Application of most-probable-number statistics to direct enumeration of microorganisms.

A novel method for rapid determination of total microbial cell numbers was investigated. The method involves the application of most-probable-number estimation statistics to direct microscopic counting of microbial cells by using a particle sizing graticule. Its accuracy and reliability were tested with computer simulations of bacterial cell distributions encountered in ecological studies. Good estimates of cell numbers were obtained when the cell density varied from 3 to 6,000 cells per field, i.e., over 3 orders of magnitude. Low levels of contagion did not markedly influence cell estimates, although high levels, corresponding to discrete scattered microcolonies, did. However, these could be recognized visually. Estimates of cell numbers in Breed smears confirmed its speed and good correlation with the standard quadrat counting technique under real experimental conditions.