A forest root less than a millimeter wide can host a bustling community of fungi and bacteria—and some of its quietest residents may help shape the whole underground neighborhood. In forests around Sugadaira, in Japan’s Nagano Prefecture, researchers used DNA data to identify root-dwelling fungi that could have an outsized influence on the communities around them.
The finding offers a new way to look for “keystone” species: organisms whose influence on an ecosystem is much larger than their numbers might suggest. Keystone species are often thought of as visible animals or plants. But forests also depend on vast communities of microbes in soil and on roots. These organisms help move nutrients and store carbon, yet it is hard to test the influence of any one species in an ecosystem as large and complex as a forest.
A Kyoto University team, led by first author Mikihito Noguchi, set out to find possible microbial keystones by studying the patterns in those communities. They collected 1,270 root samples from a cool-temperate forest in Sugadaira and analyzed DNA from fungi and bacteria living there. “Even a single plant root measuring less than one millimeter in diameter harbors a diverse array of coexisting microorganisms,” Noguchi said.
Instead of trying to test each species in isolation, the researchers used the DNA results to model how microbial communities tend to form. They then developed a method to identify species that might strongly change the structure of those communities, and ran the resulting algorithm on a supercomputer. To check whether the patterns held beyond Japan, they compared the results with a large data set from European forests.
The analysis pointed to endophytes, fungi that live within plant tissue in a mutually beneficial relationship with their hosts. The researchers found that these fungi may help reshape the relationships among plants and other microbes in roots and soil. Endophytes have received relatively little attention for their effects on forest ecosystems, but the study suggests they can influence the overall structure of fungal communities.
The results do not mean that endophytes have been proven to control every forest microbiome. Rather, they show how large-scale DNA data and statistical modeling can reveal likely influential species in a system where direct experiments are difficult. The approach could also help researchers investigate other complex microbial communities, not just those found in forests.
That broader view may matter for future ecosystem management. Forests are more than their trees: their health also depends on the organisms living out of sight, in roots and soil. By helping scientists identify which microbes may matter most, the new method could eventually inform approaches that account for these hidden communities as part of the ecosystem as a whole. The study, “Statistical inference of keystone taxa reshaping the assembly rules of forest root microbiomes,” was published in Nature Communications in 2026.