The scientists were a little tired and burned out. For two weeks, they had been aboard a research ship in the Gulf of Mexico, trying to find and analyze deep-sea communities of coral on the dark bottom, nearly a mile below.
A robot submersible was down there now. Charles Fisher, a Penn State biologist who specializes in corals, was doing other work as he kept an eye on the video feed.
Suddenly, he stopped.
They had found the reef they were searching for.
And it didn't look good. Something was wrong.
Two years after the April 20 explosion on the Deepwater Horizon drilling platform and a three-month gusher that released five million barrels of oil into the Gulf of Mexico, the laborious and meticulous process of quantifying the damage is still under way.
Much of the research by the government remains confidential because it may ultimately be part of legal proceedings against BP, the company that owned the rig.
But some results are being made public, including the recent findings of several Pennsylvania scientists studying deep-sea corals.
Last month, coral experts from Temple University and Pennsylvania State University, plus a geochemist from Haverford College, published a paper in the Proceedings of the National Academy of Sciences, showing that coral colonies at a site seven miles from the well - and in the path of the drifting oil - suffered severe effects.
The corals showed "widespread signs of stress," according to the paper.
Of 43 kinds of coral, a quarter of them showed impacts to more than 90 percent of the colony. They were covered with brown goo. They had lost tissue.
To Eric Cordes, a Temple University biologist who also was on the ship, it indicated "an ongoing process of death." When the scientists got coral samples on deck, "we could see that everywhere they had been covered, the tissue was either gone or completely degraded."
For Fisher, "this is exactly what we had been on the lookout for during all dives, but hoping not to see anywhere."
It stood to reason the brown goo was petroleum, but for proof, the scientists turned to Haverford's Helen White. She, like the others, eventually climbed into a submersible and traveled an hour to the seafloor to have a look and obtain samples.
In the lab, she was able to fingerprint the substance and determine that it was oil and was from the spill.
"It really underscores how unprecedented this spill was in terms of its size," White said. Most spills are on the surface. This one happened at depth. "It really shows this spill had the ability to impact more than a typical surface slick."
Robert Haddad, who heads the effort to assess the damage of the spill for the National Oceanic and Atmospheric Administration, said the findings were important.
"A concern we all have, and something we don't have good answers on yet, is how extensive" the damage to the corals is, he said.
"And then, what's the long-term prognosis for those corals? . . . Are they going to come back? Are they going to die? If they die, what does that mean? From our perspective, how do we value that?"
NOAA's mandate is to "restore, replace, or acquire the equivalent," Haddad said. "The question becomes: How do you restore or replace these types of natural resources?"
Deep-sea corals aren't the colorful reef animals of shallow tropical water most people think of. In the dark, eerie world they inhabit, there's no photosynthesis or algae. The corals eat small shrimp and crustaceans.
Their communities are wildly productive, attracting many fish, including species of commercial value.
Actually, long before the spill, Cordes, Fisher, and other scientists had been studying the deep-sea corals in the gulf because so little was known about them - their ecology and biodiversity - and even simply where they were. The scientists were trying to map them so drillers could stay away.
Almost ironically, the dead and dying corals had captured the evidence of what caused their demise.
"When corals are stressed, they release a lot of mucus," Cordes said. "These corals were covered in this mucus, which was basically trapping everything that went by. . . . They happened to save a record of the oil passing through the site."
Jackie Savitz, a senior scientist with the national nonprofit ocean advocacy group Oceana, said the group's conclusions were "incredibly important, but not surprising. When we saw the oil gushing out of that pipe . . . we knew that any marine life that came in contact with that oil would likely suffer. Unfortunately, it takes time for scientists to demonstrate what we generally already know."
A problem, she said, is that this study is merely a snapshot of one coral community in one place. "There are plenty more corals out there that we could go study and probably find similar effects."
Likewise, NOAA recently released information about dolphins in Barataria Bay, La., showing signs of "severe ill health." They are underweight, anemic, and show symptoms of lung and liver disease.
"Not a big surprise," Savitz said. "But, again, there are a lot of other dolphins out there that didn't get studied."
Plenty of work is going on, NOAA points out.
Among the projects:
Damage assessment and response teams have surveyed more than 4,200 miles of shoreline and documented about 1,100 miles of oiling.
As of September, researchers had collected 8,366 birds. Most of them - 6,490 - were dead or died during rehabilitation.
As of Jan. 25, nearly 50,000 contaminant chemistry samples had been collected, and by now more than 23,000 have been analyzed.
During the spill, researchers counted 574 sea turtles in the gulf's sargassum habitat, of which 464 were visibly oiled. Researchers have collected more than 500 turtle eggs for chemical and toxicological analysis. They're amassing data from satellite tags on 28 sea turtles.
For the first time in the history of NOAA's Natural Resource Damage Assessment program, said a spokesman, Ben Sherman, the agency has posted for public access the full work plans and raw data on a public website: http://www.gulfspillrestoration.noaa.gov/oil-spill/gulf-spill-data/
Haddad said he was not sure when the job would be finished. But, he said, "we don't want to wait 10 years to understand it all and then start doing restoration. We don't think that's in the best interest of the environment."
He said a likely component of the eventual settlement agreement with BP would be a clause "that allows us to come back at a future time" if they find new impacts.
What amazed Haddad as he started to map out the damage-assessment campaign was how complicated the ecosystems and their interactions were, and how little scientists actually knew.
"For a body of water that's so intimate to our nation and so important to our nation, it's surprising the ecosystem information and modeling that we did not have."
Now, they're getting it.