[Currents header graphic]

April 7, 1997

Marc Mangel issues a call to ecologists to change their methodology

By Jennifer McNulty

An ecologist with a quantitative bent, Marc Mangel has issued a call to his colleagues to change their methodology. Put simply, he wants them to use more sophisticated statistical methods to test their data against their descriptions of the world.

To that end, Mangel has prepared "a new kind of tool kit" for ecologists to use. Mangel teamed up with Ray Hilborn, a professor in the School of Fisheries at the University of Washington, to introduce these techniques in a new book called The Ecological Detective: Confronting Models with Data.

The book is designed to be a link between standard ecological modeling, or theoretical ecology, and serious statistical texts. The title of the book refers to the work ecologists do collecting clues and assembling them into a coherent picture--the very process Mangel and Hilborn would like to influence.

"Ecology is so complex that you can't derive basic laws like you can in physics," explained Mangel, a professor of environmental studies who came to UCSC from UC Davis last summer. "We're hoping that ecological detectives really will compare multiple mathematical models, which provide two or more competing descriptions, and let the data be the arbitrator."

For example, the book describes research on the Mediterranean fruit fly, or medfly, one of the most destructive agricultural pests in the world that has appeared sporadically in California during the last two decades.

Until 1991, the accepted model of medfly outbreaks was that each appearance marked a new invasion, triggered by someone accidentally bringing flies into the state and followed by detection and eradication. But a competing model has brought that view into question.

The second model asserts that the medfly has established itself in California but that conditions generally keep the population below detectable levels until something changes and the medfly population grows and spreads.

The two varying models have "enormous agricultural and socioeconomic implications," said Mangel.

"If Japan subscribed to the second model, they could demand that all fruit imported from California must be treated because it could be contaminated," said Mangel. "It will get us in lots of trouble."

Another example described in the book involves management of the wildebeest population in Africa's Serengeti ecosystem. Hilborn and Mangel show how the methods can be used to estimate the number of wildebeest that the Serengeti can support in wet and dry years and the intensity of poaching in recent years.

"The idea behind The Ecological Detective is that with two or more models, ecologists have to snoop around the way a detective does, and the data will tell us which model has more weight," said Mangel. "It seems obvious, but it's a big step from getting the picture to actually being able to do it."

Technology, particularly computing, has given ecologists and biologists the tools to do things computationally that they could never have done analytically with paper and pencil, said Mangel.

"This book is not intended for professional statisticians, who will likely find much of what we have to say obvious and belaboring the point. Our goal is to show ecologists how these ideas can be applied to a wide range of ecological problems," said Mangel, a leading conservation biologist whose own research focuses on Atlantic salmon, southern ocean krill, insect parasitoids and fruit flies, and, more recently, steelhead trout and butterflies.

Return to the Currents home page

Go to UCSC's home page