On Removing Deadwood, Part 1

Does removing dead branches truly improve the health of a tree? Photo by USDA Forest Service – Northeastern Area, Bugwood.org.

Last December, we pruned some hazardous deadwood from a large northern red oak that had been hanging over a client’s garage. She was primarily concerned with not having another hole in the roof of the garage from falling deadwood.

The customer had told us a tree service condemned the tree. The salesperson took one glance at the tree and exclaimed it had to come down. A tree of this size is not a cheap removal. It is a little embarrassing that there are salespeople in arboriculture who actually leverage fear as a sales technique.

After a brief inspection, I didn’t see a single thing that justified immediate removal. The tree was not in tip-top shape, but it didn’t show any significant defects from any viewpoint on the ground, let alone the single position from which this salesperson saw the tree.

A large, woody vine that grew up the side of the trunk had been severed at its base years ago. The vine was dead but still on the tree. It looked a little ragged, but it was not impacting the tree’s health whatsoever. There were several large, dead branches in the tree’s crown, which corresponded with a driveway having been installed over the tree’s roots about 15 years prior.

The soil level was high surrounding the trunk, too, but these are all normal things that happen to trees in urban areas.

The customer was interested in keeping her garage roof damage free, so that’s what we addressed; we reduced the length of the dead branches to a safe size. After we finished our pruning and said our goodbyes, I emailed our customer her receipt. When she replied, she said she was grateful that we “gave her tree a second chance at life.”

The sentiment was nice, but that isn’t at all what we did. All we did was reduce the residual risk of the tree and make it safer to be around. The customer was actually grateful that our professional opinion was contradictory to that of a lame salesperson, and perhaps that’s what “saved the tree.” If I were to address the tree’s health, I would tend to the soil issues first and foremost.

So that got me wondering: Do people really think removing deadwood improves the health of a tree?

The answer is, “Yes. Yes, they do.”

That’s what tree companies tell homeowners, and it’s an easy and intuitive thing to sell. It is part of the old dogma of tree work, along with the long list of other needless things sold to people for their trees (e.g., topping, thinning, excessive pruning, etc.)

I began fishing through my textbooks and online studies with intent. What I found is complicated.

From ISA podcasts to Dujesiefken’s Trees – a Lifespan Approach, or in TreeBuzz forums, everybody refers to deadwood as simply “deadwood.” It seems nobody looks further than that. In this article, I look into the ways branches die, how the tree copes with dead branches and if removing those parts truly improves the health of a tree.

Each of the processes discussed from here forward is of healthy, functioning plants. These processes can fail, but they are described as if they are functioning optimally. These processes may not work correctly in a low-energy-state, weakened or diseased tree.

Section 1


We’re familiar with leaf abscission in autumn, but the process is the same when it comes to cladoptosis (which is the natural shedding of branches), just at a larger scale (Bhat et al.) (Hirons). From here on out, I will use the terms “abscission” and “cladoptosis” interchangeably for the sake of clarity. They essentially mean the same thing in the context of this article because we’re talking about tree crowns, not tree leaves.

The tree will shed particular branches when it costs too much to keep those parts alive; those parts use more photosynthates than they produce (Bhat et al.) (Bellani et al.). In essence, trees are often modular in this way. It also is theorized that trees will undergo cladoptosis in response to drought, to reduce the surface area where water can be lost (Bhat et al.) (Hirons).

The plant detects the dysfunctional branch and the process of cladoptosis begins – it is a function of the plant to do this. The catalyst of cladoptosis is usually abiotic, not pathogens (although sometimes they are the catalyst). Excessive shading or inefficient photosynthesis, retrenchment or the expected loss of crown mass following root-mass severance can trigger branch abscission (Addicott) (Hirons) (Pallardy). Dysfunctional branches become biologically separated from the tree, although they may remain physically attached (Bellani et al).

Generally speaking, abscised branches are found in the interior and lower parts of a tree’s crown (Bellani et al.). My anecdotal experience also supports this observation, and I’m sure most arborists can agree with this. For example, here in Michigan, perfectly healthy Norway maple trees (Acer platanoides) regularly shed interior branches. Honeylocust (Gleditsia triacanthos) shares the same habit, sometimes abscising large branches, almost lion-tailing themselves.

Further notes on abscission

In Dujesiefken’s Trees – a Lifespan Approach, in a section about cladoptosis, he refers to a predetermined breaking point caused by fungi. He goes on to say, “The decay does not usually penetrate through the boundary layer into the stem, as a result of this process, known in forestry terms as ‘self-pruning.’”

He attributes this predetermined breaking location to the fungi. Contrary to his statement, both Bhat and Bellani’s studies show this is caused by the tree itself: the abscission zone. It may be more accurate to say saprophytes (fungi) will break down the wood near the abscission zone, which accelerates detachment of the part.

Deadwood may be removed to protect potential targets. Photo by Andrew Koeser, ISA, Bugwood.org.

The abscission process goes through recognizable phases (Addicott). I’m going to highlight the abscission zone, because it is the primary part of the abscission process that prevents decay ingress into the main stem.

Abscission zone

In branches that the tree abscises, a cellular layer called the abscission zone forms. This is where things get interesting and a stark difference from CODIT (compartmentalization of decay in trees) becomes clearer.

Within the abscission zone, there is a region of cells (the separation zone) that do not contain lignin (Bhat et al.), which the saprophytic fungi would ordinarily consume (“Saprophytic Fungi” fungimap.org.au). This zone functions like a gap that makes it more difficult for potentially opportunistic fungi to enter the living part of the tree while the branch is attached (Bhat et al.). Outside of this zone is a layer of cells packed with suberin and lignin, which serves as additional protection from pathogens once the limb has detached (Addicott) (Bhat et al.) (Stern).

The construction of the abscission zone prevents decay ingress while the branch is attached and after the branch is detached from the tree.

Abscission compared to CODIT

Both abscission and CODIT have functions to stop the spread of decay into the main stem, albeit in slightly differing ways.

When talking pruning practices, the application of CODIT in trees to cutting live wood is well supported and applicable (Shigo et al.). CODIT, however, does not fully explain the processes involved with the tree shedding branches on its own with no apparent wounding (Gilman). It is a different process, though it does share some overlap with abscission, which I will illustrate in the comparison section just ahead in this article (Dujesiefken et al.).

Here, I want to review the similarities and differences between branches killed by wounding/pathogen entry and branches killed by abscission. 


In abscised branches, the tree – at least in some cases – will clog the xylem with tyloses, or gums or resins, above the abscised branch to prevent vertical spread of decay (Bhat) (Hirons).

Bhat’s abscission study writes:

…the rays at this level of the branchlet, and slightly above, where the mechanical rupture occurs show dense accumulation of dark brown contents… Substances such as polyphenols, tannins and resins, often found deposited within xylem tissue, are generally believed to serve a protective function owing to their toxic or repellent properties.

Bhat here is describing how “Wall 1” functions in the CODIT model (Gilman). This is a shared characteristic between abscission and the CODIT model. Secondary growth, or basically trunk expansion, will eventually form distinctive trunk collars on limbs dead from both abscission and other means. A “Wall 4” eventually will be present on both wounds and normal abscission, though the abscission zone doesn’t depend on the presence of this fourth wall to protect the stem.


The placement of the abscission zone and the placement of the internal walls of CODIT differ – that is, the placement of their major mechanisms of preventing decay entry.

“Wall 2” exists only to prevent the inward spread of decaying organisms to the main stem. Its strength relies on the difficulty organisms have when passing through the distinct layers of growth rings (Gilman). As we’ve learned CODIT’s functions, we’re to ensure our cuts are as close to the branch collar as they can be to expedite wound closure (Gilman).

When it comes to cladoptosis, on the other hand, the abscission zone is what prevents inward spread of decaying organisms. Very interestingly, the abscission zone seems to ignore the branch-attachment zone, as shown in Image 1.

Image 1: Left – A dead branch with a collar-like swelling of living tissue as its base, formed by the stem. Right – Wood-destroying fungi create a predetermined breaking point close to the stem. From Trees – A Lifespan Approach, Dujesiefken, Fay, de Groot, de Berker.

The tissue behind the abscission zone shows slight discoloration. But the branch-attachment zone itself remains the same color as the surrounding trunk tissue, which leads me to determine that tissue is unaffected by decay. The branch shows expected discoloration as the stem dries, and the decay process continues therein.

On a branch removed in pruning, close to the collar, the CODIT walls behave with more obedience to the branch-
attachment zone. (Image 2)

Image 2: From Trees – A Lifespan Approach, Dujesiefken, Fay, de Groot, de Berker.

Once we cause damage to the tree, the remaining bit of branch will decay into that branch-attachment zone (Dujesiefken et al) (Gilman). Ideally, the process of wound closure secludes that small decay pocket, creating an airless space where fungi cannot continue to decay into the tree. The decaying organisms are allowed closer to the main stem when we cut live wood (CODIT) than when the tree kills branches on its own (cladoptosis).

CODIT describes how the tree will respond to air entry into wounds on live tissue, such as pruning or a breakage (Dujesiefken et al.). This is a fundamental difference between a branch killed by abscission and a branch killed by a poor pruning cut – the entry of air into living tree tissue initiates the CODIT process. In contrast, in abscission, there is no air entry into living tissue at all (Dujesiefken et al.).

Dujesiefken explicitly says decay does not usually penetrate through the boundary layer into the main stem in abscised branches. There is no air inside behind the abscission layer if it is healthy. Decaying organisms need oxygen to survive. And without that, decay can’t happen.

Graphic 1 best sums up the similarities and differences between the processes trees undergo to deal with damage to living wood and how trees shed their own branches.

Graphic 1: The similarities and differences between the processes trees undergo to deal with damage to living wood and how trees shed their own branches.

Section 2

Why compare abscission with CODIT?

When comparing abscission to CODIT, it seems like I’m comparing two different things. One function describes how trees kill their own limbs. The other explains what happens when healthy tree tissue is damaged. They’re two different things, right? Yes.

I compared these two plant functions because they both have mechanisms to prevent decay ingress into the tree. Removing dead, abscised limbs from trees under the notion that the tree needs to compartmentalize over the deadwood to protect itself isn’t necessarily true. Applying the CODIT principles of wound closure to branches that were abscised doesn’t hold up, because the mechanisms in place do not depend on wound closure.

As explained in the above section, the mechanisms the tree has in place to protect itself are different when the tree itself has killed off its branches, versus when an outside source causes damage to live tissue.

It would seem, then, that establishing how a branch has died is important in determining if removing it would have a positive impact on the tree. If the cause of death can be reasonably determined, you know what mechanisms are in place. Pruning abscised branches probably does not improve the overall health of the tree. But if the branch died from other means, like a bad pruning cut, the course of action may be different, because air has entered the living branch.

There are a few ways to tell the difference between dead abscised branches and dead non-abscised branches. There are many variables involved, but these three simple descriptions are worth noting.

  1. Abscised

Abscised branches are likely to be whole, with all of their large architecture intact; the small branches and twigs would fall off more easily over time. You likely won’t see old cuts on abscised branches. If you see breakages with no response wood, that limb could have broken after it abscised.

The presence of saprophytic fungi species on a dead branch with no apparent wounding could be considered benign. Granted, it is possible to have many species of fungi colonizing a single piece of wood. Regardless, in a healthy tree, the mechanisms in place should keep the main stem protected.

  1. Non-abscised

Dead branches with an obvious pruning cut heading them off probably didn’t die from abscission. The same could be said about a dead broken branch that shows response growth. In either case, that type of damage serves as an entry point for decaying organisms.

If you can positively identify pathogenic fungi on a dead branch with the presence of a mushroom, removing this branch would be beneficial because it, at least, poses a latent threat to the tree should it be in a low-energy state (Dujesiefken).

  1. Location

Determining the cause of deadwood also can be aided by the branch’s location – where in the crown is it? Generally, abscised branches are in the lower and interior part of the crown, while non-abscised limbs can be seen anywhere in the crown.

So, what if the crown has large, dead branches at the top? It presumably isn’t abscission, and it may be more reasonable to assume something more serious is happening. Have the tree’s roots recently (recent in terms of tree time) been severed? What’s the tree species? Could it be bark beetles?

If there are dead lower branches with no apparent wounding and with other noticeable defects around the limb, it may not be abscission – for example, a dead limb with an entire decay column visible beneath likely doesn’t qualify as abscission.

In Part 2, we’ll look at arguments for removing deadwood, the author’s speculations and implications and questions.

Works cited

Addicott, Fredrick T. Abscission. Univ. of California Press, 1982.

Bellani, Lorenzam., and Alessandro Bottacci. “Anatomical Studies of Branchlet Abscission Related to Crown Modification in Quercus Cerris L.” Trees, vol. 10, no. 1, 1995, doi:10.1007/bf00197775.

Bhat, K. V., et al. “Anatomy Of Branch Abscission In Lagerstroemia Microcarpa Wight.” New Phytologist, vol. 103, no. 1, 1986, pp. 177–183., doi:10.1111/j.1469-8137.1986.tb00606.x.

Chalker-Scott, Linda. “Science of Arboriculture Podcast.” What Does Science Say about Pruning Mature Trees? Oct. 20, 2017

“Deadwood.” The BuzzBoard, www.treebuzz.com/forum/threads/dead-wood.9435/.

Dujesiefken, Dirk, et al. Trees – a Lifespan Approach: Contributions to Arboriculture from European Practitioners. Fundacja EkoRozwoju, 2016.

Gilman, Edward F. An Illustrated Guide to Pruning. Delmar, 2012.

Hirons, Andrew D., and Peter Thomas. Applied Tree Biology. John Wiley & Sons, Inc., 2018.

Pallardy, Stephen G., and T. T. Kozlowski. Physiology of Woody Plants. Elsevier, 2007.

“Saprophytic Fungi.” Saprophytic Fungi, fungimap.org.au/about-fungi/saprophytic-fungi

Shigo, Alex L., et al. Compartmentalization of Decay in Trees. Forest Service, U.S. Dept. of Agriculture, 1977.

Stern, Kingsley R., and James E. Bidlack. Introductory Plant Biology. McGraw-Hill, 2008.

Jeremiah Sandler, a Certified Arborist and ISA Tree Risk Assessor Qualified, is owner/operator of Tree First Arboriculture in Royal Oak, Michigan. This article first appeared on his website, www.treefirst.org.

Leave a Reply

Your email address will not be published. Required fields are marked *

Click to listen highlighted text!