Advances in bio-logging technology for wildlife monitoring have expanded our ability to study space use and behavior of many animal species at increasingly detailed scales. However, such data can be challenging to analyze due to autocorrelation of GPS positions. As a case study, we investigated spatiotemporal movements and habitat selection in the little owl (Athene noctua), a bird species that is declining in central Europe and verges on extinction in Denmark. We equipped 6 Danish food-supplemented little owls and 6 non-supplemented owls in the Czech Republic with high-resolution GPS loggers that recorded one position per minute. Nightly home ranges, measured as 95% kernel density estimates, of Danish male owls were on average 62 ha (± 64 SD, larger than any found in previous studies) compared to 2 ha (± 1) in females, and to 3 ± 1 ha (males) versus 3 ± 5 ha (females) in the Czech Republic. Foraging Danish male owls moved on average 4-fold further from their nest and at almost double the distance per hour than Czech males. To create availability data for the habitat selection analysis, we accounted for high spatiotemporal autocorrelation of the GPS data by simulating correlated random walks with the same autocorrelation structure as the actual little owl movement trajectories. We found that habitat selection was similar between Danish and Czech owls, with individuals selecting for short vegetation and areas with high structural diversity. Our limited sample size did not allow us to infer patterns on a population level, but nevertheless demonstrates how high-resolution GPS data can help to identify critical habitat requirements to better formulate conservation actions on a local scale.

We captured 6 adult owls from both the Danish (June 2019) and Czech populations (June 2020) using mist-nets or traps baited with dead chickens (Denmark) or live mice (Czech Republic). Captured owls were sexed (based on brood patch and body mass) and fitted with a GPS backpack (Technosmart Europe srl., Rome; total mass including Teflon backpack harness of ca. 3.2 g; constituting <2.5% of the owls’ body mass). GPS units were programmed to record one position per minute, and to start recording positions the night after capture (avoiding potential capture effects).

The data shows the area (Denmark or Czech Republic), owl_ID, the latitude and longitude (GPS coordinates), the data and time, sex, distance moved between consecutive GPS positions, and distance from the nest (in m).