How deep does life go?

An expedition off Japan drilled through coal-bearing sediment to about 2.5 kilometres below the sea floor and counted cells in cores from 1.3 kilometres to the bottom of the hole. Cores from the upper few hundred metres at the same site held more than ten million per cubic centimetre. Below 1.5 kilometres, in sediment at 40 to 60 degrees Celsius, the count runs from fewer than ten per cubic centimetre to about ten thousand, far below what the global trend with depth predicts, and the peaks are in the lignite coal beds.

Those coal-bed communities resemble the organotrophs of a forest soil rather than the subseafloor above them, which is a clue about origin: the coal was laid down in settings such as coastal wetlands and deltaic lagoons and then buried, and the paper reads the resemblance as a sign that such sediment keeps members of its original community tens of millions of years after burial.

A 2018 synthesis compiled cell counts from continental subsurface sites around the world and put the population there at 2 to 6 times 10 to the 29 cells, with bacteria more abundant than archaea. Folded into the existing literature, that gives a global prokaryotic biomass of 23 to 31 billion tonnes of carbon.

It is 4 to 10 times below earlier estimates, and it is not the only figure: a census published the same year put bacteria at about 70 billion tonnes of carbon and archaea at about 7, most of both deep underground, each uncertain by a factor of ten or more. The subsurface did not shrink. The estimates have fallen as more places were counted, and the 2018 synthesis reports that neither the organic carbon in continental rock nor the cells in its groundwater predict how many cells the rock holds.

Temperature rises with depth, and life stops somewhere. The highest temperature at which cell proliferation has been demonstrated is 122 degrees Celsius, by a methanogen held at 20 megapascals: at 0.4 megapascals the same organism stops at 116. What depth that corresponds to varies enormously with the local geothermal gradient.

A borehole in the Nankai Trough reached 120 degrees at 1.2 kilometres and found cells almost all the way down, though very few. Above 45 degrees ordinary cells become a hundred times scarcer, spores outnumber them more than six thousand to one, and the inhabited zones alternate with gaps, one of them 192 metres thick, where no microbes could be detected. Methane is made and consumed biologically until the sediment reaches 80 to 85 degrees, and between 100 and 120 degrees the isotopes and a rise in cell numbers record hyperthermophiles degrading acetate.

Whether the true limit is temperature, available energy, or the rate at which damaged molecules can be repaired is not resolved. In stable, nutrient-poor settings, subsurface cells subsist on energy fluxes a thousand times below what laboratory cultures are thought to need simply to maintain themselves. Yet in the hottest Nankai sediment, incubations found a small population metabolising at rates close to those of surface sediment and laboratory cultures, apparently spending most of its energy on repairing heat damage.

In those settings, subsurface cells catabolise ten thousand to a million times more slowly than model organisms in a rich culture, and turn over their biomass on timescales of centuries to millennia. At those rates the distinction between a very slow metabolism and a long dormancy is hard to draw with the measurements available.

This is not a semantic problem. It determines whether the subsurface is a working ecosystem or an archive of things that were buried alive.

The sampling is by borehole, and boreholes are drilled where someone wanted to drill. Whole environments are barely sampled, the deep aquifers and the ocean crust among them.

Any number for the size of the deep biosphere is an extrapolation from an unrepresentative sample, and the 2018 synthesis opens by saying that the controls on how many microbes live in the continental subsurface are poorly understood.