Adequacy of in situ glass slides and direct sand extractions to assess the microbiota within sand columns used for drinking water treatment

Historically, Cholodny-Rossi buried glass slide techniques have been used to study the microbiota of subsurface environments, yet the bias of such a technique has not been compared against direct sand extraction using modern in situ probing. Over a period of 34 wk, four separate 4-m-deep sand columns receiving raw lake water were examined to compare direct extraction of sand filter biofilm material against in situ glass slide biofilms. Significantly different DAPI direct counts and fluorescent in situ hybridization signals for major phylogenetic groups were observed. Not only were lower proportions (P < 0.001) of EUB338-probed DAPI cells observed on in situ glass slides, but also fewer gamma-Proteobacteria (12%-21%) and more alpha-Proteobacteria (16%-33%) when compared to direct sand extracts. Hence, investigators of the microbial ecology of even simple sand biofilms must consider the inherent biases from "accepted" methods and seek further independent methods to identify those which may be most accurate.