Biological CleaningIrrigationLandSoil ConservationStorm WaterSwalesWater ConservationWater Harvesting

Experimenting with Overflowing Circles and Slow-Flow Swales (Panama)

The jungle garden

I am not Bill Mollison or Geoff Lawton, they will both happily report; rather, I am but a humble novice when it comes to permaculture, experimenting my way through ideas, mimicking when I can, improvising when research falls short. And, it was somewhere in between mimicry and improvisation that I came up with what I’m calling overflowing circles and slow-flow swales.

I wanted to catch water, of course. That was the number one objective. In Panama, the rains seem to come in mad fits — bursts of hurricane-like downpours that would rip through the garden and food forest, streaming into the nearby lake and leaving nothing but debris behind. Panama is a country of two seasons, which are referred to as summer (dry, November through April) and winter (wet, May through October). Knowing summer would eventually come made watching all that winter water rush away even harder. What a waste!

I had to do something, but due to factors like location, gradient and budget, this was going to have to be a shovel and pickaxe mission. It was going to require some rainy romps around the property. And, due to the quick and massive influx of water, surely there were going to be some unforeseen issues. It sounded like a blast.

Starting with circles

At first, I was really hesitant to cut into the steep hillside slope leading down to the garden forest, so I started with magic circles. This, of course, consists of digging a hole about a meter deep and roughly two meters across, piling the soil in a ring around the hole. It’s perfect for hungry papaya, plantain, and banana trees, under which I planted sweet potatoes, taro, yucca, peppers and so on. By putting these close together, just a walking path between them, it created a dense piece of forest.

High-end overflow magic circle

I knew it wasn’t going to stop the drainage problem, but I felt sure the holes would do a good job catching water and, at least, storing some underground. They worked really well. They topped up to the brim during storms then the water dispersed beneath the surface in the meantime. But, less than two months into the rainy season, the circles had had their fill. One circle’s soft soil wall washed away, taking a sapling with it and exposing a good swath of earth that had been mulched.

I had never guessed the circles would fill this way, so I had to come up with an overflow system to handle so much rain. I tried several ideas, all of which seemed to work.

  1. I left a gap. I knew the nutrient-rich water was an important aspect of the circle, but I figured it was better I direct the excess to something useful. Where the circle’s wall had washed away, I let the hole remain. on the low side of the circle, adding an overflow channel that led to a different circle.
  2. The circle that was catching overflow was new and unplanted, so I included an inlet and outlet pipe at ground level and just piled soil atop them. It meant that this circled filled up quickly but the walls didn’t burst.
  3. The other idea was leaving a ground-level gap on the high side of the circle so that the inside water level got high enough to soak into the bottom of the loose soil ring, with any possible overflow draining before it could wash away any walls. The gap ended up being a nice feature because it also allowed easy access for adding compostable stuff into the circle.

Water intake for soon-to-have papaya circle

The small swale experiment

When it came to swales, I knew I had a few problems. As the circles had demonstrated, the water sometimes came so quickly that it would take a massive swale to catch it all, but again, this had to be done by hand. The forest garden had a slant, but the hillside where the water built up momentum was steep, even difficult to clamber on. Then, I read about small swale systems, which, even if I couldn’t put in a Lawton-esque river of a swale, still seemed beneficial.

I tried one on the hillside — a trench about five meters long, a little less than a meter wide, and maybe 20 or 30 centimeters deep. It was just on the outside of the normal drainage line but still filled up easily during storms. Knowing this would be the case, I’d created a strategically placed dip into the retaining wall where water could overflow safely away from a young macadamia nut tree. It worked a charm, but of course, I was still watching the bulk of the rainwater whisk away.

The test swale and its overflow

Then, I stumbled upon an article about water gardens in which shallow ponds interacted with one another, filling up and leading into each other during good rains but essentially disappearing during the dry times. The information couldn’t have come at a better time. The overflow drainage from my circle systems had just done a bit of damage in the food forest. I was ready for another experiment.

Disappearing ponds

In order to catch more water, I decided to install one of these shallow water garden ponds where my magic circles were overflowing. I dug an area of about three or four misshapen square meters (so that I’d get more edge out of the situation). From the nearest circle, I dug a little channel to direct the water, and from the pond, I made yet another planned escape route for excess rains.

Frog pond excess water catchment (at the end of a dry week)

Unfortunately, I didn’t know what edible crops would grow well in this circumstance, but I’d noticed that local a flower, the platanillo (similar in appearance to a bird of paradise), grew really well in a similar space. The red flowers attracted hummingbirds and other wildlife, brightened things up, and protected the soil from sun and erosion, so I figured it’d work well. It did. The whole system performed yet again.

With all of the safety channels, I wasn’t getting every last drop of water that fell, but without a doubt, I was collecting a massive amount, all of which would have been lost before. I figured I could combine the ideas I’d been playing with to catch more and more water. It had me feeling confident. It was time to tackle that hillside.

The slow-flow swales

The idea behind my slow-flow swales were that I was never going to be able to create something large enough to catch all the water that falls in a tropical deluge, but I wanted to do something to harvest it. I thought the easiest way for me to do this would be to make a series of smaller, connected catchments — holes and elongated swales — that would disperse the water over the hillside and through the garden before the excess was allowed to drain into the lake.

For the swales and holes, I dug anywhere from twenty centimeters to a full meter, depending on the size and location, below the lower ground-level lip of the hole. I used some of the soil as a raised planting barrier along the bottom side of the swale and the rest on other projects (hugelkultur, raised beds and composts). Then, I connected each swale with the others using overflow trenches, leaving a small section of the swale wall slightly lower than the rest so that, when water levels rose too quickly, it would drain into the next swale, lower down, rather than cause damage.

Slow-flow swales on the hillside

It all worked with surprisingly little to adjust. In the midst of some pretty substantial storms, I watched each level slowly fill and move onto the next lower catchment until ultimately the water reached converging channels leading to a frog pond, which overflows into channels leading to the magic circles which spill into disappearing water ponds in the garden that ultimately allow anything more to reach the lake. The entire garden forest is now absorbing an abundance of water, spread out fairly evenly.

I’ll admit that this project ended up being a lot of work, but I also must say connecting all the dots turned out to be a blast. I’ve got water canals running all over the place, so the garden forest should be very fertile very soon. Now, I suppose we’ll see what happens come April, when dry season is in full swing. I guess it’s time to get to work on a grey water system. Until then….

Further Reading:

Jonathon Engels

The financially unfortunate combination of travel enthusiast, freelance writer, and vegan gardener, Jonathon Engels whittled and whistled himself into a life that gives him cause to continually scribble about it. He has lived as an expat for over a decade, worked in nearly a dozen countries, and visited dozens of others in the meantime, subjecting the planet to a fiery mix of permaculture, music, and plant-based cooking. More of his work can be found at Jonathon Engels: A Life About.


  1. I really like this! Working with high water tables here in Oregon, in preparation of the coming rains I started feeding chickens and ducks and throwing compost into a shallow pond that will overflow to swales and yard. Glad to have read this.

  2. Thanks so much for sharing your experiments with us. Have been following your project for a while now and really enjoying how it is unfolding…

    1. Hi Sheri, I’m so disappointed to say that I’ve been unable to find the site where I got this idea, else I would have linked it. The basic idea was to design linked shallow pools, which looked more or less like a normal depression in the ground, that fill and absorb the water before it has a chance to evaporate. Sorry not to be able to point you to the exact spot. I’ve searched hours for the source now.

  3. great to see the work you are doing….we have similar problems here in Trinidad…we have 30 acres that we have been working with for the last 17 years….we have quite a few ponds now that we have built with excavators but we also have many swales and disappearing ponds we have built by hand in areas we cant access with machinery…i had a question about the mini swale you had dug in your last picture…is it on contour? it looks like it is running up and down the slope on one section which may give you problems later….you are doing a good job slowing down the water…keep it up

  4. I built a series of shallow pools (not as steep as yours) that get larger in size down slope to slow the flow further. My system is also connected to a miniature stepped chinampa that also captures nutrients from a swamp cachement. Excellent work and interesting story!

  5. Hey, Thanks for sharing your ideas! Gave me lots to borrow/steal from. I am currently trying to PC an existing Plantain monoculture in Colombia that has a crazy ampitheter slope to it. I started with some raised beds and hueggelkultur beds near the top, and I’m about to try retrofitting plantain circles/triangles/etc onto the existing plantains, then tte y my luck with some swales and finally some ponds.
    J see these posts are 2 years old, so I will try to find recent photos and info on your project, but I learned some really valuable tips, thanks!

Leave a Reply

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

Related Articles

Back to top button