The Central American brocket deer is a vulnerable species. Geographically isolated populations have been affected by poaching and habitat fragmentation, leading to local extinctions. It is therefore important to understand this species' habitat characteristics, particularly of resting sites, which play a crucial role in survival and fitness. We describe the characteristics and distribution patterns of Central American brocket deer resting sites at the microhabitat and landscape scales in San

more »

... tolo Tutotepec, Hidalgo, México. We conducted eight bimonthly field surveys between November 2017 and March 2019, consisting of 32 transects of 500 m length to search for fecal pellets, footprints, scrapes, and browsed plants. At each resting site we identified, we measured canopy closure, horizontal thermal cover, protection from predators for fawns and adults, escape routes, slope from the ground, presence of scrapes, cumulative importance value of the edible plant species, and distance from the resting site to the nearest water resource to characterize the site at the microhabitat scale. At the landscape scale, we identified the type of biotope, elevation, aspect, and slope. We compared all of these parameters from resting sites with a paired randomly selected site to serve as a control. We performed a multiple logistic regression to identify the parameters associated with the resting sites and a point pattern analysis to describe their distribution. We characterized 43 resting sites and their corresponding control plots. At the microhabitat scale, resting sites were associated with higher vertical thermal cover, more concealment cover, more escape routes, more edible plant species, higher slope from the ground, and closer distance to water resources. At the landscape scale, resting sites were associated with beech forest, oak forest, secondary forest, and ravine biotopes and negatively associated with pine forest, houses, and roads. Resting sites had an aggregated spatial pattern from 0 to 900 m, but their distribution was completely random at larger scales. Our study revealed that