Keeping at the opposite scale and speed from a fungus — a tree that cannot remove decay inside its own trunk and does not heal the wound, but walls it off in wood it already had and in wood it has not grown yet, so the rot stays in old wood while new wood keeps forming outside it, for centuries
R945 closed with a fungus that spent one compartment to save a network, sealing a pore in moments. This page takes the turn it pointed to, upward in scale and outward in time. A tree does the same kind of keeping — confine the damage, accept the loss, grow on — but it does it slowly, in a trunk, and without ever undoing the injury. The standard account is CODIT: Compartmentalization Of Decay In Trees, developed by the United States Forest Service pathologist Alex Shigo and Harold Marx in the 1970s from dissection of very large numbers of wounded trees (Shigo & Marx, 1977; Shigo, 1979). The D is sometimes also read as Damage or Dysfunction, because decay is only one of the things being confined.
No healing, in the animal sense. Healing, used strictly, means restoring injured tissue in the place it occupied — repair, replace, put back. A tree, on Shigo’s account, cannot do that with wood. Wood present at the time of wounding that becomes infected stays infected; it is not cleaned out and it is not remade in place. What a tree does instead is generate: its cambium lays down new wood in new positions, outside and beyond the injured column, season after season. A wound may eventually be covered over by that new growth, and the surface may look closed, but closure is not restoration. The old, damaged wood is still in there, behind the cover. That distinction — sealed, not healed — is the whole frame.
Three walls in wood that already exists. When a trunk is wounded and micro-organisms enter, the tree alters the wood around the injury chemically and physically, and the model names the resulting boundaries by the direction they resist. Wall 1 resists spread up and down the stem: the conducting vessels above and below the wound are plugged, by tyloses, deposits, anti-fungal substances, and in conifers by closure of the pits between cells. It is the weakest wall; decay moves most readily along the grain, the way the tree itself moves water. Wall 2 resists spread inward, toward the centre: it is associated with the dense, lignin-rich last cells of a growth ring, the latewood, and it is stronger than Wall 1, though not continuous where ray cells cross it. Wall 3 resists spread sideways, around the trunk: it is formed by the ray cells that run radially through the wood like the spokes of a wheel, chemically altered after wounding, discontinuous and maze-like rather than solid. Of the three walls present at the time of wounding, Wall 3 is the strongest. Together the altered wood around an injury is often called the reaction zone: a boundary made inside tissue that was already there.
One wall in wood that does not exist yet. Wall 4, the barrier zone, is different in kind. After wounding, the cambium begins producing a band of chemically defended, anatomically distinct wood that separates everything present at the time of injury from everything the tree will make afterwards. It is the strongest of the four walls against the spread of decay, and often the only one that holds completely. It is also not callus: the callus and woundwood that roll in from a wound’s edges and may eventually cover it are a surface closure, a separate thing from this internal boundary. And the barrier zone carries a cost the other walls do not. A plane that separates old wood from new so cleanly is also a plane along which the trunk can later separate — a ring shake — so the strongest wall against decay is, structurally, a weak seam. Keeping the rot out and keeping the trunk in one piece are not the same achievement.
Success can look like a hollow tree. If Walls 1, 2 and 3 all fail and only the barrier zone holds, every cell present at the time of wounding may decay away while the tree goes on living and thickening outside Wall 4. The result is the familiar living hollow tree: a tube of sound new wood around an empty centre. In the logic of compartmentalization that is not failure; it is the bargain completed, the whole interior spent exactly as a fungal compartment is spent, only over decades instead of moments. Two cautions belong with that picture. A hollow is not, by itself, proof that a particular tree is safe to stand beside — how much sound wood remains, where, and under what load is a separate structural question. And the walls are not absolutes in any tree: they are resistances, not guarantees. Given enough time, a vigorous enough pathogen, a large enough wound, or a tree of low vigour, boundaries can be breached, even the barrier zone in time. Compartmentalizing ability also varies markedly between species, and with the tree’s condition and the wound’s size, position and season. CODIT names a pattern of response, not a promise made to every trunk.
A walk around one wound. Five stages: the wound itself, bark and cambium broken and wood exposed; Wall 1, vessels plugged above and below, the weakest boundary, resisting the vertical run; Walls 2 and 3, latewood resisting inward and rays resisting sideways, the strongest walls the old wood can offer; Wall 4, the barrier zone laid down by the cambium in wood formed only after the injury, strongest against decay and a possible plane of splitting; and the long outcome, decay confined to the wood that existed when the wound was made while new wood accumulates outside it — up to, and including, a living hollow tree. Schematic, compressed from the published model.
Walk it in order.
Schematic only: the stages compress a model drawn from many species, wounds and dissections. Labelled cautions: the four walls are a model of boundaries in wood, not four literal bricks built on command — Walls 1 to 3 are largely pre-existing anatomy chemically altered after injury, and only Wall 4 is new tissue. Do not equate the barrier zone with the callus that closes a wound’s surface. Do not call Wall 4 simply the strongest wall without its other half: strongest against decay, structurally a weak plane. Do not read a hollow trunk as either certain failure or certain safety. And do not turn the ranking 1 < 2 < 3 < 4 into a law for every species and wound; it is the model’s usual ordering, and real trees vary.