Model | Sex | HR | H(HR) | H(burst) | M | D(burst) | D(night) | E(sites) | Id | Sex*HR | Sex*H(HR) | Sex*H(burst) | Sex*M |
|---|
A: length of burst | x | Â | Â | x | x | Â | Â | Â | x | Â | Â | x | x |
B: speed of burst | x | Â | Â | x | x | x | Â | Â | x | Â | Â | x | x |
C: nightly distance | x | Â | x | Â | x | Â | x | Â | x | Â | x | Â | x |
D: number of nests | x | x | x | Â | Â | Â | Â | x | Â | x | x | Â | Â |
- Sex models the effect of the individual being a male, and thus female was set as the reference level. HR refers to home-range size (100 % MCP). H(HR) and H(burst) refer to the proportions of different habitat types within the home range and within a 25 m radius buffer along the movement burst, respectively. Used habitat types were H1–H3 (see Methods). Month (M) from March to September was included as a categorical variable, with March set as the reference level. D(burst) and D(night) refer to the log-transformed durations of the burst and the nightly tracking period, respectively, whereas E(sites) refers to the log-transformed number of days during which the nest site of the individual was recorded. Id refers to individual that was used as a random effect to control for repeated measurements. In models A–C we log-transformed the response variables and applied a normal model, whereas in model D we applied Poisson regression with the log link function
