Eurasian lynx GLCs' characteristics for classification with random forest algorithm

Kill rates are a central parameter to assess the impact of predation on
prey species. An accurate estimation of kill rates requires correct
identification of kill sites, often achieved by field-checking GPS
location clusters (GLCs). However, there are potential sources of error
included in kill site identification, such as failing to detect GLCs that
are kill sites and misclassifying the generated GLCs (e.g. kill for
non-kill) that were not field-checked. Here, we address these two sources
of error using a large GPS dataset of collared Eurasian lynx, an apex
predator of conservation concern in Europe, in three multi-prey systems,
with different combinations of wild, semi-domestic, and domestic prey. We
first used a subsampling approach to investigate how different GPS-fix
schedules affect the detection of GLCs indicating kill sites. Then, we
evaluated the potential of the random forest algorithm to classify GLCs as
non-kills, small prey kills, and ungulate kills. We show that the number
of fixes can be reduced to from 7 to 3 fixes/night without missing more
than 5% of the ungulate kills, in a system composed of wild prey. Reducing
the number of fixes per 24-h decreased the probability of detecting GLCs
connected with kill sites, particularly those of semi-domestic or domestic
prey, and small prey. Random forest successfully predicted between 73%-90%
of ungulate kills but failed to classify most small prey in all systems,
with sensitivity (true positive rate) lower than 65%. Additionally,
removing domestic prey improved the algorithm’s overall accuracy. We
provide a set of recommendations for studies focusing on kill site
detection, which can be considered for other large carnivore species
besides the Eurasian lynx. We recommend caution when working in systems
including domestic prey, as the odds of underestimating kill rates are
higher.