When Hurricane Maria roared through Puerto Rico last October, it did more than rip off roofs, destroy the power grid and kill people. It also tore up the forests that cover over half of the island. Many trees were stripped of foliage and branches; others simply blown out of the ground or snapped in half. Lush landscapes were instantly converted to expanses of naked trunks; normally cool, shady forest floors were baked by the sun. In days after, the chatter of frogs stilled, dead birds littered the ground, and bees buzzed around crazily, looking in vain for flowers.
For the past 15 years, forest ecologist Maria Uriarte has been documenting the lives of thousands of individual trees in dozens of plots spread across the island, putting her in a unique position to study the hurricane’s damage and its long-term implications. Three months after the hurricane, she had returned on a quickly organized post-hurricane study.
Uriarte, based at Columbia University’s Department of Ecology, Evolution and Environmental Biology, already has years of data on tree growth and health in response to a variety of factors including shifting rainfall and temperature. She works with other scientists studying insects, soils and other aspects of the ecosystem. Her central question: how will climate change affect these forests, and, by extension, others across the tropics and subtropics?
Hurricane Maria came and went in 24 hours, lashing the landscape with winds up to 155 miles per hour and dumping three feet of rain in places. It was the most powerful storm to strike the island since 1928. Caribbean forests, of course, have evolved to recover from giant storms. But as the climate warms, the biggest ones are projected to become more frequent, and this could change such forests dramatically. How might they operate if the catastrophic becomes commonplace, and storms that used to come every 80 or 100 years start arriving every 10 or 20?
Many biologists gravitate to the romance of so-called “virgin” forests, untouched by modern logging, fire or agriculture. Not Uriarte. She estimates that some 70 percent of tropical forests are second growth, regenerating from human disturbance since the 20th century. Even “old-growth” has almost always at some point been altered by humans, she says. Puerto Rico, for instance, was stripped centuries ago for sugar-cane plantations, grazing and subsistence farming. Starting in the 1940s, the economy shifted to manufacturing, and trees gradually took back abandoned farm land, especially in mountainous areas.
Trees play a big role in regulating climate, sucking large amounts of carbon dioxide from the air to build tissues. Worldwide, forests remove a net 1 billion to 2 billion tons of carbon from the air each year, roughly a fifth of what humans are producing. But climate-induced changes in weather could change that. For one, when trees die, they decompose and release their carbon back to the atmosphere; that means more frequent big storms could amplify the amount of carbon in the atmosphere. On the other hand, when forests are knocked down, replacement trees grow a lot faster than mature forests, gaining almost all of their eventual biomass in their first 60 or 70 years. This gives second-growth forests a potentially outsize braking impact on climate change. On the other, other hand, more frequent storms may cause species compositions to change permanently to shorter, scrubbier trees that spring up faster, but also fall down more easily in the next storm, and maybe forests will store less carbon in the long run.
Uriarte met me and two of her grad students, Jazlynn Hall and Andrew Quebbeman, at the airport outside the capital city of San Juan on a Friday. We knew the trip might be complicated; even three months after the storm, it was unclear whether roads to some field sites were open. Nearly half the island still lacked electricity. Hundreds of thousands of people had fled to the mainland. Hotel rooms were filled with emergency workers and newly homeless families.
Puerto Rico is cut lengthwise by a central east-west cordillera, dividing the island into two climatic zones. Due to moist northerly sea winds, the north side, where San Juan lies, is wet year round; that side is dominated by rain- and cloud forests. The south side, lying behind the mountains’ rain shadow, is dry much of the year, and thus populated by trees adapted to arid conditions.
Next morning, we took off in a rental car for the south. As we mounted into cordillera, the slopes seemed surprisingly green–plenty of big, tall individual trees visible, sprouting new leaves. “Yes, things are already coming back,” said Uriarte. But, she said, one doesn’t normally see individual trees–just a massive, interwoven green canopy. “Look, those are trees like children draw trees,” she said. “First the trunk. Then a couple of branches. Then a few leaves. What you’re seeing is a skeleton forest.”
We crossed the divide, and the big rain-forest trees suddenly disappeared, replaced by low, prickly dry-forest species. First stop was Guánica Forest Biosphere Preserve, a popular hiking area where Uriarte and colleagues previously measured and cataloged thousands of trees spread over 10 acres. During the 20-minute walk to the plot down a broad trail, dense clouds of butterflies started up from the vegetation as we moved through. Was this an aftereffect of the hurricane, Uriarte wondered? Remote images suggested that vegetation here had flowered abnormally after the huge rains. This may have brought a burst of life to the place.
Many of the trees in the plot were relatively small and spindly, not much of an obstacle to wind, and had suffered little damage. But the rarer bigger ones did get hit; many were snapped off or defoliated. Uriarte’s long-term concern here was more rain than wind: even with big hurricanes, all projections say the Caribbean will see a decline in rainfall as climate warms. That means that some tree species in dryer regions like this, already just hanging on, may die off. The effects could mushroom, because Guánica and similar dry forests are major stopovers for migratory birds, and the effects on them or other creatures could be widespread and unpredictable.
Next day, the team drove back up into the mountains to do a detailed survey of a rain-forest plot in Toro Negro State Park. Remote imagery suggested that this plot, 3,000 feet up on an exposed ridge, had been ripped by the full fury of the storm. Uriarte’s study involves examining imagery from satellites and aircraft, and comparing what is actually on the ground. One researcher at the University of California, Berkeley, has made a preliminary estimate based on remote sensing that Hurricane Maria killed or severely damaged 23 million to 31 million trees. It would be impossible to count every tree in the forest. So Uriarte hopes to extrapolate from a limited number of ground plots information that will allow her to interpret remote images accurately, and help answer the big questions.
A long drive up winding little roads took us past downed trees and dirt recently bulldozed to the sides. Power poles leaned at crazy angles or were broken off entirely, trailing dead wires. We reached a footpath to the plot–now a half-hour scramble through broken and uprooted trees. The 150-by-150-foot plot itself, marked off by plastic pipes driven into the ground, sprawled across a steep, muddy slope. “This used to be one of my favorite spots,” said Uriarte. It had been, she said, a cathedral-like space shaded by high, leafy canopy, with lots of room between trees. Now, debris and broken trunks lay tangled everywhere, and largely naked tree trunks jutted into open sky. Clouds scudded not far above. Soon, it began to rain.
The team got to work. Hall and Quebbeman identified dozens of big individual trees that previously had been marked with numbered metal tags. At each tree, they peered upward to call out damage: light (less than half of branches and foliage removed); moderate (more than half); or severe (90 percent or more). They announced whether the tree looked dead or alive. There were separate categories for trees that were tipped over, broken off, or simply missing in action. Uriarte stood by with a clipboard to record the observations.
Some common species, such as cecropias, with their broad, twisty canopies, had lost most of their tops. Sleeker palms, better designed for wind, had merely shed their fronds, and were already regenerating new ones. Many other still-standing trees were attempting comebacks. Those of one species, alchornea, had lost most of their tops, but their trunks were already like vertical lawns, bristling with new leaves from bottom to top. Some leaves would probably develop into new branches, and the trees might recover, said Uriarte. There appeared to be only one big tree here that remained untouched: a tall, straight sloanea that still sprouted profuse branches and leaves at its top. Maybe it had been saved by some peculiarity of wind currents, its shape, or just plain chance. “Oh, I love this tree,” said Uriarte.
Newly exposed to sunlight, the forest floor was popping with seedlings. “If we come back in a year, this place will be a thicket, impossible to move around in,” said Uriarte. She estimated that continuous canopy might redevelop here in as little as five or 10 years–but it might not consist of the same species, nor of the same size. Some of the worst damaged trees here might be 100 years old; if these didn’t regenerate, their likes might not be seen here in a human lifetime, if ever.
On succeeding days, we worked similar remote plots, cataloging up to 100 trees in each. “The generation of statistics is what brings out patterns,” said Uriarte. “Slope, aspect, tree species, wind speed, soil type, amount of rain. What factors matter when it comes to damage? If you put that all together, you start to form a picture of how these forests might develop in the future.” Tree redevelopment will in turn influence which plants, birds, amphibians and other creatures also come back, or fail.
We headed back toward San Juan. The last stop was El Yunque National Forest, a mountainous 28,000-acre tract about 30 miles from the city. Puerto Rico’s top tourist attraction, it harbors more than 200 species of trees, 50 species of birds, and many other creatures, some of them found nowhere else. Now largely defoliated and its roads and other infrastructure wrecked, it was closed to the public. We stayed at a research station that had suffered only light damage. Electricity came a few hours a day from a big gas-powered generator. From here, through the denuded landscape we could see the normally invisible concrete towers of San Juan.
Researchers including Uriarte have been studying the forest here for decades, as part of the national Long Term Ecological Research program. Uriarte’s next step: a massive survey of 40 nearby acres where thousands of trees had previously been tagged and measured. A crew of 10 volunteers was scheduled to come in the next day, and start a six-month project to examine every tree and its fate following the hurricane. The observations will be combined with images taken from satellites and aircraft, using artificial intelligence to analyze how different species fared, and what factors influenced the outcomes.
It was here that I met Grizelle Gonzalez, a soil ecologist with the U.S. Forest Service. She has been working at El Yunque for 17 years. Gonzalez walked me down a muddy path to look around. “It’s a Dr. Seuss kind of forest now,” she said of the newly simplified landscape. “Some trees have died, it’s true. But you can see some are striving to leaf up. A lot of green is coming at the surface layer–we’re seeing a lot of ferns and grasses.” She viewed masses of downed branches in a positive light; they were recycling nutrients to the forest floor, she said, and providing habitat for everything from fungi to spiders. She bent down to inspect one heap of brown, dead branches already draped with webs. Gonzalez figured lower elevation areas might regain their former look in as little as five years, though species compositions might shift. Unfortunately, she said, some of the rarer environments at higher elevations, dominated by twisty, slow-growing dwarf trees and mosses might take 100 years, if they came back at all.
Tears welled and her voice caught for a second as she described her own family’s experience of Hurricane Maria. The house shaking as she hid in the bathroom hugging her child. Rain whipping in through bolted storm shutters. Emerging afterward to see her neighborhood looking like it had been bombed. But, she said, people stuck together and helped each other; everyone got through. “Most of all,” she said, “it was amazing to see how nature kind of showed the way in this resilience. In two, three days, you could see some leaves already popping up and trees recovering. Many people made the connection between nature, the trees, and how we will also recover. It was emotional to see the immediate impact. But also it has been fabulous to be able to see nature recover every day, and how it has carried us.”
This project is supported by the U.S. National Science Foundation and The Earth Institute.