We examined variation in persistence rates of waterfowl carcasses placed along a series of transects in tundra habitats in western Alaska. This study was designed to assess the effects of existing tower structures and was replicated with separate trials in winter, summer and fall as both the resident avian population and the suite of potential scavengers varied seasonally. Carcass persistence rates were uniformly low, with <50% of carcasses persisting for more than a day on average. Persistence rate varied by carcass age, carcass size, among transects and was lowest in the fall and highest in the summer. We found little support for models where persistence varied in relation to the presence of tower structures. We interpret this as evidence that scavengers were not habituated to searching for carcasses near these structures. Our data demonstrate that only a small fraction of bird carcasses are likely to persist between searches, and if not appropriately accounted for, scavenging bias could significantly influence bird mortality estimates. The variation that we documented suggests that persistence rates should not be extrapolated among tower locations or across time periods as the variation in carcass persistence will result in biased estimates of total bird strike mortality.