New model explains why costly insect "outbreaks" are hard to predict

New model explains why costly insect "outbreaks" are hard to predict

July 14, 2004

Many insect species--notably the gypsy moth, one of North America's most devastating forest pests--produce periodic population surges, known as outbreaks, only to crash down to low levels. These outbreaks can devastate huge areas of forest, but they occur at long, irregular intervals and so are hard to predict. In the 15 July 2004 issue of Nature, three researchers present a model that appears to untangle the factors that trigger such outbreaks, and that helps to explain why outbreaks are so unpredictable.

By combining elements from two prevailing but flawed insect-outbreak theories, and supplementing the combined theory with additional data, the researchers produced a mathematical model that reproduces the unpredictable outbreaks of gypsy moth populations more accurately than any model yet.

"We used gypsy moths as our example, but this theory should apply to any forest insect that has outbreaks," said Greg Dwyer, PhD, assistant professor of ecology and evolution at the University of Chicago and lead author of the study. "The model is quite general, and there are many insects beyond gypsy moths for which this applies."

Approximately 80 species of butterflies and moths undergo outbreaks, as do some small mammals, including voles and lemmings.

Ecologists who study insect outbreaks previously fell into two camps. One camp has focused on host-pathogen or host-parasite models, which presume that outbreaks primarily depend upon the presence or absence of disease among insect populations. These models can closely approximate the average length of time between outbreaks, but they fail to explain the relatively irregular timing with which outbreaks tend to occur.

The other camp has focused on the role of "generalist" predators, such as spiders and birds. When the density of these generalists declines the insects they feed on quickly explode in numbers. But recent studies have found that generalists are so adept at stabilizing populations that these generalist-predator models predict longer periods between outbreaks than occur in the field.

Dwyer's team introduces the host-pathogen-plus-predator model, which combines the stabilizing effect of dependable predators with the effects of disease, thus accounting more accurately for the ebb and flow of forest life.

Based on painstaking lab and field research, this model succeeds, in part, because of how precisely it calculates factors that have been left out of past models, said Dwyer. The authors consider elements like the variability of weather, the presence of more than one predator for each species and variability in the ability of insect larvae to resist disease.

Taking into account so many dynamics yields a formula more complex--and more useful--than others that have been used to explain insect outbreaks.

The model offers solutions for many of the insect-outbreak mysteries that have plagued ecologists for decades. For instance, scientists have sought models that explain why outbreaks adhere to such consistent timetables. But, while the populations of 18 of the insect species that produce outbreaks do seem to follow a cycle, Dwyer's team found some surprising data when they looked more closely.

"Earlier models assume that outbreaks are very regularly spaced, but most of them are not," said Dwyer. "People focused on the few insects that have regularly spaced outbreaks, but if you look at more insects, you see that most outbreaks are not regular."

Another mystery that Dwyer's team appears to have solved is the question of why insect populations of the same species, even when they are thousands of miles away from each other, surge simultaneously. This is the only outbreak model that portrays what scientists call spatial synchrony.

Which presents another mystery: even the study's authors are unsure why their model accounts for spatial synchrony while other models do not. They suspect that the long duration between outbreaks ultimately limits the importance of the environmental distinctions between various regions.

As for applying their theory to species beyond the insect world, Dwyer and his team are fairly optimistic.

"Voles and other small mammals are famous for having outbreaks," said Dwyer. But like previous insect-outbreak models, vole-outbreak models have emphasized regularity, even though vole outbreaks are often irregular. Now that the insect-outbreak riddle might be solved, small mammals like voles and lemmings could be next in line. "Some of the ideas in our model are probably translatable, given that there are these two dynamics--predation and pathogens--for many species that have outbreaks," said Dwyer.

Grants from the U.S. National Science Foundation and the Andrew W. Mellon Foundation supported this study. Additional authors include Jonathan Dushoff of Princeton and Susan Harrell Yee of the University of Chicago.