There is increasing attention, recently, to optimizing energy consumption in IoT-based large-scale networks. Extending the lifetime of battery-powered nodes is a key challenge in such systems and their various application scenarios. This paper proposes a new zone-based and event-driven protocol for saving energy in large-scale heterogeneous WSNs called TESEES (Threshold Enabled Scalable and Energy Efficient Scheme). The proposed protocol is designed to support network scenarios deploying higher levels of heterogeneity with more than three types of sensor nodes (i.e., four, five, and more). TESEES is a reactive version of the proactive SEES protocol, in which we leverage a novel state-of-the-art thresholding model on the zone-based hierarchical deployments of heterogeneous nodes to regulate the data reporting process, avoiding unnecessary frequent data transmission and reducing the amount of energy dissipation of the sensing nodes and the entire system. With this model, we present a general technique for formulating distinct thresholds for network nodes in each established zone. This mechanism allows for individually configuring the nodes with transmission settings tailored to their respective roles, independent of the heterogeneity levels, total node count, or initial energy. This approach ensures that each node operates optimally within the network. In addition, we present an improved hybrid TMCCT (Threshold-based Minimum Cost Cross-layer Transmission) algorithm that operates at the node level and ensures effective data transmission control by considering current sensor values, heterogeneous event thresholds, and previous data records. Instead of periodical data transmission, this hybridization mechanism, integrated with a grid of energy-harvesting relay nodes, keeps the zone member nodes in the energy-saving mode for maximum time and allows for reactive data transmission only when necessary. This results in a reduced data-reporting frequency, less traffic load, minimized energy consumption, and thus a greater extension of the network’s lifetime. Moreover, unlike the traditional cluster-head election in the weighted probability-based protocols, TESEES relies on an efficient mechanism for zone aggregators’ election that runs at the zone level in multiple stages and employs various static and dynamic parameters based on their generated weights of importance. This leads to selecting the best candidate nodes for the aggregation task and, hence, fairly rotating the role among the zones’ alive nodes. The simulation results show significant improvements in the total energy saving, the lifetime extension, and the transmitted data reduction, reaching 29%, 68%, and 26% respectively, compared to the traditional SEES protocol. Also, the average energy consumption per single round has decreased by 36%.