One of the greatest aspects of bonsai for me is its’ diversity–not to be understated, in both practitioner and hobbyist but of course in the great variety of species across coniferous and deciduous trees. While a conifer may have more individual character, think yamadori with old bark or deadwood, a deciduous tree offers a soft elegance paired with the seasonality of spring, summer, fall, and winter. Deciduous trees offer a great way to experience time in the present moment and brings a sense of vigor in the garden.

To preface this entire discussion I would like to start with an old trident maple whose branches I’ve all cut off. At first glance, it seems almost sacrilegious to remove many old branches and the general shape of the tree but I humbly request to impede judgement until reading through this article!


*Large trident maple, appx 10-12 in base before cut backs


*After primary branch cut backs

For starters lets define the qualities of deciduous trees. We know that they change with the seasons (excluding broadleaf evergreens and tropicals) and notably will be laid bare in the winter. Although deciduous trees can look great year round, it is not uncommon or rather frequently preferred that they are shown without foliage. The fine branching and soft silhouette can be appreciated in this form and tells the story of the tree in how the branches were grown and developed. With this, we can establish a premise that good deciduous trees are valued for their branching rather then just a silhouette. This brings us to the second point that good quality branching are built overtime rather then styled in just one iteration which would place more emphasis on the latter aspect.

To start, I’d like to discuss the concept of branch building. Branches are built through cut backs. Cut backs enable:
1) Increase in ramification from elongation of lateral or back bud
2) As well as directional branch movement which we will discuss later

Ramifying Trees Through Cut Back

Branch building in it’s simplest form is an increase of ramification. The more opportunities of cut backs we have the more intervals of bifurcation that is possible. This is highly related to internode size or distance from bud to bud. It is no mystery why trees with small branch internode capability is sought after for bonsai.


* Early stage trident maple I previously shared, undergoing the branch building process

A way to think about this concept is to consider if we have (arbitrarily) 10 inches of canopy radius we are trying to build branches in. Although bonsai may, in theory, grow indefinitely bigger there will be some proportion relative to the trunk size, height, and style of tree that can look aesthetically optimal. So for this example, lets just state that 10 inches is this target size.

If each cut back interval is 1 inch, we have theoretically 10 times to double our branching. There are physical limitations however, relating to tree health on how many buds and branches a tree can reliably support. We are omitting that in this example.

If we double our branching 10 times, by the time we hit our target silhouette we would have over 1000 branches! What if either by health, the tree could not ramify and only elongate or the cut back process was started further out to the silhouette? Say we only did this cut back process 5 times. Upon hitting the target silhouette there would be only 32 branches or a 3200% difference in silhouette density! The above picture is a graphical example of this concept, basically to show that the difference in branching is exponential! Certainly, say for up to 20 perfectly bifurcating cut backs, there are probably no deciduous bonsai in existence with 100,000 branches. But 10,000 branches? Well lets see:


*Above photo from Bill Valavanis’s blog from the 2017 BCI exhibition in Taiwan

Not all bonsai necessarily need such a degree of density, but as a rule of thumb branches help provide scale to our bonsai. Just as we consider a tree in nature, which may in fact have thousands of branches, to maintain the relative proportion as bonsai the number of branches must remain or at least be similarly the same. This is so when we look at our bonsai, they both evoke and have the presence of a larger tree.

My bonsai friend and a fellow professional, Tony Bebb, took these photos in Japan at the Gomangoku exhibition showing an amazing degree of ramification on a trident maple. Note that the style of this tree exhibits slightly more downward branching which “natural” deciduous purists may not appreciate, but to achieve that density requires skill, consistency in application, and many many years of dedication as well as forethought.

To contrast this method of branch development, I will go over the most commonly seen branch building practice I’ve observed. Very frequently, practitioners will take long and often young primary branches and wire it out to silhouette or their envisioned tree size. As that branch elongates and breaks the silhouette shape the tips will be cut back down to the desired size–the assumption is that then latent back buds elongate which adds ramification to the tree. This works, however there are significant draw backs. Please see the following graphics for context (please excuse the poor graphic design skills, which was limited to ms paint and a laptop touchpad).

Methodology of poor branch building:

Stage 1:
A primary line is wired out to silhouette–it may have some degree of secondary or tertiary branching as well. As terminal buds elongate and break past desired design profile they are cut to shape

Stage 2:
After strong terminal tip growth is removed, latent back buds gain terminal dominance and begin to elongate. The tree gains more secondary and tertiary ramification. Process can be repeated.

Stage 3:
After every iteration of the outer silhouette tip cut back the original primary branch line, which all future ramification is derived from, will get slightly thicker with each round of growth. A way to consider this concept is how triplets or whorls of branches from one junction will cause knuckling and thickness in one spot. We can consider this to be “knuckling” of the primary line but dispersed across the length


At first glance, people may ask what is wrong with this method of branch building? For starters by cutting back only on the very tip, we limit the capacity as well as physical space for ramification. The original primary line occupies a lot of silhouette volume and (referring to the bifurcation graph) with less opportunities for cut back across a growing space we will have less branching. Second, by repeatedly cutting off terminal growth (while maintaining the same silhouette size), the branch tips will get increasingly thicker and thicker. There is a lack of taper along the branch line and softness at the branch tip which diminishes the elegance of our tree.

Invariably, with enough repetition of this process, it becomes necessary that the branches must be cut back hard to finer branch lines.


*Cutting back thick branch tips to fine branches will create taper, canopy softness and create space for more bifurcations.

Although the same result could have been achieved more quickly if the initial cut back process began smaller than the target silhouette size accommodating space to grow the canopy bigger. This is to say that deciduous trees are NOT bonsai we can make in one iteration and certainly not trees that can only be wired to shape.

It is necessary to start your silhouette smaller and build your tree to a larger silhouette. Especially in an early stage of development do not try to rush the outer silhouette volume.

Creating Branch Movement Through Cut Back

Now rewinding back a lot, I mentioned that there is a second advantage to building branches via cut back–that is, branch movement! All deciduous and broadleaf trees exhibit lateral bud placement in either a symmetric or alternating fashion. Upon cut back, elongation of this lateral bud guarantees a change of direction. The more cut backs across a built branch length, the greater capacity for movement. To provide a foil to this statement, why don’t we just wire movement in the branch instead of cut and growing, node by node, to produce movement? Well it is not that you can’t however the nature of the branch movement is different.

When wiring a fully lignified branch, even for the highest degree of bending, the movement will be of some ratio of a curve. There are some physical limitations that will result in most sharp bends cracking a branch. Whereas movement created through cutback, irrespective of the emergent angle, can have sharp lines. This seems counterintuitive, that sharp aggressive branch lines would be deemed more appealing then flowing curves. But as the branch thickens, the sharp edges soften and the branch retains the overall greater directional density then if built by wiring.

A more apt way of depicting this is through the penjing model. Primary branch lines have more aggressive movement and directional changes which are derived from “clip and growing.” Although cutback can result in increase in branch ramification they are not mutually inclusive. That is, upon cutting a branch we can thin shoots and direct the new growth into a line of our choice. The subsequent isolation of energy into one line will further thicken the branch in that direction. It is then to the artists discretion on how thick to let it run, successive cut back intervals, and at which point to start focusing on ramification.

Hwa Fong Exihibition 2022:

Here is work (a chinese elm) from Zhao Che, the penjing curator of the Huntington Gardens. He does excellent work in the “lignan” style. Notice the aggressive movement in the primary lines. They hold much more visual weight compared to some Japanese counterparts.

To contrast, a zelkova from Kokufu (2019):

I chose to contrast these two trees as they exhibit similarly high degrees of ramification but exhibit difference in the primary branch movement and net directionality of the tree. Visual weight and focus can be created through movement. We can consider the movement to be anchors in the primary lines. Primary lines are to the shape of a tree as framing is to a house. Without it, the design will collapse.

This is not to say one method is better than another, rather these are aspects in branch building that can influence how the tree looks. From my cultural experiences and understanding, in Japan there is this essence of quiet beauty. Rather than bold statement or expressions, beauty is in the subtleties. There is a criticism of bonsai in Japan as being too craft derived and through this pursuit of perfection, producing highly ramified “ultra-manicured” trees. But perhaps its this extreme juxtaposition of perfection as a craft against an organic, living, and forever changing life form that makes bonsai in the Japanese scope so widely appreciated. Think about wabi-sabi where there is this appreciation of transience and imperfection. Or in the Japanese tea ceremony where more “ordinary” everyday ceramic vessels began being employed instead of these highly valuable antique Chinese ceramics. There is a subtle beauty to be appreciated through this approach and in a sense this feeling of naturalness and belonging of the tree–through a very manipulative and unnatural means! I talked more about my general bonsai philosophy in my last post linked here: Good Bonsai Takes Time


*Ishitsuki momiji from Kouka-en, apprenticeship circa 2019

In many penjing trees stylistically there is more creativity which is observed in the primary lines and shapes. This may also relate to the study of calligraphy and by extension the emphasis and beauty in lines in the tree. The forms are more diverse, but technically no bit inferior to Japanese bonsai. There are perhaps more cultural layers and influences, but my knowledge on this end is limited. If any of my readers know a bit on the history and cultural influences of penjing I’d love to hear your thoughts!

Below photos posted from “Black Scissors Bonsai.” Unfortunately the artist names were not shared, if you know the respective owners of the work please let me know so I can credit them.

For me, this is where I believe the potential lies and creativity of the artist in how they apply the craft. Note that I make the explicit distinction of application of craft to create art. Bonsai, as a growing and changing medium that is bound to rules of horticulture, will not achieve its pinnacle being approached as an static art. What kind of lines, movement, or shape do we want? Aggressive vs soft primary lines? Thickness? When do they start tapering? Density or ramification? Silhouette softness? From our own value set, cultural, or natural environmental influences we can take our inspiration into reality.

So oh yeah, why did I cut the tree? Well in context of the previous discussion, that trident exhibited coarse branch nodes which were both thick and long. The lines are visually not as interesting and the capacity for ramification is decreased.


*Branch internode length vary from 1-3 inches. For a 12 inch canopy radius I would only have 4-8 bifurcation intervals.

If I try to maintain the tree by only cutting the tips, it may marginally increase in density but the branch ends will grow even coarser and the tree quality will actually decrease. It can be argued that there was already a general tree form in place, but in my pursuit of bonsai as a professional I am constantly seeking to increase my skill and make trees at the highest level possible. Restarting the branches shorter will enable more space for increase in ramification as well as cut derived directional movement.


Its alive and budding! I opted to not repot the tree while doing these heavy cut backs. I acquired the tree out of leaf, but from not knowing the care last season and seeing dead branch tips implied that it may have grown weakly. To maximize backbud potential I chose to leave all the roots and will do root work next season after growing it strongly. The tree will be allowed to grow freely to generate strength and for me to evaluate the branching I have. It may be necessary to do some thread grafts and cut back further later. Future updates to come as I develop this tree!

There is still so much more to be said, and managing branching and internode length is not always as simple as just cutting a branch. It will relate to the species, stage of development, and care. Additionally one should ensure a branch is sufficiently thick enough before cutting. But I hope this article will get you thinking next time you work on your deciduous trees, and for you to act with greater intentionality towards your desired end goal!