Iterative techniques for characterizing marine bird habitats with time-series of satellite images

Demonstrating long-term habitat use of marine habitats by seabirds is often complicated by short-term changes in habitat locations, persistence, and age. This paper describes iterative techniques for characterizing non-static habitats, such as meso-scale (10-100 km) ocean eddies and fronts, using time-series of satellite images that define sea surface conditions. Seabird use of satellite-detected habitats was compared using survey data and imagery from the Gulf of Alaska and southeastern United States. Time-series examination of satellite images combined with long-term seabird censuses allow 1) estimation of the successional state (age) of some marine habitats, 2) detection of recurring habitats, 3) geographically-referenced measurement of habitat location and areal extent, 4) identification of consistently-used habitats, and 5) demonstration of time-dependent use by seabirds associated with seasonal or annual variation in habitat availability.