Over the past centuries, as people have cleared fields for farms, built roads and highways and expanded cities outward, trees have been cut down. Since 1850, global forest cover has been reduced by one-third. The way forests look has also been changed: much of the world’s woodlands now exist in choppy fragments, with 20% of the remaining forest within 100 m of an edge like a road, back yard, corn field or parking lot.

Scientists have studied fragmented forests for decades, mostly to gauge their effects on wildlife and biodiversity. However, recently, two Boston University College of Arts & Sciences scientists — Andrew Reinmann, a postdoctoral research associate, and Lucy Hutyra, an associate professor of Earth and environment — have turned their attention to another issue: the effects of forest fragments on carbon storage and climate change.

Reinmann and Hutyra found that temperate broadleaf forests, like the stands of red oak common in New England, absorb more carbon than expected along their edges, but they also found that those edges are more susceptible to heat stress. The research — funded by the National Oceanic & Atmospheric Administration, the National Aeronautics & Space Administration and the National Science Foundation and published in the Dec. 19, 2016, issue of Proceedings of the National Academy of Sciences — offers some good news and bad news about forest fragmentation. It suggests that while these forests may be more valuable carbon sinks than previously thought, they may also be more sensitive to climate change.

“Having accurate estimates of what those trees on the edge are doing — how much carbon they’re taking out of the atmosphere — is really important when we think about our future climate,” Reinmann, lead author on the paper, said.

The annual atmospheric concentration of carbon dioxide, a greenhouse gas, has increased by more than 40% since the start of the Industrial Revolution and continues to rise. Forests play a critical role as a carbon sink, absorbing about 25% of carbon dioxide emissions.

Most of the understanding of forest carbon dynamics comes from studying intact rural forests like Hubbard Brook in New Hampshire’s White Mountains and Harvard Forest in Petersham, Mass., not from studying forest fragments. “When you fragment a forest, you change a lot of the growing conditions of the forest that’s left behind, but we don’t have a very good understanding of how that change affects carbon sequestration and storage,” Reinmann said.

To find out, Reinmann and Hutyra gathered data from 21 fragmented forest plots around Boston, Mass., measuring about 500 trees. In eight of those plots, they went a step further, taking sample cores from trees greater than 10 cm in diameter for a total of 420 cores from 210 trees. They used the cores and other data to calculate how fast the trees grew. A tree’s size and growth rate indicate how much carbon it can absorb and also how much stress it’s experiencing, they said.

“If this carbon sink all of a sudden shuts off, our projections for future climate will change,” Reinmann noted. “So, our current understanding and ecological models, which don’t account for this, are missing something important.”

Reinmann and Hutyra found that forest fragments grew faster along the edges than intact forests, absorbing more carbon than expected. “When you create that edge, you essentially are reducing competition and freeing up resources like light, water and nutrients for trees,” Reinmann said, noting that the effect extends in about 20 m from the forest edge. Curiously, the finding may hold true for only temperate broadleaf forests common in New England, the Appalachians, Canada and Europe. The Amazon rainforest has the opposite effect when fragmented, with lower biomass and less carbon storage along the edges, he said.

“Foresters and loggers have known this intuitively for a long time: If you go in and you reduce the competition for resources, the remaining individuals will grow faster,” Hutyra added. “The novel piece of this work was to quantify it across these edges, see how far into the forest it goes and put it into context with how much this fragmentation matters in a portion of the world — southern New England — that we know is a large net carbon sink.”

Although this seems like a win for patchy New England forests, deforestation is still bad for carbon sequestration overall. “When you fragment a forest, the remaining forest can offset a little bit of what was lost, but not completely,” Reinmann said. “So, it may not be as terrible from a carbon perspective as we thought, but it’s still bad.”

Offsetting this is the paper’s other finding: These forest edges, which are more exposed to wind and sun, grow more slowly when stressed by heat.

“You lose a lot of carbon benefit in hot years,” said Reinmann, who found that the “magic number” for New England trees is about 27°C (80.6°F), which corresponds to the average high temperature in July. “Once you get much past that threshold, the trees grow much (more slowly)," he said.

Reinmann and Hutyra are currently expanding the work to study rural forests and so far are finding even larger effects there. They are also hoping to use high-resolution imaging and more precise chemical analyses to take a closer look at core samples to see how growth and photosynthesis change over days, seasons, heat waves and other environmental stressors. More data may lead to better models, Hutyra said.

“As we continue to more actively manage our landscape, whether it be thinking about agricultural intensification in Brazil or urban expansion in China or sprawling urban development (in the U.S.), the fragmenting of the landscape is ubiquitous. It’s likely to stay, if not increase,” Hutyra said. “So, quantifying the effects of all this fragmentation is really important for understanding the long-term and short-term ability of forests to continue to take up carbon and for us to be able to accurately model that to project future climate.”