FATORES QUE INFLUENCIAM A DETECTABILIDADE DE MAMÍFEROS TERRESTRES EM UMA ÁREA DE MATA ATLÂNTICA NO SUDESTE DO BRASIL

BRAZIL: The use of presence-absence data in wild animal monitoring studies has increased rapidly in recent decades. This study uses presence-absence data obtained by camera traps over eight years in the locality of Garrafão, in Rio de Janeiro state, to analyze the effects of sampling effort and successive changes in camera trap model and record type (photo or video) on mammal species’ detection probabilities, as well as the effect of the presence of domestic animals and people on the species’ occupancy, using a static multi-species occupancy model (MSOM). We recorded five medium-sized terrestrial mammal species: Cuniculus paca , Dasyprocta leporina , Didelphis aurita , Nasua nasua and Leopardus wiedii , in addition to a small species, Guerlinguetus brasiliensis. Sampling effort and progressive replacement in camera trap models and type of record positively influenced the detection probabilities of the species, with a greater effect on smaller sized species. Only one species, Didelphis aurita , increased its occupancy over time, while the other species showed uninformative estimates. Mean occupancy of the species was not influenced by the presence of domestic cats and people, but was positively correlated with the presence of domestic dogs. The results reinforce the importance of incorporating the effect of camera trap model and its settings on detection probabilities in occupancy models, which may result in greater precision in occupancy estimates in studies where there are equipment changes, such as long-term monitoring programs.