How Much Water Is on Your Plate?

There is an avocado sitting in my kitchen.

Nutritionally, it has a pretty respectable CV. Fibre, vitamins, minerals and mostly unsaturated fats, all wrapped in its own biodegradable packaging.

Environmentally, things become more complicated.

Avocados have become one of the poster foods for another, less familiar measure of environmental impact: the water footprint.

Search online, and you'll quickly encounter eye-catching claims about how many litres of water it takes to produce an avocado. Similar calculations exist for almonds, coffee, chocolate, beef, rice, tomatoes and almost everything else we put on our plates.

We are also increasingly encouraged to think about the carbon footprint of our food, where it was produced, how it was grown and how far it travelled. Although even here, the obvious answer can mislead. For most foods, what we eat and how it is produced matter considerably more to its carbon footprint than how far it has travelled.

Before long, lunch starts to feel less like lunch and more like an environmental audit.

The obvious response is to compare the numbers and choose the food with the smallest one.

Unfortunately, that doesn't really work.

We eat a lot more water than we drink

When we think about our own water consumption, we tend to think about the obvious stuff: showers, baths, washing machines, toilets and the tap.

Compared with the water involved in producing our food, these are only part of the picture.

The UN Food and Agriculture Organisation estimates that producing one person's daily food can require between 2,000 and 5,000 litres of water. Agriculture accounts for around 72 per cent of global freshwater withdrawals.

We don't see most of that water because it has already been used long before the food reaches us.

It grew the wheat in the bread. It irrigated vegetables. It grew animal feed. It provided drinking water for livestock. It was involved in processing the ingredients.

The term often used for this is virtual water.

It is a useful concept, although it feels slightly strange to describe water as virtual. The water itself was perfectly real. We just didn't see it being used.

Water isn't carbon with a tap attached

Carbon footprints have given us a fairly familiar way of thinking about environmental impact. It is tempting to assume a water footprint works much the same way.

There is one rather important difference.

Carbon dioxide contributes to a global atmospheric problem. A tonne emitted in Britain has essentially the same effect on the climate as a tonne emitted in California or Chile.

Water is stubbornly local.

Five hundred litres of rain falling on agricultural land in a relatively wet region is one thing. Pumping 500 litres from an aquifer that is already being depleted is something quite different.

Which is why a food can look relatively good through a carbon lens and rather less comfortable through a water one.

There are other measures too. Land use matters. Biodiversity matters. Nutrition matters. Affordability certainly matters.

We would probably like all these measures to point us towards the same foods.

They don't always oblige.

A litre isn't quite a litre

Water-footprint researchers generally distinguish between three types of water.

Green water is rainfall stored in the soil and used by plants.

Blue water is surface water or groundwater consumed during production, including water used for irrigation.

Then there is grey water, which is not literally grey water in the domestic sense, but a measure of the freshwater required to assimilate pollution associated with production.

This makes those eye-catching comparisons between different foods rather less useful than they first appear.

Suppose the same crop is grown in two places and requires the same total amount of water.

One farmer gets most of that water from rainfall. The other irrigates using groundwater in an area where water resources are already under pressure.

On a simple litres-per-kilogram chart, the two crops might sit side by side.

In the real world, they are not the same proposition.

The number of litres matters, but so does where they came from.

The avocado in the room

Which brings us back to my avocado.

There are legitimate concerns about avocado production. The trees need water; commercial orchards can depend heavily on irrigation, and some important growing regions experience water stress.

But that still doesn't make “How much water does an avocado use?” a particularly useful question on its own.

An avocado grown largely with rainfall isn't the same as one dependent on irrigation. An irrigated avocado grown where water is plentiful isn't necessarily equivalent to one grown where farmers, communities, and ecosystems compete for limited supplies.

Even the country printed on the label may not tell us enough. Water availability can vary enormously within countries and between individual catchments.

This is where the seductive simplicity of the water-footprint league table starts to fall apart.

The avocado isn't innocent or guilty.

It depends where it came from and how it was grown.

But aren't we supposed to be eating more of this stuff?

This is where things become particularly awkward.

We are encouraged to eat more fruit, vegetables, pulses, nuts and other plant foods. At the same time, we hear that some of these foods have surprisingly large water footprints.

Are we being asked to choose between our own health and the environment?

Fortunately, the evidence doesn't suggest such a straightforward conflict.

A large study examining diets across the UK, France and Germany looked at 43,786 geographical areas. In every one, moving from existing eating patterns towards the recommended healthy diet containing meat reduced the dietary water footprint. Reductions ranged from 11 to 35 per cent, with even larger reductions associated with the healthy pescetarian and vegetarian diets examined.

That doesn't mean plant-based automatically means water-friendly, any more than local automatically means low carbon.

A nut grown using irrigation in a severely water-stressed region presents a different problem from a crop grown mostly using rainfall elsewhere.

This is why dividing the supermarket into environmentally good and bad foods isn't particularly helpful.

The interesting question is not just what we eat.

Increasingly, it is where and how it was produced.

Have you tried reading a food label recently?

There is another problem here; it belongs less to hydrology than to the supermarket aisle.

We already give consumers an extraordinary amount of information about food.

Calories. Protein. Fibre. Carbohydrates. Sugars. Fat. Saturated fat. Salt. Ingredients. Allergens. Portion sizes. Traffic lights.

Many of us are already standing there turning the packet over, trying to work out whether 7.3 grams of something is reassuring or alarming.

There is also growing interest in how processed our food is.

Now imagine adding carbon.

Then water.

Perhaps biodiversity after that.

There comes a point, when more information doesn't necessarily produce a better-informed consumer. It can simply produce a more confused one.

A label saying “500 litres of water” sounds wonderfully clear.

Yet it could be almost meaningless without knowing whether those litres were rainfall or irrigation, and whether the food came from Wales or a water-stressed region of southern Spain.

If we ever start putting water information on food, we may need to measure something more useful than litres.

We may need to measure water stress.

And that is considerably harder to fit on the side of a packet.

Perhaps we're asking the wrong person

There is a tendency with environmental problems to push the final decision onto the consumer.

Choose the right product. Read the label. Compare the numbers. Make the responsible decision.

But how realistic is that?

Nobody wants to conduct a quick investigation into aquifer depletion while standing in the fruit and vegetable aisle.

The curious thing is that the supply chain may already know far more than the shopper.

Retailers know where produce comes from. Suppliers can know how farms irrigate. Water stress can be mapped. Large food businesses already collect considerable information about their supply chains.

So perhaps some of the responsibility belongs there.

A supermarket deciding where to source thousands of tonnes of produce has rather more influence over water use than I do standing in front of two avocados.

That doesn't remove consumer choice. It means some of the complicated decisions are made where the information actually exists.

There is one easy bit

For all this complexity, one way to reduce the water footprint of our food doesn't require a new label, an app, or a working knowledge of hydrology.

Waste less of it.

When food goes into the bin, the resources used to produce it go with it.

The irrigation water has already been pumped. The rain has already fallen. The fertiliser has been applied. The tractor has burned its fuel. The food has been processed, transported and refrigerated.

Throwing away an avocado doesn't somehow return its water to the aquifer from which it may have been taken.

The water is invisible in the bin, but its use has already happened.

So, can I eat the avocado?

Yes, I am going to eat mine.

That isn't an endorsement of every avocado produced everywhere, nor an argument that its water footprint doesn't matter.

It simply acknowledges that food is more complicated than a league table.

We need to know more about where our food comes from and how it is produced. Retailers and manufacturers can make better sourcing decisions. Farmers can be supported in using water more efficiently. And where information genuinely helps consumers make a better choice, we should provide it.

But perhaps we should be wary of reducing every environmental question to another number on a packet.

Asking how many litres of water went into our lunch is a useful start.

The better question is what sort of water it was, where it came from and whether that water could comfortably be spared.

My avocado doesn't have those answers printed on it.

Perhaps one day it should.