Lake Malawi sediment and pore water chemistry: Proposition of a conceptual model for stratification intensification since the end of the Little Ice Age

International audienceSedimentary records of salinity indicators are largely used to reconstruct past climatic changes in lacustrine systems where chemistry is sensitive to hydroclimatic conditions. In large fresh lakes of the East African Rift such as Lakes Tanganyika and Malawi, salinity is often considered constant and other paleoclimatological proxy data are used. However, a relation between lake surface chloride concentration and hydroclimatic regime was previously demonstrated at the century scale in Lake Tanganyika. This relation is transposed to Lake Malawi on the base of similarity between hydrochemical budgets of both lakes that are computed for the whole lake and epilimnion. Whereas numerous physico-chemical difficulties make generally debatable use of lake pore water chemistry, as illustrated here by diffusion modelling, the dissolved chloride concentration profile from a core sampled in northern Lake Malawi is considered as a potential indicator of limnological–hydroclimatic condition changes for the last 200 years. A decrease in pore water chloride concentration between 1840 AD and present situation is directly associated to a metalimnetic water salinity decrease. The chronology of this event is synchronous with diatom productivity change demonstrated by Johnson et al. (2001) at the end of the Little Ice Age (LIA). A conceptual model of Lake Malawi, based on salinity, organic carbon and its “dead” watershed contribution, lake-level and productivity changes since the mid 19th century is presented. A new scenario is proposed, based on thermal stratification reinforcement at the end of the LIA. Lake productivity and chemistry depend on stratification strength, water column mixing rate and on climatic variability. During the LIA, nutrient distribution profiles were more homogeneous with depth due to the climatically (colder and drier climatic conditions than today) induced destabilisation of the mixing barrier. The productive system is then auto-supplied and does not require external silicon supply while chloride concentration is higher due to closure of the basin. Stratification has become more stable since the end of the LIA, as in nearby Lake Tanganyika where it is linked to air temperature global increase. Results demonstrated that chloride is a suitable indicator of the present and past hydroclimatic and hydrodynamic regimes. This work shows that in Lake Malawi, assessment of “dead” carbon (or old “refractive watershed carbon”) contribution to the TOC is a key parameter to set chronological frame from $^{14}$C ages