Viability and estimation of shelf-life of bacterial populations.

Mathematical concepts associated with the exponential and probit models are developed, and the similarity of the two methods is discussed. Because of its greater flexibility in design, the probit method was used to estimate the shelf-life for four bacterial populations, wet and dry spores of Bacillus anthracis and wet and dry cells of Pasteurella tularensis. On the basis of data gained by storing these organisms at high temperature, the probit method was used to predict the time at which 50% viability would occur for cells stored at 3 C. A plane passing through a three-space showing change in percentage viability of bacteria was formulated by a multiple regression method. With this functional technique, the percentage viability, expressed as a probit, was linearily related to a logarithm of storage time and storage temperature. The use of this method to study the effect of controlled variables on the viability of cells is demonstrated by comparing the effect of viability associated with three additives used prior to drying. The results of the test gave shelf-life estimates which were too low for all additives; however, the order of stability was ranked properly as confirmed by long-term tests.