When the British supermarket giant Tesco decided to start labelling its produce with ''food miles'' to let people know how far it travelled before reaching the shelf, the move was greeted with a bizarre mixture of fear, derision and relief.

The fear came from the global food industry and many primary producers, who remain worried that people will be put off by finding out how far their food has come.

The derision arose because Tesco had wandered blithely into the labyrinth of attempting to accurately measure greenhouse gas emissions. It is so fantastically complicated that a team of Oxford University climate experts commissioned by the supermarket chain to do the maths described their own findings as ''highly contentious''.

The relief stemmed from the many consumer groups, farmers and other advocates of local produce who had been crying out for more information for years. Food labelling fed a general belief in the community that people have a right to know where their food comes from and what impact it has had on the environment.

''You can't expect somebody to stand in the aisle with their mobile phone and surf the net to find the right information,'' says Jeff Angel, the director of Sydney's Total Environment Centre. ''Mostly the right information isn't there yet.''

Australia is about to start treading the food labelling path. The federal government is undertaking a comprehensive review of labelling standards, and for the first time this includes the environmental impact of food. Findings are said to be available early next year. A few supermarkets in Australia, such as Aldi, have pledged to start labelling their products with an estimated carbon footprint by next year.

The US giant Walmart plans to have ''eco-labels'' on its products within the year. In Britain, Marks & Spencer followed Tesco and has a sticker on all products that arrive by air. Coca-Cola and Pepsi are working with England's Carbon Trust to measure reductions in CO2 in their bottling process.

What we eat makes up about one-third of the total greenhouse gas emissions from an average Australian household, according to the CSIRO. This means that, in theory, a solar panel on your roof could be cancelled out by a lifetime preference for imported Italian pasta and sun-dried tomatoes. A high protein diet can be a useful for losing weight, but eating beef three times a week means about 1.5 tonnes of methane emissions a year, the CSIRO says, which could be more than the entire food footprint of a dedicated vegetarian.

But in a few cases the food miles concept could also put a dent in animal liberation arguments, because centralised farming sees fodder being shipped to a factory farm in bulk rather than delivered to a warehouse and trucked out in smaller portions to separate farms, burning more diesel.

To test the food miles concept in Australian conditions, the Herald mapped the journey of an unremarkable basket of groceries and found that it had, collectively, travelled further than some of its consumers.

Even though most of the produce was Australian, the most conservative estimates, using methodology and data being experimented with by the CSIRO and the Department of Climate Change, showed the food had travelled at least 80,919 kilometres before it reached the shelves of its central Sydney supermarket.

The lightweight basket can easily be carried in one had, yet simply getting it to the shop burnt enough fuel to run a heavily loaded car for 30 kilometres. This is the most conservative food miles figure possible, with all distances rounded down and assumptions that transport vehicles were well-maintained, fuel efficient and encountered minimal traffic on their way to market.

If the energy cost of refrigeration, snap-freezing, pre-cooking and storage are factored in, the carbon footprint of the food basket soars to alarming levels. This is without the emissions from plastic packaging, which has risen in the past five years despite signs of consumer resistance, together with the smelting of metal for canned food and the production and printing of paper labels and the manufacture of adhesives.

Billions of dollars are at stake in the food miles debate. As the world's biggest beef producer, yet isolated by long distances from its main markets, Australia has a lot to lose if the rest of the world continues paying attention to most existing carbon labels.

The meat industry is already on the counter-attack, pointing out that food miles can be highly misleading. In some cases, grass-fed beef raised and slaughtered near Albury-Wodonga but transported to and eaten in London can have lower emissions than meat from the next English county.

''Importantly, while food miles may have intuitive appeal among some consumers, the food miles concept results in less informed choices and does not reflect the carbon emission embodied in many products,'' says Dr Beverley Henry, the carbon emissions advisor to industry group Meat and Livestock Australia.

''Food miles is a misleading indicator of the carbon footprint of food products that, if widely used, would distort international agricultural markets and possibly increase global carbon emissions.''

The meat industry has a point. ''Empirical evidence indicates that food miles is an unreliable indicator of carbon emissions in the food supply chain,'' says a report from the Australian Bureau of Agricultural and Resource Economics that assessed the various food miles claims in some detail.

It cited a study from New Zealand's Lincoln University that traced the carbon footprint of exported Kiwi lamb, dairy products, apples and onions to their British equivalents.

A tonne of New Zealand lamb produced 563 kilograms of carbon dioxide, including the emissions from the fuel burnt to ship it to England, while the local lamb produced 2.85 tonnes of carbon dioxide for every tonne of meat. The proportions were repeated when other imports were compared with their local equivalent. Basically, New Zealand farmers were found to be more energy efficient.

Shaken by numbers like these, the British Soil Association, an organic certifier, began to pull back from recommending that air-freighted organic goods be forsaken in favour of strictly local produce. University studies have since concluded that a rejection of air-freighted goods could lead to higher carbon emissions, because of regional differences in farming methods.

The most common example is green beans grown in Kenya and imported in large amounts by many countries. A study in the British newspaper The Guardian suggests the Kenyan beans, which are grown with minimal irrigation and manual labour, have a smaller carbon footprint than beans produced in the Britain in fields ploughed by diesel-burning tractors and spread with chemical fertiliser. The former British minister for trade and development, Gareth Thomas, said at the time that driving to the shops burnt more fuel than flying a crate of beans to Surrey from Africa.

Sydney's food supply network is a bit like an ant colony, with a constant stream of workers fanning out across the surrounding landscape and bringing in food. Ants have solved the problem of supplying a static colony with sustenance by deploying the majority of their workforce as fetchers and carriers, whereas humans have done it by mechanisation.

It is an intricate, invisible web that is in most respects a masterpiece of commerce and logistics, in that it delivers healthy, fresh and nutritious food to the overwhelming majority of people at prices they can afford. While most citizens are asleep, about 4600 refrigerated trucks are on the move early each morning, bringing produce from regional centres to the city's main food distribution points at Wetherill Park, Bankstown, Silverwater, Blacktown and Flemington. A fleet of smaller, diesel-burning commercial vehicles collects produce and delivers it to shop shelves.

The IGA supermarket chain has two main distribution centres at Blacktown and Silverwater, but where its produce comes from before that is a trickier proposition.

''It's the suppliers' job as to where they source the food from,'' said a spokesman for Metcash, the owner of the IGA brand and distributor of its products.

''They have to comply with our policy of sourcing produce locally. We can trace back to wholesale, then the distributors can trace back to their suppliers from there. If there is a problem with one product, we can source it from somewhere else.''

Essentially, where the supermarket's fresh food comes from is a day-to-day proposition, depending on price and availability.

For instance, even though Australia produces 12,000 tonnes of fresh garlic a year, none was available when the Herald visited to buy its sample grocery basket. The garlic was from Mexico, and using the most conservative assumptions about freight travel it had travelled about 10,160 kilometres by sea and road before it arrived at IGA.

On the assumption the garlic's nautical voyage went smoothly and there was minimal traffic congestion and no mechanical trouble by road in Mexico or Sydney, the single head of Mexican garlic generated about 940 grams of carbon dioxide emissions from the fuels used to get it to the shelf.

IGA isn't the only supermarket to rely heavily on imported garlic; about 9000 tonnes are imported annually from China, according to data compiled by Australian Garlic Producers, an industry group. This accounts for half the garlic on supermarket shelves.

When tracing the food from the IGA basket, the Herald used the methods deployed by the Australian Greenhouse Office, which has since been absorbed into the Department of Climate Change and Energy. It was also informed by calculations used in a study of a basket of groceries sampled in Melbourne by staff at CERES Community Environment Park, which found that 29 common food items had travelled 70,803 kilometres. The fresh produce in the IGA basket was impossible to trace to the farm gate in every case, so average transport figures from the most likely production areas, based on trucking times and fuel use, were used.

Essentially, the distance and fuel efficiency of the type of transport to get food from A to B was divided by the weight of the food involved. As well as showing how far food has actually come, it highlights some of the shortcomings of food miles, because the weight of a particular product becomes the key determining factor. Vegetables can have a heavier food miles footprint than meat, even though meat production generally uses more energy.

It also downplays the fuel used to generate electricity for the refrigeration of food, which most analysts believe is probably more important than how far the food has travelled. Keeping food cool generates an estimated 18 million tonnes of greenhouse gas in Australia, the equivalent of putting an extra 4.3 million cars on the road each year, according to a Food Science Australia submission to a Senate inquiry into food in 2008.

The cereal products in the food basket fared relatively well in terms of food miles, with the Sanitarium Weet-Bix and loaf of Tip Top bread generating less carbon dioxide than their own weight.

Both used Australian wheat, milled locally and mostly trucked by road, with the breakfast cereal's more simple range of ingredients giving it the edge in food miles. The loaf of bread required small quantities of gluten, yeast, salt, sugar and vegetable oil in the baking process, lifting its carbon dioxide emissions to about 425 grams of CO2 for the shipping of one 650-gram loaf.

Most of the fresh vegetables were traced to their most likely sources within a 400-kilometre radius of Sydney and they also had relatively petite carbon footprints. Half a kilo of Black & Gold frozen peas required about 315 grams of carbon dioxide emissions from fuel to get it to the supermarket, while a bag of brushed potatoes took a more circuitous route from the south-west of Sydney, generating just over half a kilogram of carbon.

A bunch of broccollini, with a relatively straightforward road journey to Flemington markets and then the IGA distribution centre, came in at just 92 grams of CO2 but mandarins from just north of Gympie in Queensland travelled an estimated 1247 kilometres down the coast to Sydney, generating 448 grams of emissions from an articulated lorry.

Imported produce stood out for its lengthy journeys and relatively high emissions, though in the absence of access to a ship's log the number of nautical miles actually covered can only be based on standard shipping routes.

A 280-gram jar of Sacla brand sundried tomatoes travelled an estimated 14,680 nautical miles in a container vessel before being unloaded at Port Botany, leading to a carbon footprint of 2.72 kilograms of greenhouse gas on the way. A 100-gram wedge of Castello blue cheese managed the journey with a slightly less polluting 1.34 kilograms of CO2, not including refrigeration.

By comparison, a wedge of King Island camembert crossed the Bass Strait and completed the voyage to Sydney while managing to emit 138 grams of carbon dioxide, according to the Department of Climate Change model.

A porterhouse steak, believed to be from the northern tablelands in NSW although this could not be directly confirmed by IGA, would have travelled a minimum of 670 kilometres if the truck driver took the fastest route, generating about 225 grams of CO2 from transport.

Sourcing fish products is more complex, because trawlers cruise different fishing grounds chasing a worthwhile catch, but the swordfish fillet purchased at IGA was probably from the Brisbane grounds off the east coast. Based on average trawling practices, discussed in a Bureau of Rural Sciences report into fisheries in 2004, the unlucky swordfish could have been on a 1300 nautical mile round trip before arriving at market, its weight burning up 450 grams of carbon dioxide fuel on the way. Like the red meat, though, refrigeration costs would add significantly to the food's greenhouse burden.

Swordfish is also on Greenpeace Australia's red list as an overfished species that leads to the death of many commercially useless marine species because of trawling methods, though this is disputed by the industry.

''The supermarket needs to end its trade in overfished species and provide customers with a sustainable caught seafood and make this sustainable seafood-sourcing policy publicly available,'' said a Greenpeace spokeswoman, Genevieve Quirk.

Even advocates of local produce agree that food miles is only one tool among many for measuring the true impact of food. ''It can be useful but doesn't show the full picture,'' said Holly Vyner of the organic produce group Biological Farmers of Australia. ''In general it is much more important to look at how the food is processed.''

There are exceptions, but less packaging, less chemical fertiliser and less travel generally leaves a lighter footprint.

Ben Cubby is The Sydney Morning Herald Environment Editor.