The LoRaWAN network is widely employed in agricultural Internet of Things (IoT) applications requiring long-range wireless communication and low-energy consumption. However, challenges arise in dynamic environments like woodlands, where obstacles such as tree foliage disrupt the Fresnel zone and absorb signals. A previous study proposes a solution utilizing a LoRa Mesh Network (LoRa Meshnet) capable of establishing connections under tree canopies. Still, LoRa Meshnet is less battery-efficient than LoRaWAN. Thus, relying solely on LoRa Meshnet is inefficient in situations where LoRaWAN communication is affected by dynamic obstacles like seasonally varying foliage. To address this, we introduce a hybrid LoRa network architecture that utilizes LoRaWAN as the primary network and LoRa Meshnet as the backup, automatically switching between them. This approach ensures energy efficiency in sparse foliage situations by using LoRaWAN while guaranteeing stable data transmission in dense foliage scenarios with the use of LoRa Meshnet. Additionally, the system offers easy network deployment and cost-effectiveness.