How to Build a Dragonfly: Rycroft group reveals mathematical principles that build insect wings

September 17, 2018
Lord of the wings

Just by looking at spots and stripes, you could identify a tiger from a cheetah, or a giraffe from a zebra. For people who love insects, they can do the same. Different insects have distinct patterns on their wings. Without even seeing the body of the insect, it’s possible to pick out a dragonfly versus a mayfly or a fruitfly or a damselfly.

“In many insect species, wings are like human fingerprints,” says Christopher Rycroft, associate professor of Engineering and Applied Science at Harvard University’s John A. Paulson School of Engineering & Applied Science. “Even the left and right wings of the same individual have unique vein patterns.”

So how do insect wings produce patterns that are distinct to individuals yet reliable enough that we can easily tell the difference between species? This week, the Rycroft group released the first mathematical answer to this question.

Kathy Li, who was at the time an undergraduate in the Rycroft lab, spent her time at Harvard collecting over 500 dragon- and damselflies from 215 species and producing images of each. QBio graduate student Jordan Hoffmann then developed software to analyze images of insect wings and extract the structure of their veins. With his doctoral advisor, Rycroft, and Seth Donoughe, a graduate student in the lab of Cassandra Extavour, Hoffman then developed mathematical models which explain how insect vein patterns form.

In so doing, the authors have established an approach which broadly applies to pattern analysis of branched structures – blood vessels, neurons, branches – across species. The data and code developed to support this study are shared openly here:

“Chris and his lab have laid out a simple mathematical model which explains how insect wings can have patterns that are reliable on one scale and yet random on another,” says BN Queenan, Executive Director of Research at Harvard’s NSF-Simons Center for Quantitative Biology. “It’s extremely elegant work. And with their model, it is now possible to systematically quantify the development and evolution of these patterns across the lifetime or even the entire evolutionary history of an organism.”


Hoffmann et al 2019 PNAS A simple developmental model recapitulates complex insect wing venation patterns

More on Jordan: