Multi-source information fusion usually deals with redundant sources; the goal is to aggregate decisions, or evidence which supports one or another class. Many frameworks are then available: OWA, fuzzy integrals, etc. In this paper we are interested in managing complementary information sources. The integration is done using a fuzzy rule base. The rules are generated using a fuzzy decision tree algorithm. The application consists in the design of a crop harvest monitoring system. The three information sources are a time series of satellite images, a crop growth model and expert knowledge. The results show that the system is ready to use in an operational way.