Abstract
The Lake Tanganyika pelagic fish community consists of two short‐lived clupeid species, of which Stolothrissa tanganicae is dominant, and four much larger, long‐lived Lates species. Purseseine catches in previously unfished areas contained approximately equal amounts of clupeid prey and Lates predators. The annual production/mean biomass ratio for Stolothrissa is 3.5–4, which approximates the instantaneous mortality rate. Predator P/B ratios probably are less than 1. Prey fish biomass is about 230 kg/hectare, and production about 700 kg/hectare per year in the north of the lake (Burundi). Even in intensively fished areas, most prey production probably is converted to predator tissue. Without predators, potentially harvestable clupeid production in the Burundi sector would theoretically bc 350 kg/hectare per year, which extrapolates to 1.1 million t/year for the lake. Purse‐seining removes Lates species faster than clupeids, leaving more clupeid production for yield. Fishery models should attempt to correlate variations in Lates biomass to clupeid productivity. Limits to potential clupeid yield might be indicated by quantitative equilibria with natural predators, but it is unlikely that present fishery methods could closely simulate natural predation. With increasing dependence on clupeid species, the fishery may become more seasonal and more susceptible to annual environmental fluctuations. These and other African data strongly suggest that fish production and potential yield usually are higher in tropical than in temperate fresh water. It is suggested that predators are more prominent in tropical fish communities because of relatively high productivity of prey fish.