Episemion krystallinoron, Rainer Sonnenberg, Thomas Blum & Bernhard Y. Misof, 2006

Rainer Sonnenberg, Thomas Blum & Bernhard Y. Misof, 2006, Description of a new Episemion species (Cyprinodontiformes: Nothobranchiidae) from northern Gabon and southeastern Equatorial Guinea., Zootaxa 1361, pp. 1-20 : 6-17

Episemion krystallinoron sp. nov.

(Figs. 1-3)

Episemion callipteron in part Huber 2000: 478 (collection localities); Seegers, 1997: 57, (photograph [strain GEB 94/25]); Wildekamp, 1996: 134 (drawing [Edoum strain]).

Episemion sp., Dadaniak et al., 1995: 362-364, collection localities and syntopic fishes; Huber, 2000: 482, collection locality; Neumann, 2000: 67-68, collection localities, photograph (strain GEB 94/25).

Holotype. ZFMK 39942, male (27.9 mm SL); Gabon, northern part of the Monts de Cristal, a river near village of Nkinèn, crossing the road N5 from Médoneu to Kougouleu (0°58'06.3''N, 10°41'33.5''E), T. Blum, G. Fleck, and R. Sonnenberg, 29 July 2002 (collection locality G 02/157).

Paratypes. ZFMK 39943-39945 (1 male, 2 females), paratopotypes, collected with holotype. Finclips for DNA taken from both females (samples RS537 and RS1236).

MRAC 2006-09-P -1 (1 male) paratopotype, collected with holotype.

IRET (uncat.), (1 male), paratopotype, collected with holotype.

ZFMK 39948 (1 male), paratopotype, kept for several months in an aquarium, T. Blum, F. Bitter and P. Sewer, 24 Aug. 1999, collection locality BBS 99/23 (same as G 02/ 157). Finclip for DNA taken (sample RS279).

ZFMK 39946-39947 (1 male, 1 female), Gabon, northern part of the Monts de Cristal, river near Médouneu, at road N5 (1°00'06.0''N, 10°45'50.9''E) and kept for several months in an aquarium, T. Blum, F. Bitter and P. Sewer, 24 Aug. 1999 (collection locality BBS 99/ 22 [same as G 02/156]).

Additional material. Color pictures of live specimens of both Episemion species, including the holotype of Ep. callipteron (Radda & Pürzl, 1987a: 17; 1987b: 38), were used to compare color patterns (Dadaniak et al., 1995: 101; Neumann, 2000: 68; Seegers, 1997: 57; van der Zee, 2002: 135, 138, 139). In addition, we examined photographs of live Episemion krystallinoron from locality GEMBLJ 03/20 (courtesy of M. Juhl, Denmark). Specimens for the DNA study are listed in Table 2, with voucher numbers in tissue collection of ZFMK.

According to a statement in the description by Radda & Pürzl (1987a), the type series of Episemion callipteron was intended to be deposited in the Natural History Museum (NMW) in Vienna. However, it appears not to be present in that collection (Dr. Mikschi, pers. comm., 16 Jan. 2006) and therefore could not be studied. Measurements for Ep. callipteron were taken from the literature (Radda & Pürzl, 1987a). To our knowledge, no other specimens of either Episemion species are presently deposited in institutional collections.

Diagnosis. Episemion krystallinoron (Figs. 1-3) shares with Ep. callipteron (Fig. 4) the combination of the following characters: a black lateral stripe; a pointed snout; 11-12 dorsal fin rays and 14-16 anal-fin rays; both dorsal and anal fins posterior to mid-body, with first dorsal-fin ray situated above 6th or 7th anal-fin ray; and coloration of the unpaired fins (reddish to reddish-brown, with lighter dots or “flames” [i.e., streaks] between fin rays) in males.

Coloration of the unpaired fins in the two Episemion species is similar to that found in the genus Diapteron , but these genera differ with regard to the other characters listed above.

The new species is distinguished from Ep. callipteron by a unique combination of male color characters. This distinctness is supported by DNA data. The dorsolateral body color of adult Ep. krystallinoron males is metallic blue-green, versus metallic gold in Ep. callipteron . Ep. krystallinoron males have three or four more or less regular horizontal rows of red dots on sides of body, beginning just behind head, versus a single broad reddish-brown line on middle of side of body in Ep. callipteron (which may differ in intensity [depending on “mood” of fish]) varying from dark brown to black. In addition, there are dorsolateral red dots in Ep. callipteron , which form either a more vertical pattern (see picture of holotype in Radda & Pürzl, 1987a: 17) or an irregular pattern (see pictures of males, aquarium strain from type locality in Dadaniak et al., 1995: 101; Neumann, 2000: 68). The unpaired fins in Ep. krystallinoron contain blue to blue-green dots versus golden or whitish dots in Ep. callipteron , sometimes silver blue in anal and caudal fins. In posterior part of caudal fin the blue to blue-green dots are fused to form streaks or larger areas (see Fig. 1) in Ep. krystallinoron versus fewer and shorter streaks in Ep. callipteron .

The morphological data (see Tables 3 and 4) indicate a longer head (HL: 29.0 versus 24.6 [as percentages of standard length]) and a larger eye (E: 9.0 versus 6.2 [as percentages of standard length]) in Ep. krystallinoron , but as no Ep. callipteron could be measured by us (the voucher specimens for DNA are small and shrunken due to preservation in 96% ethanol) this might be caused by differences in measuring between the authors. These characters will be restudied and reevaluated if new preserved material of Ep. callipteron becomes available.

Description. See Figures 1-3 for overall appearance and Table 3 for morphometric data of the type series.

Episemion krystallinoron shows a strong sexual dimorphism, like most cyprinodontiform fishes. Adult males, on average, are more colorful and larger than females. A slender, laterally slightly compressed and elongate medium sized nothobranchiid species. Dorsal profile nearly straight on head, slightly convex from nape to end of dorsal fin base. Ventral profile gently convex from head to end of anal fin base. Caudal peduncle slightly concave from end of both anal and dorsal-fin bases to base of caudal fin. Snout pointed, more than in other species of Nothobranchiinae; mouth directed upwards; lower jaw longer than upper; posterior end of mouth at same level as center of eye. Dorsal and anal fins in males slightly pointed; caudal fin triangular, slightly trilobed posterior end, no filamentous extensions. Fins in females generally smaller and more rounded. Basal parts of all fins covered with an opaque mucus film. Dorsal and anal fins located posterior to mid-body. First dorsal-fin ray behind origin of anal fin, situated above 6th or 7th anal-fin ray and behind middle of body. Dorsal fin with 11-12 rays; anal fin with 14-16 rays, pectoral fins with 18-20 rays. Pectoral fins extending posteriorly past origin of pelvic fins; pelvic fins nearly reaching anal fin. Upper pectoral-fin rays longer than lower rays.

Scales cycloid; body and head entirely scaled, except on ventral surface of head. Frontal squamation of G-type; scales on mid-longitudinal series 27-29 + 2-3 on caudal fin base. Transverse rows of scales above pelvic fin 8-9; circumpeduncular scale row 11-12.

Open frontal neuromast system with two separate grooves, preopercular neuromast system with 6 pores, the apomorphic condition for the Nothobranchiinae (van der Zee, 2002).

Coloration. (Live specimens). Males (Figs.1-2). Back brownish, sometimes with irregularly distributed red dots or reticulated red pattern. A black lateral stripe situated slightly below middle of body, from lower jaw through eye and extending to base of caudal fin. Lateral stripe narrower on head region than on body and caudal peduncle. In dominant males the black lateral stripe may be hidden by a dark red to reddish-brown irregular pattern, which usually forms three stripes alternating with three to four blue to blue-green stripes, starting with a row of blue scales on the back. Striping pattern more irregular from above anal fin to base of caudal fin, with one red stripe often spliting into two. (The most irregularly colored males are found in the population near Médouneu [BBS 99/22, Fig. 2]). Belly below insertion of pectoral fins whitish to grayish.

All unpaired and pelvic fins concolorous with body (red-brown and blue-green). Dorsal fin red, with irregular blue-green dots, often fusing into streaks on posterior tip. Caudal fin with more red on anterior part and blue-green dots, more blue-green color on posterior part, with dots often fusing into streaks, upper and lower margin red. Anal fin red, with irregular blue-green dots; in some specimens anterior submarginal part of fin completely blue-green, lower margin red. Pelvic fins red at bases, middle part blue-green, often with incomplete red margin. Pectoral fins orange at base, hyaline red between fin rays, distal margins darker.

Females (Fig. 3). Coloration of females similar to stressed males and juvenile specimens; an often prominent black lateral stripe from lower jaw to base of caudal fin. Back brownish, a whitish, smaller stripe between back and lateral stripe, and a smaller whitish to grayish stripe situated below black lateral stripe. Back and sides with irregular red markings, forming an incomplete red reticulate pattern. All fins hyaline; each unpaired fin with a yellow hue; base of pectoral orange.

Color in Ethanol. Males. Back and sides brown, laterally a dark brown stripe extending from lower jaw to base of caudal fin; belly light brown to whitish. Scales on back and sides with a dark brown border, forming a reticulated pattern; in some specimens traces of red pigmentation are left. Unpaired and pelvic fins with dark pigmentation; dark brown margin especially on anal and pelvic fins. Pectoral fins hyaline, more posterior fins with some dark pigmentation.

Females. Similar to males and same as juveniles. Fins hyaline, no dark pigmentation and no dark margin on anal fin.

Etymology. This species is named after the Monts de Cristal, to which it is endemic. Translated to Greek: krystallin for Cristal and oron, plural for mountains, French Monts.

Distribution and habitat. Episemion is only known from a small number of localities in southeastern Equatorial Guinea and northern Gabon (Map 1). According to a personal comment from H. Gonin (in Dadaniak et al., 1995), Ep. krystallinoron was first collected in 1983 at Médouneu, near the airport, in the brook Ottomitan in Gabon, three years before the first collection of Ep. callipteron . Close to this place is the collection locality BBS 99/ 22. Both localities are in the Komo River system. Other localities are along the road N5 near Edoum and Nkinèn, in the Mia River system, a tributary of the Komo. All three DNA samples of Ep. krystallinoron from Equatorial Guinea were collected in small rivulets of tributaries of the Rio Mbé and Rio Ncomo (= Komo in Gabon). The Rio Mbé in Equatorial Guinea is not identical with the Mbè in Gabon; the first (together with the Rio Ncomo) is in the Komo system. The latter belongs to the next river system west of the Komo, with the watershed located approximately west of Edoum.

The distribution of Episemion callipteron is probably wider than indicated from present collections, and records indicate that it is found in at least three different river systems. First, the type locality and other known localities in Equatorial Guinea are from tributaries of the Rio Benito or Rio Uele (which to the west is known as the Rio Uolo on the official map of Equatorial Guinea), which is called Woleu in Gabon (Juhl, pers. comm.; Castello in Huber, 2000; Radda & Pürzl, 1987a). Second, the localities at Sam (Bitter, 2003) and Oveng (38 km NW of Zomoko, LEC 93/13 according to Huber (2000); locality description in Dadaniak et al., 1995) are in the Nkam system, a tributary of the Abanga, which flows southwards to the Ogooue. Third, the locality PEG 98/24 (Eberl, pers. comm.) is located in the Lara system, a tributary of the Okano River, which flows into the Ogooue at N'djolé. As no specimens of Ep. callipteron from collections outside the Rio Uele/Woleu system could be examined by the authors, we are unsure of species identities. The currently known localities indicate a possibly wider distribution of the genus Episemion than previously assumed, and not a small "relictual" distribution (Radda & Pürzl, 1987a; Huber, 1998a).

The type locality of Ep. krystallinoron is a large river (~5-6 m) up to 1 m deep. The river near Médouneu at locality G 02/156 (= BBS 99/22) is also large (~ 4-5 m) and about 80 cm deep. At both localities the water is fast flowing, with sandy bottom and no aquatic vegetation. Episemion specimens were found amongst overhanging terrestrial vegetation, as were individuals of the sympatric nothobranchiids Aphyosemion (Mesoaphyosemion) mimbon Huber, 1977 and (rarely) Aphyosemion sp. aff. herzogi Radda, 1975 (the latter is more common in smaller tributaries). In 2002, Ep. krystallinoron was not found again near Médouneu, probably due to anthropogenic alterations of the habitat.

Remarks. Episemion krystallinoron is the third endemic nothobranchiid from the Monts de Cristal, in Equatorial Guinea and Gabon, together with Aphyosemion (Mesoaphyosemion) mimbon and A. (M.) etsamense Sonnenberg & Blum, 2005 . Species diversity and number of sympatric nothobranchiid species is low within this region compared to the Ivindo basin, with which it shares some species groups (Brosset, 1982, 2003; Huber, 2000; Sonnenberg & Blum, 2005). On the other hand, Monts de Cristal has a sufficiently long-term stable environment that would allow the evolution of endemics. It can be assumed that they evolved in this area in conjunction with the last Pleistocene climatic changes. Stable refugia for these forest dwelling fishes (Kamdem Toham & Teugels, 1999) are present within the Komo and Mbè river systems. This is supported by several analyses, which indicate a forest refuge in the area of the Monts de Cristal and Monte Alen region characterized by a high plant diversity with a large number of endemics (Gioda et al., 1993; Maley, 1987, 1989, fig.1; WWF, 2003).

DNA analysis. Results within Episemion . The alignment contains 788 bp for a total of 8 specimens of Episemion : 5 Ep. krystallinoron from 4 different localities, 3 Ep. callipteron from 2 localities (including the type locality) and 12 specimens from other different nothobranchiid groups (see Table 2). Within Episemion , 114 positions are variable, 110 are parsimony informative. The degree of uncorrected pair-wise sequence divergence between Ep. krystallinoron and Ep. callipteron is 13.32-14.21% and within Ep. krystallinoron 0-0.63% and Ep. callipteron 0.25-0.63%. Sequence variation within presumed species is thus about 20 times smaller than between species. The sequences are translated into 262 amino acid positions, of which 29 are variable and 27 phylogenetically informative between both species.

In addition, a smaller PCR product was amplified only in Ep. callipteron , which seems to be a pseudogene, with only partial similarity to the functional mitochondrial sequence. This might be a potential diagnostic molecular character for Ep. callipteron , but we hesitate to use it as we don't know if it also occurs in Ep. krystallinoron . This could not be amplified due to mutations at the primer binding sites.

The phylogenetic analysis resulted in two distinct groups within Episemion , which can be assigned to the two groups diagnosed by color pattern (Fig. 5). As indicators for species status we use the distinct color pattern with respect to Ep. callipteron , and in addition the monophyletic cluster of mitochondrial haplotypes and the intra- and interspecific sequence differences. This is seen as indirect evidence for different biospecies sensu Mayr.

Phylogenetic analysis. Within the complete alignment, 374 positions are variable and 322 parsimony informative. The average uncorrected sequence divergence is 20.05%, and the maximum observed divergence 28.55% between Ep. krystallinoron and Epiplatys infrafasciatus ( Guenther, 1866) . The phylogenetic analysis shows that Episemion is nested within the Nothobranchiinae. It is here placed together with Diapteron , with the two as sister groups with Kathetys . However, only the relationship to Diapteron has a statistically significant posterior probability value above 95 (Fig. 5). Inclusion of Chromaphyosemion within the closer relationships of Episemion is not supported. The mtDNA data also exclude closer relationships to the genera Epiplatys (Epiplateinae) and Fundulopanchax (Nothobranchiinae).

Discussion

As van der Zee (2002) and Wildekamp (1996) have remarked, Episemion belongs to the complex of species groups that is subsumed by most authors under the generic name Aphyosemion . Van der Zee (2002) presented several morphological characters, including the frontal and praeopercular neuromast system, which show that Episemion does not belong to the Epiplateinae and its placement as a subgenus in Epiplatys is not supported. The pattern of the frontal neuromasts on the head has the same apomorphic state as in the Nothobranchiinae. Inclusion of Episemion in Epiplatys or in the Epiplateinae has been based solely on a superficially similar appearance, as shown after a closer examination of different morphological characters by van der Zee (2002). Within the Nothobranchiinae, a relationship with Fundulopanchax could also be excluded, based on the number of circumcaudal scales (max. 12 vs. 16 or more in Fundulopanchax , except the subgenus Paludopanchax ) (van der Zee & Wildekamp, 1994), which supports the phylogenetic placement suggested by the DNA data.

Interestingly, no Epiplatys is known to occur within the known range of Episemion , whereas in the surrounding areas different Epiplatys species are very abundant and are often the most frequently found killifish in suitable habitats (Huber, 2000; pers. obs.). The similarity in appearance to Epiplatys could be due to parallel adaptation to a similar habitat, e.g. larger streams and relatively fast-flowing water.

We concur with the conclusions of other authors (e.g. Huber, 2000; Seegers, 1997; Wildekamp, 1996; van der Zee, 2002) that Episemion is a valid genus, which is probably related to the genera Diapteron (as genus e.g. in Brosset & Lachaise, 1995; Seegers, 1980, 1997) and Kathetys (currently used as subgenus in Aphyosemion ). This is in accord with preliminary analyses of our molecular data. As the cytochrome b sequences are highly divergent between the different species groups, the results must be viewed with caution, and this topic will be readdressed in a study using a larger molecular data set (Sonnenberg & Misof, in prep.). Episemion appears to be nested within a large complex of different species groups, currently included in the genus Aphyosemion (e.g. Huber, 2000; Seegers, 1997; Wildekamp, 1993).

As there is no universally accepted genus concept in zoology, we use as a rule of thumb that species groups regarded as genera should, a) be monophyletic, b) be morphologically distinct, c) show a larger genetic gap between species groups than within; and d) have at least a slightly divergent ecology (including life history, reproduction, food and / or habitat preferences). Even if the differences in one or all of these points are small, the combination will circumscribe a unique evolutionary group, which differs from all other related groups and should be recognized as a distinct unit.

Episemion and related groups fulfill, so far as we know, all four points. All these groups are: a) monophyletic, b) morphologically distinct (which is also acknowledged by earlier authors when naming them), c) show genetic gaps between each other larger than within these groups (which is reflected in the phylogenetic tree [Fig. 5]); and d) seem to differ in their ecology, based on our own observations. Episemion is, for example, found in faster flowing and deeper rivers than Diapteron (pers. obs.) and possibly Kathetys , but all prefer flowing water (except K. bamilekorum (Radda, 1971b) , which lives in swampy waters). In addition, Diapteron appears to have a different feeding ecology compared to other nothobranchiid genera (see Brosset, 1982 and Brosset & Lachaise, 1995 for a discussion of the ecology of nothobranchiid fishes of the Ivindo Basin).

A close relationship with the genera Diapteron , Episemion , Kathetys and Chromaphyosemion (as genus in Sonnenberg, 2000 and Legros et al., 2005) was postulated by van der Zee (2002) and Huber (1998a), but this is actually not supported by mitochondrial DNA data. Detailed studies of morphological and osteological characters are needed to disclose apomorphies that might support phylogenetic placement within the DNA analysis. Van der Zee (2002) considered the dark lateral stripe of Episemion to be a possible indication for a close relationship to Chromaphyosemion , the species of which have two lateral black stripes, of which the upper one extends from the lower jaw to the base of the caudal fin, or to Kathetys (Huber 1998a), which sometimes has a dark lateral stripe in stressed specimens. On the other hand, a lateral stripe on this part of the body can also be observed in nothobranchiids (e.g. in Aphyosemion hera Huber, 1998 [Nothobranchiinae]; and in several other species of Epiplatys and Scriptaphyosemion [both Epiplateinae]), often only in stressed specimens. This indicates that a dark lateral stripe running from the lower jaw through the eye and extending to the caudal peduncle might be a plesiomorphic character of the Nothobranchiidae.