This work presents a traffic-based, mobility-aware clustering algorithm tailored for application scenarios typical of low data rate IEEE 802.15.4a networks and characterized by a hierarchical network organization. The proposed algorithm adapts to emerging traffic and mobility patterns in the network by selecting as cluster heads nodes that are the intended destinations for a large number of packets, and by associating to the same cluster nodes that send packets to the same destination. The performance of the proposed algorithm is compared by computer simulations with both a non-clustered approach and a clustering algorithm aiming at energy efficiency maximization, in a network of UWB nodes compliant to the 802.15.4a physical layer. Simulation results show that the proposed clustering algorithm improves both network throughput and energy efficiency in hierarchically organized low rate UWB networks.