Predicting how Insects, Plants Interact
News Jul 23, 2013
Two UC Davis-affiliated ecologists have developed a novel method that predicts plant/herbivore interactions before the plants arrive.
The research, involving 900 butterfly and moth species and 459 non-native plants in Europe, may lead to better screening of potential invasive plants, risk assessment, and pest management strategies, said researchers Ian Pearse and Florian Altermatt.
"Despite the growing prevalence of non-native plants, there are few effective tools for predicting the fate of non-native plants or their impacts on native communities," they wrote in newly published research, "Predicting Novel Trophic Interactions in a Non-Native World," in Ecology Letters. "We demonstrated that novel interactions between herbivores and non-native plants can be predicted based on plant evolutionary relationships and properties in the native herbivore-plant food web."
"My work has asked why some non-native plants are attacked by native herbivores while others are not," said Pearse, who completed the research while studying for his doctorate degree in entomology at UC Davis. He teamed with Altermatt, then a UC Davis postdoctoral scholar with the UC Davis Department of Environmental Science and Policy. Pearse is now a postdoctoral researcher in the Cornell Lab of Ornithology, and Altermatt is with the Swiss Federal Institute of Aquatic Science and Technology in Zurich, Switzerland.
Altermatt, interested in long-term trends in moth populations, assembled what Pearse called "one of the most extensive food webs of moth-host plant interactions, which covers a large part of Germany."
"We noticed that many non-native plants were included as hosts of native moths in that food web," Pearse said, "and we thought that we could use some of the ideas that I had been working on to explain which moths have started to eat which non-native plants."
"Herbivores, by in large, are not very adventurous in what they eat," Pearse said. "So, when a non-native plant enters their habitat, they tend to colonize those that are similar to the ones that they already eat. Plant evolutionary relationships are one of the best ways of looking at similarity between plants."
They successfully predicted the majority of novel interactions between herbivores and non-native plants. "When non-native plants enter a new ecosystem, their success and effects are mostly unpredictable," Pearse said. "However, we showed that one very predictable aspect of a non-native plant is which native herbivores can colonize it."
For instance, the larvae of the cinnabar moth (family Tyriajacobaeae), are a biocontrol agent of ragworts (Senecio), a native of Europe, but they also will colonize other plants. A geometrid moth, Eupithecia virganreata feeds on various ragworts but over the last decades, has extended its diet to invasive goldenrods (Solidago canadensis and S. gigantea).
On the basis of interactions between native hosts and insects, the researchers found "specific diet extensions of potential European pest insects to plants of forestry or agricultural interest introduced from North America, as well as the diet extension of European insects onto non-native plants that are of invasive concern."
"The goal of this approach is to correctly identify specific important interactions between a novel plant and native herbivore with the lowest possible false-positive rate, where a null model would result in a 50 percent false-positive rate," they wrote. "For example, we predicted that the tussock moth (Calliteara pudibunda) colonizes red oak (Quercus rubra; a common introduced tree throughout Europe) with a false-positive rate of only 0.7 percent. The tussock moth is an herbivorous insect of forestry concerns, having mass-outbreaks, and it is thus critical to understand its diet extension to novel host plants. Similarly, we predicted that the specialist Sessiid moth Synanthedon tipuliformis colonizes Ribes aureum, a cultivated gooseberry introduced from North America, with a false-positive rate of only 2.0 percent. S. tipuliformis is known to cause damage in agricultural gooseberry plantations, and an accurate prediction of host switch to introduced agricultural gooseberries is thus economically important."
Pearse received his doctorate in entomology from UC Davis in 2011, studying with major professor Rick Karban. Pearse's current research at Cornell "is trying to understand masting in oak trees; that is, why and how trees produce very large seed sets in some years but small ones in others."
Staphylococcus epidermidis is an ubiquitous colonizer of healthy human skin, but it is also a notorious source of serious nosocomial infections. Now, a new machine learning technique will help predict the risk of developing a serious, and possibly life-threatening S. epidermidis infection.READ MORE