When it comes to evolution, a species’ past has a massive bearing on what it might become. That’s the latest message from a 20-year-long experiment in evolution, which shows how small twists of fate can take organisms down very different evolutionary paths.
The role of history in evolution is a hotly debated topic. The late Stephen Jay Gould was a firm believer in its importance and held the view that innocuous historical events can have massive repercussions, often making the difference between survival and extinction. To him, every genetic change is an “accident of history” that makes some subsequent changes more likely and others less so. Evolution, as a result, is “fundamentally quirky and unpredictable”. In his book Wonderful Life, Gould imagined that if we replayed life’s tape from some point in the past, evolution would go down very different paths than the ones it has currently taken.
Another eminent palaeontologist, Simon Conway Morris, disagreed. He argued that life can weave its way down any number of evolutionary routes, but that its “destinations are limited”. He saw the fact that living things often converge on the same adaptations as evidence that history has very little pull on evolution. In his mind, replaying life’s tape would lead to more or less the same result, with historical contingency only altering minor details.
Of course, it’s impossible to replay life’s tape on a planetary scale but some experiments allow us the chance to do so at a more modest level. The aptly named “long-term evolution experiment” (or LTEE) is one of these. It’s the longest-running experiment in evolution ever undertaken and began in 1988, when Richard Lenski at Michigan State University bred 12 lines of the gut bacteria Escherichia coli from a single ancestor. Since then, the bacteria have been grown in twelve separate vials of sugary broth and plopped into fresh solution every day.
Every 500 generations, samples are frozen to act as a ‘fossil record’, and since the experiment’s humble beginnings, over 44,000 generations have passed. In this time, the bacteria have changed greatly. All of them are now bigger, grow faster on sugar, and take less time to establish new colonies. But recently, Lenski noticed that one lineage of bacteria have developed an extremely rare adaptation that in the entire history of the experiment has turned up only once. Why?
One in ten trillion
The bacteria are grown in broth that is low in sugar, and they usually run out by the afternoon.The broth is also rich in citrate ions but in general, E.coli cannot use these as fuel when oxygen is around. Any bacterium that evolves this ability would suddenly find itself amid a vast and exclusive energy source. But in 20 years of evolution, only one population of E.coli has managed to do this, even though all 12 strains have evolved under exactly the same conditions and all 12 have been exposed to citrate from the beginning.
Blount first discovered the citrate adaptation, when he noticed that one vial of bacteria (known as Ara-3) was much cloudier than usual, a sign that the cells that been growing on a fuel other than sugar. Unlike almost all other E.coli, some of the cells from Ara-3 were able to grow solely on citrate. To prove that these unique cells didn’t come from a contaminating strain, Blount showed that their DNA carried certain giveaway mutations that characterised earlier generations of Ara-3.
During the experiment, each population of bacteria experienced billions of mutations, and given the size of E.coli‘s genome, it’s likely that each lineage tried every typical genetic change several times. And yet, the citrate adaptation turned up only once in all of this tinkering. Based on the experiment, the odds of a bacterium developing the adaptation in any generation is just one in ten trillion! There are two possible explanations for this: it’s the result of an extremely rare mutation; or it depended on the particular history of that specific population.
Replaying the tape
To work out which, Zachary Blount from Lenski’s lab used the frozen fossil record to replay evolution from different points from this population’s past. In three separate replays, Blount cultured bacteria from the entire length of the experiment, from the original ancestor to the most recent generation. In all cases, he found that citrate users evolved much more often in samples from later generations than those from earlier ones.
These results strongly suggest that the citrate innovation was not simply the consequence of an extremely rare mutation. If that was the case, it would be equally likely to crop up at any point throughout the experiment’s history. To Blount, the fact that it happened more frequently in later generations suggests that it depended on one or more mutations that the bacteria had previously picked up, which unlocked the potential for the later innovation.
It’s possible that the ability to use citrate depends on teamwork between several genes. Any one of these could pick up the right mutation, but it would be meaningless if earlier mutations hadn’t provided the right partners. Alternatively, earlier mutations could have been directly responsible for later ones. Blount speculates that the clinching mutation might have happened in a piece of mobile DNA that had previously jumped into the right spot. Without this insertion, the critical change would never have happened. Blount and Lenski are now trying to discover this sequence of genetic events, and how they affected the bacterial cells.
One change to rule them all…
Whatever the route, one lineage of bacteria had managed to make use of citrate after about 31,000 generations. Their numbers grew quickly and by 32,500 generations, they made up 20% of the population. But 500 generations later, they had fallen back down to just 1%.
Blount believes that the first bacteria to use citrate just weren’t very good at it and reaped only marginal benefits from their innovation. As a result, they initially prospered but were soon outcompeted by other bacteria that couldn’t use citrate but had become much better at metabolising sugar. Only later did they regain their initial foothold, with further mutations that improved their unique ability and allowed them to switch seamlessly from sugar to citrate when the first fuel ran out. Eventually, these advances catapulted the citrate users to dominance. Their sugar-only peers still eke out a minority existence because they are still better at using sugar as an energy source.
So a single innovation – citrate exploitation – was enough to split a united population of bacteria into a community of two members: a specialist that focuses on sugar; and a generalist that uses both sugar and citrate.
In the team’s own words, the study clearly shows that “historical contingency can have a profound and lasting impact under the simplest conditions, in which initially identical populations evolve in identical environments. Even from so simple a beginning, small happenstances of history may lead populations along different evolutionary paths. A potentiated cell took the one less travelled by, and that has made all the difference.”
Reference: PNAS doi:10.1073/pnas.0803151105
June 2nd, 2008 at 9:42 pm
Cool! I am definitely more in the Gould camp than the Conway-Morris one so it is good to see some more solid evidence enter what often seemed to be a more philosophical debate. Another job well-done.
June 3rd, 2008 at 4:11 am
Yes. Excellent as always. But I feel more evidence needed before dismissing Conway completely.
June 3rd, 2008 at 8:12 am
Ed – Are you aware of any long-running experiments like this where the initial bacterium has accumulated sufficient mutations that in the end it would be classified as a different type (genus or something higher) of bacterium from what it started out?
I’m tired of the creationists claiming this never happens. It would be nice if there were one of these marathon bacterial studies which had something like this as a result, whether intended or not.
June 3rd, 2008 at 11:23 am
Ian:
In my experience, this wouldn’t help: the standard creationist response to speciation in fruit flies (or whatever) is “yes, but it’s still a fruit fly”. This is even more true with bacteria (after all, any invisibly-small being must be like any other, right? An amoeba is the same thing as a virus, right?)
However, I seem to recall an example of mice being brought to the Canary islands(?) by colonists in the 18th century, and evolving to the point where there are now separate genera in different parts of the same island.
June 4th, 2008 at 6:40 am
Ed – I know they’ll never accept anything as an answer(!) but it would still be nice to have something like that in my collection of evidence to pull out of the hat!
I just read on Carl Zimmer’s blog a comment by Allan Kellog(spelling?) about Myxococcus xanthus having a kingdom of its own. It might be worth following up on that, methinks.
June 4th, 2008 at 6:41 am
arensb – Sorry – I meant that last reponse of mine to be addressed to you!
June 4th, 2008 at 11:47 am
arensb: if that’s the creationists complaint, that it’s still a fruit fly, then their making a categorical error. Speciation isn’t a subjective classification. Remember, one of the criteria is the inability to reproduce fertile offspring from interspecific mating.
I was at a talk last week that demonstrated this very thing: cladogenesis in fruitfly species in the wild and lab.
June 4th, 2008 at 2:02 pm
Ian, I don’t know about experiments that have demonstrated new species, but my handy Counter-Creationism Handbook lists the following examples of new species having arisen during historical times:
1) A new mosquito species Culex molestus, isolated in London’s Underground, has speciated from Culex pipiens.
2) Helacyton gartleri is the HeLa cell culture, which evolved from a human cervical cancer in 1951. The culture grows indefinitely and has become widespread.
3) Several new species of plants have arisen via polyploidy (when the chromosome count multiples by two or more) One example is Primula kewensis.
The book also lists several examples where speciation seems imminent, where it’s in the process of happening and where it can be clearly inferred to have happened because some species only exist in environments that did not exist a few hundreds or thousands of years ago. I don’t have time to list them all, but I’d recommend getting the book.
June 4th, 2008 at 4:09 pm
Re: 1) Culex molestus — just so that you are in the clear, it is Culex so-called molestus, because it was discovered that a Southern Hemisphere species had already been granted the molestus name back in the 1860s. I don’t think they have gotten a good taxonomic replacement name yet. (The two — molestus and molestuswannabe — are completely unrelated, aside from being in the same genus.)
There is a classic of plant speciation which everyone seems to miss, though, even though it is actually very well documented. Spartina anglica (English cordgrass) is a classic hybrid+mutation speciation event — a hybridization of two very different cordgrasses, one native English — S. maritima — and one American — S. alterniflora. The first such hybrid was noticed in the 1840s, I think, around Hythe, presumably because S. alterniflora had been accidentally introduced there as a hitch-hiker in ships’ ballast. However, the original hybrid was sterile, a fact well-documented by the area’s botanists.
Subsequent to that, around 1870, an example of the hybrid underwent a polyploid mutation, and the tetraploid mutant had full fertility, and has spread ever since.
I think I know of another couple of similar examples of speciation, which do not involve hybridisation. Will go dig. Back tomorrow.
June 5th, 2008 at 11:04 am
An update:
I don’t know if it really counts as full speciation or not, but the green alga Chlorella vulgaris made a leap from a unicellular form to a multicellular organism in the lab, as a response to sustained predation by the protist Ochromonas vallescia. The slight irony in this case was that the researcher was actually studying the Ochromonas — the shift to multicellularity in the alga was noted only in passing, as a thing of little interest, in his original paper!
A few years later he went back and did another paper about the algae, though.
Boraas, M.E., Seale, D.B., Boxhorn, J.E. “Phagotrophy by flagellate selects for colonial prey: A possible origin of multicellularity” Evolutionary Ecology Vol. 12, Issue 2, 1998, Pages 153-164
(Online at http://tinyurl.com/465ekj )
Another resource that you might appreciate is the book “Frogs, Flies and Dandelions” by Menno Schilthuizen ( http://www.amazon.com/exec/obidos/ASIN/0198503938/ ) which has some excellent information on things which had been observed pre-2000, as well as some good background information on understanding speciation.
June 5th, 2008 at 11:37 am
Thanks Luna, that’s both interesting and useful.
June 5th, 2008 at 11:43 am
And the mice –
On Madeira, ordinary Mus musculus has either speciated six times in the last 500 years, or is on the verge of having done so:
http://www.genomenewsnetwork.org/articles/04_00/island_mice.shtml
And on Gough Island, there are now Evil Giant Mice eating albatross chicks, which has happened within the last 150 years:
http://www.guardian.co.uk/environment/2008/may/19/wildlife.endangeredspecies?gusrc=rss&feed=worldnews
June 5th, 2008 at 11:53 am
FWIW, the question of speciation — especially speciation events which involve the rearrangement of chromosomes or changes in chromosome structure — has interested me for a long time. If I run across any other examples I would be delighted to pitch them in your direction. (Not literally, of course, unless you *like* having mice thrown at you. I mean, some of us do…ok, shutting up now.)
March 17th, 2011 at 2:43 pm
I love Lenski’s monster experiment, and your cover of it is as always excellent. However I’m not sure whether such a result can really support Gould’s view over Conway Morris’ view. The fact that most of the evolutionary routes DID in the end achieve the same results could be actually seen as a confirmation of Morris’s view! It depends on how much importance it’s given to the exception
This is after all a manipulative experiment on a single species in controlled conditions. It sure shows that evolution has the potential to go in wildly different ways, but I would expect that genetic potential to be constrained by ecological factors. What seems evident by looking at natural communities is that, if there is a resource to be used (intended in the widest sense, could be food, habitat, etc.), most often you can be sure there is someone able to use it, and organisms adapted to the same resources, niches, whatever, tend to be similar in one regard or another – convergent evolution baby. So it seems that the phenotipical variability species can have is somewhat constrained in certain regards
Now a counter-argument could be that, since organisms themselves are a major cause of evolutionary change for other organisms, a small but important change in one lineage could have a sort of butterfly effect on others and lead to unpredictable scenarios… but still, we should agree on what we mean when we say “very different paths”. I wouldn’t suggest that, say, dinosaurs would have evolved 9 times out of 10 if evolution started back from scratch, BUT I would expect some taxon to evolve “dinosaur”-like characteristics and fill their ecological role. After all that’s exactly what happened with mammals after non-avian dinosaurs went extinct. So the Gould vs Conway Morris debate seems to me futile as long as we don’t give a meaningful definition of Gould’s “different paths” and Morris’ “minor details” (quoting your words, not sure what words they actually used). And anyway it seems to be an unresolvable issue as of now, because even Lenski’s wonderful experiment is way too simple to represent evolution in the real world