Appl Microbiol Biotechnol (1991) 36:285-288 Applied Microbiology Biotechnology © Springer-Verlag 1991 Short contribution Optimization of the fermentation of olive mill waste-waters by Asperyillus niyer Moktar Hamdi, Habib BouHamed, and Radhouane Ellouz Centre de Biotechnologie de Sfax, BP-W, 3038 Sfax, Tunisia Received 21 May 1991/Accepted 17 July 1991 Summary. The fermentation of olive mill waste-waters (OMW) by Aspergillus niger was studied. On the basis of factorial design experiments, suspended solids and concentration of OMW, nitrogen source, sulphate and size of inocula were all found to be significant by af- fecting mycelium growth and chemical oxygen demand (COD) removal. Neither the absence of yeast extract, magnesium, sodium, potassium nor of calcium limited the growth of A. niger. With media lacking additional nitrogen and sulphate, the growth was limited. The op- timal inoculum obtained was between 106 and 107 spores/g COD. The highest biomass and the greatest COD removal were obtained with removed COD to N:SO42- ratios averaging 100 to 3:1.5. Introduction The extraction of olive oil, which is carried out in small, seasonally-operating agro-industrial units in the Mediterranean region, results in the production of high-density waste-water. The amount of waste pro- duced has been said to be equivalent to that of a popu- lation of more than 17 million people. The maximum biological oxygen demand (BOD) and chemical oxygen demand (COD) reach concentrations of 100 and 220 kg/m 3 respectively (Balice et al. 1982). Many prob- lems concerning the high toxicity and the biodegrada- bility of these effluents have been encountered during aerobic and anaerobic treatment processes (Balice et al. 1988). Aspergillus niger has yielded more satisfactory re- suits for the bioconversion of olive-mill waste-waters (OMW), as it hydrolyses pectins, polyphenols, and tan- nins and degrades many phenolic compounds (Kieslich 1976). In a recent study, .4. niger was used for OMW bioconversion in small-scale experiments (Hamdi et al. 1991). The aim of the present study was to optimize the parameters affecting the fermentation of OMW by A. niger. Offprint requests to: M. Hamdi Materials and methods Fungal strain. A. niger was isolated from OMW and maintained on a medium containing 50% OMW (v/v); NHnNO3, 5 g/l; (NH4)2SO4, 5 g/l; KH2PO4, 1 g/l, and agar (Difco, Detroit, Mich., USA) 18 g/l, on slants at 4° C. Media. The media were based on olive black water, the general composition of which has been described by Balice et al. (1982). The OMW was diluted and/or substituted according the experi- mental design. The suspended solids of OMW were eliminated by filtration through a Terylene fabric sieve (100-200 p~m). Fermentation procedure. Small-scale experiments were carried out in 1000-ml erlenmeyer flasks containing 150ml medium on a shaker operating at 150 rpm. The pH of the sterilized medium was not adjusted. The OMW was inoculated with 107 spores/g COD. Cultures were incubated at 35° C. The biomass was collected by filtration through a Terylene fabric sieve and/or by centrifugation at 40009, and the mycelium washed with distilled water. Proteins and nitrogen were measured using the Kjeldahl method (N x 6.25). After separation from the biomass, the COD was de- termined as described by Knechtel (1978). Sulphate was measured gravimetrically according to standard methods (APHA 1975). Statistical methods: An experimental design (Deshayes 1980) was used to study the effect of eleven factors on the growth of A. niger on OMW. Each factor had two levels. Using Hadamard's matrix on eleven factors, twelve experiments (N= 12) were necessary. Each erlenmeyer flask contained culture medium as required by the matrix. In the matrix, the middle value of the answers is: Yl + " " + Y12 bo ~ 12 where Y1 to Y12 are the experimental responses. The results were obtained with polynomial models of the follow- ing form: Yi =bo + bi Xi , where Y~is the experimental value of experiment i and Xi is the explanatory variable value (value of parameters). The coefficient bi serves to estimate the effects of the factors tested: X,'Y~ bl = 12 ' where Xi' is the sign of the matrix (+ or -).