How a 30-Year Experiment Has Fundamentally Changed Our View of How Evolution Works

It’s 1988. A young scientist sets up a simple evolution experiment involving small colonies of bacteria growing in flasks. To be safe, he tries out two variations, each with slightly different parameters. Though they exist side by side, these two experiments are essentially parallel universes, mirrored versions of each other with just a few pieces swapped out. 

One of the experiments fails within months — the bacteria quickly stop growing like they should. But the other set of colonies prospers. The experiment eventually evolves into one of epic proportions — one that will offer unprecedented insights into the mechanics of evolution and yield a sprawling ecosystem of laboratories and researchers.

Was it the small disparities between the two experiments that made such a big difference? Or was it just dumb luck that one succeeded and one failed? Those questions, coincidentally, are exactly what the experiment’s creator, evolutionary biologist Richard Lenski, has been trying to answer.

Thirty-one years later, Lenski stands in the middle of his lab at Michigan State University. He’s tall, with a jovial face and crinkled blue eyes that flash with intensity when a topic he’s passionate about comes up. In front of him is a glass-lidded chest where 12 populations of Escherichia coli sit in flasks, gently shaken every so often by the machine inside. These small containers hold the latest iteration of the experiment that has become Lenski’s defining achievement: the Long-Term Evolution Experiment, or LTEE. 

Lenski is tracking, at the genetic level, the changes that happen to his 12 populations of bacteria. Seeing which mutations appear and linking them to physical traits lets him watch the bacteria experiment with their own biology in real time. Then he can study why one mutation might stick around while another disappears.

It’s a question with deep implications. The mechanics of evolution are responsible for every single creature that lives on the face of the Earth, and ever has or ever will. Understanding how evolution works at a fundamental level would be a major step toward answering a big question: Why does our world look the way it does?

At its core, evolution is a paradox. It is driven by a seemingly random process — the appearance of mutations — and yet the results are anything but indiscriminate. Untangling the mechanisms that carve order from chaos is the ultimate goal of Lenski’s experiment.

What’s made studying evolution difficult, though, is that it is almost impossible to isolate in nature. When scientists design an experiment, they do everything they can to make sure nothing changes except the one variable they want to study. But the mutations that drive evolution happen against the background of a complicated and constantly shifting environment. 

In the LTEE, however, where every generation of bacteria is introduced to the same set of conditions as its ancestors, there is no variability. And when they grow through more than six generations every day — the equivalent of more than 150 years of human history — the once-glacial process of evolution begins to play out on a time scale more suited for us to witness.

As organisms reproduce, mutations naturally occur. These small errors happen as genetic material is copied. Mutations are often bad, but sometimes they’ll lead to a trait that’s helpful. The process known as natural selection weeds out the damaging mutations with a simple, severe logic.

In Lenski’s experiment, as in the real world, bacteria with harmful mutations will replicate more slowly or die, so they won’t pass their mutated genes on to as many offspring. Those with new traits that make them reproduce fastest become most abundant, which means their offspring are most likely to get included in the next iteration of the experiment. This increases the odds that the most beneficial novel mutations will become a permanent fixture in the genetic makeup of the whole population.

As the bacteria stack up helpful mutations, they begin to improve, or, in the parlance of evolutionary biology, become more fit. In the LTEE, fitness is measured by how fast the bacteria reproduce.

In Lenski’s 12 glass universes, the temperature is 37 degrees Celsius, the same as your body’s 98.6 degrees Fahrenheit. It’s been 37 C for the past three decades. More than 70,000 generations of bacteria have lived and died inside flasks just like these. 

Their care and feeding proceed with precision. The bacteria live in water dosed with nutrients like glucose, the sugar they eat, as well as citrate, a compound that helps them take up iron. Every day, a lab member takes exactly 0.1 milliliter (the equivalent of about two drops) from each 10-milliliter flask (200 drops) and transfers it to a new one filled with fresh water and sugar. The bacteria that make it into that pipette each day are allowed to continue multiplying. Every 500 generations (about every 75 days), a sample from each population is stored in a super-cold freezer.

The daily transfers are currently managed by graduate student Devin Lake, though most lab members take part at some point. Lake recently inherited his position from longtime lab manager Neerja Hajela— a photo of her knighting Lake with a baseball bat hangs above the lab bench.

Multiple generations of scientists and students have participated in the daily transfer ritual. For 31 years, the goal has been the same: figure out the mechanisms by which evolution works. It’s long been obvious that beneficial traits stick around in populations. But insights into how that happens are far less common. The LTEE has begun to provide those, and it’s changed the field of evolutionary biology.

“[The LTEE] informs everything we do in experimental microbial evolution. It’s the foundational experiment,” says Michael Baym, a biologist at Harvard Medical School who studies bacterial evolution. “I’m not sure I can tell you how it’s affected my thinking, because I’m not sure I can conceive of being in this field without this experiment existing.”

Today, scientists around the world work with populations of bacteria derived from the LTEE. Multiple long-term evolutionary experiments have sprung up, inspired by Lenski’s work, and a number of his former students have gone on to establish labs of their own devoted to the study of evolution.

To track the bacteria’s progress, Lenski orchestrates competitions. This involves growing the most recent iteration of each population alongside its ancestors (resurrected from the freezer.) The goal is to see how much faster each of the 12 populations reproduces compared with its ancestors and measure their rate of improvement over time.

Today, more than 70,000 generations of growth have made winners of them all. The most recent generations of bacteria in all 12 lines have accumulated dozens of beneficial mutations that let them reproduce about 70 percent faster than their ancestors. Where it took around an hour for the ancestral group to double their numbers, it now takes about 40 minutes.

But those mutations successful enough to last are increasingly rare. Billions of changes have occurred during the course of the experiment — Lenski estimates that every single letter of DNA in the E. coli’s 4.6 million base pairs has been mutated multiple times. The few that were helpful — and lucky — enough to survive are the barest sliver of the nearly infinite possibilities.

That’s not to say that each population has followed the same path to genetic empowerment. The 12 vials contain bacteria that all look noticeably different from one another, at least at the genetic level. Though most are within a few percentage points of one another in rate of reproduction, the strategies each has acquired are quite different. These variations likely come down to a concept scientists call historical contingency — that big differences are made possible by a string of small, sometimes seemingly inconsequential, changes in the past.

The idea crops up repeatedly in the LTEE. Many of the most significant alterations to the bacteria can be traced back to seemingly inert mutations that primed their genomes for greater things. When it comes to genetics, historical contingency carries extra weight, because many traits are not the product of a single lucky mutation, but rather a chain of interconnected genetic alterations, each contingent on the last.

It’s a paradigm that’s applicable to our own lives. Should I have taken that train? What if I’d never read that book? Maybe I should have talked to him? Every choice begets its own set of alternate realities too complex to divine. It is only in retrospect that our paths become obvious.

The LTEE is far from the first experiment to watch evolution happen. As far back as the 1800s, researchers have watched microorganisms grow in an attempt to understand the dynamics behind their adaptations. One experiment in the 1940s had much in common with the LTEE, down to the strain of microbes used. Francis Ryan, the Columbia University researcher behind it, died of a heart attack before publishing any data.

More recently, researchers have kept fruit flies for thousands of generations to watch them evolve. But none of these experiments has provided the same level of insight as the LTEE.

Understanding why is difficult. But there is one factor that seems to underlie it all: Lenski himself.

“I don’t really think it can be understated how important Rich’s experiment, combined with his personality, frankly, have been in really developing this whole field,” says Michael Desai, an evolutionary biologist at Harvard University. “It’s really remarkable to me how one idea, one experiment, one person can have had such a broad impact.”

Not every period of the LTEE’s existence has been productive, so it was inevitable that Lenski would think, at some point, about shutting his experiment down. At times, the experiment seemed to be kept alive by nothing more than sheer inertia. 

“It was going gangbusters, well past 10,000 generations,” Lenski says. “But a couple of things got me thinking about possibly ending it.” At that point, roughly a decade into the experiment, the most exciting things had occurred years ago. The LTEE might conceivably have been tapped out.

Lenski came close to quitting completely in 1998, when he was on sabbatical in France and had begun playing with software called Avida. It was essentially a computerized version of his own biological experiment. It let him watch digital organisms evolve and multiply at a rate far beyond what his own experiment could achieve, and it didn’t involve the LTEE’s tedious daily transfers or freezer storage, not to mention the never-ending process of grant applications.

On the verge of abruptly ending an experiment that had been going for 10 years, he was talked into continuing by his wife and colleagues. It would turn out to be one of the most fortuitous decisions in the experiment’s history.

In 2003, one of Lenski’s students entered the lab to find what he thought was a mistake. One of the flasks had turned opaque overnight, clouded by a sudden overgrowth of bacteria. Tests showed that nothing was amiss experimentally — but somehow, the microbes inside were growing much faster than they had before.

Lenski and his students found that the bacteria had evolved the ability to eat the citrate added to their flasks to help them ingest iron. This new ability had staggering ramifications. The bacteria could now access stores of energy previously out of reach, allowing their growth to surge. None of the other populations had gained the ability before, nor have any since.

Zachary Blount, now a postdoctoral researcher in Lenski’s lab, set out to find the critical mutations that, by building on one another, had led to the bacteria’s newfound superpower.

Lenski and his team came to call these “potentiating events”: mutations that may not have resulted in noticeable changes at the time, but which made specific future mutations more likely. There were at least two of these potentiating steps, and countless other smaller events, that edged the bacteria slightly closer to the dramatic change.

“There was nothing, as far as we knew, remarkable about that population,” Lenski says of the citrate-eating bacteria. “[But] a little tinkering was going on so that when this one event happened and then another little tinkering happened, suddenly the world changed.”

They then went back and restarted the population from older samples stored in the freezer to see if the same adaptation would appear. It would, they found, but only after a certain point. A specific series of mutations needed to happen first, and there was no guarantee they would.

“It’s hard to quantify how easily at these different time points they could have gone these different ways,” Lenski says. But the odds swung dramatically from being almost impossible to much more likely as the bacteria grew and changed.

Other populations of bacteria could conceivably have gotten close to gaining the same ability, but been thrown off course by an errant mutation. Once passed, that crossroads may never return — an alternate universe whose doors have shut forever.

It’s an intriguing question for Lenski, not least because it hints at a kind of meta-narrative surrounding the project. “There’s this wonderful tension between the randomness of it all and the predictability of certain aspects, and that interplay is just fascinating,” he says. “That’s what the experiment is about and it’s written into the DNA of the experiment itself … I would argue it’s written into the experiences of all of us.”

Thirty-one years on, his experiment turns out, in a very real sense, to be explaining its own existence.

At some point along the way, Lenski realized the LTEE should never stop. It’s partly due to the fact that it took tens of thousands of generations before he discovered truly interesting things. And more surprises might await within a few months or years.

But the decision to carry on is also based on one of Lenski’s most provocative findings, one that necessitates a time span verging on infinite.

Throughout the experiment, the E. coli have gotten consistently better at reproducing. Though their growth rate has slowed over time, none of the populations has stopped improving. Based on his calculations, Lenski now believes they never will.

He began the experiment with the assumption that evolution would one day grind to a halt, he says. “Now I realize I was totally stupid to think that.”

The bacteria’s steady improvements have been so predictable that they can be mapped out mathematically. Lenski’s former student Mike Wiser found that he could explain their increasing rate of multiplication with a function known as a power law.

Graphed on paper, a power law curve looks superficially like a hyperbolic curve — a line that curves gradually to meet a ceiling after a given amount of time. But that horizontal ceiling, called an asymptote, doesn’t exist with a power law; though the upward trend slows down over time, it goes right on increasing forever.

Over thirty years in, Lenski’s data indicates that in another 30 years, or 300, the bacteria will still be replicating faster than they were the day before. The supply of beneficial mutations looks inexhaustible.

Lenski is anxious to secure funding that will ensure the LTEE’s survival over the decades to come, especially as he contemplates retirement. He says he’s already picked out a successor to take the helm, and he’s gratified by the explosion of complementary long-term evolution experiments around the world now probing similar questions.

At Harvard, Desai is conducting a similar experiment with yeast maintained by robotic lab technicians that is projected to surpass the LTEE in terms of generations in several decades. It’s an evolution, if you will, of Lenski’s project, perhaps even better fit to probe the questions he set out to answer.

Generation after generation, both within the small glass flasks that house the LTEE’s subjects and without, progress is inevitable.

Nathaniel Scharping is the associate web editor at Discover. This story originally appeared in print as "70,000 Generations and Counting."