Nothobranchius {Aphyobranchius} janpapi Wildekamp, 1977

Aphyobranchius: Subgenus of Nothobranchius proposed by R. H. Wildekamp in 1977 (Wildekamp, 1977). The name derives from the Greek roots Aphye- (small fish) and -branchion (gills/lung).

janpapi: Dedication name in honour of Jan Pap {The Netherlands}, who discovered the species in 1975 and sent some specimens to hobbyists in Holland.

First description

R. H. Wildekamp, 1977 - "Nothobranchius lourensi spec. nov. und Nothobranchius janpapi spec. nov., zwei neue Rivulinen aus Ostafrika" - Das Aquarium, Heft 98, August 1977:326-331; 7 tabs. ;6 figs.

Terra Typica

Eastern Tanzania, about 80 km west of Dar Es Salaam. First collected by Jan Pap in 1975 and then by Jan Lourens beginning July 1976, in a pool along the main highway from Morogoro to Dar Es Salaam, between the bridge over the Ruvu river and the village of Kwaraza.

Figure 1: Terra Typica of Nothobranchius janpapi near Kwaraza,
80 km west of
Dar-Es-Salaam, on the Dar-Es-Salaam to Morogoro highway.


Wildekamp, (1977) reports that the Terra Typica of N.lourensi and N.janpapi was a pool of about 10-15 meters across. Another pool, located somewhat further, on the other side of the road, was about 5-7 m in diameter. Both pools were connected over a small bridge. On July 12th ,1976, the depth in both pools measured only 70 cm. The shores were densely covered in grasses and thorny bushes. The water itself did not contain any vegetation. The water temperature on June 11th at 18:15 hours was 27C, with an air temperature of 25C. The pH ranged between 6.5 and 7.0, water hardness was not measured.

In this part of Tanzania, the rainy season lasts from November to May, with a decrease in rainfall in January and February. In 1975 Jan Pap, and in 1976 Jan Lourens, observed that these pools did entirely dry-up. N.lourensi and N.janpapi could only be found in related pools. All other pools only contained, the then as N.guentheri known, N.melanospilus and non-annual fishes (Wildekamp, 1977).

Figure 2: The meteorological profile of Tanzania (Source: FAOCLIM)


It is therefore assumed that these pools are not only filled with rainwater, but must also receive quite some water from the Ruvu River itself, when, during the large rainy season, it floods this large swampy area. According to Wildekamp (1977), the presence of non strictly annual fishes in these pools (such as Clarias mossambicus Peters, 1852; Alestes affinis Günther, 1894 and Barbus usambarae (?) Lönnberg, 1907) confirm this.

Meristic & Morphological Data

Karyotype: n= .

According to Wildekamp's 1977 first description of the species, the body height is included 4.1 to 4.9 times in the total length and 3.3 to 3.9 times in the body length. The head length is comprised 3.7 to 3.9 times in the body length and 1.1 to 1.5 times in the inter-orbital distance. The caudal peduncle is 1.7 to 1.9 times longer than deep.

The general body shape is elongated; the dorsal line only slightly curved. The greatest body depth is found at the level of the ventral fins. The dorsal fin is positioned further to the back, it starts at the level of the fifth to sixth anal fin ray. Both sexes have rounded fins. The dorsal fin in the male is long and pointed towards the back. In males, the flattened dorsal fin reaches far beyond the caudal peduncle, in females it only reaches the caudal peduncle. The mouth opening is small and directed upwards.

Under the eyes, there is only one row of scales. The head squamation approaches the regular G-type. On the head, the post- and supra-orbital neuromast grooves are interrupted, and they do not originate from the F-scales as in the Aphyosemion genus. On the dorsal lip, there is a large number of small papillae. Also on the rays of dorsal and anal fins of the males there are small papillae, in the dorsal, only on the distal part, in the anal along the distal portion.

Table 1: Meristic characteristics of Nothobranchius janpapi

{According to Wildekamp, 1977. - All measures given as % of body length}





Total length




Body height




Head length




Caudal pedicel height




Caudal pedicel length




Eye diameter




Interorbital width




Snout length




Distance Dorsal-snout tip




Distance Anal-snout tip




Distance Ventral-snout tip









N dorsal fin rays




N anal fin rays




N pectoral fin rays




N scales longitudinal line




N scales vertical over anal fin start




N scales over caudal pedicel




N scales pre-dorsal fin





The substantial morphological as well as meristical differences between N.janpapi and the Nothobranchius morpho-type warranted Wildekamp (1977) to propose a subgenus for this first representative of the new subgenus: in adult specimens of N.janpapi the cephalic squamation is regular; there is a strong similarity with the G-type cephalic squamation as observed in Aphyosemion species, but without the F- and H- scales; the scaleless pre-orbital space is small; the supra- and postorbital neuromast systems are interrupted; the four snout neuromasts are embedded into two separated V-shaped canals; in both sexes, there are numerous densely placed papillae on the upper jaw.

In addition, as compared to the Nothobranchius morpho-type, the proposed Aphyobranchius subgenus, is characterised by a lower number of dorsal finrays and the more rearward implantation of the dorsal fin (Wildekamp, 1977).

Following the description of N.willerti in 1992, the third discovered representative of this subgenus after N.luekei (Seegers, 1984), Wildekamp (1992) gave the revised diagnostic characters of the Aphyobranchius subgenus: smaller size; slender body shape; rearward positioned dorsal fin; rectangular anal fin in males; scale ctenii in males only at the lower operculum and lower preoperculum; regular cephalic squamation of the G-type, and habitat selection near the water surface.



Aphyobranchius is a subgenus of Nothobranchius composed of pelagic (= surface-dwelling) forms. The only four presently known representatives of this subgenus include the very rarely held N.luekei Seegers, 1984 (so far only found once, 40-km south of Dar-Es-Salaam, under the bridge over the Mbezi River, on the road to Kibiti, Tanzania), N.willerti Wildekamp, 1992 (found in the Lower Tana River system, coastal plains of Kenya), N.janpapi Wildekamp, 1977 (Tanzania) and N.sp. "Ifakara" TAN 95/4 (Tanzania).

According to Wildekamp (1992), N.willerti does not share the pelagic diagnostic character of the 3 other representatives of the Aphyobranchius subgenus as it prefers to occupy the middle and lower water layers. Also the diagnostic characters of slender body and rearward positioned dorsal fin are less developed in this species. Wildekamp (1992) considers that Nothobranchius {Aphyobranchius} willerti is therefore less specialised than the other representatives of the subgenus and could thus be regarded as the most primitive form of the Aphyobranchius subgenus.

N.janpapi and N.luekei are very closely related. Crossbreeding experiments between both species did not produce any offspring. From these results, Seegers (1987) decided to create a new species and named it N.(Aphyobranchius) luekei, after Lüke (Germany), who first succeeded in reproducing this difficult species.

According to Seegers (1984), N.luekei distinguishes itself however from N.janpapi by its slender and more laterally compressed body; its smaller size than N.janpapi; its even more rearwards implanted dorsal and anal fins; its dorsal which lies clearly beyond the anal origin. In N.luekei the number of rays in the fins is larger than in N.janpapi and there are more scales along the longitudinal line and along the pre-dorsal length. The dorsal has more fin rays, whilst the anal fin has less rays than in N.janpapi. Also there are colour differences between both species (Seegers, 1984).

According to Van Der Zee (1985), the egg chorion pattern in N.luekei, as compared to N.janpapi and the other species from the genus, is very different. As sole representative of the Nothobranchius species-flock, N.luekei presents an hexagonal pattern, which is commonly found in the genus Aphyosemion. Dispersed long hairy structures are visible on the egg shell of N.luekei, whilst in other species the chorion has numerous shorter and densely placed hairy structures (Van der Zee, 1985).


N. janpapi "Kwaraza " TAN 76/?; Tanzania
janpapi "Rufiji River Camp" TAN 95/07; Tanzania
janpapi "Ruhoi River"; Tanzania
janpapi "Ndundu" TAN 95/10; Tanzania
janpapi "Bagamoyo" TAN 95/13; Tanzania
janpapi "Kwaraza" TAN 95/14; Tanzania
janpapi "Kitonga North" TAN 97/9; Tanzania
janpapi "Soga" TAN 97/45; Tanzania
janpapi "Bagamoyo" TAN 97/39; Tanzania
janpapi "Ndundu Ferry" TZ 97/57; Tanzania
janpapi "Soga" TAN 98/14 (same as locality TAN 97/45); Tanzania
janpapi "Kikongo" TAN 98/09; Tanzania


Adult male (33.7 mm total length; 21.1 mm body length); collected by Rob van Haarlem, Jan Lourens, Theo Steinfort and R. H. Wildekamp on 11 and 12 June 1976, in the same pool as N.lourensi, on the Morogoro to Dar Es Salaam main highway, between the bridge over the Ruvu River and the village of Kwaraza.


Adult female (35.5 mm total length; 28.9 mm body length); collecting information as per holotype.


One adult male (40 mm total length; 32.1 mm body length); two adult females (35.2 mm and 35.6 mm total length; 28.2 and 28.6 mm body length); collecting information as per holotype.

All adult types were kept alive for 5 months in tanks. They were preserved on 13-XI-1976 and donated to the Koninklijk Museum voor Midden Afrika, Tervuren, Belgium.


Males can reach up to 50-mm whilst females remain somewhat smaller, 45-mm.



Distribution & Habitat

The Terra Typica lies in eastern Tanzania, about 80-km westwards of Dar Es Salaam, in a pool along the main highway from Dar Es Salaam to Morogoro, between the bridge over the Ruvu river and the village of Kwaraza. The pool was about 10-15 meters across. Another pool, located somewhat further on the other side of the road was about 5-7 m in diameter. Both pools were connected over a small bridge. On July 12th ,1976, the depth in both pools was only 70-cm. The shores were densely covered in grasses and thorny bushes. The water itself did not contain any vegetation. The water temperature on June 11th at 18:15 hours was 27C, with an air temperature of 25C, the pH-value was between 6.5 and 7.0 (Wildekamp, 1977).

Seegers collected 3 different Nothobranchius species living synpatric in the same pool, along the Ruhoi River. They were N.janpapi, N.melanospilus and N.eggersi "Ruhoi" (Seegers, 1985).

In 1995, Wildekamp, Sainthouse and Watters collected N.janpapi from near Bagamoyo, on the road to the ferry across the Ruvu River estuary, along an earthen embankment. The land on either side of the embankment was composed of flat grassy plains. The dense vegetation made it difficult to dip the nets into the water. Although not common, they found both N.melanospilus and N.janpapi and N.foerschi (Sainthouse, 1996)

In May 1997, Wildekamp, Watters and Cooper collected N.janpapi in the locality Kitonga North TAN 97/9. In this locality they found N.janpapi living synpatric with N.lourensi and N.fuscotaeniatus.

Figure 3: Collecting sites of N.janpapi in Tanzania:
1: "
Kwaraza" TAN 76/? And TAN 95/14; 2: "Rufiji River Camp" TAN 95/07; 3: "Ndundu" TAN 95/10 and TZ 97/57; 4: "Bagamoyo" TAN 95/13 and TAN 97/39; 5: "Soga" TAN 97/45 and TAN 98/14; 6: "Kitonga North" TAN 97/9; 7: "Kikongo" TAN 98/09; 8: "Ruhoi River" TZ?/?.


Also in May 1997, Wildekamp, Watters and Cooper collected also three different Nothobranchius species from the Soga TAN 97/45 locality. One was Paranothobranchius ocellatus, the second was N.janpapi and the third N.foerschi (Watters, June 1998, Killietalk).

In 1997, but in July, N.janpapi was found in the same pool where N.fuscotaeniatus was first collected by Seegers (1998). Seegers (1998) reports that coming from a successful collecting trip in southern Tanzania, he decided to spend some time in the Selous Game Reserve. As the smaller ferry at Utete was out of order, he had to drive 60 additional km back over Ndundu, cross the Rufiji River here, and, over a longer side way over Kibiti, reach the opposite side of the river at Utete from where he could progress stream-upwards along the northern bank, in order to enter the Selous Game Reserve.

At Nyamwage, the road to Ndundu makes a sharp bend to the north. This road was at times in a very bad shape and left and right there were numerous rice fields as well as other smaller and larger flooded areas. The road follows large road-ditches dugout at construction time. It became apparent to Seegers that he was approaching the river and the onset of the Rufiji River delta. As it was impossible to fish all the numerous pools, Seegers decided to sample only some of them. This yielded some interesting species amongst which some Nothobranchius (Seegers, 1998).

One final trial was made in a 2 x 10 m rest-pool along the road (location TZ 97/57, 2-km south of the Ndundu Ferry over the Rufiji River). The shores of this pool were covered in high grasses. There were also some patches of open water in the middle of the pool. In the deepest part, the water level reached some 60-cm. In the mid-day sun, Seegers swept unpurposely his net through the water and caught a poorly coloured female Nothobranchius. As it could not concern the widespread N.melanospilus, he got interested again and swept through the water once again. This time he caught N.janpapi, in which he was not interested, as it is almost as widespread as N.melanospilus in the area. At the third trial, he obtained a male of N.fuscotaeniatus species. In the midday sun, the fish lighted bleu-green, with some red-brown transversal bands around the body (Seegers, 1998).

More fish had to be caught in order to find the appropriate female. The next caught fish was a male N.lourensi. Females of the latter species are also colourless, which raised the question as to which species the first caught female belonged to. Therefore more colourless females were caught; but at first, no distinction could be made between them. Finally Seegers caught a female which he had not seen before and which resolved his identification problem. This female N.fuscotaeniatus was also blue-greenish, but not as pronounced as in the male. She also had distinguishable red-brown circular stripes along her sides. The uncoloured females all belonged to N.lourensi. Remarkably, in this cut-off part of the road-ditch, Seegers could not find other fish species, not even the very widespread N.melanospilus (Seegers, 1998).

Based on above collecting sites, the distribution area of N.janpapi includes the middle and coastal stretches of the Ruvu, Ruhoi, Mbezi and Rufiji River systems. This area forms in fact one large swampy area in which fish could easily get distributed from one end of the depression to the other.

The northerly flowing Ruvu River has its source in the Uluguru Mountains, near Morogoro. At first, it flows eastwards before reaching a huge swampy depression situated east of the Uluguru Mountains. Before reaching this swamp, it receives the waters of the Rudete and the Mgeta Rivers, which both also originate in the Uluguru Mountains. It then makes a bend to the north, where it reaches the Indian Ocean, north of Dar-Es-Salaam, in Bagamoyo.

Figure 4: Major river basin systems in Tanzania. (Source FAO-ALCOM)


This swampy depression is also connected to numerous smaller side rivers of the much larger Rufiji/Ruaha River system, providing ample opportunities for ichthyo-faunal exchanges between both river systems.

Similar ancient connections between the Rufiji River system and the neighbouring Ruhoi and Mbezi Rivers could also account for the occurrence of N.janpapi in these systems. The general distribution pattern of N.janpapi in eastern Tanzania is tentatively presented in following figure.

N.luekei, which is very closely related to N.janpapi (as seen above), has thus far only been caught once under the bridge over the Mbezi River (40-km south of Dar-Es-Salaam on the road to Kibiti) by Seegers and Wischmann during the year-transition of 1982 to 1983, although on 3 previous occasions, Seegers had already fished this locality without success (Van der Zee, 1985). With the collection of N.janpapi around Kitonga (locality: "Kitonga North" TAN 97/9, also under the influence of the Mbezi River) by Watters, Wildekamp & Cooper in 1997, it thus appears theoretically possible that both N.janpapi and N.luekei could be found living near each other in and around the same river system. As species definition implies, there must then be either a fierce competition between both related species whenever they occur together or they must avoid competition by having chosen specific ecological niches. A possible future synpatric collection of both species in the Mbezi River would thus appear to be of scientific value.

Figure 5: Collecting sites and presumed distribution area of N.janpapi in Tanzania:
1: "
Kwaraza" TAN 76/?
And TAN 95/14; 2: "Rufiji River Camp" TAN 95/07; 3: "Ndundu" TAN 95/10 and TZ 97/57; 4: "Bagamoyo" TAN 95/13 and TAN 97/39; 5: "Soga" TAN 97/45 and TAN 98/14; 6: "Kitonga North" TAN 97/9; 7: "Kikongo" TAN 98/09; 8: "Ruhoi River" TZ?/?.



Males: Ground colour violet-blue, over the back brownish, over the belly whitish, throat yellow. Scales carry a small reddish-brown border. Dorsal fin is yellow-green, with a few brown spots, at the base changing into grey-blue. At the distal part there is a small blue-whitish bordering band. Caudal fin yellowish-green, on the proximal part, both above and below, two red stripes. On base of anal fin yellow, changing into a deep orange-red. Pectoral fins colourless with a small light blue bordering band. The eye iris is blue.

Females: They are grey brown; belly sides and eye iris are silver; all fins are colourless.

Maintenance & Breeding

N.{Aphyobranchius} janpapi are atypical for the Nothobranchius species-flock as they preferably inhabit the upper water layers of their habitat, rather than near the bottom. Fry as well as (sub)adults can most often be seen searching for food on the surface. In their natural environments, adults preferably occupy the open pelagic zone between submerged shrubs and branches.

The species can already be reproduced in water with following characteristics: pH of 6.8 to 7.0; water hardness between 9-16 and temperature between 23 and 25C (73 and 75F). Such water can easily be obtained by adding some rainwater to ordinary tapwater. Too high maintenance temperatures will substantially reduce their live expectancy. The addition of salt [1 teaspoon per 10 litres of water] is also advised as a preventive measure against velvet.

The maintenance tank would preferably best be densely planted with java-fern or similar aquatic plants in order to provide females and subordinate males' with hiding opportunities. Frequent water changes are also necessary for the wellbeing of the fish.

Males are not very aggressive towards each other. Stronger specimens will however push weaker ones towards the lighter parts of the tank. When the tank is large enough, beaten fins and harmed fishes are seldom observed. In a large tank one can keep several males if each one of them is given about 10-15 litres of water. One has always to avoid to place into a small tank 2 male; in this case the stronger one will definitively always sue the weaker one and even kill him.

For the reproduction, one can use a trio but it is always recommended to use several females per male [3 to 4 ♀ per ♂] and to breed with groups of several specimens at once.

Peculiar to the species is also the fact that the parents do not reach for the bottom to lay their reproductive cells. Spawning can take place everywhere in the tank, preferably in the open- and mid-water. With this species there is thus no need to cover the peatmoss recipient. On the contrary, one would preferably utilise a wide-open recipient, which covers an as large surface of the bottom of the spawning tank as possible. The concern here is not to concentrate the eggs, but rather to catch them in the first place. One will have to do with the excrements in the peatmoss. Often parent will prefer to lay eggs in and above the peatmoss substrate rather than over a bare bottom tank.

A 2 to 5-mm thick peatmoss layer contained in the open plastic recipient is used as spawning medium. The fish can be left to spawn for 5 to 7 days in such a set-up, after which, the peatmoss is removed, carefully pressed and left to dry out a little further until the well known tobacco consistency is reached (outer layer becoming light brown). Finally the spawning medium with the eggs is placed into a sealed plastic bag with enough air for the long incubation period. The eggs of N.janpapi are stored away for a period of 2.5 to months at a constant temperature of around 24C (74F). I experienced successful hatching rates after 7 months of wet-dry preservation in the TAN 97/9 "Kitonga North" population.

Hatching the developed eggs can happen in ordinary aged (having stand for 1 week) and fresh (18C or 64F) tapwater. Water is poured over the peatmoss and this is well shaken in order to wet the eggs and allow them to sink to the bottom. After this the peatmoss still floating on the surface is removed in order to better enable the tiny freshly hatched fry to have easy access to the open water surface (filling the swimbladder with air).

In the final hatching recipient, which should at least have a capacity of 5 litres, the eggs are covered with a 20 to 50-mm thick water layer. At times, 24 hours of immersion are needed before the first fry appear, the last fry to hatch can take several days.

As the newly hatched fry are really very small {1.5-mm in length and smaller than N.patrizii or N.rachovii fry} one must add "green water" containing lots of infusoria and micro organisms. The biggest fry will never be able to take even the smallest freshly hatched Artemia-nauplii. It will take them at least a 2 weeks stay in "green water" before N.janpapi fry will be able to take freshly hatched Artemia-nauplii. One will thus have to be prepared for the hatching by ensuring the ready availability of Paramecium rich water or "green water" at hatching time. If micro-organism rich water is not available at hatching time, it is often much to late to quickly prepare some and the maintenance of the species is in jeopardy. It is thus always a good idea to keep somewhere in the garden or the balcony some aged water in a container with some dry leaves. Over the time, this medium will generate the necessary micro-organism for any tiny fry.

Better is however the use of the large Paramecium which were collected, somewhere in the early nineties, in Cameroun by some German hobbyists. This clean and non smelling Paramecium culture is easily maintained all the year round in plain aged tap water in an open 5 litres jar by adding 1 to 2 drops of milk every week to the culture (do not overfeed the culture as it will vanish). In this way, a readily available Paramecium culture will always be at hand for feeding the smallest fry. The medium is very rich in Paramecium, often visible by the naked eye as clouds, easy for grabs by the fry. When the number of Paramecium in the fry tank decreases, some 10-20 centilitres of the culture medium is poured directly into the tank with the fry. Using this medium will accelerate the growth rate of the fry by one week.

For hatching, I use clean aged water in a 2 litres tank. Once the tiny fry is swimming freely, it is transferred into the larger tank containing either green water collected in my daughter's turtle garden pond and brought up to the right temperature or the above mentioned Paramecium culture. The advantage of using a larger tank is that it is less prone to pollution and there is ample supply of micro-organisms which the fry can eat besides from forming a much more stable growth environment. I usually also add one or two crushed dry leaves. In addition, a large tank makes frequent water changes less urgent and the water level can be raised regularly without having to disturb the fry.

The use of a large stabilised tank provide for a safe environment for the tiny fry. The presence during several hours per day of a bright overhead light is necessary so that the fry can easily locate moving preys. In such a set-up, there is absolutely no need for any filtering or even to generate gentle water movements of the surface layer. Most micro-organisms will concentrate just below the surface, creating a murky floating layer which should not be removed. This layer is used by the fry in search of food.

After about 1 to 2 weeks, once the fry can eat the smallest freshly hatched Artemia-nauplii, most problems are generally over. It is however a slow growing species and the growth rates observed in Nothobranchius are never achieved with Aphyobranchius. After about 3-4 months the first males start to colour-up and they become sexually active at about 5-6 months.

They are valuable additions to Nothobranchius-only community tanks as they will occupy the surface and mid-water layers of the tank, just as most Aphyosemion would do.