Habitat selection by Eurasian lynx (Lynx lynx) is primarily driven by avoidance of human activity during day and prey availability during night

Abstract

The greatest threat to the protected Eurasian lynx (Lynx lynx) in Central Europe is human‐induced mortality. As the availability of lynx prey often peaks in human‐modified areas, lynx have to balance successful prey hunting with the risk of encounters with humans. We hypothesized that lynx minimize this risk by adjusting habitat choices to the phases of the day and over seasons. We predicted that (1) due to avoidance of human‐dominated areas during daytime, lynx range use is higher at nighttime, that (2) prey availability drives lynx habitat selection at night, whereas high cover, terrain inaccessibility, and distance to human infrastructure drive habitat selection during the day, and that (3) habitat selection also differs between seasons, with altitude being a dominant factor in winter. To test these hypotheses, we analyzed telemetry data (GPS, VHF) of 10 lynx in the Bohemian Forest Ecosystem (Germany, Czech Republic) between 2005 and 2013 using generalized additive mixed models and considering various predictor variables. Night ranges exceeded day ranges by more than 10%. At night, lynx selected open habitats, such as meadows, which are associated with high ungulate abundance. By contrast, during the day, lynx selected habitats offering dense understorey cover and rugged terrain away from human infrastructure. In summer, land‐cover type greatly shaped lynx habitats, whereas in winter, lynx selected lower altitudes. We concluded that open habitats need to be considered for more realistic habitat models and contribute to future management and conservation (habitat suitability, carrying capacity) of Eurasian lynx in Central Europe.

Habitat metrics based on multi‐temporal Landsat imagery for mapping large mammal habitat

“…; Filla et al. ). Accounting for such variation in habitat models can provide useful information for wildlife management and conservation (Brambilla and Saporetti ; Frans et al.…”

“…; Filla et al. ). Likewise, phenology metrics characterizing forage availability and quality were important predictors in our habitat models for red and roe deer.…”

“…; Filla et al. ). While numerous habitat studies have used snow‐cover variables from coarse‐resolution sensors such as MODIS (Hebblewhite et al.…”

“…; Filla et al. ). Since some of the radio‐collared red deer are kept in enclosures during winter (Rivrud et al.…”

“…; Filla et al. ), we fitted separate models using presence data from the entire year (hereafter: year‐round models) as well as seasonal models using presence data only from winter or summer (Table ; hereafter: summer and winter models). Summer observations included all GPS recordings collected between 1 April and 31 October, winter observations all recordings between 1 November and 31 March (Godvik et al.…”