Intercalibration of the Acridine Orange Direct Count Method of Aquatic Bacteria

Intercalibration for counting planktonic bacteria by the acridine orange direct count (AODC) method was made among ten investigators of seven laboratories in Japan. The first calibration was made for one freshwater and two seawater samples sent to seven investigators belonging to five laboratories. Each investigator estimated bacterial abundance in these samples by the AODC method according to routine procedures of each laboratory. For the freshwater sample the estimates differed by a factor of 2.3 [coefficient of variation (CV), 3b%], while for the seawater samples they varied greatly by factors of 4.0 and 11 (CV's, 46 and 94%). In the second calibration nine investigators counted bacteria on the same AODC slides under the same optical condition. The estimates of nine investigators for three freshwater samples differed by factors of 1.8-2.5 (CV's, 24-29%), while for a seawater sample they differed by a factor of 17 (CV, 89%). Sampling error associated with contagious distribution of bacteria on a filter, and difference in the criterion for discriminating bacteria from nonbacterial particles were considered to be possible sources of variation in counting bacteria.