General framework This project aimed to evaluate and understand processes of eolian and fluviatile inputs occurring in the photic zone that may lead to the formation of proxies of seasonally-controlled weather conditions.

Specific tasks RMCA 
In order to better quantify the diatom productivity, we develop innovative biomarkers in order to establish new climate proxies for Lake Baïkal. Main Results: Development of Si isotopes as a fresh water diatom activity proxy. Silicon isotopic fractionation (expressed as d29Si) by siliceous organism represents the preferential incorporation of light Si isotopes to produce opal frustules. Studies on marine diatoms have recently demonstrated the potential of such isotopic tracer to determine past silicon bio-cycle (De La Rocha et al. 1998; Varela et al. 2004), avoiding bias due to opal dissolution. Thus, the fractionation is expected to be an accurate parameter to quantify the silicon utilisation by diatoms (De La Rocha et al. 1997). However, this proxy has never been validated in freshwater environment in order to reconstruct paleoproductivity in lacustrine sediments. For this program, opal isotopic signatures and their relationship to diatom productivity have been investigated during the last interglacial periods (Holocene) in a sediment core from Lake Baikal. Freshwater samples from lake Baikal were collected in July 2003 in the northern and southern basins for Si-isotopic analyses. These results evidenced a similar nutrient type pattern as the one found in Lake Tanganyika (Heavier d29Si in surface waters than in deep waters) and confirmed the biological fractionation occurring in surface waters. Compared to Lake Tanganyika, the mean d29Si values obtained in the dissolved fraction in surface (1.08±0.07‰) and deep waters (0.90±0.04‰) are heavier. In addition, the lower Si inventory in surface and deep waters (from 25±5 to 38±7 µM) suggest an overall higher Si uptake in Lake Baikal. This large diatom productivity is probably related to the complete annual deep water mixing that increases the nutrient availability in the euphotic zone while a permanent stratification prevail in East African Great Lakes. Diatoms collected in 15 sediments trap (Neutrino/BAIK00-1 to -15) placed at various depths from 15 to 1350 m during one year from March 2000 to March 2001 have also been analyzed for Si isotopes. They display a mean Si isotopic signature of 0.56±0.07 ‰, representing the integrated signature of siliceous phytoplankton produced in surface waters over a one-year period. This average value is significantly heavier than the diatoms signature found in Lake Tanganyika confirming a probably long-term higher biologic silicon uptake in Lake Baikal. Comparing the dissolved surface water signal (1.08±0.07‰) with the diatom opal signature measured in traps indicates a difference that represents an enrichment factor value of - 0.52‰, very close to the one obtained in Lake Tanganyika and on marine diatoms.
Sediments from the Vidrino core present a Si-isotopic signature that vary from 0.06 to 0.64‰ and present a markedly good correlation (R² = 0.66) with independent measurement done for d18O (20.8 to 28.5‰) at the Univ. College of London. Such relationship may be associated to the origin of nutrient supply through riverine input or through deep water mixing that probably account for a significant fraction of total Si pool in surface waters. Consequently, long term changes in riverine supply or deep water turnover/stratification may impact the diatom Si-isotopic composition in 2 ways: 1) by changing the nutrient supply intensity to the lake, directly influencing the diatom productivity, 2) by shifting the Si-isotopic signature of the dissolved Si sources. Climate variability could affect simultaneously the productivity and the initial condition of Si-isotopic signature, which makes it more complex to interpret. Looking in details into the core Si-isotopic profile indicates that the most recent diatom sediment sample presenting an estimated age of 63 years is lower than the mean diatom trap samples collected in 2001. If we consider a constant Si supply signature over the last decades, we may suggest a higher Si utilization in recent years according to heavier Si-isotopic signal. On the other hand, a minimum value of 0.06‰, 5300 years ago, is an indication either of low diatom productivity at that time or a shift in Si source signature. Horiuchi et al. 2000 evidenced a dramatic change in BioSi content about 5000 years B.P. in sediment from the northern basin on the Academician ridge, attributed to the strong weathering of the pedosphere. We were able to evidence a strong correlation at that time between d29Si and d18O, which is influenced most probably by riverine water input changes (Morley et al., 2004). Consequently, a spectacular change in riverine nutrient inputs is the most probable hypothesis to explain the lowest d29Si values. While BioSi stay much lower before 5000 years B.P., d29Si tend to return to a higher value, which may indicates that diatom Si uptake go back to an equilibrium with respect to nutrient inputs.