Following preliminary observations in 1975 the Lake Kariba Fisheries Research Institute implemented a research program in March 1976 to investigate various aspects of the biology and population dynamics of the mussel crop of Lake Kariba, with a view to greater understanding of their status in the lake, and establishing a basis for management, if required. Aspects investigated were densities, distribution, species composition, predation by fish, reproduction, age and growth, gross biochemical composition and crude production rates. The mussel community comprised four species, a small pill clam, Corbicula africana, and three mussels, Caelatura mossarnbicensis, Aspatharia wahlbergi and Mutela dubia. Only the latter three were studied. Mussel beds occupied all the gently shelving cleared and uncleared areas, and the bulk of the populations occurred from 3 m - 9 m depth, but extending to 11 m. Predation by fish was found to be extremely low. The reproduction study showed two species bred all year C. mossambicensis and M. dubia, whereas A. wahlbergi bred seasonally during the rainy season. The two year-round breeders were repetitive spawners, capable of breeding several times a year. Indications are that A. wahlbergi is essentially a fluviatile species, with a life cycle adapted to taking advantage of riverine conditions. Sex ratios in all three species were approximately 1 : 1 , with females predominating slightly. Fecundity in one species, C. mossambicensis, was investigated. The complete parasitic cycle of M. dubia was elucidated and found to be similar to that of M. bourguignati, as described by Fryer (1961). Host species were noted amongst the cichlid and mormyrid families, and the conclusion drawn that the species is not host specific. various aspects of the life-cycles of the other two species were noted, but complete life-cycles were not successfully elucidated. population composition according to age showed the most common species, c. mossambicensis, to be comprised mainly of 2 - 5 year-old individuals, indicating a young and vigorous population. Juvenile pre-adult mussels were very scarce. In the population of A. wahlbergi, older mussels formed a greater proportion of the biomass than in the population of c. mossambicensis, and young were also scarce. M. dubia were not recorded in sufficiently large numbers to estimate age composition accurately. The production rate (whole wet mass) of the most common species, c. mossambicensis, was calculated to be 2,45 kg ha¯¹ day¯¹ in the sanyati East cleared area (30,5 km²), while the overall production rate of all three species was calculated to be 3,34 kg ha¯¹ day¯¹. Calorific values of all three species were obtained, while gross body composition in terms of water, fat, protein, ash and amino acid composition were determined. The ash component of A. wahlbergi was analysed. The Lake Kariba results were supplemented by a follow-up study of the composition of the mussel population of a eutrophic, highveld dam, Lake McIlwaine, in 1978/1979. This much smaller lake was populated by only two mussel species, M. dubia and C. mossambicensis, although A. wahlbergi and Unio caffer were recorded in the riverine upper reaches. Observations on seasonal breeding of the two lake species were made. A detailed study on the draw down zone of a gently shelving beach showed that at the time the extreme shallows had been dominated by C. mossambicensis, whereas from a depth of approximately 1,6 m M. dubia dominated very significantly. Fulling lake level was found to trigger off migratory responses, thus placing a proportion of individuals in deeper water and improving chances of survival. The mussel composition of Lake Kariba and Lake McIlwaine was considered in conjunction with preliminary observations of mussel distribution a nd composition in other water bodies, and some inferences drawn regarding the ecological factors which appear to influence the composition and diversity of the mussel fauna in rivers and lakes. / Thesis (Ph.D.)-University of Natal, Pietermaritzburg, 1980.