WattsOFF GRID
Start where you are

Pillar 03

Grow

Soil before seed. Then the bed type that suits your ground, your water and your space.

Grow — South African smallholding landscape

The short answer

Growing food in South Africa means working with conditions most published advice ignores: alkaline clay soils, long dry seasons, extreme UV, and load-shedding that makes powered systems unreliable. Start with soil biology, choose a bed type that suits the available water, and select species that actually belong in your climate. The system that feeds you is the one that survives a bad year.

Most growing advice starts with seeds. The better starting point is what is underneath them — the biology of the soil, how water moves through it, and whether the climate will let you get away with what you are planning.

South African smallholders and backyard growers face a set of conditions that make temperate-climate gardening guides actively misleading. Rainfall is seasonal and increasingly unreliable. Soils across large parts of the country are either compacted clay, shallow and rocky, or nutrient-poor sand. Summer temperatures high enough to bolt cool-season crops arrive quickly, and hard frosts are possible at altitude in winter. The seasons themselves are the mirror image of every northern-hemisphere book on your shelf.

None of that makes growing impossible. It makes the sequence matter more. Soil comes first — specifically, the living part of it. Once the biology is working, the choice of bed type determines how water and labour behave. Then species selection, which in South Africa means leaning on what genuinely thrives here rather than chasing varieties bred for English summers.

Soil

Soil is not dirt. It is a biological system, and the gap between the two is the gap between a garden that fails every dry season and one that does not. Healthy soil contains billions of organisms per teaspoon — bacteria, fungi, nematodes, protozoa — that break organic matter into plant-available nutrients, build structure, and hold water. When that biology is absent or suppressed, fertiliser substitutes for it badly and temporarily.

Most South African garden soils start with one or more structural problems. Highveld clay soils compact under foot traffic and become waterlogged in summer and brick-hard in winter. Coastal sandy soils drain so freely that irrigation runs straight through without being used. Shallow rocky soils on slopes have almost no buffer capacity at all. The first job in any of these situations is not to plant — it is to build organic matter.

Compost and mulch

Compost feeds the biology. Mulch protects it. These are not interchangeable. A layer of compost worked into the surface introduces carbon and inoculates the soil with organisms; a layer of mulch on top keeps the soil temperature stable, slows evaporation, and prevents the UV-driven crusting that kills surface biology in South African summers. Both are needed. The common error is to use compost alone and wonder why it disappears within a season — without mulch, UV and heat oxidise it faster than it can be replaced.

Practical minimum: 100mm of compost worked into the first 200mm of soil at bed establishment, then 75mm of organic mulch on the surface, renewed at the start of each growing season. Straw, wood chip, dry leaves, and lucerne all work. Lucerne is the most nutrient-dense but also the most expensive; wood chip is slow to break down but excellent for perennial beds.

Worm farms

A worm farm is the most practical way to produce a consistent supply of high-quality compost and liquid feed at small scale. Eisenia fetida (red wigglers) process kitchen scraps, paper, and garden trimmings into vermicast that is biologically richer than hot-compost at a fraction of the labour. The liquid leachate, diluted to a pale straw colour, functions as a mild foliar feed and soil drench. See Building a worm farm that actually works for the full setup guide.

Biochar

Biochar is charred biomass — not ash, but incompletely combusted carbon that is porous at the microscopic level. Mixed into soil at roughly 10% by volume, it holds water and nutrients in the root zone and persists for hundreds of years, which ordinary compost does not. It does not feed plants directly; it is a habitat and reservoir that makes other inputs more effective. South Africa produces an enormous supply of free biochar feedstock in the form of invasive wattle and rooikrans, both of which must be cleared anyway. See Making biochar from invasive wattle for the method.

Soil amendment summary
AmendmentWhat it doesWhen to applyApproximate rate
CompostFeeds soil biology, adds organic matterBed establishment and annually100mm worked in
Worm castingsConcentrated biology and nutrients, mild feedAny time, especially transplant20–30mm top-dressed
BiocharWater and nutrient retention, long-term structureBed establishment only10% by volume
Mulch (organic)Temperature buffer, moisture retention, prevents crustingAfter planting, renewed seasonally75–100mm on surface
Liquid vermicastFoliar feed and soil drench, biological inoculantFortnightly during growing seasonDilute to pale straw colour

Beds & systems

The bed type determines how water moves, how much labour is required, and whether the soil biology you have built is preserved or destroyed season after season. There is no universally correct answer — a no-dig bed in heavy clay behaves very differently from one on free-draining sand, and what works with 600mm of annual rainfall fails with 300mm.

Flat beds

A flat bed in amended soil is the lowest-effort starting point on sites with reasonable rainfall and workable soil. It preserves soil structure if it is never dug after establishment (no-dig or no-till), relies on surface mulch to protect biology, and requires no materials beyond compost and mulch. The limitation is water: on a flat surface, irrigation distributes evenly rather than concentrating at the root zone, and there is no water-harvesting advantage built into the geometry.

Raised beds

Raised beds are the right choice where the native soil is either too compacted to work, too rocky to dig, or contaminated (peri-urban sites near industry or old orchards). They allow you to import a growing medium and control it entirely. The disadvantage is cost — materials and fill must be sourced — and in a dry climate, raised beds dry out faster than ground-level beds and require more frequent irrigation. In very hot regions, dark-sided raised beds can become heat traps in midsummer. See Raised beds that work in a dry climate for design details specific to South African conditions.

Keyhole and swale beds

Keyhole beds are circular or D-shaped, accessed from a central path, and designed so the gardener never stands on the growing surface. They are highly space-efficient and integrate a central compost column that feeds the bed as it breaks down. Swale beds are dug on contour to capture and infiltrate water rather than let it run off — a critical design in any garden where rainfall is episodic and intense rather than distributed. Both belong more to permaculture design than to standalone bed choice, and they are covered in the Permaculture section below.

Choosing between them

The article Bed types: which one suits your soil, your water and your space runs through the decision matrix in full, including how to assess clay content and drainage before committing to a design.

Bed type comparison
Bed typeBest suited toWater useSetup effortMain limitation
Flat no-digWorkable soil, 500mm+ rainfallModerateLowFails on very poor or compacted soil
Raised bedPoor, rocky or contaminated soilHigh (dries fast)High (materials cost)Needs imported fill; heat trap in summer
KeyholeSmall spaces, water-efficient growingLow–moderateModerateNeeds compost input maintained
Swale bedSlopes, episodic rainfallVery low (harvests rain)Moderate–high (earthworks)Must be on contour; fails if poorly set out
Wicking bedBalconies, urban, very dry sitesVery lowModerateContainer size limits root depth

Permaculture

Permaculture is a design methodology, not a set of techniques. The distinction matters because people often arrive at it through one technique — swales, or food forests, or composting toilets — and mistake the technique for the system. The system is about observing how energy, water, nutrients, and labour flow through a site and designing so that each element provides multiple functions and waste from one process becomes input for another.

At its core, permaculture asks three questions before any intervention: What does this site already do? What do the people here need? How can those two be aligned with the least ongoing maintenance? A site that answers those questions well is more resilient than one that is simply productive, which is why permaculture design has more to offer in marginal climates — erratic rainfall, extreme heat, poor soil — than in easy ones.

The zones

Zone mapping assigns land use by frequency of access and intensity of management, from Zone 0 (the house) outward to Zone 5 (unmanaged, observed rather than intervened in). The practical value is that it prevents the common error of placing something that needs daily attention far from the house, or something that should be left alone near it. Most residential and smallholding-scale food gardens occupy Zones 1 and 2.

Water harvesting on the landscape

Swales — shallow trenches on contour — slow water movement across a slope, allow it to infiltrate, and recharge the soil moisture profile between rainfall events. They are the single most impactful permaculture intervention in a dry climate, and they are often confused with drainage channels, which do the opposite. A swale is on contour; a drainage channel is off it. Getting this wrong floods the low end of a slope.

Companion planting and guilds

Companion planting in permaculture is more specific than the folk-gardening version. A guild is a group of plants chosen because they collectively provide what a central plant needs: nitrogen fixation (legumes), pest distraction (trap crops), pollinator attraction (flowering herbs), and ground cover that suppresses competing weeds. The most-cited example is the Three Sisters — maize, beans, squash — which is genuinely effective but was developed for North American conditions and requires some adjustment for South African climates and growing calendars.

The article Permaculture principles that survive a dry climate works through which principles translate well to South African conditions and which require modification.

Hydro & aquaponics

Hydroponics and aquaponics remove soil from the growing equation and replace it with water as the nutrient-delivery medium. The appeal in the South African context is specific: a controlled system is not subject to soil degradation, water can be cycled so consumption is a fraction of open-bed irrigation, and growing can continue through both winter and summer without being limited by outdoor conditions. The costs — in capital, electricity, and management attention — are real and worth understanding before committing.

Hydroponics

In a hydroponic system, plants grow in an inert medium (perlite, coco coir, rockwool) and are fed a precisely balanced nutrient solution. The main systems in use at smallholding scale are NFT (nutrient film technique, where a thin film of solution flows over bare roots), DWC (deep water culture, where roots hang directly in aerated solution), and media-bed flood-and-drain. NFT and DWC suit leafy greens; media beds can carry fruiting crops.

The South African complication is electricity. Any active hydroponic system requires continuous or frequent pumping. Load-shedding — even brief, unpredictable interruptions — can kill a DWC system if aeration fails for more than a few hours. Designing around this means either battery backup, gravity-fed passive systems, or accepting a higher risk of crop loss during outages.

Aquaponics

Aquaponics combines fish cultivation with hydroponics: fish waste provides the nitrogen that plants need, and the plants filter the water before it returns to the fish. The system is genuinely circular and, once established, requires less nutrient input than pure hydroponics. The management complexity is higher — two biological systems (fish and plants) must be kept in balance, and a crash in one typically affects the other.

Tilapia is the most commonly recommended fish for South African aquaponics: it is heat-tolerant, fast-growing, and not illegal to keep in a closed system. It does not tolerate temperatures below about 15°C, which limits year-round production at altitude without a heated system. The article Hydroponics or soil: which is right for a smallholding compares both approaches against soil growing across the dimensions that matter most at this scale.

Hydro, aquaponics and soil: a practical comparison
FactorSoil bedHydroponicsAquaponics
Water useHigh (evaporation, runoff)Low (recirculated)Very low (recirculated, plant-filtered)
Capital to startLow–moderateModerate–highHigh
Electricity dependencyLow (drip or hand)High (pumps, aeration)High (pumps, aeration, heating)
Load-shedding riskLowHigh (crop loss risk)Very high (fish mortality risk)
Skill requiredLow–moderateModerateHigh
Species rangeWideLeafy greens and some fruiting cropsLeafy greens; fruiting crops harder
Soil building benefitYesNoneNone (outputs can feed soil)

What grows here

The most common growing failure in South Africa is choosing varieties selected for temperate climates and then wondering why they bolt, fail to set fruit, or collapse under pest pressure. The second most common failure is ignoring the difference between summer-rainfall and winter-rainfall regions and planting to an imported calendar.

Summer versus winter rainfall

South Africa's main growing regions split broadly into summer-rainfall (Highveld, Lowveld, KwaZulu-Natal, Limpopo) and winter-rainfall (Western Cape). In summer-rainfall areas, the growing year runs from spring planting after the last frost through to autumn harvest, with irrigation carrying most of summer. In the Western Cape, cool-season crops go in during autumn for winter production, and summer is a rest or heat-tolerant-crop period. Every planting guide that does not state which regime it addresses is probably wrong for your context.

Heat-tolerant and indigenous crops

The crops that consistently outperform in South African summer conditions are those bred or selected for heat: sweet potato, cowpea, okra, amaranth, Malabar spinach (Basella alba), African eggplant, and the various Cucurbit family members — butternut, gem squash, pumpkin. Indigenous leafy vegetables including imfino (Chenopodium species) and morogo (various Amaranthus species) are both highly nutritious and genuinely drought-tolerant. They are significantly underrepresented in mainstream gardening advice and significantly overrepresented in what actually works.

Cool-season crops and frost

Brassicas, peas, broad beans, spinach, and carrots all prefer cooler temperatures and can be grown in most of the country in winter. At altitude — much of the Highveld sits above 1,500m — frost is a real constraint. Light frost tolerates most brassicas; hard frost at -5°C or below will kill seedlings. Row cover and tunnel protection extend the season without requiring a full greenhouse.

Perennials and fruit

Perennial food plants — fruit trees, berry bushes, productive shrubs — are systematically underused at small-scale growing operations that focus entirely on annual vegetables. A well-chosen fruit tree delivers without significant annual input once established, and its root system builds soil biology at depth. For most of the country, the reliable producers are citrus, guava, fig, pomegranate, avocado (frost-free areas), and subtropical stone fruits. Deciduous stone fruits (peach, plum, nectarine) require a reliable chill accumulation over winter and fail at lower altitudes in the summer-rainfall regions.

Why growing is different in South Africa

Most of the gardening literature available in South Africa was written for the United Kingdom, Australia, or the United States. It is not a minor adjustment to apply it here — the fundamental growing conditions are different in ways that change what works.

  • Seasons are inverted relative to the northern hemisphere. Planting calendars from UK or US sources are wrong by six months for summer-rainfall regions and require additional regional adjustment for the Western Cape.
  • UV intensity is exceptionally high. At altitude, a South African summer day delivers solar radiation that would be unusual in most temperate countries. This destroys surface soil biology faster than compost can replace it unless mulch is consistently maintained, and it limits the effectiveness of many plastic infrastructure materials (shade cloth, irrigation fittings, polytunnel covers) that are rated for temperate UV loads.
  • Rainfall is seasonal and episodic, not distributed. A region receiving 700mm annually may receive most of it in seven or eight intense events rather than as regular rain. Soil that cannot absorb water quickly loses most of it to runoff. Beds designed for drizzle fail here; beds designed to capture and hold water are the correct baseline design.
  • Load-shedding affects any powered growing system. Irrigation timers, borehole pumps, hydroponic systems, and greenhouse ventilation fans all fail on a schedule the grid determines. Any system that cannot tolerate interruption without crop loss is a liability rather than an asset.
  • The National Water Act imposes specific obligations. Rainwater harvesting, borehole registration, and greywater reuse all carry legal dimensions that differ substantially from the position in the UK, the US, or Australia. The default assumption that you can freely harvest and reuse water on your own property is not always correct. See the water pillar for detail.
  • Invasive species are both a problem and a resource. Wattle, rooikrans, and lantana infest large areas of productive land but also represent a free supply of biomass for biochar, mulch, and hugelkultur beds. Managing them and using them is a more useful framing than managing them and burning them.
  • The soil map is highly variable at short distances. Dolomitic ground in Gauteng and the North West requires very different bed and water management from the clay soils fifty kilometres away. Generalising across "South African soil" is almost as misleading as generalising across "African soil".

Where most people go wrong

The errors below appear repeatedly. They are not exotic or obscure — they are the standard failure mode for someone who has read widely and still lost a first or second season.

  • Starting with seeds instead of soil. Planting into unamended soil and adding fertiliser is not equivalent to building soil biology. Synthetic fertiliser feeds the plant while leaving the biology absent; without biology, water retention stays poor, structure does not develop, and the bed needs more inputs every year rather than fewer.
  • Using a northern-hemisphere planting calendar. This is not a subtle error — it puts crops into the ground at the wrong time of year by six months in the summer-rainfall interior. A cool-season crop planted in spring instead of autumn goes from germination to bolt in three weeks.
  • Building a hydroponic or aquaponic system before understanding the electricity requirement. A system that requires uninterrupted power to keep fish alive or roots aerated is not appropriate as a first growing project in a context where power interruptions are a normal, scheduled feature of life.
  • Planting what they know instead of what grows here. Courgettes, runner beans, and English cucumbers are popular at nurseries and frustrating in practice across most of the summer-rainfall interior. Butternut, cowpea, and sweet potato are reliable. The less familiar crop is often the better choice.
  • Over-irrigating with a timer and under-observing the soil. Irrigation timers set in spring and forgotten through summer are one of the leading causes of both waterlogging and moisture stress. Actual water requirements change with season, crop stage, and soil condition. A finger in the soil is more useful than a programme set in October.

Where to start

  • Before any seed purchase, dig a 300mm hole in the area you plan to grow in. Look at what is there: colour (dark is organic matter, pale is not), structure (does it clump or fall apart?), drainage (does water sit or move?), and life (are there earthworms?). This takes five minutes and tells you more than any soil test kit.
  • Establish a compost system or worm farm before you plant anything. You need a production line for organic matter, not a one-off bag from the nursery. The article Building a worm farm that actually works is the right starting point for small-scale production.
  • Lay mulch immediately over any prepared bed, even before planting. Bare soil loses biology, moisture, and structure within days in South African summer conditions.
  • Choose the simplest bed type that suits your actual soil and water availability. Start with flat no-dig if the soil is workable, raised beds only if the soil is genuinely unworkable, and never start with hydroponics or aquaponics as a first project unless the electricity supply is reliable and backed up.
  • Use the correct planting calendar for your region. Most seed suppliers provide regional guides; the Kirchhoffs and Starke Ayres seasonal guides are both calibrated for South African conditions.
  • Plant at least one indigenous or heat-adapted species in your first season — morogo, cowpea, sweet potato, or African eggplant — alongside whatever familiar crops you are trying. The indigenous crop will teach you what thriving looks like in your local conditions.
  • Read Bed types: which one suits your soil, your water and your space before finalising any bed design, and Permaculture principles that survive a dry climate before making any permanent landscape interventions.

Guides

In this pillar

Each one names its sources, and says plainly where a figure could not be sourced.

Cross-section diagram of a stacked worm farm showing the upper working crate with bedding, food layer and cover, drainage holes, and the lower crate acting as a leachate collection sump with a tap.

Grow

Building a worm farm that actually works

Two stacked crates, roughly twenty minutes a week, and it turns the thing you currently throw away into the most valuable soil amendment on the property.

Cross-section of a two-barrel biochar retort showing the inner sealed barrel containing feedstock, the outer barrel with the fire, the annular gap, and the vent path where pyrolysis gases are drawn back into the flame.

Grow

Making biochar from invasive wattle

A feedstock you are legally obliged to remove, turned into a soil amendment with a buyer — and the step everyone skips, which will cost you a season.

Grow

Starting a vegetable garden

Where to put it, how to prepare the ground, and what to plant first — for a climate most gardening advice was not written for.

Diagram showing a raised bed in cross-section with a 150 mm base drainage layer of coarse material, a 300 mm growing medium above it, a drip irrigation line at root depth, and timber or steel sides retaining the bed.

Grow

Raised beds that work in a dry climate

Raised beds are sold as the universal upgrade to any garden. In a dry climate with sandy soils and strong UV, the case is more complicated — and the irrigation bill is the first thing that catches people out.

Diagram comparing a raised garden bed with soil and a hydroponic nutrient film technique channel, showing water flow and plant roots.

Grow

Hydroponics or soil: which is right for a smallholding

Both systems grow food. Only one of them survives a stage 6 blackout without a backup plan. Here is what the comparison actually looks like in South Africa.

Diagram showing cross-sections of raised beds, sunken basins, and swale configurations with soil layers and mulch depth labelled.

Grow

Permaculture principles that survive a dry climate

Permaculture is a design system, not a recipe — but most of its published recipes were tested in Britain, the Pacific Northwest, and temperate Australia. Applied uncritically in a semi-arid South African context, several of its most celebrated techniques will drain your time and your water table before they do any good.

From the journal

Recent entries on this

Dated write-ups from the journal that touch this pillar — what is being built, tested and costed right now.

The other six

Where this connects

Nothing on a smallholding is a closed system. These are the pillars this one touches most.

The Sunday Dispatch

One email every Sunday, from the ground

What we are building, what we have sourced, and what it costs — with the supplier named and the quote dated. No affiliate padding, no doom, no filler.

One email every Sunday. Unsubscribe in one click, any time.