We spent the first four months on this land planting things. Trees, perennial vegetables, soft fruit. We were enthusiastic and reasonably systematic and almost all of it struggled or failed. The second year, we spent three months doing earthworks before planting anything. The results were different. This article explains what we did and why water, not seeds or effort or good intentions, is the limiting factor in Norte Portugal.
## The Portuguese Summer Is Not What You Imagine
People who visit Portugal in July or August — or who look at tourism photographs of green rice paddies and full rivers — may not appreciate what happens to the landscape from June through September in an average year. In our location in the Minho region, annual rainfall is between 1,400mm and 1,800mm, which is high by European standards. But almost none of it falls in summer.
From late May through mid-September, we receive perhaps 40–80mm of rain in total. The soil, which holds water reasonably well in winter and spring, dries to a dusty clay-sand by August. Surface temperature on south-facing slopes can exceed 45°C. Plants without established root systems accessing subsoil moisture die or go dormant. This is the reality behind the optimism of "it rains a lot in Portugal."
A tree planted without any attention to its water situation in the first two summers has a poor probability of survival. A tree planted above a swale that captures and holds runoff has a meaningfully better one. This is not philosophy. It is observable.
## What a Swale Is
A swale is a trench cut along a contour line — horizontally across a slope, at consistent elevation across its entire length. Water that falls on the slope above the swale runs down to it and is captured. Rather than rushing off the land as runoff (carrying topsoil with it), it pools in the swale and infiltrates slowly into the soil downslope of the berm (the raised bank of excavated soil placed on the downhill side of the trench).
A swale does not drain. That is the point. It holds water in place until it sinks.
The principle is associated with permaculture design — Bill Mollison and Geoff Lawton have written extensively about it — but the technique predates the word by centuries. Traditional terraced agriculture across the Mediterranean is a variant of the same logic. We are not doing anything original. We are applying a sensible response to a dry climate.
## Mapping the Contours
Before we cut anything, we spent a week walking the land with an A-frame level — a simple timber tool that allows you to identify points at equal elevation. You mark the ground with stakes as you walk, then connect the stakes with a line of spray paint. That line is your contour. A swale cut along it will hold water level across its entire length.
We had a 4-hectare site with a complex slope — no single clean gradient, several hollows and ridges, a seasonal stream along the eastern boundary. We identified four primary contour lines for swales, each between 40m and 90m long, spaced vertically about 12–18m apart on the main orchard slopes.
We could have hired a topographic surveyor. We chose not to, partly for cost (€800–1,500 for a full survey at the time) and partly because walking the land with a hand tool teaches you things about the site that a printed contour map doesn't. After a week of this, we understood the water movement on the property in a way that informed every subsequent decision.
## What We Built
We hired a mini-excavator with operator for four days at €320/day. In that time, we cut:
- Four swales ranging from 40cm to 60cm deep and 60–80cm wide, with berm material piled on the downhill side to create a planted bank - Two small retention ponds (50m² and 30m² water surface area) at natural collection points in the topography, fed by swale overflow - A diversion drain to redirect peak runoff from an adjacent track away from the orchard
Total earthworks cost: approximately €2,800, including diesel, the operator's expertise in reading the ground, and a day of additional work regrading the track entry.
We planted the swale berms with a mix of nitrogen-fixing shrubs (*Cytisus scoparius*, *Ulex europaeus*) and fruit and nut trees including walnut, almond, fig, and pomegranate. The root systems on the berms help stabilise them; the nitrogen fixers improve soil fertility over time.
## What Changed
The first winter after the earthworks, we watched the swales fill during a heavy rainfall event in November. Water that previously sheeted down the main slope and disappeared into the stream drainage now stopped on the slope and sank over three to five days. The soil in the orchard area downslope of the upper swale was visibly wetter into December than it had been in previous years.
In the following summer, the trees planted above swale berms had markedly better survival and growth rates than comparable trees in areas without earthworks. Of 47 trees planted in swale-supported positions, 41 established well through their first summer. Of 23 trees planted in comparable species mix in areas without earthworks support, 11 were in significant stress by August and 6 died.
Those numbers are not a controlled trial. There were other variables. But they confirmed what we had expected and are consistent with what permaculture designers and agroforesters observe in comparable climates.
## The Principle
Water management precedes everything else. Before the trees, before the garden beds, before the planting plan. The question is always: where does the rain go when it lands, and can we slow it, spread it, and sink it into the soil rather than letting it run off? Answer that question through physical intervention in the land, and everything planted subsequently operates in a different context.
---
*We offer site visits for people designing ecological land systems in Norte Portugal. Contact us at [lusitanoretreat.com](https://lusitanoretreat.com).*