Establishing a Food Forest

Choosing the right method of establishing a food forest could save you a lot of time. Here Andy Cambeis offers some background and describes two different approaches – each right in its own context.

The pioneer in Western Food Forests, Robert Hart, started his Forest Garden without an establishment concept to guide him. When Bill Mollison picked up the Food Forest idea from him, there was little in the way of methodology for how to establish a Food Forest.

Geoff Lawton was, to my knowledge, the first one who documented a method for establishing a Food Forest. Geoff is used to working on difficult sites, so his method is tailored for this, and ensures you end up with a Food Food even in a very difficult location.

Geoff’s approach is based around phases of succession for every layer of the Food Forest, his goal is to slowly outgrow/out-compete the former plants with his chosen species. He starts with his own aggressive pioneer species that make room for the next species, until it’s all conquered and secured.

Geoff Lawton’s establishment approach is a perfect imitation of how natures works to establish a forest, a sure but slow process. This kind of approach will take 3-5 years of support and escort work, and this is where Geoff uses a lot of “chop & drop” to transit the existing soil into a fertile fungi dominated forest soil.

Zaytuna FF 0
Geoff Lawton on the top of a swale in the first phase of establishing a Food Forest
on a sloping site at Zaytuna Farm, Northern NSW

Zaytuna FF
The result – Geoff is standing in the knee high (not very productive) ground cover
of his Food Forest swale, and on the left side is knee high grass

But wait. . .

What if I want convert a nice flat piece of fertile, formerly farmed land into a Food Forest? Well, you can use Geoff Lawton’s method for every type of site, but why spend 3 years (in colder climates likely longer) if you have land where such an approach is not necessary?

This was my starting point.

Hawea FF 0
Hawea flat Food Forest cluster 1 before any work had been done

I had to modify Geoff Lawton’s method to meet the needs of my client, and the main differences are:

No need for the (conquer) phases of succession.
Everything goes in once the ground cover is established.
– Production plants.
– Nitrogen fixing plants.
– Dynamic nutrient accumulator class plants.
– Beneficial predator and pollinator insects attract plants.
All planted at the same time.

No need for nursing plants.
– Plants receive proper treatment at planting time and that’s it.

The establishment of a very-dense ground cover, is the key to success

Sufficient nitrogen fixing and nutrient accumulation plants at each layer, from the start.

Establishment in small (700m2) but fully functional forest clusters.
– Stackable to unlimited sizes.

Geoff Lawton’s priority is to keep the ground cover and herb layer securely occupied rather than to have these layers be productive (to humans). This is certainly something you might have to do if your Food Forest is surrounded by aggressive and invasive species. Using a productive ground cover and herb layer approach in sub-tropical bush, with it’s high pressure of invasive species, is not realistic, but you can do this in Hawea Flat or Urban areas of Waiheke.

So be aware to tailor your Food Forest establishment method to your needs, and it probably makes sense to not mix different establishment methods.

There’s no reason to spend lots of time to completely remove the grass in the subtropical bush
– You won’t be able to grow much on the ground cover layer anyway.

There’s no reason to use a time-consuming strategy of “over-stacking” with pioneer species, for a Food Forest in Hawea Flat.

Martin Crawford in his highly productive root/tuber, ground cover and herb layer

AndyCambeis Andy Cambeis was the driving force behind the establishment of the Hawea Flat Food Forest and will be at the Food Forest Hui (Sep 26-28) to share his ideas.


  1. says

    Thanks for the great article Andy.

    I am trialing the use of ‘overstacking the pioneer species’ (conquer species or nurse species) in Food Forest Design here in Coastal and Central Otago. I have observed many benefits from planting more the planned mature Food Forest number of pioneers, such as: onsite carbon harvesting for mulch material, and effective improvement of soils and transition of soils to a fungal dominated woodland type. I want to produce the needs of the Food Forest from within as much as possible.

    Kay Baxter recommends planting trees for mulch here:
    “5.) Begin with a wide diversity of legumes and mineral accumulators, including ground covers, low bushes, high bushes, low trees and canopy trees, suitable for the specific site, densely planted, keeping in mind final fruit tree spots. Many of these trees will be cut out or chopped and harvested for mulch as time goes on. They are only dominant in the juvenile stage of the forest growth. I’m specifically looking for those that also provide large amounts of high protein/fat for chicken forage, and that survive chopping and dropping over a long period.”

    Another benefit of these fast pioneer trees between the crop trees can be to create shelter and improve the site microclimate quicker than just planting the perimeter of the site. I am using this strategy in design for an orchard on a slope where the lower shelterbelt will take a long time to provide shelter to the top of the site.

    • says

      Hello Jason,

      Mulch is not a issue in Hawea Flat. We got plenty for free. Also in the future. Be aware that building mulch on site also add a considerable additional ongoing maintenance element. You need to keep your fast growing “mulch plants” under control or they can quickly impede your productive plants.

      One of my design goals for the Hawea Food Forest was to keep the maintenance work to an absolute minimum. If we observe that we have more “manpower” than we need for the present maintenance, we can add new elements that need additional ongoing maintenance. Building mulch on site may be a good option for Hawea Flat in the future, if somebody can commit to more time for care of the Food Forest.

      My approach to improve the site micro-climate is the relative small size of the cluster. You can quickly fill up 700 m2 with bushes and herb layer plants to generate a nice micro-climate.

      • Attini says

        Hi Andy, I’m interested that you see an equivalence between imported mulch to mulch grown on site. While imported surface mulch does have benefits, I do wonder which natural forest system you are attempting to mimic. It is well established now that the ecological benefits of having not only the root material and nutrient cycling that conquering/pioneer bulk living mulch provides, but also that ecological stability comes from having these CRITICAL niches filled. It is not a Pick & Mix situation. I would urge you to consider the long term plant population vector of the site. Without the support of a strong and healthy pioneer weed base the under story will inevitably degrade to a low-turnover state. I appreciate that your client has dictated that you may establish this in one fell swoop, but this kind of end-state thinking has become rather antiquated.

        • says

          Hello Attini,
          I traveled intensely through a 25000 years old *untouched* temperate forest.
          I saw no general pioneer weed base. Where did you make your observations? What I saw was a super high diversity of plants and animals forming a patchwork pattern forest.
          Some parts denser than others, but overall lighter than the average people made forest.
          You’re ignoring the fact that all tree and bush foliage and approximately 95% of the herb and ground cover biomass stays in the forest. Only a fraction of the biomass gets harvested. I’m designing way more nitrogen fixing plants into every layer than recommended. Dynamic accumulator plants have also a big place in my designs. You can only compensate for the loss of nitrogen (from harvest) with nitrogen fixing plants. You can’t fix stuff from the air that’s not part of the air. (Phosphorus, Potassium, Sulfur, Calcium, Magnesium, Sodium, Iron, Silica, Manganese, Boron, Copper, Zinc, Cobalt, Molybdenum, Selenium, Chromium, Chlorine, Vanadium, Nickel, Tin)
          You have to find a long term solution for the loss of the non-nitrogen elements anyway. We need to observe how much we get from birds and other animals and how much we have to bring in. I see no point in making ongoing labor and maintenance requirements, if my soil is not that bad that I need to do it.

          • Attini says

            Yes, bird and insect dung is an interesting and largely unexplored area! I once conceived of a fish-n-chip-smell-generator that would attract more seagulls to my area 🙂 The point I wish to explore is that mimicking an established forest but evading the notion that there are countless generations of plants that have preceded overlooks the fact that they provide critical background conditions that support the long term stability of the system. Pioneers are not simply a tool for achieving an end-state design. Their presence in the soil seed load is what ensures against the reversion toward savannah conditions.

  2. says

    Different needs for different sites and projects! At the Aro Ha Food Forest inspired orchard in Glenorchy that I am working on with Jon Foote the site is so isolated that producing mulch within the orchard is one of the goals. Its good to know the reasoning behind design and have the different options in mind for future Food Forests. Thanks for your explanation.

  3. Dave says

    Hi guys. I a in the process of establishing a food forest in Makarora. We are working slightly in reverse as the orchard trees are already there, planted somewhat haphazardly but we are working on clearing the grass and planting a productive understory. We are using 10 chickens to clear an area before moving them to an adjacent area, folding them around using rabbit netting and the old style flat warratahs to hold the fence up. Initially we had to use bird netting over the top of the rabbit netting to stop the chickens jumping the fence as they had been fully free ranging and roosting up to 4 meters up in douglas fir trees previously. We have since moved them, clipped their wings and tripled the length of the fence, and haven’t had any escapees over the fence (we are no longer using the bird netting over the top.) They could still jump the fence if they wanted to but being fully incapsulated for a week has got them a little bit institutionalized. It takes around 7 days for them to clear about 80 sq m although there’s still a fair bit of grubbing to do of the bigger grass clumps before sewing the seed. I have chicken electranet but it isn’t particularly robust and would get snagged and damaged in the rough ground so am finding that the rabbit wire netting works well.
    We are using a combination of Sepp Holzer and Geoff Lawton techniques with a bit of Robyn Gyton thrown in. I have made up two seed mixes, both based on the Koanga orchard herbal ley. One that goes in close to the tree has additional allysum and fennel seed along with garlic, spring bulbs and comfrey, and in the more open areas the mix has a few swede and turnip seeds in it, nastirtium, sunflowers and we are planting jerusalem artichokes, strawberries, raspberries blueberries, and a cranberry. We also have tagaste seedlings started for the south end of the huglebeds and some black locust for the zone 4 wood area. It has been a real eye opener moving from Wanaka to Makarora and observing the much more rapid neutrient cycling and masses of earthworms due to the increased rainfall. The soil is young silt loam on top of sand on top of gravel on a stream fan and seems quite fertile considering it is still greyish in colour and lacking humus.

  4. Juliet says

    Hi – would I be able to come and have a look at the Hawea Flats project? I’ve just come back from doing a PDC with Kay Baxter, Bob Corker and Dan Palmer and worked on a food forest design for their fruit tree collection. I live on 3 acres in Arrow Junction which needs some inspiration.

  5. Sacha says


    Does anybody know where in New Zealand, preferably the South Island (we are just near Oamaru) I could find/buy some Goumi (Elaeagnus Multiflora) plants? I believe it’s considered a pest plant, which may be why I cant find it in nurseries, but we are completely surrounded by aggressive dairy farms so there is little chance of it spreading. Good luck to it if it does! Would also be interested in Sea Buckthorn if anyone knows a source for that.


    • says

      Hello Sasha,

      there are no cultivars of Goumi (Elaeagnus multiflora) such as ‘Sweet Scarlet’, ‘Red Gem’ or ‘Just Fruits’ on sale in NZ.
      I strongly advise against growing non fruit cultivars from seed.
      If you grow them from seed (VERY challenging), the fruits (after 4-5 years waiting…) will be very likely small with a large pip, astringent, sour/bitter unpleasant.
      Fruit cultivar have large, sweet tasty fruits.
      There are better nitrogen fixing shrubs you CAN grow from seed such as the Siberian peashrub (Caragana arborescens).
      Here is most what you want know about Goumi:
      Your site need to have about 300 chilling hours for setting fruit. say you are free to import the Elaeagnus species to NZ.
      I have quite a lot plant import/export experience and would consult you with the import or even do the whole import process for you if I get 1 free plant for me.
      Plant import/export is a *very* tricky business and a lot of things can go wrong that lead to the death of your plants. are doing i.e. plant export business.
      I have 2 times imported from them. They are doing a reasonable good export job but need negotiation about the sent method and the price. They would also buy more Elaeagnus (or Hippophae rhamnoides) species from the US market and include them in the export.
      Btw. there are much more good Elaeagnus species FRUIT clones:
      Elaeagnus angustifolia var. orientalis, Elaeagnus umbellata ‘Big Red’, Elaeagnus umbellata ‘Brilliant Rose’, Elaeagnus umbellata ‘Jewel’, Elaeagnus umbellata ‘Red Cascade’, Elaeagnus umbellata ‘Sweet-N-Tart’.

      Sea-buckthorn (Hippophae rhamnoides)
      It’s almost the same with Sea-buckthorn.
      If you grow them from seed, you will get acceptable tasting fruits (in small quantities).
      But the fruits are very hard to harvest (a LOT of thorns and berries that get very hard of the twig). Most people do not bother.
      The Russians, Czech-Slovaks and to some extent also the Germans have breed FANTASTIC fruit cultivars. Very tasty, big fruits (taste like mango). Some without any thorns. Most with only little thorns. And way easier to harvest.
      Some US nurseries sell the best Russian cultivars. Same import consultant/import job offer for them.


    • says

      Hello Sylvia,

      I gifted Koanga Institute a dozen Allium ursinum seeds last year.
      They may be able to sell some. But I recommend you to buy Ramsons seedlings, if you can, because the seeds are hard to germinate (need cold stratification).

  6. Bridget says

    Hello, Does anyone know where I can buy Siberian peashrub and/or other nitrogen fixing shrubs from? Thanks.

    • says

      Hello Bridget,

      I have never seen Siberian pea shrub (Caragana arborescens) and bush clover (Lespedeza bicolor) on sale in NZ. You have to import seeds.
      I successful imported i.e. Caragana arborescens from to NZ.
      I strongly advise against growing fruit species with a long breeding history (apple, pear, grape, Chinese date, etc.) from seed.


Leave a Reply

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