Don’t Seal That Pruning Cut: CODIT Research Shows Trees Wall Off Wounds, They Don’t Heal Them

Walk into any garden center in spring and you’ll still find pruning sealant on the shelf, right next to the loppers. I stopped carrying it in my own kit years ago, after reading what the actual research says about what happens under that layer of black paint. It isn’t a matter of opinion or a newer product finally cracking the problem. Researchers tested wound sealants on 275 trees over 13 years, and the sealed wounds didn’t come out ahead. Some came out worse. The reason has nothing to do with a bad recipe and everything to do with what a tree actually does when you cut it — which isn’t heal.

Trees don’t have the cellular machinery to repair damaged wood the way an animal repairs a cut. What they do instead is called CODIT — Compartmentalization of Decay in Trees — a model developed by USDA Forest Service pathologist Dr. Alex Shigo. Understanding that one distinction explains why sealant helps nothing, and why the one situation where it’s still recommended has nothing to do with decay at all.

The Mechanism: Trees Wall Off Wounds — They Don’t Heal Them

When a branch is removed, the wood behind the cut doesn’t regenerate. It stays exactly as damaged as it was the day you cut it. What changes is that the tree builds a set of internal barriers around that damaged wood to keep decay organisms from spreading into healthy tissue elsewhere in the trunk. Shigo’s CODIT model describes four of these barriers, or “walls,” each doing a different job [1][3]:

Wall 1 plugs the vertical water-conducting vessels above and below the wound with gum and tyloses, slowing decay from moving up or down the trunk. It’s the weakest of the four. Wall 2 forms in the denser latewood growth rings, resisting inward spread toward the center of the trunk. Wall 3 forms sideways, in the radiating ray cells, using active defense chemicals to block decay from spreading around the circumference. Wall 4 — the barrier zone — is the strongest of the four and the one that matters most for pruning. It’s built by the cambium layer in the months after the cut, and its cells are packed with compounds that are directly toxic to decay fungi [1][3].

Three pre-planned garden beds, free

Stop staring at empty beds: printable plans with exact layouts, plant lists and planting calendars — yours free from the Garden Library.

Send me the plans

I agree to receive emails from Blooming Expert and accept the Privacy Policy. Unsubscribe anytime.

Wall 4 only forms correctly if the cut is made in the right place: just outside the branch collar, the slightly swollen ring of tissue where a branch meets the trunk. That collar isn’t decorative — it’s where the tree’s defense chemistry is concentrated, and it’s what allows Wall 4 to close in fast [1]. Cut into or through the collar (a flush cut) and you remove the tissue that was going to do the walling-off. There’s more on why that specific mistake matters below.

None of this happens overnight, and that’s part of why sealant felt intuitive in the first place — an exposed wound sitting bare for a season looks unfinished. Woundwood (the visible ring of new tissue rolling in from the cut edge) typically starts forming within weeks on a vigorous tree, but full closure over a several-inch cut can take multiple growing seasons, and Wall 4’s internal chemical barrier is doing its job long before the wound is visibly closed [1][3]. A young, actively growing tree in good soil will close a cut faster than an older or stressed one — vigor drives the timeline far more than wound size does, which is worth knowing before assuming a slow-closing cut means something went wrong.

What 13 Years of Testing Actually Found

The most detailed test of wound dressings is a study by Shigo and Shortle that tracked 275 experimental wounds across five species — red maple, red oak, white oak, American beech, and paper birch — for six months to seven years [2]. They didn’t just test one product. They tested bituminous (asphalt-based) paint, shellac, a rosin-acid and styrene-butadiene blend, the plant growth regulator Ethrel, and even aspirin and sucrose solutions, on the theory that something in that list might trigger a faster defensive response.

The headline finding: no material tested prevented decay, and none reliably improved wound closure compared to doing nothing [2]. What actually predicted how well a wound closed wasn’t the product — it was the individual tree’s own vigor and the amount of healthy wood behind the cut. Red oaks in the study showed minimal dieback almost regardless of what was applied; red maples varied enormously tree to tree, treatment or not [2]. That’s a genetics-and-health story, not a product story, and it’s the opposite of what a can of sealant implies on the label.

Why Sealant Makes It Worse, Not Better

It’s not just that sealant does nothing. In several of the tested materials, it actively worked against the tree. Two separate university extension sources describe the same mechanism from different angles: wound dressings “often seal in moisture and decay” rather than sealing infection out [4], and petroleum-based products specifically “impede callus tissue formation and successful compartmentalization” — meaning they slow the very process (Wall 4 formation) that’s supposed to be protecting the tree [5].

Put those two findings together and the mechanism gets clearer than either source states outright: Wall 4’s cambial cells need to stay hydrated at a normal, gradually-drying rate to synthesize their fungus-toxic compounds and physically seal the boundary [1][3]. A layer of paint or tar over the wound changes that drying curve — moisture gets held right at the cut surface instead of evaporating normally, and the small cracks that inevitably form at the paint’s edge become an entry point for the fungi it was supposed to keep out [4][5]. The wound ends up wetter, for longer, than it would have been bare. That’s the trap: the product marketed as a barrier against moisture and infection becomes the thing holding both in place against the wood.

The same logic applies to filling cavities with cement or spray foam, which some older guides still recommend for large trunk wounds. Extension research is consistent on this too — filling a cavity breaks the physical and chemical boundary the tree already built, and the plug traps moisture against a decay pocket instead of letting it dry [4][5]. If you wouldn’t caulk a scab on purpose, the same reasoning applies here.

It’s worth being specific about which sealants this applies to, since “sealant” covers a range of products people reach for. Tar-based and asphalt-based pruning paints were among the worst performers in the 13-year study, and shellac — a thinner, lighter product often assumed to be gentler — still produced no measurable decay reduction compared to an untreated wound [2]. None of the materials tested, thick or thin, showed a decay-prevention benefit worth the trade-off, which is why extension guidance across species and regions has converged on the same answer: leave the cut bare [1][2][4][5].

The Cut Matters More Than the Cover

None of this means the cut itself is inconsequential — the opposite is true. Sealant can’t fix a bad cut, and a good cut doesn’t need sealant. The distinction that actually determines outcome is whether the cut was made outside the branch collar (correct) or flush against the trunk, removing the collar (wrong). A flush cut takes out the tissue that would have formed Wall 4, leaving the trunk with an unprotected wound and no mechanism to compartmentalize it [1]. Decay organisms that would otherwise have been walled off in the branch stub instead have a clear path into the trunk itself.

This is a real, documented outcome, not a theoretical one. Cankers on dogwood are a well-known example of exactly this failure mode — a wound placed too close to the trunk, or a stub left too long past the collar, gives canker fungi an opening they wouldn’t otherwise have had, and once established the infection can spread into surrounding branches. Our dogwood pruning guide walks through exactly where that boundary sits on a dogwood specifically, because the margin for error there is smaller than on most trees.

On fruit trees, where large-diameter cuts are common, cut placement is worth getting right the first time rather than trying to compensate with a product afterward. Our apple tree pruning guide breaks down the 3-cut method professional orchardists use specifically to avoid tearing bark down into the collar on heavy limbs — a torn strip of bark is functionally the same problem as a flush cut, just caused by gravity instead of a bad angle. Get the mechanics of the cut right and there’s nothing left for a sealant to compensate for. For the full set of cut types and when to use each, see our pruning techniques guide.

The One Real Exception: Oak Wilt and Dutch Elm Disease

There is exactly one situation where painting a fresh cut is still standard advice, and it has nothing to do with decay or wound healing. Oaks and elms in regions where oak wilt or Dutch elm disease is active are vulnerable to sap-feeding beetles that carry fungal spores directly into fresh wounds. If an oak has to be pruned during the growing season — storm damage, a safety hazard, anything that can’t wait for dormancy — extension guidance is to paint the cut with ordinary latex house paint within 15 minutes of making it, before the fresh sap smell has a chance to attract beetles [6]. The paint isn’t there to help the tree compartmentalize anything; it’s a beetle deterrent, and it works by blocking scent and access, not by sealing in moisture the way sealant does elsewhere.

The simplest way to avoid needing this workaround at all is timing: prune oaks and elms in winter, from December through February, when both the beetles and the fungal spore mats that attract them are dormant [6]. Outside that narrow disease-vector situation, on any species, in any region, the research doesn’t support reaching for the paint can.

FAQ

Should I seal a large pruning wound differently than a small one?
No — the size of the wound changes how long compartmentalization takes, not whether sealant helps. A large cut on an otherwise healthy tree with good vigor will still close faster unsealed than a small one on a stressed tree with a sealant coating trapping moisture at the surface [2][4].

What about fruit trees and roses — do the same rules apply?
Yes. The compartmentalization mechanism is a general woody-plant response, not something specific to shade trees. The 13-year study included multiple species and found the same result across all of them: no material tested improved on an untreated cut [2].

Does a properly placed cut on a healthy tree really not need anything at all?
Correct — a clean cut just outside the branch collar, on a tree with reasonable vigor, is the complete protocol. The tree’s own chemistry does the rest [1][3].

If sealant doesn’t work, why is it still sold?
Mostly momentum. It was standard advice for decades before Shigo’s research changed the consensus in the 1970s and 80s, and old habits move slower through garden centers than through extension literature. The evidence against it has been consistent for over 40 years; the product just never left the shelf.

I have an old trunk cavity or hollow spot — should I fill it with foam or cement?
No. It feels like the same instinct as sealing a cut, but it’s worse: a filled cavity traps moisture directly against wood the tree has already worked to wall off, and removing the plug later to inspect or treat the area means breaking that barrier a second time [4][5]. If a cavity is a structural concern, that’s a job for a certified arborist to assess, not a DIY filling project.

Key Takeaways

Skip the sealant on routine pruning cuts — the 13-year record shows it doesn’t prevent decay and can trap the moisture that makes decay worse. Put the effort into cut placement instead: just outside the branch collar, never flush against the trunk, so the tree’s own Wall 4 barrier can form the way it’s built to. Paint is only worth reaching for on oaks and elms cut during beetle season in disease-active regions, and even then only as a fast-acting scent block, not a wound treatment. On everything else, in every season, the tree does this better without help.

Sources

• Penn State Extension — Understanding the Spread of Decay in Trees
• Shigo & Shortle, Arboriculture & Urban Forestry (ISA) — Wound Dressings: Results of Studies Over 13 Years
• University of Florida IFAS Extension — Decay Development in Trees
• UNH Extension — Should I Cover Large Pruning Wounds with a Tree Wound Dressing?
• Mississippi State University Extension — Tree Wounds: Should They Be Repaired?
• Iowa State University Extension — Pruning Oak Trees to Avoid Oak Wilt