AI Pesticide Research Part 2
From the Ag Information Network, I’m Bob Larson with today’s Fruit Grower report. Agriculture is increasingly turning to technology these days, including a study from Oregon State University that uses artificial intelligence to try to determine which pesticides are harmful to bees, wild and honey bees.
Assistant Professor Cory Simon says it’s one more method of trying to protect these pollinators from potential harm…
SIMON… “So machine learning models aren’t perfect, are they. They’re going to get some of them wrong. And so, because of that, if there was a new pesticide , okay, no one is going to use our model to say, go ahead, use this new pesticide that’s not toxic to bees, okay. You can’t trust it too much. So, at the end of the day account, you will need experiences, but I think this way of looking at our models is like a kind of pre-screening.
For pesticide companies, Simon says detox experiments can be very expensive and time-consuming…
SIMON … “So our model can kind of be used as an extra filter to quickly and cheaply filter out pesticides that might be harmful to bees, right.”
And pesticides, says Simon, change all the time…
SIMON… “You could also imagine, maybe, there’s a bunch of bugs coming in that we haven’t seen before and it’s an emergency. We have to apply a pesticide, it’s true. Then there is no time for bee experiments. So maybe our model could be used in this case to decide which pesticide we should deploy in certain emergency situations.
Simon says their findings are important because many fruit, nut, vegetable and seed crops depend on pollination.
The National Science Foundation supports this research.