habitability' as defined by liquid surface water, temperate climate, organic chemistry, stability. It's fair enough, we have to start somewhere in our search for other worlds with life, we need to stack the deck towards something that we stand a chance of recognizing. However, this is only part of the template, the bit that is convenient and tidy.
If there is one inevitable thing about life it is that particular variants, species, modes of existence, are all prone to extinction. A new work by Drake & Griffen in Nature this week makes this point rather succinctly. They show, by subjecting water flea populations to a series of unfortunate events, how the population dynamics of a species can fundamentally shift due to environmental changes. Fluctuations in population numbers occur even in stable environments, but the character and size of these fluctuations changes in degrading environments, and beyond a certain point there is no recovery. Long before it all goes down the tube there are clear statistical indicators that things are not well - population sizes drop as the tree begins to fall.
Extinction is a fact of life. When we consider what the signatures are of life on exoplanets - chemical fingerprints in an atmosphere, selective photon absorption by bio-molecular structures (red edges), seasonal variations in albedo - we typically assume things are cozy on the ground. Given that (big) extinction events seem to happen pretty fast, one would think that the odds favor us observing a planet in the calm between these periods. However, we need to be careful about convenient truths.
We are most likely to be able to sniff out the signs of life on a terrestrial-type planet when it's in full swing. Suppose a world is having a particularly fertile episode, chock-a-block with organisms, but not a stable situation. It's prime for collapse. Relative populations will swing high and swing low. At the high point for some, a planet may show the greatest bio-signatures, and make itself far more tasty for our prying telescopic eyes. Without running the numbers it's impossible to give a precise answer, but it would seem that the odds will be shifted. We may be most likely to find not the signs of normality, but the signs of a system approaching some kind of biological collapse - just like the stock market, it's all about the fluctuations.
Another way to think about this is that success breeds vulnerability. If one measure of evolutionary success is the ability to occupy as many niches as possible, with as big a population as possible, then there are many terrestrial examples. This success comes with a price though. That level of dominance may alter the global environment itself (either chemically, or in terms of restraint on other species). If something degrades, or intra-species population dynamics get out of hand, the whole system will head south.
So, the radical thought. It may be that if/when we begin to see signs of biospheres on distant worlds around other suns, these will be places on the verge of collapse, overrun by certain types of life, populations fluctuating to unsustainable heights - albeit perhaps over timescales of hundreds or thousands of years. Our limited detection sensitivity will catch only the outliers.
It's not all doom and gloom though. If a long term goal of exoplanetary science and astrobiology is to place the Earth in proper context, and to potentially help ourselves, then watching disaster unfold elsewhere may be most helpful.