Abstract
Landfills represent a unique microbial ecosystem and play a significant role in global biogeochemical processes. The study of complex ecosystems such as landfills using DNA-based techniques can be advantageous since they allow for analysis of uncultured organisms and offer higher resolution in measuring demographic and metabolic (functional) diversity. However, sample acquisition and processing from refuse is challenging due to material heterogeneity. Decomposed refuse was used to evaluate the effect of seven sample processing methods on Bacteria and Archaea community structure using T-RFLP. Bias was assessed using measured richness and by comparing community structure using multi-dimensional scaling (MDS). Generally, direct methods were found to be most biased while indirect methods (i.e., removal of cellular material from the refuse matrix before DNA extraction) were least biased. An indirect method using PO₄ buffer gave consistently high bacterial and archaeal richness and also resulted in 28 and 34% recovery of R. albus and M. formicicum spiked into refuse, respectively. However, the highest recovery of less abundant T-RFs was achieved using multiple processing methods. Results indicate differences in measured T-RF diversity from studies of landfill ecosystems could be caused by methodological (i.e., processing method) variation rather than refuse heterogeneity or true divergence in community structure.