The team developed an AI model that can predict how well seed potatoes will grow. The model combines two types of data: genetic data from microbes living on the seed potatoes and drone images of potato plants that emerged from the seed potatoes. By analyzing these two data sources together, the AI model uncovers patterns that help identify which microbes lead to the healthiest potato growth.
The researchers gathered their data by taking thousands of seed potato samples from 240 test fields. They studied the bacteria and fungi on the seed potatoes and later captured drone footage of the resulting plants during the growing season.
"This created a massive amount of data," says biologist Yang Song, one of the developers of the AI model, and first author of the paper. "By combining these data points using AI, we could pinpoint the microbes that are the best predicators of potato growth."
Microbes with big impacts
The research shows that the mix of bacteria and fungi on seed potatoes has a major influence on growth. Some bacteria, including a Streptomyces species, were found to significantly boost growth. Others, however, had the opposite effect, slowing down plant development.
"We're at the start of a revolutionary way to improve agriculture through microbiology and AI," said Berendsen. "For the first time, we can predict the quality of seed potatoes based on their microbial makeup."
Click here to see more...