Pelagic nitrogen fixation in Lake Victoria (East africa)

Between 1994 and 1998 biological nitrogen fixation was measured using the acetylene reduction method at inshore and offshore sites on northern Lake Victoria. Rates of biological N-fixation were high and often exceeded 0.5 mug N/L/h. Average rates of volumetric N-fixation at optimal irradiance were 8 times higher at inshore locations than in the offshore. Rates of annual areal N-fixation, modeled from the N-fixation light-response and light attenuation in the lake were moderate to high (1.8 to 23.1 g N/m(2)/y) depending on location within the lake, and were, on average, only twice as high inshore as offshore. Variations in the light extinction coefficient explained a small but significant proportion of the variation in the optimal N-fixation in Lake Victoria. Algal biomass and N-fixation were lower in the more deeply mixing (> 20 m) offshore waters, because of the persistently low mean light intensities (108 +/- 41 muE/m(2)/s) in the water column most of the year. N-fixation increased with increased light availability, and maximal rates occurred when the lake was shallowly mixing anti thermally stratified. At both inshore and offshore stations, minima of algal biomass and N-fixation were consistent with low light availability in July when the lake was most deeply mixing. The ratio of the mean water column irradiance (I-24) to the irradiance at which N-fixation approaches saturation (I-k) was often < 1, and provides evidence that N-fixation was light-limited, chronically so in the offshore region. Light limitation of algal growth lessens the algal demand for N, and it constrains algal biomass development anti N-fixation more in the offshore compared to the inshore surface waters of Lake Victoria. Biological N-fixation is the largest input of fixed N to Lake Victoria, greatly exceeding estimates of atmospheric deposition and river inputs of N.