Factors affecting aerobic colony counts for bottled water

The Canadian Official Method MFO-15 for aerobic colony counts of bottled waters and ice requires pour plating with plate count agar (PCA) tempered to 40–45°C and incubating 48 h at 35°C. The performance of hydrophobic grid membrane filters (HGMF), counted by computerized counter after 48 h incubation on various media, was compared against MFO-15 for 31 water samples collected across Canada, and potential sources of error in MFO-15 were investigated. HGMFs stained with triphenyl tetrazolium chloride after incubation counted more reliably than those grown on fast green FCF media. Regular PCA and tryptic soya agar gave low counts with some samples. Best results were obtained with HGMFs on half strength PCA plus 0.1% pyruvate (mPCAP/2); growth was good on R2A medium, and on one-third, and one-tenth strength PCA (plus pyruvate), but colonies stains were weaker; growth and staining were generally poor for NWRI and mHPC media. Average HGMF counts on mPCAP/2 and R2A at 35°C were 6.3 and 7.6 times the MFO-15 count, respectively, when plates were incubated in a large (“walk-in” type) incubator, and 40.6 and 39.8 times when submerged in waterbaths. HGMFs on mPCAP/2 counted the most reliably. For pour plates, agar pouring temperatures between 40–45°C affected counts by as much as 2 log8cycles, with 42°C being optimum. Incubation temperature was critical with all of the methods. Count changes were small between 33 and 35°C, but count reductions of >3 log8cycles occurred between 35 and 37°C. Counts by MFO-15 when plates were incubated submerged in a waterbath averaged 31.3 times those if plates were in a walk-in incubator. The most representative 35°C aerobic colony count of bottled water would be obtained by using HGMFs plated on mPCAP/2 or R2A, mPCAP/2 being preferable for electronic counts. Unless plates are incubated in waterbaths, count reproducibility could be improved by incubating at 33° rather than 35°C; the difference in count would be small.