A paper just published in Nature Geoscience has terrific news for anyone worried about the sustainability of agriculture. ((Davis, K. F., M. C. Rulli, A. Seveso, and P. D’Odorico. (2017). Increased food production and reduced water use through optimized crop distribution. Nature Geoscience. Published online:
06 November 2017 doi: 10.1038/s41561-017-0004-5 Behind a paywall, natch, but I’m sure you can find a copy if you’re suffiently motivated.)) It should be possible to grow 10% more calories and 19% more protein while simultaneously using 14% less rainwater and 12% less irrigation water. And that, the authors say “would feed an additional 825 million people”.
Kyle Frankel Davis and colleagues reach this happy conclusion by modelling the effect of shifting crops around to where they yield the most, taking into account things like how much water is available and which crops do best under those circumstances. Globally, you do it by increasing groundnuts, roots, soybeans, sorghum and tubers at the expense of millets, rice, sugar crops and wheat, but the details depend on where you are. In western Russia, for example, you cut down on the millets, sugar beet and sunflowers and plant rainfed sorghum, soybeans, tubers and wheat. In the Nile Delta, groundnuts, maize and sorghum replace sugar beet and wheat. In other places the substitutions involve more crops.
Their detailed look at the outputs of the models offers some important observations. For water, 42 countries, many of which currently don’t have enough for their farms, would save at least 20% of their water needs. And 63 countries that currently depend on imports for a lot of their food would see their production of protein and calories increase by at least 20%, boosting food security.
There’s bad news too: in Australia’s Murray-Darling basin, northern India and the US midwest, no choice of crops offers sustainable water use.
Optimistic optimisation
This approach is not entirely new. In 2006 Christoph Müller and his colleagues did a similar kind of optimisation exercise under which crops were allocated to the areas in which they would be most productive, ignoring trade barriers, transportation costs and subsidies. ((Müller, C., A. Bondeau, H. Lotze-Campen, W. Cramer, and W. Lucht (2006), Comparative impact of climatic and nonclimatic factors on global terrestrial carbon and water cycles, Global Biogeochem. Cycles, 20, GB4015, doi:10.1029/2006GB002742.)) That model found that you could grow all the food humanity needs on just 2 million km2, whereas you need 35 million km2 if you try and grow as much food as possible locally. ((Strangely, Davis et al. don’t cite Müller et al, but I’m sure that’s just an oversight.))
The Davis et al. study is somewhat more realistic, constraining the shifts so that there’s no loss of crop diversity, expansion of croplands or impact on nutrient and feed availability. They also say that there would not be much impact on rural livelihoods.
Both papers are making important points about the contribution of agriculture to global water and carbon cycles, although I guess the take-home is that the way agriculture is organised now is really, really inefficient. That’s clearly true.
But still …
I see no prospect of any shift to the kind of global distribution that either paper imagines.
What are people on the 33 million km2 supposed to do instead of growing food? Hang around waiting for a shipment? Which they buy with what?
And given how wedded people are to their “traditional” crops, even if those crops have been around less than a couple of hundred years, I can’t imagine them shifting just because it would be more sustainable.
Davis et al. say:
Of course, there are probably cultural barriers and dietary preferences that may limit the application of this strategy in certain ways — considerations that may be better accommodated in future analyses by constraining the production quantities of each crop.
It probably needs world peace too.