One of my goals with these articles is to spread my fascination with the microbial world. Sometimes, a topic comes along that is just so cool and so visually interesting that I don't even have to try very hard. How could you not be excited about the idea that billions of years ago the Earth might have been purple?
Since when is this less attractive than this?
Composite image made by myself and released under CC3. Original image of Kenorland courtesty of 242*1eqasdf under CC4. [Original image of Earth]
) courtesy of Lars H. Rohwedder, under CC3.
Light harvesting 101
To really understand what's going on, there's a few facts you need to know. Most of these you probably remember from science class, but let's just make sure we're all on the same page.
First, light is just a bunch of electromagnetic energy that we happen to be able to perceive because we evolved some chemicals that react in certain ways whenever the right little packets of light excite them.†. The different colors of light we perceive all have different wavelengths, ranging from red to blue, with green in the middle. (Important number alert green is at ~550 nm. To give you an example of how tiny 550nm is, if you were 550nm tall, at least four of you could lie down end to end in a coach class airline seat.††)
Second, the light hitting the Earth isn't distributed equally. To begin with, the sun does not emit all wavelengths uniformly, this is known as its emission spectrum. In addition, all that light has to pass through the atmosphere before it gets to us. Certain wavelengths of light get absorbed by different molecules and don't make it to us. By the time the light reaches the surface of the Earth, the majority of it is in the visible spectrum. This is not a coincidence - can you guess why?
Third, unless you've been living in a cave, you know that some organisms can harvest the energy from light. It's important to remember that the color they appear is from the wavelengths of light they don't use.
Wait, plants are green, that makes no sense!
Cue the record-scratch sound effect
Most plants and algae appear green, meaning they aren't using that portion of light for photosynthesis and green is the the portion of visible light we get the most of. If you're asking yourself 'what's up with that?', you're in good company.
A little over a decade ago, Shil DasSarma and colleagues proposed a colorful hypothesis suggesting what, indeed, was 'up with that'. The hypothesis states that that not using green light truly isn't the best strategy and that during the early history of Earth lots of organisms actually did use green light, and would've have appeared purple from reflecting the red and blue portions of the spectrum.
Evidence please?
That idea is at least plausible, and there's some evidence which supports it a bit. First, there are purple bugs living on Earth today. Among them is a genus called the Haloarchaea and, as the 'archaea' in the name may suggest, they have a deep link to the evolutionary past. Beyond that, they can be found growing in dense biofilm mats which, based on fossil evidence, resemble those we believe existed back then. There's also some fossil evidence that a different, but similarly colored, group of bugs, the purple sulfur bacteria, were around back then. If they were around, the sulfur containing byproducts of their lifecycle would have actually made the water toxic to green algae and cyanobacteria.
Biochemically, haloarchaea harvest light using a chemical called retinal, which is a lot simpler than the chlorophyll we're used to. Evolutionarily speaking, the simpler a molecule is, the likelier it is that it arose earlier in Earth's history (albeit that's more what'd you'd call a guideline, than a rule, really). In the specific case of retinal vs. chorlophyll, there's still a whole lot of debate about their evolutionary history.
So where did all these green things come from?
The way organisms gather energy (bioenergetics) has a lot of parallels to economics. Sure, you can go mainstream withe the most profitable strategy, but deal with a lot of competition, or you can choose a less profitable niche and have it all to yourself. Let's call this the 'South Jersey Pizza Place/Pit BBQ tradeoff'. Eventually some bugs gave up on competing for green light, and decided to harvest the lesser-used blue and red light. This required them to develop a few kinds of chlorophyll and made them appear green.
But, why is everything green now, instead of some mix?
At first, they most likely lived in microbial mats with their purple brethern, perhaps in different layers. At some point something happened to allow the green guys to outcompete the purple ones. No one is quite sure what that something was. They definitely increased the oxygen concentration of the planet, and it may be that this gave them an environmental edge (although haloarchaea actually needs some oxygen).
Other people think it may be that the early green bugs were more efficient using the light from their chosen niche, so that they actually had more net energy to work with. The mechanisms by which retinal and chlorophyll containing bugs actually eventually wind up minting ATP (the gold standard for biologial energy) from light are certainly different enough to warrant that idea.
If you want a really good idea of what goes into cholorophyll-based photosynthesis, you should definitely check out 
This could all be wrong
It's important to remember that this is a hypothesis, there are alternative explanations. For example, Dr. des Marais (and many others) point out a few flaws. The simplest of these is that when it comes to light, there can be too much of a good thing. The green light might have been avoided from the beginning because it was too much energy to manage without destroying cells.
It's also fairly damning that while the retinal system is pretty simple, it requires arginine as a precursor - and arginine production requires oxygen. If there were no green bugs around pumping out O2, there wouldn't be enough free oxygen to make arginine. However, it's possible that early biosystems didn't use retinal itself (like those purple-sulfur bacteria I mentioned earlier). Also, although the oxygen requirement kills the idea of retinal occurring before oxygen producing photosynthesis, it doesn't rule out a mutual coexistence where a few green bugs made some oxygen that was was used to make retinal.
One big point though: Being wrong is OK
Hypothesis are and probably should be wrong very often. To quote They Might Be Giants in Science is Real, "A scientific theory isn't just a hunch or guess--it's more like a question that's been put through a lot of tests".
The hypothesis is the early form of the question being put through a lot of tests. Let's assume that when we try to come up with explanations for things that we are more often wrong than not. If hypotheses weren't often proven wrong, then that would mean there was something disturbingly incorrect about the scientific method. Without this testability, we have to rely upon pure chains of logic - which can lead to argument by authority and can go notoriously wrong, such as the dogged insistence by many that the heart was the seat of intelligence (the brain was just a heat radiator for it), because Aristotle reasoned it out and said so.
One hallmark of good hypothesizin' is when the original authors are honest about it being a 'good idea, but not The Truth', which is how I interpret DasSarma's statements in interviews. They and their co-authors never say 'the Earth was purple', they say 'this could explain a few things, and it poses a few interesting questions'. This kind of attitude is especially important in areas where testability/falsification is difficult (such as some parts of paleohistory and astrophysics).
There's very good news here though. Even if a hypothesis is dead wrong, or even just difficult to test, it can still cause people to think about things from a new perspective. And that is exactly what happened here.
Astrobiology!
Sure, we can't look at Earth billions of years ago (without violating causality and developing much better telescopes), but we can look at other planets. As it turns out, one of the things we look for when trying to detect life on other planets is a characteristic brightness in infrared called the red edge and caused, specifically, by chlorophyll. The authors themselves hint in their abstract and are more explicit in interviews that this means that we might be missing important signs of life because other planets may have very different signatures.
Although the purple planet hypothesis isn't the only reason, astronomers now consider the varying emission spectra of stars, the absorption effect of different gas components, and potential alternative biological photosystems when trying to come up with fingerprints hinting at life. It does look like blue light is pretty commonly absorbed, so that might be a promising 'universal' signature.
One final cool fact, if purple isn't goth enough for you
Simulations of planets in binary star systems show that it's possible for multiple photosystems to co-exist in mutual abundance, leading to pitch black 'plant' life.
†And once the signal hits our brains, the real fun begins in how we perceive and classify them.
††The astute reader may raise reasonable objections to this comparison. However, it is technically correct.