This thesis develops, validates and applies an empirical model that provides the first spatially explicit estimates of gross and net phytoplankton production in Lake Victoria. Gross and net phytoplankton production are in turn used to estimate the maximum sustainable yield (MSY) of Lake Victoria's fishery following an empirical formula and the carbon efficiency transfer method. Chapter 2 presents results from three inshore areas where diurnal and sub-seasonal gross and net phytoplankton production was derived using an adapted version of the phytoplankton production model developed by Fee (1990). Spatial and temporal trends of chlorophyll (chl), PI parameters, the vertical attenuation of PAR (kPAR), Secchi depths (SD) and respiration rates are identified. kPAR and SD are highly correlated to chl within the euphotic zone, as well as to each other. Furthermore, the two PI parameters, PBM and aB, exhibit a strong linear relationship and both decline along an increasing chl gradient, presumably due to increased light-limitation, a taxonomic shift from diatoms to cyanobacteria with increasing chl as well as an increased need for biologically fixed nitrogen. These hypotheses are supported by observed synchronous changes in the PSII:PSI ratio of phytoplankton and changes in the chl-specific attenuation of PAR (kchl). Relationships are also derived between biomass-specific respiration rates (RB) with chl and PBM; similar to PI parameters RB decreases with increasing chl. Owing to these correlative trends, only one parameter is required to estimate gross phytoplankton production through the empirical model developed in this thesis. The empirical model predicts that gross phytoplankton production increases in a near linear fashion between chl of 0 to 10 mg. m-3, begins to flatten out as chl approaches 20 mg. m-3 and then slightly decreases when chl exceeds 40 mg. m-3 where the maximum PPG of 13. 1 g O₂. m-2. day-1 is reached and is in close agreement with a theoretical argument proposed by Talling (1965). Areal respiration and consequently net phytoplankton production are sensitive to chl within the mixed layer as well as mixed layer depths. Overall, the lakewide averages of gross and net phytoplankton production are 9. 68 and 2. 2 g O₂. m-2. day-1 respectively. Significant temporal variability was observed on sub-seasonal scales within the inshore of Lake Victoria, and changes in limnological parameters coincided with changes in water column temperatures in each of the three bays. In Fielding Bay, the availability of meteorological data revealed that strong nocturnal wind events decreased both the water column temperature and chl, while both parameters generally increased in the absence of any such wind event. Lateral exchange of water with deeper areas through strong wind events essentially flushes Fielding Bay causing the observed decreases in both the water column temperature and chl; this hydrodynamic event also influences other limnological parameters according to their respective correlative regression equations with chl. Spatial trends were also observed between inshore areas. The deepest area, Napoleon Gulf, has the lowest values of chl while the shallowest area, Inner Murchison Bay, has the highest chl as the mean depth of a bay sets an approximate upper limit on chl. With respect to diurnal variability, PI parameters decline through the day, kPAR increases over the day and no statistically valid trends were ascertained for chl and RB. Chapter three examined spatial and seasonal patterns of chlorophyll fluorescence, temperature, dissolved oxygen and water transparency from four lakewide cruises. Significant spatial variability of each parameter confirmed that lakewide data is required to generate spatially explicit estimates of phytoplankton production. Complex patterns in the thermal structure during each cruise illustrated that physical processes in Lake Victoria are at times more complex that a previously stated unidirectional hypothesis of warm water in the north and cool water in the south (Spigel and Coulter 1996), and these patterns influence spatial patterns in dissolved oxygen and Secchi depths. Similar to Chapter 2, estimates of chl within the mixed layer were highly correlated to mixed depths, while lakewide averages of chl are lower than previously reported offshore values (Mugidde 1993, 2001).

This thesis develops, validates and applies an empirical model that provides the first spatially explicit estimates of gross and net phytoplankton production in Lake Victoria. Gross and net phytoplankton production are in turn used to estimate the maximum sustainable yield (MSY) of Lake Victoria's fishery following an empirical formula and the carbon efficiency transfer method. Chapter 2 presents results from three inshore areas where diurnal and sub-seasonal gross and net phytoplankton production was derived using an adapted version of the phytoplankton production model developed by Fee (1990). Spatial and temporal trends of chlorophyll (chl), PI parameters, the vertical attenuation of PAR (kPAR), Secchi depths (SD) and respiration rates are identified. kPAR and SD are highly correlated to chl within the euphotic zone, as well as to each other. Furthermore, the two PI parameters, PBM and aB, exhibit a strong linear relationship and both decline along an increasing chl gradient, presumably due to increased light-limitation, a taxonomic shift from diatoms to cyanobacteria with increasing chl as well as an increased need for biologically fixed nitrogen. These hypotheses are supported by observed synchronous changes in the PSII:PSI ratio of phytoplankton and changes in the chl-specific attenuation of PAR (kchl). Relationships are also derived between biomass-specific respiration rates (RB) with chl and PBM; similar to PI parameters RB decreases with increasing chl. Owing to these correlative trends, only one parameter is required to estimate gross phytoplankton production through the empirical model developed in this thesis. The empirical model predicts that gross phytoplankton production increases in a near linear fashion between chl of 0 to 10 mg. m-3, begins to flatten out as chl approaches 20 mg. m-3 and then slightly decreases when chl exceeds 40 mg. m-3 where the maximum PPG of 13. 1 g O₂. m-2. day-1 is reached and is in close agreement with a theoretical argument proposed by Talling (1965). Areal respiration and consequently net phytoplankton production are sensitive to chl within the mixed layer as well as mixed layer depths. Overall, the lakewide averages of gross and net phytoplankton production are 9. 68 and 2. 2 g O₂. m-2. day-1 respectively. Significant temporal variability was observed on sub-seasonal scales within the inshore of Lake Victoria, and changes in limnological parameters coincided with changes in water column temperatures in each of the three bays. In Fielding Bay, the availability of meteorological data revealed that strong nocturnal wind events decreased both the water column temperature and chl, while both parameters generally increased in the absence of any such wind event. Lateral exchange of water with deeper areas through strong wind events essentially flushes Fielding Bay causing the observed decreases in both the water column temperature and chl; this hydrodynamic event also influences other limnological parameters according to their respective correlative regression equations with chl. Spatial trends were also observed between inshore areas. The deepest area, Napoleon Gulf, has the lowest values of chl while the shallowest area, Inner Murchison Bay, has the highest chl as the mean depth of a bay sets an approximate upper limit on chl. With respect to diurnal variability, PI parameters decline through the day, kPAR increases over the day and no statistically valid trends were ascertained for chl and RB. Chapter three examined spatial and seasonal patterns of chlorophyll fluorescence, temperature, dissolved oxygen and water transparency from four lakewide cruises. Significant spatial variability of each parameter confirmed that lakewide data is required to generate spatially explicit estimates of phytoplankton production. Complex patterns in the thermal structure during each cruise illustrated that physical processes in Lake Victoria are at times more complex that a previously stated unidirectional hypothesis of warm water in the north and cool water in the south (Spigel and Coulter 1996), and these patterns influence spatial patterns in dissolved oxygen and Secchi depths. Similar to Chapter 2, estimates of chl within the mixed layer were highly correlated to mixed depths, while lakewide averages of chl are lower than previously reported offshore values (Mugidde 1993, 2001).

Some of the major problems we face today are human induced changes to the nitrogen (N), phosphorus (P) and carbon (C) cycles. Predicted increases in rainfall and temperature due to climate change, may also increase dissolved organic matter (DOM) inflows to freshwater ecosystems in the boreal zone. N, P, C and light, are essential resources that most often limit phytoplankton (PPr) and bacterial production (BP) in the pelagic zone of lakes. PPr and BP not only constitute the total basal C resource for the pelagic aquatic food web, but also influence ecosystem function and biogeochemical cycles. In this thesis I studied how N, P, C and light affect the relative and absolute rates of PPr and BP, along a wide latitudinal and trophic gradient using published data, and in two in situ mesocosm experiments in a clear water oligotrophic lake. In the experiments I manipulated bottom-up drivers of production and top-down predation to examine how these factors interact to affect pelagic food web structure and function. The most important predictors of PPr globally (Paper I) were latitude, TN, and lake shape. Latitude alone explained the most variation in areal (50%) and volumetric (40%) PPr. In terms of nutrients PPr was primarily N-limited and BP was P-limited. Therefore bacteria and phytoplankton were not directly competing for nutrients. BP:PPr was mostly driven by PPr, therefore light, N, temperature and other factors affecting PPr controlled this ratio. PPr was positively correlated with temperature, but not BP, consequently, higher temperatures may reduce BP:PPr and hence the amount of energy mobilised through the microbial food web on a global scale. In papers II and III interaction effects were found between C-additions and top-down predation by young-of-the-year (YOY) perch. Selective predation by fish on copepods influenced the fate of labile C-addition, as rotifer biomass increased with C-addition, but only when fish were absent. Interaction effects between these top-down and bottom-up drivers were evident in middle of the food web, which is seldom examined in this type of study. Although the energy pathway from bacteria to higher consumers is generally longer than from phytoplankton to higher trophic levels, increased BP still stimulated the biomass of rotifers, calanoid copepods and YOY fish. However, this appeared to be mediated by intermediate bacterial grazers such as flagellates and ciliates. Light was an important driver of crustacean zooplankton biomass (paper IV), but the light:nutrient hypothesis was inadequate to predict the mechanisms behind the decrease in zooplankton biomass at low light. Instead, it appeared that reduced edibility of the phytoplankton community under low light conditions and reduced BP most strongly affected zooplankton biomass. Thus, the LNH may not apply in oligotrophic lakes where PPr is primarily N-limited, Daphnia is rare or absent and mixotrophic phytoplankton are abundant. N, P, C and light manipulations have very different effects on different parts of the pelagic food web. They influence the relative rates of PPr and BP, affect phytoplankton community composition, alter the biomass of higher trophic levels and change pathways of energy transfer through the pelagic food web. This thesis adds valuable information as to how major changes in these resources will affect food web structure and function under different environmental conditions and future climate scenarios. / Lake ecosystem response to environmental change

Over the period of the spring phytoplankton bloom (March-May 2009), an intensive sampling program was conducted at 2 sampling depths (0.5 m and 1% PAR) located at lacustrine zone of the canyon-shaped, meso-eutrophic Římov reservoir (Czech Republic). Changes in the production and community dynamics of epilimnetic bacterioplankton, studied by means of group-specific rRNA-targeted oligonucleotide probes, were related to shifts in composition and activity of phytoplankton, and to remarkable changes in total protistan bacterivory. It was documented that particular groups of bacterioplankton responded differently to: (i) major shifts in phytoplankton composition and its activity and to (ii) changes in overall protozoan grazing pressure.