Previous Page Table of Contents Next Page



C.K. Kapasa and P.A.M. Van Zwieten
Department of Fisheries
P.O. Box 740005


On Lake Mweru-Luapula a major fishery has developed on a species complex called ‘Chisense’ within less than one decade. Little is known about the yield and stocks and about the biology and life history of the most abundant species of the complex, Poecilothrissa moeruensis. This paper presents information on the feeding biology, fecundity, and variations between stocks in inshore and offshore waters of the open lake and the Luapula River. In addition, preliminary data on species composition of the ‘Chisense’ stocks are presented. Data on ‘Chisense’ fishing from a recent frame survey that was conducted in the area give indications of the types of fishing methods used and the size of the fishery. Recently, a study on the ‘Chisense’ complex has started and this paper presents scattered information and hypothesis on which future research could be based.


En moins de 10 ans, une grande pêcherie s'est développée sur le lac Mweru-Luapula; elle exploite un complexe d'espèces appelées “Chisense”. On sait peu de choses concernant le rendement et les stocks ainsi que la biologie et le cycle vital de l'espèce la plus abondante de ce complexe, Poecilothrissa moeruensis. Ce document contient des informations sur l'alimentation, la biologie, la fécondité et les écarts entre stocks des eaux littorales et des eaux hauturières du lac lui-même et de la rivière Luapula. On y trouve en outre des données préliminaires sur la composition spécifique des stocks de “Chisense”. Les données concernant la pêche de “Chisense” provenant d'une récente enquête cadre conduite dans la zone donnent des indications sur la dimension de la pêcherie et ses méthodes. Comme la recherche sur le complexe “Chisense” n'a commencé que récemment, le présent document présente des informations dispersées et des hypothèses sur lesquelles de futures recherches pourraient se fonder.


1.1 General

A rapid increase in the human population has been a major reason causing the exploitation of formerly unexploited pelagic fish species of African lakes. Furthermore the building of man-made lakes in Africa has created large bodies of water in which most of indigenous riverine fish species could not colonize pelagic waters due to changes in environmental conditions (Kapasa and Cowx, 1991). However these lakes have been successfully stocked with pelagic species (Junor and Begg, 1979; Marshall, 1984; Bell-Cross and Bell-Cross, 1971; Balon, 1971).

The exploitation of the small pelagic fish of Mweru-Luapula, a complex of species locally known as Chisense, and their commercial value became more apparent in the 1970's. The word “Chisense” literally means “the small fish which move in a large shoal as if it is being drifted by the wind”. Prior to 1974, the fishery on Chisense appears to have been restricted to the Northern shores of the lake. The fishing was mostly done by women using Chitenge material (meshless cloth) and stocks were not considered worthwhile to exploit. Successful fishing on the complex by the artisanal fishers was reported along the north-eastern shore of the lake after the introduction in 1984 of what is still known as “Scullion Nets”. Under the ‘Jersey Aid Project to Chisense’ 30 fishers obtained soft loans. During the same period there were about 200 fishers recorded along the northern shore who were using meshless netting materials, that is, nets made out of mosquito nets, maize meal bags, cotton cloth etc. In the subsequent years the fishery expanded to its present level with an estimated yield of 10000–15,000 tonnes per year. The Chisense fishery is now widespread all over the lake and throughout the Luapula River. It has become one of the major fisheries both in terms of production and labour input. There have been no studies on the size and yield of the fisheries. The biology of the Mweru-Luapula pelagic fish species is also not well known. This paper presents preliminary findings on the size of the Chisense fishery and the feeding biology of Poecilothrissa moeruensis, Neobola moeruensis and several other species in the Chisense by-catch.

1.2 A note on the taxonomy of the Chisense complex

Three pelagic fish species of the family CLUPEIDAE (Pellonulidae) and one species of the family CYPRINIDAE have been described from the open waters of Mweru and Luapula river system.

Jackson, (1961) recorded three species of the clupeids, namely; Potamothrissa acutirostris, Microthrissa stappersii, and M. moeruensis. CLOFFA (1990) records three species of clupeids present in Lake Mweru: Poecilothrissa moeruensis, (formerly Microthrissa moeruensis), Microthrissa stappersii, and Potamothrissa acutirostris. The FAO Fisheries Synopsis on Clupeoid fishes of the world (Whitehead, 1985) mentions only two species: Poecilothrissa moeruensis and Limnothrissa stappersii originally described as Microthrissa (Potamothrissa) stappersii. According to the FAO synopsis Potamothrissa acutirostris is present downstream the Zaire River system downstream but not in Lake Mweru.

From both the lower Luapula River and Lake Mweru, we have encountered only one clupeid species up to now, that is, Poecilothrissa moeruensis. However we have sent samples from the lake and the river for further and hopefully final identification to the Royal Museum of Central Africa, Tervuren, Belgium.

Apart from the clupeids, Chisense consists of the cyprinid Neobola (formerly Engraulicypris) moeruensis.

1.3 The lake and river system

The Mweru-Luapula fishery comprises two ecologically distinct but interdependent areas. These are Lake Mweru proper and the Lower Luapula river with its floodplains and lagoons (Fig. 1).

The lake is located at 8° 27' to 9° 31'S, and 28° 25' to 29° 10'E at an altitude of 930m. The lake is shared between Zaire and Zambia. It is fed by two main river systems, the Luapula from the south and Kalungwishi from the east. The Luvua river flows out from the north-western corner of the lake and joins the Lualaba.

Of the total lake surface area (4,650km2), 58% lies in the Zambian sector. The lake has a maximum recorded depth of 37m in the north-west near the Luvua outflow. Mean depths at the north and south are 10m and 5m respectively. The mean fluctuation in water level is 1.5m with a maximum of 5m. The Luapula river has an extensive swampy floodplain covering 1,500km2 to 2400km2. The floodplain is 160km long and 5–18km wide. Within this area, there are scattered permanent lagoons. Those located on the Zambian side are: Mofwe, Kamibombo, Chitondwe, Pemba, Kaombe and Luche while Kitshomponshi, Kuswa, Luamba and Makusi are found in Zaire. In comparison, the lake has relatively higher conductivity and chemical nutrients than Lake Bangweulu and Luapula River. A detailed description of the morphometric data, chemical and physical characteristics of the fishery system is given by Ssentongo (1990).


2.1 Sampling sites and collection of biological data

In this study, four sampling stations were selected during the period July to September 1992. These were, Isokwe inshore, Isokwe offshore, Kenani (open lake), and Kashilu which is near the mouth of the Luapula River (Fig. 1). Fishing was done using a lift net at Isokwe and Kenani while from Luapula River a boat seine was used. Both nets had a mesh size of 6mm. All the fish samples were collected by the research team based in Nchelenge. Fishing trials were done at night between 18.00–24.00 hours. The fish were attracted by pressure lamps powered by paraffin.

The samples were sorted out, counted and weighed by species. Length frequency of subsamples were done for the two Chisense species. Some specimens of Poecilothrissa moeruensis, Neobola moeruensis (Chisense) and the “by-catch” from the lake samples i.e. Haplochromis sp. “orange”, Seranochromis macrocephala, Raiamas (formerly Barilius) intermedius and Alestes macropthalmus were used for further studies. The parameters which were determined on the sub samples of Chisense and the by-catch included standard length, total weight, sex, fecundity and egg size as well as feeding habits.

Stomach fullness of the species with distinct stomachs was estimated by eye based on percentage of maximum distention (100% full) on assumption that the fish were feeding normally. Stomach contents were examined under Olympus zoom stereo microscope with variable magnification of up to ×40. The frequency of occurrence method was used to determine the percentage contribution. This was done as follows (1) volume of each food category of the total volume of food within all stomachs of that species examined (2) the percentage of individual specimens having that food item within the stomach determined.

Thirty specimens of P. moeruensis caught at Kenani in August 1991 were examined for fecundity. The diameter of the eggs was determined using the following formula:

Egg diameter = Number of eggs in a row/Length of a row (mm)

Counting of eggs was done with the aid of a stereo microscope with varying magnification up to ×40.

2.2 Frame survey

In order to determine the size of the fishery, a frame survey was conducted between June and August 1992. The fishery has been divided into four strata from Mambilima Falls to Lupiya (Bazigos, 1974). See Fig. 1 for details. During the survey, fishermen were interviewed on personal data, number of workers, number of boats, number and type of gear and movement over the lake during the year. Questions pertained to the period of June 1991 up to June 1992. A fisher was defined as a boat and/or gear owner. A worker does not own any of these. A full description of methods used can be found in Aarnink, Kapasa and Van Zweiten, (in prep). Data of Chisense fishers, that is, a person owning any one of the used Chisense fishing gear, were selected from the complete data set and are presented here.


3.1 Biological studies

3.1.1 Species composition of the catch

Samples from the Luapula River and the three lake stations consisted of 78.2 – 93.7% Poecilothrissa moeruensis (Table 1). The Luapula boat seine sample had, apart from P. moeruensis, 22 species in 13 genera. Neobola moeruensis was not found in Luapula River but comprised 12–13.3% of the catch in the lake. Barbus radiatus was caught in the lake only and it comprised 2–7.3% by weight. Barbus lukindae comprised 3.5% by weight and was caught in the river only.

Other species of fish caught are generally less numerous. In the samples used for this study, no juveniles of the commercially important tilapiine species (i.e. Oreochromis macrochir and Tilapia spp) except for those of Serranochromis macrocephala were found.

3.1.2 The size of P. moeruensis

The length frequencies (Fig. 2) of P. moeruensis show that lake specimens are generally smaller than those from the river. The modal standard length of the lake specimens was 26mm while one for the river specimens was 41mm. The largest P. moeruensis caught was 47mm.

3.1.3 Feeding biology Poecilothrissa moeruensis

The dominant food item found in the stomachs of P. moeruensis from the lake is DIPTERA larvae while P. moeruensis from the Luapula River fed mostly on aerial/terrestrial insects, predominated by ephemeropteran adults (Table 2). There was a clear difference in food composition between sites in the lake. These probably are caused by specific environmental conditions at the sites. The difference in feeding habits between P. moeruensis caught from the Luapula River and that from the lake proper cannot be attributed to size, but again is attributed to different environmental conditions. It is noted that larger specimens of P. moeruensis from the lake also seem to feed more on aerial/terrestrial insects.

Furthermore results show that plant material only forms a minor part of the stomach contents. It is possible that phytoplankton or zooplankton is eaten in early life stages of this species, but such data are not yet available. By-catch

Serranochromis macrocephala and Raiamas intermedius predominantly feed on other fish species (Table 3). In as far as it could be determined, the fish prey found in the stomachs of the two species were P. moeruensis. Alestes macropthalmus has a high intake of fish species compared to other food components. This species appears to feed on P. moeruensis as well.

However, ephemeropteran nymphs were also found in abundance. In general specimens of A. macropthalmus in the samples were small.

Further studies on feeding habits show that 92% of the food found in the stomachs of Neobola moeruensis were chitinous remains. Detritus (fine particulate organic matter) was found in the stomachs of Barbus radiatus and Haplochromis sp. (“orange”). It has also been found that Haplochromis spp is the only group in the complex that feeds on planktonic organisms (copepods).

3.1.4 Fecundity of P. moeruensis

Ripe ovaries of six P. moeruensis ranging from 26–42mm in standard length contained 872–1140 eggs each. The egg diameter varied from 0.11–0.33mm with a mean of 0.33mm (Table 4).

3.2 Frame survey

In total 9436 fishers and 7707 boats were counted during the frame survey. The first figure includes fishers operating with gill-nets (appr. 70%), Chisense gear, traps, longlines and baskets (appr. 13%) (Aarnink, Kapasa and van Zweiten, 1992). In the Chisense fishery, 1220 fishers and 5787 workers were counted (Table 5). At the time of the survey (July/August) 62% of these fishers were operating in stratum 1 and most of these came from the beaches of the northern shore of the lake, 24% from stratum 2 (southern lake), 5% from stratum 3 (Luapula mouth and islands) and 4% from stratum 4 (Luapula River). Fishing is done predominantly from plank boats although in stratum 4 canoes are used as well. Driving of boats is mostly done by paddling. Engines and sails are rare. In stratum 1 the use of engines is actively discouraged by Chief Puta. Fishing is done for approximately 20 days per month over an average of 8.5 months in a year.

As shown in Table 6 the most common fishing gear used is the so called Japan net (see Appendix 1 for a description of the various methods used) followed by the boat seine. Other gear (Mutobi, Lift nets, etc) are rarely used. In stratum 1 Japan nets form 85.1% of the gear used, while in all other strata the predominant fishing method is seining. Seining in strata 2 and 3 is done from a boat, while in the Luapula River most seining is done from the shore (beach seining).

Of all the Chisense fishing gear used 86.5% are at least made out of meshless materials like mosquito-nets, shading material, cotton cloth, maize meal bags, etc. In most cases a net made of a composite of these materials including patches of 6mm nets. Only a limited number of nets are completely made of 6mm net material (usually from Nkwazi Net Factory, Kafue). A few nets are made of netting material used for clupeid fishing on Lake Tanganyika (10mm Mesh) (Table 7).

Most fishers in stratum 1 and 2 and some of the fishers in the other strata have gill-nets as well as Chisense gear. The mesh sizes commonly used are 63mm (46%) and 76mm (21%) (Table 8).

Data on migration of fishers during the year have not been analysed yet, but it is known that many fishers move from northern camps to southern areas from April onwards. Reverse migration takes place from January to March (Fig. 3). Data on the tribal origins of the fishers show that the Bwile, Tabwa and Shile mostly operate from within their home areas (Strata 1 and 2) while Bemba/Lunda (home area in Stratum 4) are more scattered over the whole fishing area. Only about 10% of the fishers are from tribes outside the Mweru-Luapula region (Table 9).


4.1 Biology

The present data suggest that, with the possible exception of the unidentified Haplochromine species there are no planktonic feeders in the Chisense complex. However it should be noted that all specimens analysed came from the southern shallow part of the lake and the Luapula River. No specimens from the Northern, deeper part of the lake have been analysed yet.

P. moeruensis is an insectivore, feeding predominantly on aerial insects and DIPTERA larvae. This combination suggests that the latter are possibly fed upon when emerging from the bottom into the water column. N. moeruensis seems to feed exclusively on adult insects.

The lacustrine pelagic species of the complex contains at least three predators of P. moeruensis. These are S. macrocephala, A. macropthalmus and R. intermedius. Of these only S. macrocephala has some commercial importance. Juveniles of this species of about 10mm length have been found in our catches with a 6mm mesh sized net. There is no evidence yet of whether larvae/juveniles of other commercially important species (i.e. Oreochromis macrochir, Tylochromis spp.) are affected by the Chisense fishery.

Juveniles of P. moeruensis (up to about 24mm length) are not regularly caught in our catches, nor are they seen in the catches of the fishers. It seems that juveniles are not found at the usual fishing grounds of the Chisense fishers at about 2 to 5 km from the shore. This observation needs to be confirmed with length frequency data from both inshore and offshore catches. Shoals of small Chisense are often seen in the littoral zone and it is possible that these shoals gradually move into deeper waters as the individuals grow bigger. The effect of the meshless nets, predominantly used by the fishers, on juvenile stocks of both P. moeruensis and other commercially important species needs to be investigated.

At present there is no evidence available on the pattern of horizontal migrations of the species but longitudinal migrations over the lake and the Luapula River are suggested by the movements of the fishers over the lake.

4.2 Frame survey

Frame survey data combined with what is known about daily catches of fishers and our own catches give a rough estimate of a total yield of about 10,000 – 15,000 tonnes per year. Catches of fishers vary greatly from 50 to 1200 kg wet weight per night depending on the time of the year and fishing ground. This fairly conservative estimate is based on 16 days per month of fishing during 8.5 months a year with catch of 60 – 100 kg per night (= 1–2 bags of dried Chisense). There is no knowledge on the effectiveness of the various fishing methods used in the fishery both in terms of yield and cost.

The total yield of the Chisense fishery exceeds that of the traditional demersal fishery even with this conservative estimate. Since the start of the fishery in 1984 with approximately 250 fishers there has been an average increase of 125 fishers per year. Probably the fishery is still growing due to a number of factors. Most Chisense fishers, especially in the Northern camps originally are - and still are during some time of the year - gillnet fishers. The decline in catches of the demersal fishery possibly has made these fishers to diversify and start Chisense fishing. Furthermore there is a general trend all over Lake Mweru to use active fishing methods (several types of seining methods with gillnets, driving of fish into gillnets etc.) because of the decreasing catch of demersal fishes and increasing net thefts.

Chisense fishing requires quite some investments (2 to 4 paraffin pressure lamps, 1 or 2 plank boats, and a net) and organisation (4–7 workers per fishing unit) even with the present makeshift nature of the gear used. Not surprisingly most gear owners are well into their thirties on starting Chisense fishing.

The two main management questions to be answered are;

  1. what are the effects of the meshless materials used for fishing both on the Chisense stocks and other commercial species found in the by-catch, and

  2. can the stocks of Chisense withstand the present and still increasing effort of the fishery?


Aarnink, B.H.M., Kapasa C.K. and van Zwieten P.A.M. 1992. Our Children Will Suffer, present status and problems of Mweru-Luapula Fisheries. Department of Fisheries, Nchelenge, October 1992.

Aarnink, B.H.M., Kapasa C.K. and P.A.M. van Zwieten (in prep). Summary results of Frame Survey conducted between June-August 1992 on Lake Mweru with a description of methods used. Department of Fisheries, Nchelenge.

Balon, E.K. 1971. First catches of L. Tanganyika Clupeids (Kapenta - Limnothrissa miodon) in Lake Kariba. Fish. Res. Bull. Zambia, 5: 175–186.

Bazigos, G.P. 1974. The sampling design of the catch assessment survey. Rome, Italy: FAO; FI:DP/URT/71/012/1.

Bell-Cross, G and Bell-Cross, B. 1971. Introduction of Limnothrissa miodon and Limnocaridina tanganicae from Lake Tanganyika into Lake Kariba. Fish. Res. Bull. Zambia, 5: 207–214.

CLOFFA 1990. Check-list of the freshwater fishes of Africa, Volume 4. Daget, J, Gosse, J.P. & D.F.E. Thys van den Audenaerde (Eds.).

Jackson, P.B.N. 1961. The Fishes of Northern Rhodesia. A check List of Indigenous Species. The Government Printer - Lusaka.

Junor, F.J.R., and Begg, G.W. 1979. A note on the successful introduction of Limnothrissa miodon (Boulenger) “The Lake Tanganyika Sardine” to Lake Kariba. In: Newslett. Limnol. Soc. South Africa 16: pp 8–14.

Kapasa, C.K. 1983. The Sardine (Clupeid - Kapenta) of Zambia. Dip FM Dissertation, Humberside C.H. Education, Grimsby, United Kingdom.

Kapasa, C.K. and Cowx, I. 1991. Post-impoundment changes in the fish fauna of Lake Itezhi-tezhi, Zambia. Journal of Fish Biology, 39: 783–793.

Marshall, B. 1984. Kariba (Zimbabwe/Zambia). CIFA Tech. Pap. Doc. Tech. LPCA 10, pp. 105–153.

Ssentongo, G.W. 1990. A short expose on Lake Mweru exploited fish stocks. State of fisheries problems and solutions. In: Report on the Technical Consultation on Lake Mweru shared by Zaire and Zambia. RAF/87/099 - TO/11/90 M. Maes (Ed.).

Whitehead, P.J. 1985. FAO Species Catalogue: Vol. 7. Clupeoid fishes of the World Part 1: Chirocantridae, Clupeidae, Pristigasteridae, FAO Fisheries Synopsis No. 125, UNDP/FAO, Rome, pp 132–161.


A wide range of fishing gear is used to catch demersal fish species in Lake Mweru. Besides being used as passive gear, gill nets are also used as driftnets (Kasenswa), and for various forms of seining including driving of fish (Kutumpula) into them. All active methods of using gill nets are illegal in Zambia. A description of all illegal methods used can be found in Aarnink, Kapasa and van Zwieten (1992). Other methods of catching demersal fish are longlines, traps, and spears.

The methods of fishing for pelagic clupeids in Lake Tanganyika include Lusenga net (dip net), Chiromila net, beach seine net, purse seine and lift net. Kapasa (1983) presents details on how these nets are operated. In Kariba, some kind of lift nets are used.

The fishing of pelagic Chisense in Lake Mweru-Luapula is mainly an adaptation of the Tanganyika methods but for all the methods, lights are used to attract fish. The following techniques are in operation on Lake Mweru-Luapula:

1. Beach Seine net (Mukwau)

Lights are set out on floats at a distance of 10–100m offshore for about 60 minutes. The lights are then slowly pulled towards the shore. Subsequently, a boat surrounds the light with a net. Once both ends of the net are received on the shore, it is pulled onto the shore in the manner of normal beach seine. The gear requires four men to operate. It is the most common method used on the Luapula River.

This method is not allowed for demersal fish species but has been allowed for Chisense fishing. However the effect of the method on juveniles of other commercial fish species needs to be studied.

2. Boat seine (Chiromila net).

Traditionally the chiromila net was used on Lake Malawi. After some modifications, it was recommended to be used on Lake Tanganyika by the Joint Fisheries Research Organisation. The chiromila net used for Kapenta fishing differs in design from that of Chisense fishing.

For Chisese fishing two boats are required. A number of lights, usually 4 or 5, are set for 40 to 60 minutes after which they are pulled towards the boat and taken out except for one. One boat then surrounds the remaining light until the second boat is reached. The net takes the shape of a quarter sphere except that near the mid-point of the head rope the webbing almost descends vertically. Then pulling commences from one boat, while the second boat slowly moves out of the circle. The net is hauled in by hand. The speed of hauling is very important. To operate the boat seine at least five persons are required.

3. Dip net (Mutobe)

A dip net is operated in the same manner as a Lusenga net. One or more lights are set in the usual manner on floats and are approached by a single boat with a crew of three persons (two paddlers, one net handler). The net is attached to two bamboo poles. It is dipped under the light and moves through the concentration of shoaling fish two or three times. The net is then brought into the boat and emptying is done with the assistance of the second person. The procedure is repeated at short intervals.

4. Outrigger scoop net (Japan net)

This is the most commonly used fishing method for Chisense in the Lake Mweru-Luapula fishery. A bamboo outrigger is attached to a plank boat and a net is dropped vertically under the boat. Lights are set in the usual manner and are drawn after some time towards the outrigger. Once the lights are in between the bamboo frame, the net is quickly raised by pulling the ropes guided over the outer bamboo parallel to the boat. The sides and outer edge of the net are then hauled aboard. The gear is operated by two or three persons.

5. Lift net

Two types of methods of using a lift net are used. Both are not common in the fishery.

a) Lift net using a Catamaran

A catamaran is constructed using two plank boats joined by bamboo poles. A net is lowered horizontally some metres below the catamaran. Lights are set out and drawn in after some time in between the two boats. Once the lights and the accompanying shoal of fish have entered the centre of the catamaran, the net is quickly raised by pulling on all four sides and hauled aboard. The operation requires four men. Since the nets required are expensive (6mm mesh size) the method is not common among fishers.

b) Lift net between unattached boats

The sequence of operations is essentially the same as for the lift net using a catamaran, except that the nets tend to be set in shallow water and are made from meshless materials. This method is not used anymore as most fishers prefer the previously described dip net.

Table 1. Species, number of specimens and total weight (gr) of each species in 4 hauls of a lift net (lake samples) and seine net (river samples) at three different sites: Luapula River near Kashilu, and Lake Mweru inshore and offshore of Isokwe Island.

nr.weight% weightnr.weight% weightnr.weight ht% weight
Poecilothrissa moeruensis865288992093.7671353367782.02341238548278.2
Neobola moeruensis---8611494712.00203391448313.3
Barbus cf. radiatus---4818372.0486480227.3
B. cf. lukindae333033793.5------
B. trimaculatus1110.0------
B. cf. eutaenia150.0------
B. cf. unitaeniatus1170.0------
B. fasciolatus4190.0------
Alestes macropthalmus3011561.29711.0---
A. grandisquamis1800.0------
A. imberi---(8)1(52)----
A. cf. liebrechtsii------ 3 90.0
Raiamas intermedius241460.0524041.0163960.4
Serranochromis macrocephala122070.0103961.0247110.7
S. macrocephala juv.---9180.01281340.1
S. thumbergii130.0------
Tilapia rendalli31430.0------
Tylochromis mylodon4610.0------
Haplochromis(?) sp. “orange---3377181.74130.0
Chrysichthys mabusi9640.0------
Synodontis polystigma4690.0------
Gnathonemus petersii2990.8------
Hippopotamyrus discorhynch6920.8------
Marcusenius macrolepidotus193492.8------
cf. M. macrolepidotus3470.4------

1 Caught in a sample at a later date

Previous Page Top of Page Next Page