The membrane filtration (MF) technique1 can be used with a variety of different media for the detection and enumeration of many types of bacteria, including coliforms. The medium used depends on the organisms targeted for detection. The MF procedure is reproducible, can be used to test relatively large sample volumes, and usually yields numerical results more rapidly than the multiple-tube fermentation technique.2 A distinct advantage of the MF technique over using broth media is that it permits colony formation on the surface of the filter. Direct counts of bacterial colonies are useful for various readily filterable sample types including drinking water, some wastewater effluents, natural waters, and others. Also, MF permits the differentiation of the various colony morphologies of bacteria present in the water sample and allows for their subsequent subculture and identification. Additional advantages include the capability of processing larger sample volumes (providing more representative samples) and of analyzing a larger number of samples because the culture dishes require less room in the incubator.2
When used for enumerating coliforms, the MF technique is a more precise method than the multiple-tube fermentation procedure, because it uses direct counts as opposed to a statistical estimation (MPN) of results; compare Tables 9221:3 and 9221:4 with Table 9222:3. The data from these tests yield approximately the same qualitative water-quality information, although the numerical results from each method are not interchangeable.3
The MF technique has limitations, particularly when testing waters with high turbidity, because the filter can become clogged, resulting in the need for subsamples and multiple membrane filters or difficulties in counting colonies with debris on the membrane filter. In addition, if the sample being filtered contains large numbers of noncoliform (background) bacteria, the morphologies of the coliform colonies can be difficult to discern for counting.2,4 If noncoliform bacterial interference occurs, the sample result must be invalidated and a new sample collected.
Another disadvantage of the MF technique when compared with broth culture methods is that bacterial numbers are generally lower because of the stress of the filtration process on cells and because nutrients must diffuse through a filter and thus are not as readily available.5