Changes in soil C and N stocks and C:N stoichiometry 21Â years after land use change on an arable mineral topsoil

- Soil stocks of C and N increased 21 years after land use change from arable.
- Soil C and N stocks increased most under rough grassland rather than trees.
- Total stocks of C were similar between tree and rough grassland systems.
- Soil C:N narrowed in all land uses over time.
- The final C:N exhibited remarkably robust stoichiometry across land uses of 13.6.

The sequestration of excess atmospheric C into resilient and long-lasting belowground pools is of increasing global importance to mitigate greenhouse gas emissions. One land use that is particularly amenable to this type of manipulation is agricultural land, which often has high sequestration potential. However, this capacity can depend on local factors such as cropping history, soil type and climate. Critically, it may also be limited by N availability. In this study we used the retrospective (repeated measures) methodology to assess the impact on previously arable land of land use change by either afforestation with two species of broadleaf trees planted at 800 or 1600 stems ha− 1 (T800 and T1600), or reversion to rough grassland (NT) for 21 years. We quantified the concentration, distribution and total stocks of organic C and N in the upper 0-30 cm of a common soil type found in NE Scotland and investigated the robustness of C:N to land use change. Finally we estimated ecosystem stocks of C and N, and how these were partitioned between plant and soil components. We found increases in the overall concentrations of soil C from 4.6% to 5.8%, and N from 0.32% to 0.43%. The increase in soil C stocks over the experiment was in the order NT > T800 > T1600 and each treatment differed significantly. The same pattern was seen for increases in N stocks but here the increases in NT and T800 were significantly greater than for T1600. Overall, stocks were higher in the rough grassland plots than under trees by 35 Mg ha− 1 for C and 2.2 Mg ha− 1 for N. These increases in stocks were accompanied by a highly significant narrowing in C:N with time across all treatments from 14.6 to 13.6 and differences seen between upper and lower soil layers in 1991 had disappeared by 2012. From an average of 151 Mg ha− 1 in 1991, the system C stock (soil + plants) had increased to between 202 Mg ha− 1 (NT) and 221 Mg ha− 1 (T1600) by 2012, with between 96% (NT) and 73% (T1600) of the C in the soil. Concurrently the system stock of N had increased to between 14.1 Mg ha− 1 (NT) and 11.3 Mg ha− 1 (T1600), from an average of 9.4 Mg ha− 1 in 1991, with between 99% (NT) and 81% (T1600) of the system N in the soil. Although in 2012 there were significantly greater soil C stocks in NT, this was offset by C accumulation in the treatments containing trees, such that overall there were no treatment differences. However, this was not seen with system N stocks in 2012, which were significantly larger in the NT treatment than in those with trees. Mean annual rates of C and N accumulation in the systems were greatest in NT (2.7 and 0.22 Mg ha− 1 yr− 1) and least in T1600 (0.7 and 0.08 Mg ha− 1 yr− 1). These results are important in the context of land use strategies aimed at pollution mitigation, such as C sequestration and nitrate leaching. They are also relevant to the possible effects of land use change, especially reversion to agricultural use, of land previously taken out of production.