Detection histories for Lycalopex griseus, Lycalopex culpaeus, Leopardus guigna, and Canis familiaris and habitat destruction covariate data for Los Lagos Chile (2019)

In an increasingly anthropogenic world, species face multiple interacting threats. Habitat fragmentation and domestic dogs are two perturbations threatening terrestrial mammals globally. Our aim was to determine if (1) the spatial use of domestic dogs increases with habitat destruction and (2) whether domestic dogs and habitat destruction drive the spatial use of native carnivores in a heavily degraded agricultural landscape in the central valley/Andean foothills transition of Los Lagos, Chile.

We implemented a camera trap survey in a fragmented landscape comprised of native forest patches amidst a matrix of pastureland. We used single-species occupancy models to assess the impact of domestic dogs and habitat destruction on three mesocarnivores – the foxes, culpeo (Lycalopex culpaeus) and chilla (Lycalopex griseus) and the wild cat güiña (Leopardus guigna). Additionally, we compared temporal overlap of all study species (including domestic dogs).

Detection rates for both the foxes increased with domestic dog occupancy, while factors driving occupancy differed for each of the native species. We found that a 12% projected increase in domestic dog occupancy negatively impacted the spatial use of the culpeo. Habitat loss and fragmentation were positive drivers for chilla and domestic dog occupancy. The güiña did not respond to fragmentation and other habitat covariates or domestic dog occupancy. All native carnivore species were primarily nocturnal, while the domestic dog was almost entirely diurnal.

Methods
Detection histories were generated from species occurrence data gathered from a camera trap survey in the Los Lagos region of Chile in 2019. A 30-minute quite period was introduced for conspecifics. Covariate data was collected through field measurements (understory), from the camera traps (small mammal and dog trap success), and through digitization of satellite imagery.

The 'camtrapR' package was implemented in R to organize the camera data and generate the detection histories.