Carbon soil stock change in an intensive crop field near Paris reveals significant carbon losses over a decade

Abstract. Soil is a large pool of carbon (C), storing globally twice as much carbon as the atmosphere and three times as much as vegetation. Soil organic carbon (SOC) stocks are significantly impacted by land-use changes, either negatively when forests or grasslands are converted into crops or positively when the opposite occurs. This context underpins the “4per1000” initiative, which aims to promote SOC storage in soils as a mitigation strategy. However, intensive cropping and climate change may lead to losses of organic and inorganic carbon from soils, which calls for long-term observations of soil organic carbon stocks in reference ecosystems worldwide. To address this, a harmonised reference soil sampling protocol was developed for all ecosystem sites within the European Integrated Carbon Observing System (ICOS) research infrastructure, starting in 2017 with revisits planned every 5–10 years. This study presents a first case at the French cropland site FR-Gri (wheat–maize–barley–oilseed rape rotation), assessing SOC stock in 2019 with the ICOS protocol, which was combined with earlier SOC stock sampling data from the European project CarboEurope. A significant soil decompaction was observed over the 13.5 years in the 0–30 cm layer. Bulk density decreased by 22 % in the 0–5 cm layer (from 1.31 to 1.02 g cm−3) and by 5 % in the 5–30 cm layer (from 1.53 to 1.45 g cm−3), likely due to the adoption of reduced tillage since 2004. SOC content increased by 10 % in the 0–5 cm layer but declined by 6.2 % in the 5–30 cm layer. The SOC stocks based on equivalent soil mass (ESM) increased by 7.6 % in the 0–5 cm layer, but decreased by 11 % and 9 % in the 5–30 cm and 30–60 cm layers, respectively. Overall, the ESM-based SOC stock in the 0–60 cm layer decreased by approximately 0.95 ± 0.22 kg C m−2 (or 9 Mg C ha−1) between 2005 and 2019, corresponding to 0.65 % yr−1 relative to the initial SOC stock (∼ 11 kg C m−2 in the 0–60 cm layer). This leads to an average yearly decrease rate of 0.072 ± 0.017 kg C m−2 yr−1 (or 0.72 ± 0.17 Mg C ha−1 yr−1), consistent with previous studies. To further interpret this trend, we applied the soil carbon cycling model AMG to simulate soil carbon dynamics down to a 30 cm depth from 2005 onwards. Based on site-specific exports and imports and estimated residue returns, the model predicted a SOC stock decline larger than the observed one in the 0–30 cm depth, stabilising around 2028, assuming management stays the same in the future. By 2040, SOC stocks are projected to decline to 6.9 kg C m−2, representing an approximate 15 % reduction from the 2005 baseline. Furthermore, the AMG simulation was also consistent with the carbon flux balance reported by Loubet et al. (2011) for the period between 2006 and 2010. The observed decrease in SOC stocks may be attributed to a shift towards larger exports, lower residue returns, and reduced carbon imports at this site compared to past management practices. This study highlights the importance of high-quality SOC stock change monitoring, as developed within the ICOS research infrastructure.