Bacterial inoculum density and probability of para-nitrophenol biodegradability test response.

This study has been carried out to establish a model linking probability of positive response in para-nitrophenol biodegradability test to controlled variables of the test (suspended solids, SS; total bacteria, AODC; cultivable bacteria, CFU; specific biodegraders, MPN). Series of dilution of 11 raw inocula (6 activated sludges, 5 river waters) were tested. They reveal very dispersed values of biomass measured as SS, AODC, and CFU and quite comparable values of specific biodegraders for each category of inoculum (river or sludge). The proposed model fits well the empirical distribution of the experimental frequency of positive results versus inoculum density for each controlled variable. The constants k of the model, representing the fraction of biodegraders for each inoculum, were tested by the likelihood ratio test and were proven to be different from one another according to the biomass descriptor and the origin of the inoculum. The probabilistic model, in the case of para-nitrophenol biodegradation, indicates that standardized official tests (closed bottle, AFNOR, Sturm, and MITI I) are seldom optimal under those conditions. It allows the determination of which inoculum concentration can lead to a high probability (e.g., 99.9%) of observing paranitrophenol biodegradation by raw inocula.