Blog written by Patrik Karell. Read the full paper here.
For most animals, it is of crucial importance that they are not visible to their predators. This is why predation is considered one of the most important selection pressures shaping the wide variety in colouration we find among animals. For animals relying on cryptic colouration, natural selection is expected to favour the colouration or colour pattern that is least conspicuous in the natural environment where the species lives. When the environment also changes the effect of camouflage is altered, which in turn leads to changes in selection pressure on colouration. In Northern Europe winters are getting shorter due to climate change: the probability of a lasting snow cover is getting lower and the duration of the snow covered winters is decreasing. This change from whiter to darker winters is a major challenge for many species living in these regions around the year.
Our study species, the tawny owl, Strix aluco, come in a brown and a grey colour morph. The tawny owl is widespread in the Western Palearctic and individuals of the same species live in very different types of landscapes with relatively large variation in seasonal climates. Therefore, selection may favour certain colour phenotypes depending on the type of landscape (i.e. background) they live in. In Southern Finland, which is at the northernmost limit of the tawny owl’s distribution, we have previously found evidence that there is strong survival selection against the brown morph in winters with lots of snow, whereas this selection is absent in winters with no or very little snow.
In our recent paper in Ecology and Evolution we test the prediction that this snow condition-dependent selection against the brown colour morph is associated with differential camouflage of the colour morphs depending on the snow conditions. We took stuffed grey and brown tawny owls and a camera out into the woods in winter in an area in the outskirts of Helsinki where tawny owls naturally occur (and close to the area where we know survival of the brown morph is decreased in snowy conditions). We placed the owl mounts in trees in natural roosting positions in their natural environments and took photos of them. We took one set of pictures of both the grey and the brown morph when there still was no snow and an identical set of pictures after snowfall. We randomised the pictures and made a number of series of pictures to be used in the experiment.
The online “spot the owl” game
The experiment was conducted as an online game, where people could visit a website and participate. The participants were shown a series of pictures and their task was to spot the owl in the picture as fast as possible. If they could not detect the owl they were shown the next picture in the series. The online game was successful and attracted more than 5000 participants worldwide in a short time. The results were as we expected: in pictures of snowy conditions the participants were more likely to find the owl and they found them faster if the owl in the picture was a brown morph. The grey morph was therefore more cryptic in a snowy landscape. In snowless conditions this crypsis difference between morphs was much less pronounced.
Is a human observer comparable to the enemies in the natural environment?
Since birds are able to sense the ultraviolet spectrum of light it is not evident that a bird, be it a hunting goshawk or a mobbing passerine, would perceive the colour of the two tawny owl morphs in a similar way as the human observers. We therefore used an avian vision model to compare how the tawny owl morphs are perceived by a blue tit as compared to a human. As predicted, the avian vision model showed no difference between a bird’s and a human’s perception of the tawny owl colour morphs. However, this exercise showed that the plumage of the brown morph deviates more from a spruce or pine trunk than the grey plumage. Therefore, a roosting tawny owl in a boreal coniferous forest is likely to be less conspicuous if it has a grey plumage.
Is difference in camouflage related to climate-driven selection on colour in tawny owls?
We were able to show that the brown morph is more conspicuous in snowy conditions, as predicted based on previous findings of strong survival selection against the brown morph in snowy conditions. We therefore suggest that if it takes longer to detect a tawny owl or if it completely remains undetected it will have a higher probability of survival in winter. The tawny owl is nocturnal and roosts during daytime when most other (diurnal) birds are active. If a roosting tawny owl is detected in the forest by a passerine or a corvid, these potential avian prey species will harass the owl and try to chase it away. Fleeing is often the solution for a tawny owl and this requires energy, which desperately needs to be saved in order to survive when winter conditions in general are harsh in Northern Europe. Fleeing also reinforces the likelihood of being detected and killed by predators such as goshawks.
Warmer winters with less snow makes the brown tawny owl less conspicuous. Together with other potential benefits of warmer winters, such as more easily accessible small mammal prey which cannot hide under the snow and overall more favourable thermal conditions, warmer snow-free winters may benefit the brown morph. Indeed, our analyses of ringing data from Finland do show that the brown morph has steadily increased in frequency in Finland in the past 50 years.