Physicochemical and biological factors determining the patchy distribution of soil water repellency among species of dominant vegetation in loess hilly region of China

Soil water repellency (SWR) is a physical phenomenon whereby water cannot penetrate or has difficulty penetrating the soil surface. There are many factors involved in its occurrence, but the main factors controlling its emergence in loess remain unclear. In this work, we have studied numerous physicochemical and biological factors functioning in different dominant vegetations (Pinus tabulaeformis Carr., Robinia pseudoacacia L., and Hippophae rhamnoides L.) in a loess hilly region by gas chromatography-mass spectrometry (GC-MS) and high-throughput sequencing techniques. We observed that more than 75% of the soils under Robinia and Hippophae are categorized as slightly or strongly water repellent, while nearly 50% of the soils under Pinus are categorized as severely to extremely water repellent. The relative concentrations of total free lipids in the soil in the same water-repellency class were Pinus > Robinia > Hippophae, where fatty acids, alkanols, and sterols were positively correlated with SWR, whereas alkanes were not. For the abundance and diversity index of bacterial and fungal communities, the three species ranked in the following order: Robinia ≈ Hippophae > Pinus. Thus, solvent-extractable polar waxes were indicated to be better preserved in water-repellent soils under Pinus due to lower microbial diversity than Robinia and Hippophae. Here, we demonstrate polar waxes to be the principal factor controlling SWR. Moreover, the dominant phyla of fungi varied greatly than those of bacteria under three vegetation types. Correlation analysis showed that the abundance of Actinobacteria in dominant bacteria increased with SWR. Nonmetric multidimensional scaling suggested the fungal community in different water-repellent soils under Pinus to vary more than those under Robinia and Hippophae. The indicator species mainly belonged to Actinobacteria in bacteria and Basidiomycota in fungi at the phylum level; this finding was further supported by the linear discriminant analysis (LDA) effect size (LEfSe). Additionally, GC-MS identified a small amount of ergosterol, a specific biomarker of fungi under Pinus. These pieces of evidence collectively reveal that severe to extreme SWR occurs under Pinus and appears to be the most influenced by fungi and actinomycetes when the topsoil is close to air drying. However, there is a need for further testing on different plant species or land use.