Rainwater Harvesting for Gardens: Swales, Tanks & Passive Systems

The first thing most people do when they start taking water seriously is buy a tank. A big plastic barrel goes under a downpipe, and they feel like they've solved something. Then comes the first dry spell. The barrel runs out in three days. They're back to the hose, and the project feels like a failure.

Water management in permaculture isn't primarily about storage. It's about slowing, spreading, and sinking water into the landscape before it runs off. Storage is the last resort, what you reach for when the land itself can't hold any more. Design the land first. Then add tanks.

I've built water systems on clay hillsides, sandy flats, suburban plots, and everything in between. This is what I've learned about doing it in the right order.

Why Water Management Is a Permaculture Priority

Bill Mollison's instruction was simple: catch water as high in the landscape as you can and hold it there as long as possible. Every time water runs off your property, it takes topsoil, nutrients, and opportunity with it.

Most suburban and rural properties are designed to drain quickly: gutters, paved surfaces, compacted soil, shallow-rooted monocultures. Rain arrives, runs off, and is gone within hours. The result is a boom-bust cycle: flooding after rain, drought within days. The plants suffer. The soil suffers. You spend more time and money on irrigation than you should.

Permaculture water design reverses this. The goal is to build a landscape that behaves more like a sponge than a drain. When you achieve that, rainfall events top up a reservoir that your plants draw from for weeks afterward. Your irrigation need drops dramatically. Your soil biology flourishes. Your garden becomes genuinely resilient, not just to drought, but to flood.

Water also connects directly to zone and sector planning. Where water naturally flows through your property is a sector you need to map before you plant anything. I covered the full zone and sector analysis process in my permaculture zones guide. Read that first if you're starting from scratch.

Passive Earthworks: Slow It, Spread It, Sink It

Passive earthworks are changes to the land itself that change how water moves. No pumps, no pipes, no power. Just contour, slope, and gravity.

Swales

A swale is a level trench on contour. Not a drainage ditch. A drainage ditch channels water away as fast as possible. A swale is the opposite: it catches water moving down a slope and holds it there until it soaks in.

The critical word is "level." A swale dug off-contour becomes a drain. Every metre of swale must be at the same elevation so water spreads evenly along its length instead of flowing to one end. You need a dumpy level, a water level, or an A-frame level to get this right. Don't guess.

Swales are typically paired with a berm, the mound of soil excavated from the trench, placed on the downhill side. The berm slows water further and becomes your most fertile planting zone. Water-loving species go in and around the swale itself. Fruit trees, shrubs, and deep-rooted perennials go on the berm, where they'll tap into the moisture reservoir the swale creates.

Sizing a swale depends on your catchment area and rainfall intensity. A rough rule for most temperate climates: one metre of swale length for every 10 square metres of catchment above it. For high-intensity rainfall climates, be more conservative. A swale that overflows and erodes is worse than no swale.

Overflow matters. Design a spillway at each end of the swale that directs overflow to the next swale downslope, or to a stable outlet. Swales without designed overflow become erosion risk points.

Berms and Diversion Banks

Where you can't build a full swale, a diversion bank slows and redirects water without holding it. A gentle berm across a slope at a slight angle (not level, but close) channels runoff toward a swale, a pond, or a planted basin. Useful for redirecting roof runoff across a lawn before it reaches a downpipe.

Rain Gardens

A rain garden is a shallow depression planted with species that tolerate temporary flooding. It's designed to catch a specific volume of runoff, typically from a roof or driveway, and let it soak in over 24-48 hours.

Unlike a swale, a rain garden handles intermittent concentrated inputs rather than sheet flow across a slope. Direct your downpipe into a rain garden and you've taken that roof runoff out of the drainage system entirely.

The key is sizing. Measure the roof area draining to that downpipe. A rain garden should be roughly 20-30% the size of the catchment area it serves, with a depth of 15-30cm at the deepest point. The inlet side (where water enters) gets the heaviest, most flood-tolerant plantings. The outlet side (a gravel-filled overflow channel) handles the rare event where the garden fills completely.

Soil preparation matters here more than anywhere else. Rain gardens fail in compacted soil. Break up the base with a fork or broadfork, add compost, and if you're working with heavy clay, a layer of coarse sand at the base improves drainage rate.

Sheet Mulch and Soil Improvement

Every layer of compost and mulch you add to your soil increases its water-holding capacity. Clay soil at 5% organic matter holds significantly more water than clay at 1%. Sandy soil improved with biochar and compost retains moisture it would otherwise lose. This isn't glamorous water management, but it's the foundation. Sheet mulch over bare soil is a water intervention. Every food forest guild you plant is a water intervention. The deep root systems of comfrey and chicory break up hardpan and channel water downward. I covered the soil-building logic in detail in my hugelkultur guide, which explains how woody biomass underground creates a long-term moisture reserve.

Active Systems: Tanks and Barrels

Once you've done everything you can with earthworks, tanks give you a buffer for dry periods when the soil reservoir runs low.

Sizing Your Tank

The most common mistake is buying based on available budget rather than actual need. Here's a better starting point.

Calculate your catchment first. Your roof is a catchment. Multiply the footprint area (not the slope area) by your annual rainfall. For a 100m2 roof receiving 700mm of annual rain, that's 70,000 litres per year flowing off that roof, less than half captured by most first-flush systems.

Calculate your demand second. An established vegetable garden needs roughly 20-30 litres per square metre per week during summer. A small orchard needs more. Add it up.

The gap between captured supply and peak summer demand tells you how much storage you actually need. Most suburban setups land in the 5,000-20,000 litre range. Bigger isn't always better. A smaller tank that fills and empties regularly works harder than a huge tank that rarely fills.

First-flush diverters are worth the small additional cost. The first few millimetres of rain wash dust, bird droppings, and debris off the roof. A first-flush diverter catches this contaminated water and discards it, sending only the cleaner subsequent flow to your tank. Essential if you're growing food with captured water.

Placement

High is better. A tank elevated 1 metre above your garden delivers gravity-fed pressure. A tank at ground level needs a pump to deliver useful flow. Even 600mm of elevation gives you workable drip irrigation pressure without any power. Where possible, site tanks on a berm or raised pad.

Position tanks close to where the water will be used. Moving water costs energy, either your own carrying it, or a pump moving it. A smaller tank near the vegetable garden is more useful than a large tank at the other end of the property.

Interconnected Systems

Multiple small tanks linked together are often more useful than one large tank. You can place them at different points in the garden, at different elevations, and chain them so each one fills before overflowing into the next. Tanks of 1,000-2,000 litres are easily moved, repositioned, and connected as your system evolves.

Planting Along Swales

The berm on the downhill side of a swale is the most productive planting zone on a sloped property. The swale feeds it from above; the berm's slight elevation prevents waterlogging. Fruit trees, comfrey, elderberry, goumi, gooseberries, and almost any food-producing shrub do exceptionally well here.

The swale itself, and the area immediately uphill, can handle more water-tolerant species. Willows are classic swale plants. They take up enormous amounts of water, their roots stabilise the earthwork, and they produce abundant biomass for mulch and stakes. Sedges, rushes, and moisture-loving herbs like water mint fill the swale bottom.

For the berm planting, the same guild-design principles apply as anywhere else: nitrogen fixers, dynamic accumulators, ground covers, and pollinator plants supporting the central food producers. The fruit tree guild guide covers how to structure that. Swale berms are where I've had my best results with comfrey. It thrives with its roots reaching down toward the moisture reservoir the swale creates.

One thing that's easy to overlook: swale systems change over time. In year one the earthwork is bare and raw. By year three, the berm plants are establishing. By year five, you have a productive ridge running across your slope. Log observations each season. What's thriving, what's struggling, where the water is going. The system tells you things if you watch it.

Common Mistakes

Building before observing. Watch how water moves on your property through at least one full wet season before digging anything. Where does it pool? Where does it run fast? Where do existing plants go dormant earliest in dry periods? The land will show you where the interventions belong.

Digging swales on too steep a slope. Above about 15% slope, the amount of earthmoving required becomes impractical and the erosion risk during construction is high. On steep sites, work with contour-planted strips and diversion banks rather than full swales.

Forgetting the overflow. Every water-catching structure needs a designed exit for water that exceeds its capacity. An undersized swale that overflows without a spillway will cut a gully in days. Plan overflow before you dig.

Planting the wrong species in swale bottoms. Fruit trees are for the berm, not the swale. Most fruit trees do not tolerate standing water. Roots sitting in saturated soil for even a few hours during flowering or dormancy can cause significant damage. Use water-tolerant species in the wet zone; save your high-value plants for the berm and beyond.

Disconnected tank systems. A tank attached to one downpipe but not integrated with the rest of your water design captures a fraction of what's available and doesn't address peak demand. Connect your tank to your swale overflow, your grey water system, and your irrigation distribution as a single network.

Over-engineering too early. I've seen designs that call for five interconnected swales, three ponds, a reed bed, and a rooftop cistern on a quarter-acre suburban block. Build the simplest possible intervention first. A single swale, or a rain garden from one downpipe, or one tank. Watch how it performs. Add complexity only where the simple version proves insufficient.

Building Your Water Map

Before any physical work starts, sketch a water map of your property. Mark the high points and low points. Draw arrows showing where water currently flows after heavy rain. Mark the existing impervious surfaces: roof, driveway, paths. Note where it pools, where it runs fast, where it disappears.

Then design your intervention sequence from the highest point down. Catch it high first. Slow it on the slope with swales. Let it soak into planted berms. Capture the overflow in tanks. Direct the tank overflow to rain gardens. Handle the last of it in a pond or mulched basin.

Every property is different. The pattern is the same.

If you're designing your garden in PatternBase, the sector analysis layer is worth using before you plan any earthworks. It helps you map water flow, shade angles, and wind corridors as a foundation for the rest of the design. Getting water right early means every other system you layer on top of it will perform better.

Water is the first design priority in dry climates. In wet climates, it's drainage and soil absorption. In every climate, it's the thing to get right before you plant anything you actually care about.

---

PatternBase is a free permaculture garden design tool launching March 2026. Create your free account to start documenting what actually grows.