The current era, sometimes called “Anthropocene”, is characterized by extensive human-driven alterations of natural ecosystems on the Earth. Open tropical ecosystems represent an essential component of global terrestrial biodiversity, yet these systems are increasingly threatened by human disturbance and activity. Animal tolerance towards humans seems to be a one of key factors that facilitate wildlife–human coexistence. Unfortunately, traits predicting its direction and magnitude across tropical animals are poorly known, especially for tropical species. Most previous studies on human tolerance by animals came from Europe and parts of USA and Australia, with only tiny proportion of datasets coming from the Global South and tropical regions. Moreover, current biodiversity crisis urgently calls for research that can be directly translated to wildlife management practices.

Flight initiation distance is a very simple method how to estimate tolerance towards human disturbance. In a most basic form, this distance measures the distance between an individual animal and an approaching human who is walking directly towards it at the moment when the focal animal begins to escape. A large literature has confirmed that this distance represents a trade-off between the benefits of not escaping and the costs of fleeing. Thus, it tells us the willingness of animal individuals to take a risk and tolerate approaching humans. Prior work has also revealed that escape distances are relatively consistent for individuals, populations, and species when measured under similar contexts. Databases of escape distances are used to develop set-back zones to protect particularly vulnerable species and to mitigate potentially negative effects of human visitors on wildlife.

To fill existing spatial and taxonomic gaps in the current literature, we made an extensive search for papers and their authors who have studied or were studying avian escape behaviour in open tropical ecosystems. Despite our slight initial scepticism, over 40 colleagues from around the world responded to our call and were keen to share their primary data with our core team. Altogether, we quantified avian tolerance towards humans using 14,998 escape responses initiated by human observers, covering 953 species inhabiting rural and urban areas of open tropical ecosystems in Africa, South America and Australia (Fig. 1).

We then modelled associations between escape behaviour of birds and a set of a priori selected predictors using Bayesian phylogenetically- and spatially-informed mixed models. We found that the best predictors of avian tolerance toward humans were starting distance, body mass, the level of human disturbance (in terms of urbanization and human footprint index), season, and traits linked to efficient flight (migratory behaviour and wing shape) (Fig. 2). Human tolerance by birds was lower (i.e. escape distance was longer) in birds that were approached in rural than in urban populations and also in populations exposed to a lower human footprint. In addition, birds were less tolerant when approached from longer initial distances (when compared with closer approaches), larger birds escaped sooner than smaller birds, and birds during the wet season escaped at longer distances compared to the dry season. We also found that species adapted to efficient and long-distance flight escaped at longer distances.

Our study represents the first attempt to comprehensively describe and explore circumtropical variation in bird tolerance towards humans and provide results that could contribute to evidence-based conservation management. Our study emphasizes that a relatively small number of variables, which can be easily obtained in the field or literature, are solid predictors of the magnitude and the direction of avian tolerance towards humans across birds of open tropical ecosystems. Moreover, our results support the idea that some patterns of animal tolerance towards human may be general (e.g. higher tolerance of animals in human-dominated areas or shier behaviour in larger species), whereas other associations may be more geographically, taxonomically or temporally variable. Finally, I am happy that this project brought together numerous scholars and enthusiasts studying avian tolerance toward humans from around the world and created an international, diverse and inclusive web of collaborators, covering many underrepresented regions and groups in STEM.