Mass balance

For as long as anyone has been able to measure both sides of it, the ocean’s nitrogen budget has failed to balance.

You can estimate how much nitrogen gets pulled out of the atmosphere each year and converted into a form that living things can actually use.  You can also estimate how much is being consumed, buried, and handed back to the air by the microbes that run the reaction in reverse.  The two numbers have never agreed.  Demand has consistently come out ahead of the supply we can account for, by a margin too wide to shrug at, and the usual responses are the ones you’d expect.  The loss terms are probably overestimated.  The gain terms are probably underestimated.  Everything here is difficult to measure and we should be patient.

Or somebody has been fixing nitrogen all along and we hadn’t found them.

It’s worth saying why this is hard in the first place, because the chemistry is genuinely strange.  Nitrogen is not scarce.  Roughly four fifths of the air is nitrogen gas, which means every breath you’ve ever taken has been mostly nitrogen, and none of it has been the slightest use to you.  The two atoms are held together by a triple bond, one of the strongest in ordinary chemistry, and prising them apart takes either a great deal of energy or a very specialised enzyme.  Almost nothing alive has that enzyme.  Plants don’t.  Animals don’t.  It sits in a scattering of bacteria and archaea, and everything else on the planet is downstream of them.

We eventually worked out how to do it ourselves, at high temperature and pressure, in the early twentieth century.  The commonly quoted estimate is that about half the nitrogen atoms in the average person today have passed through an industrial reactor at some point.  There is a decent argument that this is the most consequential piece of chemistry anyone has ever done, and most people have never heard of it.

Before that we did it by borrowing.  Clover in the rotation, peas and beans in the garden, the whole business of resting a field - all of it is a way of renting bacteria.  Certain plants build small nodules on their roots and house a bacterial partner inside them, feed it sugar, and take fixed nitrogen in exchange.  Farmers were running this arrangement for centuries before anyone knew there was anything living in the nodules.

Out in open water, the assumption has been that nitrogen fixation is essentially a cyanobacterial business.  Photosynthetic bacteria, free-living or in filaments or tucked into other cells, doing the same job in a different setting.  That’s where people have looked, and that’s where the accepted budget comes from.

This year a group described a marine diatom carrying a nitrogen-fixing partner inside it, and the partner is not a cyanobacterium.  It belongs to the same broad group as the bacteria in those root nodules.  The oldest agricultural trick we have, running in the middle of the ocean, inside a single cell.

That alone would be a good result.  What makes it more than a curiosity is the state the bacterium is in.

Its genome has shrunk.  Badly.  It has lost a large fraction of the genes a free-living relative would carry, and what remains is skewed toward the things the host can’t do for it.  This is a well-understood process, and it’s less mysterious than it sounds.  Once a partner reliably supplies you with something, the gene you were using to make that thing yourself becomes a cost with no benefit.  Mutations that break it stop being punished.  Over enough generations the genome erodes down to whatever is still load-bearing, and the cell becomes progressively less able to leave.

Which is exactly what mitochondria look like.  And chloroplasts.  Both were free-living bacteria a very long time ago, both went inside another cell, both shed almost everything, and both are now so thoroughly integrated that we teach them as parts rather than as guests.

So the question this raises is not really “is there another symbiosis” but “how far along is it”.  And the honest answer is that we don’t have a clean threshold.  People use criteria - is it inherited reliably, does it divide in step with the host, does the host make proteins and send them in - and each of those is a matter of degree.  The word organelle implies a line that the biology doesn’t actually draw.  There’s a continuum from tenant to lodger to relative to organ, and organisms move along it slowly and untidily, and we happen to be watching this one somewhere in the middle.

That’s the part I find hard to put down.  Not the discovery itself, but the fact that a merger which we normally discuss in the past tense, as something that happened twice in the deep history of everything, is apparently still going on in ordinary seawater.

I’d be careful about how much of the budget gap this closes.  One arrangement, well described, tells you that the arrangement exists.  It doesn’t tell you how much of the ocean it accounts for, and the gap in question is enormous.  Working out whether this is a widespread contributor or a locally interesting oddity is a separate and much larger job, and until it’s done, “we found the missing nitrogen” is a headline rather than a sum.

But there’s a smaller lesson here that I think holds regardless of how the numbers land.

The nitrogen wasn’t missing.  It was being fixed the whole time, by something nobody thought to open up and look inside, because the search was built around a different kind of organism.  The failure wasn’t in the instruments or the effort.  It was in the category we were searching with.  An absence in your results is sometimes a real absence in the world, and sometimes it’s just the shape of the net.