Linear Food Forests along Hugelkultured Swales

In this design we will be planting linear-food forests all along the downhill side of each of three hugelkultured swales.   

What is a hugelkultured swale?

While the tress and shrubs are in the early stages of growing (small) we will use the open space to grow some of our annual vegetables. We will also plant some nitrogen fixing ground covers and dynamic accumulators to help build the soil.

Open areas between the linear food forests allow for access to harvesting the crops, provides full sun penetration, and creates an area for our annual gardening or small animal grazing. The nutrients from animal droppings are washed into the hugelkultured swales which in tern soaks up the nutrients to fertilize the entire linear food forest.

Please Note: WARNING!

Log Slides
Scale is an important consideration when designing. If one builds a large swale system, with big logs — and a deep swale — on a steep slope — there is the potential that when the swales fill during a large rain event, the water could literally lift the logs right out of the ground (many logs do float in water) and send them down the slope.  Our friend Jack Spirko wrote a good article for the PRI website on this exact point.

Mud Slides
We cover this in our PDC courses but wish to make a clear point here as well.  On reasonably steep slopes where there were once deep rooted trees, the soil built up over centuries with the trees anchoring the top layers of soil to the bedrock or more solid substrate below.  In heavy or prolonged rain events, even though the topsoil becomes saturated, lubricated and heavy, it stayed in place because of the deep rooted trees.  When people arrived and remove trees for either subdivisions or to make the land more ‘agriculturally productive’ the deep roots slowly rot away and now there is nothing to hold saturated soils to the steep hillside.  If one is designing swale systems on steeper slopes it would be wise to make them shallower to start with while also planting deep rooted trees immediately below them. Nut trees tend to have deeper roots–fruit trees do not.  As the trees grow over the years the swales can be deepened if desired.

Rule of thumb?  It’s better to hold the larger volumes of water in either ponds located high on the landscape (on flat ridges) and then again in swales and ponds below the keyline where the grade is much more shallow than on steeper grades.  On steep ground we prefer to just plant deeper rooted trees and shrubs. Oak (tree) and hazelnut (shrub) are good examples.

This is just some common sense. Do not let this stop you from getting started.  If holding water on the landscape is a clear benefit to the land and the overall health of the environment then put in some rain gardens, small swales or ponds. As you gain in your own experience and see the work that others have done you will deepen in your own insight and confidence. Go…go…go…!  This work is important.


[contentblock id=10]

Leave a Comment

Your email address will not be published. Required fields are marked *

12 thoughts on “Linear Food Forests along Hugelkultured Swales”

  1. I am going to do the Perc Test using the method you provided. I’ve now tested our soil type using the soil texture triangle, and it is sandy loam. I am puzzled by your comment that if the Perc rate is 4″ or more that a swale would be useless. If a good Perc rate is 4-8 inches per hour, does this mean swales will not actually hold water back if the Perc rate is over 4″ and that run off is not an issue so swales are not necessary?

    1. Good question Ed.
      Swales will hold water…IF…any water gets to them. With a perc rate of 4″ per hour however, it will have to be one heck of a rain for there to be any water running over the surface of the land as it all being soaked into the soil at a rate of 4″/hr. as the rain lands on it. Thus, no runnoff and no meaningful quantity of water making it to the swale.

      For those who don’t know how to determine what their soil perc rates are, here is a link to the post where we describe how to do the tests. You’ll have to scroll down the page a bit.

      Good Luck Everyone…

  2. Roberto Pokachinni

    I was curious about the drawn image (which is very nicely done). It looks as though in the drawing the trees are growing on the lower down slope of the hugul bed, above the pile of wood (albeit just the edge). I understand that this is not recommended/desirable because when the wood is breaking down the berm will lose stability/shape and thus destabilize the base of your largest food forest element.

    My own ideas are very similar to your many projects and I’m really grateful for your site/sharing. As my project is mostly pre-embryonic.

    I was thinking that, instead of building full length hugulkultur on the swales, I would build occasional hugul mounds (maybe twenty feet long) on the swales, with keyholes beds in the hugul mounds facing down slope (so they don’t catch frost, but do catch sun (my contours are roughly East/West). This would eliminate the probability that the mound could be saturated by a inundated swale, causing catastrophic failure to the mound. I have read and watched Jack Spirko’s work on the PRI site, and assume that you have built some heavy redundancy into your swale overflow sills for freak mega storms so that your hugulkulturs can never be compromised??

    I was thinking that the tree systems in my own swale/hugul combo sites would be a few feet down slope of the hugul mounds (which would still benefit from the sponge effect (water retention) of the hugul without the potential destabilization). The tree or shrub feeder roots could still seek the hugulkultur’s reserve of nutrients without being on it’s edge and potentially being stabilized.

    I was thinking that the large trees could be surrounded by a buffer of smaller trees, shrubs and herbaceous plants, so that the system has that much more edge and protection.

    I was thinking that smaller hugul mounds could be placed in the food forest itself, increasing surface area of the ground system, increasing frost drainage, give another dimension, and boost heat retention/nutrient reserves.

    I am reminded that hugulkultur is commonly thought of as a long term raised bed that does eventually loose shape/height, and thus needs rebuilding. I was thinking that that instead of rebuilding them, the huguls could simply be built in such a way that they can easily be added to. Toss wood on the top, and when there is excess mulch or organic debris then add this over the wood, with seeds.

    Again, great article/plan/project. Very happy to see your work. ~Roberto

    1. Hi Roberto,

      You would absolutely not want to plant trees on the slope of a hugelkultur bed as it would eventually subside and expose the roots!

      Your reasoning seems sound but it’s hard to ultimately know how a particular setup is going to perform without some testing first. As long as there is no catastrophic failure your worst case would be two unrelated but certainly useful systems.

      Our hugelswale was the result of the combination of some very specific conditions and resources that we had available to us at the time. I think the first step would be to make sure that you have the resources available to create what you’re imagining and that the soil and grade is suitable for swales. We had some windfalls nearby and had the person power to move it into the swale. The grade is also not too severe nor was the catchment that large.

      You could certainly try adding to the top of a spent hugelkultur mound but I would be worried about exposed wood. The resources you need may not appear at the moment you need them. You should test it on a small scale and see how it works!

      – Milton

  3. This is a great post and summary. Is the swale in the video about the size of those indicated in the illustration of the design concept you included? Not in terms of the length on contour that the swale can extend, but in terms of the height/width of the swale and mound itself. What length of swale did you need to be able to plant all of the categories (overstory, midstory, understory, etc.) that were in the illustration? Also, how far downhill of the berm did you plant to fit it all in?

    1. Hi Nancy…
      The illustration of the swale and the swale in the film are close in size in terms of width and height. The actual dimensions of the hugelkultured swale are 2′ wide by 1′ deep and the berm it soaks is over double that, 5’wide and 2.5′ high. The trees in the illustration are a bit shorter than they will actually be. We are intending to pollard the over story to about 15 foot….the midstory to about 10 feet… and the understory at about 3-6 feet. The spacing is wider as well than the illustration eludes. The plantings extend 15-feet from the downhill base of the berm. Hope this is helpful.

      Swale #1 is about 80 feet long, swale #2 is almost double that.

  4. I think think this is a really great way to do food forests. After time the center can be left open for gardening and grazing, or filled in with more trees depending on the needs of people around it.

    One note on the hugel beds based on sepp holzer’s teaching. You may have factored this into your design based on your site, but its worth noting. He recommends against laying out hugel beds on contour and instead lays them slightly off contour. This allows water to slow down, but still drain if theres a lot. otherwise you risk blowing out the bed. The idea of a hugel bed is that it’s light and airy, as opposed to the berm of a swale which should be a bit compacted to stop the water. Also, having it slightly off contour can help drain cool air and eliminate potential frost pockets where you’re trying to grow food. That said, next year were building some in our food forest very similar to yours pictured here. With a swale, on contour. The land has very little slope unlike the alps where Sepp is living and I’m not too concerned about it on a gentler slope.

    1. Good thoughts Jeremy. Appreciate your weighing in. We’ve designed around Sepp’s experience and concerns. Our swales on contour have spillways built into them so that the water never gets above 1/3 the height of the hugelberm. With earthworks, we are always in control of the water depth regardless of how much rain. There is no way for them to blow out and the center and upwards part of the hugel will say airy. The wood holds the water which is slowly wicked up through capillary action.
      We finally planted our first linear food forest so are looking forward to seeing how everything works together. Hope you can come and see them growing at some point Jeremy. Cheers.

        1. Good question Ed. Another question that comes up frequently is how big (deep and wide) should the swales be.
          The first thing I want to know about a piece of property is how much water is actually moving over the surface during a rain storm. Some soils, especially sandy ones and those with lots of organic matter, can often absorb 4-8 inches of water per hour. In such a case, swales would be mostly useless. If there is no water running over the surface they would never fill and thus have no reason to be there.
          But assuming the percolation rate is low and there is water to capture then the size of the swales is determined by how much of the unabsorbed water one would like to hold. If the answer is all of it you might be digging some very big swales.
          What I tell our students is to ask the question, what is the size of swale I can comfortably dig and maintain? Anything is better than nothing so do what you can conceive of and afford. There is no permaculture rule or law that says you have to collect all of the water though this may be desirable in certain situations.
          Regarding the distance apart between swales that mostly depends upon the planting system we decide to grow. If we are planting nut trees as the long-term overstory that will eventually have a radius of 50 feet, and I want some elbow room in between rows, I would put the average distance to the next swale at about 120 feet.
          In this case however, since it might be 60 years before the trees get close to that size, I might dig in another swale in between and grow fruit trees along that for 40 years until the nut trees shade them out. Make sense?
          In another design, if my largest trees will be semi-dwarf fruiting trees (8′-10′ radius) with some understory plants extending beyond those as a linear food forest with a width of 10′, and I wanted a lane to access this system (10′ wide lane) then I would put my next swale at about 30′.
          There is no set rule Ed, just multiple considerations.
          Most important… give it a try. You’re just moving some soil which can always be pushed back in if it does not work out.
          Shovel on… Bill

          1. Thank you very much for a thorough answer. My wife and I are completely new to permaculture. We have been organic gardeners for years, producing almost all of the veggies and a good amount of fruit that we eat, but we are learning much more about soil and water retention that is most interesting and something we want to try. Our soil is very sandy (North Florida) so your comments on water retention are very helpful. The area we are considering reengineering is about 100 feet, but we are still thinking about the best area to use because of the different slops on our land. I suspect your 30 ft. recommendation would be most appropriate as our first try so we would dig three swales, but not deep ones. We will be doing this with a shovel! Thank you again for your quick response. Ed