The association and distribution of fish species in Nyanza Gulf, Lake Victoria, East Africa, 1975

Graduation date: 1978A trawling survey was conducted on Nyanza Gulf, Lake\ud Victoria, East Africa during March-December 1975 to investigate\ud association and distribution of the major fish species\ud and to provide information for management of the fishery.\ud A total of 139 sites was sampled and all fishes caught were\ud counted, measured, and weighed. At each site, depth, bottom\ud type (substratum), offshore distance, season, and time of day\ud were recorded.\ud Species association and patterns of occurrence were\ud investigated by the Bray-Curtis clustering strategy and\ud interrelationships between species and environmental data\ud were analyzed by canonical correlation. Patterns of the\ud trawl data suggested that fish species responded to variation\ud in physical features of the Gulf and time of sampling.\ud These environmental factors showed significant associations\ud with species breeding and feeding habits. Depth seemed to\ud be the most important factor in species distribution and had\ud a positive correlation with substrate coarseness.\ud Although the Gulf could be divided into three geographical\ud areas (Uyoma-Naya, Homa, and Winam) according to\ud physical features, statistical analyses showed two rather\ud distinct patterns of ichthyomass; one at Uyoma-Naya dominated\ud by Tilapia nilotica, Tilapia variabilis, Lates\ud albertianus,and Clarias mossambicus; and the other at Homa-\ud Winam dominated by Haplochromis spp., Bagrus docmac, Protopterus\ud aethiopicus and Clarias mossambicus.\ud As little is known about species life histories,\ud management strategies may be based upon ichthyomass\ud patterns instead. On the Gulf, there is too much fishing\ud effort to make it practical to sample the entire catch for\ud purposes of assessing the effect of different gears and\ud regulations on the fishery. However, if sufficiently small\ud exploratory 'fishery windows' are established in each of\ud the major localized ichthyomass areas, management policies\ud could be developed experimentally without destroying the\ud assemblages. A tentative exploratory management program is\ud suggested based on the identified ichthyomass patterns, and\ud both direct and indirect management strategies are proposed