Abstract:Soil microorganisms probably represent the greatest reservoir of biological diversity in the world and play key roles in soils through regulating organic matter decomposition and plant nutrient availability. However, due to the complexity of microorganism survival condition, only 1% -5% of the total microorganisms can be isolated by cultural method in soils. Recently, the advances of molecular biological techniques, e. g., PCR DGGE (denaturing gradient gel electrophoresis) can provide information regarding soil bacterial community structure through the extraction of microbial DNA from soil, and bacterial community profiles can be generated through the PCR amplification of 16S rRNA genes. During this process, extract efficiency of soil genomic DNA is the most important step. At present, the most widely used methods to extract soil genomic DNA are frozen-thawing method and bead beating method that each has its unique property and advantage. Nevertheless, little research has been conducted to compare of the genomic DNA when dealing with different types of soil using these two methods, especially when soil is rich in humus. The objective of the work are to evaluate these two methods themselves in soil genomic DNA extract efficiency when dealing with high humus soil of Inner Mongolia steppe based on the PCR-DGGE analysis of bacterial community structure and to determine bacterial community through cloning and sequencing of the bands in the DGGE patterns. According to of the results of PCR-DGGE pattern using the bacterial primers 338F and 758R, we found that bead beating method is better than frozen thawing method in genomic DNA extract efficiency. Twenty-one bands in the DGGE pattern were selected, cloned and sequenced. Based on similarity matching, all the sequences formed five major clusters: Actinobacteria; α-, β-, γ-, Proteobacteria; Bacteriodetes; Gemmatimonadetes and Acidobacteria. Of the 21 clones obtained from DGGE patterns, YC4 exhibited 99.7% similarity to Pseudomonas sp. (DQ339153); YC5, YC18 and YC19 exhibited 99.9% similarity to Gram-positive bacterium (AB008510), Virgisporangium ochraceum (AB006162), Micromonospora chalcea (X92613), respectively.