Clustering algorithms may improve the scalability of single-level flat topology overlay multicast protocols. The adaptive hierarchical clustering algorithm (AHCA) maps a flat topology to a hierarchical tree. However, AHCA trees almost always introduce a bad degree metric and are not suitable for data distribution. A prune-relocate operation and top topologies operation are proposed in this paper to generate OM-AHCA trees. The prune-relocate operation reduces the total levels of the tree and the average number of hosts in one top cluster. The top topologies operation reduces the degree of the data source. The numerical simulation shows OM-AHCA trees are a compromise between AHCA trees and single-level topology flat protocol trees. OM-AHCA trees optimize the overall performance of the single-level flat protocol and improve the degree metric of AHCA trees with larger cost and delay.