They do look like they're hanging, don't they? I've been thinking about this too.
I think it's partly because the ponds aren't like a swimming pool: they're bowl-shaped, so the part at the edge is relatively shallow.
He's got good stabilizing trees and plants to anchor the soil, and he's got some good sub-soil structure.
He sealed the ponds well, and manages the water so the dam wall doesn't get saturated and as Paul says, "go downhill to visit the neighbors."
He's also on a mountain, good solid rock and not a ton of soil on top.
He talks about it in the "Sepp Holzer's Permaculture" book. It's a great read.
Well, that's my guess anyway. I think it's accurate, but I could of course be wrong. It's happened once before. Or maybe twice...darn memory!
Thanks Cam, do you remember where in Sepp Holzer's Permaculture he discusses this?
I know he talks in his most recent text about not "fully" sealing a pond in order to promote the restoration of the hydrological balance. In that, by allowing seeps, the water structures become more balanced across the landscape (this is my feeble reductionist paraphrase).
I wonder if the same is true for the steep slope ponds? It would seem that full sealing is "unnatural", but allowing seeps on a slope seems dangerous!
I was thinking the water would undermine such a structure, built by a less capable person than Sepp Holzer, from underneath and towards the upslope portion of the pond (with capillary action pulling the water upslope in only a limited capacity). Thereby, creating a slimy, slip layer where the subsoil and/or top soil meets the rock (or other slip layer).
I took an avalanche course a while back, maybe my explanations aren't that great, but I can see parallels (It seems snow might act like water stored in the soil in certain respects?)
I hadn't considered that the area of the pond near the "edge" would give way, since it can be compacted and planted -- something to ponder.
A lot of things come out of nowhere, so look everywhere.
The lakes at the Krameterhof were built over many years with a very well planned system. Sepp started small and expanded as he gained experience. The Krameterhof was built step by step, year by year, developed over almost 30 years. Through careful observation Sepp learned how natural systems work and how to work with them to avoid Catastrophe. He is working with the underground water on his mountain side, understanding how this underground water works is key.
It is important to start small, and such steep slopes can certainly be very dangerous. Sepp dug his first pond by hand at age 5 and built from there.