INTRODUCTION
The coexistence between species with similar ecological niches is allowed by the partitioning of the niche requirements between the species involved, being the spatial, dietary, and temporal dimensions the most common mechanisms of segregation (Schoener 1974). Temporal partitioning is more common among predators, since their prey shows peaks of activity, unlike herbivores, where plants and its parts can be found at any time (Schoener 1974). Temporal partitioning is especially important for large carnivore mammals, as they have large body-mass and, therefore, larger home ranges (Lindstedt et al. 1986), which do not allow for spatial partitioning in the current scenario of tropical forest depletion worldwide (Keenan et al. 2015). In this scenario, dietary partitioning is also unlikely, since smaller remnants support fewer species (Pardini et al. 2005; Cáceres et al. 2014) that could be a potential prey for medium- to large-sized carnivores. Therefore, temporal partitioning might be an important strategy for the coexistence between large carnivores in the current forest remnants.
The lack of an efficient niche partitioning between carnivores can lead to an Intraguild Predation (IGP) or Interspecific Killing (IK; Polis et al. 1989). IGP occurs when the dominant predator kills and eats the subordinate predator that uses similar resources, thus reducing competition, whereas IK occurs when the dominant predator kills the subordinate predator without any immediate energy gain to the dominant predator (Polis et al. 1989). The species with the highest body-mass is the dominant predator on the IGP/IK relationships and the smaller (and subordinate) predators are directly affected, a phenomenon that can directly influence the community structures (Polis et al. 1989) and may even result in local species extinction.
To avoid competition, subordinate predators can change their temporal activity, diet, and habitat use according to the specific conditions of each location at each time. This is possible because the niche of the subordinate predator is generally broader than the niche of the dominant predator, which is embedded into the niche of former species (Polis et al. 1989). However, few cases of complete inversion in the activity pattern were reported among mammals, perhaps because of the phylogenetic constrain of species, as observed in rodents by Roll et al . (2006). Usually, changes in the temporal activity or its limitation that acts as a response to the predation risks, thus reducing forage and energy gain by the subordinate competitor (Kronfeld-Schor and Dayan 2003).
The Atlantic Forest is an excellent biome for studying temporal partitioning between predators. It was greatly reduced and fragmented in the last decades, with few remaining areas of continuous native forest, and is currently composed primarily by fragments with less than 50 ha (Ribeiro et al. 2009). Among other smaller predators, there are two species of large predators that inhabit the Atlantic Forest: the jaguar (Panthera onca ) and the puma (Puma concolor ). Jaguars lost most of its historical distribution (Tôrres et al. 2008), but they are still present in some large Atlantic Forest remnants, while pumas have a broader current distribution in the biome (Beisiegel et al. 2012).
Most studies with these two large felids found great similarities between their activity patterns, as in Argentina (Di Bitetti et al . 2010), Belize (Harmsen et al . 2009; Harmsen et al . 2011), Brazil (Gómez et al . 2005; Foster et al . 2013), Mexico (Núñez et al . 2002), and Venezuela (Scognamillo et al . 2003). Conversely, there were no significant differences between their activity patterns in some cases (Romero-Muñoz et al . 2010; Blake et al . 2012). Usually, the temporal partitioning between jaguars and pumas was either associated with species avoidance (Romero-Muñoz et al . 2010) or activity synchronization with their main prey (Blake et al . 2012).
The temporal activity pattern of jaguars and pumas can also be influenced by the environment characteristics. Paviolo et al . (2009) found the higher nocturnal activity of pumas in areas where hunt pressure was higher in comparison with more protected areas, which might be a strategy to avoid conflict with humans, as also observed with other predators (Schuette et al. 2013; Massara et al. 2018). The temporal activity of predators may also be influenced by the temporal activity of their preys. The relationship between the temporal activity of predators and their prey might be linked either to the predation risk of the prey, or an alignment of the predator temporal activity to the activity peaks of their prey, which is related to coevolution strategies (Kronfeld-Schor and Dayan 2003). The synchronization with the temporal activity of preys is advantageous for predators since they can optimize its energy use foraging in a day period when their preys are more active.
There are several studies of temporal partitioning between other terrestrial carnivore mammals in the Neotropics with variated results (Di Bitetti et al. 2009; Massara et al. 2016; Ramírez-Mejía and Sánchez 2016), especially on ocelots (Leopardus pardalis ) and other smaller felids (Lucherini et al. 2009; Blake et al 2012; Massara et al. 2018; Santos et al. 2019). The temporal activity pattern of these smaller carnivores depended mainly on their habitat characteristics (i.e., prey availability and human disturbance level) and the presence of potential competitors.
We aimed to evaluate whether the coexistence of terrestrial mammal carnivores may be facilitated by temporal partitioning in an isolated Atlantic Forest remnant. We expected jaguars, the largest predator, to influence the activity patterns of all other predators, which may temporally segregate with jaguars to avoid IGP or IK. We also expected the same strategy for all the remaining predator relationships, where dominant predators may influence the temporal activity of the subordinate predators. Since some predators are naturally diurnal, we expected them to segregate temporally with nocturnal predators. We also expected nocturnal predators to be mostly or exclusively nocturnal in human-related locations, as a more nocturnal activity reduces the chances of encounters with humans that are usually more active during the daylight. Lastly, because differences in the temporal activity patterns are not necessarily relate with competition strategies to avoid dominant predators, we also evaluated whether the temporal activity of predators could be a strategy to maximize their chances to find potential preys, as observed by several studies with neotropical felids (Harmsen et al . 2011; Foster et al . 2013; Nagy-Reiset al . 2019).