The Butterfly Effect

Depending on her mood, Elizabeth Crone has two ways of introducing herself to strangers on an airplane. If she’d like to catch up on her reading, she’ll say, “I’m a statistician.” But if she feels like conversing, she’ll say, “I study butterflies.”

Both are true descriptions of Crone’s work as an associate professor of population ecology and dynamics in the Department of Biology. While the topic of butterflies will start people reminiscing about the colorful species they knew as a child, it is Crone’s skill with data sets that allows her to find clues to the continued survival of these beloved insects.

It was through number-crunching, for example, that Crone discovered that climate change is already causing a major shift in North American butterfly populations. A 2013 paper in Nature Climate Change she co-authored showed that species that typically live south of Boston are expanding their territory farther north, while those that typically live north of Boston are declining, some at a rapid rate.

That means Massachusetts is seeing more of species such as the giant swallowtail and Zabulon skipper, but fewer of the Atlantis and Aphrodite fritillaries, which seem to be withdrawing to the colder areas of the Berkshire Mountains. “It is difficult to attribute the community-wide pattern of decline in cold-adapted species and invasion and growth of warm-adapted species to any mechanism other than climate warming,” the paper concluded.

“As a conservationist, it’s both encouraging and scary,” Crone says. “Encouraging because many species are able to adapt to climate change, but scary because things are really changing. And not all species are able to adapt. And we don’t really know much about which ones can or can’t.”

Drawing on Data from Amateurs

What’s particularly interesting about the study is that Crone didn’t even need to pull out a butterfly net to get her data. She found an untapped wealth of information in the existing reports from the amateur lepidopterists of the Massachusetts Butterfly Club. This group of butterfly enthusiasts, purely for their own enjoyment, had counted the number of each species they had seen and where, and then a volunteer had compiled it all—the sightings from nearly 20,000 trips over 18 years.

When Crone, who had just moved to Massachusetts and was trying to familiarize herself with the local butterfly scene, came across the lists on the internet, her first thought was, “Wow, they have lots of data. This is amazing.”

But because the data were collected in such an informal way, questions inevitably arose. For instance, what if some trips had better spotters than others? What if one trip yielded few butterfly sightings because someone got poison ivy, and the whole group decided to turn back early?

Taking a cue from a team of ecological statisticians in Australia who had recently developed a technique for analyzing sightings by bird-watchers, Crone and her colleagues dealt with such questions by assuming the more species that are reported in a particular outing, the more thoroughly the spotters had searched an area. That gave them a metric for weighing the accuracy of the different outings.

“In ecology, we can deal with small data sets in clever ways by customizing models to fit the nature of the data,” Crone says. “So we’re better at pulling out the signal if we can account for more of the noise”—in this case, the vagaries of amateur lepidopterists. 

A Chance Encounter

Crone grew up in Alexandria, Virginia, and spent a lot of time at her family’s farm in rural Maryland. She loved being outdoors and watching nature, but unlike many of her colleagues who also knew they would study birds or beetles, she says she “never had a particular love of a particular taxon.”

Butterflies fluttered into her life on a fluke. Soon after getting her Ph.D. in botany at Duke University in 1995, she started a fellowship with the National Science Foundation; she intended to study ladybugs. She became good friends with Cheryl Schultz, who was researching Fender’s blue butterflies.

“She came into my office with some data and asked, ‘How do I make sense of this?’ I said, ‘No problem.’ And we’ve been working together for the 20 years since.”

Crone, the daughter of a mathematician and a computer scientist, has always been comfortable with numbers, but finds that many biology majors are intimidated by them and tend to underestimate their own math skills. Throughout her academic career, which included positions at the University of Calgary, the University of Montana and Harvard University before she joined Tufts in 2013, she has had a track record of training graduate students to approach problems like a mathematician.

“Biology as a discipline is changing,” she says. “It’s becoming more quantitative, for better or for worse. Cheap, fast computers have allowed us to collect and process data in ways that wouldn’t have been possible back in the 20th century.”

Why Butterflies?

Crone uses statistics and modeling in her research on bees, perennial wildflowers and sugar maples as well. But it’s the butterflies that people often want to hear about.

Why do butterflies matter? Crone sees a couple ways to answer that question. First there is the cultural importance of these beautiful, gossamer-winged insects. There’s a reason people grow wistful when they learn that the monarch butterfly population in New England is just 1/18th what it was two decades ago. And what would the little girls’ clothing industry do if it lost its most stalwart icon?  

Then there is the ecological importance, which includes carrying pollen for plants and providing food for lots of birds. But the butterfly’s role may be even broader. For example, Crone points to a symbiotic relationship with ants that was only recently revealed.

The caterpillars of Fender’s blue butterflies secrete nectar that ants like to eat. To protect this important food source, ants will defend the caterpillars from predators such as spiders. (This was illustrated in a study that one of Crone’s students conducted, where she would poke a caterpillar and then time how long it took the ant cavalry to show up.) “If you take away the butterflies, would the ants not have enough food?” Crone asks.

The butterfly effect, it seems, is not just a colorful example of chaos theory: the flap of their delicate wings really could alter the course of history, or at least ecology.

“Everything is connected,” Crone says. “If you perturb a system in one way, it’s hard to predict what is going to happen.”

Julie Flaherty can be reached at julie.flaherty@tufts.edu.