Abstract:
The objectives of this study were to investigate the efficacy of two sanitizers, i.e. hypochlorous and peracetic acid, in reducing total bacteria (TPC), coliforms and Escherichia coli levels in tomato and lettuce and to mathematically predict the relationship among the initial load, types of vegetable, types and concentration of sanitizer, and residual micro-organism levels after the washing and sanitization processes by applying a set of artificial neural networks (ANNs). Tomato fruits and lettuce leaves were inoculated with bacteria, dried and then soaked in the sanitizers for 10 minutes, in 30 repititions. The concentrations used were 25, 50 and 75 ppm for hypochlorous and 30, 40 and 50 ppm for peracetic acid. In tomato, hypochlorous at all concentrations reduced TPC load from an initial ~6.0 log[subscript 10] cfu/g to a residual level of 3.84 - 3.67 log[subscript 10] cfu/g (p> 0.05); while peracetic acid reduced TPC to 3.16 - 2.27 log[subscript 10] cfu/g (p≤0.05). Under the same conditions, coliforms showed a reduction from an initially higher level of ~6.3 log[subscript 10] cfu/g to residual levels of 3.71 - 3.22 log[subscript 10] cfu/g (p≤0.05) by hypochlorous; and 2.66 - 2.30 log[subscript 10] cfu/g (p> 0.05) by peracetic acid. The same treatments reduced E. coli from ~ 6.2 log[subscript 10] cfu/g to 3.96 - 3.16 (p≤0.05) and 2.37 - 1.75 log[subscript 10] cfu/g (p>0.05), respectively. Whereas in lettuce samples, hypochlorous and peracetic acid gave TPC reductions from ~6.8 log[subscript 10] cfu/g initial load to a residual level of 6.38 - 6.21 log[subscript 10] cfu/g (p≤0.05) and 5.65 - 5.43 log[subscript 10] cfu/g (p>0.05), respectively. Coliforms were reduced from ~6.9 log[subscript 10] cfu/g to 6.07 - 5.59 log[subscript 10] cfu/g (p≤0.05) and 4.92 - 4.57 log[subscript 10] cfu/g (p>0.05), respectively. Likewise, E. coli on lettuce leaves was reduced from ~ 6.9 log[subscript 10] cfu/g to a residual level of 5.95 - 5.67 log[subscript 10] cfu/g (p>0.05) by hypochlorous and 4.83 - 4.49 log[subscript 10] cfu/g (p≤0.05) by peracetic acid. The results also indicated that 75 ppm hypochlorous and 40 ppm peracetic acid were the most efficient concentrations for reduction of the 3 types of microorganism. The best sum square error from the artificial neural prediction of residual TPC, coliforms and E. coli were 0.72, 0.50 and 0.83, respectively, and the maximum R[superscript 2] of residual TPC, coliforms and E. coli were 0.76, 0.85 and 0.72, respectively. Only one hidden layer and four, three or five hidden neurons were respectively required for TPC, coliforms and E. coli. Factor analysis in the ANNs models supported intuition in that the residual microorganism level depends upon the initial load of microorganism, type of vegetable, type and concentration of the sanitizer used. When using 75 ppm hypochlorous and 40 ppm peracetic acid treatments on lettuce and tomato, then storing at 10 ℃] and 22 ℃], there was significant difference (p≤0.05) between the two sanitizers. On the other hand, when tomato, lettuce, kidney bean and baby corn were stored at different temperatures, it was found that at 10℃], there was no significant difference (p≤0.05) at any time (2, 4 and 8 hours); with the exception of E. coli in baby corn, which showed an increase of around 0.24 log[subscript 10]cfu/g. Whilst at 22 ℃], there was significant difference (p≤ 0.05) at any time, and all types of microorganism increased around 0.7 1.3 log[subscript 10]cfu/g. The best sum square error from the artificial neural prediction of residual TPC, coliforms and E. coli were 0.22, 0.24 and 0.22, respectively, and the maximum R[superscript 2] of residual TPC, coliforms and E. coli were 0.94, 0.92 and 0.95 respectively. Only one hidden layer and eight, six or five hidden neurons were respectively required for TPC, coliforms and E. coli. Factor analysis in the ANNs models confirmed that the final microorganism level.