Navigating the risks and rewards of scavenging in multipredator, human-impacted landscapes

Large carnivores can influence smaller scavengers through both positive and negative interactions (e.g., carrion provisioning and intraguild killing) and ultimately shape scavenging efficiency. However, we know little about this trade-off in anthropogenic landscapes where humans kill carnivores and provide carrion subsidies. In the context of wolf (Canis lupus) recolonization of human-impacted landscapes in Washington, USA, we investigated how sources of ungulate mortality (wolves, cougars [Puma concolor], and vehicles) shape scavenging efficiency, community-wide carcass visitations, and the strategies used by scavengers to navigate risk-reward trade-offs. Cougar and wolf kills mostly occurred in areas with low-to-moderate human influence, whereas roadkill typically occurred in areas with high human impact. Wolves consumed their kills most rapidly (median <4.7 days), providing fewer scavenging opportunities than cougar- and vehicle-killed ungulates, which persisted longer (median = 8.9 and 12 days, respectively). Roadkill primarily attracted avian scavengers, whereas mammalian scavengers used roadkill to a lesser degree and did so by shifting to more nocturnal foraging. The absence in winter of turkey vultures (Cathartes aura) and black bears (Ursus americanus), which are obligate and apex scavengers, respectively, coincided with a seasonal increase in scavenging by most other species. The two mammalian mesocarnivores exhibited divergent strategies: Coyotes (Canis latrans) frequently scavenged but usually for short durations and with heightened vigilance at predator kills, whereas bobcats (Lynx rufus) visited carcasses less frequently but fed for longer durations and displayed low vigilance while scavenging. These results suggest a hierarchical decision-making process whereby scavengers first choose whether to forage at a carcass before fine-tuning foraging duration, using temporal refugia, or increasing vigilance. Predator recovery in human-dominated landscapes therefore adds complexity to the spatiotemporal landscape of risks and rewards, and outcomes for scavengers will likely depend on their propensity to scavenge and vulnerability to humans and large predators.