Resting stage

Under the right conditions, mud keeps things alive.

Not in any dramatic sense.  Most of what settles into a seabed is simply dead, and stays that way.  But a lot of small organisms, when conditions get bad enough, have the option of making a resting stage - a thickened, shut-down version of themselves that does almost nothing and waits.  Some diatoms do it.  Water fleas do it, in the form of eggs with tough coats.  Plenty of bacteria do it.  The cell winds down to something close to zero and sits there.

If it sinks into sediment that nobody stirs up, it can sit there for a very long time.

And in a quiet basin where fine material accumulates in undisturbed layers, the sediment is also a calendar.  You core down, you count the layers, you date them, and you know roughly what year each depth belongs to.  Which means that if you can persuade the cells at a given depth to wake up, you can grow an organism from a known decade in a dish on your bench today.

That is an absurd thing to be able to do, and I don’t think it gets enough attention outside the small number of fields that use it.

Almost every experiment in biology compares things that exist at the same time.  You take strains from two places, or two treatments, or a mutant and its parent, and you compare them side by side, and every claim you make about change over time is an inference stitched together from things happening now.  Museums and herbaria give you dead material, which is useful, and increasingly you can get DNA out of it.  But dead material can’t be put in a flask and asked how it grows at nineteen degrees.

This can.  You can raise an organism and its own great-great-however-many-times-grandparent in the same incubator, in the same medium, under the same light, and simply look at the difference.

There are a few other versions of the same trick and they’re worth knowing about.  Someone buried twenty bottles of seeds in Michigan in 1879 to settle an argument about how long weed seeds stay viable, and the experiment is still running, dug up at intervals that have stretched out to about twenty years as it became clear the seeds were in no hurry.  Lake sediments hold egg banks of water fleas that get revived to study responses to a century of pollution.  Seed vaults are the deliberate version of the same principle, built because we noticed that the accidental version works.

What was done recently with a chain-forming coastal diatom is the marine case.  Strains were resurrected from Baltic sediment spanning about sixty years and grown together.  The modern isolates turn out to have a thermal optimum roughly one degree higher than the old ones, and they do correspondingly worse in cold water.  There’s also a shift in the proportions of the cell wall, with the modern cells built slightly differently.

One degree sounds like very little.  It isn’t nothing, and it also isn’t as big as it sounds, and both halves of that are worth holding onto.  Performance curves against temperature are not symmetrical - they rise gently and then fall off a cliff - so a shift of one degree in the peak can matter a great deal for what happens at the top end, and hardly at all in the middle.  And a population that has moved its optimum up has, in this case, given something back at the cold end.  That’s the shape of a trade-off rather than a straightforward improvement, and it’s the part I’d want in the headline.

Now the parts that make me cautious.

The first is that this is a population moving, and nothing here says what moved.  There is a difference between new variants appearing and existing variants being sorted - between a population that has changed and a population where the same range of types was always present and the warm-preferring ones are now doing better.  Both would look like this.  They have different implications for what happens next, because sorting can only continue while there’s still something left to sort.

The second is about the shape of the cells.  A change in wall proportions is interpreted as something to do with nutrient uptake, and that’s a reasonable guess from the geometry.  But it’s a guess from geometry.  Nothing was measured taking anything up.  Shape suggests function, constantly and persuasively, and it’s wrong often enough that I’d rather see the flux than the inference.

The third is the one I keep coming back to, because it applies to the whole method rather than this one study.  The archive is not neutral.  What you germinate is not a random sample of what was alive in 1960 - it’s a sample of what could make a resting stage, survive decades in mud, and then still be capable of waking up in a dish.  Every one of those is a filter, and there’s no obvious reason to assume the filter is the same for a cell from 1960 as for a cell from 2015, given that the older ones have had longer to fail.  If the ability to sit in sediment is itself something that varies, then a comparison across depths is partly a comparison of survivors with survivors, and partly a comparison of decades.  Nobody can fully separate these, which is fine.  It just needs saying out loud more often than it is.

None of that spoils it.  A living time series is still one of the most remarkable things available in this kind of work, and the plain fact that a population’s temperature preference has shifted measurably inside a human lifetime is worth sitting with.  Adaptation is usually something we reconstruct.  Here it’s something you can put in an incubator.

I’d just resist the pull towards reading it as a story about improvement.  What the sediment tells you is that something moved.  The much harder question - what moved, and whether it can keep moving - is still sitting in the mud.