Mature biocrust-covered soil carbon fluxes are dependent on their types: Moss-covered soils still serve as sinks while cyanobacteria-covered soils become sources

Soil carbon (C) fluxes in drylands are crucial for predicting future global climate change. Biocrusts covering dryland surfaces have been recognized for their roles in ecosystem functions. However, the contributions of biocrusts to soil C fluxes may vary depending on the crust-forming organisms, but evidence remains limited. In the semiarid climate of the Northern Chinese Loess Plateau, we conducted continuous measurements of soil C fluxes, including biocrust photosynthesis (P_n), soil respiration (R_s), and net C fluxes (NCF = R_s – P_n, where a positive value refers to a C source, and a negative value refers to a C sink), on two representative types of mature biocrust-covered (cyanobacteria and mosses; ∼40 years old) and bare soil over two growing seasons (2022–2023). Our results showed that mature moss crusts significantly increased P_n and R_s compared with cyanobacterial crusts. The NCF for cyanobacterial crusts was positive, whereas that for moss crusts was negative, with moss crusts showing significant net C uptake (–0.32 μmol m^–2 s^–1). Both linear regression and structural equation modelling showed that soil water content was the first determinant factor driving seasonal changes in C fluxes in biocrust-covered soil. Notably, the estimated average annual NCF from the random forest model indicated that mature moss-crusted soils had a slight net C uptake (–66 g C m^–2 yr^–1), whereas mature cyanobacteria-crusted soils (551 g C m^–2 yr^–1) exhibited remarkable net C emissions. In conclusion, mature biocrusts are important contributors to soil C fluxes in drylands; however, their contributions are highly dependent on the type of biocrust. Our study highlights that mature moss-crusted soils could increase dryland C uptake over extended periods (potentially decades or even centuries), and thus they may serve as potential ‘permanent’ C sinks in drylands, whereas cyanobacteria-crusted soils would reach C saturation sooner and subsequently act as C sources by reducing dryland C uptake. These divergent changes in dryland soil C fluxes caused by biocrusts, especially the differences among mature biocrust types, should be thoroughly considered to accurately assess dryland or global C fluxes for mitigating climate change.