This is a great thread, I want to offer some of my first hand experiences managing the hugels installed by Sepp at the Place of Gathering. Sorry for the Length. ENJOY!!
Below is a photo of an installation Sepp did in Montana at Place of Gathering. If you look closely you can see little reflections at the base of some of the hugelcultures. This water pooled at the base of these hugels after rains because they are placed on contour and not down slope. The slope is very gradual, maybe 5*. Plus the Hugels have curves and meanders in them which actually help to encourage pooling at the base of the hugels. During the management of these hugels I found these areas to be very valuable. The hugels adjacent to the puddles were obviously more moist and the areas where the water collected grew plants like parsley, carrot, parsnip, mallow (althea officianilis). Again, big fan of what placing them on contour achieved.
No erosion was evident in this scenario. In reference to frost drainage, the areas did seem to be a little cooler, which had its benefits in our hot temperate summers.
I believe hugel swales have there pros and cons. It would depend on how it was built. The important thing would be to build the hugel on the downslope side of the swale to have enough material so that as it broke down it wouldnt drop to a point that water could spill over and generate erosion. Building a mulched hugel swale would prevent the water from having much opportunities to move and generate small erosion gullies. By having the swale sloped down contour slightly at an angle of 1:300 or 1:400 like Yeoman did, then you could ensure the water from heavy rainfall events would flow down the swale and not over it. Also, we must consider that when the wood breaks down it turns into what is called brown cubicle rot. Brown cubicle rot is a stable conglomeration of carbon that will vary little in size once the wood has achieved that state. This is the end goal of the wood in a hugel and resembles biochar in structure and function. Due to its angular nature, it is actually quite structural... like building a house with bricks vs river rocks. In terms of "charging" a hugel with water, swales would undoubtedly be effective at this.
When I was talking to Sepp about how to manage the installation after he left he suggested we irrigate the hugels, contrary to the purported myth of no-irrigation necessary. However this suggestion was suited for our climate which averages only 16 in of rain a year. He suggested 3 soaker houses: one at the peak and one on each size slightly above half way up the hugel's side. When experimenting with this i found it took about 36 hours to saturate the whole hugel at a pressure rating of 25 psi and a flow rate of 7 gpm. Once the outermost layer of soil become dry the soil "wick" that allowed for evaporation of the water stored in the hugel was broken and that stored water would remain for up to a month.
In fresh hugels the wood will not store very much water, but it will release plenty of nutrients and host microbes. 2014 will be the 3rd year of our 6 foot tall alder, dogwood, hawthorn, cottonwood, aspen (fresh and dried) filled hugels. the first year the wood and churning of the soil released loads of nitrogen. The second year the wood locked it up. I suspect this third year we will begin to see a balancing effect where water and nutrients can be stored in the decomposing wood.
Adjacent to one of the ponds we did not actually put a hugel but simply a soil berm of fine sediments (see photo below). This is where we planted fruit trees and Sepp's Famous Rye (which did great, WE HAVE SEED!!!). The fine sediments allow for the effective capillary wicking. By planting the fruit trees here they are above the phreatic zone (the layer in the soil where water fills the spaces between soil particles) and are not drowned by the water but have the benefits of being in saturated soil. From my observations of the site, the coarser soil materials did not wick moisture as effectively (eg. the sandy hugels require more external watering). The tops of all of the hugels were constantly dry, constantly. Unirrigated generally the hugels were dry for the first 8 inches of the soil. Starting seeds in this dry soil is a waste of seeds as they would become seedlings and then could not reach the moist parts of the hugels and die. Just like with any other gardening it is important to give enough water to young plants that they can become established but little enough so that there roots reach deeper.
To respond to the dry tops I planted deep rooting perrenials such as rhubarb, horseradish, elecampane, and jeruseleum artichokes. These have the capacity to reach down deep into the hugels for moisture. They are also all divisible by root divisions. they all also feature massive amounts of leaves. These leaves provided chop and drop material, shade for the driest part of the hugel, prevented deer from crawling over the hugels, and captured dew from morning fog that would drip onto the hugels.
All of the trees planted on the hugels are doing INCREDIBLE. They were planted by mistake as Sepp told us to only plant bushes on them (2/3 up on the hugel). But now years later they are doing better than any other trees (except for those planted adjacent the ponds). It seems to me that planting trees on hugels is like how nature plants trees on fallen redwoods... I actually encourage it. From my experience what hugels need more than anything in order to minimize watering needs is mulch (leaves fallen from the trees planted on there tops), and shade. From growing vegetables throughout the installation i observed that they require much less light than commonly believed and did better in places where they were in the shelter of canopy (full for part of the day and dappled for part).
I was doing a restoration job with some friends that involved some thinning. We took the thinned trees and milled them for beams (a very valuable product) and took the slash and chowdered it up (put it in a pile and chainsawed it up to make it into finer materials). We then filled a 3 -5 foot deep trench with this slash and built hugels over the top of them. I feel that doing this on contour would provide a trench for the water to sink into beneath the hugel and would not compromise the structure of the hugel. This Hugels performed great. By having more thin branches rather than whole logs incorporated into the hugel we ensured there was a high proportion of nitrogen to carbon (as well as a higher proportion of cambium to support mycelial populations).