Internet of Things (IoT) and robotics technologies have reached a good maturity and they are being used in several fields, such as healthcare, manufacturing, urban planning, and agriculture. In the agro-ecology context, some solutions based on robots and sensors were proposed, for example for monitoring the status of the crops, detecting nutritional and health stress, and for management operations (such as weeding, optimal application of fertilizers, and crops’ protection treatments). Similar applications can be found in the animal farming domain. These solutions use Big Data technologies that require handling voluminous, real-time, and heterogeneous data generated by agricultural operations. Multiple architectures are proposed, with different characteristics specifically designed to address a particular farming application. Therefore, generic architecture guidelines for deploying agro-ecology application systems are missing. They could help researchers and engineers choose adequate data management tools for their farming applications. This paper discusses and presents a reference architecture of an agro-ecology application system to properly handle data needed in such a specific context.