Pedogenic controls of soil organic carbon stocks and stability beneath montane Norway spruce forests along a precipitation gradient.

Reliable data on SOC stocks in forest soils is required in the context of climate change and soil health assessments but still limited by input data availability (e.g., bulk density) and methods used for stock calculation. Relatively few studies have investigated the stability of SOC in forest soils. We investigated SOC stocks and fractionation in soils beneath Norway spruce forests and grasslands in the montane zone along a gradient of mean annual precipitation (MAP). We sampled soil cores volumetrically to 40 cm depth and measured SOC in the fractions <2 mm (fine earth), >200 µm and 200-20 µm (coarse and fine POM), and <20 µm (MAOM) along with potential pedogenic controls. Total SOC stocks beneath forests in the study region, calculated by the equivalent soil mass (ESM) approach to 40 cm depth, amount to 79.0 ± 29.9 (mean ± standard deviation) Mg ha-1 (n = 20) in the mineral soil, and to 92.9 ± 30.6 Mg ha-1 including the litter layer, with a share of 55 % associated with POM. MAOM makes up ∼41 % of SOC in the uppermost mineral layer (0-5 cm) and increases to 71 % in the subsoil (20-40 cm). Multiple regression models show that MAOM is largely controlled by ammonium oxalate extractable Al (Alo) in the forest subsoils (20-40 cm), and increases with MAP in the topsoil layers (0-20 cm). Soils on carbonate rock stand out with ∼80-100 % larger shares of MAOM in the uppermost soil layers (0-10 cm) which is likely connected to higher soil pH and MAP, supporting microbial transformation and subsequent stabilisation of organic matter, which is reflected in narrower C:N ratios in MAOM and SOC. Including the litter layers, ESM-based total SOC stocks in forest soils tend to exceed those beneath grassland (80.2 ± 21.9 Mg ha-1; n = 31) by 16 %, but only by 6.4 % if calculated by the conventional fixed-depth (FD) approach. In contrast to the forest soils, SOC stocks beneath grasslands are dominated by MAOM (75.6 %). We conclude that (coniferous) forest soils are a poor reference for establishing sequestration potentials for stable SOC. The observed large proportion of POM in forest topsoils and its increase with declining MAP (indicating water availability) suggests a risk of SOC losses in response to increasing droughts due to climate change.