关键词: 紫花苜蓿, 高通量测序, 细菌群落结构, PICRUSt功能预测

Abstract:

A field study was conducted to investigate crop land and alfalfa land established for different lenghths of time ment years （L₂₀₀₃， L₂₀₀₅ and L₂₀₁₂） on soil bacterial communities and ecological function prediction. Soil bacterial communities were identified by high-throughput sequencing 16S rRNA gene amplicons. The Illumina MiSeq high-throughput sequencing technology was used to explore the structure and diversity of bacterial communities under three treatments， and statistical methods （such as redundancy analysis） were used to explore the relationship between soil physical and chemical properties and bacterial community structure and diversity. PICRUSt software was applied to predict the ecological function of soil bacterial communities present in different treatments. The results indicated that the predominant taxa at the phylum level are Actinobacteria， Proteobacteria， Acidobacteria and Chloroflexi of bacteria in loess soil， comprising 20.34%-32.40%， 18.99%-23.14%， 12.50%-13.39%， and 11.41%-12.55% of the microbial population， respectively. The relative abundance of Actinobacteria， Proteobacteria and Chloroflexi was higher in farmland soil than in alfalfa soil. The relative abundance of Actinobacteria showed a decreasing trend with increase in the stand age， and the relative abundance ofProteobacteria and Chlorophyta initially increased and then decreased with increasing stand age. The relative abundance of Acidobacteria did not differ significantly between farmland and alfalfa soil. The dominant bacterial genera of this loess soil were Gaiella， Nitrospira， Pseudarthrobacter and Solirubrobacter， which comprised 0.65%-3.33%， 1.52%-2.34%， 1.36%-2.61% and 1.03%-2.24%， respectively， of the microbial population. Compared with farmland， the relative abundance of the Solirubrobacter genus was significantly increased in alfalfa land. Redundancy analysis showed that soil total phosphorus is the main factor affecting the soil bacterial community structure. PICRUSt function prediction analysis indicates that the bacterial microbiota of the loessial soil has a total of 46 sub-functions， of which metabolism is the most important function， accounting for 69.20% to 70.22% of activity. Soil metabolism， genetic information processing， and biological system functional gene abundance of alfalfa soils were significantly higher than those in farmland soil， and these characteristics were specifically reflected in carbohydrate metabolism， exogenous substance degradation and metabolism， terpenoid and ketone metabolism， endocrine system， nervous system and substance dependent functional genes. This study emphasizes the bacterial community structure and metabolic function in loessial soil and the findings provide a strategy to enhance the sustainability of alfalfa fields through soil microbial community management.

Key words: Medicago sativa, high-throughput sequencing, bacterial community structure, function prediction with PICRUSt