Alberta researchers have developed a tool they say helps environmental managers figure out in advance whether they can eradicate the spread of invasive water species.
The charts, or decision trees, essentially predict the chances of success when dealing with invasive plant or animal species, well before managers start doing much of the work on the ground.
The decision trees, developed with the help of a computer and artificial intelligence, determine a project’s level of success based on an analysis of invasive species data, said Mark Lewis, the Canada research chair in mathematical biology and a professor at the University of Alberta.
He said the computer analyzes the data to establish rules. Then it can predict whether something can be eradicated.
“We build rules of thumb based on our own experiences. This is a little similar to that,” he said.
“We give it (the computer) the inputs and the outputs and it decides the rules.”
For example, if managers know they are dealing with an invasive plant species and only have five years to eradicate it, the decision tree says they must use mechanical methods to contain it. As well, the species must be in an area that’s less than 1,700 sq. metres for eradication to be work.
If farmers or land managers use chemical containment methods or the area is more than 1,700 sq. metres, the decision tree will inform them that eradication attempts will likely fail.
Lewis developed the trees with Russ Greiner, a professor of computing science at the U of A, and Yanyu Xiao, who conducted her PhD on this project and is now with the University of Cincinnati.
Along with offering a five-year plan, there are also decision trees that offer one-year and 10-year eradication plans. In the one-year plan, managers will likely fail if they are dealing with a plant, but will have better success if trying to eradicate an animal.
Lewis said the decision trees allow environmental managers to fully understand if they can solve a problem before putting too many resources into it.
For instance, the decision trees would help managers figure out where to allocate resources to have a better chance of success.
Environmental managers can use the charts by accessing the research paper and no computer program is required. As well, the decision-tree charts apply to any invasive aquatic species, Lewis said.
“We’ve done the heavy lifting and got the right structure. You don’t actually need a computer.”
While these decision trees apply only to invasive aquatic species, Lewis added it’s possible that similar charts could be developed for the agriculture industry to deal with weeds.