Agricultural yield upside and population growth

This image will come out far too small on the blog, so see the full-size original connected to this short Atlantic piece titled "The Next Breadbasket?", which gets some major messages right:

Sub-Saharan Africa, despite its long history of food insecurity, is one place where yields could increase dramatically; agricultural basics such as good seed and fertilizer would go far in a region that the green revolution bypassed. “We could increase yields in sub-Saharan Africa threefold tomorrow with off-the-shelf technology,” says Kenneth Cassman, a well-regarded agronomist who researches potential yields. The problem is the continent’s long history of corruption, poor infrastructure, and lack of market access.

Agricultural investment in Africa—and in a few other high-potential places such as Ukraine and Russia—may be the world’s best bet for keeping food plentiful and cheap. This investment could bring other benefits too; the World Bank estimates that agricultural development is twice as effective at reducing poverty as other sources of growth.

However, the article's reference to Paul Ehrlich (of The Population Bomb) goes off track. Contrary the predictions of Ehrlich (and Thomas Malthus, for that matter), population growth has been slowing as fertility rates drop off with higher wealth; most experts now predict human population will stop growing some time this century. The inexorable driver of demand growth is increasingly per capita consumption, which rises with economic development (mainly calories and meat for food, but also energy, water, etc.). What is really scary is not 9 billion people, but 9 billion people with the environmental footprint of the average American.

And the map itself is a good idea, but some of its numbers seem funny to me, which makes me doubt its rigor. I buy that Sub-Saharan Africa and Eastern Europe are low, but not that Australia is equally low. I don't think India should be higher than Latin America - India's farmers are notoriously constrained for inputs and credit - and I find it very hard to believe that the U.S., whose industrial monocultures produce the best per-acre average yields in the world, is worse than Myanmar. I like the concept of the map, but the execution seems sloppy.