COMPARISON BETWEEN DIRECT COUNTS OF ACTIVE SOIL CILIATES (PROTOZOA) AND MOST PROBABLE NUMBER ESTIMATES OBTAINED BY SINGH DILUTION CULTURE METHOD

The abundance of active soil ciliates was determined by Singh's dilution culture method and by the direct counting method of Luftenegger et al. (1988) in 13 samples of 3 types of soil (meadow topsoil, spruce and beech litter). This comparative investigation revealed that the dilution culture methods are likely to overestimate active ciliates by orders of magnitude (with an average factor of 2000). One of the disadvantages of dilution methods is their difficulty meeting the assumptions of the MPN (most probable number) estimation model, on which they are based. A further serious problem is the proper separation of encysted and active cells. Due to a wide range of systematic errors, MPN estimates become seriously biased. But even if the basic assumptions could be met, the maximal precision of the estimate is far too low for a useful application within the context of soil ecology. This was revealed by the calculation of approximate confidence limits. Subjecting the data to analysis of variance showed that the direct counts can clearly differentiate ciliate abundances in forest litter versus meadow soil, whereas the dilution culture method is not precise enough to achieve this distinction. The MPN estimates did not even correlate significantly with the direct counts. Our investigation thus proved that dilution culture methods are beset with methodological deficiencies, resulting in unreliable estimates of protozoan abundances in soil.