Chromaphyosemion Radda

Rainer Sonnenberg, 2007, Description of three new species of the genus Chromaphyosemion Radda, 1971 (Cyprinodontiformes: Nothobranchiidae) from the coastal plains of Cameroon with a preliminary review of the Chromaphyosemion splendopleure complex., Zootaxa 1591, pp. 1-38 : 1-4

[[ Genus Chromaphyosemion Radda ]]

Introduction

The genus Chromaphyosemion Radda, 1971 ( Cyprinodontiformes : Nothobranchiidae ) is distributed in the coastal plains from southern Togo to northern Gabon. Its center of diversity lies within Cameroon, a pattern similar to that observed in other groups of nothobranchiids ( Epiplatys sexfasciatus group, Aphyosemion calliurum group), lampeyes of the genus Procatopus Boulenger, 1904 ( Cyprinodontiformes , Poeciliidae ), and in cichlids of the genus Benitochromis Lamboj, 2001 . Most of the species of these groups are endemic to Cameroon, which also has the richest cyprinodontiform fauna in Africa with a high proportion of endemics (species numbers and distribution data from Huber 2000 and for the cichlids Lamboj 2004).

Chromaphyosemion is diagnosed as a monophylum by characteristic features which are unique in nothobranchiids including color patterns with two parallel black stripes on the sides, a unique black ventral head pattern and several rows of scales on sides (especially dorsally) with a metallic bluish, greenish, copper or golden coloration ( Radda 1971; Sonnenberg 2000). Besides, species of the genus possess the ability of fast color changes, which for example is also known in cichlids ( Fryer & Iles 1972). This feature is unique within Nothobranchiinae, and is especially pronounced in adult males. It means that specimens can change from pale sides with clearly visible parallel black stripes within seconds into extremely colorful specimens (Figs. 1-2) and back. To some extent this can also include fin coloration and also is seen in females.

Species diagnoses in Chromaphyosemion are often based, as in most aplocheiloid species groups, on male color patterns. The morphometric and meristic characters typically used in taxonomic research of nothobranchiids show often considerable overlap and are rarely diagnostic ( Amiet 1987; Scheel 1966, 1968).

Amiet (1987) assumes that male coloration might be important in mate choice of females in nothobranchiid fishes. For this reason it may be particularly useful for taxonomic purposes because the species diagnosis would be based on the same characters which females use to distinguish between con- and heterospecific males. This is similar to the Specific Mate Recognition System (SMRS) as described by Paterson (1978). The potential applicability is basically supported by mate-choice experiments in several groups of cyprinodontiform fishes (e.g., Houde 1997; Kodric-Brown & Strecker 2001; Strecker 2006; Strecker & Kodric-Brown 1999, 2000; Brosset & Lachaise 1995) and a first study in Chromaphyosemion ( Klemme 2005) . This allows the assumption that the use of color patterns as diagnostic characters in species definitions has a real biological background and is not an artificial classification based on random variability. In addition, pheromones are also assumed to play a role in mate and species recognition in Chromaphyosemion ( Bathie 1972; Ewing & Evans 1973).

The first review of Chromaphyosemion was conducted by Scheel (1966, 1968, 1974, 1990), who studied morphology and karyotypes and carried out some crossing experiments. He disclosed, besides a conservative morphology, a confusing variety of karyotypes and crossing results ( Scheel 1974). Recent studies ( Völker et al. 2005, 2006, 2007a, in press and pers. comm.) have confirmed that karyotype variation between and within species and even within populations is not unusual. Therefore karyotype differences alone are not a sufficient indicator of species differences or boundaries, especially as their impact on reproductive isolation and speciation is not well known (e.g., Rieseberg 2001).

Crossing experiments, usually performed in a no-choice experimental setup ( Scheel 1966, 1968, 1974; Huber 1998 a, 2000), showed different degrees of hybrid infertility. A histological study of the reproductive tissues of several crosses of C. bitaeniatum ( Ahl, 1924) populations by Schröder (1967) indicated that, even within a diagnosable species, incompatibilities including hybrid sterility exist. Similar results were found by Scheel (1974) for C. riggenbachi . These results imply that crossing experiments alone are insufficient to establish taxonomic relationships (for further discussion see Kottelat 1997).

Scheel (1974) used the oldest available name, Aphyosemion bivittatum ( Loennberg, 1895) , to name what appeared to him as a number of well-defined biological species in the sense of Mayr (1963), but which he was unable to diagnose by external 'phenotype' ( Scheel 1974, p.98). Later he changed his position ( Scheel 1990) and used the names Aphyosemion (Chromaphyosemion) bitaeniatum , A. (C.) bivittatum , A. (C.) loennbergii ( Boulenger, 1903) , A. (C.) pappenheimi ( Ahl, 1924) , A. (C.) riggenbachi ( Ahl, 1924) , and A. (C.) volcanum ( Radda & Wildekamp, 1977) . Chromaphyosemion bitaeniatum sensu Scheel (1990) included the currently recognized species C. alpha ( Huber, 1998a) , C. bitaeniatum , C. kouamense ( Legros, 1999) , C. melanogaster Legros et al., 2005 , C. poliaki ( Amiet, 1991) , C. punctulatum Legros et al., 2005 , and C. splendopleure ( Bruening, 1929a) .

Currently Chromaphyosemion consists of 12 or 13 accepted species (depending on the author) and several undescribed forms ( Agnèse et al. 2006; Huber 2000; Legros et al. 2005; Legros & Zentz 2007; Sonnenberg 2000). Most of the undescribed forms are included in a polymorphic species Chromaphyosemion splendopleure , an ill-defined assemblage of probably separate species ( Eberl 1996; Eigelshofen 1998; Legros et al. 2005; Sonnenberg 2000) or as in Agnèse et al. (2006) and Legros and Zentz (2007) in the nominal taxa C. splendopleure and C. volcanum . A recent overview of this genus ( Sonnenberg 2000) distinguished 16 different forms on the basis of color pattern differences, including nine described species and seven undescribed forms. Two of the latter were recently described as valid species ( Legros et al. 2005). Additionally, there are at least two different forms known from mainland Equatorial Guinea, one of them recently described as Chromaphyosemion malumbresi ( Legros & Zentz 2007) , one from Bioko, formerly Fernando Poo ( Huber 2000; Legros & Zentz 2007; Pohlmann 2006a; Roman 1971; Scheel 1974; Thys van den Audenaerde 1967), and recently a probable new species was found in the eastern Niger delta in Nigeria ( Alders 2005).

Huber (1998 a, b) postulated an 'artificial' subdivision of Chromaphyosemion into three to four subgroups on color pattern similarities, based on his observations and data from Legros: the C. bitaeniatum subgroup, including C. splendopleure in the widest sense, C. volcanum and C. alpha (based on the absence of red or dark markings on the anal fin, more obvious dark lateral stripes, lines on the snout and a dark postopercular blotch, less and smaller dots on sides); the C. loennbergii subgroup with C. pappenheimi , C. riggenbachi and C. kouamense (based on anal, caudal and dorsal fins with more or less red streaks or dots, paler dark lateral stripes and in females an anal fin with red dots); and the C. bivittatum subgroup with C. lugens , a 'melanophoric' subgroup in which the males have black or dark red blotches on (mainly upper) sides and unpaired fins instead of smaller dots, the dark lateral stripes are conspicuous in both sexes, and the anal fin only in females without dots or streaks. He also tentatively included C. poliaki in this group, which he separated on the following characters from the previous groups: red pigmentation around scales and many red dots in fins, males with golden-yellow coloration within the reticulated pattern on sides and a conspicuous postopercular blotch.

Recent molecular genetic studies ( Agnèse et al. 2006; Legros et al. 2005; Legros & Zentz 2007) presented a phylogeny of this genus based on mitochondrial DNA data from the majority of the currently known species and undescribed forms and defined the taxa C. splendopleure and C. volcanum through genetic clusters, which are at least partially incongruent with current assumptions on species limits based on male color patterns and opinions on the validity of the taxon C. volcanum (e.g. Eberl 1996; Eigelshofen 1998; Sonnenberg 2000).

Here I present a molecular phylogeny based on a preliminary mtDNA analysis of this genus including all currently described species and most undescribed forms. I used this phylogeny based on molecular data in order to recognize the diagnostic value of (color) characters, especially for closely related species, and review some of the taxonomic conclusions of recent publications ( Agnèse et al. 2006; Legros et al. 2005; Legros & Zentz 2007). The insufficiently diagnosed taxon C. splendopleure could be restricted to a small set of closely related forms. Three forms from Cameroon that belong to different groups are described, based on diagnostic characters of male color patterns.