All techniques › Water harvesting & retention
Ollas & Subsurface Drip Irrigation
Delivering water below the soil surface, where it cannot evaporate. Strong evidence that it helps your farm, and equally strong evidence that it does not help your region.
The short answer
This works, and the evidence is good. Subsurface drip saves around 55% of water against furrow or flood irrigation. Maize yields rose 31% with nitrogen efficiency up 43–71%; tomato reached roughly 1.5× yield on 30% less water than furrow. In Spanish semiarid olive groves, buried drip beat surface drip on both yield and water-use efficiency.
But efficiency is not the same as saving. Spanish drip irrigation expanded 76% between 2001 and 2019 while total consumptive water use rose, because the saving was spent on more hectares and thirstier crops. Drip will help your farm. It will not, by itself, help your basin — and if you bank the saving as extra irrigated area, you have made your region's drought worse.
Water rights, not equipment, are the real long-term risk. A buried system lasts 12–15 years and is worthless if your abstraction gets curtailed. Check your basin's status before you invest.
Read this first: efficiency does not equal saving
The most important thing on this page is not about your soil. It is about what happens when everyone in a basin does the same sensible thing.
Making irrigation more efficient has repeatedly failed to reduce water use at basin scale. This is the irrigation efficiency paradox, a rebound effect of the kind economists call Jevons paradox, and Spain is the best-documented case in the EU.
Measured across Spanish agriculture, 2001–2019 (Agricultural Water Management):
| Change over the period | |
|---|---|
| Cultivated land | −845,982 ha (−5%) |
| Irrigated land | +413,379 ha (+13%) |
| Localised (drip) irrigation | +905,063 ha (+76%) |
| Gross blue water footprint, including conveyance and application losses | stable, thanks to drip |
| Total consumptive water use by irrigated crops | increased |
Read the last two rows together. Drip technology held gross withdrawals flat while actual consumption rose. The efficiency gain was spent on expanding irrigated area and switching to thirstier, higher-value woody crops — not on leaving water in the river. The authors flag expansion into drought-prone regions and over groundwater bodies already in poor status.
The mechanism is well theorised (Water Resources and Economics; Sustainability Science meta-study; Water 18(7):802, critical synthesis): a decision that is entirely rational for one farmer becomes collectively harmful when a whole basin makes it, producing aquifer depletion and reduced river flows. The literature's remedy is institutional — such as capping the water allocation of farmers who received efficiency subsidies — not technical.
What this means for you. Drip genuinely helps your farm, and this page recommends it. But it does not create water. If you intend it to carry you through drought, the saved water has to stay unspent. Spending it on more hectares reproduces the Spanish outcome on your own land. That is unusual advice for an agricultural site to give, and we think it is the honest version. There is a shorter summary of this on how to read this site.
What it is
Delivering irrigation water below the soil surface, into the root zone, rather than spraying or flooding it across the top.
Subsurface drip irrigation (SDI) uses perforated dripline buried 15–45 cm down, emitting slowly at low pressure. The commercially serious version.
Surface drip is the same principle with the line laid on top — cheaper, shorter-lived, more evaporative loss.
Ollas, or buried clay pots, are unglazed porous vessels buried to the neck and filled with water, which seeps through the pot wall into the surrounding soil. Ancient, self-regulating, hand-filled.
This is the logical continuation of water retention landscapes, which found that roughly 450 mm of a Mediterranean summer's pond storage evaporates away. The principle there was to keep water under the surface. This page applies it to delivery.
How it helps against drought
It eliminates evaporation from the wetted surface. Water enters below the layer where sun and wind act, so almost none is lost before roots reach it.
No wind drift or interception losses, unlike sprinklers.
It keeps the inter-row dry, which suppresses weed germination and cuts the transpiration of plants you did not plant.
It delivers precisely where roots are, allowing small frequent applications matched to demand rather than periodic flooding.
Ollas self-regulate: seepage through the porous wall slows as the surrounding soil moistens and resumes as it dries — demand-driven delivery with no controls at all.
What the evidence shows
| Finding | Magnitude | Source |
|---|---|---|
| SDI vs conventional furrow/flood | ~55% irrigation water saving | CGIAR/MEL, arid regions comparison |
| Drip vs surface methods generally | 50–80% water saving, crop- and condition-dependent | ibid. |
| Maize under SDI | yield +31%, nitrogen-fertiliser efficiency +43–71% | PMC10372093 |
| Maize grain yield, SDI at full recommended nitrogen vs furrow/flood | +15.8% | ibid. |
| Processing tomato: SDI vs sprinkler at similar applied water | higher yield under SDI | Effect of SDI on processing tomato yield, water table depth, salinity and profitability |
| Tomato: drip vs furrow | yield ~1.5× using 30% less irrigation water | ibid. |
| Semiarid olive orchards, Spain: SDI vs surface drip | SDI gave higher yield and water-use efficiency | ASCE J. Irrigation & Drainage Engineering 140(10)IR.1943-4774.0000745) |
| Winter wheat, deficit SDI | improved water-use efficiency and stabilised yield by enhancing subsoil water extraction | PMC7232564 |
| Maize silage, Po Valley, Italy | life-cycle assessment comparing surface and drip environmental performance | Sustainability 18(2):580 |
The most drought-relevant result is the last but one. Deficit subsurface drip stabilised winter wheat yield by improving subsoil water extraction. Reducing the variance between good and bad years is what an unpredictable climate actually demands — arguably more useful than raising the average.
Ollas: useful, but know the boundary
There is a genuine peer-reviewed basis here — "Buried clay pot irrigation: a little known but very efficient traditional method of irrigation" (Agricultural Water Management) — and porous-ceramic subsurface emitters have a real research literature in arid regions.
Reported figures, with health warnings attached:
- A 2013 Kenyan trial reported 97.1% water saving for maize and 97.8% for tomato against furrow irrigation, with yield per unit of water up 32.2% and 43.7%. Treat with caution: furrow is a very inefficient baseline, the trial was small-plot and non-European, and savings of that size do not transfer to field scale.
- University extension sources report 60–70% water reduction against a watering can — a garden comparison, not a farm one.
- The widely repeated claim that ollas are "ten times more efficient than sprinklers or drip" comes from gardening blogs and vendors, not research. We are not going to repeat it.
The honest boundary: ollas are hand-filled clay pots. They are excellent for market gardens, kitchen gardens, container growing, and establishing individual trees and shrubs — anywhere labour per plant is acceptable and the realistic alternative is a watering can. They are not a field-crop technique, and we found no European farm-scale trial.
They are, however, by far the cheapest and most reversible option on this page. You can dig them up and take them with you.
Where it fits
| Context | Fit | Why |
|---|---|---|
| Mediterranean high-value irrigated crops — olive, vine, almond, vegetables | Excellent (SDI) | Best evidence, and the Spanish orchard data applies directly |
| Processing tomato, maize and other irrigated row crops | Excellent (SDI) | Documented yield and efficiency gains against sprinkler and furrow |
| Winter cereals under deficit irrigation | Good | Yield stabilisation is the benefit rather than maximisation |
| Market gardens, horticulture, container growing | Ollas viable | Labour is acceptable at this scale; cheap and fully reversible |
| Tree and shrub establishment, agroforestry planting | Ollas excellent | Cheap, targeted, removable once roots establish |
| Field-scale arable | Ollas: no | Hand-filling does not scale |
| Rainfed farms with no water right | Not applicable | This page assumes you have water to apply |
| Basins already over-abstracted | Caution | Efficiency here may enable expansion and worsen a shared problem |
What it costs
We have no European installation figures. The numbers below are United States market data, undated, given only for order of magnitude, and should not be treated as European prices.
| Item (US data) | Figure |
|---|---|
| SDI installation | $1,000–4,000/acre (≈ $2,470–9,880/ha) |
| SDI premium over surface drip | 25–40% higher initial cost — trenching and specialised tubing |
| Annual maintenance | $50–150/acre — filter cleaning, emitter flushing, winterisation |
| SDI service life | commonly 12–15 years, up to 15–20 with good water quality |
| Surface drip service life | 5–10 years |
The lifespan figures transfer better than the prices, because they are engineering properties rather than market rates — and they contain the real economic argument for burying the line: roughly double the service life, plus lower maintenance, for a 25–40% higher install cost.
Water quality and filtration determine where in that range you land. Clogging is the primary constraint on SDI, and filtration is what prevents it.
Benefits arrive in the first season. Capital is recovered over the system's life, so the payback calculation is dominated by how long the line survives.
Ollas cost almost nothing, last indefinitely if not frozen or cracked, and cost labour instead.
Commitment
Moderate to high, and the lock-in is deeper than the equipment.
Buried dripline is a 12–15 year capital commitment. Removing it means lifting the whole network, and abandoned dripline leaves plastic in the soil — which sits awkwardly beside our criticism of plastic mulch on the mulching page.
The deeper commitment is that it ties the farm to irrigated production. Having invested per hectare, you are strongly incentivised to keep irrigating, which is precisely the mechanism behind the rebound effect described at the top of this page.
And the binding long-term risk is your water right, not your kit. An SDI system is worthless if abstraction is curtailed, and basins already in poor status are exactly where curtailment is most likely. That is a risk you should price before installing, not after.
Surface drip is a shorter commitment with a recurring plastic consumable. Ollas are fully reversible — the only technique on this site with essentially zero commitment.
How to start
- Check your basin's status before investing. If the groundwater body is in poor status or the basin is over-allocated, your abstraction may be curtailed within the system's lifetime. This is a bigger threat to the investment than any equipment failure.
- Reduce demand before improving supply. Mulching and organic-matter building cut irrigation need with no capital and no lock-in — and they let you install a smaller, cheaper system.
- Do not bank the saving as extra hectares. If drip is meant to see you through drought, the saved water must stay unspent.
- Test your water and specify filtration accordingly. This determines whether you get 12 years or 20.
- Buy root-intrusion protection — herbicide-impregnated, copper-oxide or physical-barrier emitters — and irrigate regularly enough that soil around the emitters never dries out. Dry soil is what draws roots into emitters.
- Start with ollas at garden or tree-establishment scale if you want the subsurface principle at near-zero cost and no commitment.
- Pair it with soil moisture measurement, so you irrigate to demand rather than to schedule. The yield-stabilising benefit of deficit irrigation depends on knowing what the soil actually holds.
Risks and failure modes
The efficiency paradox is the systemic risk — farm-level saving that becomes basin-level depletion. The most important failure mode, and invisible from any single farm.
Emitter clogging is described in the literature as the primary barrier to widespread SDI use, from physical, chemical, biological and root-intrusion causes. Emitter apertures are only 0.2–2.0 mm (Agricultural Water Management, five-year sugarcane clogging assessment).
Root intrusion into buried emitters, especially where the soil around them dries out.
Rodent damage. Voles, moles, shrews, rabbits and squirrels chew buried tubing — rodent incisors grow continuously and plastic line is ideal gnawing material. A specific, serious and frequently underestimated risk for buried systems, and hard to locate when it happens.
Leaks are invisible. A buried failure may show only as a dying patch of crop, weeks later.
Salinity accumulation at the wetting-front margin under long-term drip in arid conditions.
Capital stranded by a change in water rights, as above.
Plastic left in soil at end of life.
Ollas: labour and fragility. Hand-filling does not scale, and pots crack — especially if frozen.
Over-claiming from enthusiast sources on ollas. The "ten times more efficient" figure is vendor-adjacent, the same problem as the founder claims flagged on swales and keyline.
Claims checked. Two circulating claims about irrigation are run down to their sources: ollas are ten times more efficient than drip and drip irrigation saves water.
When and how much, not just what pipe. Scheduling, deficit strategies and the free regional services that publish a daily crop-water figure are on the precision irrigation page. The efficiency paradox above applies to all of them.