INTRODUCTION
A species geographic range is the result of complex interactions between its intrinsic characteristics and environmental factors (Brown et al., 1996). At such broad spatial scales, climate and other biophysical conditions play a major role (Lomolino et al., 2006). Additionally, recent studies have shown that anthropic factors can have an effect on species range size (Di Marco & Santini, 2015) and vulnerability to extinction (Di Marco & Santini, 2015). Such vulnerability may vary across biomes within the species range, reflecting the interaction between anthropic effects and environmental factors.
For most species, geographic ranges are composed of multiple local populations that interact with each other (Andrewartha & Birch, 1954) and are distributed across areas with high environmental heterogeneity. At the local scale (landscape, or local populations) the way in which environmental conditions and species interactions influence distribution (occupancy) is less known and continues to be debated (Araújo & Rozenfeld, 2014; Blanchet et al., 2020; Cazelles et al., 2016), especially with regards to the underlying processes that account for the observed patterns. However, there is increasing empirical evidence showing that at the local scale, species occupancy probabilities are often influenced by both environmental factors and the presence or absence of interacting species (Rota et al., 2016).
Some studies indicate that organisms can adapt differently to heterogeneous conditions in different parts of their ranges, showing plasticity (Sagarin et al., 2006), and in this sense the factors that affect their occupancy could vary across sites. In a similar fashion, the presence and magnitude of human impacts can modify positively or negatively occupancy. Human activities such as agricultural and urban expansion or hunting have been shown to affect negatively biodiversity through degradation and fragmentation of habitats, even causing local extinctions (Foley et al., 2005; Kinnaird et al., 2003; Tscharntke et al., 2012). On the other hand, human activities (voluntarily or involuntarily) can also provide food and refuge to some species, specially to those that are opportunistic (Fedriani et al., 2001; Oro et al., 2013), subsidizing their presence and abundance. These changes can, in turn, affect species interactions (trophic and competitive).
Understanding how habitat characteristics, species interactions and anthropic factors affect patterns of occupancy of species is crucial for making management and conservation decisions (Guisan & Thuiller, 2005; Williams et al., 2002). This can help to minimize the impact of human activities on biodiversity, and inform restoration processes, by focusing on factors that are more important for certain species.
Mammals and terrestrial birds play important roles in the maintenance of ecosystem stability and biodiversity in tropical forests (Jansen et al., 2010; Sekercioglu, 2006; Terborgh, 1988). Many of these species are particularly vulnerable to local extirpations due to anthropic effects, such as hunting and habitat degradation, while others can be favored by anthropic subsidies (i.e. small and medium-sized mammals) (Fedriani et al., 2001; Oro et al., 2013; Porras et al., 2016). For this reason, they represent an ideal group for understanding the variation in factors that determine their presence in different landscapes. Colombia has a high diversity of mammals and terrestrial birds that occur in different habitats subject to varying degrees of human disturbance. The Middle Magdalena (MML) and Orinoquia landscape (OL), the two landscapes studied here, present differences in their predominant ecosystems as well as their level of human influence. The MML, while originally composed of vast humid tropical forests and wetlands, is now dominated by agricultural land uses, mainly cattle farming and palm oil, with high levels of fragmentation. Forests remnants are usually of less than 100 hectares. The OL is composed of different ecosystems, mainly savannas dissected by riparian forests, it has a much lower human density, and the predominant activity of extensive cattle farming has not resulted in drastic forest loss or degradation.
We use occupancy modeling to evaluate the factors that determine species presence within these two landscapes and assess how the importance of such factors (natural vs human-mediated) vary from one landscape to another. We also evaluate the prevalence of inter-specific interactions inferred though the effect of one species in the occupancy of another, and how these vary across landscapes. Discerning how occupancy is determined by environmental and anthropic factors in landscapes with different levels of disturbance can help to understand how species are affected by or are adjusting to human influence in tropical ecosystems.