Rosella Waters Earthworks, Phase I, Part B

The Mushroom Dam overlooking the beach area

It’s taken a while to find the time to sit down and report on Part B of our earthworks here at Rosella Waters, near Cairns in far North Queensland. Phase I Part A was documented whilst the process was taking place. This latest update however will rely on memory and hurried notes made during the process, together with numerous photos. Large excavations such as the two large dams we constructed in part A are considerably easier to direct and far less time consuming than the finer detail work using smaller machinery as we experienced in putting in Part B.

Once again we had an excellent earthmover that came on the recommendation of the guys who did the two large dams. Sparky, as he is known, is a very knowledgeable and experienced earthmover, having spent a great deal of the last 40 odd years driving a 46 tonne excavator, building large scale dams, roads and “opening up new country”, as the saying round here goes. Now he runs a private earthmoving business and has at his disposal an 85HP bobcat and a 4 tonne mini excavator with numerous attachments. All of the following work was done with these two small machines.

The first part of the process in Part B was to construct a gabion rock wall at the very top of our system, in the gully that feeds our two dams. Previously, we had done a catchment analysis and based on the 1000mm of rain we receive per year, we arrived at a figure of 5,000,000 liters moving through it. We used this figure to calculate levels and engineer our spillways, level sill heights, the freeboard on the dam walls, trickle pipes, lock pipes, etc. The gully in question begins on our neighbour’s property. It is fed from the hill behind it and also from the diversion drains the road department puts in on the dirt road leading to our front gate. The catchment is predominately regrowth after being cleared 30 years ago with two dozers and a ball and chain. The catchment area is not a well functioning bio-diverse eco system and as such there is little water infiltration and a lot of sheet flow that brings top soil/sediment run off into our system. During the wet season of 2008 we did a small trial by hand building a rock wall just inside our fence line to get an idea of how much material would be trapped and how long it would take to fill up. After only 3 rain events, the small rock wall was fully backed up with silt 1.5 feet deep and the moisture remained just under the surface of that material well into our dry season. With that experience and the slight scar constructed at the back of the Lap Pool dam during its construction, we decided on a larger than first thought gabion, to (a) repair the damage caused by the construction of the Lap Pool dam (b) trap silt/top soil and sediment, preventing it washing through our system and ultimately ending up on the Great Barrier Reef, and (c) provide a small scale example of a solution to dry eroded gullies, that run like rivers in the wet, utilizing a “waste” product of local agriculture.

The “waste” product I speak of are the mountains of volcanic rock that many farms in the area have piled up in massive windrows. Farmers spend up to $4000 an acre to pull them out in preparation for planting avocados, potatoes, mangoes, bananas, sugar cane, etc. Rosella Waters sits right on the edge on an ancient lava flow so the farms that surround us are littered with such rocks, some as large as a car down to rocks as small as a grapefruit. We approached our neighbours up the top of the hill, who grow avocados and mangos, and who had recently put in a mass planting of new trees. Prior to that they had a 20 tonne excavator and dump truck working for a week to pull every rock out. They followed this by traversing over the land with a pickup and five workers pulling the grapefruit sized ones out by hand. Anyway, they were more than happy for us to go onto their property and select as many rocks as we liked from the windrows, which they had conveniently separated into different sizes.

Gabion rock wall trapping silt/sediment & top soil

The cost in building the gabion was therefore the time for Sparky to load up the individually selected rocks into his tip truck and then place them one by one with a claw on the end of his excavator arm. The process took two days in total and we estimate that it cost us close to $1800 to build. As we had large rocks to work with we decided against both “keying in” the base of the gabion wall into the side of the gully and constructing a net meshing to encase them in.

The volume and more importantly the velocity of the water coming down the gully in this case didn’t necessitate us doing either. Choosing the largest rocks first, we placed each one exactly where we wanted to create a firm base on which to construct the wall. It was built much in the same way as a dam wall is built, starting out wide at the base, six meters in this case, and tapering up to two meters wide at the top. The height of the gabion is nearly three meters. After placing each rock, Sparky would firm it down, swivel it around until it was firmly wedged. This in itself is more difficult than it might seem and does take time, but it is VERY important to get right. In all, the wall required 7 full dump truck loads of rock to construct. Once the main frame of the wall was complete we got another two loads of grapefruit sized rock which we have since placed by hand to smooth out the top of the gabion, thus providing a great access path across the gully that we can push a wheel barrow across, drive an ATV over or lead a goat and cart. To repair the scars at the side of the back of the Lap Pool Dam, just in front of the gabion wall, we placed some large rocks on the ledge and back filled behind the rocks with some top soil we had had set aside from the construction of the two dams. This was immediately cover cropped with cowpea and a crotalaria variety called gambia pea. All of the seed we used to cover crop was bought from a local seed merchant as seconds, which means there is a low strike rate (around 40%) but at $1 per kilo and having used the correct inoculant, we gained excellent coverage and stabilized the area. It’s important to remember that seed is the cheapest herbicide!

The next element we tackled was the overflow swale and spillway connected to the larger Mushroom Dam at the bottom of the property. We decided that after completing the gabion it would be best to start at the bottom of the system and then work our way back towards the front gate so that by the time it was all done, Sparky could load up and head off without risk of doing any damage with his machinery.

The first swale was only fifteen odd meters in length and had a level sill spillway half way along it that would spread the overflow of the system over a 3 meter wide area right on a broad ridge point, making it very safe to discharge and presenting no danger of causing an erosion gully. The construction of this small element proved to be a major turning point in our working relationship with Sparky. In the end it took the best part of a day to complete, due to a number of factors including our newfound language barrier. There were some important miscommunicated terms that needed clarification as we went: level sill spillway, back cut, swale, swale mound, swale dish, bottom of the swale dish and most importantly LEVEL. The idea that we wanted to construct something that didn’t run and was in fact perfectly level and on contour was quite a paradigm shift for Sparky, as in his words he had “spent his whole life draining landscapes” and what we wanted to do was quite the opposite.

The swale needed to be constructed on a steepish slope and we decided that we wanted it to hold 300mm of water in the base and have the top of the swale mound 800mm high – thus a substantial 500mm freeboard on the swale mound. The freeboard on the dam wall is one meter, so if ever there was a chance of water spilling over it would go over the swale mound first. It is unlikely to occur as we have “over engineered” things, but if it did the swale mound can be repaired with a shovel unlike the dam wall! What we soon discovered in constructing the swale was that due to the slope of the land we just wouldn’t have enough material to make the swale mound as high as we wished. The answer was to dig further up the hill from the back cut, as gently as possible, in a 1:1 cut. We didn’t want to dig too far up the hill so we adjusted the level of the swale mound back to 700mm high and with a three-meter long level sill spillway, the swale mound still wouldn’t be at risk.

First swale constructed leading off the Mushroom Dam

The data records for the region showed that the largest single 24-hour rain event in the last 30 years had been 284mm. We rounded this out to 300mm and built the spillway to be able to deal with ½ cubic meter of water per second. Together with another spillway on the swale connected to the opposite side of the dam wall, we have more than ensured the dam wall’s safety. Another safety margin we designed into the system was a 110mm lock pipe set at the bottom of the Mushroom Dam wall. The lock pipe is 27 meters long and goes right through the bottom of the wall. On the outlet side there is a butterfly valve, which can be opened wide in the event that the spillways aren’t coping. At the bottom of our system, and being our primary aquaculture dam, it also means we can drain this dam if needed. The dam also faces West, which is likely to be the direction of any fire entering our property, so in the event of a fire we have the added security of being able to drain 2.5 mega liters of water in that direction.

For ease of construction we built this first swale with the 85HP bobcat, equipped with a 1.7 meter wide tilt bucket. Time is money with earthworks, so we decided to make the swales a bucket width wide. Sparky started by running across the slope with his bucket following the back cut line we had marked out, corresponding to the high water mark of the dam. The spill was flicked down slope forming the first part of the swale mound. Once we had the basic shape and marked the position of the level sill spillway, Sparky used his tilt bucket to scrape beyond the back cut line up the slope to get the material we needed to gain the swale mound height we were after. We also took quite a bit of material from the area leading onto the dam wall, progressively cutting back to smooth out the sharpness of the cut. Sparky did a great job and we can easily drive through this area and up and onto the dam wall, giving us access to the other side of the property. The swale runs dead level at 300mm deep all the way through, from the exit point at the dam to the end of the swale itself. On the final scraping run we asked Sparky to tilt the blade slightly down slope in the swale dish, meaning that water will be predominated into the swale mound during rain events. With our first swale complete, fully seeded and earthmover trained we we’re ready to attack the rest of the design. Together with a mix of gambia pea, cow pea and pigeon pea we also planted sweet potato cuttings, aibika, cassava, pumpkin seeds, etc… giving us full cover leading into the wet. In the last few days we have started to receive our first rains in 9 months, so now we have a good base in which to begin our major plantings.

The next swale was a short one connected to the opposite side of the dam wall. It was constructed in the same fashion and care was taken again to ensure a smooth driveway leading off the dam wall for ease of access. With not much room to play with within our boundary line, the swale was extended right up to the fence line with our neighbours and the three-meter level sill spillway will serve as discharge of excess water into the creek below, and also as access to behind the dam wall and our Zone IV area of the property.

Moving further up the slope, we then tackled the 25-meter long swale connected to the Lap Pool dam. With this swale we had a few important decisions to make. Firstly it was going to be the Lap Pool’s only swale and only level sill spillway, the overflow from this leading to the Mushroom dam. The placement of this level sill was therefore vitally important as it would be the major source of water that fills the Mushroom dam and we also have future plans for structures connected to the 6m x 3m jetty we placed on the dam. We saw the opportunity for the level sill to be a feature and a potential wet/dry growing area, in close proximity to the jetty and eventual cabin connected to it. We decided to step the overflow down into a further two level sills before it entered the Mushroom dam.

The step down spillways leading overflow from the Lap Pool Dam swale into the Mushroom Dam. Jetty posts in waiting.

In this way, we slow down the volume of water, create further edge and add an aesthetic feature in the process. The level ditches are slightly wider than the level sill on the swale itself and together with generous amounts of cover crop seed, we planted clumps of vetiver grass to further stabilize the area and slow down water flow. We used the same technique on all the level sill spillways. With such an abundance of rock at hand and a couple of quite steep spillways to stabilize, we saw this as our best option. On two steep spillways, we planted out clumps of vetiver grass across the slope, starting at the top and offset all the way down. Then we placed rocks from the bottom up, starting with larger rocks in an arc, wider than the spillway, followed by smaller rocks all the way up the spillway wall face. We left a 200mm space around each of the vetiver clumps and now 3 months later we have a very stable, rock wall face to our spillways, with large clumps of green vetiver grass breaking up the brown.

Back on the Lap Pool swale we asked Sparky to dig ½ meter deep x 1 meter long x ½ meter wide ditches within the swale dish itself. These ditches will hold water for longer than the rest of the 300mm deep swale and as such become growing zones for some wet crops. We now have these ditches planted out with Taro, with water chestnuts on the edges, all of which is shaded by bananas growing at the inside edge of the swale mound.

Lap Pool swale with newly planted Taro and water chestnuts in the pits and banana on the edges.

Again the whole swale was cover cropped with cowpea, Gambia pea, pigeon pea and dotted with cassava, Aibika, sweet potato and pumpkins. The larger long-term support species and variety of fruit and nut trees are now ready to be planted. We had considered putting all of plantings in at the same time but with no rain at all for close to 9 months we decided to get cover crops and shorter term nitrogen fixers going and wait for the beginning of the first rains before putting them in. The earthworks couldn’t be put back to a more appropriate time due to the availability of machinery.

The rice paddy system was by far the biggest challenge. To look at now, it seems all we have done is push a little dirt up to make a wall and dig a couple of holes for the ducks to live near. In a sense that’s true, but the process of constructing the 1:300 diversion drain from the Lap Pool dam to a duck pond connected to a rice paddy (the overflow of which runs along a diversion drain with a 20mm fall over 20 meters, to another duck pond connected to another rice paddy, the discharge of which drops down into a 25 meter long bio-filter which is itself a level sill spillway), dropping water into the Mushroom dam wasn’t that simple! Plus, the overflow of the second duck pond, leads to a short swale with level spillway that drops down to a 20-meter long swale, the spillway of which also drops into the bio-filter before being discharged into the Mushroom dam. Phew.

The rice paddies with bio-filter below. The beach area is on the edge of the Mushroom Dam with the back side of the Lap Pool Dam wall behind it.

A great deal of gravel road base material was taken out of the rice paddy area and we used this to repair/construct a proper ringed access road, our main access road on the property. The road has now been graded correctly so that water will run into drains leading along side it directed to water storages. On the road we have placed 150mm x 50mm x 4 meter long blue gum planks in sets of two, 4 inches apart, at an angle across the road, every 10-12 meters. We first heard of this idea from Rainbow Valley Farm in New Zealand who has the same system on much steeper roads. As water runs over the road it only has a short distance to run before it drops down into these drains that run across the road at a slight angle. By not allowing the water to build up speed over the road surface the material stays on the road rather than down the bottom of the hill, with obvious benefits.

The diversion drain leading to the 1st duck pond needed to fall at 1:300 and be set low enough in the Lap Pool dam so that it was the first water to leave the dam as it filled. We can regulate this fact by capping the end of the 150mm pipe. The level at which we set the150mm diversion pipe was 450mm below the high water mark of the dam which also corresponds to the level of the level sill spillway. That is 150mm lower than the depth of the swale and the level at which water exits the dam into the swale. As I said, setting the pipe at that level ensures we can control when the water heads to the duck ponds. We have a 30,000 L concrete water tank connected to our shed with approximately 100,000 L of potential roof catchment. We needed to decide what to do with the extra 70,000L. In a minor brain wave, we came up with the idea to pipe the overflow through a 90mm pipe down the side of the tank, under the road and into the 150mm diversion pipe with a t-piece. At the entry point into to first duck pond, we have rocked the spill and next to the 150mm diversion drain pipe we have another 150mm pipe under the road that collects all the water in the drain running alongside the road. At the end of the drain along the side of the road we have dug a meter deep silt trap, concreted the base and placed a grill over the top. This will keep silt out of the duck ponds and provide another source of potting mix from the material that does ultimately come from the road.

The main issue we faced with the levels we were dealing with was to get the duck ponds as high up the slope as we could, leaving us room to put in the proposed rice paddies. The duck ponds would end up being quite small as a result and have a 800mm slope at the back of them from the ridge road. We saw this back slope as another opportunity to be creative and decided to step this down in 300mm wide ledges to the high water level of the ponds. The end result is a duck pond amphitheatre on both ponds! This stepped area will be fully planted out with duck habitat and forage, shading the ponds in the process.

Duck ponds at the back of the paddies, connected by a diversion drain. The amphitheatres at the back of the ponds are well cover cropped and stable.

The two ponds are connected by a diversion drain that runs from 1st pond to 2nd pond, with a 20mm fall over its 20-meter length. This isn’t a great deal of fall, but it’s enough. It has meant we have been able to keep the 2nd pond up as high a possible to give us room for the paddy below. The water from the duck ponds are released into the paddies by way of gates we picked up from an old rice farmer up here. They used to grow two crops a season using the channel that leads from Tinaroo Dam as a source of their water. One of the reasons they gave it up was when the cost of water went from $8 p/ML to $18 p/ML. Now they flood irrigate sugar cane instead. We swapped the four gates for a case of beer and made metal plates that slide into the 3mm gap in the concrete gates, to control the flow of water. The same gates are used at the exit end of the paddies, to discharge the nutrient rich water into the bio-filter below before it heads to the Mushroom dam.

The bio-filter that acts a level sill, taking nutrient rich water from the paddies as well as the swale in the background at the base of the chicken tractor system, overflows into the Mushroom Dam.

The two paddies are separated by a meter wide bund and surrounded by a meter wide, meter high bund with a slight grade. All of this will become a growing zone for duck forage, mulch and some soft fruits such as pawpaw and banana. The meter high bunds, once planted out, will become a living fence keeping the ducks in the paddies during the rice-growing season. We plan to grow rice using the integrated rice and duck growing system I had learnt whilst living with Takao Furuno and his family in Japan. Takao is a social entrepreneur with the world economic forum with his rice duck growing system and has an excellent book out through Tagari publications titled “The Power of Duck”.

The short swale connected to the second duck pond drops down into a longer swale, which will form part of our chicken tractor system. This 20 meter long swale lies at the bottom of the contour chicken runs and borders the Mushroom dam. It’ll take excess nutrients from the chicken system and grow some large trees on the north side of the dam, providing shade. Due to this swale being constructed on less of a slope than the first, it was built with the four tonne excavator. Working from the downward side of the swale, the bucket cut on the back cut line and the spill was dropped to create the swale mound. Following Sparky along with the laser we ensured that the swale dish was 200mm level all along. It doesn’t need to be within a mm but it does help to make the dish as level as possible so as to get an even distribution of water along the swale in lesser rain fall events. Obviously the best way to check that level is to fill the completed swale with water and adjust accordingly with a shovel. It is cheaper to do this in your own time than to pay $100 an hour for a 4 tonne excavator to do it.

“Hairy Harry” stands tall on the island at the back of the Keyhole Dam.

The final element to put in was the Keyhole dam at the entrance to our property. We named this pond the Keyhole, as it is the key to the system that connects water on both sides of the property. The Keyhole sits on a central ridge that dissects the property and the idea was to create a small water storage in our Zone 2 area that can move water through either the system described above or to future water storages on the river side of the property, or both. We decided how large a storage of water we wanted and marked out the approximate position of the dam wall for Sparky to follow. We set a target level for our high water and corresponded this to the position of the two swales that were to direct water to the Keyhole via 150mm pipes placed under the access road. The wall was built using the bobcat, layering wetted clay followed by numerous track rolls with the same machine. Using the excavator to dig the hole of the dam, material was mixed using the tilt bucket with me standing close by, hose in hand, making sure there was the right amount of moisture to make the clay bond. Dam and pond walls are all about compaction and with enough of the right clay, a little mixing if the material is good and bad, and the correct amount of moisture, things should seal. We decided to create a small island at the back of the Keyhole as an aesthetic feature, duck habitat and for the fact that the palm we’ve named “Hairy Harry” was too good looking to lose.

Once the Keyhole was built with a 400mm freeboard on it, we set about marking the back cuts of the two swales that were to connect to it. The Mediterranean swale (so named due to quite granite soils in that part of the property) leads out towards the header tank and drops its spill down into the Lap Pool dam. It is connected to the Keyhole via a 150mm pipe, under the road with a slight 20mm drop towards the pond so as to not get stagnant water sitting in the pipe.

The Mediterranean swale connects to the Keyhole Dam via a 150mm pipe under the main access road. The level sill spills water into the Lap Pool Dam below.

The end of the pipe can be capped, if we wish to keep water in the Keyhole dam and direct any overflow via the 150mm pipe under the road on the other side that connects the Council swale to the same dam. We called that one the Council swale because its main catchment comes from a slight improvement to the dirt road the council recently graded. It was graded sloping towards our fence with no drain so in large rain events we would get large sheet flows of water moving through the landscape causing unnecessary erosion. We asked Sparky if he wouldn’t mind creating a little spoon drain 100 meters up to the neighbours gate entrance and directing that water through the culvert under our road entrance. The five meters beyond the culvert to our fence line continued as a drain before entering our property where it then becomes a level swale directing a substantial volume of water through the 150mm pipe, under the road, into the Keyhole dam and ultimately through our entire system.

Considering the volume of material we are likely to receive from the dirt road, we placed a 200mm deep x three-meter wide silt trap just inside the fence line. This can be dug out by hand when necessary. The level sill spillway of this Council swale directs overflow to a gully, which in future may become a dam or a large gabion, subject to future test holes to check for clay content.
Either pipe in either swale can be capped to control the direction of water movement through our system. This small dam feature is something we are really happy with for its aesthetic beauty and complex simplicity in functionality.

This spoon drain runs 100 meters long and will direct a large amount of water through our system via the Council swale that connects the Keyhole Dam.

For our first major earthworks the complexity involved in the design was substantial. It was quite a big undertaking, made even more so by the birth of our second son Dylan smack bang in the middle of it all. At this point I must give special recognition to my darling wife Georgie who at 41 weeks pregnant, kept us fed and watered, took all the photos and spent considerable time standing there with FRED ( Forever Ridiculous Electronic Device) i.e. the lazer level staff and receiver, in 33’C tropical heat. We took close to a year observing the site, designing, listening and talking to others, re-designing and planning the earthworks and the immediate repair work after they’re done. Once the earthworks began, concept became reality and the two can be quite different no matter how good the planning. Each evening after Sparky had left we spent time talking things over and making decisions for the next day’s work. We gave our laser level a really good working over, it has been a great investment; I don’t imagine we could have done all that we did without it.

Now that the mainframe infrastructure is in place, a little water is in the dams and the site is green with cover crops, the system has literally come alive. From seemingly nowhere frogs have descended upon the water storages attracting ever-increasing numbers of birds. The place must look like a red-light sale at a discount store – a hydrated green oasis in an otherwise dry landscape.

Overlooking the system from the header tank. A transformed landscape.

A natural spring we knew existed has started to recharge with the water in the swales from irrigating the cover crops. It moves through the sub-soil leaking out into the side of the dam. Our hope is that this recharged system will help to keep the water level more constant in the Mushroom dam by offsetting any evaporation.

In all, the earthworks took close to two months to complete from start to finish with a total of 16 days of actual earthworks involved. With the start of our seasonal wet season rains upon us, the next three months or more will be spent busily planting, planting and more planting. We know Sparky is coming back when the wet really hits – we made a pact to sit down with a beer together in the pouring rain and watch the system operate in full flight. Through a local NRM group we are also planning an open day, for local farmers to come and see the system. These major earthworks are just the start of a great adventure in the development of our Permaculture demonstration site for the wet/dry tropics of Northern Australia, Rosella Waters.