Biological soil crusts decrease erodibility by modifying inherent soil properties on the Loess Plateau, China

- Biocrusts can improve soil organic matter and accumulate fine soil particles.
- Soil erodibility decreased with development of biocrusts.
- Biocrusts reduced sediment yield by about 90% through alteration of soil properties.
- Sediment loss is determined by biocrust composition and successional stage.
- Moss crust was a strong predictor of soil erodibility on the Loess Plateau.

Soil erosion and subsequent land degradation contributed to societal collapse in the past and are a leading cause of desertification in arid and semi-arid regions. Biological soil crusts (biocrusts) are ubiquitous living covers in many arid and semiarid ecosystems that have an important role in soil stabilization and erosion prevention. The “Grain for Green” ecological project improved vegetation recovery, and led to an extensive development of biocrusts across the Loess Plateau region in China, one of the most eroded regions in the world. The expansion of biocrusts was instrumental in reducing soil loss in a very large, severely eroded region of the Loess Plateau. We hypothesized that development of biocrusts would change soil organic matter (SOM) and soil particle size distribution (PSD), thereby reducing soil erodibility and soil loss. We sampled 56 sites that were passively revegetated grasslands on former croplands and 3 bare soil sites in the Loess Plateau region, and used the erosion productivity impact calculator (EPIC) model combined with simulated rainfall to test our assumption. The PSD and SOM content varied significantly among biocrust types and successional stages. The SOM content was 4 times higher in moss dominated biocrust and 1.5 times greater in cyanobacterially dominated biocrust than bare soil. More fine-particles (< 0.01 mm) and fewer coarse-particles (0.05-0.25 mm) were present in biocrusts than in bare soil. Modeled soil erodibility decreased significantly as biocrust biomass increased, mainly due to increase in SOM content, reducing the predicted soil loss by up to 90%. Finally, the prevalence of moss biocrust was a better predictor of soil erodibility than cyanobacteria in the Loess Plateau region. We conclude that biocrusts were a decisive factor for the initial reduction of soil erosion, which must be considered explicitly in models that aim to predict and manage soil loss on the Loess Plateau.