Acta Scientiarum Polonorum Technologia Alimentaria

Background. The Agaricus bisporus industrial wastewater that contains a variety of nutrients which could be used as culture for some beneficial microbiology will be one threat to our environment if the wastewater doesn’t be comprehensively utilized. Plackett-Burman is the rapid and concise ways of screening the main effective factors. Box-Behnken response surface method is used to optimize interactions between the three main factors and predict optimal fermentation conditions. This study is aimed to select the main influence fac- tors and optimize the conditions for culturing Saccharomyces cerevisiae in A. bisporus industrial wastewater by Plackett-Burman design and Box-Behnken response surface method.

Material and methods. We analyzed the total number of living S. cerevisiae in the fermentation broth using multispectral imaging flow cytometry. Plackett-Burman design was used to screen out three factors from the original six factors of processing wastewater concentration, initial pH, inoculum size, liquid volume, culture temperature, and rotation speed that affected the total number of viable S. cerevisiae. Factors significantly affecting the total number of viable S. cerevisiae, including culturing temperature, processing wastewater concentration, and initial pH were investigated.

Result. The results indicated that culture temperature (p = 0.0007) and pH (p = 0.0344) as negative factor and concentration (p = 0.0080) as positive effect were the significant factors affecting the total number of S. cere- visiae, inoculum (p = 0.1237) and shaking speed (p = 0.2112) as positive effect and loaded liquid (p = 0.4811) as negative factor were important factors. The optimum conditions for S. cerevisiae fermentation in A. bi- sporus wastewater were a rotational speed of 150 rpm, a culture temperature of 25°C, an initial pH of 6.0,   a concentration of 8.4%, a inoculation volume of 8%, and a 100 mL liquid volume in a 250 mL flask, a culture time of 48 h. Under these conditions, the concentration of total viable yeast reached 1.04 ±0.02 × 108 Obj/mL which was at the 95% confidence interval of predicted model (0.89–1.14 × 108 Obj/mL).

Conclusion. The experimental model is reliable and the experimental results are of good stability. Variance analysis is performed to determine the adequacy and significance of the linear model. Thus, Plackett-Burman