Insect infestation models may shed light on insect and disease outbreaks

Models of Larch budmoth outbreaks in the European Alps may eventually show scientists how to model a variety of disease and insect eruptions that rely on a combination of enemy, host and spatial movement to decimate populations, according to a team of ecologists.

“We use theoretical models to help understand the spatial component in these outbreaks and to predict how spatial spread occurs,” says Dr. Ottar N. Bjornstad, assistant professor of entomology and biology at Penn State. “With local outbreaks we expect a complex spread of pest species through the landscape, here, the species spreads in waves.”

The Larch budmoth feeds on larch trees, a common evergreen variety, consuming the needles and defoliating the branches. In the European Alps, the infestation moves as predictable waves from west to east completely defoliating forests beginning in the French and Italian Alps and moving across the continent through Switzerland and into Austria.

Bjornstad; M. Peltonen and A.M. Liebhold of the U.S. Department of Agriculture, and W. Baltensweiler, retired from the Swiss Institute of Technology, Zurich, report in today’s (Nov. 1) Science that “forty years of detailed surveys of defoliation caused by the Larch budmoth testify to conspicuous waves in the space-time dynamics of this system.

Creating models that predict this wave spread required consideration of the pest, its parasites, and the geographic distribution of the Larch. The historic pattern of Larch outbreaks is outbreaks that occur about every nine years and last for three to four years. The researchers confirmed that the spread of the outbreaks occurred in traveling waves and that space-time models accurately predict the geographic spread and timing of the outbreaks.

The Larch budmoth never dies out in any area of the Alps even though it totally defoliates an area. A small number of insects remain, feeding off resources left behind. While the larch trees recover, the budmoth population is kept in check by parasitic wasps that lay their eggs in the budmoth larvae. When the larva pupates, rather than a moth appearing, the adult form of the wasp emerges. Eventually, as resources improve, the budmoth population increases to reach pest levels and, because the larvae totally defoliate an area, the outbreak moves on and travels in waves through the Alps moving 125 to 186 miles per year.

“We managed to make a theoretical spatial and temporal model of the interaction between the pests, in this case the Larch budmoth, the enemy parasitic wasp and the food supply,” says Bjornstad. “Generating simple geographic models allows us to predict the spatial nature of the outbreak.”

The Penn State researcher notes that only with the use of fast, large cluster computers that can address spatially complex equations could this work be done. The models may be locally simple, but as the area covered increases, the dynamics become extremely complex. Unfortunately, a template does not exist to model the behavior of insects like this or human diseases that also spread in similar ways.

“We are currently working to generalize the model framework to other pest systems,” says Bjornstad. “With human diseases, for example, modeling the movement of people is different and more complex than modeling the movement of moths.”

While a general model will probably not be possible, a class of models researchers could use to predict insect outbreaks as well as disease epidemics such as smallpox, and measles in humans or foot-and-mouth disease in domestic animals, is slowly emerging. In each case researchers must develop a deep understanding of the specific local interactions between the pest and its resource.

“The emerging picture is that if you understand the local interactions and use a class of models that embed these on maps, you can make really good predictions,” says Bjornstad.

In essence, the researchers are developing a methodology for creating models that are applicable to infectious diseases in humans, animals and plants as well as insect outbreaks. Bjornstad is currently finalizing other research on human pathogens using very similar methodology.

“The key to understanding the spatial network of disease spread for humans is more difficult,” Bjornstad says. “Human movement patterns are more complex so the models have to be tweaked in each system to accommodate these differences. However, the overall methodology can model spatial outbreaks whether insect or human.”