Periodical collection of data from vehicles inside a target area is of interest for many applications in the context of Intelligent Transportation Systems (ITS). Long Term Evolution (LTE) has been identified as a good candidate technology for supporting such type of applications - particularly for the non-safety domain. However, a high number of vehicles intermittently reporting their information via LTE can introduce a very high load on the LTE access network. In this context, the use of heterogeneous networking technologies can yield significant offloading of LTE - here, WLAN and Dedicated Short-Range Communication (DSRC) technology can support local data aggregation. In this paper, we propose an on-the-fly distributed clustering algorithm that uses both LTE and DSRC networks in the forwarder selection process. Our results clearly indicate that it is crucial to consider parameters drawn from both networking platforms for selecting the right forwarders. In particular, we show for the first time that relying on the Channel Quality Indicator (CQI) has a substantial impact. We demonstrate that our solution is able to significantly reduce the LTE channel utilization with respect to other state of the art approaches.