We present a novel approach combining memristive hyperdimensional computing (HDC) with temporal logic reasoning for real-time agricultural intelligence systems. Our method leverages the natural parallelism and energy efficiency of memristive devices to process multi-modal agricultural data including IoT sensor readings and sentiment analysis from agricultural reviews. The proposed Enhanced Memristive-HDC framework achieves superior performance in crop classification tasks while maintaining ultra-low power consumption suitable for edge deployment. Experimental results on agricultural datasets demonstrate competitive accuracy with traditional machine learning approaches while offering significant advantages in energy efficiency, hardware scalability, and temporal reasoning capabilities. The system processes 8 different crop types using 8 IoT sensor modalities and integrates sentiment analysis from agricultural text data, making it suitable for comprehensive smart farming applications.