As a young undergraduate studying agriculture, soil science was not one of my favourite subjects, presented by lecturer who was so shy he couldn't even bear to face the class when supposedly talking to them. But I did pick up a sense that Australia was not so much farming as mining its soils: rising salinity along the breadbasket Murray-Darling basin as ancient salts were brought to the surface by irrigation water; erosion from the impact of the hard-hoofed cattle and sheep brought with white settlement; desert winds sweeping away marginal lands being ploughed up for wheat.

It was only when, as a journalist, I visited a permaculture farmer, who went into the middle of a wheat field, yellow and crispy, about ready to harvest, and with no appreciable effort pushed a spade into the earth, turning over dark, moist, hummous-rich soil crawling with earthworms (considerably better than I would have done in my own garden) that I started to grasp how alternative approaches to the industrial farming about which I had been taught might have a point. And that the soil was the foundation of all, and that the chemicals, the minerals balanced out was only the start - that soil was indeed a whole complex ecosystem of its own - microbes, invertebrates and all, a complex mix that wasn't going to be "fixed" by getting the nitrogen/phophorous balance right.

The introductory chapter of David R. Montgomery's Dirt: The Erosion of Civilization covers those basics nicely, beginning with the touching tale of how Charles Darwin, in his last book, working with his sons, demonstrated the vital role of earthworms in soil formation and the maintenance of soil fertility, concluding that "all the vegetable mould over the whole country has passed many times through, and will again pass many times through, the intestinal canal of worms". (This despite some scorn about the "minor" issue on which he was working.) He calculated, more or less correctly, that English and Scots worms moved half a billion tonnes of soil a year. (Although, Montgomery points out, he didn't know about isostasy, the process by which rocks lift to replace those eroded, so his figures on the age of mountains were wildly out.)

From that foundation, Dirt moves on to exploring the place of soil in early human civilisation. It includes the latest research in archeology and genetics — in its account of the origins of agriculture, which Montgomery places at the headwaters of the Euphrates from about 10,000BC, and finds best explored around the settlement known as Abu Hureyra. For the 3,000 years before this it had been an Edenic environment of oak forests interspersed with stands of wild grain, he says, but when there was a sudden dry period, the communities of hunter-gatherers who had become settled there chose to begin to domesticate wild varieties of rye and wheat. After about 1,000 years rainfall improved, and barley and peas, and other crops were added to the mix, allowing the village population to reach some 5,000 within a couple of thousand years, the largest concentration of humans yet known.