Primate sleeping site selection is influenced by multiple ecological factors including predation avoidance, thermoregulation, and food access. To test these hypotheses, we studied the sleeping trees used by a group of wild silky sifakas (Propithecus candidus) in Marojejy National Park, Madagascar. During this ten-month study, the group slept in 828 sleeping trees from approximately 35 genera. In support of thermoregulation, generalised linear models revealed that as temperature decreased, the number of individuals sleeping together significantly increased and they slept at further distances from the trunk. As rainfall increased, sleep site height significantly increased. Weinmannia was the most frequented tree genus, despite low abundance, accounting for 29% of all sleeping trees. In support of food access, 94.8% of sleeping trees were food trees. Weinmannia is among the most highly preferred food trees. The group slept at a mean height of 16.0 m near the top of tall trees which averaged 19.5 m. Sleep trees were significantly taller than trees in botanical plots within the sifaka’s home range. They never slept in the same trees on consecutive nights, and sleeping heights were significantly higher than daytime heights which is consistent with predation avoidance. Social sleeping in groups of two or three individuals (62.9%) was more common than solitary sleeping (37.1%). At such heights, huddling may increase vigilance and lessen the risk of predation by the fossa (Cryptoprocta ferox) while also reducing heat loss. These patterns suggest that silky sifaka sleep site choice is influenced by thermoregulation and food access in addition to predation avoidance. We suggest that understanding sleep site use can assist in conservation of species like silky sifakas by enabling researchers to find new groups, protect habitats with key tree species, and inform reforestation efforts.
