Secret soil tanks

When we first talked about moving over here, my then-two-year-old daughter was very clear that she wanted us to plant an orange tree. I thought that sounded just fine. I highly approve of kids eating oranges straight from the tree. I had visions of a miniature home-sized citrus orchard for the kids to play in and explore. The city of Gosnells is historic citrus orchard territory, though mostly on the riverine and foothills soil rather than our sand. So it all seemed feasible. The plan went a little by the wayside though when I a) realised how bad the fruit fly was here, and b) decided to make the bulk of the garden, what’s now called zone 4, unirrigated. Citrus trees thrive on sun and warmth, they need free-draining soil and hate waterlogging, but they’re really not drought tolerant. Not if you want juicy fruit. The kind of care and maintenance that would have been involved for oranges or mandarins or tangelos wasn’t feasible for what is, and will continue to be, a rental property.

I still liked the idea of citrus trees though, and the extant lemon tree proved it could be done. The dead citrus next to it proved it could also be done badly. So after much careful thinking and investigation, I bought a lemonade lemon and a West Indian lime. And I planned their beds very carefully indeed.

One of the issues with a rental property is that when you walk away and leave it in someone else’s hands, you don’t know what’s going to happen to it. No matter that I have the landlord’s approval, any infrastructure I installed to support the trees was likely to fail, or be forgotten, or lawnmowered, or accidentally dug out, or… who knows. Any infrastructure, that is, except for one type. I’d been reading and researching a little on underground soil structures. And I decided to build in secret soil tanks around the trees.

The West Indian lime bed with tank built, tree planted and straw-mulched. One peanut is growing to the left.

They have no hardware. Nothing to rust or degrade. They are simply structures of one soil type against another, hidden underneath where the trees are growing. The “tank” is sand, cleaned of weeds as far as possible, mixed with compost, organic matter and a very small amount of clay to improve the water holding ability without affecting the drainage too strongly. Water can pass straight through the tank, it’s not sealed underneath – but the idea is that the tank should be able to absorb and hold a reasonable amount of moisture as it goes past. The tanks aren’t super deep, because citrus’ main feeder root system are fairly shallow. They do have a layer of plain sand above them, a “lid” for the tank. This is to ensure that the moisture doesn’t get sucked straight out again, to protect the improved soil from sunburn/heatbake, and also to reduce the impact of surface digging/disturbance on the soil tank structure. They’re also set slightly lower than the surrounding ground, to encourage any runoff to head into them.

I built two tanks. The first, for the West Indian Lime, I did last year. It’s not large – in hindsight I should have planned a greater area for it. The tree responded really well to the secret soil tank, and in one year has already extended its dripline out to the edge of where I put the tank in. So I kinda need to build an extension already! I planted the tree into the tank in October 2012, mulched it with straw and planted peanuts around it in the tank “lid”. Watering the peanuts by hand (growing them was an experiment to learn how which I’ll post about another time) meant the tree got extra water across the whole tank space through summer.

Midsummers Day, Dec 2012. The lime is doing very well, a bare Celtic season (six weeks) after its tank was finished. (Admittedly, that one season was the right season, the best one for citrus to push their growth in Perth.) The sand around the lime has been covered with cardboard for weed control, the peanuts are thriving.

This year it did eventually get overgrown with weeds. I cleaned the area up in November, and laid down weed mat as an experiment in controlling oxalis. The weed mat covers both the original tank and the tank boundary area that the lime is growing its feeder root system into. There is rough ex-garden-clipping-type mulch around the boundary area, and macerated dried barley as a mulch across the tank’s surface area. If I hadn’t placed those boundaries in, there’d be no visible sign of where the tank is other than a slight depression in the ground around the tree. I don’t need to maintain a visible boundary, but knowing where it is is helping me with my planning and management of the whole space.

November 2013, the lime tree. The weeds have been cleared, weed mat laid and barley and mulch laid over it. No peanuts had volunteered to regrow at the time of laying this but one has since resurfaced in a gap in the weed mat left for putting a water access point into. The area between the lime and the fence has been planted with water-rarely-needed plants of a range of usefulnesses (fragrance, edible seeds/berries/leaves, pretty flowers).

The second tank took much longer to come together. It was bigger, so a larger project. Then it got hot, suddenly, and my hours available for heavy work decreased rather a lot. Plus I was stalling a little because I wanted to put in the tank water access points. Trying to “fill” the tank by watering the surface, when you’ve deliberately built a water-protective “lid” over the top, is both wasteful and counter-intuitive. I spent a lot of time looking at options before settling on 150mm joiners for stormwater pipes, that have screw-top lids. They’re set into the ground so that their base is around tank level and their top’s above the surface, and they’re half-filled with scoria gravel so that a) filling them with water doesn’t erode around them too much, and b) they can drain more slowly into the tank so that the water can spread horizontally through the tank a little and not just straight down out of root reach. You can install them afterwards, which I still have to do with the first tank, but it’s a lot easier if you put them in during construction.

Midsummers Day, December 2012, same day as the second lime photo. I have, just ten minutes earlier, finally planted the lemonade lemon. You can see the altered soil around it that makes the body of the tank, still uncovered. The little plants to left and right behind are licorice, another legume but more drought-tolerant than peanuts, planted in built-up mounds at the height that the finished bed will come to.

I finally planted the lemonade lemon on Midsummer’s Day, December 2012. I didn’t plant it into a tank, exactly. I planted it into where the tank would be when I finished it, just to get it into the ground so it could survive summer. I did get the soil structure altered so the body of the tank is there, but I didn’t put any lid or mulch over the top. So there was one tiny tree in the middle of this big hole in the ground, with (eventually) a couple of stormwater joiners sitting perched on top of the altered-soil around it. It didn’t thrive nearly as well as the lime. But it lived.

Finishing the second tank has been one of those jobs I kept putting off all year this year. I’m glad I did. The first tank I sifted all the sand back in, to reduce the weed seeds. That was hot, hard, heavy work and I really hadn’t been looking forward to doing the bigger tank. But then it turned out that there were just as many weeds in the sand I hadn’t dug out as there had been in what I did dig out. So I decided the sifting was reasonably unnecessary. That saved a lot of work and I finally finished the tank last week (December 2013).

After clearing the masses of head-high weeds out, I began the lid. I mixed a new load of compost and clay in with some of the stockpiled sand, so this lid is not just plain sand but an extra waterholding layer. The three grey lidded pipes are the water access points, with their base sitting on the top of the original tank layer.

Weed mat laid over the tank lid in an effort to control next year’s oxalis crop. This will be hindered slightly by my inability to accurately cut a hole where the tree was.

The finished bed, two weeks short of a year after planting the tree. Mulch has been spread across the tank lid, and macerated barley laid across the tree pit. The water access points are operational and now in use.

The lemonade lemon is already looking a lot happier. Of course, because the tank is underground I can’t see if the watering points are working properly. But I’m not seeing any problems yet, so I’ll just have to watch the tree and see how it does.

Advertisement