North America is blessed with a disproportionate share of the world's best agricultural soil. It is no coincidence that the U.S. is one of the few countries that’s a net exporter of food – North America has 17% of the world's arable land, but less than 7% of the world's population. As the human population approaches 10 billion people later this century, productive farmland will become an increasingly strategic resource as the amount per capita declines by as much as half.

Researching my book, Dirt: The Erosion of Civilizations, opened my eyes to how much humanity takes soil for granted – and how much trouble that has gotten society after society into in the past. A scant 3% of the world's total surface area has soil suitable for intensive cultivation.

And yet, how much do most of us know about where soil comes from? Our most essential and undervalued resource, soil is where the living world of biology meets the dead world of geology.

The world's best agricultural soils are those that developed along with the native forests (ultisols and alfisols) and grasslands (mollisols) that long covered continents in the temperate regions. In these locations the addition of organic matter from leaf fall in forests or root growth in grasslands built up fertile soils with high agricultural capacity. And the best of these soils for agriculture are those developed on deposits of loess, windblown silt such as that covering much of the American Midwest.

Central North America, eastern Europe, northern China and the Argentine Pampas are the backbone of the world's agricultural production. In all of these areas, loess from a few feet to hundreds of feet thick blankets continental bedrock. With a high proportion of finely ground, fresh mineral grains, soils developed on loess are renowned for their fertility. Loess is so fertile that it can be farmed productively even after the topsoil is eroded off. The same can't be said for soils in most of the world, where rock lies just one to several feet below ground. North America’s tremendous agricultural productivity reflects a disproportionate share of the world's loess.

But American agriculture’s foundation did not originate in place. Loess blew in on the wind. If you fly over northern Canada on a clear day you can see vast areas of naked rock, the geological scar of where it all came from. Over the past several million years, glaciers repeatedly overran most of modern Canada. As they grew and pushed south, the great ice sheets scraped off weak layers of soil and fractured, weathered rock, stripping the land down to fresh, hard rock. Bulldozed up by the ice, Canadian soils were carted to the melting front of the ice sheet and deposited as broad outwash plains built by rivers of meltwater. Strong winds then spread this fertile blanket of fine, glacier-ground silt across the American Midwest.

The windblown origin of loess helps explain why the 1930s’ Dust Bowl was so severe. Soil-forming material formerly anchored by native prairie roots was vulnerable to high winds blowing across plow-bared fields during the next severe drought.

The Dust Bowl produced dramatic soil loss and societal dislocation, but even ancient societies that mined their soil far more slowly were eventually reduced to destitution. Consider how images of modern Iraq don't exactly evoke the Garden of Eden.

As I looked back through history, I learned how the Greeks, Romans, Easter Islanders and many other societies declined as they lost their agricultural soils after cultivation practices exposed soil to accelerated erosion. Societies that plowed through their soil cashed in their future. In Syria, the foundations of Roman towns and farms standing on bare rock offer mute testimony to the former prosperity implied by Roman tax records that tell of bountiful harvests in areas now lacking soil.

Soil loss not ancient history

Yet soil loss is not just ancient history. When I compiled data from around the world on the rate of erosion from agricultural fields to compare with data on the pace of soil production, I found that tilled soils erode 10 to 100 times faster than soil forms. Every year, about half a percent of the world's arable land is degraded enough to be taken out of production.

While the pace of soil erosion makes it challenging to prioritize soil conservation efforts, thick loess soils can maintain agricultural productivity for longer than can thin soils developed in place from rock weathering. In all likelihood, North America will be feeding the world for the foreseeable future because of its natural endowment of loess.

With a limited – and shrinking – global amount of highly productive cropland, the future of humanity hinges on whether we take care of the world's best soils. What do we need to do in order to feed the world well into the future? Sustain both the physical body of the soil (prevent erosion) and the fertility of the soil. It would help to rebuild soil fertility on degraded and marginal agricultural land. How we will do all this in the coming post-cheap-oil world is a question deserving far more attention than it’s getting.

You don't have to be a geologist to realize that a century is not much time in the life of the land. And the health of the land has always been at the heart of the health of human societies. Fortunately, agriculture can also rebuild soils far faster than nature does. But to motivate doing so on a continental scale, we need to see North America's rich endowment of fine agricultural soil as an intergenerational global trust because, like us, people tomorrow will need to eat. In the end, it really is that simple.

David R. Montgomery is a professor of geomorphology at the University of Washington. He is author of Dirt: the erosion of civilizations and The rocks don't lie: A geologist investigates Noah's flood.