Spatially explicit demography in a little owl population: mapping the fitness landscape
It is widely assumed that animals select habitat in a manner that increases their reproductive output and that habitat selection enhances fitness. However, habitat quality is usually derived from spatial variation in animals’ distribution (density or intensity of space use), and is rarely inferred from spatial variation in vital rates. Animals can maladaptively select habitats where mortality is higher than survival (ecological trap or sink habitat), and fitness can therefore be a more accurate measure of habitat quality. Following a decline due to the disappearance of suitable nesting sites, the little owl population in central Europe has been in the last 20 years due to the provision of nest boxes. However, individual little owls could select for nest boxes in otherwise poor habitat, and face an ecological trap. To be able to identify and predict attractive sinks or ecological traps would therefore be invaluable for the applied conservation of the little owl. The aim of this master thesis is to link demographic data to space use data of little owl in order to identify landscape heterogeneity in reproductive success.
This project will build up on existing little owl demographic and space use data that have been collected in a population in Baden-Württemberg, Germany. In a first step, the student will refine habitat selection models evaluated using space use data (telemetry relocations in the study area) by linking them to demographic data of breeding success. The outcome will be used to identify and map possible mismatch between space use-defined habitat quality, and fitness-defined habitat quality. In a second step, the student will use experimental supplemental feeding data to understand how this could have manipulated the fitness-landscape in the study area.
This project is desktop-based, with no fieldwork component. A strong will to spend time in front of the computer analysing data is therefore mandatory. Previous knowledge of R and/or GIS (Arc, QGIS, other) would be an advantage. The student is expected to register at the University of Zürich for the thesis.