Wireless Sensor Networks (WSNs) play a critical role in Internet of Things (IoT) applications that offer automation and data exchange across various domains such as environmental monitoring, healthcare, and agriculture. WSN frequently suffers from unstable topology, energy limitations, and communication overhead that damage the network performance. To overcome these issues, this study introduced the Energy-efficient Mantis Search Optimization for Wireless Sensor Networks (EMSO-WSN) framework to enhance the routing efficiency in WSN-IoT. This model includes Fuzzy C-Means (FCM) for adaptive clustering, Adaptive Walrus Optimization (AWO) for efficient Cluster Head Selection (CHS), and Mantis Search Algorithm (MSA) for robust route optimization. This research focuses on improving the WSN in IoT applications that will enhance the network lifetime and energy efficiency. The proposed EMSO-WSN emphasizes multi-phase decision making, behavioral modelling, and soft clustering, ensures high scalability, and reduces latency. Finally, the experimental framework was simulated by Python using NS2 for fine-grain throughput and computational efficiency. The EMSO-WSN model is evaluated using key metrics such as Network Lifetime (NL), Energy Consumption (EC), delay, number of CH, Data Packet Delivery (DPD), and throughput. This shows the comparison of the proposed EMSO-WSN reduces less EC of 6 Joules at 100 nodes than that of other existing methods like SWARAM of 9 J, HHO-CFR of 10 J, and ECEEC of 11 J, respectively. The throughput of proposed EMSO-WSN achieves 58.75% higher than that of other existing like SWARAM, 45.11% higher than HHOCFR, and 22.73% higher than ECEEC. Thus, the EMSO-WSN framework is validated as a scalable and energy-efficient solution for modern WSN-IoT infrastructures