Where to make cut on this willow?

[TheTreeSpyder]

Good answer Mike. That is a big cut on that tree. Cut will also shift CG high and base taper at that point won't be as wide as it could have been; both giving less stability.

i'd squint and see the white outline as parent, and the green outline as child; with tip of moustache probably around green circle. So i'd target slightly outside the top of your cut lines and finish at red bottom.

 

[Kneejerk Bombas]

The main reason I brought it up was that I was impressed with this willows ability to recover from serious trauma.

Excellent point Tom, that's mostly what I was saying in my long winded post, a little stub has very little to do with the tree's ability to recover from damage.
Heck I even argue that leaving a stub can be better than removing it completely because you may get sprouts all over it and in time be able to trim it back into a decent limb, much the same way your whole tree has recovered.
Although in the interest of cosmetics, most will choose to cut the stubs off, along with raise the tree up to allow easy mowing and grass to grow right up to the trunk. Even though it's not what's best for the tree, it's what is appealing to some, and the tree will usually survive just fine.
What really matters is the size and location of the injury, the species of tree, the location, the health and vigor, and all those kinds of things. Leaving a little stub is not the cause of a tree's demise.
Decay does not break through barrier walls because of a stub or a wound wasn't painted or covered with wax.

 

[Bermie]

If this is a co-dominant stem then there is no branch cone, so no branch wood, so no zone to be walled off by codit.
Cut it as some have said, between red and yellow and monitor it.

 

[maxburton]

A couple points based on my conversations with Shigo:

Stubs are as bad as flush cuts.

Codominant leaders do not have collars.

No wound treatment has ever been objectively shown to affect wound response.

 

[SilentElk]

Willow? Heck I'd cut it off 4 inches above the ground. Then next fall, I would sell the wood for $160 a cord. Stuff split nice and easy and doesnt weigh much.

Over all though I would agree with all above said answers if you wanted to keep the tree. Just above the red line maybe 25% of the way between the red and yellow lines and closer to the Red.

 

[Kneejerk Bombas]

If this is a co-dominant stem then there is no branch cone, so no branch wood, so no zone to be walled off by codit.

The tree will compartmentalize the wound. Walls 1 to 3 won't be very effective, but wall 4 will be.
The only time I can think of where an injury is not followed by a CODIT response is where you break through an existing wall from the inside of the tree out, like you might if you tried to clean out a cavity.

 

[Kneejerk Bombas]

Stubs are as bad as flush cuts.

This is simply not true. There are times wen stubs are less desirable than a "proper" pruning cut, there are times when stubs are better, and mostly, it just doesn't matter.
Think about a big, beautiful old Burr or Live Oak, growing out in a forest. It will have several dead limbs of different sizes, some dead for years, some less.
If you study the collar around these dead limbs, you'll notice as the branch decays, the collar closes in, eventually the last of the limb will fall off and the collar will grow around the stub.
If you think the by coming in and removing the dead limb, that you are somehow doing the tree a favor, you're wrong. And further, if you think it matters if you cut it at the collar, 2" out, 12" out or 3/4s of the way out, you also wrong.
Now cutting into the trunk, or flush cutting, that is obviously bad.
As I have said, you can't remove all the wood associated with a limb that is being removed, unless you can drill into that cone somehow, you're not even getting close to removing it all. That tells you that another 1/2" stub of a limb isn't really going to matter.

 

[TheTreeSpyder]

i think at some points you could say that a small stub swallowed by callous didn't decrease in diameter taper at first as drastically as a 'Shigo' cut; this could give more strength, or not have as much a reduction in strength/ taper loss.

On the flip side. the longer the sealing off and surrounding with callous takes the more chance for 'infection', temporary strutural weakness etc.; and also more volume of 'expensive' tree resources need to be used/ produced to seal/ swallow the stub.