Summary: | Wolves are recolonising human-dominated landscapes, which increases conflicts as people express fear based on the perception that wolves are dangerous and unpredictable. Increasing knowledge about behavioural responses of wolves towards humans might improve appropriate management and decrease conflicts. Nowadays, behavioural responses can be monitored with high-resolution GPS data. I conducted trials in which wild, GPS-collared wolves were approached by humans. Based on these trials, I developed a standardised method for extracting variables such as flight initiation and resettling using statistical process control (SPC). Furthermore, I conducted a pilot study exploring the first results of experimental human approaches on wolves using GPS data. Experimental approach trials were conducted on six individual wolves over two years within three wolf territories along the border between Norway and Sweden. In this thesis, I compared two statistical process control methods for determining the flight initiation distance (FID) during wolf approaches: the upper control limit (UCL) and the changepoint method. The FID is often used to describe the magnitude of an animal’s response to a disturbance. Both methods performed similarly well for detecting the flight initiation. However, changepoint analyses increase reproducibility and comparability for this type of studies. Additionally, I used changepoint analyses to determine the moment of resettling. In the pilot study, I used changepoint analyses to extract both flight initiation and resettling. Human-caused disturbance could result in anti-predator behaviour, even in apex predators. The intensity of the flight response may be a function of the wolf’s perception of risk. Therefore, I measured flight distance, displacement, speed and straightness as measures of flight intensity. For 17 wolf-human interactions within this pilot study, I found a correlation between longer flight duration, distance, displacement and higher speed, indicating that these are associated with higher ...
|