I haven’t thought carefully about this question since I first read Francis Moore Lappe’s Diet for a Small Planet in the 1970s: “I have a question…about growing grains for livestock feeding: When corn, for example, is grown, it has a k/cal value, or an amount of caloric energy that I assume can be counted for human nutrition. As it is used for feed, is caloric energy lost in the transfer? In other words, is the resulting food (all the butchered and eaten parts of a cow, for example) similar in energy content to the edible feed that went into it? I assume we lose some caloric value along the way, but how much? To me, this has all kinds of implications that I’d like to ponder, including the implications of this from the (admittedly broad) prospective of global hunger.”
This is about “feed efficiency ratios”–the amount of food it takes to make a pound of animal or a pound of us. Let’s hear it for Wikipedia, which has a nice summary with a reasonable reference. It takes more grain to make beef than pigs, chicken, or fish. A lot of those animals are not usable for food (maybe half a beef carcass, for example) and we have our own problems with efficiency. The calories listed in food tables are pretty much what we can use and the better tables specify how well the meat is trimmed. Good question!