Field notes on things that run themselves
The Forest Held Up by Its Own Collapse
A four-hundred-year-old forest can look, from a low-flying plane, almost exactly the way it looked in a photograph taken a century earlier — the same rough canopy height, the same mix of species, the same broken light scattered across the floor. Ask a forester which trees are doing the holding up, though, and the honest answer is usually: mostly none of the ones in that older photo. They’re on the ground, rotting, or gone entirely, wood returned to soil years or decades ago. The forest’s shape has outlived nearly every tree that ever wore it.
Here is what actually happens, over and over, all across that forest, without a schedule: a big canopy tree dies — old age and rot, more often a windstorm or a beetle outbreak that gets there first — and it doesn’t nudge the forest toward some new equilibrium. It tears one hole straight through the roof. Sunlight that hasn’t reached the forest floor in decades pours into that single gap, and a small, temporary ecosystem switches on inside it: saplings that have spent years, sometimes decades, waiting in the shade below suddenly have enough light to bolt upward and compete for the opening. The British ecologist A. S. Watt, in a 1947 paper that became foundational to how forest ecologists think, described this as one stage — the gap phase — in a longer cycle every patch of canopy runs through: gap, then a scramble of fast young growth, then a settled mature canopy, then eventual decline, and someday another gap.
Zoom out, and the real trick appears. At any given moment, an old forest is a mosaic of small patches, each sitting in a different stage of that private cycle — this quarter-acre freshly opened, that one closed over in young growth, the next mature and quiet, another with its old canopy visibly failing. Ecologists Frederick Bormann and Gene Likens, working from decades of plot data at New Hampshire’s Hubbard Brook Experimental Forest, gave the pattern that describes the whole landscape a name: a shifting-mosaic steady state (1979). Every individual patch is always somewhere in its own cycle of collapse and repair, yet averaged across enough patches, the forest’s overall structure — its total biomass, its canopy height, its rough balance of light and shade — holds remarkably steady from one decade to the next. It’s the same trick this series has already found running a stop-and-go traffic jam (No. 9) and a spiral galaxy’s arms (No. 15): a pattern stable in aggregate, built entirely from parts that are each, individually, only ever passing through.
The repair itself moves fast, by forest standards. A gap closes mainly by two routes at once: the trees bordering it simply reach sideways into the opening — foresters call it crown expansion — while whichever sapling had already spent years banked in the shade below gets its chance and grows hard toward the light. One study of an old beech forest found roughly seventy percent of small gaps sealed within a few years this way, against barely a tenth of much larger ones, which stay open long enough for genuinely different, faster-growing pioneer trees to get a foothold before the shade-tolerant natives retake the ground.
How anyone knows any of this from a real forest, and not just a model, owes a great deal to one patch of English woodland. In 1944 the Forestry Commission set aside Lady Park Wood, forty-five hectares straddling the Wye Valley on the England–Wales border, for no purpose but to watch it develop with no management at all; a year later Eustace Jones marked out the first sample plots, and more than eighty years on, researchers led by George Peterken have tracked the birth, growth, and death of upward of twenty thousand individual trees and shrubs there. The record confirms the steady churn Watt described — but it also complicates the tidier version of the theory. A severe 1976 drought killed close to a fifth of the wood’s beech population over the following sixteen years and durably shifted the canopy toward the more drought-tolerant sessile oak: a real, synchronized step-change, bigger and messier than a scatter of independent single-tree gaps. The honest record, not the clean theory, turned out to be the better witness.
For how long can a forest keep doing this? Białowieża Forest, straddling the Poland–Belarus border, has had continuous canopy cover for close to twelve thousand years — since not long after the last glaciation retreated from that stretch of Europe, by the evidence of pollen buried in its own soil. And yet the oldest individual oaks anyone has found there are reckoned at only four or five centuries. The forest has been running this same small, unscheduled churn of collapse and regrowth for something like thirty separate lifetimes of its own oldest trees, stacked end to end, and none of them got to see more than one.
A flame holds its shape while its fuel passes through and is gone within seconds (No. 1). A skeleton holds its shape while its own calcium is dissolved and rebuilt, cell by cell, over years (No. 12). A forest holds its shape the same way, just slower and messier — patch by patch, tree by tree, with nothing about the timing coordinated by anyone. It isn’t standing despite the collapses. It’s standing on them.
One loop I’m watching
Next: a kidney, holding a steep chemical gradient inside itself that never gets to rest — two fluids running in opposite directions through the same tight space, pumping against each other without pause, for no other reason than to concentrate a fluid the body is about to let go of anyway.
Tip: the ← and → arrow keys move between issues.
Get the next issue the moment it’s live — subscribe by email or follow by RSS.
New to The Standing Wave? Start here →