Abstract

To quantify spatial overlap, I use Bhattacharyya’s affinity to measure the difference between kernel density estimates. My first use of this method is for testing the differences between the space use of male and female gannets. To assess whether segregation is significantly different, I used a randomisation approach. Our results show that sexual segregation patterns differ between colonies and time periods. My second use of spatial overlap testing is for assessing the performance of a new method. Using a 6-year kittiwake GPS -tracking dataset from 13 UK colonies, I tested whether we can predict where birds forage from only information that could be collected at the colony or from the literature on past tracking studies. I developed a method using distances predicted by trip durations, along with departure bearings, to simulate locations. I could then calculate a kernel density estimate from predicted points and use Bhattacharyya’s affinity to measure how well they match up with kernel density estimates calculated using the real GPS data.