Trees on a Hugelkulture in a Food Forest

I've read multiple times not to plant trees in hugel beds, so where does hugelkulture fit within a food forest?
My yard is heavily compacted, mostly clay, and much of the topsoil was removed a decade ago before I got here.
I'm thinking of a sun scoop shape on the southern side of my home. If I have the 7 layers of a food forest, where do mounds go within it? And how can I make the compacted clay conducive for trees if they can't be planted on hugel beds?
I love the idea of hugel beds or berms for some privacy, with trees adding to that. If the mounds are toward the back of the food forest and I plant taller trees on the edges of them, the trees would need to be on the northern side of the mounds. But then they'd be in a good bit of shade until they're at least a few feet tall. I just can't wrap my mind around what would be best practice.

I think the "don't plant trees to hugel" is referring to large trees and the biggest concern is stability. If you check Paul Wheaton's hugel videos, you can see his beds have an occasional apple or peach tree, you definitely see conifers between the berms.

When you start a new hugelculture bed it's better to concentrate on the layers closest to the soil surface, but as the berm matures, you can totally produce most (if not all) of the layers.

Living near the old growth redwood forest, a place where many wild trees start on nurselogs (check out Redwood Ecology, Noss,1998), I think it’s odd to argue to not plant trees on hugels, which basically mimic nurselogs. Of course context is key and in places where it’s never too wet and often too dry it makes sense to plant them lower on the hugel or at ground level. Here and elsewhere in the coastal NW US, high winter water tables are a major constraint for fruit trees that need to be 2-4’ above the water level. Of course a hugel cannot be allowed to think it’s a dam by blocking flood flows, but if one has to build berms to get trees above the high water level, I see no reason not to fill that berm with 1/3  wood debris. If one is worried about settling of the hugel exposing roots, just go look at wild trees that often start on nurselogs or stumps, and see the octopus like roots that form to stabilize the tree. Most trees actually benefit from some exposed roots for air access and rot prevention (Mike McGrath, YBYG podcast, Rosalie editor).

It seems Sepp Holzer (and Paul Wheaton) is in a much dryer area (1/8th our rainfall), so he plants at the base of hugels, which he also will break up and spread as compost after 15-20 years of use. I find here the elevation and excellent drainage of hugels to be of benefit for trees, but of course it’s more efficient to just plant trees on slopes that drain naturally.

If I understand right, it's about settling. The materials inside a hugel are going to shift position and decompose over time. The process is fairly slow, it takes years. Annuals and short lived perennials won't notice the position changes, but trees can live for decades.

Thanks for the replies!

Ben Zumeta wrote:  If one is worried about settling of the hugel exposing roots, just go look at wild trees that often start on nurselogs or stumps, and see the octopus like roots that form to stabilize the tree. Most trees actually benefit from some exposed roots for air access and rot prevention (Mike McGrath, YBYG podcast, Rosalie editor).

You brought up some really great points. I imagine many varieties of trees would compensate for the settling over time. Some may be lost, but others would thrive, just like in any forest. The octopus roots showing at ground level have always fascinated me. There's a popular tree in my city that is on a hill, and the whole side of the hill is exposed roots.

Forests grow on fallen forests. Most western hemlocks start on nurselogs, which provide a necessary microbe inoculated for them to grow in addition to moisture and nutrients. If they are leaning heavily due to the decomposition of their nurselog base, or more often as they reach for light and get wind blown, they may only live a few hundred years vs 1000yrs or more. Of course other tree species vary, but there are trees well over a thousand years old (cedars, redwoods, spruce) that started on nurselogs. I’ll get better pictures on my next hike.

I don't disagree with Ben's comparison of a hugel and a nurse log, but many plants started on nurse logs end up dying because enough of their roots never get to the stable ground beneath the log. I lost a bunch of volunteer shrubby plants on a nurse log last year due to the combination of the log getting too rotted, the ground being too rocky and the heat dome extreme weather.

I planted a Black Current on a on a mound that contained a lot of wood and it ended up somewhat tipped over, but I suspect this was aggravated by where the light was coming from, not just the wood decomposing (although there is no longer a mound there!) This is not any sort of a problem for a Current, and I'm risking a similar situation with a pair of Hazelnuts, but Hazels are known for being good for for copicing so if it ends up at an angle I don't like, I just have to prune off angled stems and encourage new ones that will correct by growing towards the light.  Generally with something like an apple tree, that sort of major pruning isn't done and if it's already 10 years old before it starts to tilt, the weight at the top could potentially break it off or tip it right out of the ground.

A big problem I have is that with my setting and ground, my hugels tend to end up with a *lot* of wood compared to dirt or soil. This would aggravate the issue, so I would be inclined to plant the bigger plants on the lower sections of the hugel - maybe by making the plant a bit of a terrace?

Good points Jay, and that settling issues is largely why I use a much lower proportion of wood than I used to. It also helps to minimize the thickness of wood layers, and to counterintuitively compact it as much as possible. This is done with an excavator much better than by foot. The vast majority of settling happens in the first year, so if planting perennials can wait, that’s a consideration. Again, hugelkulture is not an efficient way to plant trees in many cases, and hillsides around the world need reforestation. However, I don’t see why it’s considered implausible.

It sounds like the answer is "it depends on the type of tree and hugel bed." But it also sounds worth trying in my case, especially if I wait until the bed is at least a year old. Thanks for the things to consider!

Trees and shrubs can deal with crappy soil, but it slow growth, so I still wanted to feed them.
Spreading out the biomass kept it from breaking down quickly and fed whatever got to it first, not necessarily the tree.
Since annuals beds located near trees are often invaded by the roots of the tree, I decided to build lasagna style raised beds right next to the trees.
These are tree feeders, moisture reserves, compost piles and annual vegetable beds, in that order.

If you do the same with hugel mounds, everything that filters through the mound or slides off it will end up feeding the tree.
If/when the mound has fully decomposed,  you can shift the remains around or leave them be.
The tree will take what it needs.

In my exerience tree roots will also grow upwards quite well. I have some Rowan trees in my front garden and I made a rough bin from old corrugated steel nearby to put ash from our wood stoves in so my hubby would not have far to go with it etc. Anyhow the bin was not the best idea because it doesn't get emptied (by me) so started overflowing.... I'm just moving some down the hill to my "natural farming" area to hopefully feed and modify my acid soil slightly. As well as ground elder roots (which I hopefully won't export) I am also finding fine tree roots 18 inches up in the wood ash. It is obviously rich in potash which my soil is low in, so they are accessing the nutrients and growing into it, even though there is no soil, just ash. With nice compost in the hugel, I expect the roots will grow in even quicker.

(edited punctuation)