Data driven optimization of PGPR biomass yield for development of carrier based bioformulation for ameliorating salt...

Scale up of plant growth promoting rhizobacteria based bioformulation, remains challenging because of the complex and nonlinear nature of interactions between the components of formulation affecting plant growth promoting rhizobacteria (PGPR) biomass, resulting in suboptimal performance under stressed conditions. This study reports an integrated optimization approach using Response Surface Methodology (RSM) and Artificial Neural Network (ANN) modeling to systematically design and optimize the PGPR based bioformulation. CHRJH203, a PGPR, has high 1-aminocyclopropane-1-carboxylate (ACC) deaminase activity and its application can help ameliorate salt stress in Salvia hispanica (chia) seedlings. Different combinations of carrier materials and various concentrations of inorganic salts as magnesium sulfate (MgSO₄) and calcium phosphate (CaPO₄) were tested for optimization of its biomass. The results from optimization suggest that 1% MgSO₄, 0.125% CaPO₄, and 1.25 g Banana peel: Charcoal powder (BP: CH) was best together for the growth of bacteria with coefficient of determination (R²) observed was 0.9131. Further prediction using ANN of dataset, have showed the highest R² (0.9953) when compared to Response Surface Methodology (RSM) results and minimum root mean square error (RMSE) (0.0242) and mean square error (MSE) (0.000586) suggesting it to be the best model for development of bioformulation. The best combination of the formulation was validated in plant tests also. The bioformulation showed improved germination rate, root shoot length and overall growth of chia seedlings compared to untreated controls under same salt stress condition. This study demonstrates a scalable RSM-ANN based, data driven strategy that connects computational optimization with experimental validation, providing a simple, novel and practical approach for development bioformulation for sustainable agriculture.