Road projects are directly linked to the socioeconomic development of interconnected regions, but they also lead to various environmental impacts. The fauna is constantly affected by two of their main impacts: trampling and environment fragmentation. Primates are one of the most representative groups of the Brazilian fauna and, although they are directly affected by these impacts, they are one of the least studied groups. Thus, the objectives of this dissertation were: create a national diagnosis on the impacts of roads on primates, in order to show the current level of knowledge on the subject, evaluate which are the determining factors for the choice of crossing spots by black-tufted-ear marmosets (Callithrix penicillata) on roads, besides identifying characteristics of vegetation and landscape preferred for the occupation of this species in forest fragments impacted by roads. In the first chapter, the various impacts of roads on the arboreal fauna in general were contextualized, besides the characterization of all aspects of the biology of the species used as a model in the third chapter. In the second chapter, the diagnosis of road impacts on primates in Brazil was presented, through the application of a form with questions involving the issue of the impacts of roads on primates, so that expert researchers in primates responded, taking into consideration their experience and perception. The five impacts caused by roads with a higher degree of threat to primates were: formation of open areas, hunting, trampling, introduction of exotic species and audible noise. Altogether, 63 species of Brazilian primates were cited as demonstrably impacted by trampling. In the third chapter, we sought to evaluate which characteristics on roads are the most important for the occurrence of marmoset crossings. Therefore, four stretches of road were monitored in three different time periods, recording crossings performed by black-tufted-ear marmosets. Altogether, 200 marmoset crossings (168 aerial and 32 terrestrial) were recorded, and the presence of a canopy connection above the road was the most relevant factor to the increase in the number of crossings. In the fourth chapter, we evaluated whether marmosets avoid areas close to roads, besides identifying the determinant variables for the occupancy of the species in forest fragments impacted by roads. Our results indicate that the species occurred with a higher frequency in impacted fragments than in controls, although no difference was observed between the quadrants that formed the edge of the fragments in relation to those localized inside. The GLMMs highlighted the predictive variables CatQ (quadrant category), Zscore_Moran (index on the spatial correlation of quadrants), Tarde (time of day from 12:00 pm to 6:00 pm), Dist_rod (distance from the quadrant to the road) and T_borda (edge type which is the closest to the quadrant), as part of the explanatory models regarding the response variable (presence of marmosets in each quadrant). In this context, understanding what are the processes and factors responsible for the maintenance of the species in the fragments is of paramount importance for decision-making regarding the management of the environment for the conservation of the species.
