Description of the structural diversity of rumen microbial communities in vitro using single-strand conformation polymorphism profiles

Abstract

Changes of the rumen microbial community structure, as it can be established with a rumen simulation technique (RUSITEC) were studied using PCR and single-strand conformation polymorphism (SSCP) of small subunit rDNA genes (SSU rDNA). Four total mixed rations were incubated and two ammonia levels in the artificial saliva were applied. Three replicated vessels were used for each treatment. Mixed microbial fractions were isolated by stepwise centrifugation from the liquid fraction (reference microbes, RM) and from the solids of the feed residues (solid-associated microbes, SAM). PCR-primers targeting archaea, fibrobacter, clostridia, and bacteria, respectively, were applied to represent the individual taxonomic groups by SSCP profiles. These SSCP profiles were converted into a binary matrix and distances among treatments were visualised by non-metric multidimensional scaling. Between replicates belonging to one treatment only small differences were found, indicating a high reproducibility of the RUSITEC and the chosen SSCP method. The ammonia concentration seems to be affecting the SSCP profiles. Great differences occurred between RM and SAM, especially for profiles targeting bacteria and clostridia. Differences in the profiles of RM were also found between mixed rations that contained the same feedstuffs in different ratios and between rations with similar nutrient content but based on different feedstuffs. In conclusion, the PCR-SSCP-based technique in conjunction with non-metric multidimensional scaling was sufficiently sensitive to detect and compare changes in composition of rumen microbial community structure in vitro as affected by diet and other environmental factors.

Keywords:

• rumen
• microoganisms
• communities
• polymerase chain reaction
• scaling

Acknowledgements

This study was supported by the H. Wilhelm Schaumann Stiftung, Hamburg, which is gratefully acknowledged. The authors thank Marti J. Anderson (Department of Statistics, University of Auckland, New Zealand) for her assistance in the statistical analysis.