Soil DNA survey

There is a city under your field

One teaspoon of this soil holds around a billion bacteria, several thousand kinds of them, plus fungi, single-celled hunters, and microscopic animals that eat them. In one weedy corner of a working farm, five kinds of plant grow tangled together — a pasture grass, and four the farmer could reasonably call weeds. Nobody knows whether the life under a weed differs from the life under the grass beside it. So one July we sampled the ground beneath each of the five and read out the DNA. This is the census that came back.

The size of it

From cores of soil, each about a teaspoon

distinguishable kinds of bacteria and archaea

They came from kinds of plant in marker genes, at DNA reads a core, from one patch of ground, sampled once.

A shrink-wrapped blue rack of numbered sample tubes resting on a
                  printed handling protocol, whose list of sample names has been
                  ringed in blue pen and annotated by hand.
The samples in numbered tubes, on the sheet that lists them: tubes, one per core.

Before anything else: every number here counts DNA, not bodies and not activity. Dormant microbes leave DNA behind; so do dead ones, until it decays — and in soil the dead share is large.

What the DNA showed

The plant above picks the neighbours below

Every patch holds about the same amount of life, whichever plant is growing over it (p = ). Which organisms are there is a different answer: the cores group by the plant above them, by more than chance produces at this sample size (p = ). Plants shaping the community beneath them is well documented, in soils far beyond this one — so this is a pattern the field shares with pasture everywhere, seen here in detail.

Two cores under the same plant are more alike

Each dot is one pair of cores. The gap is real on average, and there is a great deal of overlap. The bar is the average, the whisker its 95% interval.

It is a food web, not a soup

A second marker gene picks up the organisms with more complex cells: fungi, protists that hunt and eat bacteria, and the microscopic animals that eat the protists. Something is eating something else down there, which is more than a list of residents sharing space.

Who is down there

The complex-celled share of this soil, across cores. "Other" is everything the reference database could not place, which in soil is never small.

And one number this study cannot explain

Turning ammonia into nitrate is one of soil's essential jobs, and here it is dominated by archaea — an ancient branch of life, separate from bacteria — to a degree that in farmed soil usually points at one thing: very little nitrogen going in. The study tested that reading against its own data, and it did not hold. The number and the test that undoes it are both on the research page.

What this does not establish

Thirteen cores from one field, sampled once, cannot carry much. What to know before quoting anything above.

A herd of cattle walking away up a gravel farm lane in fog, wire fencing
                  and pasture on the left, a dense hedgerow on the right.
Herding the cows up the lane to the next paddock, early and foggy. This is the farming everyone asks about, and the part this study cannot speak to — the cores came from a weedy patch, not from the grazed sward in this picture.

What this adds up to is a baseline: a close description of the soil life under five plants growing side by side, in one patch of ground on one day. Of the two real patterns in it, the plant shaping the community beneath it is something soils do everywhere — but it is worth knowing it happens between a weed and the grass next to it, over a few metres, on the same soil. The archaeal number is still unexplained. Tying either of them to the way this ground is farmed would take a different study, and one this design cannot become.

Where to go from here