We consider the problem of optimizing the number of replicas for event information in wireless sensor networks, when queries are disseminated using expanding rings. We obtain closed-form approximations for the expected energy costs of search, as well as replication. Using these expressions we derive the replication strategies that minimize the expected total energy cost, both with and without storage constraints. In both cases, we find that events should be replicated with a frequency that is proportional to the square root of their query rates. We validate our analysis and optimization through a set of realistic simulations that incorporate non-idealities including deployment boundary effects and lossy wireless links.