Does Pruning Reduce the Risk of Tree Failure

By Brian Kane
Amherst, MA (March 1, 2008)- Arborists often confront requests from concerned clients such as, “Can you storm-proof my trees?” and, “Make that tree smaller, I’m worried it’s going to fall on my house during a hurricane!” Pruning is one of the most common arboricultural practices, yet we know very little about how it affects wind loading and tree failure.


In the past, trees were topped and lion’s-tailed to “storm-proof” trees. Topped trees, the reasoning went, were smaller and therefore had less wind load. Because they had less of a sail area, lion’s-tailed trees were also believed to be less likely to fail-wind could just pass right through the canopy with all those leaves gone. Neither topping nor lion’s-tailing is good pruning, but we’ve reached that conclusion based on considerations of tree health, not because they didn’t reduce the risk of tree failure. Because so much foliage is removed, both topping and lion’s-tailing can reduce the tree’s ability to photosynthesize. Topping, in particular, also promotes decay where the topping cuts were made.
Recently, researchers have begun to examine how pruning affects wind loading of shade trees, and we’re just starting to understand this complicated interaction of wind and trees. Some common themes have been discovered, and future studies will hopefully shed more light on the subject. Here’s what we’ve learned so far.
Engineers and physicists call the force of the wind on a tree, “drag,” and the equation they use to predict drag shows that drag is related to several factors, the most important of which are the shape and area of an object and the wind speed. To demonstrate these factors, imagine standing on a tall building when the wind is blowing at 40 mph; you feel drag because your blocking the movement of air. You’d feel a lot more drag if the wind speed doubled, because drag is related to the square of wind speed. All other things being equal, if the wind speed doubles, drag would increase by four times because 22 = 4; if wind speed triples, then drag increases by nine times (32=9). Now imagine standing on the same building while you’re holding an umbrella-you experience much more drag because your area has increased, and you block more air as it moves by. You may have scared a flying squirrel out of a tree you were pruning. The squirrel doesn’t really fly, but it floats through the air and lands softly because it increased the drag on its body by extending its legs and exposing flaps of skin that increase its body area. The last main factor that affects drag is shape. We see evidence of why shape is important in the design of new cars, which have a streamlined profile. If two cars have the same area, but one has a more streamlined profile, it will experience less drag. This is one way that car manufacturers try to improve the fuel economy of their vehicles.
Since it’s usually not practical to change the wind speed around trees, arborists are left to change the size and/or shape of a tree to reduce drag. Pruning obviously reduces the canopy area of a tree, and, intuitively, we expect that drag should be reduced. This is generally true, but it turns out that predicting how much drag will be reduced from the amount of canopy area removed is challenging. The main reason is that it’s hard to measure the canopy area-it’s not like a stop sign that has a well-defined shape. Taking digital photos can improve the measurement, but this requires expensive cameras, digital imaging software, and a clear background behind the tree. Even under the best circumstances, converting a three-dimensional object like a tree into a two-dimensional image skews the area measurements. The good news is that despite these complications, researchers have shown that tree weight is a good predictor of drag. The bad news is that it’s not really practical to carry a scale around and weigh brush before feeding it to the chipper. In general, if you prune more (that is, remove more area or weight), you’ll reduce drag more.
Pruning also changes tree shape, which affects 1) whether wind can flow freely through or around the canopy and 2) the center of pressure height. In general, if the canopy is more porous, it’s easier for wind to flow through it. However, even if the canopy looks pretty open at low wind speeds, the reconfiguration of leaves and twigs often makes the canopy less porous at high wind speeds. There are very few studies that have carefully examined how air moves around (or through) an individual canopy for a good range of wind speeds, so this is something we need to understand better. Throw in the effect of pruning, and it becomes even harder to predict what will happen. For example, we’d expect that reduction pruning doesn’t really create a less porous canopy since it doesn’t remove branches from the interior of the canopy. However, by reducing the length of the branches, reduction pruning increases the stiffness of those branches since they don’t have as much leverage exerted on them. As a result, when leaves on the reduced branches begin to reconfigure, the canopy may become more porous than we expected. Thinning on the other hand may create a more porous canopy at low wind speeds, but as the wind speed increases and branches and twigs bend away from the wind, the canopy may become less porous, although its area is reduced. Understanding how air flows through the canopy is important, then, so we can figure out whether creating a more porous canopy is beneficial to reduce drag. You can see how complicated the interaction of wind and trees is, and we haven’t even considered the dynamic effects of wind, which have only very recently been investigated.
Pruning can also change the center of pressure height, which effectively changes the leverage on the tree during a wind storm. Leverage is important because the same force that acts on a longer lever creates greater torque or bending than a force that acts on a shorter lever. Anyone who’s used a rope to pull a tree as its being felled understands leverage: you don’t set the rope a few feet above the notch because you won’t get any leverage when you pull on the rope. Instead, you set the rope as high on the trunk as is safe and practical, which increases your leverage and allows one person to pull a tree that weighs many times more than they do. Of the common pruning types, reduction pruning reduces center of pressure height, since the canopy is shortened. On the other hand, raising increases the center of pressure height because only the lower branches are removed. Thinning doesn’t really change the center of pressure height, according to the handful of studies that have looked at this question.
So, how does pruning affect wind loading and the likelihood of tree failure? The following recommendations must be taken cautiously, because 1) there are only a few studies that have examined this question, 2) the trees that have been tested have been relatively small, and small trees don’t necessarily respond to wind the same way larger trees do, 3) winds have been generated artificially, which does not reflect real-world wind that is gusty, and 4) no dynamic analyses of pruning and the wind loading of shade trees have been undertaken. Of the three common pruning types from the ANSI A300 (Raise, Reduce, and Thin), reduction pruning seems to do a better job at reducing drag and bending on trees, when the weight of foliage and twigs removed has been accounted for. Obviously, if you prune out more foliage and twigs, you’ll reduce the drag more, but over-pruning adversely affects tree health. It is worthwhile noting, however, that when deciduous trees are not in leaf, they experience much less drag and bending, even more than fairly severe pruning. If you work in areas where winter storms are the most likely to cause tree failure, drastic pruning of deciduous trees might not be terribly effective at reducing the likelihood of failure. Raising has been the least effective pruning type, and this has been attributed mostly to the fact that raising increases the center of pressure height and increases leverage. So even though drag is reduced by raising, the increase in leverage offsets the reduction in drag and the tree still experiences a fair amount of bending. It’s important to remember that leaves experience much more drag than twigs and branches. Consequently, pruning types that remove proportionally more foliage than branches will be more effective at reducing drag (again, in consideration of all the caveats mentioned above).
One last factor that has not been addressed in any studies yet is the long-term effects of pruning on re-growth of trees. Although reduction pruning has been more effective at reducing drag and bending, the tests were conducted on trees right after they were pruned. It’s not guaranteed that the effect of pruning would be as beneficial once the tree begins to grow after pruning. Another area of research that is important, then, is investigating how different trees respond to different pruning types-which trees are most likely to sprout profusely after a particular pruning type, and under what conditions? For example, if reduction pruning causes excessive growth on some species, the possible benefits of pruning would only last for a short time.
Although researchers have started to attack the complicated issue of pruning and wind loads, we need to do much more research on pruning to understand all these factors better. We also need to start addressing the complex effects of gusty winds and swaying trees, which will help predict the likelihood of tree failure before and after pruning.
Related Resources:
Kane, B. and E.T. Smiley. 2006. Drag coefficients and crown area estimation of red maple. Canadian Journal of Forest Research 36:1951-1958.
Kane, B., M. Pavlis, J.R. Seiler, and J.R. Harris. In Press, Canadian Journal of Forest Research. Crown reconfiguration and trunk stress in deciduous trees.
Pavlis, M. B. Kane, J.R. Harris, and J.R. Seiler. In Press, Arboriculture & Urban Forestry. The effect of pruning on drag-induced bending moment of shade trees.
Smiley, E.T. and B. Kane. 2006. The effects of pruning type on wind loading of Acer rubrum. Arboriculture & Urban Forestry 32:33-40.