Trophic mode

One of the first pieces of biology anyone is taught is a division.  There are things that make their own food out of light and air, and there are things that eat other things.  Producers and consumers.  Green at the bottom, arrows pointing up.

It’s a good enough scaffold for a nine-year-old, and it survives more or less intact into adulthood for most people, and it starts developing cracks almost immediately if you press on it.  There are plants that catch insects.  There are plants that have given up photosynthesis entirely and now live off the roots of other plants, which is a startling thing to find in something with leaves.  Corals are usually described as animals but get most of their energy from photosynthetic tenants.  Lichens aren’t one organism at all.  And we can’t make our own vitamin B12, so we outsource that to bacteria, one way or another, and nobody thinks of us as being in a partnership.

The line is a filing system.  It was built from big, obvious organisms where the distinction is genuinely clear, and then applied downward to organisms that were never consulted.

Which brings me to a result from this year that I’ve been turning over.

A group built a detailed metabolic model of a common and widely distributed marine diatom - not a sketch, but a full reconstruction of a couple of thousand genes’ worth of reactions - and then matched it against a large set of expression data collected from actual seawater around the world.  The idea is to ask, from the pattern of which genes are switched on, which of the model’s possible metabolic states the cells appear to be in.

The answer that comes out is that purely photosynthetic living accounts for only about a fifth of the sunlit-ocean samples.  Most of the rest looks mixed - taking in some carbon as light and carbon dioxide, and some as ready-made organic molecules from the water around it.  A small remainder looks like it isn’t photosynthesising for a living at all.

If that holds, it’s a fairly serious rearrangement.  This is an organism that goes into textbooks as a primary producer, and it appears to be spending most of its time doing something less tidy.

Worth being precise about what “eating” means here, because the mental image is misleading.  There are microbes that engulf other cells whole, and that would be one kind of answer.  This is mostly the other kind: dissolved organic molecules taken up straight across the membrane, from a soup of material leaking out of everything else in the water.  Less predation, more absorption.  Sitting in a very dilute broth and drinking.

Now the part that keeps me from getting carried away.

Nobody has watched them do it.

What was actually measured is expression patterns.  What was inferred is metabolic state, by way of a model that says which states are consistent with which patterns.  That’s a legitimate and clever way to ask a question at ocean scale, where you can’t follow individual cells around.  But it is inference twice over, and there’s a straightforward experiment that would settle it: give cells organic carbon with a heavier isotope of carbon in it, and then look for the heavy label turning up inside them.  Either it does or it doesn’t.  That measurement isn’t in the paper.

The paper also floats a possibility that the signal might involve bacterial partners rather than the diatom itself, which is honest of it, and which currently rests on the two things being found in the same places.

There’s a second thing that I think matters more for people who work with these organisms, and it’s a bit uncomfortable.

Almost every stress experiment on this kind of cell is built on the assumption that light and carbon dioxide are the main inputs to its carbon budget.  That assumption is in the medium you make, the treatments you choose, the variables you decide are worth crossing, and the way you interpret anything that changes.  If a substantial part of the carbon is actually coming in as dissolved organic material, then a cell grown in clean defined medium under a lamp is not a simplified version of the wild organism.  It’s a partial one, and possibly partial in a direction that matters for exactly the responses you’re trying to measure.

I don’t want to overstate that.  The result is a prediction, not a demonstration, and plenty of laboratory work on this species will turn out to be perfectly sound.  But “we grew it under controlled conditions” is a claim about control, and control depends on knowing what the inputs are.

What I like about this one is that it doesn’t resolve into a clean replacement fact.  The old picture wasn’t wrong, exactly.  It was a category we built out of trees and cows and then handed to something microscopic that has been quietly ignoring it.  Most of the ocean’s biology is being done by organisms small enough that the distinctions we find intuitive - plant or animal, making or taking - are less like laws and more like a habit of ours.

Which is a reasonable thing to notice about a lot of the boundaries we teach, if you can stand to look at them for long.