Hybrid response surface methodology (RSM) – artificial neural network (ANN) optimization of glycerol-based rhamnolipid production by a crude-oil-adapted Pseudomonas aeruginosa SC1

Rhamnolipid production by a crude-oil-adapted Pseudomonas aeruginosa SC1 strain was statistically optimized using glycerol as the carbon source and ammonium sulfate as the nitrogen source, and the relationship between biomass formation and biosurfactant synthesis was evaluated. The strain, isolated from crude-oil-contaminated soil using Bushnell–Haas medium, was cultivated in a defined mineral medium. Response Surface Methodology (RSM) based on a Central Composite Design (CCD) was employed to optimize four independent variables: temperature (25–40 °C), agitation speed (100–500 rpm), carbon-to-nitrogen (C:N) ratio (10:1–30:1), and inoculum size (1–10% v/v). A total of 30 experimental runs were performed, with rhamnolipid concentration and biomass as response variables. Rhamnolipid production varied from 1.45 to 2.02 g L-¹, while biomass concentrations ranged from 2.5 to 3.8 g L-¹ across the CCD experiments. The optimal conditions predicted by the RSM model were 32.5 °C, 300 rpm agitation, a C:N ratio of 20:1, and a 5% inoculum size. Replicated centre-point experiments under these conditions yielded highly reproducible rhamnolipid titers of 1.95–2.02 g L-¹. A strong positive correlation between biomass and rhamnolipid concentration (r ≈ 0.80) suggested a partially growth-associated pattern of biosurfactant production under the tested fermentation conditions. The quadratic RSM model demonstrated excellent predictive accuracy, with a mean absolute deviation of approximately 0.046 g L-¹. An Artificial Neural Network (ANN) model developed using the same CCD dataset showed superior predictive performance, further validating the optimized conditions. The optimized process enabled efficient rhamnolipid production (~2.0 g L-¹) under shake-flask conditions, providing a robust foundation for scale-up and applications in hydrocarbon-contaminated environments.