This paper introduces a STIgmergy based Routing protocol (STIR) for Content Centric Delay Tolerant Networks. STIR makes the most of spatio-temporal interactions between nodes to set up bio-inspired gradient fields between content producers and content users. STIR routing follows this gradient field to efficiently route information in Delay Tolerant Networks. The validation of this protocol has been coupled with the definition of a new mobility parametric model that makes possible to easily express the preferential locations and movements commonly observed in real human mobility traces. Performance evaluations by simulations demonstrate that STIR delivers better performances than traditional protocols even in case of highly dynamic networks.