When a 10,000-year-old prairie is plowed up to plant corn or a savanna is consumed by a strip mall, these once-grassy expanses, if left to regrow, can eventually look vaguely like they did before.
But a growing body of research suggests that what comes back isn't the same as what was lost, even decades later. And now scientists are starting to understand why.
That's the takeaway of a new study of grasslands worldwide published in the journal Proceedings of the National Academy of Sciences.
The results are important because grasslands are more than just treeless expanses.
They can look barren to the untrained eye, but they have extensive underground roots that reduce water runoff, control erosion , and store a third of Earth's terrestrial carbon.
Grasslands provide fodder for livestock and habitat for unique plants and animals ranging from American bison and African wildebeests to the great bustards of Asia. Covering nearly a quarter of the land on Earth, they also help support the livelihoods of more than a billion people worldwide.
Take the vast 5,000-mile belt of grasslands that is the Eurasian steppe, which stretches from Hungary to China. The prairies of North America's Great Plains. The African savanna.
Yet over the past couple of centuries, ancient grasslands around the world have largely been converted into cropland, tree plantations or developed as cities expand.
In Brazil, for example, a species-rich savanna known as the Cerrado loses an area the size of London every three months . India's savannas shrunk from 100 million acres in 1880 to 60 million acres in 2010, according to one study . And in the grasslands of North America, less than half of their historical acreage remains.
At the same time, millions of acres of farmed lands are abandoned worldwide, providing an opportunity for secondary grasslands to spring up in their place .
Previous research by the lead author of the new study Ashish Nerlekar , a postdoctoral researcher at Michigan State University, has shown that these newer grasslands can take decades to recover the biodiversity of ancient grasslands, and that even after centuries of regrowth certain plant species are still missing.
To find out why grasslands are so slow to recover, Nerlekar and a global team of co-authors, including Michigan State University professors Lauren Sullivan and Lars Brudvig , analyzed data for 742 plant species from grasslands across six continents.
The study compared older grasslands that had never been plowed or converted to farmland with patches of younger secondary grassland at varying stages of regrowth.
The researchers found that it didn't really matter where grasslands were on the map. "Our analysis indicates that differences between old-growth and secondary grasslands are remarkably consistent around the globe," Brudvig said.
To be sure, the specific mix of species that characterize, say, the North American tallgrass prairie — think switchgrass and big bluestem — isn't the same mix of species that lives in the Brazilian Cerrado or the East African savanna.
But across the globe, old-growth and secondary grasslands showed significant differences in how their plants grew.
When an old-growth grassland is plowed up, for example, the plants that are lost tend to be long-lived perennials with tough, leathery leaves, the researchers found.
"These traits help them deal with stressful situations like droughts, or being eaten by herbivores, or burned," Brudvig said.
But many of these original residents "are also very slow-growing," Nerlekar added. "And that prevents them from competing in the post-destruction grasslands."
Instead, the new plants that grow back as the grassland recovers have features associated with faster growth and greater ability to capture resources.
Tough, drought-tolerant native perennials like wiregrass (Aristida stricta), for instance, which is characteristic of the pine savannas of the southeastern U.S., weren't a hallmark of their secondary grassland counterparts. Instead, these regrowing grasslands were distinguished by faster-reproducing annual species, such as white goosefoot (Chenopodium album), that are better at competing for water and nutrients.
Plants in regrowing grasslands also tend to be naturally taller-growing, which makes them better at stealing the sunlight they need to thrive.
"When they reproduce the next generation does the same and they quickly take over," said Nerlekar, who is currently at the Indian Institute of Science Education & Research-Pune.
Such differences can persist for a remarkably long time, the study found. In one case, plants in secondary grasslands still towered over their old-growth counterparts even 300 years later.
Taken together, the findings help explain why many grasslands, once damaged, never fully bounce back.
The scientists say more work is needed to understand what these shifts mean for the ability of secondary grasslands to do things like pull heat-trapping carbon dioxide from the air or cycle nutrients.
"But these findings clearly show that secondary grasslands are really different from old-growth grasslands," Brudvig said.
The study has direct implications for grassland restoration policies, Brudvig added.
"What these findings suggest is that we need to be taking an active hand in grassland recovery — doing things like sowing seeds and transplanting — if we want secondary grasslands to function like the old-growth ones." Brudvig said.
"Preserving and conserving existing old-growth grasslands matters because you don't quickly get back what is lost," Nerlekar said.
This research was supported by grants from the São Paulo Research Foundation FAPESP, (2016/13232-5, 2018/03755-6, 2019/07773-1, 2020/01378-0, 2020/09257-8), the Brazilian National Council for Scientific and Technological Development, (309709/2020-2), China's Gansu Provincial Science and Technology Major Projects (23ZDNA009), the Strategic Environmental Research and Development Program of the U.S. Department of Defense (SERDP Project RC-1695), the Czech Science Foundation (31-25-18351S and 25-15235S), France's Regional Conservatory of Natural Spaces and the Region Sud-Paca, French National Research Agency (ANR-23-CE02-0034 - BEF-Cerrado), DAAD German Academic Exchange Service (DBo), the Association for Tropical Biology and Conservation, the South Africa Environmental Observation Network (SAEON), the Indian Institute of Science Education and Research, Pune, India, Ndlovu Node and the Unit for Environmental Sciences and Management, North-West University, South Africa, and a Michigan State University EEB Presidential postdoctoral fellowship.
CITATION: "Grassland destruction causes shifts in plant traits that persist during recovery," Ashish N. Nerlekar, André Giles, Norbert Hölzel, Mário Cava, Julien Piqueray, Natashi Pilon, Brenda Molano-Flores, Megha Ojha, Jin Hua Li, Gregory Mahy, Omofomwan Kingsley Osazuwa, Emma Ladouceur, Thierry Dutoit, Jutta Stadler, Daniela Boecker, Rafael S. Oliveira, Xi Zhou, Renaud Jaunatre, Klára Řehounková, Wenjin Li, Deepak Barua, Michal Hájek, Soizig Le Stradic, Frances Siebert, Eszter Ruprecht, Ellen I. Damschen, Elise Buisson, Karel Prach, Forest Isbell, John L. Orrock, Giselda Durigan, Lauren L. Sullivan, and Lars A. Brudvig. Proceedings of the National Academy of Sciences, Aug. 24, 2026. DOI: 10.1073/pnas.2533967123
