The numbers of bacterial cells in field soils, as estimated by the ratio method

It has long been recognized that the method of counting bacteria in a soil sample by means of plate counts is valueless as an estimate of the total bacterial population, although such counts have undoubted value in comparing two or more samples. The numbers obtained by this method represent only a small fraction of the total population. This is due mainly to the fact that no single plating medium will enable all the diverse groups of soil bacteria to multiply and produce colonies. Indeed, many types will not grow on any of the media commonly used for plate counts, and it is very uncertain whether those which do so are of main importance in carrying out biochemical changes in the soil itself. The number of colonies which appear is probably reduced further by the fact that, in some soils, many of the bacteria occur in clumps which would not easily be broken up in the process of making the soil suspension. These limitations of the plate method make it very important that a technique should be developed for making accurate estimates of the total bacterial content of a soil sample from direct microscope counts. The counting of bacteria in dried and stained films has become a recognized technique for milk (Breed, 1911). A comparison between plate and direct microscope counts of milk samples was made by Breed and Stocking (1920) and more recently by Buice (1934) who found that the numbers obtained by two methods were of the same order. With milk, however, most, if not all, of the organisms counted were of types capable of growing upon the plating medium used. The application of direct counting methods to bacteria in a soil sample is greatly complicated by the presence of the soil particles. These introduce difficulties in staining the film, in obtaining an even distribution of bacteria over the film, and in estimating the mass of soil examined.