Nothobranchius polli Wildekamp, 1978

polli: (Latinised adjective) Dedication name in honour of Professor max Poll, Royal Museum for Central Africa (RMCA), Tervuren, Belgium.

First Description

Wildekamp, R. H. - 1978 - "Redescription of Nothobranchius brieni Poll 1938 and the description of three new Nothobranchius species from the province of Shaba, Zaire" - Rev. Zool. Afr., 92 (2): 341-354.

Terra Typica

The Dilungu swamp, near Mwadingusha (10°45'S-27°15'E), Katanga Province (presently Shaba), Congo (Zaire, Democratic Republic of Congo).

Figure 4: Collecting sites of N.polli in Katanga/Shaba province - Congo/Zaire/DRC.

Holotype

The Type specimens are preserved in the Royal Museum for Central Africa in Tervuren, Belgium.

Adult male (36-mm total length, 29-mm standard length) collected by M. Dricot in the Dilungu swamp, near Mwadingusha (10°45'S-27°15'E), Katanga Province (presently Shaba), Congo (Zaire, Democratic Republic of Congo) in February 1940.

Paratypes

• One adult male (35-mm total length) and 2 adult females (31 & 27-mm TL and 25 & 22.5-mm SL); same collection data as holotype.
• 6 adult males (25 to 46-mm TL; 20.5 to 39.5-mm SL) and 7 adult females (23 to 35.5-mm TL) collected by Magis in the Kisungu swamp along the ancient road from Jadotville (Likasi) to Mwadingusha.

Synonyms

According to Wildekamp (1978) who re-described the species the following synonyms can be encountered in literature:

• N.brieni non Poll, 1938; sensu van den Nieuwenhuizen, 1962; Magis 1963; Scheel 1968
• N. affinis taeniopygus non Hilgendorf, 1891, sensu van den Nieuwenhuizen, 1970; Poll, 1976
• N.taeniopygus non Hilgendorf, 1891, sensu Bell Cross, 1965; Jubb, 1976

According to Wildekamp (1978), N.polli is probably not more closely related to N.brieni as any other representative in the N.rachovii (Ahl, 1926) group. Most probably, there is a closer relationship with the species (identified these days as N.kafuensis Wildekamp & Rosenstock, 1989) reported by Bell Cross and Jubb from the Kafue River system in central Zambia (Bell Cross 1965; Jubb 1969) and by Jubb (1975) from the Chobe River in the Caprivi Strip in Namibia. However, in 1978, Wildekamp suggested that on the basis of bio-geographical grounds these 2 species would probably not be identical to N.polli, although there is a certain similarity in the pattern.

Meristic & Morphometric Data

D= 15-18; A= 15-19; scales along longitudinal line: 26-29+2

According to the re-description given by Wildekamp (1978), the body height 3 to 3.8 times and head length 2.7 to 3.3 times in body length. Eye diameter 3.4 to 4.1 times and interorbital width 2.8 to 3.3 times in head length. Snout length 0.8 to 1.4 times in eye diameter which in itself goes 1 to 1.35 times in interorbital width. The caudal fin is 1.3 to 1.55 times as long as its height. The dorsal fin has between 15 and 18 rays. The anal fin between 15 and 19.

There are between 26 and 19 scales along the mid-lateral line series and 2 scales on the tail peduncle. Around the body, there are 24 scales in front of the ventral fin and between 12 and 14 around the caudal peduncle.

The morpho-type of N.polli resembles closely the general Nothobranchius body shape and it has a terminal, large and slightly upwardly directed mouth opening with a slightly protruding lower jaw. All teeth are conical in shape, slightly curved towards the back and placed in irregular rows. Teeth from the first external row are always the longest.

The dorsal fin origin always starts in front of the anal fin origin. When laid down, the dorsal fin reaches beyond the basis of the caudal fin, the anal fin reaches just before the basis of the caudal fin.

A table of comparison of morphometric data for the four Nothobranchius species found in south-eastern Katanga/Shaba province, and which where all formerly identified as N.brieni, was given by Wildekamp in 1978 (values expressed in % of standard length and scales/ray counts).

Characteristics/Species	N.brieni	N.malaissei	N.polli	N.symoensi
Body height	29.2	30.1	29.1	30.4
Head length	33.8	33.7	33.4	32.2
Height caudal peduncle	13.6	13.4	12.8	12.7
Length caudal peduncle	18.2	18.7	17.8	19.4
Eye diameter	9.2	8.4	9.0	8.4
Interorbital distance	12.4	11.0	10.4	9.8
Snout length	8.8	8.0	8.0	7.3
Distance snout-dorsal fin	59.3	57.7	57.4	63.2
Distance snout-anal fin	58.3	63.5	63.0	64.8
Distance snout-ventral fin	48.5	51.7	52.5	50.9
# rays dorsal fin	15.7	16.4	16.6	18.0
# rays anal fin	17.5	16.2	16.5	17.5
# scales mid-lateral line series	28.0	29.2	27.3	29.0
# scales around body	24.0	23.7	24.0	25.0

Source: Wildekamp, 1978.

Code

POL

Size

The largest reported and preserved male specimen reached 46-mm, the largest female 35.5-mm.

Colour Description

Male: The particularly beautiful colours of living males have been described by R. H. Wildekamp (1978) on the basis of a picture taken by A. Van den Nieuwenhuizen.

In adult males, the central part of the scales is light blue with reflections and is surrounded in a large carmine red border, through which, over the entire body, a reticulation pattern is created. This carmine red to purple colour can also be found over head, throat and gillcover. On the operculum there are also a number of light blue spots. The dorsal fin is blue-grey with a dense pattern of dark red spots, which in the distal part of the fin get denser and larger; the dorsal fin shows often a thin light blue marginal hem. The base of the anal fin is light yellow to whitish, followed by one or two carmine red bands. The remaining part of the anal fin is pale yellow, which is followed by a dense pattern of red spots; finally, the marginal band is always black. The caudal fin is carmine red with several light blue spots, followed by a broad light blue to whitish band and a thin black marginal band. The ventral fins are yellowish with red spots, the pectoral fins are light orange.

Males preserved in alcohol keep their carmine red coloration but to a far lesser extend than in living specimens.

Female: The females have a uniformly grey-brown body; the scales behind the pectoral fins and over the back show light blue reflections. All the fins are further quite colourless.

History

Formerly, before Wildekamp's 1978 revision of the Nothobranchius species of the Shaba province, all Nothobranchius species from this vast region were identified as N.brieni Poll, 1938.

From analysis of the material preserved in the Royal Museum for Central Africa, Tervuren, Belgium, it appeared to Wildekamp (1978) that all male specimens of this species must have displayed a very particular colour pattern, substantially different from the colour patterns on the other known Nothobranchius species. More careful investigations of this type material by Wildekamp confirmed the existence of a distinct species, with specific morphological characteristics and colour pattern, of which the clear triangular spot over the lower caudal fin is characteristic of N.brieni.

Poll described N.brieni on the basis of 24 specimens, which were collected by Prof. P. Brien in the vicinity of Bukama town, from a number of pools interconnecting with the Lualaba river system. The description provides a good picture of the morphological data of this species as well as a good impression on the specific colour pattern. G. F. de Witte collected also N.brieni in the Sanga River and near Mabwe. Both localities lie within the Upemba National Park, next to Lake Upemba, which also belongs to the Lualaba river system and can be found some 90-km north of Bukama.

According to Wildekamp (1978), the generally accepted picture of N.polli was based on a number of colour photographs taken by A. van den Nieuwenhuizen from living specimens, collected between Likasi {former Jadotville} and Lubumbashi {former Elisabethville} by Graindorge in 1959-60. These pictures were published in several aquarium journals under the label N.brieni. This name identification was based on information received from Graindorge, the collector, who had himself received the name from Professor M. Poll. Poll identified this Likasi-Lubumbashi material from the upper Lufira River as N.brieni. Later on the basis of morphological distinctions with N.brieni and based on the presence of a wide clear sub-marginal band, which corresponded more to the pictures of N.taeniopygus, as previously published by Boulenger in 1907 and in 1914, he identified it as Nothobranchius affinis taeniopygus Hilgendorf, 1891. Van den NIEUWENHUIZEN's attention was drawn to this change by J. J. Hoedeman and in later publications the photographs appeared with the modified name.

During studies aimed at discovering the real identity of N.brieni, on the basis of published literature, it appeared rather rapidly to Wildekamp (1978) that this new name introduced a lot of confusion. In fact it even appeared that the name was being used to identify several Nothobranchius species originating from Shaba as well as from the upper Zambezi River system. J. J. Scheel doubted in his book "Rivulins of the Old World" on it's validity as a species and Bell-Cross even considered N.brieni as an under-species of N.taeniopygus and, later, even as one of its synonyms. Both, Bell-Cross and Scheel based their conclusions on the phenotypic appearance of N.brieni in Van den Nieuwenhuizen's picture and the corresponding first identification by Poll.

The same species, as the one appearing on Van den Nieuwenhuizen's photographs, had been collected by N. Magis in a number of pools between Likasi and Mwadingusha, within the middle Lufira river system. The collector gave a good description of the collecting sites. Also in this case, Poll identified these specimens as N.brieni; but later changed his opinion and gave them the name N. affinis taeniopygus. However, material collected by R. H. Wildekamp in the vicinity of Lake Manyara (Tanzania) as well as in the Bubu river system, which can be considered as a typo-typical representative for the N.taeniopygus-group, showed quite distinctive differences in colour patterns with the Shaba material.

Also, on zoogeographical grounds, it appeared impossible that both species could be identical, considering that, from the early Pleistocene period onwards, the western part of the Great Rift Valley, to which the large graben movements of Lake Tanganyika were part, formed an immense natural barrier between both geographical areas. On these grounds, R. H. Wildekamp described this species as N.polli spec. novem in 1978.

Distribution & Biotope

All presently known collecting sites of N.polli lie in Katanga/Shaba province in the Dem. Rep. Congo/Zaire, in a vast triangle formed by the towns of Jadotville/Likasi, Elisabethville/Lubumbashi and Mwadingusha dam.

Figure 6: Terra Typica and collecting sites of N.polli in south-eastern RDC/Zaire

All collecting sites lie within the Upper Lufira river system. The holotype was collected by Dricot in the Dilungu swamp, near Mwadingusha (10°45'S-27°15'E). Magis collected some additional specimens in the Kisungu swamp along the ancient road from Jadotville (Likasi) to Mwadingusha. C. Poncelet, A. de Deker and W. Verheyen also collected N.polli in the vicinity of Likasi, Mwadingusha and Lubumbashi. Graindorge collected some specimens, which provided Van den Nieuwenhuizen with some of the few rare colour photographs of the species, in 1959-1960 in a large swampy area between the towns of Likasi and Lubumbashi.

The general area in which Magis (1963) found N.polli is known as the Kisungu swamps, which is a large "Dembo"[swampy depression] originating on the foothills of the Koni Mounts. It runs along a gentle slope (from 1 to 0.1%) until it reaches the mandmade waterreservoir on the Lufira River (Mwadingusha reservoir) where it forms part of its northwestern shores. The undeep soil has a sand-clay composition and rests on a lateritic plate, which reaches the surface in many places, especially in the superior part of the "dembo". The vegetation cover presents a steppe-like caracter composed of patches of Loudetia simplex (Nees) and Monocymbium ceresiiformis (Nees) grasses, located in a light forest of Brachystegia spiciformis (Streel, 1960). Where the terrain slope declines, the steppe subsists but under the humid form with cyperacaen grasses or, in the near vicinity of the reservoir, as a swamp. In the upper part of the latter zone, where the Kisungus are crossed by the Jadotville (presently Likasi)-Mwadingusha road, some patches of dense humid forest break the monotony of the sceenery generated by the grass-steppe. According to Streel, this forest appears to be supported by nearly permanent deep-water humidity. The presence of a permanent waterhole in the village of Kasumbalesa would tend to confirm this.

Figure 7: Schematic situation at location "km 19" after Magis, 1963.

Magis (1963), who thought to be studying the ecology of N.brieni before Wildekamp's redescription of the species in 1978, found N.polli in:

• Roadside ditches along the elevated road from Jadotville to Mwadingusha, near the above mentioned patches of dense humid forest. During the rainy season, these ditches attract noumerous fishermen who capture in their traps, apart from Nothobranchius polli, Ctenopoma multispinis Peters (Anabantidae), various small barbels (Barbus sp.), Aplocheilichthys luluae Fowles and numerous small Cichlids (Tilapia sparrmani Smith and Haplochromis philander Weber). All these species, except the Nothobranchius, occurr in the reservoir and in adjacent rivers (Magis, 1961_a,b);
• In the waterholes of "Km. 19" (Figure 7): along the ancient Jadotville - Mwadingusha road, running north of the previous one. At locality "Km. 19", there are temporary pools from where field observations on the ecology of N.polli where made by Magis in 1958 (Magis, 1963; Magis, 1962);
• Waterholes very similar to the previous ones, located at "Km. 21" and "Km. 14" of the same ancient Jadotville - Mwadingusha road. However, these holes contained less abundant Nothobranchius populations;

Magis (1963), who followed the seasonal evolution (from May 1958 to December 1959) of some Nothobranchius pools located at "Km. 19" along the ancient Jadotville - Mwadingusha road, provided a description of the collecting site (Figure 7). The 3 small and shallow waterpools extend on both sides of the road. They are neither linked to a river nor to a larger pool or to a spring. These totally isolated temporay pools reconstitute themselves each year as a result of rain and runn-off waters accumulating in depressions resulting from a localised collapse of the laterite plate. The bottom of the pools is covered with a 15-cm thick layer of black mud. These sediments originate primarily from the vegetation cover, which, during the rainy season, abundantly colonizes these waterholes: wild rice, waterlillies, lotus, Polygonum and Myriophyllum. During the dry season, this mud layer hardens and cracks nearly over its entire depth.

Magis (1963) further reports that the water in which these Nothobranchius live is weakly acid [Ph of 6.3 to 6.5], very weakly mineralised (electric conductibility at 18°C x 10⁶ 26 to 45) and presents a very limited buffering capacity. The hardness is in the order of 1 to 2° French and the temperature varies between 15 and 30°C.

The sequential inventory of captured specimens in pools N°1 and N°2 is presented in Table 1. The water recolonisation happened suddenly after 3 months of complete desiccation. However in 1959, this happened sequentially: pool N°1, downstreamly placed, already contained water in its deepest central part, whilst in pool N°2, the bottom mud was just simply humid. On 7-XII-59, Magis (1963) found 8 juveniles in pool N°1 [19.1 mm average length], whilest in pool N°2 he only could find fingerlings with a 5 time smaller length [about 4 mm long]. No specimens were found in pool N°3 during the entire survey period. No adult specimens and no other species were found just after water recolonisation, indicating that no connexions with other waters existed and that the existence of durable eggs had to be the way of survival for the species. In addition, comparisons of sizes at capture between 7.XII.58 and 16.I.59, and between XI.59 and XII.59, revealed that growth is rapid and that sexual maturity is quickly reached.

Table 1: Inventory of "Km.19" collections (after Magis, 1963)

	Pool n°1					Pool n°2
Date	` `	a a			Sex ratio	` `	a a	Sex ratio
15-V-1958	# 14 <47 mm> 1 of 50 mm	# 47 <35 mm>			0.297	# 8 <35 mm>	# 71<25 mm>	0.112
14-VI-1958	# 19 <50 mm> 2 of 28 mm	# 20 - <37 mm> 1 of 24 mm			1.000	# 11<35 mm> 1 of 24 mm 2 of 40 mm	# 26 <32 mm>	0.538
12-VII-1958	# 8 <50 mm> 1 of 24 mm	# 27 <37 mm>			0.333	# 5 <35 mm> 2 - 43,48 mm 2 - 28,31 mm	# 26 <32 mm>	0.346
15-VIII-1958	Dried-up pool					Dried-up pool
15-XI-1958	Beginning of water recolonisation					Bottom mud just simply humid
7-XII-1958	8 juveniles: 17,17,18,18,19,19,22,23 mm 19,1 mm average length (± 0,791)					6 fingerlings: 4 mm
16-I-1959	6 juveniles: 19,20,20,21,21,24 mm 20.8 mm average length (± 0,40) 1 ` of 27 mm with sexual colours					9 juveniles: 13,13,14,14,15,15,15,16,17 mm 14.7 mm average length (± 0,44) 2 a a (?) of 18,19 mm
15-II-1959	# 9 - 26 to 33 mm <29.7 mm>(± 0,88)	# 14 - 24 to 34 mm <27.5 mm>(± 0,82)		0.642		# 13 - 23 to 32mm <26.5 mm>(± 0,82)	# 20 - 19 to 31 mm <25.6 mm>(± 0,57)	0.650
14-III-1959	# 2 - 38,42 mm	# 3 :26,29,37mm				# 1 - 34 mm	# 20 - 21 to 30 mm <25.0 mm> (± 0,44)	0.050
Between 18-I and 14-III-59, pools 1 and 2 were interconnected through a shallow (3 to 5 cm deep) water flow. A tiny rivulet also interconnects pool N°2 to pool N°3, but the Nothobranchius do not distribute making use of it.
18-IV-1959		# 4 :36,37,37,38 mm					# 2 :26,28 mm
16-V-1959	# 1: 42 mm	# 6 - 26 to 32 mm <29.5 mm> (± 0,90)					# 8 - 26 to 30 mm <27.75 mm> (± 0,42)
13-VI-1959		# 3 : 28,30,32 mm				3 dead found in exundated parts
15-VII-1959	Dried-up pool					Dried-up pool
6-XI-1959	Brutal filling of pools following heavy downpour					Brutal filling of pools following heavy downpour
15-XI-1959	8 juveniles: 10 mm					Not inspected
26-XII-1959	# 1: 40 mm	# 1: 32 mm				Not inspected

Note: # = number of specimens captured - <length> = average total length

Magis (1963) also noticed, from field observations made in December 58 and in January 59 in pools N°1 and 2, that even over a very small geographic area, one can find Nothobranchius populations of remarkably different ages. The age difference resulted primarily from different hatching times of the durable eggs, which can either consist of resting fry (diapause at the finalised embryo stage) or resting egg (diapause at a primitive embryo stage) which itself is a direct consequence of the morphology of the temporay pools. However, confinement of Nothobranchius populations is probably one reason besides from others, which are still unkown. When pools N°1 and 2 got interconnected at the height of the rainy season (from 18.I to 14.III.59), one could hardly observe a markable change in length frequency distribution, suggesting that cross-overs from one pool to the next did not occur. The permanent absence of Nothobranchius in pool N°3 has already previously been mentioned. According to Magis (1963), it thus appears as if Nothobranchius populations are strongly confined to the habitat they occupy.

Another observation whih arose from Magis' field study indicates that sexual colouration in male specimen occurr when they reach a length of about 25 mm. It happens first along the anal fin with the appearance of horizontal coloured bands, followed by the body scales, which, initially greyish-green, become dark-violet. Some very few males never appear to colour out. The inhibition or lack of colour development in these few males does not appear to be final as Magis (1963) could experience with a colourless male kept in a tank and who suddenly developped its colours.

Besides from differences in colour patterns, sexes also distinguish themselves by their size. Field data gathered by Magis (1963) indicate that male are systematically larger than females (Table 1). This is also shown in Table 2, which represents the length frequency distribution of Nothobranchius polli, this time collected in roadside ditches along the new road from Jadotville to Mwadingusha on May 31^st, 1958, just prior to dessication.

Statistical comparison with t-Student test on averages was highly significant ["t-Student"=8.44 with P<0.01], indicating that differences between males and females of a population is not linked to a particular environment but is a biological feature of the species. Magis (1963) also mentions that the length of specimens collected in Bukama, Lualaba River during the same period [May 1937] and which he believed to belong to the same species, had a size ranging between 34 and 58 mm (Poll, 1938). With Wildekamp's redescription of various Nothobranchius species from Shaba in 1978, the larger specimens appeared to belong to another species and to be Nothobranchius brieni, the smaller ones studied in the field by Magis belonged to Nothobranchius polli.

Magis (1963) further mentions that the intestinal content in 2 males and 5 females revealed a strong percentage of cuticular debris originating from insect larvae. In lower proportions, he also identified numerous Chydorids and at times debris of Cyclops, Ostracods, Rotifers (Bracionus falcatus Zacharias) and fragments of mollusc shells.

Concerning the evolution of ovaries in females, Magis (1963) noted that on 12-VII-1958, more than half of the females collected at locality "Km. 19" presented an enlarged belly. They appeared to have ovaries full of brilliant and translucent yellow ovules of sizes ranging from 1.0 to 1.5 mm in diameter. On March 14^th 1959, Magis observed well-differentiated ovaries, filled with ovules ranging from 0.5 to 0.75 mm in diameter. On Mai 16^th the ovules had increased in size, their diameter being in the order of 0.5 to 1.0 mm. They still had their brilliant and translucent yellow coloration.

According to Van den Nieuwenhuizen (1962) who received the information from Graindorge who himself collected N.polli in the wild between Likasi and Lubumbashi, there is a large swampy area between both towns. This swamp is partly connected to running water and another part is, on the contrary, entirely dependent of the water table.

In Shaba/Katanga, the dry season lasts from May to October and the rainy season from November to April. During the latter period, one can find almost everywhere numerous pools and pounds. These vary in size from 1.5 to 10-m across and from 30-cm to 1.5-m deep. These pools subsist the entire rainy season. The rainy season starts about mid-October or beginning of November with heavy downpours. In December, rainfall is on the average 150 mm, in January around 200 mm, in February and March 300 mm, in April somewhat more than 100 mm, then the rain end. During the rainy season the total rainfall lies between 1200 to 1300 mm. The air temperature during the rainy season averages 27°C in the sun and 16-19°C in the shadow.

It still remains a wonder that fish eggs, which from May to November remain in the cracked dry bottom mud, hatch after having survived substantial temperature changes. During the winter months, air temperature can drop from 23°C during the day to about 2°C at night. In valleys, it can even reach freezing point. Then the temperature increases again in July. At the end of the dry season one observes the first changes. The colour of the soil becomes red-brown. In August reigns a heat of around 30°C. In September/October the maximum reaches 33-35°C. Even though the sun still burns relentlessly, the first new leaves appears on the trees. Then the downpours reappear in October, and by December, the scenery has entirely changed: everything is green again.

According to the Graindorge family (in van den Nieuwenhuizen, 1962), which collected N.polli in the wild, one can find in December 2 to 3-cm long N.polli fry in the new pools. They are then being caught in places where there is hardly any current. These pools are overgrown with waterlilies (Nymphea) and Othelia which all flourishes in February. One can also find Salvinia and numerous Urticularia and, in the larger pools and in running water, Eichhornia crassipes.

Water temperatures in smaller pools appear to be high. Graindorge recorded in mid-March, at the surface and in the morning, water temperatures of 25°C; they reached 27°C around midday. In a layer, 80 to 100-cm deeper, temperature was one degree lower. In May, water temperatures were much lower; at the surface of not yet entirely dried-up pools and in stagnant water and at mid-day, the temperature reached 20.5°C; one meter deeper, it was one degree less.

The water hardness was low; in March the marsh water measured 0.5 DH and in May between 2 and 5°DH. pH-value changed from March to May from 6.3 to 7. The water was relatively acid to neutral. These measures apply also to the running part of the swamp. In these waters, Graindorge could also catch following accompanying species: several Barbus species, Ctenopoma, Amphillius, Aplocheilichthys, small Synodontis and other species. The water in the swamp and in the ditches was clear allowing an identification of the species swimming in it.

When catching N.polli and the other fish, it appeared that the young fish often stayed in the open water, whilst the older specimens were found closer inshore, between the grasses and the edge vegetation. In the small pools it was even possible to catch the fish by hand or with a very simple net.

In 1960, A. van den Nieuwenhuizen obtained eggs from E. Graindorge through airmail and spread the content over three incubating plates. The eggs, laid in February and March, originated from wild caught specimens, which had been kept by Graindorge in a tank. When he inspected some eggs in May, these were in the first stage of their development. It was only at the end of the summer that van den Nieuwenhuizen could add some water. When hatching the fry he observed something remarkable. It appeared as if all the young fish were belly sliders, but they all grew properly and after 4 weeks they measured around 1.5-cm in length. He obtained about 60 young fry, which he distributed over 3 tanks in order to minimise risks. In spite of these precautions, the fifth week was a real battlefield; one young died after the other, until all had passed away. They all presented fast respiratory movements and were hanging on the water surface.

In the meantime, the situation in Congo got out of hand (independence scuffles) but before leaving to Belgium, Graindorge collected some new eggs and gave them once again to van den Nieuwenhuizen, this time with more success. However, also these fish totally disappeared from the hobby.

Zoogeography

The fish fauna of Shaba province {formerly Katanga} in south-eastern Democratic Republic of Congo/Zaire has been the subject of several investigations. Prof. Max Poll (1963) carried out a study on the relationship between hydrological basins and the distribution patterns of the fish fauna. He recognised three major distinct river systems in south-eastern Congo: the Kasai-, Luapula- and Lualaba-river systems. Surprisingly, at first glance, the Kasai river system, running westerly into Shaba province, presents large similarities in its fish fauna with the northerly running Zaire/Congo river system. The Luapula river in the east, to which also Lake Bangwuelu and Lake Moero belong, presents a fish fauna resembling more the southerly running Zambezi River system.

The Lualaba River system, which runs in the middle displays an ichthyofauna resembling on the one hand the fish fauna of the central Congo/Zaire River system and on the other hand the ichthyofauna of the upper Zambezi River system. A particularity of the Lualaba River system is that in its higher parts, south of the "Portes d'Enfer" falls, it contains a fish fauna which presents striking affinities with the Upper Nile River system. These "Portes d'Enfer" {= "doors to hell"} constitute a series of rapids where supposedly previous links with the Nile River system have existed. However, the Lufira River, a tributary of the Lualaba, shows in its higher system a particular fish fauna, whereby also a certain level of similarity can again be established with the upper Zambezi ichthyofauna.

The east African Cyprinodont-genus Nothobranchius presents a large distribution within the Zambezi River system, whereby a distinction has to be made between the fishes of the upper river system and those belonging to the coastal lower river system. The differences observed by Poll (1938, 1963) in the ichthyofauna of the upper Zambezi and the Lualaba-Lufira River systems on the one hand and that of the Luapula-Bangwuelu-Moero systems on the other hand, can also be observed for the genus Nothobranchius.

Maintenance & Breeding

Maintenance and breeding aspects reported here are based on experience gained and reported by Graindorge and van den Nieuwenhuizen (1962).

N.polli lives in temporary pools and swamps densely covered with vegetation and sometimes in small rivulets, which can dry-out entirely once a year. Therefore, the fishes can best be kept in an aquarium with dense vegetation, which provides at the same time numerous hiding opportunities to the females, which are constantly being followed by the males. The size of the tanks will or course be in relation to the number of fishes kept, but can seldomly be too large.

When inspecting the eggs of N.polli after only 6 weeks of incubation, one observes no development at all. This is easily understandable considering that in the wild the eggs of this species undergo a dry period from about June to end October/mid-November, or about 5 months. In this species, one can also find a higher proportion of resting eggs, which starts to develop at an even later stage. Because of this phenomenon, in case of early or short rains, these eggs have not yet reached the hatching stage. In this way, nature protects the species' survival. Best hatching results are thus obtained when the incubation time is kept at 6 to 8 months as in N.rachovi.

Van den Nieuwenhuizen maintained the sexes separated (for about one week) in smaller tanks and fed the fish normally with different food sources (mosquito larvae, Tubifex, small beetles, etc.) and brought them together for a group spawning for short periods of time (one to two weeks). Or one can keep females together with only one to two males, which are being changed very regularly in order to, secure a correct exchange of genes in the population. Several males can be brought together provided there is a dense vegetation cover and/or when sufficient hiding possibilities have been introduced into the tank. Male are in general not very aggressive towards each other but like to display, as most Nothobranchius species do.

A tank with following characteristics: 40x25x25-cm and with a water depth of about 10-cm is already sufficient (van den Nieuwenhuizen, 1962). The water can be soft or hard, as, according to Van den Nieuwenhuizen's experience, the water hardness does not seem to play any major role. The water temperature should be kept between 22 and 24°C.; the bottom is covered with a litter of some 2-cm properly cooked and rinsed peatmoss.

It is best to breed with more fishes, 1 male for two to three females. After a stay of one to two weeks in such an environment, the females need to be put to rest and can be replaced with new ones. After 2 weeks the water is removed and the bottom content is discharged into a net and slightly pressed in the hand. The peatmoss containing the eggs is left to dry-out a little further on paper until the typical tobacco consistency is reached and the outer layers turn light-brown. Eggs and peatmoss are then stored into a sealed plastic bag with name label and both dates (dry and hatching). For the long journey, there should also be enough air trapped together with the peatmoss and the eggs.

According to van den Nieuwenhuizen (1962), one should not attempt to leave peatmoss and eggs in the tank and let it dry out from there, as the peatmoss would definitely become far to dry and the eggs would disintegrate. With N.polli, the eggs will surely desiccate in such a set-up even though in nature, one could find similar conditions when the bottom of the pool starts to crack.

The plastic bags are best kept at a temperature of 23° C. At temperatures between 28 and 30°C, all went wrong. During the incubation period, it is advisable to change from time to time the air in the sealed plastic bag.

After the incubation period, the peatmoss and the developed eggs are wetted. At hatching time, one should not use a too high water depth with N.polli in order to avoid belly sliders. Van den Nieuwenhuizen (1962) could not initially understand this until he observed the hatching process in 15-cm of water. All fry, being initially belly sliders after hatching, tried to reach for the 15-cm higher surface but in vain. When he decided to lower the water surface to 4-cm, the situation changed. All belly sliders managed to fill their airbladder and to swim normally.

Before hatching the eggs, one must ensure to have set-up, one or two days in advance, an Artemia culture for the young fry, with which the fry can immediately be fed. The fry should be fed regularly and often in order to secure a steady grow.

N.polli growths very rapidly. In the wild, according to information provided by E. Graindorge, the fish reach maturity in only 55 days after the start of the rains. With adequate feeding they measure about 1.5-cm, in only 2 weeks. During growth, regular partial water changes are recommended as well as frequently changing the food sources and progressively increasing the size of the food. A large aquarium is also preferable to small and crowded tanks.

Van den Nieuwenhuizen (1962) supposed that wild caught fish were infected with parasites, as he could not explain why the entire first batch died almost together in a matter of days. He was certain the fish had not been infected with Oodinium (velvet).

As a final note, Van Den Nieuwenhuizen (1962b) remembers that too high water temperatures are not to be recommended for this species. He notes that water temperature above 27°C and a strong feeding will definitively lead to hydropsy and death.

Literature

• Bell Cross, G. - 1965 - "Additions and amendments to the checklist of the fishes of Zambia" - The Puku, Occ. Papers; Department of Game & Fisheries, Zambia, 3.
• Boulenger, G. A. - 1907 - "Zoology of Egypt, Fishes of the Nile"; London.
• Boulenger, G. A. - 1914 - "Catalogue of the freshwater fishes of Africa"; Vol. III, Trustees, London.
• Jackson, P. B. N. - 1961 - "The fishes of Northern Rhodesia", Government printer, Lusaka.
• Jubb, R. A. - 1969 - "The Nothobranchius of Southern Africa", Ann. Cape Province Museum, 8-1.
• Jubb, R. A. - 1975 - "The Jubb Papers", J.A.K.A. KN., VIII-6.
• Jubb, R. A. - 1976 - "Three interesting species of Nothobranchius", B.K.A. Killi News, 127.
• Magis, N.; Goorts, P. & Wilmet, J. - 1961_a - "Les aspects biologiques, humains et économiques de la pêche dans le lac de barrage de la Lufira (Katanga)" - Fulreac, édit., 127 pp., Liège
• Magis, N. - 1961_b - "La pêche dans les lacs de retenue de Koni et de N'Zilo I (Haut Katanga)" - Fulreac, édit., 53 pp., Liège, Belgique.
• Magis, N. - 1962 - "Etude limnologique des lacs artificiels de la Lufira et du Lualaba (Haut Katanga). I. Le régime hydraulique, les variations saisonnières de la température" - Int.Rev.gest. Hydrobiol., XLVII, N°1, p.33
• Magis, N. - 1963 - "Renseignements écologiques sur le Nothobranchius brieni Poll (Cyprinodontidae, Rivulini)" - Bull. Soc. Sci. Liège, 32 (1-2):174-184;
• Poll, M. - 1938 - "Poissons du Katanga récoltés par le Prof. Brien", Revue Zool. Bot. Afr., XXX-n°4, p.389.
• Poll, M. - 1963 - "Zoogéographie ichthyologique du cours supérieur du Lualaba" in Coloque sur les problèmes biogéographiques du Parc National de l'Upemba; Publ. Univ. Elisabethville, 6.
• Poll, M. - 1976 - "Exploration du Parc National de l'Upemba, Poissons" Publ. Inst. Parcs Nat. Bruxelles, 1976.
• Rothe, S. - 1993 - Es geht schon wieder los: Nothobranchius symoensi Wildekamp, 1978 - DKG-Journal, 25(8):124-126,1993; 2 figs.
• Tait, C. C. 1975 - "Notes on the species Nothobranchius (Cyprinodontidae)", The Puku, Occ. Papers; Dept. Game & Fisheries; Zambia, n° 3.
• van den Nieuwenhuizen, A. - 1962 - Die natürliche Umwelt unserer Aquarienfische im Kongo (I) - DATZ; 15(7); Juli 1962:202-204; 7 figs.
• van den Nieuwenhuizen, A. - 1962 - Die natürliche Umwelt unserer Aquarienfische im Kongo (II) - DATZ; 15(8); August 1962:225-228; 7 figs.
• van den Nieuwenhuizen, A. - 1962 - Die natürliche Umwelt unserer Aquarienfische im Kongo (III) - DATZ; 15(9); September 1962:262-264; 6 figs.
• Wildekamp, R. H. - 1978 - Redescription of Nothobranchius brieni Poll, 1938 and the description of three new Nothobranchius species (Pisces, Cyprinodontidae) from the province of Shaba, Zaïre - Rev. Zool. Afr. 92(2): 341-354.
• Wildekamp, R. H. - 1978 - "Nothobranchius species auf der Shaba Provinz - Zaïre" - D. K. G. Journal, 10(8): 121-129, 11 figs.
• Wildekamp, R.H. & J.Rosenstock - 1989 - Anmerkunge zu den Nothobranchius-Arten Sambias mit der Beschreibung von Nothobranchius kafuensis spec. nov. (Cyprinodontiformes, Nothobranchiinae) - D.A.T.Z. 42(7): 413-419.