Growth modeling of a γ-hexachlorocyclohexane-degrading microbial consortium based on chloride release kinetics

Nonlinear growth modeling can offer a more robust approach in curve fitting exercises compared to the traditional linear regression for modelling microbial growth processes. The models utilized include Huang, Baranyi-Roberts, modified Gompertz, Buchanan-3-phase, modified Richards, modified Schnute, modified Logistics, von Bertalanffy, MMF (Morgan Mercer Flodin), and the study evaluated these primary growth models to describe a bacterial consortium growth on lindane, with growth measured indirectly by chloride release. The results show that choosing the best model using visual inspection is inadequate for distinguishing between models, as all evaluated models demonstrated acceptable fits. A statistically and information-criterion-based discriminatory approach demonstrated distinct performance disparities. The Huang model consistently demonstrated superior performance compared to competing models, characterized by the lowest error values, the highest explanatory power, favorable information criteria, and minimal systematic bias. However, the modified Richards and modified Logistics models also exhibited competitive performance under specific criteria. The MOORA multi-criteria decision-making approach was utilized to mitigate the uncertainty associated with the majority voting approach. MOORA demonstrates that the Huang model is the most robust overall, with the modified Richards and modified Logistics models following closely behind. The parameters obtained from the Huang model for the chloride release kinetics at 10 μM lindane were Lag period (d or day), Ymax, and μm values of -3.132 (d), 9.235, and 0.934 (d-1), respectively. The utilization of the modelling exercise yielded important parameters for future secondary modelling exercises and preliminary prediction of performance and limitations in field studies.

Text 127262.pdf - Published Version Available under License Creative Commons Attribution. Download (709kB) Official URL or Download Paper: https://journal.hibiscuspublisher.com/index.php/JE...

Actions (login required)

View Item