Lake Kariba Fisheries Research Institute
P.O. Box 75
The sardine Limnothrissa miodon was introduced to Lake Kariba where it now comprises 80% of the total fish catch. It was stocked in 1967/68 and had colonised the entire lake by 1970, thus demonstrating its capacity for reproduction and dispersal. The smallest fish are found in shallow water and they move out into deeper water as they get older. Breeding apparently occurs in shallow water and appears to take place throughout the year although there are some variable peaks of activity. The fish in Lake Kariba are very small in comparison to those from Lakes Tanganyika and Kivu and mature very much earlier. Their small size is probably caused by their very high mortality rates which are, in turn, a consequence of food shortages.
La sardine d'eau douce Limnothrissa miodon a été introduite dans le lac Kariba où elle représente aujourd'hui 80 pour cent du total des captures de poissons. Introduite en 1967/68, elle avait dès 1970 colonisé tout le lac, ce qui démontre sa grande capacité de reproduction et de dispersion. Les poissons les plus petits se rencontrent dans les eaux peu profondes et se déplacent avec l'âge vers des eaux plus profondes. La reproduction s'effectue apparemment toute l'année, dans les eaux peu profondes, avec cependant des pointes d'activité variables. Les poissons du lac Kariba sont de très petite taille comparés à ceux des lacs Tanganyika et Kivu et parviennent à maturité beaucoup plus tôt. Leur petite taille est probablement due à leur très fort taux de mortalité qui est lui-même provoqué par des pénuries d'aliments.
Lake Kariba was formed when the Zambezi River was dammed in 1958 and the fish population of the new lake grew rapidly as nutrients were released from the drowned vegetation. These fish were most abundant in shallow water down to about 10 m in depth (Coke, 1968; Balon, 1974). Few of them were able to live in the deeper waters which constitute the bulk of the lake and the possibility of introducing a pelagic species was considered. After investigations into the biology of the two clupeids in Lake Tanganyika, Stolothrissa tanganicae and Limnothrissa miodon, it was concluded that the latter would be the most suitable for stocking into Lake Kariba (Matthes, 1965–66a).
Techniques for handling and transporting these fish were developed and they were introduced in 1967 and 1968 (Matthes, 1965–66b; Bell-Cross and Bell-Cross, 1971). The first evidence of their survival came when a number were found in the stomach of a tigerfish, Hydrocynus vittatus, caught in 1968. By May 1970 they were present throughout the lake and had been reported from the stilling pool below the dam wall (Begg, 1974). Those that escaped from Lake Kariba through the hydroelectric turbines were able to survive in the Zambezi River and invade lake Cahora Bassa. This species has also been introduced into Lake Kivu and, recently the man-made Lake Itezshi-Tezshi in Zambia.
The environments in which Limnothrissa now occurs differ considerably from each other and present a challenge for a species adapted for life in Lake Tanganyika. In order to survive it had to modify its life history strategy and that it could do so demonstrates its adaptability. Aspects of its life history in Lake Kariba are considered in detail in this paper and compared with data from elsewhere.
The fact that the sardines occupied the whole lake within two years of their introduction (Begg, 1974) shows that they are capable of rapid dispersal. However, their distribution in the lake is very patchy (Marshall, 1988; Lindem, 1988), possibly because the zooplankton is also patchily distributed (Magadza, 1980).
Vertical migrations occur in Limnothrissa which spends the day in tight shoals in deep water, rising to the surface at dawn and dusk while spending the rest of the night more widely dispersed throughout the pelagic waters (Begg, 1974; Cochrane, 1978; Lindem, 1988). This pattern is similar to those found in Lakes Tanganyika and Cahora Bassa (Coulter, 1960; Gliwicz, 1984) but rather different from that in Lake Kivu where the fish move into the upper layers during the early morning and late afternoon (de longh et al., 1983).
One explanation for these movements is that they follow those of the zooplankton (de longh et al, 1983; Gliwicz, 1984) but in Lake Kariba they do not, in fact, match closely (Begg, 1976). It seems that the fish and the plankton respond to the same stimuli and are therefore found at the same place at the same time of day. This might be an endogenous response to light intensity which provides protection against predators, as in herrings at sea (Balls, 1951).
Two incidents suggest that light may be important in the vertical migration of Limnothrissa. Begg (1974) noted that they came to the surface when the sun was obscured by cloud and in Lake Cahora Bassa Gliwicz (1984) found that they were nearer the surface in parts of the lake shaded by mountains and that over the full moon they remained near the surface rather than dispersing throughout the water column.
Their depth distribution is determined by the depth of the thermocline and the availability of dissolved oxygen (Begg, 1974). The water below the thermocline is usually anoxic and so from November to April the fish are limited to about 20 m depth. The only data on horizontal movements relates to their migration into deep water as they grow larger, with the smallest and youngest fish being found in very shallow water (Fig. 1.). Adult fish probably move into shallow water to breed but it is not known how long they remain there.
Limnothrissa is an omnivorous species, although it feeds predominantly on plankton taking both zoo- and phytoplankton. Begg (1974) found that Bosmina longirostris was their main prey, constituting about 80% of the food items in their stomachs. By contrast, they comprised on average only 26% of the food items in samples collected by Cochrane (1978) while Mesocyclops sp. made up 56% of the items. Cochrane's work showed that the stomach contents of the fish matched the availability of the plankton; for example, they utilised Ceriodaphnia dubia extensively when it was most abundant in the lake. Limnothrissa may have caused a decline in the abundance of some larger zooplankton species, such as Diaptomus, Ceriodaphnia and Diaphanosoma, which were more abundant before 1971 but since no stomach contents were taken at that time it is not certain that it caused them to decline1.
Other food items include insects (e.g. chironomids, ephemeropterans, trichopterans and hemipterans) and in one sample 55% of the food items were of terrestrial origin (Mitchell, 1976). Insects are most likely to be taken at dusk when the sardines come to the surface as fish captured at that time had plankton in the hind gut but insects in the stomach (Begg, 1974).
There is some evidence that they may prey on other fish species since scales, vertebrae and fin rays have been found in their stomachs. In one case, a Pseudocrenilabrus philander was found in the stomach of a 64 mm sardine (Mitchell, 1976). It has been suggested that in Lake Kariba they are cannibalistic, as they are in Lake Kivu. Fish have been found in the stomachs of individuals over 80 mm in length and Limnothrissa can ingest specimens of half their own length; a 57 mm fish was recorded in the stomach of one measuring 125 mm and one of 46 mm in a 93 mm fish. However, Gliwicz (1984) pointed out that the behaviour of fish is abnormal under light attraction and evidence of cannibalism might reflect these abnormalities. He found no evidence of cannibalism in specimens captured without light attraction in Lake Cahora Bassa.
According to Begg (1974) the condition of the fish was lower in Lake Kariba than it was in Lake Tanganyika which he attributed to differences in their productivity. However, pelagic primary productivity in the two lakes is similar (Hecky and Fee, 1981; Machena, 1983) so this may not in fact be the explanation. Their condition was lowest after the breeding season (May-June) and in October-December when the availability of food was reduced after the population had reached its maximum abundance in July-August. The condition of the fish was reflected by changes in their lipid composition (Cochrane, 1978, 1984).
The relationship between the abundance of fish and zooplankton is still unclear. There is evidence that poor river flows lead to a reduced sardine biomass (Marshall, 1982), presumably because the nutrient supply to the lake is reduced. Limnothrissa might, under these circumstances, feed more frequently in shallow water. This aspect requires further investigation.
1 Editor's note. Details of the effects of Limnothrissa on the zooplankton of Lake Kariba can be found in Marshall (1991). Changes in its composition are entirely consistent with size-selective predation.
Limnothrissa in Lake Kariba are smaller than they are in Lakes Tanganyika and Kivu and fewer than 1∙0% of the 130 000 specimens processed by Marshall (1987) were more than 70 mm in length. In 1987 and 1988 the average length of fish in samples collected every fortnight was 57∙1 and 56∙1 mm respectively.
The growth parameters of the Lake Kariba fish obtained by three authors vary considerably (Table 1), although the resultant growth curves are similar (Fig. 2). Limnothrissa in the lake have the potential to grow to a large size and large fish occur rarely; in one exceptional case a fishing vessel caught 400 kg of fish measuring 110–120 mm (unpublished observation). The low estimates of L∞ obtained by Cochrane (1984) and Marshall (1987) are therefore unrealistic and are probably a consequence of the method used to determine their age (length-frequency analysis). By contrast, the higher value given by Chifamba (1992) was obtained by counting growth rings on the otoliths which makes it possible to examine large specimens.
Why do so few fish grow to a larger size when they obviously have the capacity to do so? High mortality may be one explanation as Cochrane (1978) suggested that there was catastrophic mortality from starvation in September-October when the availability of food was lowest. Since dead fish are seldom found predation might also contribute to mortality.
The lack of larger prey might also affect their size. Cochrane (1978) suggested that the absence of atyid shrimps, which Limnothrissa feeds on in Lake Tanganyika (Matthes, 1965–66a), was the reason why the fish were small. This problem had been anticipated and shrimps (Limnocaradina sp) from Lake Tanganyika were introduced at the same time as Limnothrissa (Bell-Cross and Bell-Cross, 1971)2.
The small size of Limnothrissa in Lake Kariba could be a response to an unstable environment (Marshall, 1987). Their biology is typical of an r-selected species which will be small and rapidly-growing in an unstable environment. Instability in lake Kariba is brought about by the high level of predation by tigerfish and its low water retention time (3 years compared to 1000 years in Lake Tanganyika). Limnothrissa is also very small in Lake Cahora Bassa which has a water retention time of about 9 months (Gliwicz, 1984)3.
The total mortality rate of Limnothrissa in Lake Kariba ranged from 0.816 month-1 in 1978 to 1∙149 in 1983 (Marshall, 1987) compared to Lake Tanganyika it varied between 0∙37 and 0∙81 (Moreau et al., 1991). The rates of natural mortality, estimated by Pauly's empirical method, were 0∙44 and 0∙30 month-1 in Lakes Kariba and Tanganyika respectively. The relationship between fishing effort and mortality in Lake Kariba gave a much higher estimate of 0∙731 month-1 (Marshall, 1987). There also appeared to be different mortality rates in two areas of the lake but the cause of these differences is unclear at present.
2 Editor's note. The fate of these shrimps is unknown; it is said that some survive in the lake but this has not been documented.
3 Editor's note. The small size of Limnothrissa in the man-made lakes has been reviewed in a paper that appeared after the symposium was held (Marshall, 1993).
Limnothrissa has displayed a high degree of adaptability in relation to its age of maturity. In Lake Kariba mature males and females of 48 mm have been recorded compared to 61 and 78 mm in Lake Tanganyika and 64 and 75 mm in Lake Kivu (Begg, 1974; Ellis, 1971; Spliethoff et al., 1983). Most of the population in the open waters of Lake Kariba is from 50–60 mm in length and therefore already mature which reduces the danger that the stock will be unable to breed before they being caught.
According to Begg (1974) breeding takes place throughout the year since larvae, juveniles and ripe adults were found throughout the year. Recent work has confirmed the presence of larvae throughout the year (M.Z. Mtsambiwa, personal communication). Begg identified two breeding peaks that coincide with periods of increased nutrient availability; the first after turnover (June-August) and the second when the tributary rivers are flowing (January-March). Cochrane (1978) suggested that most breeding activity took place between September and March, as it does in Lake Tanganyika (Ellis, 1971). The differences between the findings of Begg and Cochrane might be caused by year-to-year variations in the breeding cohort.
In Lake Kariba few ripe females were found in the pelagic waters which indicated that they do not breed there. The high incidence of females in the inshore waters suggests that breeding took place there and sandy areas with gently shelving bottoms seemed to be preferred (Cochrane, 1978). The eggs have never been collected so exactly where they are laid is uncertain. Cochrane speculated that they were laid between 10–40m along steep rocky shores but there is no evidence to support this view. The fact that fish less than 2∙5 mm long occur in water with an average depth of 1∙5 m (Mitchell, 1976) suggests that the eggs are actually laid in very shallow water. The fry appear to prefer clear water with a rocky or sandy bottom and also occur in areas with steep shores (Cochrane, 1978).
Their fecundity increases with size; a fish 46 mm had 600 eggs whilst one measuring 114 cm had 14 044 eggs (Begg, 1974). This is lower than in Lake Tanganyika where a fish of 140 mm had 55 000 eggs (Matthes, 1965–66a). They breed from about September to February and enter the fishery at about 40 mm in length when they are 3–4 months old. The bulk of those caught in the fishery are 40–50 mm long and about 5 months old. The population will generally increase steadily from February to August and then fall because of high mortality and decreased recruitment.
Because the sardine fry live in shallow water they might compete for food with the fry of many other species, including H. vittatus, Brycinus lateralis, B. imberi and Micralestes acutidens. It was thought that B. lateralis would compete with Limnothrissa in deep water (Balon, 1974) but this not the case. B. lateralis only occurs in any numbers close to the lake's margin since it seems to rely on the cover provided by aquatic vegetation.
The most important predator on Limnothrissa is the tigerfish, whose population increased following the sardine introduction. Their diet changed and by 1971 as much as 70% of the food eaten by tigerfish consisted of Limnothrissa (Kenmuir, 1971). Tigerfish predation appeared to be most intense at dawn and dusk and they were seen feeding on sardines at these times (Begg, 1974). Because the tigerfish had moved out into open water to feed on Limnothrissa it was felt that it would become a valuable by catch in the fishery but this hope was never realised (Marshall et al., 1982).
Other fish that prey upon Limnothrissa include Tilapia rendalli, Synodontis zambezensis and Schilbe intermedius. More of the latter are caught in the upper reaches of the lake (unpublished data) which suggests that they have a greater impact in that area. Some birds also feed on Limnothrissa, notably the White-winged Black Tern, Chlidonias leucoptera and the Pied Kingfisher Ceryle rudis (Begg, 1973; Junor, 1972).
Seasonal abundance, reflected in the catches, changes in relation to food availability. The fish are generally most numerous in August-September although there is an earlier peak in April-May. there appears to be a relationship between biomass and fishing effort, as the biomass was 10% higher in 1974, when fishing began, than it was in 1985 (Marshall, 1988).
I am most grateful to my colleagues at the Lake Kariba Fisheries Research Institute for their contributions to the preparation of this paper. Special thanks are due to Dr Cecil Machena whose suggestions were incorporated in the writing up of this paper.
Balls, R., 1951. Environmental changes in herring behaviour. A theory of light avoidance, as suggested by echosounding observations in the North Sea. J. Conseil. Int. Explor. Mer, 17: 274–94.
Balon, E.K., 1974. Fish production of a tropical ecosystem. In E.K. Balon and A.G. Coche (eds), Lake Kariba: a man-made tropical ecosystem in central Africa, pp. 249–573. The Hague; Dr W Junk.
Begg, G.W, 1973. The feeding habits of the Whitewinged Black Tern on Lake Kariba. Ostrich, 44: 149–53.
Begg, G.W, 1974. Investigations into the biology and status of the Lake Tanganyika sardine, Limnothrissa miodon (Boulenger), in Lake Kariba, Rhodesia. Lake Kariba Fish. Res. Inst., Proj. Rept. 17: 151pp.
Begg, G.W, 1976. The relationship between the diurnal movements of some of the zooplankton and the sardine Limnothrissa miodon in Lake Kariba, Rhodesia. Limnol. Oceanogr., 21: 529–39.
Bell-Cross, B. and G. Bell-Cross, 1971. Introduction of Limnothrissa miodon and Limnocaradina tanganicae from Lake Tanganyika into lake Kariba. Fish. Res. Bull. Zambia, 5: 207–14.
Chifamba, P.C., 1992. Daily rings on otoliths as a method for ageing the sardine Limnothrissa miodon in Lake Kariba. Trans. Zimbabwe Sci. Ass., 66 (in press).
Cochrane, K.L., 1978. Seasonal fluctuations in the catches of Limnothrissa miodon (Boulenger) in Lake Kariba. Lake Kariba Fish. Res. Inst., Proj. Rept. 29: 163 pp.
Cochrane, K.L., 1984. The influence of food availability, breeding seasons and growth on commercial catches of Limnothrissa miodon (Boulenger) in Lake Kariba. J. Fish Biol. 24: 623–35.
Coke, M., 1968. Depth distribution of fish on a bush-cleared area of Lake Kariba, central Africa. Trans. Am. Fish. Soc., 97: 460–5.
Coulter, G.W., 1960. Research results - Lake Tanganyika. Ann. Rept. Joint Fish. Res. Org. 10: 7–30. Lusaka; Government Printer.
de longh, H.H., P.C. Spliethoff and V. Frank, 1983. Feeding habits of the clupeid Limnothrissa miodon (Boulenger) in lake Kivu. Hydrobiologia, 102: 113–22.
Ellis, C.M.A., 1971. The size at maturity and breeding seasons of sardines in southern Lake Tanganyika. Afr. J. Trop. Hydrobiol. Fish., 1: 59–66.
Gliwicz, Z.M., 1984. Limnological study of Cahora Bassa reservoir with special regard to sardine fishery expansion. FAO/GCP/MOZ/006/SWE Field Doc. 8: 71pp.
Hecky, R.E. and E.J. Fee, 1981. Primary production and rates of algal growth in Lake Tanganyika. Limnol. Oceanogr., 26:532–47.
Junor, F.J.R., 1972. Offshore fishing by the Pied Kingfisher Ceryle rudis at Lake Kariba. Ostrich, 43: 185.
Kenmuir, D.H.S., 1971. Some aspects of Hydrocynus vittatus Castelnau (tigerfish) research at Lake Kariba. Newslett. Limnol. Soc. sth. Afr. 17:33–8.
Lindem, T., 1988. Results from the hydroacoustic survey, Lake Kariba, September, 1988. Report to the SADCC Zambia/Zimbabwe fisheries project. Machena, C., 1983. A study of the primary productivity of the Sanyati Basin of Lake Kariba and the application of a regression model to estimate fish yields in the lake. M.Sc. thesis, University of Zimbabwe.
Magadza, C.H.D., 1980. The distribution of zooplankton in the Sanyati bay, Lake Kariba: a multivariate analysis. Hydrobiologia, 70: 57–67.
Marshall, B.E., 1982. The influence of river flow on pelagic sardine catches in Lake Kariba. J. Fish Biol., 20: 465–69.
Marshall, B.E., 1987. Growth and mortality of the introduced Lake Tanganyika clupeid, Limnothrissa miodon, in Lake Kariba. J. Fish Biol., 31: 603–15.
Marshall, B.E., 1988. A preliminary assessment of the biomass of the pelagic sardine, Limnothrissa miodon, in Lake Kariba. J. Fish Biol., 32: 515–24.
Marshall, B.E., 1991. The impact of the introduced sardine Limnothrissa miodon on the ecology of Lake Kariba. Biol. Conserv., 55: 151–65.
Marshall, B.E., 1993. The biology of the African clupeid Limnothrissa miodon with special reference to its small size in artificial lakes. Rev. Fish Biol. Fish., 3 (in press).
Marshall, B.E., F.J.R. Junor and J.D. Langerman, 1982. Fisheries and fish production on the Zimbabwean side of Lake Kariba. Kariba Studies 10: 175–231.
Matthes, H., 1965–66a. Preliminary investigations into the biology of the Lake Tanganyika Clupeidae. Fish. Res. Bull. Zambia, 4: 39–48.
Matthes, H., 1965–66a. Report on the catching, handling, keeping and transportation of Lake Tanganyika clupeids. Fish. Res. Bull. Zambia, 4: 111–4.
Mitchell, S.A., 1976. The marginal fish fauna of Lake Kariba. Kariba Studies 8: 109–62.
Moreau, J., J. Munyandorero and B. Nyakageni, 1991. Evaluation des paramètres démographiques chez Stolothrissa tanganicae et Limnothrissa miodon du lac Tanganyika. Verh. Internat. Verein. Limnol., 24: 2552–8.
Mtsambiwa, M.Z., 1989. The age structure of pre-recruits of the sardine Limnothrissa miodon in Lake Kariba based on daily increments in otoliths. M.Sc. thesis, University of Wales.
Spliethoff, P.C., H.H. de longh and V. Frank, 1983. Success of the introduction of the freshwater clupeid Limnothrissa miodon (Boulenger) in Lake Kivu. Fish. Mgmt., 14: 17–31.
Table 1. The von Bertalanffy growth parameters of Limnothrissa miodon in Lakes Kariba, Tanganyika and Kivu.
|Tanganyika: Moreau et al. (1991)||153.2||0.104|
|Kivu: Spliethoff et al.(1983)||145.0||0.100|
Figure 1. The mean length (circles) and age (squares), with standard errors, of Limnothrissa miodon collected at various depths in Lake Kariba. Based on data in Cochrane (1978) and Mtsambiwa (1989).
Figure 2. Growth curves of Limnothrissa miodon in Lake Kariba, based on the data given in Table 1.