An investigation of the ecology of water distribution systems.

Animal infestations of water distribution systems are a cause of considerable concern for the water industry. The appearance of an animal at a consumer's tap often gives rise to complaints. This study, in Anglian Water, Oundle Division, Northamptonshire, U.K. investigated the ecology of the infesting community and its management. Mains sampling used sequential flushes at the hydrant and a model for animal removal was constructed. Tap sampling collected animals from filters at the tap and an index of emergence (density in a day's consumption) and degree of infestation (based on density and the ability to cause complaints) were devised. Reservoir sampling used a perspex trap to sample the water column and sediments. Methods to determine the environmental and operational characteristics of the system were also devised. Community clustering on the basis of animal density determined five distinct community types. Operational and environmental parameters did not coincide with any of the types, though correlations indicated that highest densities occurred at sites further (in time) from treatment with a low turnover time. Sites fed by slow sand filters were not recolonised by chironomid larvae nor to the degree of other sites by other taxa. No successional sequence was found and it was concluded that animals recolonised on the basis of a 'competitive lottery'. The relationship between density at the tap and in the main was considered and seasonal changes in the degree of infestation between sites confirmed that water from rapid gravity filtration leads to more severe animal problems, particularly during the summer. Benthic animals penetrated treatment in low numbers, but reservoirs near treatment were dominated by limnetic animals. Many died and became an indirect food source for infesting animals, but some survived and colonised pipes. Reservoirs at the extremes of the system were not influenced by these taxa and were 'extensions' of the distribution system.