Structure and interspecific dynamics in mammalian communities

Mammalian communities are shaped by the combined effects of environmental conditions, species interactions, and human disturbance, which jointly influence patterns of species activity, distribution, and diversity. Large mammals play key roles in these dynamics. Apex predators can affect prey and mesocarnivores through direct or non-lethal effects. At high densities, large ungulates may modify habitat structure, becoming ecologically dominant within communities and potentially generating cascading effects across trophic levels. Landscape features and anthropisation may modulate these processes, often mediating or overriding biotic interactions. Understanding how these drivers interact across ecosystems is critical, but information is still limited especially for European ecosystems, characterised by a significant human presence, by ongoing environmental changes, and by the recent recolonisation by predators. This thesis investigated how environmental, anthropogenic and biotic factors shape the spatial and temporal organisation of mammalian communities, with a specific focus on the role of the wolf Canis lupus and dominant ungulates. By combining long-term camera-trapping data, experimental approaches and statistical modelling, I analysed community-level patterns and species-specific responses across two contrasting ecosystems, a Mediterranean (Maremma Regional Park, central Italy) and an Alpine (Gran Paradiso National Park, northern Italy) one, with the aim of disentangling the relative contribution of these drivers in shaping community structure. In scientific Chapter 2 and Chapter 3, I investigated the determinants of spatial variation in mammalian species richness in Maremma and in Gran Paradiso, respectively. In the Mediterranean ecosystem, the spatial variation of species richness was primarily shaped by habitat heterogeneity and by human activity, increasing in structurally complex and ecotonal sectors and decreasing in sites with higher human visitation rates. Detection rates of the wolf and dominant ungulates were positively associated with local species richness, suggesting interspecific spatial convergence in high-quality habitats rather than top-down regulation of community structure. In the Alpine ecosystem, species richness patterns were mainly driven by environmental gradients, particularly elevation, and by landscape-level human footprint, reflecting strong environmental filtering. In particular, while direct human presence negatively affected richness at finer spatial scales, the Human Footprint Index – reflecting also the legacy of old infrastructures – was associated with a greater richness, potentially related to the formation of ecotones in/around abandoned villages. In both ecosystems, species-specific effects of biotic interactions were supported on species activity, in line with facilitation (red fox Vulpes vulpes spatially associated to wolf, in Maremma), competition (roe deer Capreolus capreolus negatively associated to fallow deer Dama dama, Maremma) and predator avoidance (chamois Rupicapra rupicapra negatively associated to wolf, in Gran Paradiso). Then I aimed at identifying finer scale mechanisms favouring interspecific coexistence and spatial partitioning. I focused on the well-studied ecosystem of Maremma, where previous work failed to detect strong signals of major spatial avoidance of predators by smaller carnivores or by prey. In scientific Chapter 4, I assessed whether spatial partitioning between the wolf and mesocarnivores occurred at night, and whether it was modulated by a proxy for nocturnal visibility, i.e., nocturnal illumination indexed by estimated moonlight. While badger Meles meles and Martes spp. showed a negative spatial relationship with the wolf, there was no support to nocturnal brightness mediating this partitioning. The red fox reduced its nocturnal activity under brighter conditions, with no support to interaction with the wolf. Thus, results suggest that predator avoidance was not mediated by proxies of nocturnal visibility. In scientific Chapter 5, I addressed whether partitioning could be mediated by olfactory cues associated with predator presence. I used a multi-species field experiment based on predator-derived olfactory cues to investigate fine-scale behavioural responses to perceived predation risk. Exposure to wolf cues did not induce changes in site-specific visitation rates or in temporal distribution of activity. Fallow deer showed increased vigilance in sites with higher wolf activity, with a weak support for olfactory cues emphasising vigilance. Red fox detection rates were positively associated with those of the wolf, in line with previous works showing facilitative interactions. While confirming the absence of a strong support to spatial avoidance, results indicate that spatial partitioning operates through complex, finer behavioural mechanisms involving a spatial modulation of nocturnal activity or vigilance. My results also support that in anthropised landscapes mammalian communities are largely structured by environmental and anthropogenic drivers, while biotic interactions, including those involving apex predators, tend to operate through context-dependent and species-specific patterns. Fine-scale behavioural responses to predation risk emerge without necessarily translating into large-scale changes in community structure, reflecting scale-dependent ecological processes. These findings also emphasise the importance of integrating multiple spatial and temporal scales across contrasting ecological contexts to better understand species coexistence and community dynamics in rapidly changing ecosystems.