Our forest garden design: the future is forested and foodlike

| Forest Gardening, Permaculture, Permaculture Design | comments | Author :

Over the last couple of months, we’ve been cooking up a re-design of our top food forest with Harris. Once we move into our tinyhouse, this food forest will be right outside our back door. So we want to get serious about making it a gorgeous place that drips with fun and food.

Up until now, however, it’s been hard to prioritize this project over everything else (including the building of that back door), so things have been rather slow. But thanks to Harris, we now have a design. And that takes us a lot closer to realizing this particular patch of abundance.

Whether in our high-and-dry climate, or in costal Cringila, some things hold true in all food forests: "when at the edge, plant comfrey" for example...

Following on from food forest design time, below is the consultation and site design that Harris has come up with. It’s not too complicated and it’s step by step, something we really need here at Milkwood, with so many other aspects of the farm whirling around our heads.

I’ve included the full design below as I could see some of this info being very helpful to others wanting to design and implement forest gardens in temperate climates on less-than-ideal sites. Read on and use the info freely, in the spirit of open source permaculture.

This design will inevitably change a bit, but as someone once said, the beginning is all

Consultation and Site Design for Zone 2 Forest Garden for Milkwood Permaculture Farm

by Dan Harris-Pascal

Our existing food forest plantings, as mapped by Harris. To be expanded and edited in the coming years.

Initial Site Assessment:

The “Food Forest” site at Milkwood farm is located on a NE facing slope above the tiny house site between the road and the chicken coop. Up slope from the site are a dam, beehive sites, suntrap and intern accommodation. The goals of the forest garden have not been clearly articulated at this stage and desired yields and how many people the forest is for is unknown at this stage. The area may be netted to prevent crop losses to birds. The average slope of the forest area is a fall of 4.8m over 24 m giving a slope percentage of 20%. The proximity to the chicken coop means that it would make sense to design into the forest a way to graze the chickens in different areas of the forest at appropriate times. Running the chickens up and down the slope in the forest will lead to erosion and loss of materials so it would be best to cell them along the contours of the hill.

To the NW of the site is an established Eucalypt forest consisting of Yellow Box. This site is currently providing shade and wind protection to the NW planted side of site. The Eastern Side of the Forest is very exposed and plants planted previously have not done particularly well as they get much more summer sun. The site is very exposed to winds, the base soil consists of clay and rocks and has a pH of 6.

The NW established side of the forest is at the “forest edge” of the Box Forest and has fared much better due to edge effects. Working with the idea of gardening at the forest edge you can see that the plants that have been the most successful have been those sheltered by the forest edge from the westerly sun. This area has been cultivated for a number of years with the addition of mulch and humanure. The surface soil is not particularly aggregated but it does have a neutral pH. Much of the mulch that has been previously applied has been washed down the slope. This area also has a wide variety of opportunist species that have filled the empty niches of the ground cover layers which need to be removed or controlled so as not to compete with the trees or with the desired ground covers. Interestingly I could find no macro soil life to speak of (worms, slaters, earwigs) and I wonder if it is leaving because it is too dry, being washed away or being eaten by birds. There is extensive geese manure, some kangaroo poo and evidence that something has scratched up the sheet mulch (possibly to eat worms).

The fruit trees in this established are have been planted too close together for a productive garden and many of the trees obstruct paths or access to the humanure hacienda. Just replanting the trees that are there you will stack the canopy layer out enough to have a forest garden with all the canopies touching and no support species in the canopy. Paths need to be at least 1m wide to allow adequate access along them with branching paths to allow access to each plant. Further, there has been no thought put towards spread of products or desired yields from the forest so the established trees may not be the best to keep in the long run.

Brainstorming the sections and purposes of our food forest. Existing plantings are printed graphics, overlayed with notes, sketches and thoughts for the future of the forest

Design and Development


Access paths and better definition of the area is the first priority of the garden space. Two or three 1-1.2m wide paths on the contours of the slope would help to slow water down so that it can infiltrate and also reduce the losses of mulch washing away and make it easier to move around the area so one is not always moving up or down the slope. If designed properly the substantial runoff that flows from the suntrap to the road could be effectively harvested and used to get deep infiltration along the slope. Given the substantial number of rocks on the site it may not be possible to get paths exactly on the contour but provide the drop is small is shouldn’t matter and may prevent pooling. If paths are made to the contour they can be lined with mulch to provide a weed-free path surface.


While built in water harvesting and soil improvements aims to provide the bulk of water needs in the forest garden an effective irrigation system is helpful for getting trees established. Irrigation mainlines should follow the established paths so they are easy to locate. The least hassle irrigation system is dripline as it doesn’t have drippers which can break off and doesn’t clog as easily. The flow rate from the dam needs to be calculated to determine what size polypipe is required and what spacing of drippers. With clay soils less drippers are needed every meter as the water more readily spreads across the soil. The Dripline you buy needs to be pressure compensated to cope with the pressure difference going down the slope to ensure that water is applied equally to the forest garden and a pressure regulator may be useful as well. A vacuum breaker is needed 30cm higher than the first dripper to prevent the drippers clogging from sucking up earth. A flush valve is needed at the lowest point to clean the system. For ideal drip irrigation on your soils drip lines should be placed every ~50cm in the garden beds. You can design the system smaller to begin with but with expansion designed into it.


Strong winds come to the garden from the South and East due to landscape effects on wind direction. The longest length of the garden is 45m. This equates to a required windbreak height of approximately 6m to shelter the forest area. However, a break of this height may place the zone of turbulence on the house site. An additional break on the other side of the road may mediate this as, would wind tolerant canopy trees in food forest area or a higher windbreak.

If the windbreak is to double as a bee forage area I believe that a double row would be beneficial as this will give the bees more shelter on the NW side of the break. Ideally windbreak species will be a mix of bee fodder, chook feed and nitrogen fixers. The windbreak will be ~30m long to allow vehicle access through to the savannah area at the uphill end of the chicken coop. However, thought should be put towards other plantings uphill that are staggered to prevent wind funneling at the edge of the break.

For the first establishment of the windbreak I would recommend getting tree guards and tree tonic planting powder from global land repair as these appear to aid in tree establishment. Initial plantings could be done every 2 meters with large trees in the row furthest from the chicken coop. The planting could then be thinned as the trees establish if required. Closer planting at 1.5 might be better to allow for tree losses.  Alders, stone pines, casuarinas and poplars are all possible to include in this row.

This planting plan would equate to planting 20 large trees for the primary row and could be composed of a mix of species but not to many as the edge needs to be uniform. The shrub row, planted 1m from the tree row on the coop side of the windbreak can focus more on bee forage. Shrubs should be planted 60cm to 1m apart and would benefit from being a mix of nitrogen fixers and forage crops from the chickens, bees and humans. Closer plantings will again accommodate possible losses as the breaks establish. Hazelnuts, Acacias, Elders, Crab apples and Hawthorns would be good in this row. At a planting distance of 60cm you would require 50 shrubs.

Once the break is up and growing, a row of Ribes could be added onto the break from cuttings taken from other established individuals.

Although it has not been designed, plantings on the NE side of the sun trap would also be ideal as a windbreak and forage crops.

I would also consider extensive planting of Australian plants that will be beneficial for insectivorous birds and potential bee forage. These natives can also act as soil improvers such as members of the proteaceae (banksias) which mine phosphorus. Natives I would look into would include:

Grevillias, Hakeas, Banksias, Casuarinas, Pitosporums, non-toxic pea shrubs (Pultenaea sp.), Eriostemon myoporoides (wax flower), Myoporum parvifolium (ground cover), Sida petrophila (rock sida), Leptospurmum trivalve (inland tea tree), Kunzea spp., Melaleuca erubescen, Calytrix sp. (heath myrtle), Westringia sp., Eremophila longifiolia, Cassinia spp. And other herbaceous Apiaceae (umbels) and Asteraceae (daisy) families.

Established area

Given that there is an established forest edge to work to I feel that you should focus your initial efforts at improving this area by transplanting some of the canopy trees, planting N-fixing shrubs and fruiting shrubs that can thrive in the dappled shade and developing ground covers of dynamic accumulators and soil improvers that will feed the soil food web and also work to de-compact the soil. This area would also benefit from the addition of mulch pits behind the larger trees so that rainwater can be effectively harvested.

Most forest gardeners recommend leaving at least 20% of a tree’s diameter as the space between canopies to create a woodland garden and allow enough light to get through so the shrub layer and ground covers can flourish. However, Jackie French recommends canopies touching because the Australian Sun is so intense. I would err on the side of having the canopies touching but including a mix of N-fixers in the canopy that can feed the soil and provide shelter from frosts and winds. If the canopy is to be opened up by removing the nitrogen fixing plants this will feed the soil. Denser plantings may make harvesting harder. You could trial two areas of the food forest with different densities.

Ground prep of the eastern area should aim to widen and establish paths and then transplanting some canopy trees. A few large trees should be transplanted to the new forest edge and the chicken coop to make space. The trees that will do best on the forest edge will be the plums and the peaches. Plums tend to get larger and so should be higher up the slope than the peaches. Following thinning, mulch pits should be dug for remaining trees. This should be followed by de-compacting and feeding the soil in the beds (rockdust, aeration). A shrub layer of fruiting plants and support species can be planted and should have the best chance of establishing while the canopy trees grow. The establishment of these shrubs in this zone will provide you with a forest garden nucleus and more and more of these berries can be grown from cuttings each year.

Perennial ground covers in this area should aim to provide habitats and food for beneficial insects as well as a number of soil improvers and dynamic accumulators which will enrich the soil. This layer needs to be stacked full to prevent unwanted plants moving into empty niches. The more diverse you can make these plantings, the more bio-diverse your forest mimic will be. These plants can also be bee forage and if they establish well in the first area then they can be divided and transplanted out as the forest edge grows. However, if you just try and plant everything you can think of in each bed you will lose track of what is what and the most robust species will end up swamping the others. Each canopy tree and shrub be should only be planted with a few species and will become more diverse over time.

Around each canopy tree you can plant aromatic ground covers that help confuse pests including mints or salvias. Also good to include would be clumping dynamic accumulators including comfrey, yarrow, sorrel or tansy for example. Clumping ground covers should be planted in islands or drifts and the spreading understorey can move through the spaces and fill the gaps. Different assemblages of the following species could be used in different beds. A conscious focus on Mediterranean herbs that require less water may make the garden more robust. Many of these herbs have ecological analogs from drier climates and this will help the forest garden succeed.

Another important element of forest garden architecture are the spring Ephemerals that survive through winter as bulbs. These plants start growing early in the season before deciduous trees have leafed out and so capture nutrients and kick start the soil food web earlier in the season. The alliums also provide benefits to canopy trees by preventing disease and confusing pests. Rather than just garlic and onion there are a whole range of different kinds of onions and garlic from around the world. Find ones to suit your climate. As these grow they can be divided and spread through the forest. Bulb flowers can perform this function in a forest garden too with plants such as winter irises and daffodils.

There are many annual plants that have roles in the forest garden. Although they only live one season seed can be collected or allowed to wild sow to increase the prevalence of these species in the forest garden. Many of these annuals will be crucial during the establishment of the canopy while sunlight is high on the understory. They will prevent evaporation and feed the soil as well as providing habitat for beneficial insects.

The current forest layout has a number of areas that could be colonized by vines or trellised shrubbing crops. The NW face of the chicken coop is ideal for this and it could be planted out with deciduous plants such as raspberries and other Rubus spp. These will shade the chickens in the summer but let sun in the winter. Hardy kiwi’s could also be used but these tend to be less wind tolerant.

Fungi should be incorporated and encouraged in the forest garden soil. By placing some inoculated logs in the forest zone this should encourage mycelium to grow into the woody mulch on the forest floor and colonize the tree roots.

New Forest Edge

The area that is the Eastern reach of the garden is becoming the new forest edge as the forest garden develops. Rather than planting canopy trees straight in, the forest will benefit long term from mimicking the process of succession. This is the process where the growth of plants in an area builds up the ecosystem by mining soil nutrients, feeding the soil food web and attracting wildlife. As the system builds the conditions are right for new species to colonize the site and the form of the ecosystem develops to a new stage. This is commonly seen as cleared land – scrubland – Initial forest – Climax Forest. For a forest garden we are aiming for an initial forest successional stage as these appear to be the most productive. We call this the horizon habitat. By working with successional processes we cultivate the conditions for a healthier forest.

This area should be effectively cleared of the grasses that are currently growing in the area so they do not compete with the new plantings. This process may take 6 months to a year as the area is mulched with woven black plastic or thick cardboard. This area will be where some of the canopy trees in the established zone will be transplanted but a greater range of support trees are required to feed the soil and shelter the plants from winds and sun. Potential pioneer and support species are shown below.

When these support species are planted they could be planted around desired canopy trees or planted a year or so before the canopy trees so that the soil gets a kick start. Planting over a number of years is more successful and provides structural diversity to the forest. Further as you observe the forest growing you can tailor your strategy based on feed back.

A similar process can be undertaken for the ground covers. In the early stages of forest establishment it is crucial to cultivate and de-compact the soil and we can use plants to do this work. Green manures and other transitional ground-covers can be used effectively. These can be broadcast sown between the pioneer trees and shrubs.

As this planting plan is implemented on this patch it can be modified if necessary and expanded across the slope each year.

Road Edge

The NE section of the garden is steeper than the other parts and also bordered by the road which provides an exposed edge. Because of this the successional plan for this area is to try and convert it to a Thicket. Thickets are dominated by a shrub overstorey and have a closed canopy. Initially a mix of herbaceous plants tolerant of dry sites will be planted and also pioneers tolerant of exposure. Over time this area can be planted to thickets of hazelnuts. These are planted densely but in planting patterns that are accessible for harvesting.

Sketching out canopy reaches in 5 years time...

Successional Plan By Year

To achieve effective implementation of this design it needs to be developed over a number of years focusing on soil improvements, ameliorating the winds, lack of water and access to the site. If you work to the forest edge and engage in successional Aikido then you will be on track to growing a healthy and robust forest garden. In addition to this site I would recommend designing tree crops into lower areas of the property such as below the house dam.

Any thoughts, questions or comments on the above plan are welcome, as always. Got an extra species we should consider?

We’re running a Food Forest Workshop with Dan Harris-Pascal and Nick Ritar in September .

We’ll be installing various nucleus plantings for the above design, and working through what is required to design and implement a resilient forest garden for a range of climates. You’re most welcome to join us.

Related posts:

See the comments

Related Posts

Foraging Ice Cream Beans – Inga edulis

It's the kind of thing you might dream as a kid to find in a bac . .
Read More

Making a Nature Drawer: to help kids explore their world

I love nature tables. Collecting bits of nature and observing the . .
Read More

Permaculture Design Process with Dan Palmer: Woodend Example

So what does a permaculture design process look like on the groun . .
Read More


21 responses to “Our forest garden design: the future is forested and foodlike

  1. Be careful planting some of the species that have been recommended. Hawthorn, Sorbus and Sambucus can all be invasive in some areas.
    Some of the annuals are also invasive (eg. St Johns Wort). Chanomeles and Robinia have a very extensive and invasive root system. If you are planting this anywhere that you may be digging or scratching in the dirt – don’t plant it! You will get root suckers coming up everywhere……
    Some of these would be considered noxious or at least highly undesirable, and may upset the neighbours if they get away in a rural area. I think there would other species that would be better options.
    Otherwise, lots of good ideas!

  2. Hi guys, thanks for the feedback.

    Ria – do you get frosts where you are in Belgium? We had a -9 frost last night and it tends to keep things like nettles and St. John’s wort annual as far as I’m aware. If you’ve know of frost tolerate perennial nettles I’d love to know.

    Meg – Thanks for your words of caution, I think they should always be kept in mind when introducing new plants to an area. I prefer use the term noxious to describe poisons and chemicals that are biocides, not plants.

    It is a hard line to tread in setting up a forest system. You want pioneer plants that can survive the harsh conditions of a deforested area and establish micro-climates suitable for planting your desired tree crops (introduced exotic plants such as Apples, Plums, Chestnuts, Hazelnuts, etc). However, it is these hardy plants that can be pioneers that are often opportunistic and can become rampant. However, in the context of this milkwood food forest, the area is small enough and close enough to the house that it can be fairly heavily managed and so the spread of these species could be controlled by removing fruit before they are ripe and cutting back the plants to control growth. This cutting back serves a dual purpose of feeding the soil food web and generating mulch for the system.

    Given the rampant deforestation of the Mudgee area and all of Australia I feel it is imperative to start re-foresting as much as possible. More trees = more rain, more habitat, more carbon fixed, feeding the soil food web, more water infiltrated. When I worked for a ranger group in the Northern Territory and the rangers would go out for “Invasive species control” I always had to wonder “surly the mimosa isn’t as bad for the flood plain and wetlands (degraded by pigs and buffalo) as the poisons and herbicides being applied.” At least the mimosa wasn’t killing the frogs, lizards and snakes.

    I personally feel that any trees that are growing are beneficial and nature is not the enemy. If there was any one thing that I think would benefit the planet most it would be growing as many trees as we could as quickly as possible. It might be a little out of ecological balance in the short term and become rampent but it will get sorted out in ecologicial time. When I worked in conservation biology in Hawai’i we were in areas that had had a species extinction of 95%, this means 95% invasive species but we were still trying to remove patches of plants and plant in native species. However, because the damage had been done years ago, many of the pollinators and plant allies had also gone extinct so it seemed like a loosing battle. What is required in these areas is a new eco-synthesis where novel plants adapt and are adapted too in the ecosystem and a new stronger system is born. This has been the ecological process in the past, however we humans have sped it up as well as unleashed a disturbance regime unparallelled in our history.

    If you interested in reading about discussions of opportunistic species there is a great article written by David Holmgren in the Permaculture International Journal available on his website:

    Another interesting writer on the subject is David Theodoropoulous from California, he is a conservation biologist with a solid critique of invasion biology. http://dtheo.org/

    Thank you for your comments on the design. I’m really excited about teaching this edible forest garden course in September with Nick and in documenting the establishment and management of the Milkwood Food Forest.


  3. Hi Guys, just want to tell you I am enjoying your site – I am doing a small scale semi-self sufficient setup in Tassie and it’s great to see others going through similar process. I love that ‘food forest’ concept in permaculture where everything is interacting and lush and growing. Can’t wait to get my garden up and running – at the moment it’s 600 square metres of bare clay after the clearing. The possums around here seem especially rapacious, stripping the only 2 plants I’ve tried, a salvia and a hardenbergia. Do you have issues with possums where you are?

  4. Hello, congratulations on beginning your set up in Tassie. Great place for one, just make sure that you have enough space in the canopy to allow light to get into the understory. This will vary depending on the amount of canopy coverage that you’re looking for but a rule of thumb leave a gap about a quarter of the total canopy diameter between two trees of the same diameter. If you had two trees with a 4m spread (common for lots of stone fruit and pome fruit on nursery rootstock) then you would plant them 5 meters apart. Each tree would have ~2m of canopy and leave you with a 1m gap. Shrubs and a herbaceous understory can be planted in the gap.

    I haven’t seen any possums at the milkwood site yet. One food for them becomes available they might show up. There are roos and wallabies around though. Depending on what you want to grow on your site it might be worth thinking about netting trees or getting some really good tree guards for the first few year. Possums are less likely to eat wood plants or ones high in tannins so maybe if you start with woodier plants then the possums will be distracted by the time you add your herbs. Good luck, happy growing

  5. Just a few tips we have learned from our experience. Avoid wattles as they attract aphids and sooty mould. We had them indigenous here and anything we planted near them did poorly. We even reduced the numbers and everything under those did poorly too. We’ve just never had any success with integrating wattles into the garden.

    They just became too much of a maintenance issue that we decided to go with other natives and good old pidgeon peas. Good native plants for us have been grevileas and melalucas.

    With the grevileas, we found a mixture of taller trees (ie: honey gem) and smaller shrubs (ie: with spider flowers) made for attracting a diversity of birds. The larger nectar eaters got the larger flowers, while the smaller honey eaters, got the protection of the dense shrubs.

    Melalucas are good (especially paperbarks) because they create their own mulch. Great leaf little for reptiles and insects which inhabit the understory. Another good tree is the native peanut tree as this is semi deciduous during winter. Which allows the opportunity for sunlight to enter the canopy.

    The best native tree I have found are the ironbarks. They are more forgiving to plants growing underneath, plus they have the advantage of attracting more insects. They also tend to lose less limbs than the eucalyptus during drought.

    Other good native plants to consider are Davidson Plums, Burdikin Plums and paperbark figs. Some of those are indigenous to NSW, so should integrate successfully. Not sure if I caught it in your list or not, but other plants I wouldn’t go without are lomandras, dianellas and native gingers as understory and erosion control measures. They draw in bees and native animals like nothing else.

  6. Hi Chris, thanks for your advice. Can I ask what climate you are in?

    The Milkwood Food Forest already has an abundance of wattles growing, as yet they don’t seem to have attracted aphids but I’ll keep my eyes peeled. I’d also like to try Siberian pea shrub or some more low growing acacias. However, their role is more of a stop-gap, generating a micro-climate and then being mulched under. I know John Champagne at Brogo Permaculture Gardens has successfully integrated Acacia floribunda as a nurse plant allowing the cultivation of Avocados and other sub tropicals.

    Do you know what species the peanut tree is? I have been wanting to try some kurrajongs as they grow as a street tree in Canberra and have edible seeds.

    Left Lomandra and Dianella off the list but they’ve been a hallmark of smaller forest gardens I have designed and we’ve got a bunch of specimens we divided waiting in the shade house. They also both have edible roots in a pinch. Do you know if native ginger is frost hardy?

    I’d like to stack out the herbaceous layer with native plants from the Daisy and Umbel families but the biggest limitation I have found is access. Most nurseries only stock shrubs and flowers for native gardens with less of a focus on the herbs. Any ideas? Otherwise its into the bush for some wildcrafting.


  7. Hi Harris, we’re located in South-East Qld, in a very small pocket called Upper Lockyer. It’s like a heat sink in summer, so we get a lot of powedery mildew problems, hence the problems with wattles. Whatever grew near the acacias, just got the sotty mould problem automatically. They are such great miners of nutrients too, that anything remotely ornamental never managed to thrive. Even other hardy natives didn’t like growing underneath them.

    I’m talking specifically black wattles which seem to be indigenous to this region. I’ve heard talk around botany circles they’re thinking of changing the scientific classification of some acacias, as the genus has proven to be more diverse than first anticipated. So it’s even possible (unless Milkwood’s acacias are also black wattles) we could be discussing different genus. 🙂

    The species for the rainforest peanut tree is Sterculia quadrifida, but I’ll provide a link from the daleys fruit tree nursery. I do believe they’re located somewhere in NSW, so if you want more info they’d be the perfect people to contact. They seem to encourage a lot of enquiries from people wanting more info.

    Here’s the link: http://www.daleysfruit.com.au/bushfood/peanut.htm

    The native ginger is indigenous to Northern Queensland, so it’s not a great lover of frost. With the correct microclimate however, they can survive winter without carking it. Because they’re an understory plant, they do well under trees, which can also help provide the right microclimate. They would probably be a good addition to the Forest, once all the sweet-spots are discovered. I’m going to be trying mine in the chook yard, as I hear they are good forage for them. I’ll have to see what the chickens think though.

    Lomandra Hystrix is a great ground cover for our slopes, because we have the horrible clay soils which kill pretty much everything else. They don’t mind partial shade either. Lomandra Longifolia tends to prefer sun more.

    Herbs are great for quck yeild and cover. You can’t go past Warrigal greens (Tetragonia tetragonioides) as a ground cover. We’ve had it on our dry clay soils in full sun, and in shady spots with moisture. Some people say it has weed potential but we’ve managed to eradicate it from areas we didn’t want it, just by hand pulling. It does great in the old compost pile too.

    There are quite a few places which sell native seed though, if you’re finding it difficult to locate the particular variety you’re after. I’ve got a link to a place which sells quite a variety of native seed.


    I haven’t purchased from them before, but find the “grow notes” very helpful for germinating native seed. 🙂

  8. thanks everyone for such great comments… one thing on the acacias… as the toughest trees in the district (and nitrogen fixing to boot) I can’t see how they wouldn’t continue to be a part of our ongoing design… if it grows and performs its function, we’re interested! Looking forward to the pigeon pea however –

  9. You guys rock!!

    It’s so enjoyable and inspiring to read the living discussion between ecological engineers (permaculturalistst) who really know their stuff!! The free and generous sharing of information is indeed encouraging (just like ecological processes, … just doing what they can and do, … giving freely) Your fertile musings give hope and inspiration, … the future does indeed look forested and foodlike!!

    yay to you’s all!

Leave a Reply