Hillslopes in action? Combining in situ 14C- and 10Be-derived insights into the spatio-temporal variability of soil dynamics on Swiss alpine hillslopes

Rapidly eroding mountain ranges act as major sources for the export of clastic sediments, solutes and organic matter from the continents to the oceans, and thereby affect the global climate via the sequestration of atmospheric CO₂ into geological reservoirs. At the same time, mountainous landscapes experienced Quaternary variations in climate and vegetation – including the waxing and waning of glacial ice masses – and were therefore exposed to temporally varying erosional regimes. In the specific case of soil-mantled hillslopes, changing environmental conditions (e.g., temperature, precipitation) along with spatially variable topographic parameters (e.g., slope, aspect) are expected to modify the geomorphic processes responsible for the mobilization and downslope movement of soil material, such that soil erosion might represent a function of both space and time. Furthermore, transient soil erosion is likely coupled to alterations of the soil thickness and hence to changes in the production of soil material from the underlying bedrock. In this study, we investigate the spatio-temporal variability of soil production and erosion on different soil-covered alpine landforms within the Goms area (Switzerland), based on paired measurements of the in situ ¹⁴C and ¹⁰Be concentrations within quartz separates of saprolite (n=12) and till-derived (n=2) samples. The short half-life of ¹⁴C compared to ¹⁰Be provides this approach with a unique sensitivity for short-term changes of soil production and/or soil thickness on Lateglacial to Holocene time scales. For multiple sites, the apparent soil production rates and exposure ages calculated from the cosmogenic nuclide concentrations reveal a constant exposure of the saprolite-soil interface, without any (detectable) signs of soil production or changes to the thickness of the overlying soil column since the Early to Mid-Holocene. The absence of soil production in combination with a constant vertical extent of the soils further implies that the erosional downslope transport of soil material is inactive at these locations, despite their generally steep slope angles (10 to 38°). Remarkably, the estimated exposure ages also differentiate between relatively young samples (¹⁰Be: 5.5 to 7.8 ka; ¹⁴C: 5.4 to 7.9 ka) from tributary hillslopes and older samples (¹⁰Be: 10.2 to 12.0 ka; ¹⁴C: 9.6 to 12.7 ka) from the hillslopes of the main valley and erosional ridges, which represent important landforms within the investigated landscape. Additional results from Markov chain Monte Carlo inversion modelling indicate the potential occurrence of instantaneous losses of the soil cover (9.3 to 14.7 m at 7.6 to 8.7 ka) or the step-like decline of soil production rates (11- to 12-fold at 12.4 to 13.1 ka) for most of the remaining samples. Overall, these observations demonstrate the co-occurrence of ‘static’ sites without active soil production and erosion and more ‘dynamic’ sites characterized by transient, non-zero soil production and erosion rates, attesting to the spatial variability of hillslope soil dynamics between an active and an inactive state.