| Summary: | This study deals with the ecology of ciliate and bacterial populations in operational slow sand filter beds used as a stage in treatment of river water for drinking water supply. It differs from earlier research in this college on the meiofauna and protozoa of similar filter beds in the source of water, the use of micro-strainers for pre-treatment, the coarser composition of the sand and in the study of bacteria. Three beds were studied but only one of these was studied for 9 runs, starting immediately after resanding in March 1985 and finishing in December 1985. The other beds were investigated comparatively to assess the effects of excluding sunlight. The sampling programme was designed to provide a time-course of depth distribution of densities of the ciliate protozoan and bacterial populations as well as the abundance of particulate organic carbon and chlorophyll a in relation to measures of the bed's performance such as head loss and filtration rates, which were monitored by Thames Water. Fourteen species of Ciliatea were commonly recorded and were counted alive after extraction from the sand at different depths and cold sedimentation. The same depths were used to determine the concentration of particulate organic carbon and chlorophyll a as well as the density of bacteria shaken off the sand grains and counted by incidence epifluorescence microscopy. As only single cores were taken on each sampling occasion due to the need to sample frequently, at many depths and to count live organisms, a special study was undertaken to estimate the horizontal variability of protozoan and bacterial densities. Tests were made on the extraction efficiency ofthe protozoan and bacterial techniques. Scanning electron microscopyof the sand grain surfaces from different depths was used to determine the sizes and shapes of the bacterial forms present as well as to assess the efficiency of the shaking technique for their removal. Thelinear dimensions of the different ciliate species were also measured. This permitted the calculation of cell volume of each common species, using formulae for the most appropriate geometric shape. Three phases of development in ciliate depth distributions occurred in most runs: during the first four days, ciliate densitieswere low and uniform throughout the depths; this was followed by increased densities in the top few cms and a sharp decline with depth; after the third week of the run, a marked surface avoidance was observed. The bacterial populations and the concentrations of POC and chlorophyll a showed similar patterns of depth distribution with time, with the exception of the phase with surface avoidance. The bed exposed to normal levels of sunlight was more productive than the other covered beds which was shown by the bacteria, ciliates, chlorophyll a and POC. Peak ciliate densities occurred during the third week of a run in all three beds but the highest bacterial densities appeared earlier (13-14 days) in the unshaded bed compared with the shaded ones (31 days). In all three beds, a marked decline in ciliate densities coincided with an increased development of the bacterial populations. During the first ten days of a run when head loss increased slowly, the ciliate and bacterial populations were able to increase their densities exponentially accompanied by an increasing cumulation of POC and chlorophyll a. This was followed by a period when the head loss increased more markedly when the ciliates, bacteria and chlorophyll a concentrations respond in a variety of ways in different runs but the POC continued to cumulate until the end of the run. These changes were accompanied by a succession in the species of ciliate which was numerically dominant and there was a change-over from predominantly bacterivore to carnivore forms. Scanning electron-micrographs of the sand grain surfaces showed that the bacteria were attached by strong filaments and that apparently anaerobic fusiform bacteria appeared towards the end of a run and during periods of high temperature. The micrographs were also used to assess the efficiency of the shaking technique at removing bacteria from the sand grain surfaces. A laboratory filter model was used to perturb experimentally the simulated sand filter system by changing the filtration rates experimentally and determining the depth distributions of the ciliate fauna and bacterial flora. In general, the densities of organisms in the model filter were less than those found in lit operational beds,perhaps because of the difficulty of provding realistic levels of radiation.
|
|---|
