A New Killifish from North-Eastern Gabon - Aphyosemion abacinum nov. spec. (Atheriniformes, Cyprinodontidae, Rivulinae).


By J.H. Huber*




Among the species belonging to the genus Aphyosemion, there are few species characterized by a cross-bar color pattern. Those that have such a pattern can be classified into three groups :

1) The group of A. walkeri Boulenger, 1911, from Bokitsa Mine in southern Ghana, and the related (or synonymous) A. spurrelli (Boulenger, 1913), from the Tano river in western Ghana, as well as the new phenotype (temporarily considered A. walkeri) from the Ivory Coast, and the species A. GH2/74 recently caught in Ghana.

2) The A. exiguum-bualanum group, known by numerous populations in Eastern Nigeria and in Cameroon.

3) A. sjoestedti Lönnberg 1895, an annual species found in the coastal zones of Nigeria and Western Cameroon.


The small species studied in this paper shows alternatively blue and red cross-bars on the body and fins.


Considering the difficulties of this expedition, this species was collected in only one locality in the Mékambo region, not far from the north-eastern frontier of Gabon. Its name alludes to the characteristic color pattern: abacinum, which means "mosaic", is derived from abacus, which refers to a system of parallel lines.


Aphyosemion abacinum


HOLOTYPE: male, Standard length: 23.6 mm. Total length: 28.4 mm.


TYPE LOCALITY: About 30 km to the north of Mékambo (Gabon) on the cycle-path leading to the Congo (path Ego-Bouomo-Sembé). ColI. J.H. HUBER. August 22. 1976. Collection locality #JH 42.


PARATYPES: males and females, 27 specimens caught at the same locality as the holotype, of which 8 have been studied with holotype.

The nine specimens studied will be deposited in Paris, France (Muséum national d'Histoire naturelle, MNHN), and the rest at the Musée de Zoologie de l'Université et de la Ville de Nancy.


Table I contains proportions and meristic data of the holotype and eight of the paratypes.




Morphological characteristics:

Some of these characteristics are found in three species from the same area.

Listed in the relative order of importance that I see, these characteristics are :

- small size, less than most of the other species of the genus,

- ventral fins are set far back,

- type G frontal scalation pattern,

- frontal neuromasts in open-type canals, trapezoid,

- superimposed dorsal and anal fins; the dorsal sometimes in front of the anal fin,

- great number of rays in the pectoral fins,

- rays of the fins are frequently divided into two, 3/5 of the length,

- on the contrary, no filament prolongation was observed on the fins, although aquarium strains of the three species just cited (with a bar pattern) do have short ones.


Color Pattern:

The color is unique in the group of the species that have the same morphology as A. abacinum.


In preserved specimens:

The types were preserved in a 10% solution of the commercial 40% formaldehyde stock. After a few hours almost all coloration disappeared, but the basic pattern was still largely identifiable. Six to twelve cross-bars could be seen on the brownish body of the males. Three red cross-bars, however, remained on the caudal fin, and to some degree also on the anal and dorsal fins. All the rays were dark. The females had also a brownish body color, but only a few cross-bars were visible. Only the dorsal fin had black cross-bars.


In life:

The coloration of the male is shown in Fig. 1. When frightened, the cross-bars tend to disappear and only a brownish body color remains. The pattern of females is similar to that of the preserved specimen, except that the cross-bars are even less distinct.



A. abacinum is known from only one locality which is on a cycle path from the northern district to the frontier of the Congo, about 30 km from Mékambo (2 and a half hour bike ride, including a stretch by boat on the Djadié river… and the painful mosquito -"fourous"- bites!).

The Mékambo region is furrowed by small tributaries of the Djadié and the Djoura. Many are stagnant, but the creek inhabited by this new species runs fairly rapidly through the forest between the Djadié and the Djoura. Its width ranges from 0.3 m to 5 m, and during the dry season there is also a criss-cross of rivulets. Although the depth of the water was at most 30 cm, during the rainy season (starting around September 15) it must be considerably deeper, since most of the forest becomes flooded.

On August 22, 1976, when these fish were collected, the air temperature was 23°C and the water temperature 20°C. The water, flowing on a bed of leaves was brown. Little light penetrated the vault of trees. Chemical and physical water measurements are given in Table II.

The creek can be divided into four zones:

1) Banks with regular forest plants that fall into the water. Also found are sections where amphibian plants of the genus Anubias grow. This biotope is a refuge for Cyprinidae, Characidae, and young Cichlidae (a species of each was later found in the Djadié).

2) A zone where the water was rapidly flowing. Here Grasseichthys gabonensis Géry and an unidentified species of Procatopodinae were observed.

3) A zone with slow or no current at all, containing Epiplatys aff. sangmelinensis.

4) The edges where the water was calm. There were no plants, but thick (up to 15 cm.) layers of dead leaves. Two species of Aphyosemion and occasionally a species of Mormyrid were found in this area. The two species are A. kunzi Radda 1976 (present in couples, or trios) and our new species, which is very abundant.


It should be noted that five out the nine species that inhabit this heavily populated creek have a more or less regular pattern of cross-bars. This phenomenon may probably be attributed to convergent evolution.




If there are some species (cited in the introduction) that have a pattern of alternating cross-bars, it is also clear that from a phylogenetic viewpoint, A. abacinum is quite different.


The species is morphologically close to three other species in the Ivindo basin : A. georgiae Lambert & Géry 1967, A. cyanostictum Lambert & Géry 1967, and A. fulgens Radda 1976, which I raise here to the full species level.



Table III contains a summary of collection information of this group and other sympatric species of the genus. Since these species can be collected only along certain roads, it is difficult to establish the actual geographic distribution of these species. It is apparent that from year to year the four species of the group replace each other allopatrically, sometimes in the same stream. Lambert and Géry 1967 (op. cit) note that the first two species were sympatric in two places, close to Bélinga. My study extends in an important way their distribution. The group in question is not close to any other and consequently they seem to occupy an isolated position among the Aphyosemion. We need to gain better geographic knowledge of the northern region. Unfortunately, there is a lack of roads.


Considering the preceding data, I propose the following identification key, based on coloration and pattern :

1 - Brick-red body, blue dots

1a) No white bars on the caudal fin. Fins uneven red, blue dots. A. cyanostictum Lambert & Géry.

1b) A double, lower band, white and orange on the caudal ; anal fin orange. A. georgiae Lambert & Géry.

1c) A lower band and an upper band, orange on the caudal ; anal fin orange. A. fulgens Radda (then, as subsp. georgiae fulgens).

2 - Body and fins, uneven brick-red, blue cross-bars, A. abacinum nov. sp.



A new species of the genus Aphyosemion (Atheriniformes, Cyprinodontidae, Rivulinae) from Gabon is described: A. abacinum nov. sp. shows a distinct color pattern of blue cross-bars on a red body and fins. The data on biology and chemistry of the type locality are given and the systematic position of the new species, with an identification key is proposed.




Lambert, (J.) & Géry (J.), 1967: Biologica Gabonica, 3(4): 291-318 (pIs. I-III).

Radda, (A.C.), 1976: BKA Separatum : 1-20.


* Musée de Zoologie, 34 rue Ste-Catherine, 54000 Nancy, France

First published in French: Rev. Fr. Aquariol., 3 (15), November 1976

Translation by Gari R. Miller, published in JAKA, 11 (6), November 1978

Translation revised by the author, 2001


Additional reading of this author's publications ? Select any ONLINE article.