Risks From Trees - a short discussion

Although many sources will state that the risks from trees to people are very low, there are a lot of caveats to that. This rating is a result of an average death rate of 5-6 persons/year in the UK caused by tree failures (Ball & Watt, 2013) – which, given the current population, puts a crude estimate of about 1 in 10 million chance that you’ll be killed by a tree or branch falling in any given year. However, there are many aspects that are not covered, like damage to property and infrastructure, injury to person (not death) and that one cannot take an average and apply it to all situations. So – the risks from trees in general to cause death of a person in the UK is very low – but the risks associated with one problematic tree to person and property can be very high, due to its poor or deteriorated structure and where it is situated.?

This high variation in risks from trees justifies hiring arborists to assess trees carefully (Dunster et al., 2017), and it also justifies further science to investigate what tree structures are capable of bearing, with or without defects and/or decay. The ideal would be that we could accurately assess the ‘factor of safety’ of component parts of a tree, much as engineers do for built structures: we are still far from achieving this, but we should strive towards this goal.?

Hollows and ‘central tendency’

There has been a lot of theories and debate about how to assess trees with hollowing stems over the last four decades, in terms of how likely they are to fail. We have moved from assessing the basic proportions of wood remaining versus the extent of the cavity/hollow, to currently considering the section modulus of the hollow section compared to a reference section set a little above the column of decay (Rinn, 2011). However, we still have quite a long way to go – because we also need to consider wood quality and the loading of the tree’s structure.?

Figure 1: Decay in a tree stem is often associated with additional growth (or ‘swelling’). To make a fair comparison of the load bearing capacity, one needs to compare the bulged and decayed section with a non-decayed reference section of the tree’s stem – as shown.?

In terms of wood quality, it is mature trees that tend to hollow out, so their outer stem wood consists mostly of mature wood, with a low proportion of juvenile wood (Nawrot et al., 2014). In addition, decay agents struggle to invade functional sapwood, which, again, lies at the outer edge of the tree’ stem, with more easily invaded dysfunctional sapwood (and potentially heartwood) situated further into the stem (Boddy, 2021). The outer wood, forming the ‘shell’ around the hollow may also have qualities of reaction woods – denser, tougher, stronger, and more resistant to splitting that the ‘normal’ wood that a tree lays down (Gardiner et al., 2014).?

These facts help to explain why you get to see so many hollow trees persisting in our landscape. One effect of these heightened defensive qualities of the outer wood of such hollowing trees is that one often gets to see what I call ‘central tendency’ – that, when decay occurs in the stem of a tree, it tends to feed upon the inner dysfunctional wood, up and down the stem, hollowing it out so it becomes like a pipe. Examples of this tendency of fungal decay to eat out the centre of living trees can be seen where there is one-sided root decay in a tree, but just higher up the stem, the cavity is often approximately centrally placed in the stem.?

Decay at the edge?

Arguably, theorists influencing the industry have spent too much effort on assessing the likelihood of failure of hollowing stems, when there are a much wider range of tree failures that occur which are relatively unexplored by science – and the failure of hollow stems is not in the least the most common form of tree failure. Trees more often fail by being uprooted, failure of their branches, or the failure of bark-included junctions.?

Even if we stick to the topic of stem decay, decay at the edge of a tree stem has received only very limited assessment, compared to when a tree is hollowing (Smiley et al., 2012). This is a shame, as decay at the edge of a tree’s stem or branch can have a lot more influence on that component’s factor of safety, potentially lowering it substantially. Smiley et al. (2012) progressively cut away outer sapwood from trees during pulling tests, finding that wood removed at the outer edge of the tree’s stem had double the effect of lowering the bending resistance of a tree stem, compared with wood loss due to an internal cavity. ?However, it would be good to go further and to assess real specimens with real decay scenarios, rather than rely on this one-off artificial experiment involving only a small number of trees.?

Figure 2: Failure of a branch of a semi-mature field maple (Acer campestre) due to previous squirrel damage. Although the associated dysfunction and decay looks superficial, decay at the edge of a limb or tree stem can substantially reduce the factor of safety of the damaged component.?

Mutual shelter?

Another neglected area of tree risk in arboriculture is that trees are often growing alongside other trees – in avenues, in groups, in shelterbelts and in woodlands. As a forestry student, mutual shelter within a forestry plantation was a primary topic that I was taught (Miller, 1985) – but the influence of mutual shelter is rarely brought up in arboricultural teaching. Unfortunately, there is no scientific work I can cite on the heightened risk of branch or tree failure in an urban tree when a neighbouring tree of a similar size is cut down.?

It's not hard to experience this effect of lost mutual shelter, though. I know of one woodland where two mature trees were removed – and within four years, two large limbs failed in adjacent trees: probably not a coincidence. Similarly, I have often had the experience of one tree failing at its roots – then, in another storm a couple of years later, the adjacent tree also being windthrown.?

Guidance on this is not often given – but we should all be made aware that removing one mature tree situated right next to the crown of another mature tree is very likely to increase the wind loading to the remaining tree, and thus increase the likelihood of a component failure in that tree (at least for some time, until the tree acclimates to its new conditions).

Figure 3: Domino-like failure of spruce in a plantation. Mutual shelter helps to keep these trees standing upright and when one fails during a storm, others may well follow.?

Too brittle to climb?

Dead and dying trees decay at different rates – the species of that dead tree can have a big effect – but at some point, they become too precarious to climb. This has come to our industry’s attention recently because of ash dieback (causal agent: Hymenoscyphus fraxineus), where ash trees that decline severely or die, their branches can become problematically brittle over a relatively short period of time due to white rot occurring as branches and stems of these infected ash trees die.?

However – at least in the past – awareness of some arborists that a tree pruning, or removal task may need the use of a MEWP has been lacking. A few years ago, I got to hear of a local tree surgery firm that failed their AA accreditation on the basis that they started to climb a tree that was clearly severely decayed – and should not have been climbed.?

This awareness of the brittleness of dead and decayed trees needs to spread wider – to tree owners – so if they want their dead tree removed or made into a snag, they act in due time. This is one simple message that can be given out by the industry that will help achieve safer tree operations. Essentially, we need a ‘Dead Tree Awareness Week’ – as it is quite common for laypeople to have no idea that a tree under their ownership is dead and thus it needs reducing or removing.?

Figure 4: A large dead tree in a suburban garden. The longer the owner takes to get hold of an arborist to cut this tree, the more of a problem it will be to remove, as it becomes further decayed and more brittle.?

Research enlightens?

The work of several scientists has informed changes in the ‘hollow tree debate’ over the years: as another example, my own research work has dispelled several myths about branch junctions – with or without bark included in them. I have dedicated twelve years to studying and experimenting with branch junctions – and a much more clarified picture of branch junction anatomy, likelihood of their failure and the weakening effect of included bark can now be taught to arborists (Slater, 2022).?

The biggest breakthrough in my research was to find that ‘natural braces’, which form in many trees, are the primary cause of many bark-included junctions. This not only explains why a bark-inclusion has been formed in many cases, but it also gives us some good ‘tools’ about how to avoid these weakened junctions forming or failing, depending on the timing of our tree management.

?Figure 5: At the base of this grey alder (Alnus incana) is a bark-included junction. However, the likelihood of the failure of this weak junction is low, because the two stems connected via this junction are connected by a natural brace higher up, as seen. This is no coincidence: the natural brace formed has caused the bark-included junction lower down in the tree – as that junction has experienced little-to-no movement, so not enough mechanical stimulus to grow to be a strong junction.?

Lessons Learnt?

Two scientific methods combined are the way forward in understanding tree failures and associated risks. Large cohort studies of trees, especially surveying trees after storms have passed through, can help to highlight what are the most frequent forms of failure and what visible signs (if any) there may have been of use to the arborist to predict the failure. Controlled scientific experiments also have their role to play, testing component parts of trees and improving modelling so that one can calculate the factor of safety of a suspected defective part. Essentially, we need more people in the industry to dedicate their time to filling our knowledge gaps on the likelihood of failure of different component parts of trees – to the extent we can confidently give a factor of safety to an inspected component of a tree.?

Do bear in mind, though, that even if all this research work is done, on an individual basis, tree risk predictions will remain inaccurate and that is because a key element that causes tree failure is inherently unpredictable – the weather!

?

References

Ball, D.J. and Watt, J., 2013. The risk to the public of tree fall.?Journal of Risk Research,?16(2), pp.261-269.?

Boddy, L., 2021. Fungi and trees: their complex relationships. Stroud, England: Arboricultural Association.?

Dunster, J.A., Smiley, E.T., Matheny, N. and Lilly, S., 2017.?Tree risk assessment manual. Ed. 2. Champaign, Illinois, USA: International Society of Arboriculture.?

Gardiner, B., Barnett, J., Saranp??, P. and Gril, J. eds., 2014.?The biology of reaction wood?(p. 274). Heidelberg, Germany: Springer.?

Miller, K.F., 1985. Windthrow Hazard Classification. Forestry Commission Leaflet No. 85. London: HMSO.?

Nawrot, M., Pazdrowski, W., Walkowiak, R., Szymański, M. and Ka?mierczak, K., 2014. Analysis of coniferous species to identify and distinguish juvenile and mature wood.?Journal of Forest Science,?60(4), pp.143-153.?

Rinn, F., 2011. Basic aspects of mechanical stability of tree cross–sections.?Arborist News,?20(1), pp.52-54.?

Slater, D., 2022. Guidance Note 14: Branch junctions: a classification system for arborists. Stroud, England: Arboricultural Association.?

Smiley, E.T., Kane, B., Autio, W.R. and Holmes, L., 2012. Sapwood cuts and their impact on tree stability.?Arboriculture & Urban Forestry,?38(6), pp.287-292.?

About the Author?

Dr. Duncan Slater is a senior lecturer in arboriculture at Myerscough College, Lancashire. He holds six university degrees, including an MSc in Resource Management, an MSc in Environmental Management and a PhD in Plant Sciences. However, there is still much more to learn about trees, ecology, mycology and the good care of this planet.

***This Article First Appeared in ProArb Spring Edition 2024***