Search for a command to run...
The landscape of fear (LOF) framework predicts that prey adapt their behavior to mitigate predation risk, yet the framework's expression in resource-limited, unfenced systems remains poorly understood. Across seven large herbivore species in an unfenced arid system, space use is governed by bottom-up constraints while predator risk is buffered in time, producing consistent low diel overlap with nocturnal carnivores and trait-dependent moderation of spatial responses. We used camera trap data from northwest Namibia to examine how five herbivore species (gemsbok <i>Oryx gazella</i>, southern giraffe <i>Giraffa giraffa</i>, greater kudu <i>Tragelaphus strepsiceros</i>, Hartmann's mountain zebra <i>Equus zebra</i>, springbok <i>Antidorcas marsupialis</i>) and two megaherbivores (black rhinoceros <i>Diceros bicornis</i>, African bush elephant <i>Loxodonta africana</i>) navigate bottom-up environmental constraints and top-down predation risk from lions (<i>Panthera leo</i>) and spotted hyenas (<i>Crocuta crocuta</i>). We tested four hypotheses: (1) that bottom-up environmental constraints would explain more variance in herbivore space use than predator covariates; (2) that herbivores would reduce diel overlap (Δ) with nocturnal predators rather than vacating resource patches; (3) that megaherbivores would show minimal spatiotemporal avoidance of predators, relative to smaller herbivores; and (4) that herbivores would respond reactively to recent predator presence, rather than proactively by avoiding areas of chronic predator activity. Our findings support a hierarchical LOF in which bottom-up constraints (dry-season progression, visibility for grazers, habitat structure for browsers) contribute to herbivore space use and top-down predator pressures guide temporal activity: most species exhibited strongly diurnal activity, resulting in low temporal overlap with nocturnal carnivores (Δ ≈0.11-0.21 for grazers; Δ ≈0.06-0.26 for browsers). By contrast, spatial responses to predator presence were modulated by environmental context, reinforcing the importance of spatiotemporal plasticity. These results advance understandings of predator-prey dynamics, particularly in dryland ecosystems. We suggest refinements to the LOF framework for multi-predator, resource-constrained landscapes.