We know that temperatures are heating up around the globe — but climate change is also responsible for all kinds of other changes on the planet, including increases in storms, fires, floods and, not surprisingly, drought.

It seems obvious that forests, which are full of water-guzzling trees, are among the ecosystems likely to be most affected by droughts. But while scientists know that drought is stressful on trees, “what we didn’t really know is what happened after the drought was alleviated,” said William Anderegg, a postdoctoral fellow at Princeton University’s Environmental Institute.

“One scenario is that once the water comes back into the soil and atmosphere, trees recover, in essence, immediately — kind of like your house plants might if you water them after a couple days of not watering them,” Anderegg said. But the other possibility is that drought causes “legacy effects,” or damage to trees that lingers long after the drought is over. This kind of lasting damage could have important implications for Earth’s climate, as healthy trees are able to store large amounts of carbon that would otherwise go into the atmosphere. Long-lasting damage could hurt this carbon-storing ability.

Currently, many models that focus on climate and vegetation assume that trees recover immediately, Anderegg said. He and his colleagues suspected that this was not actually the case. And in a new study, published Thursday in the journal Science, they demonstrate that legacy effects actually do exist.

It turns out trees don’t recover from drought immediately, as models often assume they do. Trees may slow their growth for up to four years following a severe drought, the researchers discovered. They used an international database to examine data on tree growth from 1,338 sites around the world, mostly non-tropical forests in the Northern Hemisphere. Legacy effects in general lasted two to four years, with growth being the most affected (about 9 percent slower than normal expected growth) in the first year following a severe drought. That said, after the first year or two, the effects are generally “very modest,” cautioned Robert Jackson, a professor of environmental science at Stanford University, who was not involved with the study.

The effects seemed to be worse for some areas and some types of trees than others. Damage was worse in drier ecosystems, and the effects also seemed to be a little more pronounced in gymnosperms — a classification including cone-bearing trees, such as pines — than angiosperms, or flowering trees, such as oaks.

The reasons behind the lengthy growth delays following drought aren’t totally clear, but Anderegg and his colleagues have some theories. Drought could cause leaves to shrink or trees to lose certain carbohydrates that assist in growth. Droughts are also sometimes followed by an increased burden of pests or diseases, which could weaken trees and limit their growth. But one of the likelier explanations, according to Anderegg, is that drought causes damage to trees’ water transport systems. If trees become less effective at shuttling water into their leaves, their growth will be stunted.

More research is likely needed on which causes apply to which types of trees. Regardless, the researchers maintain that representing these legacy effects in models can help scientists better understand how drought might affect forests’ ability to store carbon, which could have important implications for Earth’s climate future.

Forests generally serve as “carbon sinks,” meaning they trap carbon that would otherwise end up in the atmosphere. Bigger trees and larger forests can suck up more carbon, so if a forest’s growth is compromised, its status as a carbon sink could also be damaged.

The researchers even make some predictions about the way drought could affect carbon storage in forests, based on their results. For example, they write, if they base their predictions on the way carbon is stored in forests in the Southwestern United States, “legacy effects could lead to 3 percent lower carbon storage in semi-arid ecosystems over a century, equivalent to 1.6 metric gigatons of carbon when considering all semi-arid ecosystems across the globe.”

“We want to try to predict which types of forests are going to be most resilient in a changing climate,” said Anderegg, research that more accurate models can help facilitate. But Jackson, the Stanford professor, cautioned that there could be some difficulties implementing the results in global climate-vegetation models, partly because these models are so large-scale.

“One of the limitations for the models is that whatever you do has to be done on a large regional or global basis,” Jackson said. “You have to treat most trees the same. You can’t tune a response for a given species, or even genus, in a global model.” Since different species may react to drought in different ways and experience slightly different effects, this means it would be hard to end up with a perfect global model.

However, just knowing that legacy effects do, in fact, exist can help improve the models. For example, Jackson said, “One way to do it would be to just dial down growth a little bit after a drought in subsequent years.”

The paper also highlights the need for further research in related areas. The study largely left out tropical forests, whose growth is more difficult to study because many tropical species don’t produce annual rings in their trunks — the way growth and age is often measured in trees. More research should be done to see if tropical trees react to drought in similar ways, Anderegg said.

And even more important than drought’s effect on tree growth is its effect on tree mortality, Jackson added, which has “far bigger implications for the carbon cycle than these legacy effects of the drought.” Dead trees can release large amounts of carbon into the atmosphere, and there’s a possibility that more frequent and severe droughts could cause mass die-off events in forests, especially if they’re still in recovery from a previous drought, said Anderegg.

Events that release large amounts of carbon into the atmosphere can contribute to what’s known as a feedback loop: More trees dying and releasing carbon into the atmosphere could accelerate climate change, which could cause more changes on Earth, such as more severe droughts. More droughts could then cause more forest die-offs, perpetuating the vicious cycle. Similar climate loops exist in other aspects of the planet — for example, the polar ice caps, which reflect sunlight away from the Earth and help keep temperatures from getting too high. Human-caused climate change, however, is causing the ice to melt, which means more sun is absorbed by the Earth, causing temperatures to rise. This accelerated climate change, in turn, causes more ice to melt.

Research in these areas is important for understanding the ways forests and climate change effects interact and could affect one another. It’s a learning opportunity, and also a reminder that humans still have some control over the events that influence the Earth, Anderegg said.

“One of the final things I’d like to emphasize is that the future of a lot of these forests really rests in our hands,” Anderegg said. “The sooner and the more effectively we address climate change, the less risks forests will face.”