Large East African Rift lakes have been changing rapidly during the last decades. They typically have a relatively high productivity compared to large temperate lakes and have active fisheries providing local populations with a relatively cheap source of proteins. However, human-induced changes, including climate change, can have significant effects on primary production of these lakes, as shown for Lake Tanganyika. It is likely that these primary production decreases have affected upper trophic levels and fisheries, but, before being able to predict the extent of the primary productivity changes and how they affect whole ecosystem production, an improved understanding of ecosystem function and food web processes is required. For instance, food web efficiency may greatly depend on the amount of organic carbon transiting through the microbial food web, known to be important in these lakes.

Lake Kivu, located north of L. Tanganyika, has undergone recent changes induced by alien species introduction and possibly climate change, which have affected the lake's biodiversity, productivity and ecosystem resources. Future industrial methane harvesting additionnaly threatens sustainable development of ecosystem resources. Lake Kivu may provide an adequate model for studying responses of large tropical lake to human-induced changes: indeed, despite its physical and geochemical peculiarities, the limnological and ecological processes in its pelagic waters are subjected to the same forcing as in other great lakes of the same region, as shown by studies conducted in recent years. In addition, the simple pelagic food web of the lake facilitates our understanding of ecosystem functioning and of human-induced alterations. Some past changes in the lake have been revealed by analyses of the sediments, which can be further improved by studies of an array of proxies, by development of new proxies and by inference from present ecological processes taking place in the mixolimnion.

In this project, we will exploit the important database acquired in the period 2002-2009 (WP1). The existing data base includes limnological variables, plankton (diversity, biomass and production of phyto- and zooplankton), fish abundance and meteorological data. This data base will be completed by sediment archives (biogeochemical and biological proxies) and by remote sensing of phytoplankton biomass and surface hydrological features, for taking into account spatial and temporal heterogeneity.

New in situ studies will be conducted in order to extend the database and increase our present understanding of ecosystem biodiversity and functioning (WP2). This part of the project will include monitoring of the mixolimnion by regular sampling and measurements and by continuous recording of temperature (vertical profiles) and of chlorophyll a and phycoerythrin concentration, field experiments for determining the carbon and nutrient pathways within the planktonic food web, studies of proxies (incl. Si fractionation as a proxy of diatom productivity) in the water column and in the sediments, and surveys of fish stock and of fisheries yield. Laboratory studies (WP3) will be undertaken to determine ecophysiological requirements of key diatoms isolated from L. Kivu and for studying Si isotopic fractionation by different species under conditions mimicking those prevailing in the lake (light, temperature, nutrient concentration).

The final step will be devoted to data processing and modeling (WP4), in order to:

* link atmospheric forcing and the lake's physics: relation between atmospheric conditions and lake temperature and water column structure, aiming to understand/simulate the variability of seasonal mixing processes and forecast long term changes

* link physical processes with biological and ecological (e.g. nutrient availability) processes: diversity and biomass of plankton, fate of primary production in the planktonic food web, fish abundance and fisheries yield

* predict future changes of ecosystem processes and resources, as a result from fisheries management, exploitation of methane from the deep waters, and climate change (linking global climate to regional climate).

Finally, we will examine how the methodologies used and developed in the research project could be applied to other large African lakes, particularly Lake Tanganyika.