We propose a novel scheme that aims to determine the optimal number of clusters based on the field conditions and the positions of mobile sink nodes. In addition, we merge algorithms of tree-based index structures to form an energy-efficient cluster structure. A performance evaluation shows that the proposed method produces highly-balanced clusters that are energy efficient and achieves up to 1.4 times higher survival rates than the previous clustering schemes, under various operational conditions.