THE ACCURACY OF THE PLATING METHOD FOR ESTIMATING THE NUMBERS OF SOIL BACTERIA, ACTINOMYCES, AND FUNGI IN THE DILUTION PLATED

During the crop seasons of 1936, 1937, and 1938, 1465 samples of field soil, held in the laboratory one day after crushing, were plated by the recognized technique in four replicates of one dilution for counts of fungi, and of a higher dilution for counts of bacteria and of actinomyces. A χ2 value was calculated for each set or counts. These values for each group of micro-organisms were distributed into classes, and the number in each class was compared with the theoretical for the Poisson series. The data for each year indicate that the fungal counts conform to expectancy on the basis of random sampling, and show that the method provides a reasonably accurate estimate of the population in the dilution plated capable of developing under the conditions of growth. Too many sets of counts of bacteria in each year yield high χ2 values. The counts of actinomyces conform to expectancy.In an attempt to determine the cause of this abnormality for sets of counts of bacteria, samples were plated on the afternoon of the day they were taken from the field. Three hundred and four samples plated in six replicates of one dilution, and another 100 samples plated in four replicates, yield χ2 values whose distributions conform to expectancy. Accordingly, the plate method provides a satisfactory estimate of the bacterial population of soil in the dilution plated if the procedure is carried out within six hours after sampling. Data on 88 samples plated on the day of sampling, on 88 samples held one day, on 88 samples held two to five days, and on 88 samples held eight to thirteen days show that the discrepancy between the actual and theoretical distributions of χ2 values becomes progressively greater at each successive period of holding the samples. Further, the data indicate that the area sampled, the season, the medium used and the technique of plating bear no relation to the abnormal variation in counts of bacteria on replicate plates.A record was kept of the presence of abnormal types of bacterial colonies and various genera of fungi on all plates from 468 samples plated one day after sampling and crushing during 1938. The data show that sets having pin-point colonies or spreading colonies of the Mucorales on one or more plates usually have high χ2 values. Counts on such plates should be excluded from the estimate of the mean number of bacteria in the sample. Likewise, the number of actinomyces colonies on each plate from these samples was recorded. The χ2 values for these counts were found to conform to expectancy, indicating that the factor or factors associated with a large number of high χ2 values for counts of bacteria does not affect the count of actinomyces in the same way.Percentage moisture and P values corresponding to χ2 values for the counts of bacteria obtained each year were correlated. The data yield coefficients that are not significant in each case.