『Abstract
　Seasonal changes in river chemistry offer the potential to assess how weathering processes respond to changing meteorological parameters and ultimately how chemical weathering might respond to climatic parameters. Systematic seasonal variations in magnesium isotope ratios (the ²⁶Mg/²⁴Mg ratio expressed as δ²⁶Mg in per mil units) are reported in stream waters from a mono-lithological granitic, weathering-limited, first order catchment from the Swiss Alps (Damma glacier). Rain, ground, and pore-waters, in addition to plants, rocks, mineral separates and soil are also reported. The concentration response of the river waters is attenuated compared to the large changes in discharge. However, the systematic trends in the isotope data imply that either the source of the Mg changes in a systematic manner, or that the process by which Mg is released into solution changes as a function of discharge.
　The two first order observations in the data that need to be explained are 1) the systematic enrichment in ²⁴Mg in the stream waters compared to the granitic rocks they drain and 2) a systematic increase in δ²⁶Mg in the waters during the summer melt season. Both observations (which are similar to many other rivers draining silicate rock) can either be accounted for by 1) conservative mixing between at least two different sources of Mg (in addition to precipitation inputs), or 2) process related fractionation. If the stream water compositions can be rationalised by multi-component mixing, there is at least one unidentified component with a δ²⁶Mg＜-1.2‰. This is considered unlikely. Multiple physicochemical processes could fractionate Mg isotope ratios such as 1) preferential leaching of ²⁴Mg, 2) exchange of Mg onto (or from) mineral surfaces and into interlayer sites of clays, 3) uptake by plants, and 4) ²⁶Mg could be preferentially retained during the formation of secondary phases, such as clays, amorphous phases or oxides. These processes are not mutually exclusive and distinguishing between them at a field scale is not trivial, but significant biological uptake is improbable at this site. Unless there is a non-identified external input of Mg, ²⁶Mg must be accumulating in solid phase residues in the catchment because of at least one physicochemical process. Such processes are likely well described, at least in the first order by a Rayleigh distillation model. Simple calculations illustrate how much ²⁶Mg would accumulate in the catchment per unit time. In the first order, the isotopic enrichments in the solids are so small that they would not be detectable for the time-scales that are relevant to this field site, in spite of the marked impact on the water chemistry. The seasonal signal detected by mg isotope ratios is promising for using them (with a better understanding of fractionation mechanisms) to quantify how specific weathering processes impact upon both export fluxes, and retention of elements within catchments.

Keywords: River geochemistry; Magnesium; Mg isotopes; Chemical weathering』

1. Introduction
2. Study area and sample collection
3. Analytical procedure
4. Results
　4.1. Correcting for precipitation inputs
5. Seasonal variations in stream water δ²⁶Mg and causal mechanisms
　5.1. Conservative mixing as a control of stream water δ²⁶Mg
　5.2. Ion exchange as a control on mg isotopes?
　5.3. Biological forcing of seasonal mg isotope variations
　5.4. Preferential release of ²⁴Mg during dissolution
　5.5. Formation of secondary minerals and mg isotope variations
　5.6. Implications and semi-quantitative constraints on seasonal variations
6. Conclusions and perspectives
Acknowledgements
References