Mammals tolerate harmless human presence: Lessons from COVID-19 lockdown on Barro Colorado Island, Panamá

Human presence in protected forests impacts wildlife, but investigating such impacts is challenging because it is rare to isolate human presence from other anthropogenic factors. The COVID-19 lockdowns in 2020 provided a quasi-natural experiment that reduced human activity on Barro Colorado Island, a tropical forest isolated from most human footprints in Panamá. We used trail-based camera trap data from mammal species to compare a lockdown period (April-July 2020) versus non-lockdown (2019). For all observed species, we tested the hypotheses that human presence impacts activity level for 14 species, and for focal species we also tested diel activity, predator-prey dynamics, group cohesiveness, scent-marking, foraging, and vigilance. To assess lockdown effect, we analyzed our data using negative binomial, logistic and recurrent event analysis, and we contrasted null and alternative models. We also estimated diel activity patterns and used confidence intervals to examine lockdown effects. Based on camera trap observations, human presence on BCI forest was 9 times lower, while 16.2 times lower based on safety book records. Results showed no significant changes in activity level (rate of events) and diel activity for any species; in foraging duration of agouti, collared-peccary, red-brocket deer, and white-nosed coati; in predator-prey dynamic between agouti and ocelot; and in scent-marking of agouti. However, group cohesiveness and vigilance of white-nosed coati and collared-peccary were higher during lockdown. Overall, under lockdown, animal activity and diel activity patterns remained unchanged, although agouti, peccary, coati, and ocelot's diel activity slightly increased during typical human-active hours. Our results indicate that mammals, living on a managed forest with low anthropogenic impact and disturbance, can tolerate non-consumptive human presence.