David Montgomery‘s Dirt: The Erosion of Civilizations reminded me a lot of When the Rivers Run Dry by Fred Pearce, except that instead of looking at how we have allocated our water resources globally, it focuses on the way humanity has husbanded (or not) its soil resources throughout history, through a vast array of case studies in what we got wrong. It also reminded me a little bit of Energy at the Crossroads, insofar as the last chapter or two, instead of being a concrete, level-headed outline of what we need to do if we actually want to solve the problem which has been presented, it devolves a little bit into a lament. You’ve convinced me there’s a problem. Clearly you have some idea of what the solution looks like. Please don’t be afraid to put that idea into words, even if you think the plausible solutions are so far removed from our current way of doing things that someone is going to think you’re crazy. I think a lot of the most credible solutions to our sustainability problems sound “crazy” to “normal” people these days… but that’s just the way it is. We still need to know what the available solutions look like, or at the very least, what characteristics one can sketch out which any available solution has to have.
The basic problem that Montgomery lays out is the combination of the plough, topography, and gravity. Ploughing inevitably disturbs soil. Disturbed soil on a slope goes downhill, especially in places where the rain isn’t gentle. If the net flux of the soil going downhill is greater than the rate of soil production in situ in the long term, you’re fucked. And the net flux of soil going downhill is greater than the rate of soil production, just about everywhere. About a third of the Earth’s inventory of topsoil has been washed into the oceans as a direct result of human deforestation and agriculture.
The simplest description of soil breaks it into 3 layers or horizons. Topsoil, or the A horizon, is the relatively organic rich layer containing most of the living matter in the soil. The subsoil, or B horizon, is poor in organic material, somewhat weathered chemically, and made up of relatively small particles. The C horizon, or parent rock, is largely unweathered boulders or bedrock. Slowly over millennia water and oxygen and biological forces penetrate into the crust dissolving and fracturing the rock. The rate of this penetration is roughly the rate of soil formation, and it’s of order millimeters to centimeters per century, whereas using industrial agriculture techniques on slopes can result in erosion rates of 10-100 cm per century. Topsoil thickness in a particular place will eventually come to equilibrium under a given land use. The problem is that with most agriculture today, that equilibrium thickness is zero.
Forests allow thick, rich soils to form even in sloping places by reducing the erosion rate significantly. They soften the impact of rains, and their roots physically hold the soils in place. Their roots also contribute to the mechanical weathering of the parent rock. In the short term, clearing forests to open up new farmland seems like a good idea, but ultimately, it ends up being more like soil mining, as without the trees, the equilibrium thickness is significantly reduced, and the soils end up in the bottomlands (and may even temporarily improve farming down there… before getting washed out to sea). Deforestation agriculture in the (flat) tropics generally fails for a somewhat different reason. When temperatures are high, and rainfall is plentiful, all the soluble mineral nutrients that plants require get leached out of the soils, and the entire available inventory accumulates in the local biota, and increases only very slowly with time. Remove that biota, and you’ve removed most of what made the area capable of being lush, and you’ll be forced to import all the nutrients your crops need if you want to keep them producing. In contrast, irrigated agriculture in arid climates (net evaporation) destroys soil productivity in place by depositing mineral salts over time (cf. Mesopotamia, the Indus and the San Joaquin Valley).
The only way to make sloping farmland work long term seems to be fairly labor intensive terracing… which is really just a way to make sloping farmland flat. The relatively closed nutrient loop terraced rice paddy system that has developed in Bali seems to have been stable on a 1000 year timescale… which is pretty unusual. Carefully managed floodplain agriculture in non-arid climates also seems to be workable potentially long term, but historically has proven unstable because of population growth, and expansion of the practicing civilizations into more marginal upland areas, with the associated deforestation. Large scale loess deposits like are found in the periglacial plains of midwestern North America, northern China, and parts of the former Soviet Union also may be able to support long term agriculture, but only if they are protected from heavy rains and winds like those that created the Dustbowl, meaning that you can never till the soil and leave it bare, and must instead use continuous cover crops, and in all likelihood no-till agricultural methods. As it turns out, these methods can also potentially sequester significant amounts of carbon in the soils, which would be a good thing.
In the end, the message seems to be that certainly using current methods, we are far beyond Earth’s sustainable food production. Since in the US we currently throw away half of our food, and a lot of that is unnecessarily produced meat anyway, there’s a lot of slack to work with. In places like India and China, I have to imagine there’s less food waste, and I know there’s far less meat consumption and waste, so this slack is not globally distributed. There are more labor intensive (and less profitable) methods of agricultural production that get much more food per unit soil (and less per man hour), but as of yet, there seems to be limited interest in getting more people tending the land. There are surely also approaches to this problem which would involve genetic engineering. Not the proprietary and dishonest kind that (e.g.) Monsanto generally hocks, but the open and altruistic kind that (e.g.) Pamela Roland is advocating. Is it without issues? Of course not. But neither are any of the other options on offer.
In the fullness of time, everything about our agricultural system will be detached from the mines. We will not mine the soil, and send it to the sea. We will not mine fossil fuels to fertilize, or bring in rock phosphate by the trainload. We will not mine the vast post-glacial aquifers. We will not do these things because they will not be options any more. What we have to decide is whether we’d like to live at a modest level of agricultural consumption, investing more labor and intelligence and less materials and energy in the production of our food, or whether we would like to continue the historical agricultural boom-and-bust pattern, periodically dropping into famine and despair while the forests regrow, and the soils rebuild, as the sedimentary records indicate they did in pre-Roman Europe, and again after Rome fell, for centuries. Choosing the former stable solution within any system resembling our current economy will require deciding that in matters of soil conservation, the right discount rate is zero. This is true for many long term sustainability issues, and I just don’t see anybody working toward it in a generalized way within the finance world. Hopefully I’m wrong.
One thought on “Dirt: The Erosion of Civilizations by David Montgomery”