recently collected in southern Sudan as C. bozasi ( Häckel 2016: 517, fig. 19, here Plate 9, Fig. 69 left) and two males from Gambela State, western Ethiopia ( Plate 9, Fig. 69 right). When Alluaud (1930: 4) described C. bozasi , he was unaware of Burgeon’s description of C. uelensis published in the same year ( Burgeon 1930b: 159 ). I suspect that when Basilewsky synonymized C. uelensis with C. savagei (without comment), he united two very similar taxa, which however differ in pronotum proportions and convexity and sculpture of the elytra ( Plate 9, Figs 62, 63 and Figs. 67, 70). The finding of a new specimen of C. bozasi (candidate for neotype ) in southern Sudan supports the notion that C. bozasi and C. uelensis are conspecific. Burgeon‘s (1930) work was published on 30 July and that of Alluaud (1930) on 15 September . The priority is thus clear, C. bozasi is a synonym of C. uelensis . The name of the eastern subspecies of C. reflexus thus must be C. reflexus uelensis Burgeon, 1930 . All aedeagi of the studied populations of C. reflexus have the same shape ( Plate 11, Figs. 72–76). While studying Basilewsky’s material at MRAC I had the opportunity to observe similar variability in the size of the pronotum and elytral pattern in C. stanleyi Alluaud, 1930 . This species differs from the previous not only in the elytral pattern but also in the shape of the aedeagus (Plate 11, Fig. 77). The variability of the elytral pattern is limited, which permits to distinguish C. stanleyi from the similarly variable C. reflexus at first sight. Examination of many specimens at MRAC and of other recent collections indicates that the specimen found in the Alluaud- Jeannel collection at MNHN labeled C. cf. stanleyi (Häckel 2016: 517, plate 3, fig. 21, here Plate 10, Fig. 70c) is actually a male of that species. Also the type of C. lebaudyi Alluaud, 1932 , which was not found at MNHN and is known only by a drawing (Häckel 2016: 517, plate 3, fig. 24), apparently belongs to a taxon conspecific with C. stanleyi . All differences in the description and the drawing (Alluaud 1932: 9) fall within the limits of intraspecific variation (Plate 10, Figs 70, 71). C. lebaudyi Alluaud, 1932 is therefore hereby synonymized with C. stanleyi Alluaud, 1930 . A final proposed synonymy is aimed to follow Basilewsky’s intention, that could not be carried out due to his passing in 1993. Basilewsky labeled one specimen collected in Kivu ( DR Congo ), the type locality of C. ethmoides Alluaud, 1930 , as “ C. ethmoides All. = C. impictus (Boh.) “. The study of a series of C. impictus (at MRAC) from Uganda and/or northeastern DR Congo , and the comparison with the holotype of C. ethmoides Alluaud, 1930 plus a labeled Alluaud’s specimen at MNHN collected in Kivu (Häckel 2016: 510, 518: Fig. 27), suggest that C. ethmoides Alluaud, 1930 should be synonymyzed with C. impictus (Boheman, 1848) , as already proposed by Basilewsky.