<p>All people on the planet expected agriculture to provide them with food to survive. The process of producing agricultural goods while integrating logistical elements is labor-intensive. Additionally, the lack of water exacerbated several issues with agricultural product production. The researchers did not adequately address the issues surrounding the lack of water over food chains. For supply chain management, this paper presented a novel deep learning based optimization mechanism to address the agricultural production challenges under water scarcity. Hence, this study proposed a Dynamic Salp Swarm Optimization (DSS) algorithm for feature selection. The DenseNet121 based Modified Runge–Kutta Optimization (MRKO) is used to predict the water usage. For supply chain optimization, unique technique called Beetle Swarm Optimization (BSO) proposed. In addition, this reduces production costs and time while increasing profitability. The Python simulator is used to perform simulations and implement the processes. According to the similar study, the proposed strategy outperforms all other methods in terms of supply chain manufacturing cost, time, accuracy, and recall for water consumption predictions. The overall sustainability enhanced and the agricultural supply chain water scarcity impact mitigated with these findings and it describes the potential of proposed work compared to state-of-art. From the comparative study, the proposed work demonstrated 96.81% accuracy, 95.7% precision and 96.8% recall with3.11% MSE, 2.10% MAE, and 3.10% RMSE results.</p> Graphical Abstract <p></p>