What in the world caused this?
Early in my career, I followed an older colleague through a dark forest. He’d been telling me highly suspect stories of Sasquatch sightings in that very forest, when suddenly his voice quieted. “We’re almost there,” he whispered. We slipped into a light-filled gap. Looming above us was a white pine tree, long dead, its bark long gone. “Look at the trunk,” he whispered. I saw my first argyle tree.
Since then, I’ve seen plenty of trees with argyle patterns on their trunks—white pines and ponderosa pines, sometimes in person but usually in photos that get forwarded to me, all of which come with the question, “What in the world caused this?” along with some awkward suggestions. One such suggestion: “Sasquatch using nunchucks.”
I don’t think so.
As a professor of forest biology in Oregon, I’ve been able to ask a good number of other forest researchers and wood scientists, “What in the world?” We have a best guess, but (of course, waving our arms) caution that we’d know a lot more if we actually made some systematic observations, dissections, and mechanical tests. (I have a bunch of suggestions, if you’re interested!) But lacking those, here’s what we think: The cross-hatches help lower stresses when the trees twist in the wind.
What made us think that?
We came to that conclusion after making a lot of simplifying assumptions. If a tree were a column made of something like cheese that has the same strength in all direction, and you twisted that cheese column, the highest stresses would be at +45° and -45° to the direction of the column—exactly where we see these X’s. The extra cheese material in the X’s would provide more material in a local area to bear the stress, and so the stress at any one point in the cheese would be lower.
Even though there are many ways in which a tree trunk is not like cheese, there is evidence in other cases having to do with tree support that trees develop a shape that will make the stresses more-or-less even within the trunk it is “loaded” (under stress). For example, trees rock in the wind, causing a local stress concentration at the tree’s base—and in response, trees are wider at their bases. That butt swell (the actual technical term!) decreases the stress concentration by giving more wood for the stress to be spread over. People have used finite-element modeling in many situations to predict the stem shape that will confer uniform stresses throughout (such as where a trunk and branch meet, forked trunks, and trunks growing around fixed objects like rocks). The modeled shapes are usually remarkably close to what trees make.
What assumptions did we make that might be wrong?
Trees don’t only alter the amount of wood they put in different places as they grow, but they also alter the mechanical properties of that wood—so they could, for example, make stronger wood in some places and more flexible wood in other places. If so, then their growth could result in uniform stresses using a combination of material properties and shape. We made this best guess on shape alone.
And wood doesn’t have the same properties in every direction the way cheese probably does. Wood is much stronger along its main axis (up and down) than across it. In that way, a tree trunk is more similar to a bundle of uncooked spaghetti noodles than cheese. If you yank the noodles lengthwise, they don’t stretch appreciably. If you grab the two sides of the bundle and yank outward, the bundle falls apart. But even so, if you twist the column of noodles, you will cause stresses on the surface at +45° and -45°.
And wood fibers don’t necessarily run straight up and down the tree; they may run in spirals. In a tree with such “spiral grain,” the tree is strongest when pulled along the same angle as the fibers, not straight up and down. But argyle trees do not have spiral grain: their fibers run up and down, as can be seen from the vertical cracks in this dried tree.
And the wood fibers can vary significantly in their mechanical factors from place to place in the tree. The variation occurs through different amounts of cell wall (which alters wood density) or through changes in the orientation of the strength elements within the walls of the wood cells. These wood cells are about the size of an eyelash and hollow, and have several layers in their walls. Each layer has internal reinforcements (strands of super-strong cellulose) at a characteristic angle, like steel-belted radial tires. If the reinforcements in the thickest layer are at a flat angle, the wood is weaker when it’s pulled up and down than if the reinforcements run more vertically. But even so, these potential differences in material properties are unlikely to have much effect on where the primary stresses are on the trunk’s surface when it twists.
On top of all of the differing wood properties in different places, trees aren’t untapered columns and their cross-section isn’t necessarily circular. And they have branches.
But even in the face of these potential wrong assumptions, a column-like structure that is twisted will have its highest principle stresses on the outside of the trunk at + 45° and –45°: exactly where the X’s are situated.
Why are there X’s rather than just ribs?
Twisted stems untwist, too. When a tree rotates in one direction, it develops high principle stresses at +45°. When it untwists, the trunk will have high principle stresses at -45°.
These X’s are only on one side of the tree-trunk though, and in some trees, the X’s do trail off into ribs that are longer in one direction than the other. I presume the assymetrical lengths of the ribs in the X’s are related to higher average forces in one direction than the other.
Why would a tree twist?
A tree would be vulnerable to twisting if it grew in a windy site, and a) had larger branches on one side of the trunk than the other, or b) if its environment had more wind on one side of the canopy than the other. Either way, the wind’s force on branches on one side of the tree is greater than on the opposite side, and the trunk twists.
Why are the X’s in those particular spots on the trunk?
In trees that have fainter X’s, the X’s are centered where a branch used to be. The branches would have been the levers that the wind pushed on, and so the high stress concentrations—and the center of these X-s—would be at these branch junctions. I would guess that all the X’s originated at the base of branches.
Why do only a few trees in a stand have argyle stems?
Like people, individual trees have different genotypes. The trees that make argyles apparently have a propensity to react to the wind stress by developing argyles. Other individuals in the stand may have different reactions—perhaps active ones, such as having shorter branches or not growing so tall, and perhaps passive ones, such as breaking more often.
Are argyle trees associated with habitat?
Yes. Argyle trees seem to develop in places that are prone to high winds, such as in open stands on hillsides and near ridges. In Oregon, you can see them from Highway 20 near Hoodoo where the B&B Fire Complex burned 90,000 acres in 2003. The bark has fallen from these dead trees in the wind-exposed pass area, so there are a lot of trunks to look at. (Be sure to look at sites where ponderosa pine is coming back, not other species.)
How about tree size?
I’ve only seen the argyle pattern in fairly large-diameter trees (at least a couple feet in diameter near the base). I expect the development will not happen until the tree has been growing long enough for its stem of, say, 6” diameter to reach into the windier zone of the atmosphere.
So that’s our best guess, but only the trees know, deep down, what causes argyle stems. I’m arm-waving. We need branch-waving instead. Until we get that, I’d love to hear your ideas and observations!
Some work examining the uniform stress hypothesis:
- Morgan J, Cannell MGR. 1994. Shape of tree stems—a re-examination of the uniform stress hypothesis. Tree Physiology 14:49-62.
- Mattheck C. 1995. Biomechanical optimum in wood stems. In Plant Stems: Physiology and Functional Morphology (ed. B. L. Gartner). Springer, pages 75-90.
- Dean TJ, Roberst S, Gilmore D, Maguire DA, Long JN, O’Hara, KL, Seymour RS. 2002. An evaluation of the uniform stress hypothesis based on stem geometry in selected North American conifers. Trees 16:449-568.
Thanks for discussion—especially to Dean DeBell for showing me my first argyle tree, to Bob Leichti, who coined the term “argyle stem,” to John Nairn who was patient with all my ideas, and to Everett Hansen for tolerating slow trips when I kept jumping out of the car or off the trail to take pictures.