Routine monitoring of bacterial levels is an essential part of monitoring the quality of laboratory animal drinking water. The classic way to enumerate bacteria in water is to do a plate count which is to spread a known volume of sample on the surface of a laboratory medium and count the number of visible colonies that develop after a period of time. However, it should be recognized that plate counts may underestimate the total number of bacteria present in a watering system.

Most bacteria are in biofilms
Water samples only collect planktonic or free-floating bacteria. Free-floating bacteria in animal drinking water are either sloughed off of the biofilm or pass through from the incoming water supply. If a plate count test is low, one shouldn’t assume that bacteria are not present in the watering system. More than 99% of the bacteria in water systems are in biofilms attached to pipe surfaces. If the integrity of a mature biofilm hasn’t been disrupted by recent flushing or sanitization, it may not slough off many cells into the drinking water, but it is still there. As Smith (1987) puts it,

"When you take a water sample you are sampling only the "strays" and not the main "herd" of bacteria in the system."

Plate counts don’t detect all viable bacteria
Plate counts are based on the ability of bacteria in a sample to grow on a defined nutrient medium. When bacteria grow on a nutrient, they form distinct colonies. Theoretically, a colony is derived from a single bacteria cell. Some underestimation of bacteria is caused by clumps of bacteria that form only one colony.

Another reason viable counts can be too low in nutrient-poor purified water is that the bacteria are in a starved state and cannot grow on rich nutrient media. Rich laboratory media are toxic to bacteria adapted to living in high-purity water systems. To get higher bacterial recoveries from purified waters, special media (R2A agar), decreased incubation temperatures, and increased incubation times are sometimes used.

Understanding particle showers
Sometimes the results of bacterial plate count testing seem very erratic. Samples taken from one point in the system may vary from less than 10 cfu/ml to TNTC (too-numerous-to-count). Or maybe the counts are usually low, but occasionally a high count appears. Some of this variability can be explained by understanding that biofilms periodically "shed", causing bacterial counts to skyrocket. According to Patterson (1991), "Sudden failure of the integrity of the biofilm at specific locations will result in bacteria and particle showers that occur randomly in time."

Next Page