Growth of the clupeid fishes, Stolothrissa tanganicae and Limnothrissa miodon, in the Zambian waters of Lake Tanganyika

The age and growth of two endemic Lake Tanganyikan clupeid fishes, Stolothrissa tanganicae and Limnothrissa miodon , were estimated from analysis of otolith microstructure in specimens collected in Zambian waters from October to December 1990. Both species had relatively clear, concentric otolith increments which were believed to represent daily increments based on their similarity to daily increments found in some marine clupeid species. The growth rates estimated from length-at-age data were: for S. tanganicae, L_t = 104·0 (1 – exp (– 0.00512 ( t – 5·8))); for L. miodon, L_t = 155.4 (1 – exp (– 0.00236 ( t +8.1))). In most cases, the growth parameters obtained from otolith analyses generally agreed with those previously estimated from analyses of length-frequency distributions.     

Genetic population structure in the Lake Tanganyika sardine Limnothrissa miodon

There was significant morphological differentiation among samples of Tanganyika sardine Limnothrissa miodon , indicating non-random association of fish. Individuals within schools appeared to be unrelated as high mtDNA haplotype diversity demonstrated the presence of many maternal lineages in each school. Nevertheless, there was evidence from allozyme analysis for genetic differentiation on a very small geographical scale, possibly even among individual schools, without any clear geographical boundaries among populations. Similar microgeographical differentiation at the allozyme level was found in several marine pelagic species, suggesting a general pattern of random genetic structure in pelagic schooling fish. Such genetic patchiness may arise from biased reproductive success in localized spawning events and long-term stability of schools, resulting in genetic differentiation among schools. While the present results are preliminary, the far-reaching implications both for fisheries management and for evolutionary processes in pelagic species warrant further research on microgeographical genetic patterns using more sensitive markers.     

Feeding habits of the clupeid Limnothrissa miodon (Boulenger), in Lake Kivu

A limnological sampling programme implemented during 1980 in Lake Kivu, Rwanda, confirmed a seasonal increase of the pelagic plankton biomass at the end of the dry season during the windy period of August/September. From plankton samples taken in the littoral and pelagic zones of the Lake it was concluded that a successful introduction of the freshwater clupeid Limnothrissa miodon Boulenger, during 1958–1960, had not obviously influenced the species composition of the Copepoda in the Lake. Examination of stomach contents of different length classes of Limnothrissa proved that the Kivu clupeid is not a strict planktonphagic species. Its diet is adapted to its life cycle and inshore/offshore migration patterns. In pelagic waters of the lake it is an exclusive plankton feeder, while in littoral waters it has more littoral bound feeding habits. Cannibalism was observed among individuals larger than 100 mm, near the margins of the lake. Echosoundings and periodic observations on stomach contents support the hypothesis that Limnothrissa feeds in the late afternoon and digests its food during the night, while a second feeding/digestion cycle starts in the early morning, at sunrise.     

Seasonal and annual variations in the abundance of the clupeid Limnothrissa miodon in Lake Kivu

There was no clear seasonal cycle in the abundance of the sardine Limnorhrissa miodon in Lake Kivu hut average values suggest that the fish are most numerous in the south from October-April and in the north from April4ctoher. The possibility that there may be a migration from south to north is considered. The catch per unit effort appears to be independent of the present level of fishing effort but may be determined by the annual rainfall two years earlier. In this respect, Lake Kivu is similar to Lake Kariba where nutrient supply is determined by river flow.     

A preliminary assessment of the biomass of the pelagic sardine Limnothrissa miodon in Lake Kariba

The sardine Limnothrissa miodon was very patchily distributed in Lake Kariba in 1981–1983 and its mean biomass was 59 23 kg ha⁻¹, declining from 90–91 kg ha⁻¹ in 1981 to 38–66 kg ha⁻¹ in 1983. In addition, the mean size of the fish fell from 55–2 mm in 1981 to 49–9 mm in 1983. These changes are attributed to the effects of drought and reduced river flows as well as to the high level of commercial fishing for this species. There was a correlation between commercial catches and biomass estimates which allowed estimation of biomass in previous years from fishing statistics. Fishing effort appeared to be a major influence on sardine abundance: total biomass in 1985 was about 10% of what it was in 1974 when the fishery began. When expressed in terms of unit area, as is usual for African lakes, the pelagic biomass appeared to be greater in Lake Tanganyika, where Limnothrissa is endemic, than in Lake Kariba, although there is little difference in their primary productivity, but consideration in terms of volume rather than area shows that this was not the     

Morphological and genetic differentiation of the African clupeid Limnothrissa miodon 34 years after its introduction to Lake Kivu

The evolutionary consequences of an artificial introduction of the clupeid Limnothrissa miodon from Lake Tanganyika into Lake Kivu, East Africa were examined. In 1959, 57 400 fry (mixture of Limnothrissa and the related clupeid, Stolothrissa tanganicae ), were released into Lake Kivu to boost fisheries production. Comparisons were made between respective source and transplant populations 34 years later (1993) using morphometrics ('truss' method), allozymes (29 enzyme-coding loci) and mitochondrial (mt) DNA variation (RFLP analysis of PCR-amplified ND5/6 genes). Significant morphological and genetic differentiation between source and transplant samples was detected, with a distinct clustering of Kivu Limnothrissa on respective dendrograms, especially at the morphometric and mtDNA levels. Differentiation within Lake Tanganyika was, however, consistently higher than that between lakes. Allozymic diversity was similar in samples from both lakes (Lake Tanganyika: heterozygosity = 0.0658, mean number of alleles=1.44; Lake Kivu: heterozygosity = 0.0655; mean number of alleles = 1.48), however, a significantly lower mtDNA haplotype diversity was detected in Lake Kivu (Lake Tanganyika: 0.905; Lake Kivu: 0.755). Data suggest that high post-introduction mortality and various demographic factors reduced the effective population size of the introduced population to tens rather than thousands of individuals, resulting in a reduction in genetic diversity and founder effect.     

Biology of the African clupeid Limnothrissa miodon with reference to its small size in artificial lakes

Summary Limnothrissa miodon is a clupeid from Lake Tanganyika which has been introduced to Lakes Kivu and Kariba and which invaded Lake Cahora Bassa. These lakes differ considerably from one another but the biology of Limnothrissa is similar in many respects in all of them. Similarities include its feeding and breeding biology, whilst its populations fluctuate, on both an annual and a seasonal basis, in relation to environmental changes. The major differences between the lakes concern the size to which Limnothrissa grows. Their average length in the two natural lakes, Lakes Tanganyika and Kivu, is about twice the average length in the two artificial lakes, Lakes Kariba and Cahora Bassa. This in turn affects their age of maturity and they breed during their first year in the reservoirs but during their second in the natural lakes. Their growth rates and performance are comparable up to about 6 months of age and the difference in their size seems to be due to high mortality in the artificial lakes where few Limnothrissa survive for more than a year. Possible explanations for their small size include the genetic composition of the original introduction and the unpredictability of the environment. Neither of these explanations is supported by evidence at present. Deficiencies in the available food and the effects of intense predation are probably the most likely causes. In both natural lakes, Limnothrissa are predatory or cannibalistic from 100 mm in length and 12 months in age, and fish in their diet may enable them to survive into their second year. The fish in Lakes Kariba and Cahora Bassa, in contrast, are neither predatory nor cannibalistic to any degree. There are no predatory fish species in Lake Kivu but there is a diverse community of pelagic piscivores in Lake Tanganyika which tends to contradict the predation hypothesis. However, the position of Limnothrissa in the reservoirs probably resembles that occupied in Lake Tanganyika by Stolothrissa which is also a small, annual species. The significance of this phenomenon is that Limnothrissa can maintain a high biomass and productivity in the face of intense predation. This trait may be of importance to fisheries management because it means that their yield can be greatly increased.     

The influence of food availability, breeding seasons and growth rate on commercial catches of Limnothrissa miodon (Boulenger) in Lake Kariba

Commercial catches of Limnothrissa miodon in Lake Kariba show clear seasonal fluctuations. Recruitment to the pelagic fishing areas was investigated. A von Bertalanffy growth curve for the species was calculated and the breeding season determined from the calculated hatching dates of observed alevins and juveniles. Distribution of the different size classes was determined by a horizontal transect. Analysis of body composition showed a decrease in total lipids at the time when catches were dropping. This was shown to be due to a decline in triglyceride fat reserves apparently resulting from food stress and possibly resulting in increased natural mortality. Recruitment to the fishing areas and the total mortality rate are believed to cause the seasonal fluctuations in catch. The annual cycle of production may be caused by the summer inflow of nutrients from the rivers. Growth is apparently stunted in comparison to the species' growth rate in Lake Tanganyika, probably due to the lower nutrient status of Lake Kariba.     

Growth,mortality and reproduction of Sargochromis codringtonii in Lake Kariba,Zimbabwe

Age determination of Sargochromis codringtonii was undertaken using scalimetry. Marginal scale analysis, used for age validation, indicated that annulus formation occurred in October and November. The possible reasons for this are discussed. S. codringtonii is relatively long-lived, surviving up to 8 years. It is sexually dimorphic with males growing at a faster rate and attaining a larger size than females. It is suggested that the mouth brooding habits of the female fish is one possible reason for sexual dimorphism. A higher total mortality rate (Z) was recorded for females as compared to males. S. codringtonii reproduces throughout the year, but with a peak during the hot wet months. The right ovary is atrophied and the possible significance of this is discussed.     

Seasonal changes in fat and fatty acid composition of Limnothrissa miodon Boulenger (Pisces: Clupeiformes) from Lake Tanganyika,with reference to their nutritive value

The fat content and fatty acid composition of whole sun-dried Limnothrissa miodon Boulenger caught during 1973 have been measured. The fat content seems to increase during the dry seasons. The fatty acids however showed no defined seasonal variations.     

Mercury in tigerfish (Hydrocynus vittatus, Castelnau), green happy (Sargochromis codringtonii, Boulenger) and kapenta (Limnothrissa miodon, Boulenger) from Lake Kariba, Zimbabwe

Concentrations of total mercury were determined in Hydrocynus vittatus (Castelnau), Sargochromis codringtonii (Boulenger), and Limnothrissa miodon (Boulenger) from two localities in Lake Kariba, Zimbabwe.
The mean concentrations of total mercury in H. vittatus from Basin 5 and Basin 2 were 0.08 mg kg⁻¹ and 0.094 mg kg⁻¹, respectively. In S. codringtonii, mean concentrations were 0.004 mg kg⁻¹ and 0.026 mg kg⁻¹ for Basins 5 and 2, respectively. No mercury was detected in L. miodon from Basin 2 while samples from Basin 5 had a mean concentration of 0.069 mg kg⁻¹ (wet weight). Total mercury concentrations were also determined on a dry weight basis.
Within each sampling area, total mercury concentrations were significantly different among species ( P  < 0.05). For H. vittatus and S. codringtonii, total mercury concentrations (in the same species) were not significantly different between the two localities ( P  < 0.05).
The factors causing the observed differences in total mercury between similar species from different localities and among different species in the same locality (sampling area) are discussed. From the observed low levels of mercury in all three species, it was concluded that the mercury constituted 'background levels'. These levels are below the maximum concentrations permissible in human fish foods.     

Validation of daily increment deposition in otoliths of juvenile Limnothrissa miodon (Clupeidae)

Daily deposition of otolith increments was validated for juvenile Limnothrissa miodon in the Cahora Bassa reservoir, Mozambique, by the means of two successive chemical immersion markings.     

Fish production in Lake Kariba,reconsidered

Synopsis Revised total and available production, yield and mean biomass per ha were calculated for each species in Lake Kariba, and for the whole lake. The revision was undertaken because (1) the original value 1 g for W₀ for each species was too high, (2) B_i+1 was occasionally used instead of B_i in the calculation of mean biomass for an interval i to i + 1, and (3) species' contributions to mean parameter values for the whole lake were not weighted according to their mean standing crop. Revised values are, A = 1224, P = 720, Y_A = 400, Y_{Pp, = 202 and B} = 827 kg ha^–1y–1. These correspond to 38.2%, 66.7%, 107.1% and 25.8% respectively of the values calculated initially.     

Temperature, selective mortality and early growth in the short-lived clupeid Spratelloides gracilis

Six cohorts of the silver-stripe round herring Spratelloides gracilis , a fast-growing and short-lived tropical clupeid, were collected as juveniles and then as adults during austral summers from November to February in 1998–1999 and 1999–2000, using light traps in the Dampier Archipelago, Western Australia. Otolith analysis allowed backcalculation of size and growth rate at age to examine the relative influences of selective mortality and water temperature on early growth. Negative size-selective mortality and growth-selective mortality between the juvenile and the adult stages was found only in the cohort that was the smallest and slowest growing in the period immediately following hatching. Selective mortality preferentially removed members of this cohort that were smaller from age 0 to 15 days, and slower growing from 0 to 10 days, resulting in an elevation of size at age to, or even above, that of cohorts that had not undergone this process. Size and growth rate at 5 day age intervals in the first 20 days after hatching differed among cohorts within and between summers and were strongly and positively correlated ( r ²= 0·61–0·83) with water temperature.     

Comparison of fish yields from prediction models between Lakes Tanganyika and Kariba

Fish yield from catch data and from prediction models in Lake Kariba is compared to that in Lake Tanganyika. Fish yields from catch data in L. Tanganyika are above expectation from the rates of algal production, and this disparity is not experienced in L. Kariba. Of the two hypotheses that have been forwarded to explain the situation in L. Tanganyika — high trophic efficiency and high bacterial production, it is suggested that, with the experience from L. Kariba, only the hypothesis of high bacteria production could be important.This paper is published with the permission of the Director, National Parks and WildLife Management, Zimbabwe.     

Diatom associations and succession in Lake Kariba,South Central Africa

Diatom Associations formed by 1% or more of species, found by the Thomasson analysis method are discussed in relation to the morphology and limnology of the four basins of Lake Kariba. A brief comparison is made with associations in two stretches of the Zambesi river. A dominant association formed by certain species of Melosira demonstrated a pattern of succession, in relation to the thermal cycle in the lower, middle and upper lake by variations of the percentage occurrence of species forming it: Melosira granulata dominated during homothermy and remained so throughout stratification in the lacustrine conditions at the barrier or wall; M. granulata v. angustissima dominated during overturn and became the dominant species in the riverine conditions of the two uppermost basins. Results from the Central basin illustrated the intermediate position of this area of the lake. The pattern was disrupted in the far N.E sub-basin of the Sanyati basin, which in general showed individuality and where accidental and ambivalent plankters became co-dominant with the Melosira association during homothermy.A second association, formed by species of Cyclotella, Fragilaria and Synedra was a regular feature of the diatom population, it was sub-dominant to the melosira association at the barrier but gradually decreased up-lake and was superseded by accidental and ambivalent planktonic species. Except for the N/E Sanyati sub-basin, the diatom associations showed a gradual change from upper to lower lake, thus basin individuality was not seen.     

Diets of large mammals in the woodlands around Lake Kariba,Rhodesia

Summary Large mammals that formerly occupied the Kariba basin of the Middle Zambezi Valley have, since it was flooded, been forced to live in a restricted range of deciduous woodland habitats. Feeding records in a study area on the shore of Lake Kariba showed that most herbivore species were browsers, and that only the tree components of their diets varied significantly between vegetation types. The common species varied a proportion of their diet seasonally. However, each depended upon a small range of food staples which differed from those of other species and acted as food refuges for part of the year. Diets overlapped during the wet season because of diversification, and in the late dry season because of common use of a restricted range of plants remaining green. There was a good correlation between the ability of a species to avoid dietary overlap and its biomass in the study area. The chance of interspecific competition occurring increased in the late dry season when most of the species would formerly have migrated from the study area to the flood plain. Despite the enforced occupation of only part of their former annual range the more common herbivores maintained a considerable degree of ecological separation through utilisation of different foods.     

Crustacean parasites of fishes in Lake Kariba,Zimbabwe, preliminary results

During a survey of fish ectoparasites in Lake Kariba (Zimbabwe), two crustaceans were often encountered. Dolops ranarum (Branchiura), parasite of the body, mouth and gill chamber, infected mostly cichlids, the catfish Clarias gariepinus, Synodontis zambezensis and two Mormyrids. Among cichlids clear host preferences were shown. No seasonal variations could be detected in the infections. D. ranarum parasitized predominantly fish above 15 cm in standard length. Lamproglena monodi (Copepoda) parasitized only the gills of cichlids with a preference for Serranochromis codringtonii, S. macrocephalus and Tilapia rendalli. Small fish were seldom infected. Seasonal variations of the prevalence occurred, the hot season being the less favourable. Few other crustaceans were found during the survey: a single specimen of Lamproglena hemprichii on the gills of the tigerfish, Hydrocynus vittatus; three specimens of Afrolernaea longicollis, one found on the gills of Hippopotamyrus discorhynchus and two on the gills of Mormyrops deliciosus; a hundred of Ergasilus mirabilis in a specimen of H. discorhynchus.     

The feeding habits of Sargochromis Codringtonii in Lake Kariba, Zimbabwe

The food of adult Sargochromis codringtonii in Lake Kariba was dominated by prosobranch snails with pulmonate snails forming an insignificant dietary component. S. codringtonii does not show particular preference for any snail species, feeding on those species that are readily available. Morphometrical analysis of the feeding structures of S. codringtonii showed that the fish is well adapted to handling a molluscivorous diet. The pharyngeal jaws with their flat crowned teeth are well adapted to crushing molluscs. The daily food consumption rate of S. codringtonii was estimated over different seasons. Estimates of food consumption indicate that the fish consumed 13.96% of their dry body weight per day in summer falling to 4.46% in winter. The low consumption of pulmonate snails suggests that S. codringtonii can not be a successful biological control agent against the vectors of schistosomiasis.     

The inshore fish yield potential of Lake Kariba, Zimbabwe

Four surplus yield models were used to estimate the yield potential of the Kariba inshore fishery. The classical methods gave maximum sustainable yield (MSY) values as high as 1333 t y-¹. These values are regarded as over-estimates in view of the present average annual catch of 619.4 t y-¹. The Schnute (1978) model gave an MSY of 719t y-¹ and a failure index close to 1 indicating that the data do not fit well into the model. Walter's (1986) graphical approach of fitting an approximate equilibrium yield curve to catch and effort predicts an MSY between 550 and 570t y-¹. It is suggested that the Kariba inshore fishery on the Zimbabwean side should be managed within this range, in Basins 4 and 5.

Résumé

Quatre modèles de rendement excédentaire ont été employés pour estimer le rendement potentiel de la pêcherie dans le lac Kariba. Les méthodes classiques donnaient un rendement maximal supportable (RMS) de 1333 ta-¹. On considere que ces valeurs sont surestimées au vu de la capture annuelle moyenne de 619 ta^-1. Le modèle de Schnute (1978) donnait un RMS de 719 ta^-1 et un risque d'erreur proche de 1, indiquant que les données ne s'intègrent pas bien dans le modèle. L'approche graphique de Walter (1986) qui tend à ajuster un niveau de rendement en équilibre approximatif entre I'effort et la prise prévoit un RMS compris entre 550 et 570 ta^-1. On suggère que la pêcherie dans le lac Kariba devrait être gérée dans cet ordre-là du côté zimbabwéken.