Wireless sensor networks (WSNs) consist of battery-limited sensor nodes which have the ability of sensing the environment, communicating with other nodes and processing the data. Large number of sensor node deployment over a geographical area imposes some constraints on the retrieval of the data. The use of mobile sinks (e. g., unmanned aerial vehicle, UAV) is an effective solution method for such large-scale networks. However, depending on the path and altitude of the UAV, and the type of radios in use, coverage problem arises where some nodes cannot get connected to the UAV. In this paper, the coverage problem is examined where UAV is used as mobile sink node. On the basis of our analysis, a dynamic and distributed clustering approach is proposed. Evaluations are performed with a realistic simulation environment. Performance results show that proposed approach reduces the energy-consumption and construct more stable and well balanced clusters that connect the uncovered nodes to the UAV.