How old growth survives

The earlier primers defined old growth as a structurally complex system that develops only when nothing interrupts a forest for several centuries. That definition is correct but incomplete. In every old-growth forest on Earth, something has interrupted parts of the system — wind, fire, insect outbreak, landslide, eruption — repeatedly, across the centuries the forest has stood. The forest is old growth not because nothing happened, but because enough survived each event to carry the forest forward.

This primer is about that survival. What gets through a forest’s worst day, by what mechanism, and what those survivors do over the decade after the event. The science is global; the canonical natural experiment that taught the field most of what it now knows is one mountain in the Cascades.

Disturbance as a window into how forests work

Catastrophic disturbance is one of the few honest measurements ecology has for forest dynamics. The event provides a measurable beginning state. The recovery unfolds at a pace researchers can document. And what survives, where, and why is a direct readout of which traits matter under which kinds of stress. Storm Gudrun in southern Sweden in 2005 (around 75 million m³ of windthrow), Hurricane Hugo in the Caribbean and US southeast in 1989, Australia’s 2019–20 Black Summer fires, and the May 18, 1980 lateral blast and pyroclastic flows of Mount St. Helens are among the most-cited modern cases, and they have been studied with the same scientific apparatus^[1].

Mount St. Helens is the most intensively documented of them. The USDA Forest Service Pacific Northwest Research Station began tracking biological response inside the blast zone within weeks of the eruption, and has continued; the long-term ecological research published its 35-year retrospective in 2018^[2][3]. The reason this primer is anchored there is not parochial. It is that the science itself is anchored there.

What survives, by mechanism

Three mechanisms account for almost all surviving stems in a catastrophic disturbance. They are global; the species change, the mechanism doesn’t.

1. Topographic shielding. Where a ridge, a deep valley, or a coastal lee separates a stand from the source of the disturbance, mature trees survive as living individuals or whole patches. After Mount St. Helens, the Bear Meadow ridge on the eastern perimeter held mature Douglas-fir survivors that watched the eruption happen and recovered. After Storm Gudrun, sheltered north-facing slopes in Götaland held intact stands while exposed slopes were scoured. After Hurricane Hugo, leeward valleys in Puerto Rico’s Luquillo mountains preserved canopy that windward ridges lost. The mechanism is the same; the geography is local.

2. Trait-driven survivors at the disturbance fringe. Where traits buy survival under stress, the species carrying those traits do better than their neighbours. The traits that matter for a fire and heat-pulse event differ from those that matter for blowdown, but the principle is consistent. In the Pacific Northwest matrix, three genus groups carry most of the trait-driven survival load — and each has parallels worldwide:

• Firs (Abies). Pacific silver fir (A. amabilis) and noble fir (A. procera) survive at higher elevations through a combination of cold-tolerance and snowpack protection (see refugia, below). Subalpine fir (A. lasiocarpa) is shaped to cold-disturbance regimes. European silver fir (A. alba) and the Asian Abies species play parallel roles in their respective ranges.
• Hemlocks (Tsuga). Mountain hemlock (T. mertensiana) is a high-elevation survivor across the Pacific Northwest; western hemlock (T. heterophylla), the lower-elevation climax, is shade-tolerant and successional but has thin bark and tall, narrow stems that fare poorly in lateral blast and high-intensity fire. The contrast within the same genus is itself instructive: trait, not lineage, predicts who gets through.
• Cedars (Thuja, Calocedrus, Cedrus). Western redcedar (Thuja plicata) carries thick fibrous bark that buffers the heat pulse, and is uniquely long-lived (individuals over 1,000 years are common, and the species reaches roughly 1,500 years). Incense-cedar (Calocedrus decurrens) and the true Mediterranean cedars (Cedrus libani, C. atlantica) carry the same thick-bark survival pattern in fire-prone systems on three continents.

Underneath the species list is a mechanical story. Bark thickness insulates the cambium from heat. Deep rooting holds against lateral force. Tall, narrow stems bend or snap in blowdown; short, deep, branchy stems bend and recover. Old, large individuals carry more of all three traits than young ones, and that is one reason — not the only one — old growth is more resilient than secondary forest to the disturbance regimes of its own region.

3. Refugia. Biological pockets that escape the disturbance entirely. In the Pacific Northwest, May 18, 1980 came late in a heavy snow year; high-elevation stands buried under two to three metres of snowpack were physically shielded from the lateral surge and the heat pulse, and Pacific silver fir and mountain hemlock at those elevations survived in place^[2]. In fire-disturbed landscapes, the equivalent refugia are riparian corridors and damp lee slopes. In hurricane-disturbed landscapes, the equivalent is leeward coastal valleys. In every system, the refugium is doing the same job: providing a small geography in which the disturbance does not reach.

Goat Marsh — a stand that has survived a continuum

One Pacific-Northwest example illustrates how all three mechanisms can compound across human time scales. The Goat Marsh Research Natural Area, on the south flank of Mount St. Helens in the Gifford Pinchot National Forest, was established in 1974 — six years before the 1980 eruption — to protect what remained of a 350-year-old noble-fir / Douglas-fir stand at low elevation, alongside a mosaic of marshlands and lodgepole-pine flats whose hydrology was itself shaped by earlier eruptions of Mount St. Helens between roughly 1550 and 1700 CE^[4].

That stand survived 1980 by topographic geography: the lateral blast went north, and the south flank received pyroclastic and lahar impact but not the direct surge. It carries some of the largest noble-fir and Douglas-fir specimens on Earth — the “Goat Marsh Giant” at 272 ft (83 m) and unnamed Douglas-firs measured at over 300 ft (91 m) — and stand biomass at Goat Marsh is among the highest in the temperate world outside the redwood belt of northern California^[4]. It is a useful illustration of the layered survival logic. The same stand has now survived a continuum of disturbance events, each by a different mechanism, across centuries it has been measurable to no one.

What didn’t, and why

The honest counter-list matters. Western hemlock at lower elevations in the Mount St. Helens blast zone was almost entirely killed — the species’ thin bark and tall, narrow profile do not survive lateral blast, and its lower-elevation distribution did not benefit from the snowpack that protected its mountain-hemlock relatives upslope. Similar species-by-species honesty applies elsewhere: thin-barked hardwoods in fire-disturbed Mediterranean systems, shallow-rooted spruce in storm-disturbed Sweden, narrow-canopied pioneer species in hurricane-disturbed Caribbean systems. The reason this primer lists what didn’t survive alongside what did is that retention forestry, which is built on this knowledge, is only honest if the trait-survival story is honest.

Biological legacies

Jerry Franklin, walking inside the Mount St. Helens blast zone in the weeks after the eruption, named the survivors he found biological legacies^[5]. The term covers more than living trees: surviving root systems with intact mycorrhizal networks, surviving soil with its organic-matter horizon, surviving seed banks, surviving downed wood, surviving small-mammal and amphibian populations sheltered in burrows or refugia. Each is a head start for what regrows.

Franklin and his colleagues then watched, over twenty years, what those legacies actually did. Recovery rates inside the blast zone correlated with biological-legacy density. Patches that retained living overstory or root systems regrew differently and faster than patches that had been stripped to mineral soil. Mycorrhizal continuity — the surviving fungal partners ready to re-form symbioses with returning seedlings (see the mycorrhizal network) — was a measurable input to recovery rate, not a metaphor for it.

The general statement, now consensus across the disturbance-ecology literature: the rate at which forest function returns scales with what survives, not with what gets planted.

From observation to silvicultural prescription

Once forest science had the biological-legacy concept, working forests had a question: can we leave the same things standing during deliberate harvest that catastrophic disturbance leaves standing by accident? The answer, developed across the 1990s and codified globally since, is variable-retention harvest. Instead of clearfelling, leave a deliberate fraction of the stand — typically 10–30%, in dispersed singletons or aggregated patches — standing across the harvest unit, in a pattern that retains the structural features biological legacies provide^[6][7].

The science is now global. British Columbia codified variable retention on coastal Crown forest in the late 1990s. Sweden’s national forestry authority published retention guidance after Storm Gudrun in 2005. Tasmania uses variable retention in buffer zones around the Tasmanian Wilderness World Heritage Area. The International Tropical Timber Organization publishes retention guidelines for tropical-forest concessions^[7]. In the Pacific Northwest, the 1994 Northwest Forest Plan codified retention on 9.7 million hectares of federal land in the spotted-owl range — a legal framework that exists in the shape it does because of what was learned at Mount St. Helens^[8].

What this means for the surviving old growth

The retention story changes what “protecting old growth” means in a working landscape. It is not only about the few stands that have escaped human harvest entirely. It is also about the much larger area of mature and recently-mature forest that is being harvested, and could be harvested in a way that preserves biological legacies — versus a way that does not. The first protects future old growth. The second forecloses it.

That distinction is the active rulemaking question on US federal land in 2026. Executive Order 14072 in 2022 commissioned the first national inventory of mature and old-growth forests on federal lands; the inventory was completed in 2024; the follow-up rulemaking that would govern how those forests are managed has been repeatedly drafted and remains incomplete^[8]. The same shape of question is open in Canada, in the EU under the Deforestation Regulation, and in the UN Forest Declaration framework for the tropics.

What you can do

If you live in the United States, the highest-leverage action this month is to submit a comment to the USDA Forest Service Mature and Old-Growth Inventory rulemaking process — not in support of more inventory, but in support of preventive measures that keep mature and old-growth forests standing. The inventory is the technical basis for protection; the protection rule is the political fight. Specific, civil, sourced comments asking for retention requirements with measurable biological-legacy thresholds are the kind of input the agency reads^[8].

If you live elsewhere, find your country’s national forestry or environment ministry, find its retention or primary-forest guidance, and ask the same question: does the rule preserve the biological legacies the science says matter, or does it not? The Forest Stewardship Council certification standard is a useful third-party reference for what retention should require in a working forest^[7].

If you own forest land at any scale, write a retention plan before you ever cut. Decide which biological legacies you would retain — old large stems, snags, downed wood, riparian buffers, intact soil patches — and commit to that retention in writing. The plan does most of its work by existing.