Mohamed Esivan, Siti Marsilawati (2022) Optimisation of co-culture fermentation of lactobacillus casei and propionibacterium jensenii in rice bran extract. PhD thesis, Universiti Teknologi Malaysia.
![]() |
PDF
566kB |
Official URL: http://dms.library.utm.my:8080/vital/access/manage...
Abstract
Co-culture fermentation is a fermentation process involving two defined microorganisms growing together in the same culture. A co-culture of lactic acid-producing bacteria (LAB) and propionic acid-producing bacteria (PAB) is beneficial in producing direct-fed microbial (DFM) products. The synergistic activity between LAB and PAB in co-culture fermentation can improve the survival of LAB and the growth of PAB. On this basis, the objectives of this study are two-fold. Firstly, the optimisation of co-culture fermentation involving Lactobacillus casei and Propionibacterium jensenii in the agricultural waste extract. Secondly, the development of an artificial neural network (ANN) predictive model for predicting the cell biomass concentration and the co-culture-specific growth rate. In the preliminary phase, two different substrates, namely rice bran and banana peel, were used in this study. This step was conducted to select the suitable carbon source for L. casei to grow and produce lactic acid for P. jensenii consumption. From the observation, rice bran was found more suitable as a carbon source and fermentation medium. Next, the co-culture optimisation of L. casei and P. jensenii fermentation was conducted using the one-factor-at-a-time approach. The fermentations were optimised for rice bran at concentration of 5% to 25% w/v; incubation temperature (30? to 42?); inoculation ratio (1:1 to 1:10 % v/v) and the initial pH (5.0 to 7.0). The optimum fermentation condition was obtained at 20% w/v rice bran concentration, incubated at 35? with an inoculation ratio of 1:4 % v/v and initial pH of 6.5. The optimum growth (2.74 g dry cell weight/L) was recorded after 96 hours of incubation. The highest viable cell counts for L. casei and P. jensenii were 9.10 log CFU/mL and 9.42 log CFU/mL, respectively. The optimum specific growth rate, µ obtained, was 0.41 h-1. The growth of L. casei and P. jensenii was compared to its monoculture fermentation, and it was found that the co-culture did not affect the growth of L. casei but helped maintain its survival. Moreover, P. jensenii gained benefits in the co-culture system, as its growth improved compared to during its monoculture. The ANN predictive model was developed using the multilayer perceptron and trained using the Levenberg-Marquardt training algorithm. Five input parameters, incubation time (h), the concentration of total reducing sugar (g/L), pH culture, incubation temperature (?) and inoculation ratio (% v/v), were used to train the network for the prediction of cell biomass concentration (g/L) and the co-culture specific growth rate, µ (h-1). The model has a low mean square error and high regression coefficient (R2) for the training and testing set, indicating the model is fit to predict the cell biomass produced and its specific growth rate during the co-culture of L. casei and P. jensenii. The structure obtained for ANN predictive model consist of five inputs, eight hidden nodes and two outputs, 5-8-2. The optimum predicted cell biomass concentration and the specific growth rate, µ, were 2.24 g dry cell weight/L and 0.51 h-1, respectively. In conclusion, this work provides a strategy to produce multispecies DFM through co-culture fermentation using rice bran and presented the first predictive ANN model to predict the cell biomass concentration and the co-culture-specific growth rate of L. casei and P. jensenii.
Item Type: | Thesis (PhD) |
---|---|
Uncontrolled Keywords: | lactic acid-producing bacteria (LAB), direct-fed microbial (DFM), artificial neural network (ANN) |
Subjects: | T Technology > TP Chemical technology |
Divisions: | Chemical and Energy Engineering |
ID Code: | 101589 |
Deposited By: | Widya Wahid |
Deposited On: | 26 Jun 2023 06:50 |
Last Modified: | 26 Jun 2023 06:50 |
Repository Staff Only: item control page