I’ve been working on an interesting little project lately in Dorset, looking after a newly-designed garden which is learning to settle back in to its landscape after some fairly major re-sculpting of a steep hillside into handsome stone-built terraces.

The (extremely nice) owners are keen gardeners but not knowledgeable, so it’s up to me to come up with solutions to the inevitable little problems that crop up when you drop a garden onto a hillside and leave the evolution into maturity till afterwards.

The owners mentioned to me that three little ornamental Japanese cherries planted on one of the terraces hadn’t been thriving this summer. We worked through the usual possible ailments: drought in this year’s relatively dry summer, lack of shelter (the prevailing wind blows directly onto this terrace), silverleaf disease.

But then I went and actually looked at them, close-up. And it was blindingly obvious what the problem was: the above (bar the silverleaf) may have played their part but the clincher was the way these trees have been staked.

It’s hard to know where to start.

Bamboo canes. Far too weak, flimsy and flexible to hold a growing tree steady against gales and the knocks and bumps of everyday life. They weren’t stuck into the ground all that deeply either – you could flap them about with one hand. No support whatsoever.

That flexitie. I have issues with this stuff for all sorts of reasons: but in this case, again, it’s far too stretchy for holding a young tree in position.  It’s also tied so loosely around the tree that it’s simply not doing its job.

All this meant these little trees might just as well have been planted without stakes at all. And you could see the damage: the root balls were clearly lifting out of the ground.

The reason you stake a tree while it’s getting established is to hold the rootball firmly in place. If you don’t, when the top of the tree whips about in the wind it will also pull at the rootball, which (because you haven’t staked it properly) is free to move within the soil. That rips away those delicate feeder roots a tree puts out to explore and colonise the surrounding soil, effectively repeatedly preventing the roots from anchoring the tree in the ground. This keeps the root ball loose, and because the tree can’t develop a better root system than it had in the pot, it cannot grow. That’s why these trees were suffering.

So here’s what I did:

This is a proper tree stake: around 7.5cm (3″) diameter, sturdy round wooden pole about 1.5m (5ft) long (you can use squared timber, as long as it’s good and robust). It dwarfs the tree trunk a little but that’s the point: it’s meant to be stronger.

I’ve driven it in to the ground with a mallet at a 45° angle, firmly enough that you can’t move it easily by hand. Opinion is divided about whether stakes should be parallel with the trunk to about halfway up, or like this: I’ve always favoured the 45° approach as it holds the rootball while letting the top of the tree move, and you’re driving it in a little way away from the rootball itself so you don’t have to damage any roots by sinking it closer to the trunk.

You’ll notice it’s pointing away from the wall, into the wind: this is deliberate. When the wind blows at this tree, the force is pushing the tree against the stake’s anchor, so it shouldn’t move. If I’d pointed it the other way (or straight upwards) the wind is effectively pulling at the tree rather than pushing it into the stake – much less stable.

And last but not least: a proper, collared tree tie. This holds snugly around the trunk and around the stake, the collar making sure the two don’t rub, and is stretchy enough to allow growth but not stretchy enough to give in to the wind. I’ll be checking that tie each summer as the trees grow and loosening it off if it’s needed: hopefully, in a couple of seasons’ time, this tree will be fully recovered from its bad start, well rooted and growing on so strongly I can pull the stake out and let it strike out on its own.

Problem solved!