Milk Drinkers Conquer the World?

I went on a bike ride to Minneapolis yesterday and listened to the last hour or so of an audiobook called The 10,000 Year Explosion: How Civilization Accelerated Human Evolution by Gregory Cochran and Henry Harpending (2010). I had begun listening quite a while ago and had been moderately interested, but not excited enough to finish the book quickly. It was a moderately interesting account of recent discoveries in genetics along with some speculations about history and culture as sources of selective pressure on particular groups. I do find it interesting to imagine that with contributions from epigenetics, there may be recognizable evolutionary effects happening at historical scales (generations, centuries) as well as over multi-millennia.

One thing that jumped out at me was a discussion of the Proto-Indo-Europeans and their apparent success spreading their genes, language, and culture. The authors speculated that in addition to potential technological differences with their neighbors, these people may also have benefited from the European lactose-tolerance mutation, which they identify as the 13910-T allele. They said:

Initially, selection favored individual carriers of the lactose-tolerance mutation, but the mutation was rare and had little social effect. Cattle were used for plowing and pulling wagons, for their beef, and as a source of secondary products like leather and tallow. But when the lactase-persistence allele became common, so that a majority of the adult population could drink milk, a new kind of pastoralism became possible, one in which people kept cattle primarily for their milk rather than for their flesh. This change is very significant, because dairying is much more efficient than raising cattle for slaughter: It produces about five times as many calories per acre.17 Dairying pastoralists produce more high-quality food on the same amount of land than nondairy pastoralists, so higher frequencies of lactose tolerance among Indo-Europeans would have caused the carrying capacity of the land to increase—for them. (p. 181)

The thing that jumped out at me was the claim that dairying is five times more efficient than raising cattle for beef. I made a mental note of that and looked into it a bit when I got home from my ride. Contemporary estimates of feed efficiency do suggest that about 24% of feed is "effectively converted" to milk, while only 1.9% becomes beef. This claim is illustrated in a pretty graphic provided on Oxford University's "Our World In Data" website.

I'm not sure what this means, however. Are the people who compiled this data into this helpful chart saying that over the lifetime of the animal, only 1.9% of the feed they ingest produces meat but that 24% produces milk? I don't see how this could possibly be true. It strikes me as logical that a certain percentage of the feed any animal eats is going to supply energy for basic metabolism. Then another percentage will be used to produce milk, after the needs of staying alive have been met. Although I imagine cattle use a lot of food energy in growing to their adult size, it doesn't strike me that once they are full size, they would need to spend lots of additional calories growing "meat". Maybe putting on fat, so the steaks will be well-marbled. But I just don't see how the claim that milk-producing is 24% efficient while meat producing is 1.9% efficient makes any actual sense. If anybody can tell me what I'm missing, please do!

This doesn't entirely explain to me, either, where the authors were coming from with their claim of a 5x advantage for milk-producing pastoralists in the ancient world. There was a footnote in the passage and I was able to find the source, which turned out to be a USDA Farmer's Bulletin from 1917. The bulletin summarizes its findings at the start, saying "corn, on a given area, will produce more human food than any other crop commonly grown on American farms." It goes on to say "the dairy cow is the most efficient of farm animals in production of human food (milk) and the hog is the most efficient in the conversion of grain into meat, producing five times as much per acre of crops as does any other animal."

So there's a 5x advantage mentioned there. It's about hogs, though, not cattle at all. So if the authors were using that as the source of their claim in a comparison of milk with beef, they missed the mark completely. But even if this wasn't the source of the error, the data and assumptions that went into the bulletin's findings are still problematic. The measurements of per-acre efficiency in milk and meat production come with a laundry list of assumptions about what combinations of grains and silage and clover hay and cottonseed meal were produced on the acre to feed the animal. It was not about animals grazing on grasses. So the conclusions the bulletin came to were really applicable only to this type of feedlot conversion of acres of crop outputs into a "mix" fed to livestock. I don't think it's even remotely possible to generalize from this data to a claim that the ancient Proto-Indo-Europeans would have had a five times advantage in available calories per acre after switching from eating beef to drinking milk.

Neither the Proto-Indo-Europeans nor their neighbors lacking the lactose-digestion gene were operating feedlots. They were probably both pastoralists. Or maybe the Aryans were pastoralists and their neighbors were settled farmers. In either case, the biggest differences between milking a cow and butchering, it seems to me, have to do with sustainability. To whatever extent you can get milk from a grazing cow (I assume a cow eating grass and wandering around the ancient world would produce less milk daily than a cow living in a barn or feedlot on a combination of grain feeds and silage today), you would have a source of protein and fat that renews itself daily and that you don't have to kill the animal to take advantage of. Pastoralists who could benefit from the nutrients in milk because they were able to metabolize it would certainly have an advantage over people living the same lifestyle without this mutation. This and the additional benefits of being mobile pastoralists rather than settled farmers, may have made enough of a difference to allow the Aryans to expand across continents. But it wasn’t because milk production is 5x more efficient per acre than meat production.

This type of blatant misuse of contemporary data (even if that data wasn’t a bit suspect to start!) really causes me difficulty in taking anything else in the book seriously, which is unfortunate. The authors had some other interesting ideas about the ways biological evolution may have interacted with culture; but I now feel like I have to take them with a big handful of salt.

PS. I also suspect there’s a bit of contemporary bias in the whole meat-bad, milk-good argument in the Oxford data. I’m reminded of another book from 2010, Meat: A Benign Extravagance, by Simon Fairlie.