Methodology for estimating numbers of free-living and attached bacteria in estuarine water.

A fundamental problem in estuarine microbiology studies is the accurate determination of the density in the water column of both free-living bacteria and those attached to suspended particulate matter. When a water sample is filtered and the filter is viewed by epifluorescence microscopy, counts can be made of the numbers of bacteria which are seen on the filter background (free-living) and those which appear to lie on sediment particles (both free-living and attached). With only the additional knowledge of the proportion of the filter area covered by particles (a quantity that is straightforwardly determined by stereological point counting), results from geometric probability were used to determine the expected number of bacteria which are hidden by particles and hence to provide an estimation scheme for the true densities of free-living and attached bacteria. Variance equations based on a Taylor series are given, and a partial check of the method is attempted with controlled mixtures of bacteria and sediment. An alternative procedure is also proposed, in which the natural attached/free-living ratio is altered by an intervention experiment, allowing an estimation which is less model dependent but more labor intensive. Both methods are applied to a series of samples from the Tamar estuary, United Kingdom, taken in April 1985. A notable conclusion is that there are always more free-living than attached bacteria in the water column throughout the estuary.