Published in Science and featured on the cover of its latest issue, the research, conducted with the involvement of the UAB and the CVC, a global study highlights how ecological context and multiple factors determine whether camouflage or warning colouration provide the greatest evolutionary advantage.
Natural selection has led some animals to evolve two alternative antipredator colour strategies: camouflage, in which cryptic colouration helps them blend into their surroundings, and bright warning colouration, which advertises danger if eaten (aposematism). The co-existence of these strategies demonstrates that each can be advantageous in different circumstances, yet until now it has remained unclear what determines the success of one over the other.
Now, a global experiment conducted in 21 countries across 6 continents and published in the journal Science reveals that context is crucial for deterring predators, and that several factors determine whether camouflage or warning signals work best.
The research, featured on the cover of the journal's latest issue, was led by Iliana Medina of the University of Melbourne and William Allen of Swansea University. Amongst its main authors is Olivier Penacchio from the Department of Computer Science at the Universitat Autònoma de Barcelona (UAB) and the Computer Vision Centre (CVC).
The experiment involved over 15,000 artificial prey in the form of moths, designed with three distinct colour patterns: a classic orange-and-black warning pattern, a dull brown camouflage pattern, and an unusual turquoise-and-black pattern. The artificial moths were pinned to trees with a mealworm reward in various woods and forests, while researchers monitored the behaviour of wild avian predators.
"For a long time, scientists have wondered why some animals use one defence over the other – and the answer turns out to be complicated. The predator community, prey community and habitat are all influential. This helps explain why we see camouflaged and warning coloured animals all over the world", says William Allen.
The study revealed that the predator community, including factors such as diversity and density, had the strongest influence on which prey colour pattern was most successful in avoiding predation. The results suggest that when predators compete intensely for food, they are more likely to risk attacking prey that might be toxic or unpalatable. Therefore, camouflage proved most effective in areas with a high predator density.
However, being cryptic does not always confer an advantage. In bright environments, the experiment showed that camouflaged prey were more visible than in darker settings and were attacked more frequently than prey with classic warning colours. Familiarity with particular prey colour strategy also plays a role: in areas where camouflaged prey are abundant, hiding becomes less effective because predators are better at detecting cryptic prey.
Overall, the results show how multiple mechanisms determine which strategy is more advantageous in a particular circumstance. "Some questions in ecology involve such a wide range of variables that only global collaboration and replication can bring us closer to understanding how nature works. It was a pleasure to work with such a diverse group of colleagues who made this research possible", says Iliana Medina.
The study included significant contributions from the UAB and the CVC in the fields of sensory ecology and computational visual neuroscience. "To predict and understand how visible the prey would be to predators in each context, we built a generic avian vision model and applied it to the image database generated for the 21 experimental locations. The computational predictions were crucial to validating the field observations", explains Olivier Penacchio.
The researchers say their findings will help build a better understanding of the evolution and global distribution of the most common antipredator colour strategies in animals.
Original article: Iliana Medina et al. «Global selection on insect antipredator coloration». Science 389,1336-1341 (2025). DOI:10.1126/science.adr7368
