Leaf-shaped wings in the delicate yellow-green of the brimstone butterfly; brown wings with spots that resemble eyes and scare off predators, like in the owl butterfly; or shimmering wings in green, blue and turquoise like those of the peacock swallowtail from Indonesia: butterflies and moths have wing patterns with, in the words of Charles Darwin, “endless forms most beautiful and most wonderful.” However, the evolutionary origins of this diversity are still not fully understood.

“If you think of an idealized butterfly or moth wing and all the possible patterns that can be painted on it, it really does seem endless,” said Moritz Lürig, a postdoctoral researcher in the Porto lab at the Florida Museum of Natural History. Lürig has been awarded a group starter grant by the European Research Council to create an atlas of butterfly wing patterns using machine learning.

The atlas will be a powerful tool for researchers who study the evolution of Lepidoptera, the group that contains all butterflies and moths. Because, said Lürig, although it may seem like nature is a dazzling lifescape of endless diversity, there are invisible forces that restrict what can actually exist.

Predation is one such force. Many butterflies accumulate toxins from their host plants ,and birds learn to recognize the distinct color pattern of their wings so they can avoid them in the future. Other toxic and non-toxic butterflies in the same area that mimic the color pattern also benefit from this defense system, resulting in separate species that look indistinguishable to the avian or untrained eye.

According to Lürig, machine learning has the potential to illuminate these invisible forces on a scale that would have been impossible just a couple of years ago. “Why are things as diverse and manifold as they are? It’s a big question to ask, and we’ve only touched the tip of the iceberg,” he said.

His goal is to create a comprehensive database of the wing color patterns for all the major groups of Lepidoptera, the second largest insect order after beetles.

“This digital ‘wing atlas’ will enable me to tackle three questions: first, how color patterns have changed during evolution and whether they are related to speciation; second, whether color and shape evolved independently of each other; and third, how environmental factors such as light, habitat or temperature shape the global distribution of wing color patterns.”

Lürig will be transitioning to the University of Bonn next spring. There, he’ll use funds from the grant to build his group and continue working on the atlas, which entails the photographic analysis for millions of preserved butterflies and moths from natural history museums around the world.

“Natural history museums are valuable research archives and publicly accessible repositories of knowledge. Their digitization opens up new ways of making biological diversity visible and more accessible,” Lürig said.

---

Source: Moritz Lürig, moritz.luerig@gmail.com
Media contacts: Katrin Piecha, kpiecha@uni-bonn.de;
Jerald Pinson, jpinson@flmnh.ufl.edu, 352-294-0452