Can vein-filling speleothems constrain the timing of deep-seated gravitational slope deformation? A case study from the Vinschgau (Italian Alps)

The steep, south facing slopes of the central Vinschgau (Val Venosta, South Tyrol, northern Italy) are characterized by deep-seated compound rockslides affecting 51 km² and reaching beneath the Quaternary valley fill. Morphological features include double ridges, trenches, scarps, and counterscarps, whereby the extent of displacement rises from W to E. Near the toe of these slopes, abundant fractures are present whose orientation and spatial strongly suggest that they are related to the gravitational slope deformations. These fractures host carbonate speleothems which are associated with local tufa occurrences at the surface. Given that the metamorphic host rock is essentially carbonate-free, these carbonate deposits require extensive water-rock interactions to form. Modern springs along the toe of the slope are also mostly supersaturated with respect to calcite and locally also aragonite. The invariant chemical composition, very low tritium levels, and in particular the low oxygen isotope values indicate that this groundwater infiltrates at high elevation and is characterized by a long residence time. On the other hand, local infiltration in the lower reaches of the slope is very limited due to the semiarid climate and is incompatible with the measured groundwater stable isotope composition and the high degree of mineralization. We therefore propose a conceptual model which reconciles the deep-seated gravitational slope deformation with the occurrence of mineralized groundwater and associated carbonate precipitates in fractures near the toe of the slope. These deposits provide a means to place constraints on the timing of rock fracturing and hence the slope movements. U/Th dating of 34 samples from eight sites shows that speleothem deposition started 14.2 ka BP and occurred semi-continuously throughout the Holocene. Thus, gravitational displacements likely commenced with a lag of ~3–4 ka after the deglaciation of the valley, which is also consistent with the few available dates on the onset of other deep-seated slope deformations in the Alps.