Trees Over Time Part Six: Branch Junctions

Figure 1 (above): Cup union in common ash (Fraxinus excelsior L.), the junction’s bark ridge clearly growing in length (and thus the apex of the junction has risen higher) over the five years of this time-lapse.

A Unique Article

There are two aspects of ‘uniqueness’ I find interesting. First, there is that saying: “You’re unique… just like everybody else!” Second, many English teachers insist there cannot be gradations of uniqueness: something is unique or it is not: and they can boil-over when one of their pupils says something is “quite unique” or “very unique”. 

The former usage of the term is a great ‘leveller’ that can be helpful to keep in check that ‘special person’ who might be getting too big for their boots: the latter rule, however, I think is over-pedantic. As all objects in the universe have a different history and are different in terms of their position in time and space, then every object is unique if you accept that second definition literally – and that, in turn, makes the term ‘unique’ quite meaningless. For example, 1,000 printed leaflets carrying the same text and images would all be unique – as they all contain different molecules of different origins and orientations, and each has a different position in the space/time continuum. I don’t think that’s what we mean when we say something is ‘unique’ – we usually mean something that is a one-off, different to the extent that nothing else is similar to it, although some things might be a bit similar. In which case, there are arguably gradations of uniqueness that are worth expressing. 

I have spent a considerable proportion of my life studying branch junctions in trees, determining their anatomical features, finding that natural bracing affects their development, mastering how to make bark inclusions artificially and determining that the twisting dense wood formed in the axil of most branch junctions is a reaction wood (axillary wood). I have also acquired a quite unique collection of ‘Trees over Time’ (ToT) images of branch junctions changing as they develop and grow. I can’t find that any other researcher has recorded visual changes in 50+ branch junctions in trees using photographic methods – so, perhaps ‘quite unique’ is under-selling this collection of useful arboricultural teaching aids. Unfortunately, understanding branch junctions is a highly niche technical topic, so it would not, of itself, make a good book, despite the years I have spent researching this topic. At best, it makes a chapter in a book about trees – perhaps an interesting chapter if pitched well. I have batched up my ToT images here as a set of ‘lessons’ that can be learnt by watching branch junctions changing over time – you’ll have to judge for yourself whether they are interesting – or not. 

By the way, there are a lot of terms that can be used for the point where two or more branches conjoin in a tree: for example, quite a few Americans will use the term ‘crotch’ for a branch junction – and that’s really going to prove unacceptable to most readers of this article, as who wants to look at images described as “over-wet crotches”, “diseased crotches”, “crotches that have split open” etc… You would be very much in danger of landing on unwholesome web pages if you ‘googled’ such terms! To be fair, though, when I ran a ‘fork workshop’ in New Zealand in 2017, I decided to rename the session as ‘Crotch Clinic’. You might be surprised how many NZ males turned up to my crotch clinic on a Saturday morning! Along similar lines, a considerable percentage of my “professional” life has been reliant on making puns based on the word ‘fork’. Again, it’s easy to take this too far – too forking easily, in my experience. So – for this article, I’ve stuck to the term ‘branch junctions’ – less humorous, but safer… 

Lesson One: Thinking in 4D

To interpret branch junctions well, you need the ability to think in 4D. Unfortunately, that has proven difficult for the human mind upon many an occasion but, if you’ve made it this far through this somewhat philosophical article, you may well be capable of it.

We are all familiar with the way that a tree trunk expands in diameter over time – through the process of secondary thickening; easily evidenced in the temperate region by annual growth rings formed in the tree’s wood. But perhaps you haven’t thought so much about secondary thickening of a branch junction: how does that section of the vascular cambium which lies in the axil of a branch junction grow? (or does it grow at all?!) Well, that part of the vascular cambium does still grows outward: so, if your branch junction is orientated so its axil is horizontal, then its increment growth will be vertical: or, in other words, the junction’s bark ridge grows longer as it develops (Fig.s 1 & 2).

Figure 2: Development of a normally formed branch junction in silver birch (Betula pendula Roth). From the markings to the side of the junction apex, you can see the junction apex has increased in height above the ground. It’s not grown by much in this small tree, as it is growing in a restricted soil volume. 

Within the axil of the growing branch junction (without a bark inclusion), under the bark ridge, a section of dense reaction wood (axillary wood) is typically formed. This axillary wood not only develops and improves the tensile strength of branch junctions but, being a reaction wood, it is also capable of ‘posture control’ – of posing the branches in different directions – much more noticeably when the two branches are young and flexible. It is thus a mistake to think that if two arising branches from a young branch junction are set at a tight angle to each other that a bark inclusion will automatically occur: the growth of the axillary wood can become extensive to prevent the two branches ever coming to touch each other. 

Lesson Two: Thinking Again 

Where bark is included in a branch junction, it is common for bulges to form either side of the bark inclusion: within these bulges can be found axillary wood, helping to reinforce such a weakened junction. As proven by Slater & Ennos (2015) (and I reconfirmed this finding in a further experiment recently), a cup-shaped branch junction is a stronger form of branch junction than one with a long seam of included bark – and its greater bending strength comes from the development of these bulges so that they come to cover over much of the included bark (Fig. 3).

Figure 3: I love it when one of my ToT images shows multiple changes. Here, not only has the junction’s bark ridge in this semi-mature common ash grown over 250 mm in the eight years of this time-lapse, but the pruning wound above the junction and the stem wound below the junction have both fully occluded. A growth crack has also appeared to the left in the latter image – a common feature in fast-growing trees where the bark changes to being fissured when mature. 

Previous theories relating to bark-included junctions have been that they inevitably force themselves apart due to internal growth pressures (Mattheck & Breloer, 1994) and that they are always a weak point in a tree: however, the development of cup-shaped branch junctions and how commonly they are found in our trees identifies that most bark inclusions do not split themselves apart by their own growth – and that most cup unions do not fail in the life-time of the tree. When one investigates these previous theories in some depth, you find that they are not backed up by scientific studies nor mechanical testing. I urge all arborists to at least treat with scepticism any theories that have no science behind them. Which means, in the case of branch junctions, we really need to ditch all previous unevidenced theories, apply some common sense and try to set in the correct context any reputable peer-reviewed research. 

Probably the most well-known suggestion related to branch junctions in the UK is one propagated by Mattheck, that if a branch junction has ‘big ears’ then it is more likely to fail (Mattheck, 1998). Is there any scientific evidence behind this suggestion? No. And yet, how many trees have been felled, pruned or braced on the basis of this suggestion in the UK? – probably tens of thousands of trees, if not hundreds of thousands. My research identifies the bulging around branch junctions to be compensatory growth – not a sign of imminent failure (Slater, 2018; Slater, 2020). My ToT images of branch junctions help to confirm this latter interpretation (e.g. Fig.s 4 & 5).

Figure 4: A bulge developing at the side of a bark inclusion in silver birch – with the bulge growing in size and height over eleven years. As the bulge is now larger than it was, does this mean the junction is more likely to fail at this later stage? This suggestion seems improbable as it conflicts with what we know about how trees respond to strain through the process of thigmomorphogenesis. If more wood is added to a branch junction, surely that branch junction, in general, will become stronger, not weaker.

Figure 5: An example of a ‘big bulge’ or ‘big-eared’ bark included junction in a common beech (Fagus sylvatica L.). Over the eleven years of this time-lapse, the bulge has grown slightly, particularly at its top, to reinforce the junction. My most recent research identifies that these bulges reliably predict that there is included bark inside the join – but they do not reliably predict either the bending strength of the junction, nor their likelihood of failure. Unfortunately, a myth has been generated (and then taken as fact) that these bulged junctions are the most dangerous type to be found, even though it has probably taken this tree around thirty to forty years to develop this bulge in compensation for the bark inclusion inside the junction. As is often the case with my ToT images, one of the worst things that could happen to this tree is an arborist with the wrong ideas in her/his head turning up to assess it. If you are concerned about a junction like this, you need evidence to justify that concern: citing old theories that were not based on science is not good justification for any action to this tree – or trees like this one.

Please spread the word: “big ears” are not necessarily a bad sign at a branch junction: ironically, for three decades, we have been advised to take action when finding a branch junction that has, at least to some extent, gone through substantial strengthening growth. We have been targeting branch junctions that were mostly recovering, rather than those most at risk of failure. If we are driven by opinions and suggestions rather than science, such poor advice inevitably will be acted upon by some arborists. We need to question it – we need to think again… 

Lesson 3: Quando? Quando? Quando? 

I have now run a series of over forty ‘fork talks’ around the UK over three years, from August 2016 to August 2019. This was a great experience – especially, I met so many friendly arborists and ‘tree people’. By my count, I’ve trained over 1,200 UK-based arborists in natural bracing. In the first eleven workshops in this series, I made all the delegates sit down at the start and complete a questionnaire about branch junctions. It was a good way for attendees to get into the topic of the workshop – and created some useful data for a paper (Slater, 2019). I’m very grateful to all my 348 respondents, who willingly filled in my questionnaire – partly, at least, because the person sitting next to them was dutifully filling it in: positive peer pressure decidedly drove the very high response rate for this research. 

From that questionnaire data it can be determined that around 21% of attendees believed that bark inclusions failed because of internal growth pressures: and this answer was linked strongly to a belief that the failure of a bark-included branch junction was inevitable. As discussed above, such failure is not inevitable – and the idea that bark inclusions can push themselves apart is, again, one that is not supported by any scientific research. As a scientist, one can ‘never say never’ – there may be some branch junctions in the world that have pushed themselves apart – but, from my research, this seems an unlikely scenario: a fact that I have discussed in detail in my previous Arb Magazine article ‘A Pressing Point’ [DS1] and in my workshop notes (Slater, 2016).  

My ToT images add further evidence against this old idea (Fig.s 6 & 7). If these bark inclusions fail because of internal growth pressures, when does that happen? In three years? In seventy years? Why would it not happen in the first couple of growing seasons after the two branches came to touch each other? Asking this ‘Quando?’ question highlights that this belief lacks any evidence-base, for no-one holding that belief can answer it citing data. The current answer is that, if failure does occur due to internal growth pressures at bark inclusions, it occurs very rarely: it is certainly not a common cause of failure at bark-included junctions in trees – and thus failure of a bark inclusion is also not inevitable.

Figure 6: Time-lapse of a bark inclusion in a wild cherry (Prunus avium L.) over an eleven-year period. The two stems were pressing together when I took the first image: why has the junction not failed from internal growth pressures over this time period, then? If one does the maths, it is easy to calculate that growth of the cambium within this junction would not generate enough pressure to break it apart. In addition, for the most part, the majority of inner bark is asphyxiated when it becomes pressed together in this sort of junction – it dies, applying NO pressure to split the junction apart.

Figure 7: Eight year time-lapse of the fracture surface of a bark-inclusion failure in common oak (Quercus robur L.). Included bark runs for around 450 mm down into the junction, but only the top 50 mm remains alive and growing – which becomes more obvious on the return visit. Once the inner bark is squeezed into a bark inclusion, not only is it likely to die, it has little stimulation to continue growing. Two main stimulating factors for secondary growth in trees are mechanical perturbation and the need to create new sap-carrying tissues. When sections of the inner bark are occluded into the main structure of the tree, these stimulating factors decrease greatly – little to no growth would be ‘the norm’.

 Lesson Four: Crossing the Threshold 

Natural bracing in trees is a threshold concept (Mayer & Land, 2003). Once you see that relationship between natural bracing higher up in a tree and the presence of bark-included junctions formed lower down in the tree, you can pass over the threshold to another way to look at, understand and assess branch junctions – and, once you have that knowledge, you’d never really want to go back to your previous state of ignorance. I’m still waiting, however, for natural bracing to make its way into a pruning standard or a book on tree pruning: it is very important that this concept is incorporated in such documents, as at least two key lessons can be learnt from it: 1) by formative pruning that removes potential natural braces in young trees, one can greatly lessen the number of bark inclusions being formed; and 2) when pruning mature trees, the removal of natural braces can heighten the likelihood of failure of the bark inclusions formed below them unless other remedial actions are also taken.

My ToT images of branch junctions with included bark have often led me to discover, on my return to them, that they were (and still are) naturally braced. Although this is no longer a revelation for me – I have a personal collection of over 2,100 photographs of different trees with natural braces set above bark-included junctions, at time of writing – it’s all still very interesting. I still enjoy, again and again, finding confirmation that natural bracing is strongly associated with weak junctions: and the associated logic: that, if a branch junction is not getting the proper ‘exercise’ it should be getting, one should not expect it to grow to be a strong junction. Figures 8 and 9 show typical examples of revisits to bark inclusions, only to find, now I have crossed the threshold and understand the influence of natural bracing can have on branch junctions, that this influence is self-evident and obvious in many cases.

Figure 8: A severe, non-bulged bark inclusion in a semi-mature beech tree, time-lapsed over eleven years. Note that the apex of the join has again risen upwards over that time. On the revisit, two natural braces were found set above this bark-included junction – both formed by side branches – the one shown probably the more important of the two in terms of restricting the junction’s normal movement. Many of my revisited bark inclusions without any associated bulging have proven to have been naturally braced, as predicted by my research work (Slater, 2018).

Figure 9: The common beech, as a species, is definitely the ‘Queen of Natural Bracing’ – it forms natural braces so often – and very often they persist. This shorter time-lapse over eight years shows a slightly bulged bark inclusion in a mature beech tree: the first image taken in ignorance, as I didn’t know about natural bracing at that time: the last image in this triptych showing one of two natural braces that are set above this branch junction, providing the ‘explanation’ for the junction’s current weak form. As both the bark inclusion and the natural braces are physical entities, one cannot deny the obvious mechanical relationship between these features in a tree like this. It sets a particularly challenging puzzle to the arborist in assessing the tree and determining a course of action, though – especially if the natural braces involved are unlikely to persist as the tree continues to grow and develop (Slater, 2018). 

Lesson Five: Changes Can Happen Very Slowly…

 One of the key outcomes of my branch junction-related research is that, when released from any natural bracing, a bark-included junction tends to bulge and thus compensate for its weakness over time, once it goes back to a ‘normal exercise regime’. Putting together these ToT images has been a revelation, in that there is potentially a very long timescale for a good repair of a bark inclusion in a semi-mature or mature tree: explaining, perhaps, why so many fail. Why does the tree not plunge in many of its resources to quickly address this weak point in its structure? (and make us all so much happier!) Research needs to go further to determine why this return to structural stability is proving so slow for trees like the ones I have been tracking (Fig.s 10-12).

Figure 10: A ten-year time-lapse of a wild cherry which was situated outside the office of my PhD supervisor, Prof. Roland Ennos, in the University of Manchester when I first took a picture of it (ahh… the irony…). The dark bark of the bulge at the base of this bark inclusion suggests there is no associated natural bracing now evident in the tree – and that is the case, should you go to visit this tree. Over the ten years, the bulge has slightly increased in size and height, raising the apex of the join slightly. Why so slow in such a large tree? Perhaps the fact that this tree is situated in the courtyard of a six-storey building may be a factor – giving the tree substantial shelter – and giving its main branch junction less ‘exercise’ than one growing in an open area.

Figure 11: Very little morphological change to this cup-shaped bark inclusion in a mature beech tree: however, this form of bark inclusion tends to trap water and detritus, making it a ‘plant pot in the air’. As can be seen, three plants have taken advantage of this little offering of a growing space, namely a common ash sapling, a holly sapling (Ilex aquifolium L.) and an ivy plant (Hedera helix L.).

Figure 12: An epicormic side branch on a wild cherry, time-lapsed over eleven years. No natural brace evident on my return. Note how the epicormic side branch has become more integrated into the structure of the tree over this period and that a bulge has formed at the apex of the join between these two branches. 

Lesson Six: …And Changes Can Happen Very Quickly 

I have captured quite a large number of bark-inclusion failures with my camera – including many that exhibited evidence of natural bracing, since my findings of 2016. Failure of a branch junction can be sudden, unexpected and dramatic: however, many branch junctions crack under loading but still remain in the tree, like they are waiting to be ‘finished off’ by the next storm. Figure 13 highlights this subtle change from a bark inclusion to a cracked junction.

Figure 13: A time-lapse of a bark-included junction in a mature holm oak (Quercus ilex L.) – it’s subtle, perhaps, in this image, but something major has changed. If you look closely, the junction has cracked – particularly noticeable where the two stems are now more obviously separated at the top of the image. The likelihood of failure has been dramatically increased due to it cracking in this way: one could argue that it has failed as a structure already – but I only count failures in trees that hit other objects or the ground. Not shown in the images here is a significant decay cavity in the rear part of the right-hand stem. When I found this change to this junction, I shared it with the local tree officer, who then shared it with his council team. Whilst the tree officer was considering possible solutions, the in-house tree gang felled the tree, I’m told, which irked the officer somewhat, who may have favoured bracing this park tree as at least a temporary measure, as it sheltered other trees from coastal winds. 

Not many of the trees I have been tracking with bark included junctions have been pruned to deal with the bark inclusion. Figure 14 is an interesting exception, where someone doing some pruning to a line of trees decided to ‘single’ this junction down to just one arising branch. It took me some time to identify that this was the same tree, thirteen years later.

Figure 14: A surprising change in this bark-inclusion in a common ash – as, at some time between the initial image and my later image, thirteen years later, this branch junction has been ‘singled’, so there is only one arising branch. A branch cavity has developed at the cut, which is unfortunately common when cutting one branch off from a codominant junction: best done when the branch is of a small diameter and thus the extent of any decay will be limited. 

Lesson Seven: Failed, but not fallen 

I have also been surprised by the number of cracked branch junctions that I took a picture of over a decade ago – and that are persisting, as found through the ToT process (Fig.s 15-18). In my previous experiments relating to my PhD, the mean annualised ‘factor of safety’ of a cracked branch junction was around 1.0 – meaning that there was around a fifty/fifty chance of it failing in just one year. Perhaps the fact that these persisting cracked junctions are all set within woodland is a key factor in their continuing survival: I certainly cannot recommend leaving a large diameter cracked junction in a tree by a main highway based on my research and personal experience.  

It is interesting in some of these examples, though, to see the rapid growth of axillary wood to ‘stitch back together’ the crack – much more obvious than the slow growths of the bulges of bark inclusions (Fig.s 10-12). Note that this growth of axillary wood only occurs from the point where the vascular cambium straddles the join – so, where the vascular cambium experiences the tensile strain associated with being torn apart at a Y-shaped junction. I hope to keep tracking these junctions to see what proportion of them come to a satisfactory repair over time: I had previously considered that a very unlikely outcome but I’m thinking now, perhaps it isn’t that uncommon in woodland trees, at least.

Figure 15: A split bark-included branch junction in wild cherry, within a shelterbelt (on the leeward side). Over the nine years of this time-lapse, there has been only limited further development of the axillary wood that is forming from the base of the crack upwards. I hope to keep time-lapsing many of these types of cracked forks, to see to what extent they are able to compensate for the initial major defect: I suspect that some in my collection will still end up failing in strong winds – particularly with compounding factors, such as decay ingress into the crack. Would I do nothing if I found this tree next to the A6 trunk road? – No – I would definitely plan some work to this tree if that was the case.

Figure 16: A cracked junction in a semi-mature common ash (again, in a farmland shelterbelt), time-lapsed over nine years. Note the extensive development of axillary wood, starting at the base of the crack and working upwards, in response to the high levels of strain experienced by the cells of the inner bark in the axil of the join. Hopefully this tree will prove somewhat resistant to ash dieback disease and I can get to see continuing repair and recovery from this initial cracking of the branch junction (which had bark included in it).

Figure 17: Very little change to this cracked bark-included junction in a native oak (Quercus x rosacea) in a woodland setting over the last twelve years. If you get a hammer or stick and tap at the area around this junction, it makes a great resonating sound, due to the long crack that lies within the junction and down the main stem of this tree. At a guess, it probably cracked around 25-40 years ago and the bulging around this crack has created a sufficient ‘repair’ for the tree to persist in its current surroundings. Heightened likelihood of failure will occur should this tree lose some of its neighbours from thinning operations or storm damage.

Figure 18: Considerable repair at a cracked branch junction in a young (changing to semi-mature) horse chestnut (Aesculus hippocastanum L.) – where some surrounding branches have been pruned off during this eleven-year period – and the tree’s stem has also suffered damage from bleeding canker (most probably caused by Pseudomonas syringae). It’s still a bark-inclusion but with much of the initial crack occluded by further secondary growth. With that sort of back-history, it’s difficult to predict the junction’s relative strength or factor of safety. Unlike a lot of previous testing that scientists (including myself) have done, there has been no work on previously cracked branch junctions with different levels of ‘repair’, to the best of my knowledge: one would need to test many of such specimens to produce any kind of meaningful model, as each junction will be unique in its morphology and in the factors that have affected its development over time. 

Lesson Eight: No-one Wants to Inspect Over-Wet Diseased Crotches for a Living

 As anyone attending my workshops on natural bracing will be familiar with, some types of natural bracing are pretty permanent features in a tree (e.g. fused branches, entwined stems) and other types are fickle and ephemeral: they can cause the problem (i.e. they make a bark inclusion) then the branch(es) that formed the natural brace get shaded out, die, decay and then disappear from the scene of the crime. Even with the very permanent forms of natural bracing, in a street tree, or a tree that one cares about, it’s still probably best to have prevented any natural bracing occurring in the first place. Why? Well, my view on this is probably affected by where I live: the damp North West of England. Here, it is quite common to find bark inclusions that are weeping out water on a regular basis, and this can often lead to bleeding cankers within the junctions (especially in beech (Fagus sylvatica L.) sycamore (Acer pseudoplatanus L.) and lime (Tilia spp.)). These bleeding cankers are then often followed up by secondary decay and eventual failure. Although it’s quite possible for normally formed branch junctions to suffer from bleeding cankers, I’ve found it far more common for that to occur where there is included bark in the junction (Figs. 19 & 20).

Figure 19: Time-lapse of a bark-inclusion in a mature common beech tree, where a water-run was evident. Unfortunately, this is all-too-common in the NW of England – and, revisiting eight years later, an area of inner bark has died associated with the water-run from this junction (I have panned out the second image so you can more clearly see the extent of the damage below the junction). Frequently, the next stage is the colonisation of decay fungi which will further weaken the junction. Whether naturally-braced or not, a bark inclusion is often still a defect in a tree – and can be a court for disease and subsequent decay – particularly in wetter biomes (like Lancashire!)

Figure 20: A pathway to failure. In this sycamore (Acer pseudoplatanus L.), a fairly-well compensated bark inclusion has unfortunately fallen victim to a bleeding canker, which then exposed a lot of the sapwood to invasion by fungal decay agents. The junction failed eleven years later, with obvious white rot associated with a large bracket of Ganoderma australe evident just below the junction. The limb that failed was hung-up in another tree in this wooded area, until the whole tree was felled – hence how I got the latter image of the diseased and decayed junction splitting apart but the limb not falling to the ground.

Conclusions 

There are many positives to be gained by throwing out the old unevidenced theories on branch junctions and coming to a more rational point of view. I have found over fifty tree specimens with natural bracing present in association with bark-included junctions at Myerscough College, a selection of which I use to teach our students – and the occasional interested visitor. Whatever one might think of this ‘new theory’ of natural bracing in general, for these fifty trees at Myerscough – and the other 2,000+ I have found over the last few months, it’s not a theory but a reality: there’s a natural brace (or more than one) set above a bark inclusion in each of these trees – which sets a particular challenge in thinking out what to do with each of them for the arborist asked to assess them (Fig. 9). Probably the biggest positive is that these shared research findings really give powerful knowledge to the arborist in such situations, clearing away all that previous foggy conjecture, some of which I have touched upon in this article. I sincerely hope that soon our pruning standards and pruning books will get updated in the light of this research. 

From my somewhat unique collection of branch junctions changing over time, I’m still learning some lessons that the trees are teaching: I will reiterate the two that I found rather surprising: a) that an unrestricted bark inclusion in a mature tree only subtly changes its shape and form over a decade or so (rather than a more rapid ‘repair’) and b) that some cracked junctions can persist for over a decade or so – in part, probably, as I have recorded them on woodland trees in sheltered locations – and in part, probably as the tree’s response growth to a crack will improve the low factor of safety from ≈ 1.0 to ≈ 1.5 quite quickly – meaning it needs a degree of abnormal loading to fail it. 

Yours forkfully, Duncan.

 *** THIS ARTICLE FIRST APPEARED IN THE SPRING EDITION OF THE ARB MAGAZINE, PUBLISHED BY THE ARBORICULTURAL ASSOCIATION *** 

?References

 Mattheck C and Breloer H (1994) The Body Language of Trees: A Handbook for Failure Analysis; London: TSO.

Mattheck C (1998) Design in Nature: Learning from trees. Berlin: Springer-Verlag.

Meyer J H F and Land R (2003) Threshold concepts and troublesome knowledge – linkages to ways of thinking and practising; in Rust C (Ed.) Improving Student Learning – Ten Years On. Oxford: OCSLD.

Slater D and Ennos A R (2015) The level of occlusion of included bark affects the strength of bifurcations in hazel (Corylus avellana L.); Arboriculture and Urban Forestry 41, 194-207.

Slater D (2016) Assessment of Tree Forks; Course notes published by the Arboricultural Association.

Slater D (2018) Natural bracing in trees: Management recommendations; Arboricultural Journal 40 (2), 106-133.

Slater D (2019) Current opinion within the UK arboricultural industry on the management of bark-included junctions in trees; Arboricultural Journal 41 (1), 10-34.

Slater D (2020) The mechanical effects of bulges developed around bark-included branch junctions of hazel (Corylus avellana L.) and other trees; submitted to Trees: Structure & Function, 23.08.2019.