Soil formation

How does bare rock become soil deep enough to support a forest?

A forest floor is soft, dark, full of water and roots and living things. A granite outcrop is none of those, and yet every forest soil began as rock. What strikes me is that soil is not simply rock in smaller pieces: crush granite to powder and you have sand, not soil. So something must be added. What, and from where?

Hold that question. Its answer is surprising, and it is why the whole process takes as long as it does.

Begin with the obvious part. Exposed rock cracks: water gets into fissures and freezes, expanding; the surface heats and cools daily until its outer skin flakes; grit abrades it. None of this changes what the rock is. What does it change? Surface area — no small thing, since every later step happens on surfaces.

Now chemistry can begin. Rain carries dissolved carbon dioxide, which makes it faintly acidic, and it attacks the minerals themselves. Feldspars, the bulk of granite, break down into clay minerals; iron-bearing minerals rust. Why does clay matter so much? Because clay particles are minute, sticky, and electrically charged at their surfaces, so they hold water against gravity and nutrient ions against being washed away. Powdered rock cannot do that. Clay is the first genuinely new substance in the story.

Life arrives earlier than most people picture. Lichens — fungi living with algae or cyanobacteria — colonise bare stone within years and secrete acids that etch grains loose; mosses catch dust and hold moisture. What do they leave when they die? Organic matter: the second new substance, and the one that turns pale grit into dark, crumbly, water-holding soil.

Now return to the question we held back. Rock supplies phosphorus, potassium, calcium, magnesium. But plants need a great deal of nitrogen, and ordinary rock has almost none. So where does a forest's nitrogen come from? Not from below. It comes from the air, and only certain microbes can convert it — some free-living, some partnered with pioneer plants such as alders. That is the hidden bottleneck: young soils are typically nitrogen-poor, and the pace of enrichment sets the pace at which real vegetation can establish.

Once plants take hold the process turns on itself. Roots pry cracks wider; litter feeds fungi and invertebrates that shred and rework it; burrowing animals mix organic matter downward. Percolating water carries fine clay and dissolved iron out of the upper layer and deposits it lower, so the soil separates into horizons — dark organic top, leached layer, enriched layer, broken parent rock. Distinct horizons are the mark of a soil that has been running a long time.

How long? That is the striking part. Lichens within years, a thin organic mat within decades, appreciable dark topsoil within centuries — accumulation is commonly estimated at something like a centimetre per century, though that figure varies by an order of magnitude either way. A deep, fully differentiated profile takes millennia to tens of millennia. And these numbers travel badly: warm, wet, vegetated ground weathers far faster than cold or arid ground, and basalt far faster than quartz-rich granite. Where forest appears on new land within a century or two — the classic case is ground uncovered by retreating glaciers at Glacier Bay, Alaska — the head start is that glacial till is already ground-up rock, not solid bedrock.

The most useful correction is that soil is not a substance but a system with five controls: parent material, climate, organisms, topography and time. Change one and the same rock yields a different soil — basalt gives deep red clay in the wet tropics and a thin stony soil on a cold slope.

The second correction concerns sequence. It is tempting to read the stages as a relay: breakdown, then chemistry, then life. They overlap almost entirely. Lichens work on bare rock while frost is still cracking it, and chemical weathering continues at depth long after a forest stands. What changes with time is not which process runs but which one is limiting.

• Why the same parent rock produces such different soils under rainforest, prairie and boreal conifers.
• How chemical weathering of silicate rock draws carbon dioxide from the air over geological time.
• What soil chronosequences on lava flows and glacial forelands reveal that experiments cannot.

Every term the collection defines is gathered in the glossary.