The importance of fine root protection in topsoil carbon and nitrogen sequestration following land-use changes on sloping karst ecosystems.

[Display omitted] • SOC in disturbed grassland was similar to shrubland, while higher than in woodland. • Fine root protection was dominant process for soil C changes, instead of C inputs. • The dominant processes for soil N sequestration were similar to soil C. • The change in TN concentration was weak and lagged compared to that in SOC. • Complex and closed N cycling, and N 2 -fixing plant input caused lagged change in TN. Soil carbon (C) and nitrogen (N) sequestration are important for recovery and sustainability of degraded ecosystems. However, the contributions of different processes to soil C and N sequestration following land-use changes in karst areas remain unclear, hindering our ability to recover degraded ecosystems. Here, we selected three typical land-use types (natural shrubland; woodland: Cornus wilsoniana wanger , Cyclobalanopsis glauca , Eriobotrya japonica, and Zenia insignis plantations; disturbed grassland: prescribed burning grassland, aboveground vegetation removal grassland, and complete vegetation removal grassland) to analyse the relative contribution of leaf and fine root inputs and fine root protection to 10 cm topsoil C and N sequestration after a 12-year land-use change in karst slopes. The results showed that (1) soil C concentration in grassland with aboveground vegetation removal was insignificantly different compared to natural shrubland, while woodland with high-intensity preplanting disturbance decreased soil C concentration; (2) soil C dynamic changes from 2006 to 2017 were not significant in natural shrubland and disturbed grassland due to low C inputs rather than high C turnover, which was indicated by the C and N isotopic signature of plant and soil; (3) structural equation model confirmed that fine root protection was the dominant process for the changes in soil C following land-use changes, instead of aboveground or root C inputs derived from replanting vegetation; and (4) the dominant processes for soil N sequestration were similar to soil C. However, the changes in soil N concentration were weaker and lagged compared to soil C during land-use changes because N cycling was relatively complex and closed, especially with increased N inputs derived from N 2 -fixing plants, which were beneficial to increase soil N sequestration. Altogether, our results indicate that primary soil C and N protection strategy, combined with active N input, may be more advantageous for land restoration in karst areas. [ABSTRACT FROM AUTHOR]