Credit: NASA
Credit: NASA

Is Everything in the English Channel (and Everywhere Else)?

The English Channel is home to seals, whales, dolphins, seabirds, a railway line, and the occasional celebrity on a fundraising swim. It’s also home to all the world’s oceanic bacteria (or, at least, the vast majority of them). Species that are usually found in mangrove swamps are here. So are those you’d typically associate with super-hot deep-sea vents. So are those from the open ocean.

Sean Gibbons from the University of Chicago discovered these overlaps by repeatedly sequencing the DNA of bacteria from the Channel to detect the signs of the rarest species, and comparing them to a global database of microbes. His work uncovered evidence that marine bacteria exist as a global “seed bank”. That is, many species (most? all?) are found throughout the ocean’s habitats but at different levels in each one.

While animals are each restricted to certain habitats or parts of the world, “we found that most known [bacterial] species can be found everywhere,” says project leader Jack Gilbert. The question is not whether they are present or absent, but whether they are common or rare. And that depends on their environment—temperature, acidity, predators, and so on.

This concept was first put forward by Dutch scientist Lourens Baas Becking, who summed it up in a famous quote: “Everything is everywhere, but the environment selects.” That was in 1934, and it’s only now that genetic technology is becoming powerful, cheap and fast enough to truly test Becking’s idea.

Here’s the problem: If you collect a sample of seawater, you can sequence the genetic material of whatever’s living in it and work out their identities. But you can’t sequence everything in one go. On your first pass, you’d only spot the most common residents. Try it again, and you’d pick up a few more. To get everything, including the really rare species, you need to do this millions, perhaps billions, of times. That’s called deep sequencing. It’s extremely deep sequencing. It’s like trying to track a city’s people using a few security cameras. Look at a day’s worth of footage and you’d spot the regular commuters. Give it a week, and new people would show up. Give it a decade, and even agoraphobes would turn up.

Gilbert’s team tried this with a set of water samples collected every month from an English Channel research station, between 2003 and 2008. (“Because it’s the most important place on Earth, and the fish and chips are really good,” says Gilbert.) In earlier shallow-sequencing studies, they had found that the Channel’s microbe radically change across the seasons, so that only 12 of 8,000 species were there all year round. But when they deep-sequenced just the sample from January 2007—going over it 10 million times—they found all the species that turned up in the entire 6- year period. They rise, they fall, but everything’s there all the time.

Now, Gibbons has compared the deep sequences from the single Channel sample to the International Census of Marine Microbes—356 batches of data glommed together to give a global inventory of sea-going bugs. This worldwide census represents all sorts of watery realms, from estuaries to the deep abyss to hot smoking vents. On average, 44 percent of the species from any of these habitats were also found in the English Channel in January 2007.

“It’s an unprecedented overlap,” says Gilbert. Even places like deep-sea vents and mangroves, where unusual conditions should lead to highly specialised residents, are home to bacteria that are also found in the Channel.

Now, 44 percent is far from “everything”, but it’s still an astonishing result. The Channel sample was around 2 litres of seawater, collected at an arbitrarily chosen point in time, and it contained almost half of all the microbe species in the entire ocean? That’s amazing!

The critical point is the more deeply Gibbons sequenced the sample, the greater the overlap with the global census, with no signs of slowing down. He stopped at a depth of 10 million reads and got an overlap of 44 percent. With even deeper sequencing, he might have been on track to reveal the entire global checklist of species within the English Channel. However, deeper sequencing in other habitats might also have revealed the presence of rare and unique species.

The study supports the idea that most of the species in microbe communities are actually very rare, according to Stuart Jones from Michigan State University, who came to the same conclusion through his own work. “These low-abundance populations can serve as a reservoir of genetic diversity, or a seed bank,” he says. However, Jones adds that the team’s conclusion that they would see all of the ocean’s bacterial diversity if they sequenced enough is “based on extreme extrapolation of their data.”

If there actually is a 100 percent overlap, the team estimate that they’d only find it if they went over their samples almost 200 billion times—a feasible goal, given the pace of progress in sequencing technology. “That’s the hypothesis,” says Gilbert. “Give me $100,000 and I will test it.”

How could microbes be so widely dispersed? The key, as Bass Becking originally said, is that microbes are small and easy to move via currents or wind or animals. “Any bit of water in the oceans can be anywhere else once every 10,000 years,” says Gilbert, “and these bugs have lived for 3 billion years.” They’ve had plenty of time to get around.

“This means that if the environmental conditions suit the growth of a bug, it is sitting there waiting to grow,” says Gilbert. Think about the Channel. As the seasons change and the water conditions shift, different members of the same ever-present community rise to the fore, while others drop away. The same thing happens at much longer timescales. As entire ecosystems come and go—say, a deep-sea vent rises and dies—there should be a rare and resting group of bacteria waiting to take advantage of it.

Reference: Gibbons, Caporaso, Pirrung, Field, Knight & Gilbert. 2013. Evidence for a persistent microbial seed bank throughout the global ocean. PNAS http://dx.doi.org/10.1073/pnas.1217767110