EFFECTS OF SEDIMENTS ON ESTIMATES OF BACTERIAL DENSITY

We compared direct epifluorescent microscopic bacterial counts using either acridine orange [AO; 3,6-bis(dimethylamino)acridinium chloride] or DAPI (4',6-diamidino-2-phenylindole) in the presence and absence of fine sediments. Differences in a time-series of population estimates obtained using either AO or DAPI direct-count methods were greater in sediment-amended cultures. In the presence of sediments, DAPI yielded significantly lower estimates of bacterial densities than AO. A sediment-addition experiment indicated that DAPI counts may be more sensitive to sediment presence than AO counts, although interaction between treatments (stain, sediment) clouds statistical interpretation. Variation between replicate cultures was greater than between subsamples of preserved combined samples. Mean differences attributable to choice of stain may be as great as differences attributable to potential masking effects of sediments. Despite culture variability, DAPI and AO methods yielded count estimates agreeing within an order of magnitude. Greater correspondence between counts was observed for turbid natural water samples spanning a more limited range of bacterial concentrations.