Winter Pruning Cordon Apple Trees, 'Short Spur' Method

Last week I spent a very happy hour or so pruning a dozen cordon-grown apple trees down on our allotment. Aside from just wanting to say how satisfying a job it was – pruning fruit trees is my all-time favourite horticultural operation, although grafting new fruit trees is a very close second – I thought it might be useful to describe some of the decision-making processes involved in pruning these particular trees, varying as they do in size and complexity.

Please note, as per my stepover-trained apple trees, none of these cordons were pruned in the summer (or, indeed, in any of the past four summers), for the same reasons explained in my articles on winter pruning stepovers, years one, two and three. The very short version of which would be:

These trees are growing on an allotment and have no deliberate aesthetic purpose, so they don’t need to be kept neat and tidy through the year. They’re all about maximising the production of good fruit in a small space.
Keeping as many leaves as possible on the tree increases the potential for photosynthesis – carbohydrate production, the tree’s source of energy – which then helps to develop and ripen the apples and also allows for late-season secondary growth; the thickening and strengthening of stems¹.
Keeping as many leaves as possible into autumn also increases the rate of dormant-season nitrogen storage within the tree. In autumn, the chlorophyll in the leaves is broken down – which is why they turn brown – and reclaimed nitrogen is stored in the tree’s roots. This stored nitrogen is then readily available to the tree the following spring, helping it to break fruit buds into blossom and develop strong early leaves; kick-starting the new season’s growth and development.

Cordon Pruning Basics

A cordon-trained fruit tree, whether grown as a vertical cordon, a horizontal cordon (a.k.a. ‘stepover’) or at an angle (oblique cordon) is grown as a column, or collection of columns, of one or more main stems, with fruiting spurs growing out along the stem(s) to carry the fruit. It’s a highly artificial form of tree growth that is impossible to maintain without regular intervention on the part of the orchardist – mainly via pruning, but also by tying-in stems to a support structure – to control the general shape of the tree. With time, practice, expertise and no small amount of good fortune with regard to bud-break and stem growth, it’s possible to create some highly artistic and rather wonderful-looking tree shapes:

Complex ‘compound cordon’ tree shapes, via Paul Champagnat (trans. N. B. Bagenal), *The Pruning of Fruit Trees* (1954)

The height of the cordon tree is decided by the orchardist but is usually determined the height of the support structure, or the wall, against which the cordon tree is grown. Support structures are usually necessary, mainly to guide the direction of growth for the first few years of the tree’s life, until each main stem has lignified and thickened to the point where it’s no longer flexible.

The width of each main stem is determined by the growth pattern of the fruiting spur network along that stem, but is intended to be kept relatively narrow. The shape of the overall tree structure is, as you can see from the diagrams above, determined by the type of cordon you’re growing; whether it’s a single stem, double stem, or part of a much larger arrangement.

The mode of pruning that’s common to all cordon shapes is the trimming back or outright removal of the previous season’s excess vegetative growth: the long, usually quite straight, non-fruit-bud-bearing stems, often called ‘watershoots’ or ‘watersprouts’²:

Apple ‘Belle de Boskoop’ double-oblique cordon showing extremely vigorous regrowth of ‘watershoot’ sub-stems.

The same image, with red marks to indicate likely pruning cuts

As with any pruning operation, the first thing to look for is the ‘3-Ds’: dead, damaged or diseased material that requires priority attention. Once any died-back stem tips and snapped stems have been trimmed, and any sections of the tree affected by apple canker – usually the most likely culprit in the ‘diseased’ category – cut out, the next step is to assess the remaining growth and start making decisions as to where else to make pruning cuts.

‘Short-Spur’ Pruning Method

I generally prefer the short-spur (a.k.a. ‘Courtois’) pruning method³ for apple cordons which are, after all, meant to be kept compact. Starting at the base of the cordon and working upwards, the process centres on identifying a stem that needs to be cut back and then selecting a bud on that stem to cut back to. Under the ‘short-spur’ system the aim is to cut back to three buds up from the base of the stem. There’s often a small cluster of two or three buds ringing the very base of the stem, but these can be counted as a single bud for pruning purposes.

N.B. the third bud – which will become the terminal bud – is highly likely to develop into a new, vegetative regrowth shoot, so it’s very important that this bud is well placed to re-grow in a suitable direction. If the third bud is inward-pointing – meaning regrowth is likely to result in a crowded canopy – or outward-facing but likely to cause interference with a neighbouring tree, it’s better to move out to four or even five buds, or back to two instead, in order to find the best-placed bud to cut to. If the stem is particularly long it’s a good idea to cut it back in sections, to reduce the risk of accidental damage as it bends and drops away, before making a finishing cut⁴ at the selected bud.

Every finishing cut should be made with a sharp, clean pair of secateurs if the stem is thin enough, or a sharp, clean pruning saw if it’s grown too thick for secateurs⁵. It’s also important to clean your tools between trees, or even between pruning cuts if the area you’re working on has obvious disease damage. The quickest way to transfer apple canker spores or viral infections between trees is via your pruning tools. Keep a disinfectant spray or wipes handy and apply as frequently as you think you need to.

The cut should be made just above the chosen bud – far enough out from the bud to avoid inadvertent damage, but close enough to avoid leaving a stub of stem, which is likely to die back and could then rot down into the healthy wood – and at a sloping angle. An angled cut helps rain or irrigation water to run off and away from the exposed end of the stem. A horizontal cut risks droplets of water settling on the exposed end, which could cause the stem to die back.

Looking at the photo above, I think the cut I made on the right-hand stem is perhaps a little too close to the bud, I probably should have taken a bit more time, thought about it a bit more and come out another couple of millimetres before cutting. An easy mistake to make – and probably no great harm done in the long-run – but also an easy one to avoid. Take the process slowly – it’s always best to “think twice, cut once” – position yourself so you can see exactly what you’re doing, place the secateurs carefully before you make each cut, and you’ll be fine.

Fruit Bud Development

During the growing season following pruning, the two remaining buds closest to the base of the pruned stem are likely to develop into potential fruit buds, as per:

A = The point at which stem regrowth began, from a bud left just below a winter-pruning cut that was made last year (2022).
B = Two potential fruit buds that have developed during last year’s (2022) growing season, back from the pruning cut made at A. These buds will hopefully produce fruit in this year’s (2023) growing season.
C = A new pruning cut made this year (2023), just above the third bud on last year’s new stem growth. This leaves two more buds that will hopefully develop into potential fruit buds during the 2023 growing season. The third bud, just below the pruning cut, is now the terminal bud and is more likely to produce a regrowth stem.

All being well⁶ those buds will produce blossom and fruit in the following year. The successful growth and maturation of fruit should then result in the development of ‘bourse’ shoots or ‘knobs’, which are in turn good sources of further potential fruit buds and/or short, fruit-bud-bearing stems in future years.

When a bourse shoot produces a fruit bud that successfully fruits and results in the development of another bourse shoot, then a ‘bourse on bourse’ structure begins to develop. This is a good basis for the growth of a fruiting spur system, as per:

A = Two-year-old bourse shoot, formed in 2021.
B = One-year-old bourse shoot, formed in 2022 from a fruit bud that developed on the 2021 bourse shoot.
A + B = A ‘bourse on bourse’ structure, or fruiting spur.
C = Another fruiting ‘bourse on bourse’ / fruiting spur structure that has produced:
D = A short, potential fruit bud bearing stem.
E = Potential fruit buds, each of which could feasibly produce blossom, apples and new bourse shoots in 2023, thereby extending the fruiting spur system.

In some cases, it might be appropriate to leave un-pruned a long, thin stem that has originated from a previous year’s bourse shoot⁷, but only if it has obvious potential fruit buds or more bourse shoots at the tip. For example this stem, with a double-bourse plus potential fruit buds structure at its tip, could be left un-pruned⁸:

A = Two-year-old bourse shoot, formed in 2021.
B = Brindille shoot, grown from the bourse at A in 2021.
C = One-year-old bourse shoots, formed in 2022.
D = Potential fruit buds, which could produce blossom, apples and new bourse shoots in 2023.

Two important points to note in all of the above: the first is that I’ve used the term ‘potential fruit buds’ because although the indicated buds are the right size and shape to produce blossom, they’re not guaranteed to produce blossom this year. Whether they do, or remain dormant, or revert to producing vegetative growth – new stems and leaves – instead, will be dependent on the weather at blossom time, the available nutrients within the tree, and a whole list of other variable factors.

The second point is that the act of pruning does not in itself cause those potential fruit buds to develop. Yes, it might encourage their development, by altering the balance of plant hormones within that particular stem. But it’s also likely that, even if the stem had not been pruned, those buds could have developed into potential fruit buds anyway, due to their position towards the base of the stem and the length of the stem pre-pruning⁹.

The pruning is necessary though, in order to reduce the length and volume of last season’s stem regrowth, otherwise the cordon shape will be lost as the tree reverts to a more natural growth pattern.

That, then is the basic theory of winter pruning cordon apple trees. I do invite and welcome questions, or discussion of the relative merits of summer vs winter pruning – I’m planning writing a number of articles exploring my understanding of the implications and impacts of both methods at a future date and I’m always happy to stand corrected if better information can be provided – via the comments or by email. But if you happen to disagree strongly with any of the above, I do request that you provide some sort of research-based evidence to back up your position if at all possible. Opinions along the lines of “I’ve always done X and it’s always worked for me”, or “it says to do X in all the books” are fine as far as they go, but received wisdom without evidence, or at least an explanation of your understanding of the processes and reactions involved, isn’t going to be anywhere near as helpful as a good explanation.

Meet Our Cordon Apple Trees

Cordon apple ‘Keswick Codlin’ in blossom, April 2022

The cordon apples on our allotment have been planted at various times over the past four years. The earlier ones are on MM106 rootstock – which is usually intended for small standard trees of around 3m to 4m in height – so they’ve thrown up some interesting challenges in the volume-of-regrowth department¹⁰. The later ones are on much more sensible M27 or M9 dwarfing rootstocks and as a result the trees are a lot more manageable. They’re all planted along the right-hand long edge of the plot, down at the front. Some of the trees are somewhat shaded by a neighbour’s plum tree from mid afternoon onwards, but they all receive a reasonable amount of light at various points during the day and the majority of them seem to be growing well.

The general gist of this year’s pruning was, as I’ve detailed above, to remove excessive vegetative growth – in some cases this meant cutting back stems of up to five feet in length – and then trim up smaller side-shoots to encourage the development of fruiting spurs. I ended up using a bit of a mash-up of the ‘short-spur’ and ‘long-spur’ pruning methods, in response to the of size, shape and likely onward growth pattern of individual stems.

Cordon Apple Pruning, Winter 2023

Here are a few photos of our twelve cordon trees, both before I started working on them and then after pruning, along with brief notes on how I approached the pruning of each tree. Click on any image to see a larger version, which may help with making out the details¹¹.

Section One consisting of:
– double oblique ‘Belle de Boskoop’ (MM106)
– single oblique ‘Elstar ??’ (MM106)
– double oblique ‘Howgate Wonder’ (MM106)

Apple ‘Belle de Boskoop’ is a hugely vigorous triploid cultivar and this specimen is grafted onto MM106 rootstock, which is definitely not a good combination for a cordon tree¹². It produces masses of lengthy extension growth every year; one regrowth stem I cut off was around five feet (1.5m) in length and had already developed potential fruit buds about halfway along, which is highly unusual in all but the most vigorous cultivars. I did consider removing one entire arm of this tree, but in the end decided to keep both, hoping that the tree would put more of its excess energy into forming fruit, and therefore less of it into stem re-growth, if it had more fruit buds to divide that energy between. But then again, it might just respond to the hard pruning with yet more excessive regrowth instead. Time will tell whether I need to remove this tree completely, on the grounds that it’s just unsuitable for the space it’s in. Or I could re-graft a scion from this tree onto a dwarfing rootstock and start it off again.

The second tree is meant to be ‘Elstar’ but the fruit it produced last year were decidedly un-‘Elstar’-like and a positive i.d. remains elusive; although the closest match I’ve been able to find to the fruit that it produced last year is probably ‘Falstaff’. Depending on what fruit it produces this year, whether I can i.d. it accurately, and whether I actually like it, this tree could be another candidate for replacement. It also has some damage – which could be canker, or could be the site of a torn-off branch – about two thirds of the way along the main stem, so if the affected area begins to rot back into the main stem, that will help me finalise the decision to remove it. In the meantime, I’ve trimmed it back to fruiting wood in the hope of getting some identifiable fruit to work with.

The ‘Howgate Wonder’ is another triploid apple but thankfully not quite as vigorous as ‘Belle de Boskoop’. As I’ve got a ‘Howgate Wonder’ stepover that produced some absolute champion apples last year, as per…

https://twitter.com/OrchardNotes/status/1583765455017500672

…I decided to sacrifice one arm of this double oblique cordon and took it back to a single stem, in order to create a bit more room for light and air to circulate through the reduced canopy. I’ll be interested to see how it regrows this year and whether it starts to produce any champion apples of its own.

Section Two consisting of:
– oblique (tiny, almost invisible) ‘Rosemary Russet’ (M9)
– double oblique ‘Keswick Codlin’ (MM106)
– double oblique ‘Tydeman’s Early Worcester’ (MM106)

The ‘Rosemary Russet’ was newly-grafted in 2021 and planted in 2022 to replace a second ‘Howgate Wonder’ tree that I’d had remove due to an apple canker infection. The new cordon hasn’t put on much in the way of new growth at all just yet, so I’ve removed all the visible buds except the apical bud at the tip of the leader. Hopefully this will encourage some onward and upward growth this year. If it doesn’t respond as hoped-for, I’ll take the tree out again to make room for another one, as I’ve planted another, much stronger ‘Rosemary Russet’, again as a cordon tree, on the opposite side of this section of the plot.

‘Keswick Codlin’ is a tree that tends to throw out a lot of side-shoots and produce a lot of fruit buds; it definitely needs to be properly fruit-thinned every June or so. This particular cordon is no exception. I decided to remove a couple of sections of the tree that were starting to grow outwards from the main stems and trimmed the rest of the tree back hard. I’ve kept both arms of the double oblique for now though, as ‘Keswick Codlin’ is such a good apple it would be a shame not to have as many as possible to enjoy¹³.

I decided to keep the double-oblique shape of the ‘Tydeman’s Early Worcester’ as well because, as its name suggests, it produces an early crop of handsome apples that I think taste great, so the more the merrier as far as I’m concerned. You might have spotted that a couple of the long stems at the top of the cordon have remained uncut; this is because I plan to trim them later on and use them as scions to graft another couple of trees, which I’ll hopefully be able to grow on as two more cordons. As I say, the more the merrier.

Section Three consisting of:
– oblique ‘Simister Seedling’ (M27)
– oblique ‘Greensleeves ??’ (M27)
– oblique ‘Rajka’ (M27)

‘Simister Seedling’ is, as its name suggests, a seedling apple that I think has some promise. I’m trying to develop this cordon to the point where I can take further scions from it at a later date. It has suffered a serious setback though: the main leader was showing signs of canker so I had to remove it, leaving me with just those two thin side-stems to work with. Let’s hope one or both of them develop well this year, otherwise I’ll have to go back to the parent tree, re-graft and start all over again¹⁴.

Again, there’s some doubt over whether the ‘Greensleeves’ tree growing here is actually the right cultivar. The one or two apples it produced last year definitely didn’t look like the golden-green fruit that ‘Greensleeves’ is meant to bear. Again, I’ll see what it does this year and then decide whether or not to keep it. In the meantime, I’ve removed one section that was developing into a second main stem and trimmed back to wood that might develop fruit buds this year.

‘Rajka’ is a pleasant, modern apple that’s meant to be reasonably disease resistant. This was another straightforward case of pruning back the outgrown stems to encourage the development of potential fruit bids and fruiting spurs. I also trimmed back the leader to the height of the support, which should result in some branching at the top of the tree, but nothing that can’t be addressed next year.

Section Four consisting of:
– oblique ‘Duke of Devonshire’ (M27)
– oblique ‘Wareham Russet’ (M27)
– oblique ‘Winter Gem’ (M27)

‘Duke of Devonshire’ is an old nineteenth century variety that I’ve been involved in growing for a few years now, initially via my old job at Ordsall Hall, where there are a couple of trees in the heritage orchard¹⁵. Pruning this cordon was another simple case of trimming back to encourage the development of fruiting spurs.

‘Wareham Russet’ is a fairly rare apple from Weaverham in Cheshire, and as I’m a firm believer that anything with ‘russet’ in its name is well worth maintaining, I’m doing my best to keep at least one or two trees going. This cordon should start producing a decent amount of fruit this year, and again, I’ve trimmed back the side-stems whilst leaving a couple uncut in case I want to take them as scions and graft this tree again this year.

‘Winter Gem’ is another modern apple that’s meant to ripen late in the season. Again, it was a straightforward trim-back for this one. It only produced one apple last year, but I’m hopeful for a fair few more this autumn, if the number of potential fruit buds on the tree is anything to go by.

So, there you have it: an overview of how these twelve cordon apple trees were pruned to reduce last season’s growth and hopefully encourage the formation of more fruiting wood and more fruit buds this year, which should then result in more fruit in years to come.

Some of the pruning involved making large-scale decisions about whether or not to remove entire limbs amounting to half the tree, or take off potential fruiting wood to reduce crowding and improve access via the narrow paths in front of the cordons. The rest of the job was a series of micro-decisions about which bud to cut to on a particular stem in order to encourage outward directed regrowth, prevent over-crowding and hopefully encourage the development of potential fruit buds.

Pruning any fruit tree, whatever the species, size or configuration of the tree, involves the same sort of repeated decision making process and working on cordons is a great way to practice and hone the method. By focusing on a single, small tree, it’s easy to see how a single decision and resulting pruning cut affects the overall shape and balance of the tree, and how each decision is influenced and directed by the cuts that go before. Done properly, pruning can be a very enjoyable, mindful experience, and is always incredibly satisfying. Or at least, it is for me.

How about you? Do you grow cordon-trained fruit trees? What time of year do you prune your cordons, and why? What’s your preferred pruning method? How much do you enjoy the process of pruning? Do please let me know, via the comments. (As always, if you’ve not commented before, your comment will be held in the moderation queue until I can click the ‘approve’ button, but I promise I’ll respond to all of them as soon as I can.)

Footnotes

1
The effects of pruning on both fruit ripening and primary / secondary growth are topics that I’ll be returning to in detail at a later date.
2
There’s a Wikipedia entry describing what a watersprout actually is, but there’s no explanation of the origin of the term. If anyone knows and can point me in the direction of a source document or reference, I’d be very grateful; please do leave a comment below…
3
This method, as far as I can work out – without much in the way of primary source material to refer to, and all of it in French, which I’m not entirely au fait with – was developed from the 1860s through the 1880s by a monsieur Jules Courtois. I haven’t had a chance to examine the source material in detail, or to even begin to work out how it differs from the rather more famous ‘Lorette System’, but both the ‘Courtois’ and the ‘Lorette’ were originally developed for pruning pears and subsequently adapted for use on apples. One day I’ll find the time to explore and discuss both systems in more detail, and a few others besides.
4
By ‘finishing cut I mean the final cut on a stem, the one that needs to heal quickly and cleanly, as opposed to a rough cut that’s just intended to reduce the length of the stem.
5
Some folks might suggest using loppers on thicker stems, but unless they’re brand new and/or extremely sharp, it’s very difficult to get a clean finishing cut and some crushing damage to the remaining stem usually occurs. Loppers are great for cutting back large stems, particularly when a whole branch section is being pruned or cut out, but for a clean, finishing cut I’d always opt for secateurs or a saw if at all possible.
6
Assuming weather conditions are favourable; light levels are sufficient; the tree can take up the required nutrients; pollination is successful; no accidental, insect or animal damage to the buds occurs… etc. etc. etc.
7
Technically known as a ‘brindille’.
8
Over the course of this season, if that stem produced full-sized apples then I’d expect the weight of the fruit to bend the stem downwards quite significantly. Depending on whether or not the stem then lignified and fixed in its new position, and how that position related to the rest of the tree, it might have to be removed next year. But by then it will have done its job – produced one or two good apples – and new fruiting spurs will hopefully have developed elsewhere on the tree to replace it. The fact that this stem wasn’t bent over last year suggests that the apples that formed the bourse shoots at the end of the brindille didn’t grow to maturity; they may have been thinned out, predated, or removed before they were heavy enough to bend the stem.
9
This is another rather complex and technical topic and again, I’ll be exploring it in much more detail at a later date.
10
Definitely a case of “if I knew then what I know now, I’d have done it very differently”.
11
These were taken with a half-decent Canon EOS4000D camera, but I’m not a good enough photographer to know how to make the foreground detail more prominent, or the background detail less so. It’s even worse when I use my phone camera, which always wants to take pictures of woodchip mulch, and hardly ever focuses on the actual tree…
12
Note to self, and others: proper research at the planning stage can save a lot of time and effort down the line.
13
Even though I do have another ‘Keswick Codlin’, a stepover, that fruited prolifically last year, even after rigorous fruit thinning in June, and will probably do so again this year.
14
Which isn’t a problem per se, except it’s growing on the edge of a plot of land that’s ear-marked to become a construction depot for a nearby motorway junction extension, so depending on where they put the access road, it might be chopped down. Perhaps I’d better re-graft sooner rather than later, just in case.
15
They didn’t fruit particularly well during my 5 year stint, but apparently had a bumper year as soon as I’d left.

Winter Pruning Cordon Apple Trees, ‘Short Spur’ Method