In September, 2009, Erik Cordes, a deep-sea biologist, sat in a dark control room aboard the research vessel Ronald H. Brown, which was floating in the Gulf of Mexico. Fourteen hundred metres beneath him, a remotely operated vehicle (R.O.V.) was transmitting live video and data from the Mississippi Canyon lease area, a region just southeast of Louisiana that has long been pursued by oil companies. At first, the video feed showed little more than brown mud pockmarked by burrowing worms and snails. But, a few hours into the dive, screens in the control room suddenly lit up with red, orange, yellow, and purple life forms. The R.O.V. had stumbled upon a garden of deep-water corals. “It was like floating through a forest,” Cordes told me. “It was beautiful—a different planet.” An elaborate ecosystem swirled around the coral stalks: brittle stars, glass sponges, sea urchins, shrimp, fish. Cordes was early in his career, looking forward to studying the dizzying biodiversity of the Gulf’s hidden coral communities.

Seven months later, just eleven kilometres from the coral garden, a blowout on BP’s Deepwater Horizon drilling rig caused an explosion that killed eleven workers and sent oil gushing up from the seafloor. By the time the spill was finally capped, three months later, seven hundred and eighty million litres of crude had billowed into the water—the largest marine oil spill in history. Cordes was horrified, but he hoped that his coral forest would be safe: oil floats, because it’s lighter than water. When his team returned to the deep-sea site, however, its brilliant colors were smothered under a blanket of fluffy gray scum. Mucus oozed out of the coral stems, a telltale response to trauma. The researchers realized that a mixture of oil, plankton, and a chemical dispersant—used to break up slicks before they choke coastal ecosystems—had rained down onto the seafloor. “It was shocking,” Cordes told me. “Everything kind of stopped.”

Cordes’s career took a dramatic turn. He would no longer study the unspoiled biodiversity of the Gulf’s corals. Instead, he thought of the tainted ecosystems as E.R. patients; he would conduct a forensic damage assessment. “This was disaster response,” he told me. “We needed to see who had survived. We needed to get to work.” In the years since then, Cordes and his team have followed nearly three hundred individual coral colonies within sixteen kilometres of the spill, recording minuscule changes in color, branch lengths, and physical integrity. One of the researchers on the project found that coral colonies died when they were more than half-covered by the petrochemical goo. According to Cordes’s estimates, twenty-five per cent of the corals he studied have either died or show no signs of recovery. Some were probably centuries old.

Cordes’s cohort of scientists helped to show that the devastation of oil spills goes far beyond our coasts, where birds and fish are the most visible victims. In the surface waters of the open ocean, plankton can absorb oil compounds, compromising their ability to photosynthesize and produce oxygen. When they die and sink through the water column, other marine animals eat their remains, toxins and all. Fish embryos that grow in the presence of oil may be born with heart defects, as well as spine and skull deformities. Still, he took comfort in the idea that American offshore drilling—and therefore many oil spills—might eventually be relegated to history. The Deepwater Horizon disaster sparked new regulations. Oil companies began to turn away from the areas he studied and toward deep-water deposits in other places, including the southern Caribbean and the West African coast. “We were ramping down, at least in U.S. waters,” Cordes said. Then Donald Trump took office for the second time, and in his Inaugural Address he told the nation, “We will drill, baby, drill.”

This past November, the U.S. Department of the Interior released plans to lease up to 1.27 billion acres of public waters for new offshore-drilling efforts. (Such plans don’t automatically translate into new rigs; leases had been available for years, including under the Biden Administration, but regulatory costs and low oil prices had limited their appeal.) The Center for Biological Diversity soon warned that, if the Trump Administration carried out its plans, it could cause more than four thousand oil spills—not including large-scale disasters like Deepwater Horizon. “These catastrophic incidents will become more likely as the Trump administration rolls back offshore drilling safety rules,” the Center said on January 6th. The Trump Administration went on to rescind key regulations from the National Environmental Policy Act, and to exempt drilling projects in the Gulf from Endangered Species Act requirements.

In the past month, since the U.S. and Israel began waging war in Iran, oil prices have spiked, which incentivizes fossil-fuel companies to drill new wells. On March 9th, the Bureau of Ocean Energy Management released an estimate of undiscovered oil and gas reserves in U.S. waters; two days later, it sold twenty-five leases, covering a hundred and forty-one thousand acres in the Gulf of Mexico, at record-low royalty rates. Two days after that, the agency greenlit Kaskida, a new five-billion-dollar ultra-deep drilling project southwest of New Orleans, which could start pumping eighty thousand barrels of oil a day as early as 2029. The Gulf is open for business again, and Cordes is bracing himself. “The more drilling we do, the more oil we’re going to release into the environment,” he said. “It’s really that simple.” In the past fifty years, he said, North American waters have seen three catastrophic spills: Ixtoc I, in 1979; Exxon Valdez, in 1989; and Deepwater Horizon, in 2010. “It’s been fifteen years since we’ve had one,” Cordes said. “You can do the math. We’re just about due for another.”

The first offshore oil wells were drilled in California, where black tar has been oozing out of the ground for eons. In what is now the La Brea Tar Pits, sticky pools of asphalt famously trapped dire wolves and sabre-toothed cats. More recently, Indigenous peoples such as the Chumash and the Yokuts used asphalt to make face paint, game pieces, glue, and waterproofing caulk for boats and baskets. When the British captain George Vancouver sailed through the Santa Barbara Channel, in 1792, he noted that “the surface of the sea, which was perfectly smooth and tranquil, was covered with a thick, slimy substance.” The era of oil drilling began not long after James Young, a Scottish chemist, collected a sample of petroleum near an English coal mine in the mid-nineteenth century. He heated the liquid in a flask and collected the condensate, which burned cleaner and was cheaper than whale oil, the standard fuel of the day.

Summerland Beach is a mile-long crest of sand just down the coast from Santa Barbara. The Santa Ynez Mountains rise directly to the north, pinning Highway 1 tightly to the coast. On most days, the surf is loud enough to mask the steady purr of cars. In the eighteen-nineties, oil drillers tapped into pools beneath the sand; new wells crept all the way to the surf’s edge, and eventually into the water. Droves of workmen were hired to build sturdy piers. By 1896, the offshore rigs were operational; their pipes extended down through several metres of water and a couple hundred more of seafloor sediment.

A bust inevitably followed. In 1903, a vicious winter storm reduced most of the piers to splinters, and by 1906 offshore oil production at Summerland had all but ceased. Still, a threshold had been crossed: Offshore wells proliferated. Steel piers replaced wooden structures, and rigs reached farther from shore. Along the Gulf of Mexico coast, drilling ships allowed for mobile “overwater” operations. Floating platforms moved into deeper waters. Between 1954 and 1971, offshore oil production in the United States expanded more than tenfold. Off the coast of Summerland, standalone platforms named Hazel, Hilda, and Heidi were erected in California’s waters, which extended five and a half kilometres from shore. Beyond that, in federal waters, were Hogan, Houchin, and Platform A.

On January 28, 1969, a drill extending eleven hundred metres into the sea floor beneath Platform A punched through a layer of rock and into a pocket of oil. When the crew retracted the drill to replace its bit, an overpowering jet of oil fountained from the well. They managed to plug the pipe, but growing subterranean pressure created cracks in the sea floor. Oil rushed through the sediment and rock and blackened the water. Eleven million litres of oil spread across an area of two thousand square kilometres.

Even after the leaks were plugged with cement, rivulets of oil persisted for months, and the oil spill’s ecological and cultural impacts lasted even longer. Dead seals and dolphins washed ashore. Fishermen found lobsters and crabs painted black and weighed down by oil. It was the birds, though, that seized the public’s attention and launched a movement. From Ventura to Santa Barbara, gulls, pelicans, murres, and grebes staggered along beaches, unable to fly. Locals mobilized to save them; a nearby zoo recommended feeding the birds butter to emulsify and flush out the oil in their throats. According to a Los Angeles Times report, birds fleeing their would-be rescuers instinctively waddled toward the water, and, “falling into the black liquid, they lay in the ooze, crying weakly.” Cormorants that tried to clean each other with their beaks died after ingesting the viscous muck. Others expired from hypothermia: the oil compromised their feathers’ water-repelling properties. Kathryn Morse, a professor of history and environmental studies at Middlebury College, has written that images of these birds marked a turning point in society’s relationship with the oil industry: “They contested older visual narratives of oil as abundant and powerful.” Around the same time, the Washington Post decried “the systematic fouling of our nest,” and the New York Times called the pursuit of petroleum “a survival issue both for sea life and for man himself.”

The political fallout was lasting. President Richard Nixon walked on the beach and flew over the slick in a helicopter. He vowed to take “more effective control” over the oil industry, and opined that “preserving beaches is more important than economic considerations.” The oil spill helped inspire the Environmental Protection Agency, the National Oceanic and Atmospheric Administration, the Clean Water Act, and the Endangered Species Act. California declared a moratorium on offshore-drilling projects. In a blow to the fledgling environmental movement, however, a report by Nixon’s science adviser, Lee DuBridge, called for more drilling, not less. “The situation which makes leaks possible,” DuBridge wrote, “is the fact there is oil down there. The only way to prevent future leaks is to get the oil out.” In the summer of 1969, less than six months after the spill, several additional wells were drilled from Platform A. An entirely new rig was towed from an Oakland shipyard and installed just to the east of the platform. The offshore-oil industry had weathered the environmental reckoning and emerged intact, and arguably stronger.

More spills polluted California’s beaches. In 1990, a tanker ship ran over its own anchor; a quarter century later, a pipeline on land ruptured, sending a river of oil straight to the sea. In 2010, when I was in graduate school in California, I took a barefoot walk on Summerland Beach at dusk. It was only the next morning, when my soles stuck to the floorboards of my apartment, that I noticed quarter-size globs of tar stuck to my feet, like wads of black gum. I couldn’t find any news online about a local oil spill. Instead, I discovered that such tar balls are so common that they’ve inspired a product called Oil Slick—a beach-tar-remover spray. Oil had permanently altered California’s coast. And yet, when viewed through the lens of today’s oil industry, the drilling efforts in the Santa Barbara Channel may seem quaint. In the Gulf of Mexico, an even more profound transformation has taken place.

The Gulf of Mexico formed about two hundred million years ago, when the supercontinent Pangea split up. During periods of high sea levels, salt water flooded into the basin between North and South America and formed a shallow sea where algae thrived. When the water periodically evaporated, it left thick deposits of salt. Then rivers carried eroded particles—from the Rockies, the Appalachians, and other mountain ranges that have come and gone—into the basin, covering the salt and decaying algae beneath kilometres of sediment. The weight of the sediment had the effect of a slow-motion pressure cooker. Biological molecules were deconstructed and flattened into a tangled mess of mostly carbon and hydrogen. This goo is called kerogen, and geologists estimate that it contains about ten thousand times as much carbon as all the biomass on earth today.

Kerogen is more a category than a specific thing, like jazz, or casseroles. Each molecule can contain more than two hundred carbon atoms, twisted into a dizzying array of rings and folds. At the greatest depths, where temperatures and pressures are highest, the kerogen tends to split into smaller hydrocarbons such as methane, the main component in natural gas. At more moderate pressures and temperatures, oil forms. What makes the Gulf of Mexico such a bonanza of fossil-fuel production isn’t just the sheer quantity of algae that was cooked to the right specifications but also the layers that it’s encased in. Oil and gas infuse the pores of sandstone; layers of impermeable caprock trap it inside.

In 2014, four years after the Deepwater Horizon catastrophe, I boarded the research vessel Atlantis in New Orleans for an expedition into the Gulf. I was early in my career as a microbiologist, and I was hoping to study places where methane naturally seeps from the seafloor, supporting remarkable ecosystems of deep-sea life. At dawn, we began to wind through the misty coils of the Mississippi Delta, passing pelicans that sat on mossy wooden docks. As we entered the Gulf, we suddenly came upon an unseemly jumble of cranes and metal scaffolding—a vast oil rig. More were soon visible all around us, ringing the horizon.

It’s nearly impossible to definitively determine the number of offshore oil rigs in the Gulf of Mexico. They are constantly built and removed as the oil market ebbs and flows. NOAA estimates that about six thousand structures have been built over the water since 1942, and many of them remain viable. Some are not much more apparent than a telephone pole. The rigs we saw from Atlantis were more akin to floating cities. There are likely a couple hundred such platforms currently operating in state and federal waters off the coasts of Alabama, Mississippi, Louisiana, and Texas.

Our destination was the Mississippi Canyon—an incision, now underwater, that formed roughly thirty thousand years ago, when the Mississippi flowed more than a hundred kilometres past its modern-day delta. My goal was to inspect calcium carbonate rocks, found near methane seeps at the base of the canyon walls, in search of methane-eating microbes. I imagined that they might be unsung heroes of carbon sequestration in the deep sea, turning methane—a potent greenhouse gas—into rock.

One morning, a few of my colleagues crawled into the ship’s submersible, Alvin, and descended to the seafloor, where they found “microbial mats” growing on mounds of craggy gray rock. At one site, in a trough, they found frozen methane hydrate—bone-white ice—and a thin stream of bubbles. Using Alvin’s robotic arm, they plucked a few rocks from the seafloor. Later, in the ship’s main lab, I placed those rocks in sterile plastic bins and readied my tools; my plan was to chisel off fist-size fragments in search of microbes. Before I could lift my hammer, however, I noticed pinpricks of black liquid covering the rocks. The dots soon became acrid splotches of oil. At first, I mistook the splotches as remnants of the Deepwater Horizon spill. But then I realized with a start that I was seeing a natural process. The rocks, recently depressurized after their journey from the seafloor, were oozing petroleum.

In the end, I found the hydrocarbon-eating microbes that I was looking for. They’re a type of archaea that perish in the presence of oxygen—and they form intricate relationships with bacteria that breathe sulfur compounds. My colleagues, meanwhile, used forceps to pick glistening scale worms and peanut worms from the rocks. The life amid the oil was a stark reminder that, for all of petroleum’s destructive power, it’s still a natural substance, capable of fuelling ecosystems as well as cars. It’s as much a product of nature as orchids or tree trunks. In some parts of the Gulf, hydrocarbons gush directly out of the seafloor. Mud volcanoes ooze liquified brown sediment and emit shimmering curtains of methane bubbles. Oil stalagmites emerge from the seafloor, pulled upward by their buoyancy. Asphalt ribbons form “tar lilies” that resemble toothpaste squeezed from a tube.

Considering the quantity of hydrocarbons that are naturally entering Gulf of Mexico waters, what exactly constitutes an environmental disaster? If some oil sloshes off a ship or leaks out of a pipeline, what’s the difference? Mandy Joye, an oceanographer specializing in microbiology at the University of Georgia, has spent decades characterizing microbes in hydrocarbon-rich habitats across the Gulf. Her discoveries have underscored a key biological principle: virtually every flavor of hydrocarbon can be eaten by something. However, the microbes that digest them are often found at low abundances: they are members of the “rare biosphere,” and, under typical conditions, they can take months to replicate. Ten times the normal rate of hydrocarbon seepage, Joye told me, is “probably the cutoff between a natural process these ecosystems can handle, and an anthropogenic disaster that they maybe can’t.” Above that threshold, the microbes can’t keep up.

The other problem is when the oil goes places it shouldn’t. The oil industry amounts to a vast program of oil relocation and transformation. It sucks crude from subsurface rocks into metal pipes, tanks, and refinery reservoirs, spitting carbon dioxide into the atmosphere all the while. When this relocation process breaks down, and viscous black liquid flows into marine environments, plants and animals are evolutionarily unprepared. The Center for Biological Diversity estimated that the Deepwater Horizon disaster harmed or killed more than eighty thousand birds, six thousand sea turtles, and twenty-five thousand dolphins and whales. Oil coated twenty-one hundred kilometres of shoreline; it suffocated salt-marsh grasses and oysters and, by weakening root systems, accelerated erosion. One study valued the environmental damage at $17.2 billion. These assessments focussed on surface waters and coastlines—the part of the Gulf we interact with most frequently. Yet much of the oil never made it to shore; it fell into the deep sea. What, exactly, happened down there?

There’s a theory in the study of microbial ecology, attributed to the Dutch polymath Lourens Baas Becking, that “everything is everywhere, but the environment selects.” The first part—“everything is everywhere”—argues that the swirl of wind and water transports organisms to every habitable nook and cranny on Earth. The second—“the environment selects”—suggests that environmental conditions produce microbial winners and losers. (A photosynthetic algae may sink to the seafloor, but it’s not going to survive in complete darkness.) The hypothesis is imperfect, but one could see its principles at work in the aftermath of Deepwater Horizon. Scientists have understood since 1913 that certain microbes are capable of degrading oil; many of them make up a tiny minority of the microbes in a given ecosystem, but they were present in the Gulf.

In May, 2010, Joye was the first microbiologist on the scene of the spill. She was part of a monitoring effort that sent a water-sampling device within a hundred metres of the discharge zone. When they pulled water samples on board, the bitter scent of hexane—a component of petroleum—permeated the respirator that Joye was wearing. “Those samples were the nastiest thing I’ve ever seen,” she told me. She watched with disgust as a colleague’s protective gloves disintegrated around his fingers. Over the years, she and her fellow-scientists sampled the water dozens of times. They assembled a detailed account of the microbial population during different phases of the oil’s degradation.

About half of the oil from Deepwater Horizon accumulated in a soupy plume that hovered at a depth of around eleven hundred metres. Within about a month, a lineage of Oceanospirillales microbes proliferated, feasting on cyclohexane. (Microbial populations inside the oil plume were about two orders of magnitude higher than in the uncontaminated water, and ninety-nine per cent of them had oil-degrading properties.) A few weeks later, Cycloclasticus stormed onto the scene, fuelled by a range of polycyclic aromatic hydrocarbon molecules. Then came Colwellia, feeding off short chains of oil-derived hydrocarbons such as ethane and propane. Once the well was capped, the community shifted again: microbes from the Flavobacteriaceae and Rhodobacteraceae families mopped up longer-chain hydrocarbons. In the water above the blowout, Baas Becking’s aphorism proved true: the key players were lurking the whole time. As the plume’s chemical profile changed, different types of microbes bloomed and then receded in a sort of ecological relay race. Joye estimates that thirty to forty per cent of the oil was degraded by microbial communities—a feat for which we can only be grateful.

Even so, tens of millions of litres of oil swirled through the Gulf. Much of it was lofted by tiny gas pockets into the upper water column, only to fall back downward after the bubbles fizzed into the atmosphere. The effect of the sinking oil, Joye told me, “was like dropping flypaper through the water column—all of the biology was just getting stripped out and moved to the seafloor, as were the nutrients associated with those organisms.” In some places, fish catches declined precipitously in medium depths, which Joye attributes to nutritionally barren conditions.

The effect on deep-sea organisms was even more dire. Five centimetres of flocculent “oil fluff” blanketed the seafloor, the equivalent of thousands of years of sediment dumped in just a few months. “It was completely absurd,” Joye said. For crabs, worms, and other invertebrates, “there was nowhere to run and hide.” Cordes, the coral researcher, said that some of the oil-coated corals recovered, but most didn’t. They lost branches, and their withering skeletons were colonized by hydroids. Deep-sea corals operate on different time scales than their shallow-water relatives; they can live for thousands of years. As the oil settled, centuries of growth were undone. These ancient ecosystems “are easy to remove,” Cordes told me, “but impossible to replace.”

Conservationists have become creative in an effort to rescue damaged habitats. The area around an oil spill can be seeded with “probiotics” of naturally occurring, hydrocarbon-eating microbes, which help shallow-water corals fight oil damage three times faster than corals left to their own devices. Artificial corals can give deep-water marine life a place to live; cat sharks, for example, wrap necklace-like egg cases around any structure, natural or not, that sticks up from the seafloor.

Even the infrastructure of the oil industry can become a habitat. A “rigs to reef” strategy converts derelict wells into scaffolds for new reefs—empty apartment buildings, essentially, welcoming new denizens to move in. In the Gulf, more than six hundred offshore oil-and-gas platforms have been “reefed” since the eighties, and many are now oases for fish such as amberjack, filefish, spadefish, and red snapper. (They’re also often overrun by invasive orange-cup corals and lionfish, which crowd out native fauna.) The deep sea may wind up full of these prosthetic habitats—ecosystems that function in some ways, but that remain diminished and hollow compared to their natural state.

Even after studying deep-water corals for three decades, Cordes continues to make new discoveries. By 2024, his team had mapped nearly all of an enormous reef off the Carolina coast, perched a thousand metres down. “It was this incredibly diverse, beautiful coral reef,” he told me. There were some species he’d never seen before. He often wonders how many hidden coral ecosystems could be destroyed—and how many were lost before we even knew they existed.

One of his most painful findings emerged among the gray, oil-coated corals near the Deepwater Horizon blowout site. Purple-pink brittle stars still clung to the corals, their tendril-like arms wrapping around the branches like yarn. In late 2010, Cordes and his team noticed that some parts of the corals had cleared up: the brittle stars were removing the oil, revealing vibrant yellow tissue beneath. It was the first direct evidence of a mutualistic relationship between the two species. The brittle stars got an elevated perch from which to feed; the corals got a cleaning.

Deep-sea scientists rarely have a chance to return to the same places, but Cordes continued to revisit his Gulf corals. Nearly a year later, he was in the control room of the Ronald H. Brown, watching live video from a decimated coral garden known as MC294. He leaned forward as the R.O.V. approached an intricate coral colony that he’d come to know well. When he spotted its distinctive dendritic branches, its off-kilter lean, and its patches of healthy yellow tissue, he was heartened: the coral was on the road to recovery. But something was missing—brittle stars. After cleaning so much oily residue off the corals, Cordes concluded, they had died. “It was tragic,” he told me. “It felt like, in order to see these fundamental truths, something awful had to happen.” ♦

This is drawn from “The Dark Frontier.”