Predators are key drivers of ecosystem function. Animals must continuously assess the risk of predation and adopt strategies to minimise danger by modifying behavioural and functional traits such as habitat use, foraging time, and prey selection. Humans, however, represent a unique class of predators and exert disproportionate lethal pressure across ecosystems, while their non-lethal activities such as recreation or tourism further alter animal behaviour through disturbance and perceived risk. My thesis investigates how direct and indirect human interactions shape animal behaviour and, by extension, ecosystem processes.
In Chapter 1, I conducted a systematic review and meta-analysis synthesising global evidence of behavioural responses of wild animals to both lethal and non-lethal human activities. Lethal activities such as hunting and fishing elicit strong and consistent anti-predator responses from wild animals, including heightened vigilance and reduced foraging. Non-lethal interactions such as tourism and recreation produce similar but weaker effects while passive activities such as roads produce more variable behavioural effects. These findings suggest that the magnitude and direction of behavioural responses depend on the type, intensity, and context of human activity, though significant geographic and taxonomic gaps remain.
In Chapter 2, I developed an agent-based model to explore the population-level consequences of humans as lethal versus non-lethal “superpredators” within a multi-trophic system. The model incorporates different types of predator and prey agents that interact through both consumptive and non-consumptive pathways. I compared scenarios involving human-mediated lethality (targeting mesopredators, prey, or both) against non–lethal interactions. Results show that lethal interactions profoundly alter population dynamics, increasing coexistence probabilities and reducing extinction risk, whereas non-lethal interactions alone exert minimal long-term influence. This theoretical framework highlights that the consumptive component of human predation drives ecological restructuring to a far greater extent than fear of humans alone.
In Chapter 3, I investigated the indirect effects of human activity through selective predator removal by fishing. I compared predator assemblages and prey behaviour across protected and unprotected coral reefs in the South Andaman Islands. Predator abundance, coral cover, and structural complexity were significantly higher within marine protected areas (MPAs). Correspondingly, prey fish in MPAs displayed increased vigilance and modified foraging consistent with a foraging–safety trade-off, whereas individuals in fished reefs exhibited homogenised behavioural patterns indicative of reduced predation pressure.
In Chapter 4, I tested for erosion of anti-predator responses through an in-situ behavioural experiment employing 3D-printed decoys of predatory fish. Prey fish in protected reefs responded strongly to predator models by increasing vigilance and foraging rates, while those in unprotected reefs showed limited responses. These results provide direct evidence that sustained predator removal may in turn alter perceived predation risk and weaken anti-predator behaviour among prey individuals.
Collectively, my thesis integrates global synthesis, theoretical modelling, and empirical field experiments to demonstrate that human predation fundamentally reshapes animal behaviour and ecosystem function. Lethal human activities generate potent landscapes of fear, while chronic predator removal leads to behavioural homogenisation. Humans as superpredators thus alter ecosystem processes both directly and indirectly. However, the overall impacts of human activities depend strongly on the intensity and nature of human pressure, local ecological context, and management regimes. Understanding these context-dependent outcomes is therefore essential to balance conservation goals with sustainable human use of natural ecosystems.