Jump to content

Log in or register to remove this advert

  • Cassian Humphreys

    Article: Tree Welding & Natural Grafts - The Magic of Myrtaceae



    Tree Welding - Natural Grafts a form of Self-Optimisation of Woody Plants


    Anyone that knows my work knows that study of tree mechanics of SE Qld trees is an ongoing personal/business interest and passion. I have been observing and gathering data on natural grafts for four years with a view to this article. According to Science natural grafting is common place with lower plants (non vascular plants), this project is a study of natural grafts amongst higher plants (with study of the family Myrtaceae) which according to my reviewer has limited scientific record.




    This is my second article written subject to review by a professional associate, the reviewer for this article is Scott McKenzie a Sydney based consulting Botanist BSC-Biological, ADV Cert Hort (Urban). Scott like me is a keen Shigoist and is a very seasoned consultant and report writer for a Sydney based arboricultural company, he is an experienced Resistograph user and is a rare operator in that as well as being a qualified horticulturist he is a scientist, making his report writing skills and ability to quantify scientific information be considerably beyond the scope of an arborist. Scott has been reviewing my Tree Culture Course Manual since its conception in 2004, his encouragement and support has been integral to my personal growth in the area of science and as an arborist.


    In honour of my mentor (Shigo) and reason for developing a scientific brain readers may notice my articles are becoming more scientific in their content, though I do not wish to alienate some readers this is necessary and is evidence of an arborist following the fathers (Shigo) call to develop brain to match the muscle, like Shigo I encourage all arborists to do the same.



    As a result of my study of the body language of trees and the science of Shigo and Mattheck I believe that the genuses Eucalyptus and Corymbia are doing some unusual things with their bodies (in reference to VTA), unusual in the context of current science, but normal in the context of the evolution of biological systems or plants. For the sake of this article I collectively refer to these genuses

    as `gums’.


    Grafting of vegetation as a horticultural craft was born as a result of natural grafts observed in species such as Citrus and Prunus.

    I have seen little data on natural grafts and after a limited response following my posting photographs of a natural graft (between two different species of Eucalypt) on a world wide forum I was inspired to write this article and to also introduce a rather astonishing tree to our profession, a tree nick-named the Lionel tree.


    We know that trees are reactive generating organisms that are products of their environment, made of elements and designed by the elements.

    As self optimising organisms trees adapt into a range of environments from forest giant to bonsai.


    Science (Shigo 91/Mattheck 94) has documented a range of symptoms that quantify a trees mechanical status and are indicators of reaction wood development.

    Mechanically successful trees with or without mechanical constraints are successful because of their wood production or `muscle’.

    Reaction wood production with symptoms such as ribs, ropes, pillars, growth striations, increment strips, welds, feet and grafts are all created by trees in response to their distribution of load bearing stress (compressive and tensile) and are evidence of woody muscle.


    This article and the photographic evidence within is an example of what Australian trees are doing within their structures – or for the simplicity of this article to be referred to as their bodies. I see natural grafts as being evidence of self optimisation or reaction wood development and believe that arborists and tree managers should be considering working with trees as a means to strengthen mechanical constraints via encouraging natural grafting to take place, this concept is to be covered further in a future article.


    Trees with crown structures which cross over and touch have the capacity to fuse together, this occurs as a result of the generation of incremental layers of sapwood.

    Generally (amongst arborists) we see reaction wood as a being reactive wood tissue formed by the cambium in response to load bearing stress (usually in ribs, ropes or pillars). I believe that trees are reactive with triggers other than just load as a catalyst with grafting being a probable example.


    Before we can comprehend grafts as a form of optimisation we must first consider and comprehend wood.




        1.  Cambium layer – trunk meristem - cell generator producing xylem (on the inside) and phloem cells (on the outside), the cambium at the centre of the symplast also acts as the chemical `brain’ of the tree.


        2.  Phloem & Xylem – phloem cells transport simple sugars (glucose) as carbohydrate, with the woody xylem cells transporting water and elements as well as make up tree structure (deal with tensile/compressive load) via thickened cell walls.


        3.  Cork cambium – suberises outer bark.


        4.  Bark - outer layer of spent suberised phloem cells for waterproofing and protection.


        5.  Wood is formed by the cambium layer as sapwood in layers or increments, as sapwood cells age they lignify (cell walls become infused with lignin), the primary role of new sapwood or the most current tree or NOW tree (to quote Cowan – ISAAC Brisbane Conference 2008) is for transporting water/elements. As sapwood ages its primary role is support though it still also pipes water The now tree as the most recent layer of sapwood is laid down by the cambium in response to growth (the synthesis of carbohydrate, water & elements) and the latest environmental loading.


        6.  With sequential growth increments, time, age and wood differentiation sapwood becomes heartwood (according to Shigo there are a number of types ofhardwood). As trees reach an advanced age they become hindered by previous increments (past trees) which detract from the mechanics of the most recently designed tree - the now tree. It is widely understood that a cylinder (assuming it has sufficient wall thickness) is stronger than a solid. Though heartwood still plays a functional role (transporting water/minerals) as trees become of an advanced age it is beneficial for them to shed their ballast. Trees achieve this via forming columns of decay through their heartwood (as with mychorrizal nutrient absorption this is achieved via fungal association).


        7.  Medullary rays are responsible for conveying information (from the cambium layer) and substances laterally through wood, wood rays also mechanically strengthen wood by helping bind wood increments together.


    There are many examples of trees as generators growing `their bodies’ over other bodies such as other woody plants, fence palings and even motorbikes (Figs 2-7).




    In the course of my career having worked in Britain, Germany and Australia as an arborist I have observed only one genus of tree that can achieve the act of `regeneration’ (a play on words - trees are generators and do not regenerate) over their felled stumps via a stump sprout – Eucalyptus.




    I have observed forest red gum (E. tereticornis), white mahogany (E. umbra), grey gum (E. propinqua), Gum topped box E, moluccana and tallowood (E. microcorys), though suspect other species of Eucalypt and species of Corymbia capable of stump sprout generation. In both these trees you can see the original tree (stump) in the body of the tree.


    It is a natural part of a trees growth to use whatever is available in its environment to succeed biochemically and biomechanically. Like a modern defence force has a range of strategies available to deal with conflict, in a sense a tree has a range of strategies to maintain a state of optimisation, of which I see grafting as being one, however in reality trees have one strategy that is they grow reaction wood, I see natural grafting as a form of adapted reaction wood development.

    Because trees (like any life form) are evolving organisms some species and some individuals (as with plant genera and plant families) are more successful at grafting than others, and the family Myrtaceae with the Genuses Corymbia and Eucalyptus appears to be leading the way.


    Grafting occurs as a portion of a tree such as a branch crosses over another stem or branch then touches and abrades `itself’ (its neighbour or neighbouring portion of itself) causing wounding both to the thickening limb and the thickening stem.

    As these portions continue to thicken (make incremental growth) and abrade sapwood to sapwood they squeak and rub until (due to compression between their surfaces) they can move no more. The cambium layers of the branch and stem then generate sufficient woundwood over both wounds to fuse and grow together. In the case of a successful graft the fused parties compartmentalise their wounds and share photosynthate. And at a more advanced stage of wood generation the grafted branch and stem also share mechanical load.




    Fig 1 and Fig 12 are examples of Corymbia citriodor a maculata (spotted gums) which are at advanced stages of grafting where they are now sharing mechanical load as well as have the ability to share common properties such as water and nutrients.


    Generative optimisation of rootplates – Buttress and root grafts


    It is widely recognised by science, the horticultural and arboricultural professions that root grafting occurs between trees. Root grafting between trees is a brilliant means for trees to greatly increase their surface area for nutrient absorption. Woody roots are also mechanical optimisers and are known to use rocks as underground bollards (Mattheck 04) by growing slings of wood around them. Grafting below ground likewise plays a significant mechanical role.




    Fig 13 - Water Gum Tristaniopsis laurina – ancient riverine forest north of Cooktown Qld


    Thanks to a presentation (ISAAC Brisbane 08) run by UK arborist/researcher/ecologist Andrew Cowan of Arbor Ecology I had my attention bought to the phenomenon of the development of the `cone’ as a means for trees to increase their mechanical footprint.

    This occurs when trees reaching an advanced age generate interconnecting wood tissue over and between buttresses and roots – this involves grafting and is typical of the root plates of advanced age Brush Box (Lophostemon confertus) and trees of the genera Ficus.


    Fig 13 – has grafted a curtain of roots into one mass of wood, generating a huge footprint to better manage compressive loading, this tree like its neighbours is growing in boulders in

    a dry creek bed.




    Figs 14 and 15 are of a spotted gum (C.citriodora maculata) with an optimal shear root plate (soil controlled). Lateral roots where entirely severed on three sides when the road was graded. The tree generated a new root system over its damaged root plate which over time has grafted together to form a woody cone which is still growing. This tree is a good 35m tall with a spread of 20m

    growing in a very exposed area and is a testament to the mechanical optimisation of the gum.



    A naturally grafted prop on a multi-trunked Moreton bay ash Corymbia tessellaris at Mackay Botanical Gardens




    I see all successful grafts as astonishing acts of self optimisation, this example is brilliant. The stem (Fig 16) marked with a red arrow was a leader which became grafted to its neighbouring stem (blue arrow) its growing point was then removed with the wound becoming occluded (closed following incremental generation of wound wood).


    The stem though having lost its crown still receives energy as it is now supplied by its neighbouring stem whilst drawing on its own supply of water and elements. This may or may not be seen as a form of optimisation by us, but it is a classic example of a tree using its resources to maintain a state of sustainability (bio-chemically and bio-mechanically). When we consider Shigo’s tree as a giant oscillating pump (or his model for dynamic equilibrium) then we have to be impressed with this tree.


    This example is living evidence of mechanical adaptation with the growth of a natural graft, in the course of my work I have been finding and seeking out trees which use self grafting as a means to strengthen mechanical constraints (I prefer this term to the term `defect’). My most significant finds so far (like rolling ribs – turnbuckles/horseshoes) are gums which have formed grafts directly above V forks. I refer to these grafts as natural bridge grafts.



    Natural bridge graft unions above V forks – assisting gum trees to stabilise, compartmentalise and transform V forks into U forks




    These samples of Corymbia (species uncertain from Rockhampton), Lophostemon (confertus – brush box – Sunshine Coast) and Melaleuca (bracteata – the black tea tree - Moggill) all have grown/are growing grafts within 1-2m above V forks.




    The Melaleuca (both sides of V fork) with a natural bridge graft on one side and a rib on the other.



    Natural Bridge graft above a V fork on a Sydney Blue gum Eucalyptus saligna Ocean View Qld






    Scribbley gum - Eucalyptus racemosa - with evidence of a natural bridge graft in process directly above a V fork.







    Forest red gum Eucalyptus tereticornis – (Rocksburg Qld) with evidence of a natural bridge graft above a fork, in this case a U fork with the graft union (on one side) having been negated by parrot damage.

    Damage to young mechanically robust trees caused by parrots is a common side effect to loss of habitat caused by development and removal of old habitat trees.



    Psidium guajava – Yellow Guava




    The Yellow Guava - native to Central America and like gum trees being of the Myrtaceae family likewise commonly self grafts. This exampleis interesting as following a branch break out failure (yellow arrow – fig 32) on a cluster (whorl) of three branches the wood of the remaining two subsided (bowing open – red ar

    rows – Figs 31&32) opening along an inclusion. The remaining stems then grafted together (blue arrow – Fig 31).




    Though subject to future decay the tree is generating wound wood about the past failure and has such a limited lever arm is very unlikely to fail, the graft has saved the day.



    Eucalyptus Crebra Red Iron Bark Brookfield Brisbane


    Co-dominant trunks with a shear crack and evidence of self optimisation


    Crown height – 38m

    Crown spread – 17.5m N to S

                               14.6m E to W

    Trunk DBH – 120 cm

    Trunks width 80 cm and 60 cm diameters


    Crack opened up by 18cm on top and 6cm at ground level

    Crack length 150 cm

    Crack depth 70 cm


    Based on my study of this tree I estimate the crack has been present for 20 to 30 years




    Inside of open shear crack (top of split fork) wound wood has filled the top of the crack with wood tissue equivalent to a weld.



    Opposite side of Red Iron Bark




    This is a classic example of a tree which has managed a significant mechanical constraint for years via production of reaction wood which has grown in the form of ropes, rib and weld. To my mind a tree like this has to have arborists everywhere re-consider the nature of trees and tree mechanics in respect to human assumption of trees and their structures.


    Generally we do what we are taught to, in a rapidly evolving profession where tree science is itself an evolving precept we must

    always challenge what we are told. We as arborists must be open to other possibilities when it comes to managing and interpreting trees with or without `defects’. Until we have the science to match our muscle (to quote Shigo) we are incapable of being potent as tree care professionals as we are unable to quantify what trees are actually doing.

    Based on established arboricultural opinion this tree is to be condemned, based on its body language (and history) I stand by this tree as a tree that is likely to stand for many more years to come.




    Utility arborist and colleague John Bevelander giving dimension to the wood rib.



    Natural graft between two different species of Eucalypt – Bark samples left & Right




    I have yet to climb the trees to get closer photographs, or identify the species (possibly different genuses) though can confirm based on VTA that this is a graft union meaning shared biochemistry and mechanical load.


    This example is the first I have witnessed of a natural graft between two different species of tree. Found by utility vegetation manager and arborist Matthew Palmer these

    trees are an exceptionally rare discovery. Following posting of Fig 39 on a world wide forum for arborist’s no-one was in the position to acknowledge a comparison; this must be seen as a very unusual scientific phenomenon.





    Introducing the Lionel Tree – Natural Graft between two different species of Corymbia


    Most certainly the most astonishing tree I have ever observed - the Lionel tree is a natural graft on Corymbia citriodora maculata (spotted gum) of Corymbia intermedia (Pink bloodwood)

    I found this tree whilst working away. I made the request to the ETS crew I was training to keep an eye out for any unusual trees; never in my widest dreams could I have conceived of this tree. The tree is now named after the fencing contractor who found it.




    Standing on a bloodwood branch attached to a spotted gum




    Above and below shots of the bloodwood limb – me up tree and ETS colleague Greg Dale taking picture below








    Spotted gum (C. citriodora maculata) attempting to graft itself to a white mahogany (E. carnea) on 3 different locations




    At three different locations on the white mahogany the spotted gum appears to be attempting to graft itself, the spotted gum is certainly projecting (generating) large portions of reactive wood at or around the smaller

    tree. One could almost state it looks like the spotted gum is attempting to eat its neighbour.




    Location 1 – From ground level to approximately 1.5m up the stem - the base of the Spotted Gum is generating itself around the trunk of the White Mahogany.
    North western side – location 1
    Location 2 – From approximately 4m up the stem a major projection of reaction wood (akin to a rib) has grown out and around the White Mahogany (Figs 51-53)
    Location 3 - Approximately 9m up the stem a projection of reaction wood (akin to a rib) is growing out and around the base of a limb – Fig 54.

    Occluded projections (stubs?) of this nature are not uncommon on spotted gum but are difficult to quantify, I have seen mistletoe be occluded in this way – Fig 55



    Conclusion –

    This article is based on careful consideration of what natural grafts appear to be doing based on a developed VTA perspective, the question I pose is `are the generation of natural grafts/welds in response to load bearing stress and therefore reactive – evidence of reaction wood development and mechanical optimisation, or is this some form of opportunism’? Personally I believe in the former as when we consider all the other parts and processes trees are running to succeed then grafting as a part of the self optimisation process is very likely.

    Considering the science of trees Mattheck has discussed natural grafts in the context of tree mechanics. I believe his work (like Shigo’s) has not yet been fully integrated by the Australian profession of arborists as a collective, that VTA is yet to be integrated in Australia as a scientific discipline. However I have been developing an Australian context for VTA for myself (and my business) and see rolling ribs (Turnbuckles and Horse shoes – Edition 2 AAA P36 August/September edition 08) and natural grafts as examples of the self optimisation process in trees with gum trees (and possibly the family Myrtaceae) leading the evolutionary way. This kind of data collection is generally beyond the scope of current arboricultural training or (available to our profession) research, as what trees are doing `with their bodies’ does not yet appear to attract a dollar value. However I hope that a university some where considers this article and my contribution to data collection on our native gum trees as what we have is unique to Australia and that has to be worth something.


    With all trees in this article being of the same family I pay homage to the Magic of Myrtaceae with yet another half Turnbuckle from Australian leading Arboricultural personality and trainer – Brett Hamlin and his latest contribution (Fig 56) a cross section of a V fork with a linear rib and a rolling rib. This one the first of a new genus with a Broad-leaved Paperbark – Melaleuca quinquenervia, I believe that we all have something tremendous to learn from this family of plants.





    • Like 3

      Report Article

    User Feedback

    Recommended Comments

    Great Post! I am a big fan of fusing though I have never come across two species inosculating. Anyway, I thought I would share these pics of not fusing but butt regrowth which makes me think these guys can probably fuse quite well. The original stem is 30-40 cms in dia. All the raised bits are alive and the rest is sand: Bombax ceiba https://photos.app.goo.gl/jP2C2mR9zS8d3CHP2

    Share this comment

    Link to comment
    Share on other sites


    are the generation of natural grafts/welds in response to load bearing stress and therefore reactive – evidence of reaction wood development and mechanical optimisation, or is this some form ofopportunism

    Over 40 years working with trees and exploring forests, most examples of natural grafts I saw fit neither of those suggestions.


    When branches get close and trunks rub, it's understandable that cambium area tissue from both will merge into one mass cambium, producing a common mass of wood and bark. 


    It's interesting that the question I quoted, is so far apart from the simplicity of what's happening. Grafts are generally not a common function like growing roots for anchor and nutrition or leaves growing to face the light. Grafts are more like coincidence, and not necessity.


    It's not impossible for a graft to reinforce a weak tree by bonding at a higher point. But it would be by chance and not by purpose or opportunity. To stretch for much beyond this simplicity would be like the articles the past couple years where writers were suggesting albino redwoods were saviors of the forest (when the opposite it true).



    Edited by mdvaden

    Share this comment

    Link to comment
    Share on other sites

    Create an account or sign in to comment

    You need to be a member in order to leave a comment

    Create an account

    Sign up for a new account in our community. It's easy!

    Register a new account

    Sign in

    Already have an account? Sign in here.

    Sign In Now

  • Featured Adverts

  • Topics

  • Blog Entries


Arbtalk.co.uk is a hub for the arboriculture industry in the UK.  
If you're just starting out and you need business, equipment, tech or training support you're in the right place.  If you've done it, made it, got a van load of oily t-shirts and have decided to give something back by sharing your knowledge or wisdom,  then you're welcome too.
If you would like to contribute to making this industry more effective and safe then welcome.
Just like a living tree, it'll always be a work in progress.
Please have a look around, sign up, share and contribute the best you have.

See you inside.

The Arbtalk Team

Follow us

  • Create New...

Important Information

We have placed cookies on your device to help make this website better. You can adjust your cookie settings, otherwise we'll assume you're okay to continue.