First Steps to a Food Forest in the Southern Tablelands

by Dominick ter Huurne & Inke Falkner

“We acknowledge and pay respects to the Dhurga people of the Yuin Nation, the traditional custodians of the Country where we live.”

Having found the bush block we had long been searching for, a protracted settlement period gave us plenty of time to decide exactly what we wanted to do with it. At 40 hectares the property was much larger than we had ever envisaged buying, but we fell in love with the diversity of wildlife and vegetation, seduced by the possibilities it offered. Establishing an orchard was a major priority, and having recently been introduced to permaculture gave us a chance to put many ideas into practice. So, armed with a lot more enthusiasm than experience, this is how Inke and I began the transformation of one small pocket into a food forest.

The property is located in the NSW Southern Tablelands. It is approximately 80% woodland, with a mixture of box gum, scribbly gum, and melaleuca, with large open sections of Kunzea parvifolia. Rainfall is fairly evenly distributed throughout the year, averaging around 700mm annually. The land is undulating, with a small creek from the south-west feeding a dam before continuing through the property. There is quite an established ecosystem in and around the dam, with water dragons, wombats, roos and wallabies, echidnas, a thousand frogs, yabbies and fish all thriving. Black cockatoos, gang-gangs, galahs, rosellas, cuckoos, wrens, ducks, herons, kookaburras, forest kingfishers, and a pair of whistling kites are regular visitors.

A small area along the south western boundary was the obvious site to establish a food forest (shaded in yellow in the above photo). It was mostly cleared — there remained some mature and dead trees that could provide shade and shelter for the fledgling orchard as well as providing habitat for wildlife — and it was close to the creek, and runoff from the road could be easily harvested for additional water. We would also be managing this on weekend commutes down from Sydney, so this was as large as we could comfortably go at this stage.

Contour map of the site with contours at 10m intervals. Average elevation is
690m above sea level. The land is flat near the road, but becomes progressively
steep towards the creek in the south-west (left) corner,
where it forms a small gully.

There were two obvious physical constraints on the site. The first was the existing wattles in the centre that we mostly wanted to retain, as they would provide shade and shelter for seedlings that required extra protection, as well as serving as a partial wind break, and secondly, the granite that is common in the area was quite pervasive in the northern half, which we naturally wanted to avoid. This left an area of approximately 80m x 60m to play with. The site is quite flat near the road, becoming steeper towards the creek with a slope of approximately 12%.

Food forest site looking north from the road

Food forest site looking south from the creek

By this time Inke has just completed the PDC in Melbourne in September and we felt more confident about what we wanted to achieve. We decided to keep things fairly simple to start with. Given the existing vegetation, the slope of the land, and our desire to fit in as many fruit trees as possible without compromising the system, we aimed for 3-4 swales spaced a minimum of 6m apart at the valleys. So, armed with our bunyip level, we set off to mark out the contours at (we later realised) very fastidious 2m intervals.

Inke with the bunyip level

What a wonderful way to get to know your land! It takes a while to get started but you quickly learn to appreciate how varied the land really is. Having decided to remove a few trees, the end result was three longer swales around 75m each, and another three shorter ones of 40m.

With everything in place we were dying to have an overview of our work, but without a GPS we couldn’t think of how to map it out. Finally we printed out a Google map of the site, and resorted to a tape measure and triangulation to mark the co-ordinates of each stake. Then, using a Google map overlay in Sketchup, we were able to approximate an aerial view of the orchard, using program tools to then calculate the catchment area for each swale. We knew it was rough but we felt like any analysis was better than none.

Food forest created in Sketchup. Swales in brown (although swale width
not drawn). Existing trees in green, and catchment in blue.

Note that the swale width (shown in brown) is not to scale, this just shows their location. The distance marked between swales is therefore inclusive of the swale itself. If we had drawn the swale width to scale then we would have realised that some of them were too close together, which led to a few complications later on.

Now we had to decide just how deep to make the swales. For a newcomer to permaculture this was something we agonised over a long time. On the one hand permaculture is very general – just put in some swales and plant into the mounds; easy — only to be followed by specifics that we could hardly imagine possible: slopes of no more than 1 in 6000, how could this be? Finally we decided it didn’t matter too much, as long as we tried to back up our reasoning.

Looking back through the rainfall statistics we noted that even though the heaviest rain event might not be much more than 60-70mm on one day, this was sometimes preceded and followed by one or two days with 10-30mm.

So, for example, based on a swale width of 2.5m with a 45 degree back cut and 1m flat bottom, length of 30m, a maximum rainfall event of 80mm, and a catchment of 500m2, then the swale depth would be roughly 760mm. With a further reduction for infiltration, we settled on a depth of 500mm. Was this right? Who knows, but it was exciting stuff.

The time had come to get the excavator in. By now we had been planning everything for three months and it was mid-October. We called the operator, set a date, and then it started raining. Worse, NSW was flooding and the works were on hold indefinitely. The rain halted the work for two months, until at the last moment there was a break in the weather in the week before Christmas.

The big day finally arrived and Christmas had indeed come early. It was incredible how fast the whole event took place. Forget about ever thinking you will save money by doing the job yourself, a proficient operator is worth every dollar. We watched proceedings for the first couple of hours, making sure everything went to plan.

Unfortunately we had to leave the site for a couple of hours. By the time we returned, the operator was on the fourth swale. All was fine, with one exception: the granite outcrops extended further than we had thought. Even though we had dug a few holes before commencing excavation to check this, we were not thorough enough. The result was that the lower two swales were split into two sections divided by the rock line. The problem was that the swale in the rocky section was only 30-40cm deep, with no chance of being level given all the rocks, while the remainder of the swale continued at 50cm, since the excavator stuck to the original design instructions. Had we been on site, we may have altered the design of these swales to the shallower dimension. With the only options now being to fill the deeper sections back in or leave it as it was, we opted for the latter. We decided to effectively maintain a dual system in these two swales, creating a small wall between the rocky and non-rocky sections so that water did not all flow to the deeper areas. This way with light rainfall the two sections would still receive roughly the same amount of water, with the swale operating as one only if it rained heavily.

The other bonus was that we now had a lot of rocks with which to build a pizza oven.

Hauling rocks from the swales. Useful material for the next earthworks

The final stroke of genius from Inke was to create a drain at the top of the property diverting water from the road into the first swale. As you can see in one of the pictures below, this two minute operation proved to have the most significant effect on the site, effectively doubling the amount of water harvested for the second swale.

Overview of the bottom four swales

With rain looming again, we were eager to sow the first seeds. We chose Japanese millet and cow pea as our cover crops with another 60 odd species of plants for support. That was the easy part. Next we mulched each swale with a straw/grass mix, using approximately one bale per 30m2. Tough work but nothing like impending rain to drive you on.

Seeded and mulched swale mounds

Then, finally, a week later, we were fortunate enough to be able to test out the works. Commissioning earthworks as a beginner is exciting but nerve racking. You have altered the landscape drastically, so it had better be right. Within two weeks of completion a large storm was heading our way, arriving early one morning, dumping 20mm in the first hour and a half, and 50mm altogether that day.

We ran out to check the swales first thing and could hardly believe it. The theory actually worked. Every swale had at least 150mm of water and rising. All except the top swale that was catching the road runoff. Here the water was gushing in, creating a mini waterfall as it entered the swale. The water depth was already 30cm and would rise to 40cm by the evening. One thing was for certain, we needed more of these small diversions!

Swales after 20mm fell in one and a half hours

Runoff from road. Inke is standing behind the dam wall that diverts
water into a drain leading to our second swale

Runoff pooling in the drain before it enters the property

Waterfall created by road runoff into top swale

The second swale filled to twice the depth of the other swales
due to road runoff channelled in.

We were hooked and immediately began planning the next set of works, walking everywhere looking for water patterns in the earth. We would build a much more comprehensive system of dams and drains, spillways and berms, not wasting a single drop. Clearly this was a longer term project, requiring a lot more reading and observation, not to mention a clearer picture of what we wanted to achieve. Brad Lancaster has quickly become a trusted friend.

Of course not everything was perfect with our new swales. Some levelling out was required and we need to plan our spillways for a major downpour given the extent to which they filled with just 50mm.

The other issues we have, and would be grateful for any feedback on:

1. The water took longer to infiltrate than we had thought. We had 50mm the first day, 7mm the second, 5mm the third, and drizzle for another 2 days. It was a further two days, i.e.six days after rain commenced, before all the water was gone. Was this due to the volume of rain? I cannot say how long it takes for water to infiltrate with a small rain event as I have not been able to observe one yet.
2. The soil has a high clay content commencing 10cm below the bottom of the swale. Could this be impeding infiltration and would it have been better to make the swale shallower for this reason?
3. Are there any other issues if water infiltrates very slowly? How long until mosquitoes start to breed?
4. Can swales be too deep? The only real drawback we could think of was that the water infiltrates far below the root level while plants are young but this would be the case for most small rainfall events. Any others?
5. We thought about seeding the swale itself with a cover crop to enhance infiltration, but also to prevent the grass from growing back, perhaps with a clover and lucerne mix. Is this a good idea?

Finally, here are our tips and lessons learnt:

1. Ask your excavator how they are going to make sure it is level. At least this way he might understand exactly what you mean. A lot of operators have laser levels for when they excavate house foundations. Make sure they bring it with them.
2. It’s amazing what you can do with Google maps and Sketchup. Sketchup is extremely simple to use, it will take half an hour to learn and you can download it for free. You can import a Google map straight into Sketchup and then draw your earthworks, food forest, garden, house site, etc., straight over the top of it.
3. The small incisions you make are just as important as the large ones. Capturing the road runoff was literally two minutes work but had a greater impact than anything else we did.
4. It’s only hard work if you don’t enjoy it. And if you don’t enjoy it you probably won’t be doing this.
5. Read volume two of Brad Lancaster’s Rainwater Harvesting series. The calculations in there are extremely basic and will give you the confidence to believe in what you thought of as common sense anyway.
6. When using a bunyip level, take a pipette with you. It’s much easier to refill the hose with a small funnel than pouring from a water bottle.
7. Neither Inke nor myself are horticulturalists (although Inke has just enrolled in Horticulture at TAFE). We are, however, very practical people. Inke is a biologist and part-time florist, and I am a cabinet maker, and together we have a lot of combined experience outdoors and on building sites. Use your knowledge and apply it where you can and talk to as many people as possible. If you are not aiming to make a commercial venture from your works then I do believe you can plan a lot of your food forest yourself. If you have ever levelled floor joists before laying a timber floor, you can easily mark swales on contour.

The results 10 days after planting

And just because I couldn’t resist, here’s a picture of our new toilet. It was built in our cabinet-making factory from timber off cuts I couldn’t use for furniture, and then assembled onsite.