The effect of permafrost on time-split soil erosion using radionuclides (¹³⁷Cs, ²³⁹ ⁺ ²⁴⁰Pu, meteoric ¹⁰Be) and stable isotopes (δ¹³C) in the eastern Swiss Alps

Purpose: Global warming is expected to change the thermal and hydrological soil regime in permafrost ecosystems which might impact soil erosion processes. Erosion assessment using radionuclides can provide information on past and ongoing, i.e. time-split, processes. The focus of this work was to find out if permafrost soils in the Swiss Alps differ in their medium- and long-term erosion rates from non-permafrost soils and if rates have accelerated during the last few decades. Materials and methods: Using cosmogenic (meteoric ¹⁰Be) and anthropogenic radionuclides (¹³⁷Cs, ²³⁹ ⁺ ²⁴⁰Pu), a time-split approach was achieved by determining erosion activities on the long (millennia; ¹⁰Be) and medium term (decades; ¹³⁷Cs, ²³⁹ ⁺ ²⁴⁰Pu). Additionally, the stable isotope δ¹³C signature in soil organic matter was used as a qualitative indicator for soil disturbance patterns. We compared soil erosion processes in permafrost soils and nearby unfrozen soils in the alpine (sites at 2,700 m asl, alpine tundra) and the subalpine (sites 1,800 m asl, natural forest) range of the Swiss Alps (Upper Engadine). ¹³⁷Cs, ²³⁹ ⁺ ²⁴⁰Pu and δ¹³C measurements were performed at the alpine sites only. Results and discussion: Depending on the calculation procedure (profile distribution model or inventory method), the ¹³⁷Cs measurements revealed soil accumulation rates of 1–3 t/km²/year in permafrost soils and 34–52 t/km²/year in non-permafrost soils. However, due to snow cover and subsequent melt-water runoff during ¹³⁷Cs deposition after the Chernobyl accident, caesium does not seem to be an appropriate soil erosion tracer on the investigated alpine sites. With ²³⁹ ⁺ ²⁴⁰Pu, more reliable results were achieved. ²³⁹ ⁺ ²⁴⁰Pu measurements provided erosion rates of 31–186 t/km²/year in permafrost soils and accumulation rates of 87–218 t/km²/year in non-permafrost soils. Erosion and accumulation were relatively low and related to the vegetation community. The long-term (¹⁰Be) soil redistribution rates (erosion rates up to 49 t/km²/year and accumulation rates up to 4 t/km²/ year) were low with no significant differences between permafrost and non-permafrost sites. The δ¹³C signature indicated soil disturbances in permafrost and non-permafrost soils compared to the reference site. Conclusions: Our results highlight that soil redistribution rates have increased during the last few decades. However, whether the higher medium-term erosion rates obtained for the last decades are the result of the ongoing climate warming and related accelerated soil erosion or if other factors (e.g. measurement uncertainties) have been responsible for such an increase could not fully be clarified